1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.393 2006/09/19 11:28:44 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.
39 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
41 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
43 PH/08 An error is now given if message_size_limit is specified negative.
45 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
46 to be given arbitrary names.
52 SC/01 Use a glob alias rather than an array ref in eximstats generated
53 parser. This improves both readability and performance.
55 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
56 Don't display local sender or destination tables in eximstats unless
57 there is data to show.
58 Added average volumes into the eximstats top table text output.
60 SC/03 Collect data on the number of addresses (recipients) as well
61 as the number of messages in eximstats.
63 TF/01 Correct an error in the documentation for the redirect router. Exim
64 does (usually) call initgroups() when daemonizing.
66 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
67 with consistent privilege compared to when running as a daemon.
69 TF/03 Note in the spec that $authenticated_id is not set for local
70 submissions from trusted users.
72 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
73 Thanks to Dean Brooks <dean@iglou.com> for the patch.
75 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
76 by adding some example configuration directives to the default
77 configuration file. A little bit of work is required to uncomment the
78 directives and define how usernames and passwords are checked, but
79 there is now a framework to start from.
81 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
82 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
83 without this. I don't know how relevant this is to other LDAP libraries.
85 PH/02 Add the verb name to the "unknown ACL verb" error.
87 PH/03 Magnus Holmgren's patch for filter_prepend_home.
89 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
91 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
92 directory not expanded when it should be if an expanded home directory
93 was set for the address (which is overridden by the transport).
95 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
98 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
99 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
100 because it is too late at that time, and has no effect.
102 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
103 security issue with \' (bugzilla #107). I could not use the
104 PQescapeStringConn() function, because it needs a PGconn value as one of
107 PH/08 When testing addresses using -bt, indicate those final addresses that
108 are duplicates that would not cause an additional delivery. At least one
109 person was confused, thinking that -bt output corresponded to deliveries.
110 (Suppressing duplicates isn't a good idea as you lose the information
111 about possibly different redirections that led to the duplicates.)
113 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
114 systems where poll() doesn't work, in particular OS X.
116 PH/10 Added more information to debugging output for retry time not reached.
118 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
119 operations in malware.c.
121 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
124 PH/13 If write_rejectlog was set false when logging was sent to syslog with
125 syslog_duplication set false, log lines that would normally be written
126 both the the main log and to the reject log were not written to syslog at
129 PH/14 In the default configuration, change the use of "message" in ACL warn
130 statements to "add_header".
132 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
133 not followed by a command (e.g. "seen endif").
135 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
136 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
139 PH/17 Added extra conditions to the default value of delay_warning_condition
143 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
144 { match{$h_precedence:}{(?i)bulk|list|junk} } \
145 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
148 The Auto-Submitted: and various List- headers are standardised, whereas I
149 don't think Precedence: ever was.
151 PH/18 Refactored debugging code in route_finduser() to show more information,
152 in particular, the error code if getpwnam() issues one.
154 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
155 This is apparently needed in addition to the PH/07 change above to avoid
156 any possible encoding problems.
158 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
159 but not after initializing Perl.
161 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
162 output them only if debugging. By default they are written stderr,
163 apparently, which is not desirable.
165 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
168 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
171 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
173 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
174 authenticated or an ident call has been made. Suppress the default
175 values for $authenticated_id and $authenticated_sender (but permit -oMai
176 and -oMas) when testing with -bh.
178 PH/24 Re-jigged the order of the tests in the default configuration so that the
179 tests for valid domains and recipients precede the DNS black list and CSA
180 tests, on the grounds that those ones are more expensive.
182 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
183 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
184 This bug exists in every version of Exim that I still have, right back to
187 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
188 However, an attempt to turn on case-sensitivity in a regex key by
189 including (?-i) didn't work because the subject string was already
190 lowercased, and the effects were non-intuitive. It turns out that a
191 one-line patch can be used to allow (?-i) to work as expected.
197 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
198 other effects) broke the use of negated acl sub-conditions.
200 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
203 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
204 "Deny" causes Exim to reject the incoming connection with a 554 error.
205 Unfortunately, if there is a major crisis, such as a disk failure,
206 tcp-wrappers gives "deny", whereas what one would like would be some
207 kind of temporary error. A kludge has been added to help with this.
208 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
209 554 error is used if errno is still zero or contains ENOENT (which occurs
210 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
213 PH/03 Add -lutil to the default FreeBSD LIBS setting.
215 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
216 errors. Otherwise a message that provokes a temporary error (when other
217 messages do not) can cause a whole host to time out.
219 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
220 addresses were routed directly to files or pipes from a redirect router.
221 File deliveries just didn't batch; pipe deliveries might have suffered
224 PH/06 A failure to get a lock for a hints database would erroneously always say
225 "Failed to get write lock", even when it was really a read lock.
227 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
228 of 0600. This has been changed to use the value of the lockfile_mode
229 option (which defaults to 0600).
231 PH/08 Applied small patch from the Sieve maintainer.
233 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
234 folder from quota calculations, a direct delivery into this folder messed
235 up the contents of the maildirsize file. This was because the regex was
236 used only to exclude .Trash (or whatever) when the size of the mailbox
237 was calculated. There was no check that a delivery was happening into an
238 excluded directory. This bug has been fixed by ignoring all quota
239 processing for deliveries into excluded directories.
241 PH/10 Added the maildirfolder_create_regex option to appendfile.
247 PH/01 The code for finding all the local interface addresses on a FreeBSD
248 system running IPv6 was broken. This may well have applied to all BSD
249 systems, as well as to others that have similar system calls. The broken
250 code found IPv4 interfaces correctly, but gave incorrect values for the
251 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
252 that it would not match correctly against @[] and not recognize the IPv6
255 PH/02 The ipliteral router was not recognizing addresses of the form user@
256 [ipv6:....] because it didn't know about the "ipv6:" prefix.
258 PH/03 Added disable_ipv6.
260 PH/04 Changed $reply_address to use the raw form of the headers instead of the
261 decoded form, because it is most often used to construct To: headers
262 lines in autoreplies, and the decoded form may well be syntactically
263 invalid. However, $reply_address has leading white space removed, and all
264 newlines turned into spaces so that the autoreply transport does not
267 PH/05 If group was specified without a user on a router, and no group or user
268 was specified on a transport, the group from the router was ignored.
270 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
271 for visible compile-time settings that can be used to change these
272 numbers, for those that want even more. Backwards compatibility with old
273 spool files has been maintained. However, going back to a previous Exim
274 release will lost any variables that are in spool files.
276 PH/07 Two small changes when running in the test harness: increase delay when
277 passing a TCP/IP connection to a new process, in case the original
278 process has to generate a bounce, and remove special handling of
279 127.0.0.2 (sic), which is no longer necessary.
281 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
282 be the same on different OS.
284 PH/09 Moved a debug statement in filter processing to avoid a race problem when
287 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
288 whether --show-vars was specified or not
290 JJ/02 exipick: Added support for new ACL variable spool format introduced
293 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
294 syntactically invalid From: or Reply-to: line, and a filter used this to
295 generate an autoreply, and therefore failed to obtain an address for the
296 autoreply, Exim could try to deliver to a non-existent relative file
297 name, causing unrelated and misleading errors. What now happens is that
298 it logs this as a hard delivery error, but does not attempt to create a
301 PH/11 The exinext utility has a -C option for testing purposes, but although
302 the given file was scanned by exinext itself; it wasn't being passed on
303 when Exim was called.
305 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
306 an end-of-file indication when reading a command response.
308 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
309 compiled. In many other places in Exim, IPv6 addresses are always
310 recognized, so I have changed this. It also means that IPv4 domain
311 literals of the form [IPV4:n.n.n.n] are now always recognized.
313 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
314 used if the router is not running as root, for example, when verifying at
315 ACL time, or when using -bh. The debugging output from this situation was
316 non-existent - all you got was a failure to exec. I have made two
319 (a) Failures to set uid/gid, the current directory, or a process leader
320 in a subprocess such as that created by queryprogram now generate
321 suitable debugging ouput when -d is set.
323 (b) The queryprogram router detects when it is not running as root,
324 outputs suitable debugging information if -d is set, and then runs
325 the subprocess without attempting to change uid/gid.
327 PH/15 Minor change to Makefile for building test_host (undocumented testing
330 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
331 additional section of a DNS packet that returns MX or SRV records.
332 Instead, it always explicitly searches for A/AAAA records. This avoids
333 major problems that occur when a DNS server includes only records of one
334 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
335 fixed another bug: if SRV records were looked up and the corresponding
336 address records were *not* found in the additional section, the port
337 values from the SRV records were lost.
339 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
340 using the correct key (the original address) when searching the retry
341 rules in order to find which one to use for generating the retry hint.
343 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
344 adding the default one. Similarly, if it contains a Reply-To: header, the
345 errors_reply_to option, if set, is not used.
347 PH/19 When calculating a retry time, Exim used to measure the "time since
348 failure" by looking at the "first failed" field in the retry record. Now
349 it does not use this if it is later than than the arrival time of the
350 message. Instead it uses the arrival time. This makes for better
351 behaviour in cases where some deliveries succeed, thus re-setting the
352 "first failed" field. An example is a quota failure for a huge message
353 when small messages continue to be delivered. Without this change, the
354 "time since failure" will always be short, possible causing more frequent
355 delivery attempts for the huge message than are intended.
356 [Note: This change was subsequently modified - see PH/04 for 4.62.]
358 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
359 $1, $2, $3) because the numerical variables can be reset during some
360 expansion items (e.g. "match"), thereby losing the authentication data.
362 PH/21 Make -bV show the size of off_t variables so that the test suite can
363 decide whether to run tests for quotas > 2G.
365 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
366 mailbox_size, and mailbox_filecount in the appendfile transport. If a
367 filecount value is greater than 2G or if a quota value is greater than 2G
368 on a system where the size of off_t is not greater than 4, a panic error
371 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
372 never match. The debug and -bh output now contains an explicit error
373 message indicating a malformed IPv4 address or mask.
375 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
376 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
379 PH/25 Do not write to syslog when running in the test harness. The only
380 occasion when this arises is a failure to open the main or panic logs
381 (for which there is an explicit test).
383 PH/26 Added the /no_tell option to "control=freeze".
385 PH/27 If a host name lookup failed very early in a connection, for example, if
386 the IP address matched host_lookup and the reverse lookup yielded a name
387 that did not have a forward lookup, an error message of the form "no IP
388 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
389 could be logged. Now it outputs the IP address instead of "NULL".
391 PH/28 An enabling patch from MH: add new function child_open_exim2() which
392 allows the sender and the authenticated sender to be set when
393 submitting a message from within Exim. Since child_open_exim() is
394 documented for local_scan(), the new function should be too.
396 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
397 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
398 results in an empty string is now treated as unset.
400 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
402 PH/31 Added sender_verify_fail logging option.
404 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
405 needed by RFC 822 but not by RFC 2822 was commented out. I have now
406 tidied the source and removed it altogether.
408 PH/33 When a queue run was abandoned because the load average was too high, a
409 log line was always written; now it is written only if the queue_run log
410 selector is set. In addition, the log line for abandonment now contains
411 information about the queue run such as the pid. This is always present
412 in "start" and "stop" lines but was omitted from the "abandon" line.
414 PH/34 Omit spaces between a header name and the colon in the error message that
415 is given when verify = headers_syntax fails (if there are lots of them,
416 the message gets confusing).
418 PH/35 Change the default for dns_check_names_pattern to allow slashes within
419 names, as there are now some PTR records that contain slashes. This check
420 is only to protect against broken name servers that fall over on strange
421 characters, so the fact that it applies to all lookups doesn't matter.
423 PH/36 Now that the new test suite is complete, we can remove some of the
424 special code in Exim that was needed for the old test suite. For example,
425 sorting DNS records because real resolvers return them in an arbitrary
426 order. The new test suite's fake resolver always returns records in the
429 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
430 bounces) except when queue_only is set, to avoid logging races between
431 the different processes.
433 PH/38 Panic-die if .include specifies a non-absolute path.
435 PH/39 A tweak to the "H" retry rule from its user.
437 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
438 a label. They prevented compilation on older perls.
440 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
441 a warning to be raised on newish perls.
443 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
444 on queue. Changes to match documented behaviour of showing count of
445 messages matching specified criteria.
447 PH/40 Changed the default ident timeout from 30s to 5s.
449 PH/41 Added support for the use of login_cap features, on those BSD systems
450 that have them, for controlling the resources used by pipe deliveries.
452 PH/42 The content-scanning code uses fopen() to create files in which to put
453 message data. Previously it was not paying any attention to the mode of
454 the files. Exim runs with umask(0) because the rest of the code creates
455 files with open(), and sets the required mode explicitly. Thus, these
456 files were ending up world-writeable. This was not a big issue, because,
457 being within the spool directory, they were not world-accessible. I have
458 created a function called modefopen, which takes an additional mode
459 argument. It sets umask(777), creates the file, chmods it to the required
460 mode, then resets the umask. All the relevant calls to fopen() in the
461 content scanning code have been changed to use this function.
463 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
464 to 24 hours. This avoids potential overflow problems when processing G
465 and H retry rules. I suspect nobody ever tinkers with this value.
467 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
469 PH/45 When the plaintext authenticator is running as a client, the server's
470 challenges are checked to ensure they are valid base64 strings. By
471 default, the authentication attempt is cancelled if an invalid string is
472 received. Setting client_ignore_invalid_base64 true ignores these errors.
473 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
474 they are received. Thus, the responses can be made to depend on the
475 challenges. If an invalid string is ignored, an empty string is placed in
478 PH/46 Messages that are created by the autoreply transport now contains a
479 References: header, in accordance with RFCs 2822 and 3834.
481 PH/47 Added authenticated_sender_force to the smtp transport.
483 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
485 PH/49 Installed latest patch from the Sieve maintainer.
487 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
488 was also set, the appendfile transport was unnecessarily scanning a
489 directory of message files (e.g. for maildir delivery) to find the count
490 of files (along with the size), even though it did not need this
491 information. It now does the scan only if it needs to find either the
492 size of the count of files.
494 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
496 PH/52 Two bugs concerned with error handling when the smtp transport is
499 (i) Exim was not creating retry information for temporary errors given
500 for individual recipients after the DATA command when the smtp transport
501 was used in LMTP mode. This meant that they could be retried too
502 frequently, and not timed out correctly.
504 (ii) Exim was setting the flag that allows error details to be returned
505 for LMTP errors on RCPT commands, but not for LMTP errors for individual
506 recipients that were returned after the DATA command.
508 PH/53 This is related to PH/52, but is more general: for any failing address,
509 when detailed error information was permitted to be returned to the
510 sender, but the error was temporary, then after the final timeout, only
511 "retry timeout exceeded" was returned. Now it returns the full error as
512 well as "retry timeout exceeded".
514 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
515 do this, and (what is worse) MTAs that accept it.
517 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
518 will now be deprecated.
520 PH/56 New os.c-cygwin from the Cygwin maintainer.
522 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
523 formats (previously only available in exim formats via -bpr, -bpru,
524 and -bpra. Now also available in native and exiqgrep formats)
526 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
527 with very large, slow to parse queues
529 JJ/08 exipick: added ! as generic prefix to negate any criteria format
531 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
533 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
534 responses to authentication challenges, though it was showing the
535 challenges; (ii) I've removed the CR characters from the debug output for
538 PH/58 Allow for the insertion of a newline as well as a space when a string
539 is turned into more than one encoded-word during RFC 2047 encoding. The
540 Sieve code now uses this.
542 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
543 data_4xx, lost_connection, tls_required.
545 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
546 message was being sent as an SMTP response.
548 PH/61 Add -l and -k options to exicyclog.
550 PH/62 When verifying, if an address was redirected to one new address, so that
551 verification continued, and the new address failed or deferred after
552 having set something in $address_data, the value of $address_data was not
553 passed back to the ACL. This was different to the case when no
554 redirection occurred. The value is now passed back in both cases.
556 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
557 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
558 use login_cap.h, so on its own it isn't the distinguishing feature. The
559 new name refers directly to the setclassresources() function.
561 PH/65 Added configuration files for NetBSD3.
563 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
565 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
566 is preferred over IPv4.
568 PH/68 The bounce_return_message and bounce_return_body options were not being
569 honoured for bounces generated during the reception of non-SMTP messages.
570 In particular, this applied to messages rejected by the ACL. This bug has
571 been fixed. However, if bounce_return_message is true and bounce_return_
572 body is false, the headers that are returned for a non-SMTP message
573 include only those that have been read before the error was detected.
574 (In the case of an ACL rejection, they have all been read.)
576 PH/69 The HTML version of the specification is now built in a directory called
577 spec_html instead of spec.html, because the latter looks like a path with
578 a MIME-type, and this confuses some software.
580 PH/70 Catch two compiler warnings in sieve.c.
582 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
583 function verify_get_ident() calls ip_connect() to connect a socket, but
584 if the "connect()" function timed out, ip_connect() used to close the
585 socket. However, verify_get_ident() also closes the socket later, and in
586 between Exim writes to the log, which may get opened at this point. When
587 the socket was closed in ip_connect(), the log could get the same file
588 descriptor number as the socket. This naturally causes chaos. The fix is
589 not to close the socket in ip_connect(); the socket should be closed by
590 the function that creates it. There was only one place in the code where
591 this was missing, in the iplookup router, which I don't think anybody now
592 uses, but I've fixed it anyway.
594 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
595 well as to direct DNS lookups. Otherwise the handling of names in host
596 lists is inconsistent and therefore confusing.
602 PH/01 Two changes to the default runtime configuration:
604 (1) Move the checks for relay_from_hosts and authenticated clients from
605 after to before the (commented out) DNS black list checks.
607 (2) Add control=submission to the relay_from_hosts and authenticated
608 clients checks, on the grounds that messages accepted by these
609 statements are most likely to be submissions.
611 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
613 (1) Generate an error if the third argument for the ${prvs expansion is
616 (2) Treat a missing third argument of ${prvscheck as if it were an empty
619 (3) Reset the variables that are obtained from the first argument of
620 ${prvscheck and used in the second argument before leaving the code,
621 because their memory is reclaimed, so using them afterwards may do
624 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
625 one (it's much easier than Tom thought :-).
627 (5) Because of (4), we can now allow for the use of $prvscheck_result
628 inside the third argument.
630 PH/03 For some reason, the default setting of PATH when running a command from
631 a pipe transport was just "/usr/bin". I have changed it to
634 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
635 anything to be listed in the output from -bV.
637 PH/05 When a filter generated an autoreply, the entire To: header line was
638 quoted in the delivery log line, like this:
640 => >A.N.Other <ano@some.domain> <original@ddress> ...
642 This has been changed so that it extracts the operative address. There
643 may be more than one such address. If so, they are comma-separated, like
646 => >ano@some.domain,ona@other.domain <original@ddress> ...
648 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
649 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
650 not being looked up in the rDNS to get a host name, Exim was showing the
651 IP address twice in Received: lines, even though the IP addresses were
652 identical. For example:
654 Received: from [1.2.3.4] (helo=[1.2.3.4])
656 However, if the real host name was known, it was omitting the HELO data
657 if it matched the actual IP address. This has been tidied up so that it
658 doesn't show the same IP address twice.
660 PH/07 When both +timestamp and +memory debugging was on, the value given by
661 $tod_xxx expansions could be wrong, because the tod_stamp() function was
662 called by the debug printing, thereby overwriting the timestamp buffer.
663 Debugging no longer uses the tod_stamp() function when +timestamp is set.
665 PH/08 When the original message was included in an autoreply transport, it
666 always said "this is a copy of the message, including all the headers",
667 even if body_only or headers_only was set. It now gives an appropriate
670 PH/09 Applied a patch from the Sieve maintainer which:
672 o fixes some comments
673 o adds the (disabled) notify extension core
674 o adds some debug output for the result of if/elsif tests
675 o points to the current vacation draft in the documentation
676 and documents the missing references header update
680 o fixes a bug in processing the envelope test (when testing
681 multiple envelope elements, the last element determinted the
684 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
685 Electronic Mail") by including:
687 Auto-submitted: auto-generated
689 in the messages that it generates (bounce messages and others, such as
690 warnings). In the case of bounce messages for non-SMTP mesages, there was
691 also a typo: it was using "Auto_submitted" (underscore instead of
692 hyphen). Since every message generated by Exim is necessarily in response
693 to another message, thes have all been changed to:
695 Auto-Submitted: auto-replied
697 in accordance with these statements in the RFC:
699 The auto-replied keyword:
701 - SHOULD be used on messages sent in direct response to another
702 message by an automatic process,
704 - MUST NOT be used on manually-generated messages,
706 - MAY be used on Delivery Status Notifications (DSNs) and Message
707 Disposition Notifications (MDNs),
709 - MUST NOT be used on messages generated by automatic or periodic
710 processes, except for messages which are automatic responses to
713 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
714 to the default Received: header definition.
716 PH/12 Added log selector acl_warn_skipped (default on).
718 PH/13 After a successful wildlsearch lookup, discard the values of numeric
719 variables because (a) they are in the wrong storage pool and (b) even if
720 they were copied, it wouldn't work properly because of the caching.
722 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
723 checking when decoding. Apparently there are clients that generate
724 overlong encoded strings. Why am I not surprised?
726 PH/15 If the first argument of "${if match_address" was not empty, but did not
727 contain an "@" character, Exim crashed. Now it writes a panic log message
728 and treats the condition as false.
730 PH/16 In autoreply, treat an empty string for "once" the same as unset.
732 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
733 extension "envelope-auth". The code is finished and in agreement with
734 other implementations, but there is no documentation so far and in fact,
735 nobody wrote the draft yet. This extension is currently #undef'ed, thus
736 not changing the active code.
738 Print executed "if" and "elsif" statements when debugging is used. This
739 helps a great deal to understand what a filter does.
741 Document more things not specified clearly in RFC3028. I had all this
742 sorted out, when out of a sudden new issues came to my mind. Oops."
744 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
747 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
748 canonical form (as documented). However, after a host name lookup from
749 the IP address, check_host() was doing a simple string comparison with
750 addresses acquired from the DNS when checking that the found name did
751 have the original IP as one of its addresses. Since any found IPv6
752 addresses are likely to be in abbreviated form, the comparison could
753 fail. Luckily, there already exists a function for doing the comparison
754 by converting both addresses to binary, so now that is used instead of
757 PH/20 There was another similar case to PH/19, when a complete host name was
758 given in a host list; looking up its IP address could give an abbreviated
759 form, whereas the current host's name might or might not be abbreviated.
760 The same fix has been applied.
766 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
767 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
770 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
771 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
773 PH/03 Typo: missing ".o" in src/pcre/Makefile.
775 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
776 header line, restrict the check to what is listed in RFCs 2369 and 2929.
777 Also, for "Auto-Submitted", treat anything other than "no" as
778 non-personal, in accordance with RFC 3834. (Previously it treated
779 anything starting "auto-" as non-personal.)
781 TF/01 The control=submission/name=... option had a problem with syntax
782 errors if the name included a slash character. The /name= option
783 now slurps the rest of the string, so it can include any characters
784 but it must come last in the list of options (after /sender_retain
787 PH/05 Some modifications to the interface to the fake nameserver for the new
795 TK/01 Added the "success_on_redirect" address verification option. See
796 NewStuff for rationale and an example.
798 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
800 PH/02 Patch to exigrep to allow it to work on syslog lines.
802 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
803 fread() to skip over the body file's header line, because in Cygwin the
804 header line is locked and is inaccessible.
806 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
807 co-exist for some time) to make it clear that it is the Exim ID that is
808 referenced, not the Message-ID: header line.
810 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
811 string_format() function, because snprintf() does not exist on all
814 PH/06 The use of forbid_filter_existstest now also locks out the use of the
815 ${stat: expansion item.
817 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
818 protocol synchronization error", to keep the pedants happy.
820 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
821 well as for IRIX systems, when gcc is being used. See the host.c source
824 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
826 PH/10 Named domain lists were not working if used in a queue_smtp_domains
829 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
830 transport and to the smtp transport in LMTP mode.
832 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
834 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
835 run a filter in a subprocess. This could lead to confusion in subsequent
836 lookups in the parent process. There should also be a search_tidyup() at
837 the end of the subprocess.
839 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
840 only if the host matched helo_try_verify_hosts, which caused the
841 verification to occur when the EHLO/HELO command was issued. The ACL just
842 tested the remembered result. Now, if a previous verification attempt has
843 not happened, "verify = helo" does it there and then.
845 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
847 TK/03 Fix log output including CR from clamd.
849 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
850 exist provoked a memory error which could cause a segfault.
852 PH/15 Installed PCRE 6.2
854 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
856 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
857 of the problem. Specifically, suggested +O2 rather than +O1 for the
860 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
862 PH/20 If a delivery was routed to a non-standard port by means of an SRV
863 record, the port was not correctly logged when the outgoing_port log
864 selector was set (it logged the transort's default port).
866 PH/21 Added support for host-specific ports to manualroute, queryprogram,
867 fallback_hosts, and "hosts" in the smtp transport.
869 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
870 host errors such as "Connection refused".
872 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
873 authentication with radiusclient 0.4.9:
875 - Error returned from rc_read_config was caught wrongly
876 - Username/password not passed on to radius server due to wrong length.
878 The presumption is that some radiusclient API changes for 4.51/PH/17
879 were not taken care of correctly. The code is still untested by me (my
880 Linux distribution still has 0.3.2 of radiusclient), but it was
881 contributed by a Radius user.
883 PH/24 When doing a callout, the value of $domain wasn't set correctly when
884 expanding the "port" option of the smtp transport.
886 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
887 while reading a MIME header. Thanks to Tom Hughes for a patch.
889 PH/24 Include config.h inside local_scan.h so that configuration settings are
892 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
893 that in an ACL for RCPT (for example), you can examine exactly what was
896 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
897 commands, but it was not correctly comparing the address with the actual
898 client host address. Thus, it would show the EHLO address in Received:
899 header lines when this was not necessary.
901 PH/27 Added the % operator to ${eval:}.
903 PH/28 Exim tries to create and chdir to its spool directory when it starts;
904 it should be ignoring failures (because with -C, for example, it has lost
905 privilege). It wasn't ignoring creation failures other than "already
908 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
911 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
912 because an input line was too long, either on its own, or by virtue of
913 too many continuations, the temporary file was not being removed, and the
914 return code was incorrect.
916 PH/31 Missing "BOOL" in function definition in filtertest.c.
918 PH/32 Applied Sieve patches from the maintainer.
920 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
922 PH/33 Added "verify = not_blind".
924 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
925 Local/Makefile (with some defaults set). These are used in built scripts
926 such as exicyclog, but they have never been used in the exim_install
927 script (though there are many overriding facilities there). I have
928 arranged that the exim_install script now takes note of these two
931 PH/35 Installed configuration files for Dragonfly.
933 PH/36 When a locally submitted message by a trusted user did not contain a
934 From: header, and the sender address was obtained from -f or from an SMTP
935 MAIL command, and the trusted user did not use -F to supply a sender
936 name, $originator_name was incorrectly used when constructing a From:
937 header. Furthermore, $originator_name was used for submission mode
938 messages from external hosts without From: headers in a similar way,
939 which is clearly wrong.
941 PH/37 Added control=suppress_local_fixups.
943 PH/38 When log_selector = +received_sender was set, and the addition of the
944 sender made the log line's construction buffer exactly full, or one byte
945 less than full, an overflow happened when the terminating "\n" was
948 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
949 when the result of a list match is failure because a DNS lookup failed.
951 PH/40 RM_COMMAND is now used in the building process.
953 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
954 the "build-* directories that it finds.
956 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
957 address in a domain literal was a prefix of an interface address.
959 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
960 when verifying a sender address, unless rewrite_headers is false.
962 PH/44 Wrote a long comment about why errors_to addresses are verified as
963 recipients, not senders.
965 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
966 the ratelimit ACL was added.
968 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
970 PH/46 Added extra information about PostgreSQL errors to the error string.
972 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
973 avoiding the need to install special zones in a real server. This is
974 backwards compatible; if it can't find the fake resolver, it drops back.
975 Thus, both old and new test suites can be run.
977 TF/02 Added util/ratelimit.pl
979 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
980 clock is set back in time.
982 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
983 Candler <B.Candler@pobox.com>.
985 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
986 OK for addresses that are the result of redirections.
988 PH/48 A number of further additions for the benefit of the new test suite,
989 including a fake gethostbyname() that interfaces to the fake DNS resolver
992 TF/06 The fix for widen_domains has also been applied to qualify_single and
993 search_parents which are the other dnslookup options that can cause
996 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
999 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1001 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1002 service" if it encountered an MX record with an empty target hostname.
1003 The message is now "an MX or SRV record indicated no SMTP service".
1005 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1006 if the DNS of the sending site is misconfigured. This is quite a
1007 common situation. This change restores the behaviour of treating a
1008 helo verification defer as a failure.
1010 PH/51 If self=fail was set on a router, the bounce message did not include the
1011 actual error message.
1017 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1019 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1020 command itself ended in error, the underlying message about the transport
1021 filter timeout was being overwritten with the pipe command error. Now the
1022 underlying error message should be appended to the second error message.
1024 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1026 PH/02 Reduce the amount of output that "make" produces by default. Full output
1027 can still be requested.
1029 PH/03 The warning log line about a condition test deferring for a "warn" verb
1030 was being output only once per connection, rather than after each
1031 occurrence (because it was using the same function as for successful
1032 "warn" verbs). This seems wrong, so I have changed it.
1034 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1035 it should not have, which might have caused a crash in the right
1036 circumstances, but probably never did.
1038 PH/04 Installed a modified version of Tony Finch's patch to make submission
1039 mode fix the return path as well as the Sender: header line, and to
1040 add a /name= option so that you can make the user's friendly name appear
1043 TF/03 Added the control = fakedefer ACL modifier.
1045 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1046 Mark Lowes for thorough testing.
1048 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1050 TK/03 Merged latest SRS patch from Miles Wilton.
1052 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1053 with the definition in sysexits.h (which is #included earlier).
1054 Fortunately, Exim does not actually use EX_OK. The code used to try to
1055 preserve the sysexits.h value, by assumimg that macro definitions were
1056 scanned for macro replacements. I have been disabused of this notion,
1057 so now the code just undefines EX_OK before #including unistd.h.
1059 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1060 in the smtp transport. When a block could not be written in a single
1061 write() function, the timeout was being re-applied to each part-write.
1062 This seems wrong - if the receiver was accepting one byte at a time it
1063 would take for ever. The timeout is now adjusted when this happens. It
1064 doesn't have to be particularly precise.
1066 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1067 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1069 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1070 without a domain if the check to <postmaster@domain> fails.
1072 SC/01 Eximstats: added -xls and the ability to specify output files
1073 (patch written by Frank Heydlauf).
1075 SC/02 Eximstats: use FileHandles for outputing results.
1077 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1079 SC/04 Eximstats: fixed display of large numbers with -nvr option
1081 SC/05 Eximstats: fixed merging of reports with empty tables.
1083 SC/06 Eximstats: added the -include_original_destination flag
1085 SC/07 Eximstats: removed tabs and trailing whitespace.
1087 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1089 TK/06 MBOX spool code: Add real "From " MBOX separator line
1090 so the .eml file is really in mbox format (even though
1091 most programs do not really care). Patch from Alex Miller.
1093 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1094 The latter is generated from $received_to and is only set if the
1095 message has one envelope recipient. SA can use these headers,
1096 obviously out-of-the-box. Patch from Alex Miller.
1098 PH/08 The ${def test on a variable was returning false if the variable's
1099 value was "0", contrary to what the specification has always said!
1100 The result should be true unless the variable is empty.
1102 PH/09 The syntax error of a character other than { following "${if
1103 def:variable_name" (after optional whitespace) was not being diagnosed.
1104 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1105 accidental colon was present, for example, could give incorrect results.
1107 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1108 result is used (not including appendfile, where other changes are about
1111 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1112 This involved changing a lot of size variables from int to off_t. It
1113 should work with maildirs and everything.
1115 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1116 spamd dying while we are connected to it.
1118 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1121 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1122 function for BDB 4.3.
1124 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1126 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1127 the sources for PCRE have been re-arranged and more files are now
1130 PH/15 The code I had for printing potentially long long variables in PH/11
1131 above was not the best (it lost precision). The length of off_t variables
1132 is now inspected at build time, and an appropriate printing format (%ld
1133 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1134 to be "long long int" or "long int". This is needed for the internal
1135 formatting function string_vformat().
1137 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1138 the configuration file to be ":syslog", then the script "guesses" where
1139 the logs files are, rather than using the compiled in default. In our
1140 case the guess is not the same as the compiled default, so the script
1141 suddenly stopped working when I started to use syslog. The patch checks
1142 to see if log_file_path is "". If so, it attempts to read it from exim
1143 with no configuration file to get the compiled in version, before it
1144 falls back to the previous guessing code."
1146 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1147 implementing BATV in an Exim configuration. See NewStuff for the gory
1150 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1151 Makefile that are specific to HP-UX.
1153 PH/18 If the "use_postmaster" option was set for a recipient callout together
1154 with the "random" option, the postmaster address was used as the MAIL
1155 FROM address for the random test, but not for the subsequent recipient
1156 test. It is now used for both.
1158 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1159 patch removes a few documentation additions to RFC 3028, because the
1160 latest draft now contains them. It adds the new en;ascii-case comparator
1161 and a new error check for 8bit text in MIME parts. Comparator and
1162 require names are now matched exactly. I enabled the subaddress
1163 extension, but it is not well tested yet (read: it works for me)."
1165 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1166 rework some of the code of TK/09 above to avoid the hardwired use of
1167 "%lld" and "long long". Replaced the call to snprintf() with a call to
1170 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1171 records point to non-existent hosts", "retry timeout exceeded", and
1172 "retry time not reached for any host after a long failure period".
1174 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1175 experimental DomainKeys support:
1177 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1178 (2) On an error such as an illegally used "control", the wrong name for
1179 the control was given.
1181 These problems did NOT occur unless DomainKeys support was compiled.
1183 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1185 PH/24 Added ${if match_ip condition.
1187 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1188 databases so that it will be absolutely obvious if a crash occurs in the
1189 DB library. This is a regular occurrence (often caused by mis-matched
1192 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1193 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1194 picked up on a user's system that detects such things. There doesn't seem
1195 to be a gcc warning option for this - only an attribute that has to be
1196 put on the function's prototype. It seems that in Fedora Core 4 they have
1197 set this on a number of new functions. No doubt there will be more in due
1200 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1201 specify a transport. However, if an address that was verified by such a
1202 router was the subject of a callout, Exim crashed because it tried to
1203 read the rcpt_include_affixes from the non-existent transport. Now it
1204 just assumes that the setting of that option is false. This bug was
1205 introduced by 4.51/PH/31.
1207 PH/28 Changed -d+all to exclude +memory, because that information is very
1208 rarely of interest, but it makes the output a lot bigger. People tend to
1209 do -d+all out of habit.
1211 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1212 code in os-type was giving problems when libc.so lives in lib64, like on
1215 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1216 aren't the modern standard, and it seems that some systems' include files
1217 don't always have them. Exim was already checking for some of the newer
1218 ones like T_AAAA, and defining it itself. I've added checks for all the
1219 record types that Exim uses.
1221 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1222 not automatically generating a new one, as it is supposed to. This
1223 prevented TLS from working. If the file did exist, but contained invalid
1224 data, a new version was generated, as expected. It was only the case of a
1225 non-existent file that was broken.
1227 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1228 with a change in libdomainkeys > 0.64.
1230 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1231 from DNS. If the selector record carries the flag, it now has
1232 precedence over the domain-wide flag.
1234 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1236 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1237 the use of a port name that isn't defined in /etc/services) occurred, the
1238 message was deferred as in a normal delivery, and thus remained on the
1239 spool, instead of being failed because of the mua_wrapper setting. This
1240 is now fixed, and I tidied up some of the mua_wrapper messages at the
1243 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1244 the messages in a hash of arrays rather than using individual hashes.
1245 This is a bit cleaner and results in dramatic memory savings, albeit
1246 at a slight CPU cost.
1248 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1249 as requested by Marc Sherman.
1251 SC/10 Eximstats: added histograms for user specified patterns as requested
1254 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1256 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1257 fopen() in the content-scanning modules that did not already have it.
1263 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1264 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1266 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1268 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1270 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1271 to test Sieve filters that use "vacation".
1273 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1274 that changes the way the GnuTLS parameters are stored in the cache file.
1275 The new format can be generated externally. For backward compatibility,
1276 if the data in the cache doesn't make sense, Exim assumes it has read an
1277 old-format file, and it generates new data and writes a new file. This
1278 means that you can't go back to an older release without removing the
1281 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1282 work if there are any delivery delays because "one_time" forces the
1283 parent to be marked "delivered", so its unseen clone is never tried
1284 again. For this reason, Exim now forbids the simultaneous setting of
1287 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1288 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1289 entry if you want to know the details. The fix, however, neglected to
1290 consider the case when local delivery batching is involved. The test for
1291 "previously delivered" was not happening when checking to see if an
1292 address could be batched with a previous (undelivered) one; under
1293 certain circumstances this could lead to multiple deliveries to the same
1296 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1297 in its include files, and this causes problems building Exim.
1299 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1300 header_syntax) but Exim was just ignoring anything given after a slash.
1301 In particular, this caused confusion with an attempt to use "verify =
1302 reverse_host_lookup/defer_ok". An error is now given when options are
1303 supplied for verify items that do not have them. (Maybe reverse_host_
1304 lookup should have a defer_ok option, but that's a different point.)
1306 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1307 defined by RFC 821) to 2048, because there were problems with some AUTH
1308 commands, and RFC 1869 says the size should be increased for extended
1309 SMTP commands that take arguments.
1311 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1314 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1315 "unknown" error; now it says that the functionality isn't in the binary.
1317 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1318 an address' error message when a string expansion fails (syntax or
1319 whatever). Otherwise the password may appear in the log. Following change
1320 PH/42 below, there is no longer a chance of it appearing in a bounce
1323 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1325 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1326 of its addresses was ever tried. (Bugzilla bug #2.)
1328 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1329 the result incorrectly in the debug output. (It correctly added a newline
1330 to what was transported.)
1332 TF/01 Added $received_time.
1334 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1335 commented out examples of how to interface to a virus scanner and to
1336 SpamAssassin. Also added commented examples of av_scanner and
1337 spamd_address settings.
1339 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1340 and controls are allowed in which ACLs. There were a couple of minor
1341 errors. Some of the entries in the conditions table (which is a table of
1342 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1343 negation of where the condition IS allowed.
1345 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1347 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1348 header file does not have a version number, so I've had to invent a new
1349 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1350 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1351 radiusclient), but it was contributed by a Radius user.
1353 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1354 files or directories, for OpenSSL.
1356 PH/19 When an Exim process that is running as root has to create an Exim log
1357 file, it does so in a subprocess that runs as exim:exim so as to get the
1358 ownership right at creation (otherwise, other Exim processes might see
1359 the file with the wrong ownership). There was no test for failure of this
1360 fork() call, which would lead to the process getting stuck as it waited
1361 for a non-existent subprocess. Forks do occasionally fail when resources
1362 run out. I reviewed all the other calls to fork(); they all seem to check
1365 PH/20 When checking for unexpected SMTP input at connect time (before writing
1366 the banner), Exim was not dealing correctly with a non-positive return
1367 from the read() function. If the client had disconnected by this time,
1368 the result was a log entry for a synchronization error with an empty
1369 string after "input=" when read() returned zero. If read() returned -1
1370 (an event I could not check), uninitialized data bytes were printed.
1371 There were reports of junk text (parts of files, etc) appearing after
1374 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1376 PH/22 Added support for macro redefinition, and (re)definition in between
1377 driver and ACL definitions.
1379 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1380 forgetting to use the resulting value; it was using the unexpanded value.
1382 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1383 hadn't been configured. The fix is from Juergen Kreileder, who
1384 understands it better than I do:
1386 "Here's what I see happening with three configured cyrus_sasl
1387 authenticators configured (plain, login, cram-md5):
1389 On startup auth_cyrus_sasl_init() gets called for each of these.
1390 This means three calls to sasl_listmech() without a specified mech_list.
1391 => SASL tests which mechs of all available mechs actually work
1392 => three warnings about OTP not working
1393 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1395 With the patch, sasl_listmech() also gets called three times. But now
1396 SASL's mech_list option is set to the server_mech specified in the the
1397 authenticator. Or in other words, the answer from sasl_listmech()
1398 gets limited to just the mech you're testing for (which is different
1400 => the return list contains just 'plain' or 'login', 'cram-md5' or
1401 nothing depending on the value of ob->server_mech.
1403 I've just tested the patch: Authentication still works fine,
1404 unavailable mechs specified in the exim configuration are still
1405 caught, and the auth.log warnings about OTP are gone."
1407 PH/25 When debugging is enabled, the contents of the command line are added
1408 to the debugging output, even when log_selector=+arguments is not
1411 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1412 answer is "GNU", and only if the return is "GNU/something" is the answer
1415 PH/27 $acl_verify_message is now set immediately after the failure of a
1416 verification in an ACL, and so is available in subsequent modifiers. In
1417 particular, the message can be preserved by coding like this:
1419 warn !verify = sender
1420 set acl_m0 = $acl_verify_message
1422 Previously, $acl_verify_message was set only while expanding "message"
1423 and "log_message" when a very denied access.
1425 PH/28 Modified OS/os.c-Linux with
1427 -#ifndef OS_LOAD_AVERAGE
1428 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1430 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1431 nomenclature these days.)
1433 PH/29 Installed patch from the Sieve maintainer that adds the options
1434 sieve_useraddress and sieve_subaddress to the redirect router.
1436 PH/30 In these circumstances:
1437 . Two addresses routed to the same list of hosts;
1438 . First host does not offer TLS;
1439 . First host accepts first address;
1440 . First host gives temporary error to second address;
1441 . Second host offers TLS and a TLS session is established;
1442 . Second host accepts second address.
1443 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1444 and peerdn, if requested) that were in fact used only for the second
1447 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1448 attention to any local part prefix or suffix that was matched by the
1449 router that accepted the address. It now behaves in the same way as it
1450 does for delivery: the affixes are removed from the local part unless
1451 rcpt_include_affixes is set on the transport.
1453 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1454 timeout during the DATA phase of an incoming message.
1456 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1457 applied a patch sanctioned by the Sieve maintainer.
1459 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1460 the uid or gid is negative. A case of a negative gid caused this to be
1461 noticed. The fix allows for either to be negative.
1463 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1464 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1467 PH/36 The change PH/12 above was broken. Fixed it.
1469 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1470 the grounds that routing the same address twice would always produce the
1471 same answer. This might have been true once, but it is certainly no
1472 longer true now. Routing a child address may depend on the previous
1473 routing that produced that child. Some complicated redirection strategies
1474 went wrong when messages had multiple recipients, and made Exim's
1475 behaviour dependent on the order in which the addresses were given.
1477 I have moved the duplicate checking until after the routing is complete.
1478 Exim scans the addresses that are assigned to local and remote
1479 transports, and removes any duplicates. This means that more work will be
1480 done, as duplicates will always all be routed, but duplicates are
1481 presumably rare, so I don't expect this is of any significance.
1483 For deliveries to pipes, files, and autoreplies, the duplicate checking
1484 still happens during the routing process, since they are not going to be
1487 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1488 It corrects a timeout issue with spamd. This is Ian's comment: "The
1489 background is that sometimes spamd either never reads data from a
1490 connection it has accepted, or it never writes response data. The exiscan
1491 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1492 blindly assumes that writes won't block so it may never time out."
1494 PH/39 Allow G after quota size as well as K and M.
1496 PH/40 The value set for $authenticated_id in an authenticator may not contain
1497 binary zeroes or newlines because the value is written to log lines and
1498 to spool files. There was no check on this. Now the value is run through
1499 the string_printing() function so that such characters are converted to
1500 printable escape sequences.
1502 PH/41 $message_linecount is a new variable that contains the total number of
1503 lines in the message. Compare $body_linecount, which is the count for the
1506 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1507 bounce and delay warning messages, except in certain special cases, which
1510 (a) An SMTP error message from a remote host;
1511 (b) A message specified in a :fail: redirection;
1512 (c) A message specified in a "fail" command in a system filter;
1513 (d) A message specified in a FAIL return from the queryprogram router;
1514 (e) A message specified by the cannot_route_message router option.
1516 In these cases only, Exim does include the error details in bounce and
1517 warning messages. There are also a few cases where bland messages such
1518 as "unrouteable address" or "local delivery error" are given.
1520 PH/43 $value is now also set for the "else" part of a ${run expansion.
1522 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1523 being worked on, but at least Exim now implements the latest version to
1526 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1527 process to complete was treated as a delivery failure, a timeout while
1528 writing the message to the pipe was logged, but erroneously treated as a
1529 successful delivery. Such timeouts include transport filter timeouts. For
1530 consistency with the overall process timeout, these timeouts are now
1531 treated as errors, giving rise to delivery failures by default. However,
1532 there is now a new Boolean option for the pipe transport called
1533 timeout_defer, which, if set TRUE, converts the failures into defers for
1534 both kinds of timeout. A transport filter timeout is now identified in
1537 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1538 systems where "make" and "gmake" are different, calling "gmake" at top
1539 level broke things. I've arranged for the value of $(MAKE) to be passed
1540 from the Makefile to this script so that it can call the same version of
1544 A note about Exim versions 4.44 and 4.50
1545 ----------------------------------------
1547 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1548 changes of various kinds. As a consequence, a big documentation update was
1549 needed. This delayed the release for rather longer than seemed good, especially
1550 in the light of a couple of (minor) security issues. Therefore, the changes
1551 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1552 release. So 4.44 and 4.50 are in effect two different branches that both start
1555 I have left the 4.50 change log unchanged; it contains all the changes since
1556 4.43. The change log for 4.44 is below; many of its items are identical to
1557 those for 4.50. This seems to be the most sensible way to preserve the
1558 historical information.
1564 1. Minor wording change to the doc/README.SIEVE file.
1566 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1567 computation of the current number of files was incorrect.
1569 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1570 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1573 4. Give more explanation in the error message when the command for a transport
1574 filter fails to execute.
1576 5. There are several places where Exim runs a non-Exim command in a
1577 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1578 was being done only for the command run by the queryprogram router. It is
1579 now done for all such subprocesses. The other cases are: ${run, transport
1580 filters, and the commands run by the lmtp and pipe transports.
1582 6. Added CONFIGURE_GROUP build-time option.
1584 7. Some older OS have a limit of 256 on the maximum number of file
1585 descriptors. Exim was using setrlimit() to set 1000 as a large value
1586 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1587 systems. I've change it so that if it can't get 1000, it tries for 256.
1589 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1590 was an oversight, and furthermore, ever since the addition of extra
1591 controls (e.g. 4.43/32), the checks on when to allow different forms of
1592 "control" were broken. There should now be diagnostics for all cases when a
1593 control that does not make sense is encountered.
1595 9. Added the /retain_sender option to "control=submission".
1597 10. $recipients is now available in the predata ACL (oversight).
1599 11. Tidy the search cache before the fork to do a delivery from a message
1600 received from the command line. Otherwise the child will trigger a lookup
1601 failure and thereby defer the delivery if it tries to use (for example) a
1602 cached ldap connection that the parent has called unbind on.
1604 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1605 of $address_data from the recipient verification was clobbered by the
1606 sender verification.
1608 13. The value of address_data from a sender verification is now available in
1609 $sender_address_data in subsequent conditions in the ACL statement.
1611 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1613 15. Added a new option "connect=<time>" to callout options, to set a different
1616 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1617 was its contents. (It was OK if the option was not defined at all.)
1619 17. A "Completed" log line is now written for messages that are removed from
1620 the spool by the -Mrm option.
1622 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1623 information about exactly what failed.
1625 19. Added -dd to debug only the daemon process.
1627 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1628 handles timeouts, both on the server side and network timeouts. Renamed the
1629 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1631 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1632 It is now set to "smtps".
1634 22. $host_address is now set to the target address during the checking of
1635 ignore_target_hosts.
1637 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1638 being passed; this would have caused $sender_host_name to have been used if
1639 matching the list had actually called for a host name (not very likely,
1640 since this list is usually IP addresses). A host name is now passed as
1643 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1644 code that specifies a non-restarting handler (typically sigaction() in
1645 modern systems) in an attempt to fix a rare and obscure crash bug.
1647 25. Narrowed the window for a race in the daemon that could cause it to ignore
1648 SIGCHLD signals. This is not a major problem, because they are used only to
1649 wake it up if nothing else does.
1651 26. A malformed maildirsize file could cause Exim to calculate negative values
1652 for the mailbox size or file count. Odd effects could occur as a result.
1653 The maildirsize information is now recalculated if the size or filecount
1656 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1657 support for a long time. Removed HAVE_SYS_VFS_H.
1659 28. Installed the latest version of exipick from John Jetmore.
1661 29. In an address list, if the pattern was not a regular expression, an empty
1662 subject address (from a bounce message) matched only if the pattern was an
1663 empty string. Non-empty patterns were not even tested. This was the wrong
1664 because it is perfectly reasonable to use an empty address as part of a
1665 database query. An empty address is now tested by patterns that are
1666 lookups. However, all the other forms of pattern expect the subject to
1667 contain a local part and a domain, and therefore, for them, an empty
1668 address still always fails if the pattern is not itself empty.
1670 30. Exim went into a mad DNS loop when attempting to do a callout where the
1671 host was specified on an smtp transport, and looking it up yielded more
1672 than one IP address.
1674 31. Re-factored the code for checking spool and log partition space into a
1675 function that finds that data and another that does the check. The former
1676 is then used to implement four new variables: $spool_space, $log_space,
1677 $spool_inodes, and $log_inodes.
1679 32. The RFC2047 encoding function was originally intended for short strings
1680 such as real names; it was not keeping to the 75-character limit for
1681 encoded words that the RFC imposes. It now respects the limit, and
1682 generates multiple encoded words if necessary. To be on the safe side, I
1683 have increased the buffer size for the ${rfc2047: expansion operator from
1686 33. It is now permitted to omit both strings after an "if" condition; if the
1687 condition is true, the result is "true". As before, when the second string
1688 is omitted, a false condition yields an empty string. This makes it less
1689 cumbersome to write custom ACL and router conditions.
1691 34. Failure to deliver a bounce message always caused it to be frozen, even if
1692 there was an errors_to setting on the router. The errors_to setting is now
1695 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1696 canonical form (fully expanded) before being placed in
1697 $sender_host_address.
1699 36. The table in the code that translates DNS record types into text (T_A to
1700 "A" for instance) was missing entries for NS and CNAME. It is just possible
1701 that this could have caused confusion if both these types were looked up
1702 for the same domain, because the text type is used as part of Exim's
1703 per-process caching. But the chance of anyone hitting this buglet seems
1706 37. The dnsdb lookup has been extended in a number of ways.
1708 (1) There is a new type, "zns", which walks up the domain tree until it
1709 finds some nameserver records. It should be used with care.
1711 (2) There is a new type, "mxh", which is like "mx" except that it returns
1712 just the host names, not the priorities.
1714 (3) It is now possible to give a list of domains (or IP addresses) to be
1715 looked up. The behaviour when one of the lookups defers can be
1716 controlled by a keyword.
1718 (4) It is now possible to specify the separator character for use when
1719 multiple records are returned.
1721 38. The dnslists ACL condition has been extended: it is now possible to supply
1722 a list of IP addresses and/or domains to be looked up in a particular DNS
1725 39. Added log_selector=+queue_time_overall.
1727 40. When running the queue in the test harness, wait just a tad after forking a
1728 delivery process, to get repeatability of debugging output.
1730 41. Include certificate and key file names in error message when GnuTLS fails
1731 to set them up, because the GnuTLS error message doesn't include the name
1732 of the failing file when there is a problem reading it.
1734 42. Allow both -bf and -bF in the same test run.
1736 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1738 44. The "Exiscan patch" is now merged into the mainline Exim source.
1740 45. Sometimes the final signoff response after QUIT could fail to get
1741 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1742 < 0 before doing a fflush(). This bug looks as though it goes back to the
1743 introduction of TLS in release 3.20, but "sometimes" must have been rare
1744 because the tests only now provoked it.
1746 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1747 (this can affect the format of dates).
1749 47. exim_tidydb, when checking for the continued existence of a message for
1750 which it has found a message-specific retry record, was not finding
1751 messages that were in split spool directories. Consequently, it was
1752 deleting retry records that should have stayed in existence.
1754 48. Steve fixed some bugs in eximstats.
1756 49. The SPA authentication driver was not abandoning authentication and moving
1757 on to the next authenticator when an expansion was forced to fail,
1758 contradicting the general specification for all authenticators. Instead it
1759 was generating a temporary error. It now behaves as specified.
1761 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1762 (the order specifies the preference for clients). The order is now AES256,
1763 AES128, 3DES, ARCFOUR128.
1765 51. Small patch to Sieve code - explicitly set From: when generating an
1768 52. Exim crashed if a remote delivery caused a very long error message to be
1769 recorded - for instance if somebody sent an entire SpamAssassin report back
1770 as a large number of 550 error lines. This bug was coincidentally fixed by
1771 increasing the size of one of Exim's internal buffers (big_buffer) that
1772 happened as part of the Exiscan merge. However, to be on the safe side, I
1773 have made the code more robust (and fixed the comments that describe what
1776 53. Now that there can be additional text after "Completed" in log lines (if
1777 the queue_time_overall log selector is set), a one-byte patch to exigrep
1778 was needed to allow it to recognize "Completed" as not the last thing in
1781 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1782 patch that reportedly fixes this has been added. I am not expert enough to
1783 create a test for it. This is what the patch creator wrote:
1785 "I found a little strange behaviour of ldap code when working with
1786 Windows 2003 AD Domain, where users was placed in more than one
1787 Organization Units. When I tried to give exim partial DN, the exit code
1788 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1789 But simultaneously result of request was absolutely normal ldap result,
1790 so I produce this patch..."
1792 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1793 REFERENCE, so I have modified the code to exclude the patch when that macro
1796 55. Some experimental protocols are using DNS PTR records for new purposes. The
1797 keys for these records are domain names, not reversed IP addresses. The
1798 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1799 leaves it alone. Component reversal etc. now happens only for IP addresses.
1802 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1804 57. Double the size of the debug message buffer (to 2048) so that more of very
1805 long debug lines gets shown.
1807 58. The exicyclog utility now does better if the number of log files to keep
1808 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1810 59. Two changes related to the smtp_active_hostname option:
1812 (1) $smtp_active_hostname is now available as a variable.
1813 (2) The default for smtp_banner uses $smtp_active_hostname instead
1814 of $primary_hostname.
1816 60. The host_aton() function is supposed to be passed a string that is known
1817 to be a valid IP address. However, in the case of IPv6 addresses, it was
1818 not checking this. This is a hostage to fortune. Exim now panics and dies
1819 if the condition is not met. A case was found where this could be provoked
1820 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1821 components; fortuitously, this particular loophole had already been fixed
1822 by change 4.50/55 above.
1824 If there are any other similar loopholes, the new check in host_aton()
1825 itself should stop them being exploited. The report I received stated that
1826 data on the command line could provoke the exploit when Exim was running as
1827 exim, but did not say which command line option was involved. All I could
1828 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1829 running as the user.
1832 61. There was a buffer overflow vulnerability in the SPA authentication code
1833 (which came originally from the Samba project). I have added a test to the
1834 spa_base64_to_bits() function which I hope fixes it.
1837 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1838 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1840 63. The daemon start-up calls getloadavg() while still root for those OS that
1841 need the first call to be done as root, but it missed one case: when
1842 deliver_queue_load_max is set with deliver_drop_privilege. This is
1843 necessary for the benefit of the queue runner, because there is no re-exec
1844 when deliver_drop_privilege is set.
1846 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1847 This has been fixed.
1849 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1850 was in use, was not putting the data itself into the right store pool;
1851 consequently, it could be overwritten for a subsequent message in the same
1852 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1855 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1857 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1858 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1859 the function were treating the return as a boolean value, which happened to
1860 work because 0=false and not-0=true, but is not correct code.
1862 68. The host_aton() function was not handling scoped IPv6 addresses (those
1863 with, for example, "%eth0" on the end) correctly.
1865 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1866 negated items (that is, ~something) in unsigned ints. Some compilers
1867 apparently mutter when there is no cast.
1869 70. If an address verification called from an ACL failed, and did not produce a
1870 user-specific message (i.e. there was only a "system" message), nothing was
1871 put in $acl_verify_message. In this situation, it now puts the system
1874 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1875 session; change 4.31/43 added the unwanted input to the log line - except
1876 that it did not do this in the start of session case. It now does.
1878 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1879 This is wrong and can cause the other end to generate a synchronization
1880 error if it is another Exim or anything else that does the synchronization
1881 check. A QUIT command is no longer sent after a timeout.
1883 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1884 during host lookups.
1886 74. The defer_ok option of callout verification was not working if it was used
1887 when verifying addresses in header lines, that is, for this case:
1889 verify = header_sender/callout=defer_ok
1891 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1892 those file descriptors could be used for SMTP connections. If anything
1893 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1894 could be sent to the SMTP client, causing chaos. The daemon now opens
1895 stdin, stdout, and stderr to /dev/null when it puts itself into the
1898 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1899 log by default. The user can override this with suitable Perl magic.
1901 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1902 the log message when discard triggers, was not working for the DATA ACL or
1903 for the non-SMTP ACL.
1905 78. Error message wording change in sieve.c.
1907 79. If smtp_accept_max_per_host was set, the number of connections could be
1908 restricted to fewer than expected, because the daemon was trying to set up
1909 a new connection before checking whether the processes handling previous
1910 connections had finished. The check for completed processes is now done
1911 earlier. On busy systems, this bug wouldn't be noticed because something
1912 else would have woken the daemon, and it would have reaped the completed
1915 80. If a message was submitted locally by a user whose login name contained one
1916 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1917 It caused a spool format error. I have fixed the spool reading code. A
1918 related problem was that the "from" clause in the Received: line became
1919 illegal because of the space(s). It is now covered by ${quote_local_part.
1921 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1924 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1926 83. Patch from Sieve maintainer to fix unterminated string problem in
1927 "vacation" handling.
1929 84. Some minor changes to the Linux configuration files to help with other
1930 OS variants using glibc.
1932 85. One more patch for Sieve to update vacation handling to latest spec.
1935 ----------------------------------------------------
1936 See the note above about the 4.44 and 4.50 releases.
1937 ----------------------------------------------------
1943 1. Change 4.43/35 introduced a bug that caused file counts to be
1944 incorrectly computed when quota_filecount was set in an appendfile
1947 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1948 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1951 3. Give more explanation in the error message when the command for a transport
1952 filter fails to execute.
1954 4. There are several places where Exim runs a non-Exim command in a
1955 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1956 was being done only for the command run by the queryprogram router. It is
1957 now done for all such subprocesses. The other cases are: ${run, transport
1958 filters, and the commands run by the lmtp and pipe transports.
1960 5. Some older OS have a limit of 256 on the maximum number of file
1961 descriptors. Exim was using setrlimit() to set 1000 as a large value
1962 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1963 systems. I've change it so that if it can't get 1000, it tries for 256.
1965 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1966 was an oversight, and furthermore, ever since the addition of extra
1967 controls (e.g. 4.43/32), the checks on when to allow different forms of
1968 "control" were broken. There should now be diagnostics for all cases when a
1969 control that does not make sense is encountered.
1971 7. $recipients is now available in the predata ACL (oversight).
1973 8. Tidy the search cache before the fork to do a delivery from a message
1974 received from the command line. Otherwise the child will trigger a lookup
1975 failure and thereby defer the delivery if it tries to use (for example) a
1976 cached ldap connection that the parent has called unbind on.
1978 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1979 of $address_data from the recipient verification was clobbered by the
1980 sender verification.
1982 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1983 was its contents. (It was OK if the option was not defined at all.)
1985 11. A "Completed" log line is now written for messages that are removed from
1986 the spool by the -Mrm option.
1988 12. $host_address is now set to the target address during the checking of
1989 ignore_target_hosts.
1991 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1992 being passed; this would have caused $sender_host_name to have been used if
1993 matching the list had actually called for a host name (not very likely,
1994 since this list is usually IP addresses). A host name is now passed as
1997 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1998 code that specifies a non-restarting handler (typically sigaction() in
1999 modern systems) in an attempt to fix a rare and obscure crash bug.
2001 15. Narrowed the window for a race in the daemon that could cause it to ignore
2002 SIGCHLD signals. This is not a major problem, because they are used only to
2003 wake it up if nothing else does.
2005 16. A malformed maildirsize file could cause Exim to calculate negative values
2006 for the mailbox size or file count. Odd effects could occur as a result.
2007 The maildirsize information is now recalculated if the size or filecount
2010 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2011 support for a long time. Removed HAVE_SYS_VFS_H.
2013 18. Updated exipick to current release from John Jetmore.
2015 19. Allow an empty sender to be matched against a lookup in an address list.
2016 Previously the only cases considered were a regular expression, or an
2019 20. Exim went into a mad DNS lookup loop when doing a callout where the
2020 host was specified on the transport, if the DNS lookup yielded more than
2023 21. The RFC2047 encoding function was originally intended for short strings
2024 such as real names; it was not keeping to the 75-character limit for
2025 encoded words that the RFC imposes. It now respects the limit, and
2026 generates multiple encoded words if necessary. To be on the safe side, I
2027 have increased the buffer size for the ${rfc2047: expansion operator from
2030 22. Failure to deliver a bounce message always caused it to be frozen, even if
2031 there was an errors_to setting on the router. The errors_to setting is now
2034 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2035 canonical form (fully expanded) before being placed in
2036 $sender_host_address.
2038 24. Updated eximstats to version 1.33
2040 25. Include certificate and key file names in error message when GnuTLS fails
2041 to set them up, because the GnuTLS error message doesn't include the name
2042 of the failing file when there is a problem reading it.
2044 26. Expand error message when OpenSSL has problems setting up cert/key files.
2047 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2048 (this can affect the format of dates).
2050 28. exim_tidydb, when checking for the continued existence of a message for
2051 which it has found a message-specific retry record, was not finding
2052 messages that were in split spool directories. Consequently, it was
2053 deleting retry records that should have stayed in existence.
2055 29. eximstats updated to version 1.35
2056 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2057 1.35 - bugfix such that pie charts by volume are generated correctly
2059 30. The SPA authentication driver was not abandoning authentication and moving
2060 on to the next authenticator when an expansion was forced to fail,
2061 contradicting the general specification for all authenticators. Instead it
2062 was generating a temporary error. It now behaves as specified.
2064 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2065 (the order specifies the preference for clients). The order is now AES256,
2066 AES128, 3DES, ARCFOUR128.
2068 31. Small patch to Sieve code - explicitly set From: when generating an
2071 32. Exim crashed if a remote delivery caused a very long error message to be
2072 recorded - for instance if somebody sent an entire SpamAssassin report back
2073 as a large number of 550 error lines. This bug was coincidentally fixed by
2074 increasing the size of one of Exim's internal buffers (big_buffer) that
2075 happened as part of the Exiscan merge. However, to be on the safe side, I
2076 have made the code more robust (and fixed the comments that describe what
2079 33. Some experimental protocols are using DNS PTR records for new purposes. The
2080 keys for these records are domain names, not reversed IP addresses. The
2081 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2082 leaves it alone. Component reversal etc. now happens only for IP addresses.
2085 34. The host_aton() function is supposed to be passed a string that is known
2086 to be a valid IP address. However, in the case of IPv6 addresses, it was
2087 not checking this. This is a hostage to fortune. Exim now panics and dies
2088 if the condition is not met. A case was found where this could be provoked
2089 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2090 components; fortuitously, this particular loophole had already been fixed
2091 by change 4.50/55 or 4.44/33 above.
2093 If there are any other similar loopholes, the new check in host_aton()
2094 itself should stop them being exploited. The report I received stated that
2095 data on the command line could provoke the exploit when Exim was running as
2096 exim, but did not say which command line option was involved. All I could
2097 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2098 running as the user.
2101 35. There was a buffer overflow vulnerability in the SPA authentication code
2102 (which came originally from the Samba project). I have added a test to the
2103 spa_base64_to_bits() function which I hope fixes it.
2106 36. The daemon start-up calls getloadavg() while still root for those OS that
2107 need the first call to be done as root, but it missed one case: when
2108 deliver_queue_load_max is set with deliver_drop_privilege. This is
2109 necessary for the benefit of the queue runner, because there is no re-exec
2110 when deliver_drop_privilege is set.
2112 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2113 was in use, was not putting the data itself into the right store pool;
2114 consequently, it could be overwritten for a subsequent message in the same
2115 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2118 38. Sometimes the final signoff response after QUIT could fail to get
2119 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2120 < 0 before doing a fflush(). This bug looks as though it goes back to the
2121 introduction of TLS in release 3.20, but "sometimes" must have been rare
2122 because the tests only now provoked it.
2128 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2129 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2130 Now it returns an integer. A number of calls were still expecting a T/F
2131 return. Fortuitously, in all cases, the tests worked in OK situations,
2132 which is the norm. However, things would have gone wrong on any write
2133 failures on the smtp file descriptor. This function is used when sending
2134 messages over SMTP and also when doing verify callouts.
2136 2. When Exim is called to do synchronous delivery of a locally submitted
2137 message (the -odf or -odi options), it no longer closes stderr before doing
2140 3. Implemented the mua_wrapper option.
2142 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2144 5. Implemented the functions header_remove(), header_testname(),
2145 header_add_at_position(), and receive_remove_recipient(), and exported them
2148 6. If an ACL "warn" statement specified the addition of headers, Exim already
2149 inserted X-ACL-Warn: at the start if there was no header name. However, it
2150 was not making this test for the second and subsequent header lines if
2151 there were newlines in the string. This meant that an invalid header could
2152 be inserted if Exim was badly configured.
2154 7. Allow an ACL "warn" statement to add header lines at the start or after all
2155 the Received: headers, as well as at the end.
2157 8. Added the rcpt_4xx retry error code.
2159 9. Added postmaster_mailfrom=xxx to callout verification option.
2161 10. Added mailfrom=xxxx to the callout verification option, for verify=
2164 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2165 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2167 12. Inserted some casts to stop certain compilers warning when using pointer
2168 differences as field lengths or precisions in printf-type calls (mostly
2169 affecting debugging statements).
2171 13. Added optional readline() support for -be (dynamically loaded).
2173 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2174 same clock tick as a message's arrival, so that its received time was the
2175 same as the "first fail" time on the retry record, and that message
2176 remained on the queue past the ultimate address timeout, every queue runner
2177 would try a delivery (because it was past the ultimate address timeout) but
2178 after another failure, the ultimate address timeout, which should have then
2179 bounced the address, did not kick in. This was a "< instead of <=" error;
2180 in most cases the first failure would have been in the next clock tick
2181 after the received time, and all would be well.
2183 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2184 being recognized when the domain list was tested by the match_domain
2185 condition in an expansion string.
2187 16. Added the ${str2b64: operator.
2189 17. Exim was always calling setrlimit() to set a large limit for the number of
2190 processes, without checking whether the existing limit was already
2191 adequate. (It did check for the limit on file descriptors.) Furthermore,
2192 errors from getrlimit() and setrlimit() were being ignored. Now they are
2193 logged to the main and panic logs, but Exim does carry on, to try to do its
2194 job under whatever limits there are.
2196 18. Imported PCRE 5.0.
2198 19. Trivial typo in log message " temporarily refused connection" (the leading
2201 20. If the log selector return_path_on_delivery was set and an address was
2202 redirected to /dev/null, the delivery process crashed because it assumed
2203 that a return path would always be set for a "successful" delivery. In this
2204 case, the whole delivery is bypassed as an optimization, and therefore no
2207 21. Internal re-arrangement: the function for sending a challenge and reading
2208 a response while authentication was assuming a zero-terminated challenge
2209 string. It's now changed to take a pointer and a length, to allow for
2210 binary data in such strings.
2212 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2214 23. Exim was not respecting finduser_retries when seeking the login of the
2215 uid under which it was called; it was always trying 10 times. (The default
2216 setting of finduser_retries is zero.) Also, it was sleeping after the final
2217 failure, which is pointless.
2219 24. Implemented tls_on_connect_ports.
2221 25. Implemented acl_smtp_predata.
2223 26. If the domain in control=submission is set empty, Exim assumes that the
2224 authenticated id is a complete email address when it generates From: or
2225 Sender: header lines.
2227 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2228 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2229 chown and chgrp in /bin and hostname in /usr/bin.
2231 28. Exim was keeping the "process log" file open after each use, just as it
2232 does for the main log. This opens the possibility of it remaining open for
2233 long periods when the USR1 signal hits a daemon. Occasional processlog
2234 errors were reported, that could have been caused by this. Anyway, it seems
2235 much more sensible not to leave this file open at all, so that is what now
2238 29. The long-running daemon process does not normally write to the log once it
2239 has entered its main loop, and it closes the log before doing so. This is
2240 so that log files can straightforwardly be renamed and moved. However,
2241 there are a couple of unusual error situations where the daemon does write
2242 log entries, and I had neglected to close the log afterwards.
2244 30. The text of an SMTP error response that was received during a remote
2245 delivery was being truncated at 512 bytes. This is too short for some of
2246 the long messages that one sometimes sees. I've increased the limit to
2249 31. It is now possible to make retry rules that apply only when a message has a
2250 specific sender, in particular, an empty sender.
2252 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2253 it possible to be selective about when SMTP synchronization is enforced.
2255 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2257 32. Implemented hosts_connection_nolog.
2259 33. Added an ACL for QUIT.
2261 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2264 35. Added mailbox_size and mailbox_filecount to appendfile.
2266 36. Added control = no_multiline_responses to ACLs.
2268 37. There was a bug in the logic of the code that waits for the clock to tick
2269 in the case where the clock went backwards by a substantial amount such
2270 that the microsecond fraction of "now" was more than the microsecond
2271 fraction of "then" (but the whole seconds number was less).
2273 38. Added support for the libradius Radius client library this is found on
2274 FreeBSD (previously only the radiusclient library was supported).
2280 1. When certain lookups returned multiple values in the form name=value, the
2281 quoting of the values was not always being done properly. Specifically:
2282 (a) If the value started with a double quote, but contained no whitespace,
2284 (b) If the value contained whitespace other than a space character (i.e.
2285 tabs or newlines or carriage returns) it was not quoted.
2286 This fix has been applied to the mysql and pgsql lookups by writing a
2287 separate quoting function and calling it from the lookup code. The fix
2288 should probably also be applied to nisplus, ibase and oracle lookups, but
2289 since I cannot test any of those, I have not disturbed their existing code.
2291 2. A hit in the callout cache for a specific address caused a log line with no
2292 reason for rejecting RCPT. Now it says "Previous (cached) callout
2293 verification failure".
2295 3. There was an off-by-one bug in the queryprogram router. An over-long
2296 return line was truncated at 256 instead of 255 characters, thereby
2297 overflowing its buffer with the terminating zero. As well as fixing this, I
2298 have increased the buffer size to 1024 (and made a note to document this).
2300 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2301 when Exim is waiting for an SMTP response from a remote server, Exim
2302 restarts its select() call on the socket, thereby resetting its timeout.
2303 This is not a problem when such interrupts are rare. Somebody set up a cron
2304 job to run exiwhat every 2 minutes, which is less than the normal select()
2305 timeout (5 or 10 minutes). This meant that the select() timeout never
2306 kicked in because it was always reset. I have fixed this by comparing the
2307 time when an interrupt arrives with the time at the start of the first call
2308 to select(). If more time than the timeout has elapsed, the interrupt is
2309 treated as a timeout.
2311 5. Some internal re-factoring in preparation for the addition of Sieve
2312 extensions (by MH). In particular, the "personal" test is moved to a
2313 separate function, and given an option for scanning Cc: and Bcc: (which is
2314 not set for Exim filters).
2316 6. When Exim created an email address using the login of the caller as the
2317 local part (e.g. when creating a From: or Sender: header line), it was not
2318 quoting the local part when it contained special characters such as @.
2320 7. Installed new OpenBSD configuration files.
2322 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2323 try to make them clearer.
2325 9. Callout options, other than the timeout value, were being ignored when
2326 verifying sender addresses in header lines. For example, when using
2328 verify = header_sender/callout=no_cache
2330 the cache was (incorrectly) being used.
2332 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2333 only the Cygwin environment.
2335 11. When return_path_on_delivery was set as a log selector, if different remote
2336 addresses in the same message used different return paths and parallel
2337 remote delivery occurred, the wrong values would sometimes be logged.
2338 (Whenever a remote delivery process finished, the return path value from
2339 the most recently started remote delivery process was logged.)
2341 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2342 lines when AUTH and/or TLS are in use. This is the "received protocol"
2343 field. Exim used to use "asmtp" for authenticated SMTP, without any
2344 indication (in the protocol name) for TLS use. Now it follows the RFC and
2345 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2346 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2347 names appear in log lines as well as in Received: header lines.
2349 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2350 extensions, and comparison tests, and to fix some bugs.
2352 14. Changes to the "personal" filter test:
2354 (1) The test was buggy in that it was just doing the equivalent of
2355 "contains" tests on header lines. For example, if a user's address was
2356 anne@some.where, the "personal" test would incorrectly be true for
2358 To: susanne@some.where
2360 This test is now done by extracting each address from the header in turn,
2361 and checking the entire address. Other tests that are part of "personal"
2362 are now done using regular expressions (for example, to check local parts
2363 of addresses in From: header lines).
2365 (2) The list of non-personal local parts in From: addresses has been
2366 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2367 taken from the Sieve specification recommendations.
2369 (3) If the message contains any header line starting with "List-" it is
2370 treated as non-personal.
2372 (4) The test for "circular" in the Subject: header line has been removed
2373 because it now seems ill-conceived.
2375 15. Minor typos in src/EDITME comments corrected.
2377 16. Installed latest exipick from John Jetmore.
2379 17. If headers_add on a router specified a text string that was too long for
2380 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2381 of string_sprintf() is now avoided.
2383 18. $message_body_size was not set (it was always zero) when running the DATA
2384 ACL and the local_scan() function.
2386 19. For the "mail" command in an Exim filter, no default was being set for
2387 the once_repeat time, causing a random time value to be used if "once" was
2388 specified. (If the value happened to be <= 0, no repeat happened.) The
2389 default is now 0s, meaning "never repeat". The "vacation" command was OK
2390 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2391 (I found it when inspecting the code).
2393 20. There is now an overall timeout for performing a callout verification. It
2394 defaults to 4 times the callout timeout, which applies to individual SMTP
2395 commands during the callout. The overall timeout applies when there is more
2396 than one host that can be tried. The timeout is checked before trying the
2397 next host. This prevents very long delays if there are a large number of
2398 hosts and all are timing out (e.g. when the network connections are timing
2399 out). The value of the overall timeout can be changed by specifying an
2400 additional sub-option for "callout", called "maxwait". For example:
2402 verify = sender/callout=5s,maxwait=20s
2404 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2405 to the end before writing, but this should make it even safer).
2407 22. Exim was forgetting that it had advertised PIPELINING for the second and
2408 subsequent messages on an SMTP connection. It was also not resetting its
2409 memory on STARTTLS and an internal HELO.
2411 23. When Exim logs an SMTP synchronization error within a session, it now
2412 records whether PIPELINING has been advertised or not.
2414 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2415 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2416 rather than long int.
2418 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2420 26. Added the never_mail option to autoreply.
2426 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2427 crash if the getsockname() call failed; this can happen if a connection is
2428 closed very soon after it is established. The problem was simply in the
2429 order in which certain operations were done, causing Exim to try to write
2430 to the SMTP stream before it had set up the file descriptor. The bug has
2431 been fixed by making things happen in the correct order.
2437 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2438 before the connection was closed, thus losing the rejection response.
2440 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2441 some early Solaris releases, but causes trouble in current releases where
2442 socklen_t is defined.
2444 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2447 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2450 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2451 that releasing the top of it at the end releases what was used for sub-
2452 expansions (unless the block got too big). However, discard this block if
2453 the first thing is a variable or header, so that we can use its block when
2454 it is dynamic (useful for very large $message_headers, for example).
2456 6. Lookups now cache *every* query, not just the most recent. A new, separate
2457 store pool is used for this. It can be recovered when all lookup caches are
2458 flushed. Lookups now release memory at the end of their result strings.
2459 This has involved some general refactoring of the lookup sources.
2461 7. Some code has been added to the store_xxx() functions to reduce the amount
2462 of flapping under certain conditions.
2464 8. log_incoming_interface used to affect only the <= reception log lines. Now
2465 it causes the local interface and port to be added to several more SMTP log
2466 lines, for example "SMTP connection from", and rejection lines.
2468 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2470 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2472 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2473 could be overwritten at the end of the current message (or the start of a
2474 new message if it was set in a HELO ACL). The value is now preserved for
2475 the duration of the SMTP connection.
2477 12. If a transport had a headers_rewrite setting, and a matching header line
2478 contained an unqualified address, that address was qualified, even if it
2479 did not match any rewriting rules. The underlying bug was that the values
2480 of the flags that permit the existence of unqualified sender and recipient
2481 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2482 non-local messages, and by -bnq for local messages) were not being
2483 preserved with the message after it was received.
2485 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2486 "next input=" as part of the text comprising the host identity instead of
2487 the correct text. The code was using the same buffer for two different
2488 strings. However, depending on which order the printing function evaluated
2489 its arguments, the bug did not always show up. Under Linux, for example, my
2490 test suite worked just fine.
2492 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2493 doesn't work with some older versions of Perl. It has been changed to "my",
2494 which in any case is probably the better facility to use.
2496 15. A really picky compiler found some instances of statements for creating
2497 error messages that either had too many or two few arguments for the format
2500 16. The size of the buffer for calls to the DNS resolver has been increased
2501 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2502 for addresses that have a lot of PTR records. This alleviates a problem; it
2503 does not fully solve it.
2505 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2506 buffer now truncates the list and logs the incident, which is the same
2507 action as happens when Exim is looking up a host name and its aliases.
2508 Previously in this situation something unpredictable would happen;
2509 sometimes it was "internal error: store_reset failed".
2511 18. If a server dropped the connection unexpectedly when an Exim client was
2512 using GnuTLS and trying to read a response, the client delivery process
2513 crashed while trying to generate an error log message.
2515 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2516 string, for example:
2518 warn message = H1: something\nH2: something
2520 the text was added as a single header line from Exim's point of view
2521 though it ended up OK in the delivered message. However, searching for the
2522 second and subsequent header lines using $h_h2: did not work. This has been
2523 fixed. Similarly, if a system filter added multiple headers in this way,
2524 the routers could not see them.
2526 20. Expanded the error message when iplsearch is called with an invalid key to
2527 suggest using net-iplsearch in a host list.
2529 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2530 ACLs are no longer actually imposed (and a message to that effect is
2533 22. If a "gecos" field in a passwd entry contained escaped characters, in
2534 particular, if it contained a \" sequence, Exim got it wrong when building
2535 a From: or a Sender: header from that name. A second bug also caused
2536 incorrect handling when an unquoted " was present following a character
2537 that needed quoting.
2539 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2540 was not being matched caselessly.
2542 24. Arranged for all hyphens in the exim.8 source to be escaped with
2545 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2546 properly. Recipient callout cache records were still being keyed to include
2547 the sender, even when use_sender was set false. This led to far more
2548 callouts that were necessary. The sender is no longer included in the key
2549 when use_sender is false.
2551 26. Added "control = submission" modifier to ACLs.
2553 27. Added the ${base62d: operator to decode base 62 numbers.
2555 28. dnsdb lookups can now access SRV records.
2557 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2558 the configuration file.
2560 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2561 (-v) mode. This makes the output for a verbose queue run more intelligible.
2563 31. Added a use_postmaster feature to recipient callouts.
2565 32. Added the $body_zerocount variable, containing the number of binary zero
2566 bytes in the message body.
2568 33. The time of last modification of the "new" subdirectory is now used as the
2569 "mailbox time last read" when there is a quota error for a maildir
2572 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2574 35. Added +ignore_unknown as a special item in host lists.
2576 36. Code for decoding IPv6 addresses in host lists is now included, even if
2577 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2578 address was recognized as an IP address, but was then not correctly decoded
2579 into binary, causing unexpected and incorrect effects when compared with
2586 1. Very minor rewording of debugging text in manualroute to say "list of
2587 hosts" instead of "hostlist".
2589 2. If verify=header_syntax was set, and a header line with an unqualified
2590 address (no domain) and a large number of spaces between the end of the
2591 name and the colon was received, the reception process suffered a buffer
2592 overflow, and (when I tested it) crashed. This was caused by some obsolete
2593 code that should have been removed. The fix is to remove it!
2595 3. When running in the test harness, delay a bit after writing a bounce
2596 message to get a bit more predictability in the log output.
2598 4. Added a call to search_tidyup() just before forking a reception process. In
2599 theory, someone could use a lookup in the expansion of smtp_accept_max_
2600 per_host which, without the tidyup, could leave open a database connection.
2602 5. Added the variables $recipient_data and $sender_data which get set from a
2603 lookup success in an ACL "recipients" or "senders" condition, or a router
2604 "senders" option, similar to $domain_data and $local_part_data.
2606 6. Moved the writing of debug_print from before to after the "senders" test
2609 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2610 problems for message scanning, either using a data ACL, or using
2611 local_scan() because the Received: header was not generated till after they
2612 were called (in order to set the time as the time of reception completion).
2613 I have revised the way this works. The header is now generated after the
2614 body is received, but before the ACL or local_scan() are called. After they
2615 are run, the timestamp in the header is updated.
2621 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2622 before starting a queue runner without re-exec. This happened only when
2623 deliver_drop_privilege was set or when the Exim user was set to root. The
2624 effect of the bug was that timeouts during subsequent deliveries caused
2625 crashes instead of being properly handled. The handler is now left at its
2626 default (and expected) setting.
2628 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2629 message, again when deliver_drop_privilege is set or Exim is run as root.
2630 The bug described in (1) was not present in this case, but the tidying up
2631 of the other signals was missing. I have made the two cases consistent.
2633 3. The ignore_target_hosts setting on a manualroute router was being ignored
2634 for hosts that were looked up using the /MX notation.
2636 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2639 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2640 operated on the sender address. After changing the $sender_address to <>
2641 for the sender address verify, Exim was re-instated it as the original
2642 (before rewriting) address, but remembering that it had rewritten it, so it
2643 wasn't rewriting it again. This bug also had the effect of breaking the
2644 sender address verification caching when the sender address was rewritten.
2646 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2647 This has been changed so that if the ip literal address matches
2648 ignore_target_hosts, the router declines.
2650 7. Added expansion conditions match_domain, match_address, and match_local_
2651 part (NOT match_host).
2653 8. The placeholder for the Received: header didn't have a length field set.
2655 9. Added code to Exim itself and to exim_lock to test for a specific race
2656 condition that could lead to file corruption when using MBX delivery. The
2657 issue is with the lockfile that is created in /tmp. If this file is removed
2658 after a process has opened it but before that process has acquired a lock,
2659 there is the potential for a second process to recreate the file and also
2660 acquire a lock. This could lead to two Exim processes writing to the file
2661 at the same time. The added code performs the same test as UW imapd; it
2662 checks after acquiring the lock that its file descriptor still refers to
2663 the same named file.
2665 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2666 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2669 11. Added the smtp_active_hostname option. If used, this will typically be made
2670 to depend on the incoming interface address. Because $interface_address is
2671 not set up until the daemon has forked a reception process, error responses
2672 that can happen earlier (such as "too many connections") no longer contain
2675 12. If an expansion in a condition on a "warn" statement fails because a lookup
2676 defers, the "warn" statement is abandoned, and the next ACL statement is
2677 processed. Previously this caused the whole ACL to be aborted.
2679 13. Added the iplsearch lookup type.
2681 14. Added ident_timeout as a log selector.
2683 15. Added tls_certificate_verified as a log selector.
2685 16. Added a global option tls_require_ciphers (compare the smtp transport
2686 option of the same name). This controls incoming TLS connections.
2688 17. I finally figured out how to make tls_require_ciphers do a similar thing
2689 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2690 before starting the TLS session.
2692 18. Tabs are now shown as \t in -bP output.
2694 19. If the log selector return_path_on_delivery was set, Exim crashed when
2695 bouncing a message because it had too many Received: header lines.
2697 20. If two routers both had headers_remove settings, and the first one included
2698 a superfluous trailing colon, the final name in the first list and the
2699 first name in the second list were incorrectly joined into one item (with a
2700 colon in the middle).
2706 1. Added -C and -D options to the exinext utility, mainly to make it easier
2707 to include in the automated testing, but these could be helpful when
2708 multiple configurations are in use.
2710 2. The exinext utility was not formatting the output nicely when there was
2711 an alternate port involved in the retry record key, nor when there was a
2712 message id as well (for retries that were specific to a specific message
2713 and a specific host). It was also confused by IPv6 addresses, because of
2714 the additional colons they contain. I have fixed the IPv4 problem, and
2715 patched it up to do a reasonable job for IPv6.
2717 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2718 delivery, the log line now contains "pipelined" if PIPELINING was used.
2720 4. An SMTP transport process used to panic and die if the bind() call to set
2721 an explicit outgoing interface failed. This has been changed; it is now
2722 treated in the same way as a connect() failure.
2724 5. A reference to $sender_host_name in the part of a conditional expansion
2725 that was being skipped was still causing a DNS lookup. This no longer
2728 6. The def: expansion condition was not recognizing references to header lines
2729 that used bh_ and bheader_.
2731 7. Added the _cache feature to named lists.
2733 8. The code for checking quota_filecount in the appendfile transport was
2734 allowing one more file than it should have been.
2736 9. For compatibility with Sendmail, the command line option
2744 and sets the incoming protocol and host name (for trusted callers). The
2745 host name and its colon can be omitted when only the protocol is to be set.
2746 Note the Exim already has two private options, -pd and -ps, that refer to
2747 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2748 "s", but I don't think that's a major issue.
2750 10. A number of refactoring changes to the code, none of which should affect
2753 (a) The number of logging options was getting close to filling up the
2754 32-bit word that was used as a bit map. I have split them into two classes:
2755 those that are passed in the argument to log_write(), and those that are
2756 only ever tested independently outside of that function. These are now in
2757 separate 32-bit words, so there is plenty of room for expansion again.
2758 There is no change in the user interface or the logging behaviour.
2760 (b) When building, for example, log lines, the code previously used a
2761 macro that called string_cat() twice, in order to add two strings. This is
2762 not really sufficiently general. Furthermore, there was one instance where
2763 it was actually wrong because one of the argument was used twice, and in
2764 one call a function was used. (As it happened, calling the function twice
2765 did not affect the overall behaviour.) The macro has been replaced by a
2766 function that can join an arbitrary number of extra strings onto a growing
2769 (c) The code for expansion conditions now uses a table and a binary chop
2770 instead of a serial search (which was left over from when there were very
2771 few conditions). Also, it now recognizes conditions like "pam" even when
2772 the relevant support is not compiled in: a suitably worded error message is
2773 given if an attempt is made to use such a condition.
2775 11. Added ${time_interval:xxxxx}.
2777 12. A bug was causing one of the ddress fields not to be passed back correctly
2778 from remote delivery subprocesses. The field in question was not being
2779 subsequently used, so this caused to problems in practice.
2781 13. Added new log selectors queue_time and deliver_time.
2783 14. Might have fixed a bug in maildirsizefile handling that threw up
2784 "unexpected character" debug warnings, and recalculated the data
2785 unnecessarily. In any case, I expanded the warning message to give more
2788 15. Added the message "Restricted characters in address" to the statements in
2789 the default ACL that block characters like @ and % in local parts.
2791 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2792 Three changes have been made:
2794 (a) There was a serious bug; a negative response to MAIL caused the whole
2795 recipient domain to be cached as invalid, thereby blocking all messages
2796 to all local parts at the same domain, from all senders. This bug has
2797 been fixed. The domain is no longer cached after a negative response to
2798 MAIL if the sender used is not empty.
2800 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2803 (c) A new callout option, "use_sender" has been added for people who want
2804 the modified behaviour.
2810 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2813 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2814 indeed breaks things for older releases.
2816 3. Added additional logging to the case where there is a problem reading data
2817 from a filter that is running in a subprocess using a pipe, in order to
2818 try to track down a specific problem.
2820 4. Testing facility fudge: when running in the test harness and attempting
2821 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2822 getting "No route to host". Convert this to a timeout.
2824 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2827 6. Some OS don't have socklen_t but use size_t instead. This affects the
2828 fifth argument of getsockopt() amongst other things. This is now
2829 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2830 can be set for individual OS. I have set it for SunOS5, OSF1, and
2831 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2832 some earlier ones do not.
2834 7. Change 4.30/15 was not doing the test caselessly.
2836 8. The standard form for an IPv6 address literal was being rejected by address
2837 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2838 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2839 this, as well as the form without the "IPv6" on the front (but only when
2840 address literals are enabled, of course).
2842 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2844 10. Exim crashed if a message with an empty sender address specified by -f
2845 encountered a router with an errors_to setting. This could be provoked only
2846 by a command such as
2850 where an empty string was supplied; "<>" did not hit this bug.
2852 11. Installed PCRE release 4.5.
2854 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2855 remained set. It is now erased.
2857 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2858 times from message ids (which are base 36 rather than the normal 62).
2860 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2861 were being counted as actual protocol errors, and logged if the log
2862 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2863 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2864 and DATA following a set of rejected RCPTs do not count as protocol errors.
2865 In other words, Exim assumes they were pipelined, though this may not
2866 actually be the case. Of course, in all cases the client gets an
2867 appropriate error code.
2869 15. If a lookup fails in an ACL condition, a message about the failure may
2870 be available; it is used if testing the ACL cannot continue, because most
2871 such messages specify what the cause of the deferral is. However, some
2872 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2873 that caused an old message to be retained and used if a later statement
2874 caused a defer, replacing the real cause of the deferral.
2876 16. If an IP address had so many PTR records that the DNS lookup buffer
2877 was not large enough to hold them, Exim could crash while trying to process
2878 the truncated data. It now detects and logs this case.
2880 17. Further to 4.21/58, another change has been made: if (and only if) the
2881 first line of a message (the first header line) ends with CRLF, a bare LF
2882 in a subsequent header line has a space inserted after it, so as not to
2883 terminate the header.
2885 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2886 unnecessarily, used atoi() instead of strtol(), and didn't check the
2887 termination when getting file sizes from file names by regex.
2889 19. Completely re-implemented the support for maildirsize files, in the light
2890 of a number of problems with the previous contributed implementation
2891 (4.30/29). In particular:
2893 . If the quota is zero, the maildirsize file is maintained, but no quota is
2896 . If the maildir directory does not exist, it is created before any attempt
2897 to write a maildirsize file.
2899 . The quota value in the file is just a cache; if the quota is changed in
2900 the transport, the new value overrides.
2902 . A regular expression is available for excluding directories from the
2905 20. The autoreply transport checks the characters in options that define the
2906 message's headers; it allows continued headers, but it was checking with
2907 isspace() after an embedded newline instead of explicitly looking for a
2910 21. If all the "regular" hosts to which an address was routed had passed their
2911 expiry times, and had not reached their retry times, the address was
2912 bounced, even if fallback hosts were defined. Now Exim should go on to try
2915 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2916 equivalent code in the SMTP transport. Some hosts send humungous responses
2917 to HELO/EHLO, more than 1024 it seems.
2919 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2920 gives compiler warnings in some environments. I've now done it "properly",
2923 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2924 (because of problems with the built-in one) was declared to return uschar *
2925 instead of char *, causing compiler failure.
2927 25. Fixed a file descriptor leak when processing alias/forward files.
2929 26. Fixed a minor format string issue in dbfn.c.
2931 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2933 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2934 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2937 29. When a sender address is verified, it is cached, to save repeating the test
2938 when there is more than one recipient in a message. However, when the
2939 verification involves a callout, it is possible for different callout
2940 options to be set for different recipients. It is too complicated to keep
2941 track of this in the cache, so now Exim always runs a verification when a
2942 callout is required, relying on the callout cache for the optimization.
2943 The overhead is duplication of the address routing, but this should not be
2946 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2947 to be verified with callout=postmaster, and the main callout worked but the
2948 postmaster check failed, the verification correctly failed. However, if a
2949 subsequent RCPT command asked for sender verification *without* the
2950 postmaster check, incorrect caching caused this verification also to fail,
2953 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2954 it was not caching the DNS options (qualify_single, search_parents) that
2955 were used when the lookup failed. A subsequent lookup with different
2956 options therefore always gave the same answer, though there were cases
2957 where it should not have. (Example: a "domains = !$mx_any" option on a
2958 dnslookup router: the "domains" option is always processed without any
2959 widening, but the router might have qualify_single set.) Now Exim uses the
2960 cached value only when the same options are set.
2962 32. Added John Jetmore's "exipick" utility to the distribution.
2964 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2965 than a timeout (e.g. a certificate is required, and is not provided), an
2966 Exim server now closes the connection immediately. Previously it waited for
2967 the client to close - but if the client is SSL, it seems that they each
2968 wait for each other, leading to a delay before one of them times out.
2970 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2971 maintained 0.8.x compatibility because I don't think many are using it, and
2972 it is clearly obsolete.
2974 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2977 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2978 client certificate was expired. A simple patch fixes this, though I don't
2979 understand the full logic of why the verify callback is called multiple
2982 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2983 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2984 which causes problems with some clients (such as the Certicom SSL Plus
2985 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2986 disables the coutermeasure allowing Eudora to connect."
2988 38. Exim was not checking that a write() to a log file succeeded. This could
2989 lead to Bad Things if a log got too big, in particular if it hit a file
2990 size limit. Exim now panics and dies if it cannot write to a log file, just
2991 as it does if it cannot open a log file.
2993 39. Modified OS/Makefile-Linux so that it now contains
2995 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2997 The two -D definitions ensure that Exim is compiled with large file
2998 support, which makes it possible to handle log files that are bigger than
3001 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3002 instance) a domain was checked against a named list that involved a lookup,
3003 causing $domain_data to be set, then another domain was checked against the
3004 same list, then the first domain was re-checked, the value of $domain_data
3005 after the final check could be wrong. In particular, if the second check
3006 failed, it could be set empty. This bug probably also applied to
3009 41. The strip_trailing_dot option was not being applied to the address given
3010 with the -f command-line option.
3012 42. The code for reading a message's header from the spool was incrementing
3013 $received_count, but never initializing it. This meant that the value was
3014 incorrect (doubled) while delivering a message in the same process in which
3015 it was received. In the most common configuration of Exim, this never
3016 happens - a fresh exec is done - but it can happen when
3017 deliver_drop_privilege is set.
3019 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3020 it now includes up to 150 characters of the unexpected data in the log
3023 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3024 and building data strings. The size of both of these buffers was 10 000
3025 bytes - far larger than anybody would *ever* want, thought I. Needless to
3026 say, somebody hit the limit. I have increased the maximum line length to
3027 20 000 and the maximum data length of concatenated lines to 100 000. I have
3028 also fixed two bugs, because there was no checking on these buffers. Tsk,
3029 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3030 buffer is too small.
3032 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3033 lsearch lookups. Now it does.
3035 46. When parsing a route_list item in a manualroute router, a fixed-length
3036 buffer was used for the list of hosts. I made this 1024 bytes long,
3037 thinking that nobody would ever have a list of hosts that long. Wrong.
3038 Somebody had a whole pile of complicated expansion conditions, and the
3039 string was silently truncated, leading to an expansion error. It turns out
3040 that it is easier to change to an unlimited length (owing to other changes
3041 that have happened since this code was originally written) than to build
3042 structure for giving a limitation error. The length of the item that
3043 expands into the list of hosts is now unlimited.
3045 47. The lsearch lookup could not handle data where the length of text line was
3046 more than 4095 characters. Such lines were truncated, leading to shortened
3047 data being returned. It should now handle lines of any length.
3049 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3050 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3053 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3054 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3055 no longer happens in comment lines. A list of replacements is now placed
3056 at the head of all of the source files, except those whose only change is
3057 to replace PERL_COMMAND in the very first #! line.
3059 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3060 messages on the queue, with a bottom-up merge sort, using code contributed
3061 by Michael Haardt. This should make operations like -bp somewhat faster on
3062 large queues. It won't affect queue runners, except when queue_run_in_order
3065 51. Installed eximstats 1.31 in the distribution.
3067 52. Added support for SRV lookups to the dnslookup router.
3069 53. If an ACL referred to $message_body or $message_body_end, the value was not
3070 reset for any messages that followed in the same SMTP session.
3072 54. The store-handling optimization for building very long strings was not
3073 differentiating between the different store pools. I don't think this
3074 actually made any difference in practice, but I've tidied it.
3076 55. While running the routers to verify a sender address, $sender_address
3077 was still set to the sender address. This is wrong, because when routing to
3078 send a bounce to the sender, it would be empty. Therefore, I have changed
3079 it so that, while verifying a sender address, $sender_address is set to <>.
3080 (There is no change to what happens when verifying a recipient address.)
3082 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3083 target A or AAAA records (if not already returned) without resetting the
3084 qualify_single or search_parents options of the DNS resolver. These are
3085 inappropriate in this case because the targets of MX and SRV records must
3086 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3087 target that, when qualified, matched something in the local domain. These
3088 two options are now turned off when doing these lookups.
3090 57. It seems that at least some releases of Reiserfs (which does not have the
3091 concept of a fixed number of inodes) returns zero and not -1 for the
3092 number of available inodes. This interacted badly with check_spool_inodes,
3093 which assumed that -1 was the "no such thing" setting. What I have done is
3094 to check that the total number of inodes is greater than zero before doing
3095 the test of how many are available.
3097 58. When a "warn" ACL statement has a log_message modifier, the message is
3098 remembered, and not repeated. This is to avoid a lot of repetition when a
3099 message has many recipients that cause the same warning to be written.
3100 Howewer, Exim was preserving the list of already written lines for an
3101 entire SMTP session, which doesn't seem right. The memory is now reset if a
3102 new message is started.
3104 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3105 debugging output unless log_rewrite was also set.
3107 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3108 of (int)(handle) when we know that handle contains (void *)(-1).
3110 61. The Exim daemon panic-logs an error return when it closes the incoming
3111 connection. However "connection reset by peer" seems to be common, and
3112 isn't really an error worthy of noting specially, so that particular error
3115 62. When Exim is trying to find all the local interfaces, it used to panic and
3116 die if the ioctl to get the interface flags failed. However, it seems that
3117 on at least one OS (Solaris 9) it is possible to have an interface that is
3118 included in the list of interfaces, but for which you get a failure error
3119 for this call. This happens when the interface is not "plumbed" into a
3120 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3121 failure of the "get flags" call assumes that the interface is down.
3123 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3124 makes life easier for people who are doing arithmetic on fields extracted
3125 from dates, where you often get leading zeros that should not be
3126 interpreted as octal.
3128 64. Added qualify_domain to the redirect router, to override the global
3131 65. If a pathologically long header line contained very many addresses (the
3132 report of this problem mentioned 10 000) and each of them was rewritten,
3133 Exim could use up a very large amount of memory. (It kept on making new
3134 copies of the header line as it rewrote, and never released the old ones.)
3135 At the expense of a bit more processing, the header rewriting function has
3136 been changed so that it no longer eats memory in this way.
3138 66. The generation of the Received: header has been moved from the time that a
3139 message starts to be received, to the time that it finishes. The timestamp
3140 in the Received: header should now be very close to that of the <= log
3141 line. There are two side-effects of this change:
3143 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3144 logged header lines no longer include the local Received: line, because
3145 it has not yet been created. The same applies to a copy of the message
3146 that is returned to a non-SMTP sender when a message is rejected.
3148 (b) When a filter file is tested using -bf, no additional Received: header
3149 is added to the test message. After some thought, I decided that this
3152 This change does not affect the value of $received_for. It is still set
3153 after address rewriting, but before local_scan() is called.
3155 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3157 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3158 gave an unhelpful panic error message, and a defer error. I have managed to
3159 change this behaviour so that it now rejects any supplied certificate,
3160 which seems right, as the list of acceptable certificates is empty.
3162 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3163 gave an unhelpful defer error. I have not managed to make this reject any
3164 supplied certificates, but the error message it gives is "no certificate
3165 supplied", which is not helpful.
3167 70. exigrep's output now also includes lines that are not associated with any
3168 message, but which match the given pattern. Implemented by a patch from
3169 Martin Sluka, which also tidied up the Perl a bit.
3171 71. Recipient callout verification, like sender verification, was using <> in
3172 the MAIL FROM command. This isn't really the right thing, since the actual
3173 sender may affect whether the remote host accepts the recipient or not. I
3174 have changed it to use the actual sender in the callout; this means that
3175 the cache record is now keyed on a recipient/sender pair, not just the
3176 recipient address. There doesn't seem to be a real danger of callout loops,
3177 since a callout by the remote host to check the sender would use <>.
3178 [SEE ABOVE: changed after hitting problems.]
3180 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3181 temporary errors. However, in the case of such a code being given after
3182 the end of a data transmission (i.e. after ".") Exim was failing to write
3183 a retry record for the message. (Yes, there was some broken host that was
3184 actually sending 8xx at this point.)
3186 73. An unknown lookup type in a host list could cause Exim to panic-die when
3187 the list was checked. (An example that provoked this was putting <; in the
3188 middle of a list instead of at the start.) If this happened during a DATA
3189 ACL check, a -D file could be left lying around. This kind of configuration
3190 error no longer causes Exim to die; instead it causes a defer errror. The
3191 incident is still logged to the main and panic logs.
3193 74. Buglet left over from Exim 3 conversion. The message "too many messages
3194 in one connection" was written to the rejectlog but not the mainlog, except
3195 when address rewriting (yes!) was being logged.
3197 75. Added write_rejectlog option.
3199 76. When a system filter was run not as root (that is, when system_filter_user
3200 was set), the values of the $n variables were not being returned to the
3201 main process; thus, they were not subsequently available in the $sn
3204 77. Added +return_path_on_delivery log selector.
3206 78. A connection timeout was being treated differently from recipients deferred
3207 when testing hosts_max_try with a message that was older than the host's
3208 retry timeout. (The host should not be counted, thus allowing all hosts to
3209 be tried at least once before bouncing.) This may have been the cause of an
3210 occasionally reported bug whereby a message would remain on the queue
3211 longer than the retry timeout, but would be bounced if a delivery was
3212 forced. I say "may" because I never totally pinned down the problem;
3213 setting up timeout/retry tests is difficult. See also the next item.
3215 79. The ultimate address timeout was not being applied to errors that involved
3216 a combination of host plus message (for example, a timeout on a MAIL
3217 command). When an address resolved to a number of possible hosts, and they
3218 were not all tried for each delivery (e.g. because of hosts_max_try), a
3219 message could remain on the queue longer than the retry timeout.
3221 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3224 81. Fixed an obscure SMTP outgoing bug which required at least the following
3225 conditions: (a) there was another message waiting for the same server;
3226 (b) the server returned 5xx to all RCPT commands in the first message so
3227 that the message was not completed; (c) the server dropped the connection
3228 or gave a negative response to the RSET that Exim sends to abort the
3229 transaction. The observed case was a dropped connection after DATA that had
3230 been sent in pipelining mode. That is, the server had advertised PIPELINING
3231 but was not implementing it correctly. The effect of the bug was incorrect
3232 behaviour, such as trying another host, and this could lead to a crash.
3238 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3239 and daemon.c were passed as pointers to ints; they should have been
3240 pointers to socklen_t variables (which are typically unsigned ints).
3242 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3245 3. Fixed a really odd bug that affected only the testing scheme; patching a
3246 certain fixed string in the binary changed the value of another string that
3247 happened to be identical to the end of the original first string.
3249 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3250 name", it returns that address as the IP address. On some operating
3251 systems (e.g. Solaris), it also passes back the IP address string as the
3252 "host name". However, on others (e.g. Linux), it passes back an empty
3253 string. Exim wasn't checking for this, and was changing the host name to an
3254 empty string, assuming it had been canonicized.
3256 5. Although rare, it is permitted to have more than one PTR record for a given
3257 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3258 all the names associated with an address, because they do in Solaris.
3259 However, it seems that they do not in Linux for data that comes from the
3260 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3261 I found this out when I moved to a new Linux workstation and tried to run
3262 the Exim test suite.
3264 To get round this problem I have changed the code so that it now does its
3265 own call to the DNS to look up PTR records when searching for a host name.
3266 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3267 addresses that are only in /etc/hosts are still found.
3269 This behaviour is, however, controlled by an option called host_lookup_
3270 order, which defaults to "bydns:byaddr". If people want to use the other
3271 order, or indeed, just use one or the other means of lookup, they can
3272 specify it in this variable.
3274 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3275 some operating systems, this comes back from gethostbyaddr() as an empty
3276 string, and this is what Exim used to test for. However, it seems that in
3277 other systems, "." is yielded. Exim now tests for this case too.
3279 7. The values of check_spool_space and check_log_space are now held internally
3280 as a number of kilobytes instead of an absolute number of bytes. If a
3281 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3282 kilobyte. This means that much larger values can be stored.
3284 8. Exim monitor: an attempt to get the action menu when not actually pointing
3285 at a message produces an empty menu entitled "No message selected". This
3286 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3287 no entries in it ("Shell widget menu has zero width and/or height"). So I
3288 have added a single, blank menu entry in this case.
3290 9. Added ${quote_local_part.
3292 10. MIME decoding is now applied to the contents of Subject: header lines when
3295 11. Now that a reference to $sender_host_address automatically causes a reverse
3296 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3297 host lookup before query-style lookups in lists that might use this
3298 variable. This has therefore been abolished, and the "net-" prefix is no
3299 longer necessary for query-style lookups.
3301 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3302 have been a typo for LFLAGS, so it has been changed.
3304 13. The install script calls Exim with "-C /dev/null" in order to find the
3305 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3306 to be output. Howeve, since Exim outputs its version number before the
3307 error, it didn't break the script. It just looked ugly. I fixed this by
3308 always allowing "-C /dev/null" if the caller is root.
3310 14. Ignore overlarge ACL variable number when reading spool file - insurance
3311 against a later release with more variables having written the file.
3313 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3314 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3315 this, as well as the form without the "IPv6" on the front.
3317 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3318 OS/Makefile-Darwin file.
3320 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3321 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3323 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3324 variable. However, some OS use ints for inodes. I've added cast to long int
3325 to get rid of the compiler warning.
3327 19. I had forgotten to lock out "/../" in configuration file names when
3328 ALT_CONFIG_PREFIX was set.
3330 20. Routers used for verification do not need to specify transports. However,
3331 if such a router generated a host list, and callout was configured, Exim
3332 crashed, because it could not find a port number from the (non-existent)
3333 transport. It now assumes port 25 in this circumstance.
3335 21. Added the -t option to exigrep.
3337 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3338 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3339 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3340 them is set, LOOKUP_LSEARCH is forced.
3342 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3344 24. Added sender and host information to the "rejected by local_scan()" log
3345 line; previously there was no indication of these.
3347 25. Added .include_if_exists.
3349 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3350 Linux. It turns out that not all file systems support this. Apparently some
3351 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3352 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3353 not supported on the file descriptor, is now ignored when Exim is trying to
3354 sync a directory. This applies only to Linux.
3356 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3358 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3359 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3362 29. Added support for maildirsize files from supplied patch (modified a bit).
3364 30. The use of :fail: followed by an empty string could lead Exim to respond to
3365 sender verification failures with (e.g.):
3367 550 Verification failed for <xxx>
3368 550 Sender verify failed
3370 where the first response line was missing the '-' that indicates it is not
3371 the final line of the response.
3373 31. The loop for finding the name of the user that called Exim had a hardwired
3374 limit of 10; it now uses the value of finduser_retries, which is used for
3375 all other user lookups.
3377 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3380 33. Exim was neglecting to zero errno before one call of strtol() when
3381 expanding a string and expecting an integer value. On some systems this
3382 resulted in spurious "integer overflow" errors. Also, it was casting the
3383 result into an int without checking.
3385 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3386 did not work. The code looks as if it has *never* worked, though it appears
3387 to have been documented since at least releast 1.62. I have made it work.
3389 35. The "timeout_DNS" error in retry rules, also documented since at least
3390 1.62, also never worked. As it isn't clear exactly what this means, and
3391 clearly it isn't a major issue, I have abolished the feature by treating it
3392 as "timeout", and writing a warning to the main and panic logs.
3394 36. The display of retry rules for -brt wasn't always showing the error code
3397 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3398 timeout_connect_A, timeout_connect_MX.
3400 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3401 to the empty sender.
3403 39. The daemon was not analysing the content of -oX till after it had closed
3404 stderr and disconnected from the controlling terminal. This meant that any
3405 syntax errors were only noted on the panic log, and the return code from
3406 the command was 0. By re-arranging the code a little, I've made the
3407 decoding happen first, so such errors now appear on stderr, and the return
3408 code is 1. However, the actual setting up of the sockets still happens in
3409 the disconnected process, so errors there are still only recorded on the
3412 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3413 connections (as happens on some IP stacks) was logged at start up time as
3414 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3415 it from "IPv6 and IPv4", which means that two separate sockets are being
3418 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3419 says whether AF_INET or AF_INET6 was passed as an argument.
3421 42. Exiwhat output was messed up when time zones were included in log
3424 43. Exiwhat now gives more information about the daemon's listening ports,
3425 and whether -tls-on-connect was used.
3427 44. The "port" option of the smtp transport is now expanded.
3429 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3430 silently ignored. Now an error message is written to the main and panic
3433 46. There's a new ACL modifier called "logwrite" which writes to a log file
3434 as soon as it is encountered.
3436 47. Added $local_user_uid and $local_user_gid at routing time.
3438 48. Exim crashed when trying to verify a sender address that was being
3441 49. Exim was recognizing only a space character after ".include". It now also
3442 recognizes a tab character.
3444 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3445 extracting the relevant information from the specification. The man page no
3446 longer contains scrambled data for the -d option, and I've added a section
3447 at the front about calling Exim under different names.
3449 51. Added "extra_headers" argument to the "mail" command in filter files.
3451 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3454 53. Installed eximstats 1.29.
3456 54. Added transport_filter_timeout as a generic transport option.
3458 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3459 Cc: header lines. This was required by RFC 822, but it not required by RFC
3462 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3463 incoming messages that did not have them. Now it does so only if the
3464 message originates locally, that is, if there is no associated remote host
3465 address. When Resent- header lines are present, this applies to the Resent-
3466 lines rather than the non-Resent- lines.
3468 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3469 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3471 58. Messages for configuration errors now include the name of the main
3472 configuration file - useful now that there may be more than one file in a
3473 list (.included file names were always shown).
3475 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3476 for those rare installations that do not start the daemon as root or run it
3477 setuid root. I've cut out the call to initgroups() if the daemon is not
3480 60. The Exim user and group can now be bound into the binary as text strings
3481 that are looked up at the start of Exim's processing.
3483 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3485 62. Added $mailstore_basename variable.
3487 63. Installed patch to sieve.c from Michael Haardt.
3489 64. When Exim failed to open the panic log after failing to open the main log,
3490 the original message it was trying to log was written to stderr and debug
3491 output, but if they were not available (the usual case in production), it
3492 was lost. Now it is written to syslog before the two lines that record the
3493 failures to open the logs.
3495 65. Users' Exim filters run in subprocesses under the user's uid. It is
3496 possible for a "deliver" command or an alias in a "personal" command to
3497 provoke an address rewrite. If logging of address rewriting is configured,
3498 this fails because the process is not running as root or exim. There may be
3499 a better way of dealing with this, but for the moment (because 4.30 needs
3500 to be released), I have disabled address rewrite logging when running a
3501 filter in a non-root, non-exim process.
3507 1. The buildconfig auxiliary program wasn't quoting the value set for
3508 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3509 not defined. This bug was masked in 4.22 by the effect that was fixed in
3512 2. Some messages that were rejected after a message id was allocated were
3513 shown as "incomplete" by exigrep. It no longer does this for messages that
3514 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3516 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3517 have allow_domain_literals set, the ID did not get logged in the <= line.
3518 Domain literals are now always recognized in Message-ID: header lines.
3520 4. The first argument for a ${extract expansion item is the key name or field
3521 number. Leading and trailing spaces in this item were not being ignored,
3522 causing some misleading effects.
3524 5. When deliver_drop_privilege was set, single queue runner processes started
3525 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3526 same command in the process it spins off) were not dropping privilege.
3528 6. When the daemon running as "exim" started a queue runner, it always
3529 re-executed Exim in the spun-off process. This is a waste of effort when
3530 deliver_drop_privilege is set. The new process now just calls the
3531 queue-runner function directly.
3537 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3540 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3541 ignored. Though the use of -oP was forcing the writing of a pid file, it
3542 was always written to the default place.
3544 3. If the message "no IP address found for host xxxx" is generated during
3545 incoming verification, it is now followed by identification of the incoming
3546 connection (so you can more easily find what provoked it).
3548 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3550 5. Added some features to "harden" Exim a bit more against certain attacks:
3552 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3553 be put in Local/Makefile. This is like the never_users runtime option,
3554 but it cannot be overridden. The default setting is "root".
3556 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3557 prefix string with which any file named in a -C command line option
3560 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3561 is retained for -C and -D only if the caller of Exim is root. Without
3562 it, the exim user may also use -C and -D and retain privilege.
3564 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3565 command line option is disabled.
3567 6. Macro names set by the -D option must start with an upper case letter, just
3568 like macro names defined in the configuration file.
3570 7. Added "dereference=" facility to LDAP.
3572 8. Two instances of the typo "uknown" in the source files are fixed.
3574 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3575 the Configure-Makefile script screwed up while processing it.
3577 10. Incorporated PCRE 4.4.
3579 11. The SMTP synchronization check was not operating right at the start of an
3580 SMTP session. For example, it could not catch a HELO sent before the client
3581 waited for the greeting. There is now a check for outstanding input at the
3582 point when the greeting is written. Because of the duplex, asynchronous
3583 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3584 way, but not yet received, when the check is performed.
3586 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3587 on TCP/IP sockets, because this apparently causes some broken clients to
3590 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3591 unchanged) from the Cygwin maintainer.
3593 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3594 was supported instead of testing for mbx. Effectively a typo.
3596 15. The spa authenticator server code was not checking that the input it
3597 received was valid base64.
3599 16. The debug output line for the "set" modifier in ACLs was not showing the
3600 name of the variable that was being set.
3602 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3604 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3605 lookup on its own. Something else was needed to trigger the lookup. For
3606 example, a match in host_lookup or the need for a host name in a host list.
3607 Now, if $sender_host_name is referenced and the host name has not yet been
3608 looked up, a lookup is performed. If the lookup fails, the variable remains
3609 empty, and $host_lookup_failed is set to "1".
3611 19. Added "eqi" as a case-independent comparison operator.
3613 20. The saslauthd authentication condition could segfault if neither service
3614 nor realm was specified.
3616 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3617 error message that was logged was misleading, and incoming SMTP
3618 connections were dropped. The message is now more accurate, and temporary
3619 errors are given to SMTP connections.
3621 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3622 (except RSET) with a 421 error, until QUIT is received. However, it was
3623 failing to send a response to QUIT.
3625 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3626 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3627 ACL is now run after the helo_try_verify_hosts code.
3629 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3630 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3631 case-independent, so other case variants are also recognized). Apparently
3632 some systems use these upper case variants.
3634 25. If more than two messages were waiting for the same host, and a transport
3635 filter was specified for the transport, Exim sent two messages over the
3636 same TCP/IP connection, and then failed with "socket operation on non-
3637 socket" when it tried to send the third.
3639 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3641 27. The extern definition of crypt16() in expand.c was not being excluded when
3642 the OS had its own crypt16() function.
3644 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3645 as a preferred synonym for return_size_limit, both as an option and as an
3648 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3650 30. Changed the default configuration ACL to relax the local part checking rule
3651 for addresses that are not in any local domains. For these addresses,
3652 slashes and pipe symbols are allowed within local parts, but the sequence
3653 /../ is explicitly forbidden.
3655 31. SPA server authentication was not clearing the challenge buffer before
3658 32. log_message in a "warn" ACL statement was writing to the reject log as
3659 well as to the main log, which contradicts the documentation and doesn't
3660 seem right (because no rejection is happening). So I have stopped it.
3662 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3663 However, I am unable to do any testing of this.
3665 34. Fixed an infelicity in the appendfile transport. When checking directories
3666 for a mailbox, to see if any needed to be created, it was accidentally
3667 using path names with one or more superfluous leading slashes; tracing
3668 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3670 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3671 contents are added to the log line that is written for every discarded
3672 recipient. (Previously a log_message setting was ignored.)
3674 36. The ${quote: operator now quotes the string if it is empty.
3676 37. The install script runs exim in order to find its version number. If for
3677 some reason other than non-existence or emptiness, which it checks, it
3678 could not run './exim', it was installing it with an empty version number,
3679 i.e. as "exim-". This error state is now caught, and the installation is
3682 38. An argument was missing from the function that creates an error message
3683 when Exim fails to connect to the socket for saslauthd authentication.
3684 This could cause Exim to crash, or give a corrupted message.
3686 39. Added isip, isip4, and isip6 to ${if conditions.
3688 40. The ACL variables $acl_xx are now saved with the message, and can be
3689 accessed later in routers, transports, and filters.
3691 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3692 strings in the file are not string-expanded.
3694 42. If a MAIL command specified a SIZE value that was too large to fit into an
3695 int variable, the check against message_size_limit failed. Such values are
3696 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3697 day this will have to be increased, but I don't think I want to be around
3698 when emails are that large.
3705 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3706 iconv() is not standard in FreeBSD.
3708 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3709 IPv6 enabled. The observed symptom was a segmentation fault on return from
3710 the function os_common_find_running_interfaces() in src/os.c.
3712 3. In the check_special_case() function in daemon.c I had used "errno" as an
3713 argument name, which causes warnings on some systems. This was basically a
3714 typo, since it was named "eno" in the comments!
3716 4. The code that waits for the clock to tick (at a resolution of some fraction
3717 of a second) so as to ensure message-id uniqueness was always waiting for
3718 at least one whole tick, when it could have waited for less. [This is
3719 almost certainly not relevant at current processor speeds, where it is
3720 unlikely to ever wait at all. But we try to future-proof.]
3722 5. The function that sleeps for a time interval that includes fractions of a
3723 second contained a race. It did not block SIGALRM between setting the
3724 timer, and suspending (a couple of lines later). If the interval was short
3725 and the sigsuspend() was delayed until after it had expired, the suspension
3726 never ended. On busy systems this could lead to processes getting stuck for
3729 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3730 process, before it forks any delivery processes. The open lookup caching
3731 mechanism meant that the open file or database connection was passed into
3732 the delivery process. The problem was that delivery processes always tidy
3733 up cached lookup data. This could cause a problem for the next delivery
3734 process started by the queue runner, because the external queue runner
3735 process does not know about the closure. So the next delivery process
3736 still has data in the lookup cache. In the case of a file lookup, there was
3737 no problem because closing a file descriptor in a subprocess doesn't affect
3738 the parent. However, if the lookup was caching a connection to a database,
3739 the connection was closed, and the second delivery process was likely to
3740 see errors such as "PGSQL: query failed: server closed the connection
3741 unexpectedly". The problem has been fixed by closing all cached lookups
3742 in a queue runner before running a delivery process.
3744 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3745 seem to have the "const" qualifier which it has on other OS. I've
3748 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3749 *of the same type* (client or server) with the same public name that an
3750 error should be diagnosed.
3752 9. When Exim looked up a host name for an IP address, but failed to find the
3753 original IP address when looking up the host name (a safety check), it
3754 output the message "<ip address> does not match any IP for NULL", which was
3755 confusing, to say the least. The bug was that the host name should have
3756 appeared instead of "NULL".
3758 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3759 user that is built into the binary, and the -C or -D options is used, root
3760 privilege is dropped before the configuration file is read. In addition,
3761 logging is switched to stderr instead of the normal log files. If the
3762 configuration then re-defines the Exim user, the unprivileged environment
3763 is probably not what is expected, so Exim logs a panic warning message (but
3766 However, if deliver_drop_privilege is set, the unprivileged state may well
3767 be exactly what is intended, so the warning has been cut out in that case,
3768 and Exim is allowed to try to write to its normal log files.
3774 1. smtp_return_error_details was not giving details for temporary sender
3775 or receiver verification errors.
3777 2. Diagnose a configuration error if two authenticators have the same public
3780 3. Exim used not to create the message log file for a message until the first
3781 delivery attempt. This could be confusing when incoming messages were held
3782 for policy or load reasons. The message log file is now created at the time
3783 the message is received, and an initial "Received" line is written to it.
3785 4. The automatically generated man page for command line options had a minor
3786 bug that caused no ill effects; however, a more serious problem was that
3787 the procedure for building the man page automatically didn't always
3788 operate. Consequently, release 4.20 contains an out-of-date version. This
3789 shouldn't happen again.
3791 5. When building Exim with embedded Perl support, the script that builds the
3792 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3793 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3795 6. The freeze_tell option was not being used for messages that were frozen on
3796 arrival, either by an ACL or by local_scan().
3798 7. Added the smtp_incomplete_transaction log selector.
3800 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3801 was accepting AUTH without a new EHLO.
3803 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3804 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3807 10. Logging of TCP/IP connections (when configured) now happens in the main
3808 daemon process instead of the child process, so that the TCP/IP connection
3809 count is more accurate (but it can never be perfect).
3811 11. The use of "drop" in a nested ACL was not being handled correctly in the
3812 outer ACL. Now, if condition failure induced by the nested "drop" causes
3813 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3814 or "require"), the connection is dropped.
3816 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3817 that yield "discard" can now be used with an "accept" or a "discard" verb,
3818 but an error is generated for any others (because I can't see a useful way
3819 to define what should happen).
3821 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3822 are now processed in the same way as lines in the Exim configuration file.
3823 In particular, continuation lines are supported.
3825 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3827 15. Added -ti meaning -t -i.
3829 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3830 domains, and warn by logging if others are found.
3832 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3833 returned from GIFCONF when Exim is trying to find the list of interfaces on
3834 a host. The code in os.c has been modified to copy each ifreq to an aligned
3835 structure in all cases.
3837 Also, in some cases, the returned ifreq's were being copied to a 'struct
3838 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3839 means the last couple of bytes of an IPv6 address could be chopped if the
3840 ifreq contained only a normal sockaddr (14 bytes storage).
3842 18. Named domain lists were not supported in the hosts_treat_as_local option.
3843 An entry such as +xxxx was not recognized, and was treated as a literal
3846 19. Ensure that header lines added by a DATA ACL are included in the reject log
3847 if the ACL subsequently rejects the message.
3849 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3850 MD5 (which is deprecated).
3852 21. When testing a filter file using -bf, Exim was writing a message when it
3853 took the sender from a "From " line in the message, but it was not doing so
3854 when it took $return_path from a Return-Path: header line. It now does.
3856 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3857 with a valid header line field name (a series of printing characters
3858 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3860 23. Changed "disc" in the source to "disk" to conform to the documentation and
3861 the book and for uniformity.
3863 24. Ignore Sendmail's -Ooption=value command line item.
3865 25. When execve() failed while trying to run a command in a pipe transport,
3866 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3867 could be confused with a return value of 69 from the command itself. This
3868 has been changed to 127, the value the shell returns if it is asked to run
3869 a non-existent command. The wording for the related log line suggests a
3870 non-existent command as the problem.
3872 26. If received_header_text expands to an empty string, do not add a Received:
3873 header line to the message. (Well, it adds a token one on the spool, but
3874 marks it "old" so that it doesn't get used or transmitted.)
3876 27. Installed eximstats 1.28 (addition of -nt option).
3878 28. There was no check for failure on the call to getsockname() in the daemon
3879 code. This can fail if there is a shortage of resources on the system, with
3880 ENOMEM, for example. A temporary error is now given on failure.
3882 29. Contrary to the C standard, it seems that in some environments, the
3883 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3884 program. Exim now does this explicitly; it affects the formatting of
3885 timestamps using strftime().
3887 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3888 complaints. I've now initialized all the variables, to avoid this.
3890 32. Header lines added by a system filter were not being "seen" during
3891 transport-time rewrites.
3893 33. The info_callback() function passed to OpenSSL is set up with type void
3894 (*)(SSL *, int, int), as described somewhere. However, when calling the
3895 function (actually a macro) that sets it up, the type void(*)() is
3896 expected. I've put in a cast to prevent warnings from picky compilers.
3898 34. If a DNS black list lookup found a CNAME record, but there were no A
3899 records associated with the domain it pointed at, Exim crashed.
3901 35. If a DNS black list lookup returned more than one A record, Exim ignored
3902 all but the first. It now scans all returned addresses if a particular IP
3903 value is being sought. In this situation, the contents of the
3904 $dnslist_value variable are a list of all the addresses, separated by a
3907 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3908 to accept a host name and all its aliases if the forward lookup for any of
3909 them yielded the IP address of the incoming connection. Now it accepts only
3910 those names whose forward lookup yields the correct IP address. Any other
3911 names are discarded. This closes a loophole whereby a rogue DNS
3912 administrator could create reverse DNS records to break through a
3913 wildcarded host restriction in an ACL.
3915 37. If a user filter or a system filter that ran in a subprocess used any of
3916 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3917 the wrong values were passed to the pipe command ($thisaddress had the
3918 value of $0, $0 had the value of $1, etc). This bug was introduced by
3919 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3921 38. Improved the line breaking for long SMTP error messages from ACLs.
3922 Previously, if there was no break point between 40 and 75 characters, Exim
3923 left the rest of the message alone. Two changes have been made: (a) I've
3924 reduced the minimum length to 35 characters; (b) if it can't find a break
3925 point between 35 and 75 characters, it looks ahead and uses the first one
3926 that it finds. This may give the occasional overlong line, but at least the
3927 remaining text gets split now.
3929 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3930 file descriptors might be low, and that setting 1000 would always raise it.
3931 It turns out that in some environments, the limit is already over 1000 and
3932 that lowering it causes trouble. So now Exim takes care not to decrease it.
3934 40. When delivering a message, the value of $return_path is set to $sender_
3935 address at the start of routing (routers may change the value). By an
3936 oversight, this default was not being set up when an address was tested by
3937 -bt or -bv, which affected the outcome if any router or filter referred to
3940 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3941 log only when "message" or "log_message" are set. However, if one of the
3942 conditions was an address verification, or a call to a nested ACL, the
3943 messages generated by the underlying test were being passed through. This
3944 no longer happens. The underlying message is available in $acl_verify_
3945 message for both "message" and "log_message" expansions, so it can be
3946 passed through if needed.
3948 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3949 new expansion $bh_ to give the encoded byte string without charset
3950 translation. Translation happens only if iconv() is available; HAVE_ICONV
3951 indicates this at build time. HEADERS_CHARSET gives the charset to
3952 translate to; headers_charset can change it in the configuration, and
3953 "headers charset" can change it in an individual filter file.
3955 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3956 that creates RFC 2047 encoded "words" labels them as that charset instead
3957 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3958 expansion operator; (ii) when Exim creates a From: line for a local
3959 message; (iii) when a header line is rewritten to include a "phrase" part.
3961 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3962 buggy, causing it to skip the first lines of messages whose message ID
3963 ended in 'D'. This would not have bitten before Exim release 4.14, because
3964 message IDs were unlikely to end in 'D' before then. The effect was to have
3965 incorrect size information for certain domains.
3967 45. #include "config.h" was missing at the start of the crypt16.c module. This
3968 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3971 46. If there was a timeout during a "random" callout check, Exim treated it as
3972 a failure of the random address, and carried on sending RSET and the real
3973 address. If the delay was just some slowness somewhere, the response to the
3974 original RCPT would be taken as a response to RSET and so on, causing
3975 mayhem of various kinds.
3977 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3978 when I implemented it. It didn't allow for the fact that some option values
3979 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3980 the right test for positive values.
3982 48. Domain names in DNS records are case-independent. Exim always looks them up
3983 in lower case. Some resolvers return domain names in exactly the case they
3984 appear in the zone file, that is, they may contain uppercase letters. Not
3985 all resolvers do this - some return always lower case. Exim was treating a
3986 change of case by a resolver as a change of domain, similar to a widening
3987 of a domain abbreviation. This triggered its re-routing code and so it was
3988 trying to route what was effectively the same domain again. This normally
3989 caused routing to fail (because the router wouldn't handle the domain
3990 twice). Now Exim checks for this case specially, and just changes the
3991 casing of the domain that it ultimately uses when it transmits the message
3994 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3997 50. If a filter generated a file delivery with a non-absolute name (possible if
3998 no home directory exists for the router), the forbid_file option was not
4001 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4002 the existing equality matching.
4004 52. Exim was using ints instead of ino_t variables in some places where it was
4005 dealing with inode numbers.
4007 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4008 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4009 called TMPDIR, and if it finds it is different, it changes its value.
4011 54. The smtp_printf() function is now made available to local_scan() so
4012 additional output lines can be written before returning. There is also an
4013 smtp_fflush() function to enable the detection of a dropped connection.
4014 The variables smtp_input and smtp_batched_input are exported to
4017 55. Changed the default runtime configuration: the message "Unknown user"
4018 has been removed from the ACL, and instead placed on the localuser router,
4019 using the cannot_route_message feature. This means that any verification
4020 failures that generate their own messages won't get overridden. Similarly,
4021 the "Unrouteable address" message that was in the ACL for unverifiable
4022 relay addresses has also been removed.
4024 56. Added hosts_avoid_esmtp to the smtp transport.
4026 57. The exicyclog script was not checking for the esoteric option
4027 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4028 will work only if exicyclog is run under the appropriate euid.
4030 58. Following a discussion on the list, the rules by which Exim recognises line
4031 endings on incoming messages have been changed. The -dropcr and drop_cr
4032 options are now no-ops, retained only for backwards compatibility. The
4033 following line terminators are recognized: LF CRLF CR. However, special
4034 processing applies to CR:
4036 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4037 nor a local message in the state where . is a terminator.
4039 (ii) If a bare CR is encountered in a header line, an extra space is added
4040 after the line terminator so as not to end the header. The reasoning
4041 behind this is that bare CRs in header lines are most likely either
4042 to be mistakes, or people trying to play silly games.
4044 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4045 was being incorrectly given as 18 bytes larger than it should have been.
4046 This is a VOB (very old bug).
4048 60. This may never have affected anything current, but just in case it has:
4049 When the local host is found other than at the start of a list of hosts,
4050 the local host, those with the same MX, and any that follow, are discarded.
4051 When the list in question was part of a longer list of hosts, the following
4052 hosts (not currently being processed) were also being discarded. This no
4053 longer happens. I'm not sure if this situation could ever has previously
4056 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4059 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4060 header. This is something that is recommended in a new Internet Draft, and
4061 is something that is documented as being done by Sendmail. There are two
4062 possible values. For messages generated by the autoreply transport, Exim
4065 Auto-Submitted: auto-replied
4067 whereas for all other generated messages (e.g. bounces) it adds
4069 Auto-Submitted: auto-generated
4071 63. The "personal" condition in filters now includes a test for the
4072 Auto-Submitted: header. If it contains the string "auto-" the message it
4073 not considered personal.
4075 64. Added rcpt_include_affixes as a generic transport option.
4077 65. Added queue_only_override (default true).
4079 66. Added the syslog_duplication option.
4081 67. If what should have been the first header line of a message consisted of
4082 a space followed by a colon, Exim was mis-interpreting it as a header line.
4083 It isn't of course - it is syntactically invalid and should therefore be
4084 treated as the start of the message body. The misbehaviour could have
4085 caused a number of strange effects, including loss of data in subsequent
4086 header lines, and spool format errors.
4088 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4089 client host had authenticated. This control can now be exercised by an ACL
4090 for more flexibility.
4092 69. By default, callouts do not happen when testing with -bh. There is now a
4093 variant, -bhc, which does actually run the callout code, including
4094 consulting and updating the callout cache.
4096 70. Added support for saslauthd authentication, courtesy of Alexander
4099 71. If statvfs() failed on the spool or log directories while checking their
4100 size for availability, Exim confusingly gave the error "space shortage".
4101 Furthermore, in debugging mode it crashed with a floating point exception.
4102 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4103 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4104 serious problem, Exim now writes to the main and panic logs when this
4105 happens, with details of the failure. It then refuses to accept the
4106 incoming message, giving the message "spool directory problem" or "log
4107 directory problem" with a 421 code for SMTP messages.
4109 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4110 0, 1, and 2 exist, because some OS complain for execs without them (see
4111 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4112 descriptors. However, the code omitted to check that the open succeeded,
4113 causing mysterious errors if for some reason the permissions on /dev/null
4114 got screwed. Now Exim writes a message to the main and panic logs, and
4115 bombs out if it can't open /dev/null.
4117 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4118 interact so that it is all more flexible. It is supposed to remain
4119 backwards compatible. Also added extra_local_interfaces.
4121 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4122 to bomb out with an assertion failure - to the client this appears as a
4123 connection drop. This problem occurs in the part of the code that was taken
4124 from the Samba project. Fortunately, the assertion is in a very simple
4125 function, so I have fixed this by reproducing the function inline in the
4126 one place where it is called, and arranging for authentication to fail
4127 instead of killing the process with assert().
4129 75. The SPA client code was not working when the server requested OEM rather
4130 than Unicode encoding.
4132 76. Added code to make require_files with a specific uid setting more usable in
4133 the case where statting the file as root fails - usually a non-root-mounted
4134 NFS file system. When this happens and the failure is EACCES, Exim now
4135 forks a subprocess and does the per-uid checking as the relevant uid.
4137 77. Added process_log_path.
4139 78. If log_file_path was not explicitly set, a setting of check_log_space or
4140 check_log_inodes was ignored.
4142 79. If a space check for the spool or log partitions fails, the incident is now
4143 logged. Of course, in the latter case the data may get lost...
4145 80. Added the %p formatting code to string_format() so that it can be used to
4146 print addresses in debug_print(). Adjusted all the address printing in the
4147 debugging in store.c to use %p rather than %d.
4149 81. There was a concern that a line of code in smtp_in.c could overflow a
4150 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4151 initially expressed, the concern was not well-founded, because trailing
4152 spaces are removed early. However, if the trailing spaces were followed by
4153 a NULL, they did not get removed, so the overflow was possible. Two fixes
4156 (a) I re-wrote the offending code in a cleaner fashion.
4157 (b) If an incoming SMTP command contains a NULL character, it is rejected
4160 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4161 runs initgroups(), so that if the Exim user is in any additional groups,
4162 they will be used during message reception.
4168 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob