1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.390 2006/09/12 10:35:56 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.
43 SC/01 Use a glob alias rather than an array ref in eximstats generated
44 parser. This improves both readability and performance.
46 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
47 Don't display local sender or destination tables in eximstats unless
48 there is data to show.
49 Added average volumes into the eximstats top table text output.
51 SC/03 Collect data on the number of addresses (recipients) as well
52 as the number of messages in eximstats.
54 TF/01 Correct an error in the documentation for the redirect router. Exim
55 does (usually) call initgroups() when daemonizing.
57 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
58 with consistent privilege compared to when running as a daemon.
60 TF/03 Note in the spec that $authenticated_id is not set for local
61 submissions from trusted users.
63 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
64 Thanks to Dean Brooks <dean@iglou.com> for the patch.
66 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
67 by adding some example configuration directives to the default
68 configuration file. A little bit of work is required to uncomment the
69 directives and define how usernames and passwords are checked, but
70 there is now a framework to start from.
72 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
73 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
74 without this. I don't know how relevant this is to other LDAP libraries.
76 PH/02 Add the verb name to the "unknown ACL verb" error.
78 PH/03 Magnus Holmgren's patch for filter_prepend_home.
80 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
82 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
83 directory not expanded when it should be if an expanded home directory
84 was set for the address (which is overridden by the transport).
86 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
89 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
90 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
91 because it is too late at that time, and has no effect.
93 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
94 security issue with \' (bugzilla #107). I could not use the
95 PQescapeStringConn() function, because it needs a PGconn value as one of
98 PH/08 When testing addresses using -bt, indicate those final addresses that
99 are duplicates that would not cause an additional delivery. At least one
100 person was confused, thinking that -bt output corresponded to deliveries.
101 (Suppressing duplicates isn't a good idea as you lose the information
102 about possibly different redirections that led to the duplicates.)
104 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
105 systems where poll() doesn't work, in particular OS X.
107 PH/10 Added more information to debugging output for retry time not reached.
109 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
110 operations in malware.c.
112 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
115 PH/13 If write_rejectlog was set false when logging was sent to syslog with
116 syslog_duplication set false, log lines that would normally be written
117 both the the main log and to the reject log were not written to syslog at
120 PH/14 In the default configuration, change the use of "message" in ACL warn
121 statements to "add_header".
123 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
124 not followed by a command (e.g. "seen endif").
126 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
127 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
130 PH/17 Added extra conditions to the default value of delay_warning_condition
134 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
135 { match{$h_precedence:}{(?i)bulk|list|junk} } \
136 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
139 The Auto-Submitted: and various List- headers are standardised, whereas I
140 don't think Precedence: ever was.
142 PH/18 Refactored debugging code in route_finduser() to show more information,
143 in particular, the error code if getpwnam() issues one.
145 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
146 This is apparently needed in addition to the PH/07 change above to avoid
147 any possible encoding problems.
149 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
150 but not after initializing Perl.
152 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
153 output them only if debugging. By default they are written stderr,
154 apparently, which is not desirable.
156 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
159 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
162 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
164 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
165 authenticated or an ident call has been made. Suppress the default
166 values for $authenticated_id and $authenticated_sender (but permit -oMai
167 and -oMas) when testing with -bh.
169 PH/24 Re-jigged the order of the tests in the default configuration so that the
170 tests for valid domains and recipients precede the DNS black list and CSA
171 tests, on the grounds that those ones are more expensive.
173 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
174 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
175 This bug exists in every version of Exim that I still have, right back to
178 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
179 However, an attempt to turn on case-sensitivity in a regex key by
180 including (?-i) didn't work because the subject string was already
181 lowercased, and the effects were non-intuitive. It turns out that a
182 one-line patch can be used to allow (?-i) to work as expected.
188 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
189 other effects) broke the use of negated acl sub-conditions.
191 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
194 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
195 "Deny" causes Exim to reject the incoming connection with a 554 error.
196 Unfortunately, if there is a major crisis, such as a disk failure,
197 tcp-wrappers gives "deny", whereas what one would like would be some
198 kind of temporary error. A kludge has been added to help with this.
199 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
200 554 error is used if errno is still zero or contains ENOENT (which occurs
201 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
204 PH/03 Add -lutil to the default FreeBSD LIBS setting.
206 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
207 errors. Otherwise a message that provokes a temporary error (when other
208 messages do not) can cause a whole host to time out.
210 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
211 addresses were routed directly to files or pipes from a redirect router.
212 File deliveries just didn't batch; pipe deliveries might have suffered
215 PH/06 A failure to get a lock for a hints database would erroneously always say
216 "Failed to get write lock", even when it was really a read lock.
218 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
219 of 0600. This has been changed to use the value of the lockfile_mode
220 option (which defaults to 0600).
222 PH/08 Applied small patch from the Sieve maintainer.
224 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
225 folder from quota calculations, a direct delivery into this folder messed
226 up the contents of the maildirsize file. This was because the regex was
227 used only to exclude .Trash (or whatever) when the size of the mailbox
228 was calculated. There was no check that a delivery was happening into an
229 excluded directory. This bug has been fixed by ignoring all quota
230 processing for deliveries into excluded directories.
232 PH/10 Added the maildirfolder_create_regex option to appendfile.
238 PH/01 The code for finding all the local interface addresses on a FreeBSD
239 system running IPv6 was broken. This may well have applied to all BSD
240 systems, as well as to others that have similar system calls. The broken
241 code found IPv4 interfaces correctly, but gave incorrect values for the
242 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
243 that it would not match correctly against @[] and not recognize the IPv6
246 PH/02 The ipliteral router was not recognizing addresses of the form user@
247 [ipv6:....] because it didn't know about the "ipv6:" prefix.
249 PH/03 Added disable_ipv6.
251 PH/04 Changed $reply_address to use the raw form of the headers instead of the
252 decoded form, because it is most often used to construct To: headers
253 lines in autoreplies, and the decoded form may well be syntactically
254 invalid. However, $reply_address has leading white space removed, and all
255 newlines turned into spaces so that the autoreply transport does not
258 PH/05 If group was specified without a user on a router, and no group or user
259 was specified on a transport, the group from the router was ignored.
261 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
262 for visible compile-time settings that can be used to change these
263 numbers, for those that want even more. Backwards compatibility with old
264 spool files has been maintained. However, going back to a previous Exim
265 release will lost any variables that are in spool files.
267 PH/07 Two small changes when running in the test harness: increase delay when
268 passing a TCP/IP connection to a new process, in case the original
269 process has to generate a bounce, and remove special handling of
270 127.0.0.2 (sic), which is no longer necessary.
272 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
273 be the same on different OS.
275 PH/09 Moved a debug statement in filter processing to avoid a race problem when
278 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
279 whether --show-vars was specified or not
281 JJ/02 exipick: Added support for new ACL variable spool format introduced
284 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
285 syntactically invalid From: or Reply-to: line, and a filter used this to
286 generate an autoreply, and therefore failed to obtain an address for the
287 autoreply, Exim could try to deliver to a non-existent relative file
288 name, causing unrelated and misleading errors. What now happens is that
289 it logs this as a hard delivery error, but does not attempt to create a
292 PH/11 The exinext utility has a -C option for testing purposes, but although
293 the given file was scanned by exinext itself; it wasn't being passed on
294 when Exim was called.
296 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
297 an end-of-file indication when reading a command response.
299 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
300 compiled. In many other places in Exim, IPv6 addresses are always
301 recognized, so I have changed this. It also means that IPv4 domain
302 literals of the form [IPV4:n.n.n.n] are now always recognized.
304 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
305 used if the router is not running as root, for example, when verifying at
306 ACL time, or when using -bh. The debugging output from this situation was
307 non-existent - all you got was a failure to exec. I have made two
310 (a) Failures to set uid/gid, the current directory, or a process leader
311 in a subprocess such as that created by queryprogram now generate
312 suitable debugging ouput when -d is set.
314 (b) The queryprogram router detects when it is not running as root,
315 outputs suitable debugging information if -d is set, and then runs
316 the subprocess without attempting to change uid/gid.
318 PH/15 Minor change to Makefile for building test_host (undocumented testing
321 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
322 additional section of a DNS packet that returns MX or SRV records.
323 Instead, it always explicitly searches for A/AAAA records. This avoids
324 major problems that occur when a DNS server includes only records of one
325 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
326 fixed another bug: if SRV records were looked up and the corresponding
327 address records were *not* found in the additional section, the port
328 values from the SRV records were lost.
330 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
331 using the correct key (the original address) when searching the retry
332 rules in order to find which one to use for generating the retry hint.
334 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
335 adding the default one. Similarly, if it contains a Reply-To: header, the
336 errors_reply_to option, if set, is not used.
338 PH/19 When calculating a retry time, Exim used to measure the "time since
339 failure" by looking at the "first failed" field in the retry record. Now
340 it does not use this if it is later than than the arrival time of the
341 message. Instead it uses the arrival time. This makes for better
342 behaviour in cases where some deliveries succeed, thus re-setting the
343 "first failed" field. An example is a quota failure for a huge message
344 when small messages continue to be delivered. Without this change, the
345 "time since failure" will always be short, possible causing more frequent
346 delivery attempts for the huge message than are intended.
347 [Note: This change was subsequently modified - see PH/04 for 4.62.]
349 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
350 $1, $2, $3) because the numerical variables can be reset during some
351 expansion items (e.g. "match"), thereby losing the authentication data.
353 PH/21 Make -bV show the size of off_t variables so that the test suite can
354 decide whether to run tests for quotas > 2G.
356 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
357 mailbox_size, and mailbox_filecount in the appendfile transport. If a
358 filecount value is greater than 2G or if a quota value is greater than 2G
359 on a system where the size of off_t is not greater than 4, a panic error
362 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
363 never match. The debug and -bh output now contains an explicit error
364 message indicating a malformed IPv4 address or mask.
366 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
367 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
370 PH/25 Do not write to syslog when running in the test harness. The only
371 occasion when this arises is a failure to open the main or panic logs
372 (for which there is an explicit test).
374 PH/26 Added the /no_tell option to "control=freeze".
376 PH/27 If a host name lookup failed very early in a connection, for example, if
377 the IP address matched host_lookup and the reverse lookup yielded a name
378 that did not have a forward lookup, an error message of the form "no IP
379 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
380 could be logged. Now it outputs the IP address instead of "NULL".
382 PH/28 An enabling patch from MH: add new function child_open_exim2() which
383 allows the sender and the authenticated sender to be set when
384 submitting a message from within Exim. Since child_open_exim() is
385 documented for local_scan(), the new function should be too.
387 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
388 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
389 results in an empty string is now treated as unset.
391 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
393 PH/31 Added sender_verify_fail logging option.
395 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
396 needed by RFC 822 but not by RFC 2822 was commented out. I have now
397 tidied the source and removed it altogether.
399 PH/33 When a queue run was abandoned because the load average was too high, a
400 log line was always written; now it is written only if the queue_run log
401 selector is set. In addition, the log line for abandonment now contains
402 information about the queue run such as the pid. This is always present
403 in "start" and "stop" lines but was omitted from the "abandon" line.
405 PH/34 Omit spaces between a header name and the colon in the error message that
406 is given when verify = headers_syntax fails (if there are lots of them,
407 the message gets confusing).
409 PH/35 Change the default for dns_check_names_pattern to allow slashes within
410 names, as there are now some PTR records that contain slashes. This check
411 is only to protect against broken name servers that fall over on strange
412 characters, so the fact that it applies to all lookups doesn't matter.
414 PH/36 Now that the new test suite is complete, we can remove some of the
415 special code in Exim that was needed for the old test suite. For example,
416 sorting DNS records because real resolvers return them in an arbitrary
417 order. The new test suite's fake resolver always returns records in the
420 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
421 bounces) except when queue_only is set, to avoid logging races between
422 the different processes.
424 PH/38 Panic-die if .include specifies a non-absolute path.
426 PH/39 A tweak to the "H" retry rule from its user.
428 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
429 a label. They prevented compilation on older perls.
431 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
432 a warning to be raised on newish perls.
434 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
435 on queue. Changes to match documented behaviour of showing count of
436 messages matching specified criteria.
438 PH/40 Changed the default ident timeout from 30s to 5s.
440 PH/41 Added support for the use of login_cap features, on those BSD systems
441 that have them, for controlling the resources used by pipe deliveries.
443 PH/42 The content-scanning code uses fopen() to create files in which to put
444 message data. Previously it was not paying any attention to the mode of
445 the files. Exim runs with umask(0) because the rest of the code creates
446 files with open(), and sets the required mode explicitly. Thus, these
447 files were ending up world-writeable. This was not a big issue, because,
448 being within the spool directory, they were not world-accessible. I have
449 created a function called modefopen, which takes an additional mode
450 argument. It sets umask(777), creates the file, chmods it to the required
451 mode, then resets the umask. All the relevant calls to fopen() in the
452 content scanning code have been changed to use this function.
454 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
455 to 24 hours. This avoids potential overflow problems when processing G
456 and H retry rules. I suspect nobody ever tinkers with this value.
458 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
460 PH/45 When the plaintext authenticator is running as a client, the server's
461 challenges are checked to ensure they are valid base64 strings. By
462 default, the authentication attempt is cancelled if an invalid string is
463 received. Setting client_ignore_invalid_base64 true ignores these errors.
464 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
465 they are received. Thus, the responses can be made to depend on the
466 challenges. If an invalid string is ignored, an empty string is placed in
469 PH/46 Messages that are created by the autoreply transport now contains a
470 References: header, in accordance with RFCs 2822 and 3834.
472 PH/47 Added authenticated_sender_force to the smtp transport.
474 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
476 PH/49 Installed latest patch from the Sieve maintainer.
478 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
479 was also set, the appendfile transport was unnecessarily scanning a
480 directory of message files (e.g. for maildir delivery) to find the count
481 of files (along with the size), even though it did not need this
482 information. It now does the scan only if it needs to find either the
483 size of the count of files.
485 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
487 PH/52 Two bugs concerned with error handling when the smtp transport is
490 (i) Exim was not creating retry information for temporary errors given
491 for individual recipients after the DATA command when the smtp transport
492 was used in LMTP mode. This meant that they could be retried too
493 frequently, and not timed out correctly.
495 (ii) Exim was setting the flag that allows error details to be returned
496 for LMTP errors on RCPT commands, but not for LMTP errors for individual
497 recipients that were returned after the DATA command.
499 PH/53 This is related to PH/52, but is more general: for any failing address,
500 when detailed error information was permitted to be returned to the
501 sender, but the error was temporary, then after the final timeout, only
502 "retry timeout exceeded" was returned. Now it returns the full error as
503 well as "retry timeout exceeded".
505 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
506 do this, and (what is worse) MTAs that accept it.
508 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
509 will now be deprecated.
511 PH/56 New os.c-cygwin from the Cygwin maintainer.
513 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
514 formats (previously only available in exim formats via -bpr, -bpru,
515 and -bpra. Now also available in native and exiqgrep formats)
517 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
518 with very large, slow to parse queues
520 JJ/08 exipick: added ! as generic prefix to negate any criteria format
522 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
524 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
525 responses to authentication challenges, though it was showing the
526 challenges; (ii) I've removed the CR characters from the debug output for
529 PH/58 Allow for the insertion of a newline as well as a space when a string
530 is turned into more than one encoded-word during RFC 2047 encoding. The
531 Sieve code now uses this.
533 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
534 data_4xx, lost_connection, tls_required.
536 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
537 message was being sent as an SMTP response.
539 PH/61 Add -l and -k options to exicyclog.
541 PH/62 When verifying, if an address was redirected to one new address, so that
542 verification continued, and the new address failed or deferred after
543 having set something in $address_data, the value of $address_data was not
544 passed back to the ACL. This was different to the case when no
545 redirection occurred. The value is now passed back in both cases.
547 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
548 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
549 use login_cap.h, so on its own it isn't the distinguishing feature. The
550 new name refers directly to the setclassresources() function.
552 PH/65 Added configuration files for NetBSD3.
554 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
556 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
557 is preferred over IPv4.
559 PH/68 The bounce_return_message and bounce_return_body options were not being
560 honoured for bounces generated during the reception of non-SMTP messages.
561 In particular, this applied to messages rejected by the ACL. This bug has
562 been fixed. However, if bounce_return_message is true and bounce_return_
563 body is false, the headers that are returned for a non-SMTP message
564 include only those that have been read before the error was detected.
565 (In the case of an ACL rejection, they have all been read.)
567 PH/69 The HTML version of the specification is now built in a directory called
568 spec_html instead of spec.html, because the latter looks like a path with
569 a MIME-type, and this confuses some software.
571 PH/70 Catch two compiler warnings in sieve.c.
573 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
574 function verify_get_ident() calls ip_connect() to connect a socket, but
575 if the "connect()" function timed out, ip_connect() used to close the
576 socket. However, verify_get_ident() also closes the socket later, and in
577 between Exim writes to the log, which may get opened at this point. When
578 the socket was closed in ip_connect(), the log could get the same file
579 descriptor number as the socket. This naturally causes chaos. The fix is
580 not to close the socket in ip_connect(); the socket should be closed by
581 the function that creates it. There was only one place in the code where
582 this was missing, in the iplookup router, which I don't think anybody now
583 uses, but I've fixed it anyway.
585 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
586 well as to direct DNS lookups. Otherwise the handling of names in host
587 lists is inconsistent and therefore confusing.
593 PH/01 Two changes to the default runtime configuration:
595 (1) Move the checks for relay_from_hosts and authenticated clients from
596 after to before the (commented out) DNS black list checks.
598 (2) Add control=submission to the relay_from_hosts and authenticated
599 clients checks, on the grounds that messages accepted by these
600 statements are most likely to be submissions.
602 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
604 (1) Generate an error if the third argument for the ${prvs expansion is
607 (2) Treat a missing third argument of ${prvscheck as if it were an empty
610 (3) Reset the variables that are obtained from the first argument of
611 ${prvscheck and used in the second argument before leaving the code,
612 because their memory is reclaimed, so using them afterwards may do
615 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
616 one (it's much easier than Tom thought :-).
618 (5) Because of (4), we can now allow for the use of $prvscheck_result
619 inside the third argument.
621 PH/03 For some reason, the default setting of PATH when running a command from
622 a pipe transport was just "/usr/bin". I have changed it to
625 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
626 anything to be listed in the output from -bV.
628 PH/05 When a filter generated an autoreply, the entire To: header line was
629 quoted in the delivery log line, like this:
631 => >A.N.Other <ano@some.domain> <original@ddress> ...
633 This has been changed so that it extracts the operative address. There
634 may be more than one such address. If so, they are comma-separated, like
637 => >ano@some.domain,ona@other.domain <original@ddress> ...
639 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
640 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
641 not being looked up in the rDNS to get a host name, Exim was showing the
642 IP address twice in Received: lines, even though the IP addresses were
643 identical. For example:
645 Received: from [1.2.3.4] (helo=[1.2.3.4])
647 However, if the real host name was known, it was omitting the HELO data
648 if it matched the actual IP address. This has been tidied up so that it
649 doesn't show the same IP address twice.
651 PH/07 When both +timestamp and +memory debugging was on, the value given by
652 $tod_xxx expansions could be wrong, because the tod_stamp() function was
653 called by the debug printing, thereby overwriting the timestamp buffer.
654 Debugging no longer uses the tod_stamp() function when +timestamp is set.
656 PH/08 When the original message was included in an autoreply transport, it
657 always said "this is a copy of the message, including all the headers",
658 even if body_only or headers_only was set. It now gives an appropriate
661 PH/09 Applied a patch from the Sieve maintainer which:
663 o fixes some comments
664 o adds the (disabled) notify extension core
665 o adds some debug output for the result of if/elsif tests
666 o points to the current vacation draft in the documentation
667 and documents the missing references header update
671 o fixes a bug in processing the envelope test (when testing
672 multiple envelope elements, the last element determinted the
675 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
676 Electronic Mail") by including:
678 Auto-submitted: auto-generated
680 in the messages that it generates (bounce messages and others, such as
681 warnings). In the case of bounce messages for non-SMTP mesages, there was
682 also a typo: it was using "Auto_submitted" (underscore instead of
683 hyphen). Since every message generated by Exim is necessarily in response
684 to another message, thes have all been changed to:
686 Auto-Submitted: auto-replied
688 in accordance with these statements in the RFC:
690 The auto-replied keyword:
692 - SHOULD be used on messages sent in direct response to another
693 message by an automatic process,
695 - MUST NOT be used on manually-generated messages,
697 - MAY be used on Delivery Status Notifications (DSNs) and Message
698 Disposition Notifications (MDNs),
700 - MUST NOT be used on messages generated by automatic or periodic
701 processes, except for messages which are automatic responses to
704 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
705 to the default Received: header definition.
707 PH/12 Added log selector acl_warn_skipped (default on).
709 PH/13 After a successful wildlsearch lookup, discard the values of numeric
710 variables because (a) they are in the wrong storage pool and (b) even if
711 they were copied, it wouldn't work properly because of the caching.
713 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
714 checking when decoding. Apparently there are clients that generate
715 overlong encoded strings. Why am I not surprised?
717 PH/15 If the first argument of "${if match_address" was not empty, but did not
718 contain an "@" character, Exim crashed. Now it writes a panic log message
719 and treats the condition as false.
721 PH/16 In autoreply, treat an empty string for "once" the same as unset.
723 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
724 extension "envelope-auth". The code is finished and in agreement with
725 other implementations, but there is no documentation so far and in fact,
726 nobody wrote the draft yet. This extension is currently #undef'ed, thus
727 not changing the active code.
729 Print executed "if" and "elsif" statements when debugging is used. This
730 helps a great deal to understand what a filter does.
732 Document more things not specified clearly in RFC3028. I had all this
733 sorted out, when out of a sudden new issues came to my mind. Oops."
735 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
738 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
739 canonical form (as documented). However, after a host name lookup from
740 the IP address, check_host() was doing a simple string comparison with
741 addresses acquired from the DNS when checking that the found name did
742 have the original IP as one of its addresses. Since any found IPv6
743 addresses are likely to be in abbreviated form, the comparison could
744 fail. Luckily, there already exists a function for doing the comparison
745 by converting both addresses to binary, so now that is used instead of
748 PH/20 There was another similar case to PH/19, when a complete host name was
749 given in a host list; looking up its IP address could give an abbreviated
750 form, whereas the current host's name might or might not be abbreviated.
751 The same fix has been applied.
757 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
758 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
761 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
762 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
764 PH/03 Typo: missing ".o" in src/pcre/Makefile.
766 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
767 header line, restrict the check to what is listed in RFCs 2369 and 2929.
768 Also, for "Auto-Submitted", treat anything other than "no" as
769 non-personal, in accordance with RFC 3834. (Previously it treated
770 anything starting "auto-" as non-personal.)
772 TF/01 The control=submission/name=... option had a problem with syntax
773 errors if the name included a slash character. The /name= option
774 now slurps the rest of the string, so it can include any characters
775 but it must come last in the list of options (after /sender_retain
778 PH/05 Some modifications to the interface to the fake nameserver for the new
786 TK/01 Added the "success_on_redirect" address verification option. See
787 NewStuff for rationale and an example.
789 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
791 PH/02 Patch to exigrep to allow it to work on syslog lines.
793 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
794 fread() to skip over the body file's header line, because in Cygwin the
795 header line is locked and is inaccessible.
797 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
798 co-exist for some time) to make it clear that it is the Exim ID that is
799 referenced, not the Message-ID: header line.
801 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
802 string_format() function, because snprintf() does not exist on all
805 PH/06 The use of forbid_filter_existstest now also locks out the use of the
806 ${stat: expansion item.
808 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
809 protocol synchronization error", to keep the pedants happy.
811 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
812 well as for IRIX systems, when gcc is being used. See the host.c source
815 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
817 PH/10 Named domain lists were not working if used in a queue_smtp_domains
820 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
821 transport and to the smtp transport in LMTP mode.
823 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
825 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
826 run a filter in a subprocess. This could lead to confusion in subsequent
827 lookups in the parent process. There should also be a search_tidyup() at
828 the end of the subprocess.
830 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
831 only if the host matched helo_try_verify_hosts, which caused the
832 verification to occur when the EHLO/HELO command was issued. The ACL just
833 tested the remembered result. Now, if a previous verification attempt has
834 not happened, "verify = helo" does it there and then.
836 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
838 TK/03 Fix log output including CR from clamd.
840 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
841 exist provoked a memory error which could cause a segfault.
843 PH/15 Installed PCRE 6.2
845 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
847 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
848 of the problem. Specifically, suggested +O2 rather than +O1 for the
851 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
853 PH/20 If a delivery was routed to a non-standard port by means of an SRV
854 record, the port was not correctly logged when the outgoing_port log
855 selector was set (it logged the transort's default port).
857 PH/21 Added support for host-specific ports to manualroute, queryprogram,
858 fallback_hosts, and "hosts" in the smtp transport.
860 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
861 host errors such as "Connection refused".
863 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
864 authentication with radiusclient 0.4.9:
866 - Error returned from rc_read_config was caught wrongly
867 - Username/password not passed on to radius server due to wrong length.
869 The presumption is that some radiusclient API changes for 4.51/PH/17
870 were not taken care of correctly. The code is still untested by me (my
871 Linux distribution still has 0.3.2 of radiusclient), but it was
872 contributed by a Radius user.
874 PH/24 When doing a callout, the value of $domain wasn't set correctly when
875 expanding the "port" option of the smtp transport.
877 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
878 while reading a MIME header. Thanks to Tom Hughes for a patch.
880 PH/24 Include config.h inside local_scan.h so that configuration settings are
883 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
884 that in an ACL for RCPT (for example), you can examine exactly what was
887 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
888 commands, but it was not correctly comparing the address with the actual
889 client host address. Thus, it would show the EHLO address in Received:
890 header lines when this was not necessary.
892 PH/27 Added the % operator to ${eval:}.
894 PH/28 Exim tries to create and chdir to its spool directory when it starts;
895 it should be ignoring failures (because with -C, for example, it has lost
896 privilege). It wasn't ignoring creation failures other than "already
899 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
902 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
903 because an input line was too long, either on its own, or by virtue of
904 too many continuations, the temporary file was not being removed, and the
905 return code was incorrect.
907 PH/31 Missing "BOOL" in function definition in filtertest.c.
909 PH/32 Applied Sieve patches from the maintainer.
911 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
913 PH/33 Added "verify = not_blind".
915 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
916 Local/Makefile (with some defaults set). These are used in built scripts
917 such as exicyclog, but they have never been used in the exim_install
918 script (though there are many overriding facilities there). I have
919 arranged that the exim_install script now takes note of these two
922 PH/35 Installed configuration files for Dragonfly.
924 PH/36 When a locally submitted message by a trusted user did not contain a
925 From: header, and the sender address was obtained from -f or from an SMTP
926 MAIL command, and the trusted user did not use -F to supply a sender
927 name, $originator_name was incorrectly used when constructing a From:
928 header. Furthermore, $originator_name was used for submission mode
929 messages from external hosts without From: headers in a similar way,
930 which is clearly wrong.
932 PH/37 Added control=suppress_local_fixups.
934 PH/38 When log_selector = +received_sender was set, and the addition of the
935 sender made the log line's construction buffer exactly full, or one byte
936 less than full, an overflow happened when the terminating "\n" was
939 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
940 when the result of a list match is failure because a DNS lookup failed.
942 PH/40 RM_COMMAND is now used in the building process.
944 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
945 the "build-* directories that it finds.
947 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
948 address in a domain literal was a prefix of an interface address.
950 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
951 when verifying a sender address, unless rewrite_headers is false.
953 PH/44 Wrote a long comment about why errors_to addresses are verified as
954 recipients, not senders.
956 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
957 the ratelimit ACL was added.
959 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
961 PH/46 Added extra information about PostgreSQL errors to the error string.
963 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
964 avoiding the need to install special zones in a real server. This is
965 backwards compatible; if it can't find the fake resolver, it drops back.
966 Thus, both old and new test suites can be run.
968 TF/02 Added util/ratelimit.pl
970 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
971 clock is set back in time.
973 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
974 Candler <B.Candler@pobox.com>.
976 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
977 OK for addresses that are the result of redirections.
979 PH/48 A number of further additions for the benefit of the new test suite,
980 including a fake gethostbyname() that interfaces to the fake DNS resolver
983 TF/06 The fix for widen_domains has also been applied to qualify_single and
984 search_parents which are the other dnslookup options that can cause
987 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
990 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
992 TF/07 Exim produced the error message "an SRV record indicated no SMTP
993 service" if it encountered an MX record with an empty target hostname.
994 The message is now "an MX or SRV record indicated no SMTP service".
996 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
997 if the DNS of the sending site is misconfigured. This is quite a
998 common situation. This change restores the behaviour of treating a
999 helo verification defer as a failure.
1001 PH/51 If self=fail was set on a router, the bounce message did not include the
1002 actual error message.
1008 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1010 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1011 command itself ended in error, the underlying message about the transport
1012 filter timeout was being overwritten with the pipe command error. Now the
1013 underlying error message should be appended to the second error message.
1015 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1017 PH/02 Reduce the amount of output that "make" produces by default. Full output
1018 can still be requested.
1020 PH/03 The warning log line about a condition test deferring for a "warn" verb
1021 was being output only once per connection, rather than after each
1022 occurrence (because it was using the same function as for successful
1023 "warn" verbs). This seems wrong, so I have changed it.
1025 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1026 it should not have, which might have caused a crash in the right
1027 circumstances, but probably never did.
1029 PH/04 Installed a modified version of Tony Finch's patch to make submission
1030 mode fix the return path as well as the Sender: header line, and to
1031 add a /name= option so that you can make the user's friendly name appear
1034 TF/03 Added the control = fakedefer ACL modifier.
1036 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1037 Mark Lowes for thorough testing.
1039 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1041 TK/03 Merged latest SRS patch from Miles Wilton.
1043 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1044 with the definition in sysexits.h (which is #included earlier).
1045 Fortunately, Exim does not actually use EX_OK. The code used to try to
1046 preserve the sysexits.h value, by assumimg that macro definitions were
1047 scanned for macro replacements. I have been disabused of this notion,
1048 so now the code just undefines EX_OK before #including unistd.h.
1050 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1051 in the smtp transport. When a block could not be written in a single
1052 write() function, the timeout was being re-applied to each part-write.
1053 This seems wrong - if the receiver was accepting one byte at a time it
1054 would take for ever. The timeout is now adjusted when this happens. It
1055 doesn't have to be particularly precise.
1057 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1058 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1060 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1061 without a domain if the check to <postmaster@domain> fails.
1063 SC/01 Eximstats: added -xls and the ability to specify output files
1064 (patch written by Frank Heydlauf).
1066 SC/02 Eximstats: use FileHandles for outputing results.
1068 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1070 SC/04 Eximstats: fixed display of large numbers with -nvr option
1072 SC/05 Eximstats: fixed merging of reports with empty tables.
1074 SC/06 Eximstats: added the -include_original_destination flag
1076 SC/07 Eximstats: removed tabs and trailing whitespace.
1078 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1080 TK/06 MBOX spool code: Add real "From " MBOX separator line
1081 so the .eml file is really in mbox format (even though
1082 most programs do not really care). Patch from Alex Miller.
1084 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1085 The latter is generated from $received_to and is only set if the
1086 message has one envelope recipient. SA can use these headers,
1087 obviously out-of-the-box. Patch from Alex Miller.
1089 PH/08 The ${def test on a variable was returning false if the variable's
1090 value was "0", contrary to what the specification has always said!
1091 The result should be true unless the variable is empty.
1093 PH/09 The syntax error of a character other than { following "${if
1094 def:variable_name" (after optional whitespace) was not being diagnosed.
1095 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1096 accidental colon was present, for example, could give incorrect results.
1098 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1099 result is used (not including appendfile, where other changes are about
1102 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1103 This involved changing a lot of size variables from int to off_t. It
1104 should work with maildirs and everything.
1106 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1107 spamd dying while we are connected to it.
1109 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1112 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1113 function for BDB 4.3.
1115 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1117 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1118 the sources for PCRE have been re-arranged and more files are now
1121 PH/15 The code I had for printing potentially long long variables in PH/11
1122 above was not the best (it lost precision). The length of off_t variables
1123 is now inspected at build time, and an appropriate printing format (%ld
1124 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1125 to be "long long int" or "long int". This is needed for the internal
1126 formatting function string_vformat().
1128 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1129 the configuration file to be ":syslog", then the script "guesses" where
1130 the logs files are, rather than using the compiled in default. In our
1131 case the guess is not the same as the compiled default, so the script
1132 suddenly stopped working when I started to use syslog. The patch checks
1133 to see if log_file_path is "". If so, it attempts to read it from exim
1134 with no configuration file to get the compiled in version, before it
1135 falls back to the previous guessing code."
1137 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1138 implementing BATV in an Exim configuration. See NewStuff for the gory
1141 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1142 Makefile that are specific to HP-UX.
1144 PH/18 If the "use_postmaster" option was set for a recipient callout together
1145 with the "random" option, the postmaster address was used as the MAIL
1146 FROM address for the random test, but not for the subsequent recipient
1147 test. It is now used for both.
1149 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1150 patch removes a few documentation additions to RFC 3028, because the
1151 latest draft now contains them. It adds the new en;ascii-case comparator
1152 and a new error check for 8bit text in MIME parts. Comparator and
1153 require names are now matched exactly. I enabled the subaddress
1154 extension, but it is not well tested yet (read: it works for me)."
1156 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1157 rework some of the code of TK/09 above to avoid the hardwired use of
1158 "%lld" and "long long". Replaced the call to snprintf() with a call to
1161 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1162 records point to non-existent hosts", "retry timeout exceeded", and
1163 "retry time not reached for any host after a long failure period".
1165 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1166 experimental DomainKeys support:
1168 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1169 (2) On an error such as an illegally used "control", the wrong name for
1170 the control was given.
1172 These problems did NOT occur unless DomainKeys support was compiled.
1174 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1176 PH/24 Added ${if match_ip condition.
1178 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1179 databases so that it will be absolutely obvious if a crash occurs in the
1180 DB library. This is a regular occurrence (often caused by mis-matched
1183 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1184 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1185 picked up on a user's system that detects such things. There doesn't seem
1186 to be a gcc warning option for this - only an attribute that has to be
1187 put on the function's prototype. It seems that in Fedora Core 4 they have
1188 set this on a number of new functions. No doubt there will be more in due
1191 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1192 specify a transport. However, if an address that was verified by such a
1193 router was the subject of a callout, Exim crashed because it tried to
1194 read the rcpt_include_affixes from the non-existent transport. Now it
1195 just assumes that the setting of that option is false. This bug was
1196 introduced by 4.51/PH/31.
1198 PH/28 Changed -d+all to exclude +memory, because that information is very
1199 rarely of interest, but it makes the output a lot bigger. People tend to
1200 do -d+all out of habit.
1202 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1203 code in os-type was giving problems when libc.so lives in lib64, like on
1206 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1207 aren't the modern standard, and it seems that some systems' include files
1208 don't always have them. Exim was already checking for some of the newer
1209 ones like T_AAAA, and defining it itself. I've added checks for all the
1210 record types that Exim uses.
1212 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1213 not automatically generating a new one, as it is supposed to. This
1214 prevented TLS from working. If the file did exist, but contained invalid
1215 data, a new version was generated, as expected. It was only the case of a
1216 non-existent file that was broken.
1218 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1219 with a change in libdomainkeys > 0.64.
1221 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1222 from DNS. If the selector record carries the flag, it now has
1223 precedence over the domain-wide flag.
1225 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1227 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1228 the use of a port name that isn't defined in /etc/services) occurred, the
1229 message was deferred as in a normal delivery, and thus remained on the
1230 spool, instead of being failed because of the mua_wrapper setting. This
1231 is now fixed, and I tidied up some of the mua_wrapper messages at the
1234 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1235 the messages in a hash of arrays rather than using individual hashes.
1236 This is a bit cleaner and results in dramatic memory savings, albeit
1237 at a slight CPU cost.
1239 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1240 as requested by Marc Sherman.
1242 SC/10 Eximstats: added histograms for user specified patterns as requested
1245 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1247 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1248 fopen() in the content-scanning modules that did not already have it.
1254 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1255 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1257 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1259 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1261 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1262 to test Sieve filters that use "vacation".
1264 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1265 that changes the way the GnuTLS parameters are stored in the cache file.
1266 The new format can be generated externally. For backward compatibility,
1267 if the data in the cache doesn't make sense, Exim assumes it has read an
1268 old-format file, and it generates new data and writes a new file. This
1269 means that you can't go back to an older release without removing the
1272 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1273 work if there are any delivery delays because "one_time" forces the
1274 parent to be marked "delivered", so its unseen clone is never tried
1275 again. For this reason, Exim now forbids the simultaneous setting of
1278 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1279 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1280 entry if you want to know the details. The fix, however, neglected to
1281 consider the case when local delivery batching is involved. The test for
1282 "previously delivered" was not happening when checking to see if an
1283 address could be batched with a previous (undelivered) one; under
1284 certain circumstances this could lead to multiple deliveries to the same
1287 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1288 in its include files, and this causes problems building Exim.
1290 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1291 header_syntax) but Exim was just ignoring anything given after a slash.
1292 In particular, this caused confusion with an attempt to use "verify =
1293 reverse_host_lookup/defer_ok". An error is now given when options are
1294 supplied for verify items that do not have them. (Maybe reverse_host_
1295 lookup should have a defer_ok option, but that's a different point.)
1297 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1298 defined by RFC 821) to 2048, because there were problems with some AUTH
1299 commands, and RFC 1869 says the size should be increased for extended
1300 SMTP commands that take arguments.
1302 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1305 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1306 "unknown" error; now it says that the functionality isn't in the binary.
1308 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1309 an address' error message when a string expansion fails (syntax or
1310 whatever). Otherwise the password may appear in the log. Following change
1311 PH/42 below, there is no longer a chance of it appearing in a bounce
1314 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1316 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1317 of its addresses was ever tried. (Bugzilla bug #2.)
1319 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1320 the result incorrectly in the debug output. (It correctly added a newline
1321 to what was transported.)
1323 TF/01 Added $received_time.
1325 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1326 commented out examples of how to interface to a virus scanner and to
1327 SpamAssassin. Also added commented examples of av_scanner and
1328 spamd_address settings.
1330 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1331 and controls are allowed in which ACLs. There were a couple of minor
1332 errors. Some of the entries in the conditions table (which is a table of
1333 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1334 negation of where the condition IS allowed.
1336 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1338 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1339 header file does not have a version number, so I've had to invent a new
1340 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1341 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1342 radiusclient), but it was contributed by a Radius user.
1344 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1345 files or directories, for OpenSSL.
1347 PH/19 When an Exim process that is running as root has to create an Exim log
1348 file, it does so in a subprocess that runs as exim:exim so as to get the
1349 ownership right at creation (otherwise, other Exim processes might see
1350 the file with the wrong ownership). There was no test for failure of this
1351 fork() call, which would lead to the process getting stuck as it waited
1352 for a non-existent subprocess. Forks do occasionally fail when resources
1353 run out. I reviewed all the other calls to fork(); they all seem to check
1356 PH/20 When checking for unexpected SMTP input at connect time (before writing
1357 the banner), Exim was not dealing correctly with a non-positive return
1358 from the read() function. If the client had disconnected by this time,
1359 the result was a log entry for a synchronization error with an empty
1360 string after "input=" when read() returned zero. If read() returned -1
1361 (an event I could not check), uninitialized data bytes were printed.
1362 There were reports of junk text (parts of files, etc) appearing after
1365 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1367 PH/22 Added support for macro redefinition, and (re)definition in between
1368 driver and ACL definitions.
1370 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1371 forgetting to use the resulting value; it was using the unexpanded value.
1373 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1374 hadn't been configured. The fix is from Juergen Kreileder, who
1375 understands it better than I do:
1377 "Here's what I see happening with three configured cyrus_sasl
1378 authenticators configured (plain, login, cram-md5):
1380 On startup auth_cyrus_sasl_init() gets called for each of these.
1381 This means three calls to sasl_listmech() without a specified mech_list.
1382 => SASL tests which mechs of all available mechs actually work
1383 => three warnings about OTP not working
1384 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1386 With the patch, sasl_listmech() also gets called three times. But now
1387 SASL's mech_list option is set to the server_mech specified in the the
1388 authenticator. Or in other words, the answer from sasl_listmech()
1389 gets limited to just the mech you're testing for (which is different
1391 => the return list contains just 'plain' or 'login', 'cram-md5' or
1392 nothing depending on the value of ob->server_mech.
1394 I've just tested the patch: Authentication still works fine,
1395 unavailable mechs specified in the exim configuration are still
1396 caught, and the auth.log warnings about OTP are gone."
1398 PH/25 When debugging is enabled, the contents of the command line are added
1399 to the debugging output, even when log_selector=+arguments is not
1402 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1403 answer is "GNU", and only if the return is "GNU/something" is the answer
1406 PH/27 $acl_verify_message is now set immediately after the failure of a
1407 verification in an ACL, and so is available in subsequent modifiers. In
1408 particular, the message can be preserved by coding like this:
1410 warn !verify = sender
1411 set acl_m0 = $acl_verify_message
1413 Previously, $acl_verify_message was set only while expanding "message"
1414 and "log_message" when a very denied access.
1416 PH/28 Modified OS/os.c-Linux with
1418 -#ifndef OS_LOAD_AVERAGE
1419 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1421 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1422 nomenclature these days.)
1424 PH/29 Installed patch from the Sieve maintainer that adds the options
1425 sieve_useraddress and sieve_subaddress to the redirect router.
1427 PH/30 In these circumstances:
1428 . Two addresses routed to the same list of hosts;
1429 . First host does not offer TLS;
1430 . First host accepts first address;
1431 . First host gives temporary error to second address;
1432 . Second host offers TLS and a TLS session is established;
1433 . Second host accepts second address.
1434 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1435 and peerdn, if requested) that were in fact used only for the second
1438 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1439 attention to any local part prefix or suffix that was matched by the
1440 router that accepted the address. It now behaves in the same way as it
1441 does for delivery: the affixes are removed from the local part unless
1442 rcpt_include_affixes is set on the transport.
1444 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1445 timeout during the DATA phase of an incoming message.
1447 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1448 applied a patch sanctioned by the Sieve maintainer.
1450 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1451 the uid or gid is negative. A case of a negative gid caused this to be
1452 noticed. The fix allows for either to be negative.
1454 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1455 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1458 PH/36 The change PH/12 above was broken. Fixed it.
1460 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1461 the grounds that routing the same address twice would always produce the
1462 same answer. This might have been true once, but it is certainly no
1463 longer true now. Routing a child address may depend on the previous
1464 routing that produced that child. Some complicated redirection strategies
1465 went wrong when messages had multiple recipients, and made Exim's
1466 behaviour dependent on the order in which the addresses were given.
1468 I have moved the duplicate checking until after the routing is complete.
1469 Exim scans the addresses that are assigned to local and remote
1470 transports, and removes any duplicates. This means that more work will be
1471 done, as duplicates will always all be routed, but duplicates are
1472 presumably rare, so I don't expect this is of any significance.
1474 For deliveries to pipes, files, and autoreplies, the duplicate checking
1475 still happens during the routing process, since they are not going to be
1478 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1479 It corrects a timeout issue with spamd. This is Ian's comment: "The
1480 background is that sometimes spamd either never reads data from a
1481 connection it has accepted, or it never writes response data. The exiscan
1482 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1483 blindly assumes that writes won't block so it may never time out."
1485 PH/39 Allow G after quota size as well as K and M.
1487 PH/40 The value set for $authenticated_id in an authenticator may not contain
1488 binary zeroes or newlines because the value is written to log lines and
1489 to spool files. There was no check on this. Now the value is run through
1490 the string_printing() function so that such characters are converted to
1491 printable escape sequences.
1493 PH/41 $message_linecount is a new variable that contains the total number of
1494 lines in the message. Compare $body_linecount, which is the count for the
1497 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1498 bounce and delay warning messages, except in certain special cases, which
1501 (a) An SMTP error message from a remote host;
1502 (b) A message specified in a :fail: redirection;
1503 (c) A message specified in a "fail" command in a system filter;
1504 (d) A message specified in a FAIL return from the queryprogram router;
1505 (e) A message specified by the cannot_route_message router option.
1507 In these cases only, Exim does include the error details in bounce and
1508 warning messages. There are also a few cases where bland messages such
1509 as "unrouteable address" or "local delivery error" are given.
1511 PH/43 $value is now also set for the "else" part of a ${run expansion.
1513 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1514 being worked on, but at least Exim now implements the latest version to
1517 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1518 process to complete was treated as a delivery failure, a timeout while
1519 writing the message to the pipe was logged, but erroneously treated as a
1520 successful delivery. Such timeouts include transport filter timeouts. For
1521 consistency with the overall process timeout, these timeouts are now
1522 treated as errors, giving rise to delivery failures by default. However,
1523 there is now a new Boolean option for the pipe transport called
1524 timeout_defer, which, if set TRUE, converts the failures into defers for
1525 both kinds of timeout. A transport filter timeout is now identified in
1528 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1529 systems where "make" and "gmake" are different, calling "gmake" at top
1530 level broke things. I've arranged for the value of $(MAKE) to be passed
1531 from the Makefile to this script so that it can call the same version of
1535 A note about Exim versions 4.44 and 4.50
1536 ----------------------------------------
1538 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1539 changes of various kinds. As a consequence, a big documentation update was
1540 needed. This delayed the release for rather longer than seemed good, especially
1541 in the light of a couple of (minor) security issues. Therefore, the changes
1542 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1543 release. So 4.44 and 4.50 are in effect two different branches that both start
1546 I have left the 4.50 change log unchanged; it contains all the changes since
1547 4.43. The change log for 4.44 is below; many of its items are identical to
1548 those for 4.50. This seems to be the most sensible way to preserve the
1549 historical information.
1555 1. Minor wording change to the doc/README.SIEVE file.
1557 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1558 computation of the current number of files was incorrect.
1560 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1561 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1564 4. Give more explanation in the error message when the command for a transport
1565 filter fails to execute.
1567 5. There are several places where Exim runs a non-Exim command in a
1568 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1569 was being done only for the command run by the queryprogram router. It is
1570 now done for all such subprocesses. The other cases are: ${run, transport
1571 filters, and the commands run by the lmtp and pipe transports.
1573 6. Added CONFIGURE_GROUP build-time option.
1575 7. Some older OS have a limit of 256 on the maximum number of file
1576 descriptors. Exim was using setrlimit() to set 1000 as a large value
1577 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1578 systems. I've change it so that if it can't get 1000, it tries for 256.
1580 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1581 was an oversight, and furthermore, ever since the addition of extra
1582 controls (e.g. 4.43/32), the checks on when to allow different forms of
1583 "control" were broken. There should now be diagnostics for all cases when a
1584 control that does not make sense is encountered.
1586 9. Added the /retain_sender option to "control=submission".
1588 10. $recipients is now available in the predata ACL (oversight).
1590 11. Tidy the search cache before the fork to do a delivery from a message
1591 received from the command line. Otherwise the child will trigger a lookup
1592 failure and thereby defer the delivery if it tries to use (for example) a
1593 cached ldap connection that the parent has called unbind on.
1595 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1596 of $address_data from the recipient verification was clobbered by the
1597 sender verification.
1599 13. The value of address_data from a sender verification is now available in
1600 $sender_address_data in subsequent conditions in the ACL statement.
1602 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1604 15. Added a new option "connect=<time>" to callout options, to set a different
1607 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1608 was its contents. (It was OK if the option was not defined at all.)
1610 17. A "Completed" log line is now written for messages that are removed from
1611 the spool by the -Mrm option.
1613 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1614 information about exactly what failed.
1616 19. Added -dd to debug only the daemon process.
1618 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1619 handles timeouts, both on the server side and network timeouts. Renamed the
1620 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1622 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1623 It is now set to "smtps".
1625 22. $host_address is now set to the target address during the checking of
1626 ignore_target_hosts.
1628 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1629 being passed; this would have caused $sender_host_name to have been used if
1630 matching the list had actually called for a host name (not very likely,
1631 since this list is usually IP addresses). A host name is now passed as
1634 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1635 code that specifies a non-restarting handler (typically sigaction() in
1636 modern systems) in an attempt to fix a rare and obscure crash bug.
1638 25. Narrowed the window for a race in the daemon that could cause it to ignore
1639 SIGCHLD signals. This is not a major problem, because they are used only to
1640 wake it up if nothing else does.
1642 26. A malformed maildirsize file could cause Exim to calculate negative values
1643 for the mailbox size or file count. Odd effects could occur as a result.
1644 The maildirsize information is now recalculated if the size or filecount
1647 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1648 support for a long time. Removed HAVE_SYS_VFS_H.
1650 28. Installed the latest version of exipick from John Jetmore.
1652 29. In an address list, if the pattern was not a regular expression, an empty
1653 subject address (from a bounce message) matched only if the pattern was an
1654 empty string. Non-empty patterns were not even tested. This was the wrong
1655 because it is perfectly reasonable to use an empty address as part of a
1656 database query. An empty address is now tested by patterns that are
1657 lookups. However, all the other forms of pattern expect the subject to
1658 contain a local part and a domain, and therefore, for them, an empty
1659 address still always fails if the pattern is not itself empty.
1661 30. Exim went into a mad DNS loop when attempting to do a callout where the
1662 host was specified on an smtp transport, and looking it up yielded more
1663 than one IP address.
1665 31. Re-factored the code for checking spool and log partition space into a
1666 function that finds that data and another that does the check. The former
1667 is then used to implement four new variables: $spool_space, $log_space,
1668 $spool_inodes, and $log_inodes.
1670 32. The RFC2047 encoding function was originally intended for short strings
1671 such as real names; it was not keeping to the 75-character limit for
1672 encoded words that the RFC imposes. It now respects the limit, and
1673 generates multiple encoded words if necessary. To be on the safe side, I
1674 have increased the buffer size for the ${rfc2047: expansion operator from
1677 33. It is now permitted to omit both strings after an "if" condition; if the
1678 condition is true, the result is "true". As before, when the second string
1679 is omitted, a false condition yields an empty string. This makes it less
1680 cumbersome to write custom ACL and router conditions.
1682 34. Failure to deliver a bounce message always caused it to be frozen, even if
1683 there was an errors_to setting on the router. The errors_to setting is now
1686 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1687 canonical form (fully expanded) before being placed in
1688 $sender_host_address.
1690 36. The table in the code that translates DNS record types into text (T_A to
1691 "A" for instance) was missing entries for NS and CNAME. It is just possible
1692 that this could have caused confusion if both these types were looked up
1693 for the same domain, because the text type is used as part of Exim's
1694 per-process caching. But the chance of anyone hitting this buglet seems
1697 37. The dnsdb lookup has been extended in a number of ways.
1699 (1) There is a new type, "zns", which walks up the domain tree until it
1700 finds some nameserver records. It should be used with care.
1702 (2) There is a new type, "mxh", which is like "mx" except that it returns
1703 just the host names, not the priorities.
1705 (3) It is now possible to give a list of domains (or IP addresses) to be
1706 looked up. The behaviour when one of the lookups defers can be
1707 controlled by a keyword.
1709 (4) It is now possible to specify the separator character for use when
1710 multiple records are returned.
1712 38. The dnslists ACL condition has been extended: it is now possible to supply
1713 a list of IP addresses and/or domains to be looked up in a particular DNS
1716 39. Added log_selector=+queue_time_overall.
1718 40. When running the queue in the test harness, wait just a tad after forking a
1719 delivery process, to get repeatability of debugging output.
1721 41. Include certificate and key file names in error message when GnuTLS fails
1722 to set them up, because the GnuTLS error message doesn't include the name
1723 of the failing file when there is a problem reading it.
1725 42. Allow both -bf and -bF in the same test run.
1727 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1729 44. The "Exiscan patch" is now merged into the mainline Exim source.
1731 45. Sometimes the final signoff response after QUIT could fail to get
1732 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1733 < 0 before doing a fflush(). This bug looks as though it goes back to the
1734 introduction of TLS in release 3.20, but "sometimes" must have been rare
1735 because the tests only now provoked it.
1737 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1738 (this can affect the format of dates).
1740 47. exim_tidydb, when checking for the continued existence of a message for
1741 which it has found a message-specific retry record, was not finding
1742 messages that were in split spool directories. Consequently, it was
1743 deleting retry records that should have stayed in existence.
1745 48. Steve fixed some bugs in eximstats.
1747 49. The SPA authentication driver was not abandoning authentication and moving
1748 on to the next authenticator when an expansion was forced to fail,
1749 contradicting the general specification for all authenticators. Instead it
1750 was generating a temporary error. It now behaves as specified.
1752 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1753 (the order specifies the preference for clients). The order is now AES256,
1754 AES128, 3DES, ARCFOUR128.
1756 51. Small patch to Sieve code - explicitly set From: when generating an
1759 52. Exim crashed if a remote delivery caused a very long error message to be
1760 recorded - for instance if somebody sent an entire SpamAssassin report back
1761 as a large number of 550 error lines. This bug was coincidentally fixed by
1762 increasing the size of one of Exim's internal buffers (big_buffer) that
1763 happened as part of the Exiscan merge. However, to be on the safe side, I
1764 have made the code more robust (and fixed the comments that describe what
1767 53. Now that there can be additional text after "Completed" in log lines (if
1768 the queue_time_overall log selector is set), a one-byte patch to exigrep
1769 was needed to allow it to recognize "Completed" as not the last thing in
1772 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1773 patch that reportedly fixes this has been added. I am not expert enough to
1774 create a test for it. This is what the patch creator wrote:
1776 "I found a little strange behaviour of ldap code when working with
1777 Windows 2003 AD Domain, where users was placed in more than one
1778 Organization Units. When I tried to give exim partial DN, the exit code
1779 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1780 But simultaneously result of request was absolutely normal ldap result,
1781 so I produce this patch..."
1783 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1784 REFERENCE, so I have modified the code to exclude the patch when that macro
1787 55. Some experimental protocols are using DNS PTR records for new purposes. The
1788 keys for these records are domain names, not reversed IP addresses. The
1789 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1790 leaves it alone. Component reversal etc. now happens only for IP addresses.
1793 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1795 57. Double the size of the debug message buffer (to 2048) so that more of very
1796 long debug lines gets shown.
1798 58. The exicyclog utility now does better if the number of log files to keep
1799 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1801 59. Two changes related to the smtp_active_hostname option:
1803 (1) $smtp_active_hostname is now available as a variable.
1804 (2) The default for smtp_banner uses $smtp_active_hostname instead
1805 of $primary_hostname.
1807 60. The host_aton() function is supposed to be passed a string that is known
1808 to be a valid IP address. However, in the case of IPv6 addresses, it was
1809 not checking this. This is a hostage to fortune. Exim now panics and dies
1810 if the condition is not met. A case was found where this could be provoked
1811 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1812 components; fortuitously, this particular loophole had already been fixed
1813 by change 4.50/55 above.
1815 If there are any other similar loopholes, the new check in host_aton()
1816 itself should stop them being exploited. The report I received stated that
1817 data on the command line could provoke the exploit when Exim was running as
1818 exim, but did not say which command line option was involved. All I could
1819 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1820 running as the user.
1823 61. There was a buffer overflow vulnerability in the SPA authentication code
1824 (which came originally from the Samba project). I have added a test to the
1825 spa_base64_to_bits() function which I hope fixes it.
1828 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1829 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1831 63. The daemon start-up calls getloadavg() while still root for those OS that
1832 need the first call to be done as root, but it missed one case: when
1833 deliver_queue_load_max is set with deliver_drop_privilege. This is
1834 necessary for the benefit of the queue runner, because there is no re-exec
1835 when deliver_drop_privilege is set.
1837 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1838 This has been fixed.
1840 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1841 was in use, was not putting the data itself into the right store pool;
1842 consequently, it could be overwritten for a subsequent message in the same
1843 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1846 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1848 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1849 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1850 the function were treating the return as a boolean value, which happened to
1851 work because 0=false and not-0=true, but is not correct code.
1853 68. The host_aton() function was not handling scoped IPv6 addresses (those
1854 with, for example, "%eth0" on the end) correctly.
1856 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1857 negated items (that is, ~something) in unsigned ints. Some compilers
1858 apparently mutter when there is no cast.
1860 70. If an address verification called from an ACL failed, and did not produce a
1861 user-specific message (i.e. there was only a "system" message), nothing was
1862 put in $acl_verify_message. In this situation, it now puts the system
1865 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1866 session; change 4.31/43 added the unwanted input to the log line - except
1867 that it did not do this in the start of session case. It now does.
1869 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1870 This is wrong and can cause the other end to generate a synchronization
1871 error if it is another Exim or anything else that does the synchronization
1872 check. A QUIT command is no longer sent after a timeout.
1874 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1875 during host lookups.
1877 74. The defer_ok option of callout verification was not working if it was used
1878 when verifying addresses in header lines, that is, for this case:
1880 verify = header_sender/callout=defer_ok
1882 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1883 those file descriptors could be used for SMTP connections. If anything
1884 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1885 could be sent to the SMTP client, causing chaos. The daemon now opens
1886 stdin, stdout, and stderr to /dev/null when it puts itself into the
1889 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1890 log by default. The user can override this with suitable Perl magic.
1892 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1893 the log message when discard triggers, was not working for the DATA ACL or
1894 for the non-SMTP ACL.
1896 78. Error message wording change in sieve.c.
1898 79. If smtp_accept_max_per_host was set, the number of connections could be
1899 restricted to fewer than expected, because the daemon was trying to set up
1900 a new connection before checking whether the processes handling previous
1901 connections had finished. The check for completed processes is now done
1902 earlier. On busy systems, this bug wouldn't be noticed because something
1903 else would have woken the daemon, and it would have reaped the completed
1906 80. If a message was submitted locally by a user whose login name contained one
1907 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1908 It caused a spool format error. I have fixed the spool reading code. A
1909 related problem was that the "from" clause in the Received: line became
1910 illegal because of the space(s). It is now covered by ${quote_local_part.
1912 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1915 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1917 83. Patch from Sieve maintainer to fix unterminated string problem in
1918 "vacation" handling.
1920 84. Some minor changes to the Linux configuration files to help with other
1921 OS variants using glibc.
1923 85. One more patch for Sieve to update vacation handling to latest spec.
1926 ----------------------------------------------------
1927 See the note above about the 4.44 and 4.50 releases.
1928 ----------------------------------------------------
1934 1. Change 4.43/35 introduced a bug that caused file counts to be
1935 incorrectly computed when quota_filecount was set in an appendfile
1938 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1939 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1942 3. Give more explanation in the error message when the command for a transport
1943 filter fails to execute.
1945 4. There are several places where Exim runs a non-Exim command in a
1946 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1947 was being done only for the command run by the queryprogram router. It is
1948 now done for all such subprocesses. The other cases are: ${run, transport
1949 filters, and the commands run by the lmtp and pipe transports.
1951 5. Some older OS have a limit of 256 on the maximum number of file
1952 descriptors. Exim was using setrlimit() to set 1000 as a large value
1953 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1954 systems. I've change it so that if it can't get 1000, it tries for 256.
1956 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1957 was an oversight, and furthermore, ever since the addition of extra
1958 controls (e.g. 4.43/32), the checks on when to allow different forms of
1959 "control" were broken. There should now be diagnostics for all cases when a
1960 control that does not make sense is encountered.
1962 7. $recipients is now available in the predata ACL (oversight).
1964 8. Tidy the search cache before the fork to do a delivery from a message
1965 received from the command line. Otherwise the child will trigger a lookup
1966 failure and thereby defer the delivery if it tries to use (for example) a
1967 cached ldap connection that the parent has called unbind on.
1969 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1970 of $address_data from the recipient verification was clobbered by the
1971 sender verification.
1973 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1974 was its contents. (It was OK if the option was not defined at all.)
1976 11. A "Completed" log line is now written for messages that are removed from
1977 the spool by the -Mrm option.
1979 12. $host_address is now set to the target address during the checking of
1980 ignore_target_hosts.
1982 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1983 being passed; this would have caused $sender_host_name to have been used if
1984 matching the list had actually called for a host name (not very likely,
1985 since this list is usually IP addresses). A host name is now passed as
1988 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1989 code that specifies a non-restarting handler (typically sigaction() in
1990 modern systems) in an attempt to fix a rare and obscure crash bug.
1992 15. Narrowed the window for a race in the daemon that could cause it to ignore
1993 SIGCHLD signals. This is not a major problem, because they are used only to
1994 wake it up if nothing else does.
1996 16. A malformed maildirsize file could cause Exim to calculate negative values
1997 for the mailbox size or file count. Odd effects could occur as a result.
1998 The maildirsize information is now recalculated if the size or filecount
2001 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2002 support for a long time. Removed HAVE_SYS_VFS_H.
2004 18. Updated exipick to current release from John Jetmore.
2006 19. Allow an empty sender to be matched against a lookup in an address list.
2007 Previously the only cases considered were a regular expression, or an
2010 20. Exim went into a mad DNS lookup loop when doing a callout where the
2011 host was specified on the transport, if the DNS lookup yielded more than
2014 21. The RFC2047 encoding function was originally intended for short strings
2015 such as real names; it was not keeping to the 75-character limit for
2016 encoded words that the RFC imposes. It now respects the limit, and
2017 generates multiple encoded words if necessary. To be on the safe side, I
2018 have increased the buffer size for the ${rfc2047: expansion operator from
2021 22. Failure to deliver a bounce message always caused it to be frozen, even if
2022 there was an errors_to setting on the router. The errors_to setting is now
2025 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2026 canonical form (fully expanded) before being placed in
2027 $sender_host_address.
2029 24. Updated eximstats to version 1.33
2031 25. Include certificate and key file names in error message when GnuTLS fails
2032 to set them up, because the GnuTLS error message doesn't include the name
2033 of the failing file when there is a problem reading it.
2035 26. Expand error message when OpenSSL has problems setting up cert/key files.
2038 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2039 (this can affect the format of dates).
2041 28. exim_tidydb, when checking for the continued existence of a message for
2042 which it has found a message-specific retry record, was not finding
2043 messages that were in split spool directories. Consequently, it was
2044 deleting retry records that should have stayed in existence.
2046 29. eximstats updated to version 1.35
2047 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2048 1.35 - bugfix such that pie charts by volume are generated correctly
2050 30. The SPA authentication driver was not abandoning authentication and moving
2051 on to the next authenticator when an expansion was forced to fail,
2052 contradicting the general specification for all authenticators. Instead it
2053 was generating a temporary error. It now behaves as specified.
2055 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2056 (the order specifies the preference for clients). The order is now AES256,
2057 AES128, 3DES, ARCFOUR128.
2059 31. Small patch to Sieve code - explicitly set From: when generating an
2062 32. Exim crashed if a remote delivery caused a very long error message to be
2063 recorded - for instance if somebody sent an entire SpamAssassin report back
2064 as a large number of 550 error lines. This bug was coincidentally fixed by
2065 increasing the size of one of Exim's internal buffers (big_buffer) that
2066 happened as part of the Exiscan merge. However, to be on the safe side, I
2067 have made the code more robust (and fixed the comments that describe what
2070 33. Some experimental protocols are using DNS PTR records for new purposes. The
2071 keys for these records are domain names, not reversed IP addresses. The
2072 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2073 leaves it alone. Component reversal etc. now happens only for IP addresses.
2076 34. The host_aton() function is supposed to be passed a string that is known
2077 to be a valid IP address. However, in the case of IPv6 addresses, it was
2078 not checking this. This is a hostage to fortune. Exim now panics and dies
2079 if the condition is not met. A case was found where this could be provoked
2080 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2081 components; fortuitously, this particular loophole had already been fixed
2082 by change 4.50/55 or 4.44/33 above.
2084 If there are any other similar loopholes, the new check in host_aton()
2085 itself should stop them being exploited. The report I received stated that
2086 data on the command line could provoke the exploit when Exim was running as
2087 exim, but did not say which command line option was involved. All I could
2088 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2089 running as the user.
2092 35. There was a buffer overflow vulnerability in the SPA authentication code
2093 (which came originally from the Samba project). I have added a test to the
2094 spa_base64_to_bits() function which I hope fixes it.
2097 36. The daemon start-up calls getloadavg() while still root for those OS that
2098 need the first call to be done as root, but it missed one case: when
2099 deliver_queue_load_max is set with deliver_drop_privilege. This is
2100 necessary for the benefit of the queue runner, because there is no re-exec
2101 when deliver_drop_privilege is set.
2103 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2104 was in use, was not putting the data itself into the right store pool;
2105 consequently, it could be overwritten for a subsequent message in the same
2106 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2109 38. Sometimes the final signoff response after QUIT could fail to get
2110 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2111 < 0 before doing a fflush(). This bug looks as though it goes back to the
2112 introduction of TLS in release 3.20, but "sometimes" must have been rare
2113 because the tests only now provoked it.
2119 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2120 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2121 Now it returns an integer. A number of calls were still expecting a T/F
2122 return. Fortuitously, in all cases, the tests worked in OK situations,
2123 which is the norm. However, things would have gone wrong on any write
2124 failures on the smtp file descriptor. This function is used when sending
2125 messages over SMTP and also when doing verify callouts.
2127 2. When Exim is called to do synchronous delivery of a locally submitted
2128 message (the -odf or -odi options), it no longer closes stderr before doing
2131 3. Implemented the mua_wrapper option.
2133 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2135 5. Implemented the functions header_remove(), header_testname(),
2136 header_add_at_position(), and receive_remove_recipient(), and exported them
2139 6. If an ACL "warn" statement specified the addition of headers, Exim already
2140 inserted X-ACL-Warn: at the start if there was no header name. However, it
2141 was not making this test for the second and subsequent header lines if
2142 there were newlines in the string. This meant that an invalid header could
2143 be inserted if Exim was badly configured.
2145 7. Allow an ACL "warn" statement to add header lines at the start or after all
2146 the Received: headers, as well as at the end.
2148 8. Added the rcpt_4xx retry error code.
2150 9. Added postmaster_mailfrom=xxx to callout verification option.
2152 10. Added mailfrom=xxxx to the callout verification option, for verify=
2155 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2156 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2158 12. Inserted some casts to stop certain compilers warning when using pointer
2159 differences as field lengths or precisions in printf-type calls (mostly
2160 affecting debugging statements).
2162 13. Added optional readline() support for -be (dynamically loaded).
2164 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2165 same clock tick as a message's arrival, so that its received time was the
2166 same as the "first fail" time on the retry record, and that message
2167 remained on the queue past the ultimate address timeout, every queue runner
2168 would try a delivery (because it was past the ultimate address timeout) but
2169 after another failure, the ultimate address timeout, which should have then
2170 bounced the address, did not kick in. This was a "< instead of <=" error;
2171 in most cases the first failure would have been in the next clock tick
2172 after the received time, and all would be well.
2174 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2175 being recognized when the domain list was tested by the match_domain
2176 condition in an expansion string.
2178 16. Added the ${str2b64: operator.
2180 17. Exim was always calling setrlimit() to set a large limit for the number of
2181 processes, without checking whether the existing limit was already
2182 adequate. (It did check for the limit on file descriptors.) Furthermore,
2183 errors from getrlimit() and setrlimit() were being ignored. Now they are
2184 logged to the main and panic logs, but Exim does carry on, to try to do its
2185 job under whatever limits there are.
2187 18. Imported PCRE 5.0.
2189 19. Trivial typo in log message " temporarily refused connection" (the leading
2192 20. If the log selector return_path_on_delivery was set and an address was
2193 redirected to /dev/null, the delivery process crashed because it assumed
2194 that a return path would always be set for a "successful" delivery. In this
2195 case, the whole delivery is bypassed as an optimization, and therefore no
2198 21. Internal re-arrangement: the function for sending a challenge and reading
2199 a response while authentication was assuming a zero-terminated challenge
2200 string. It's now changed to take a pointer and a length, to allow for
2201 binary data in such strings.
2203 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2205 23. Exim was not respecting finduser_retries when seeking the login of the
2206 uid under which it was called; it was always trying 10 times. (The default
2207 setting of finduser_retries is zero.) Also, it was sleeping after the final
2208 failure, which is pointless.
2210 24. Implemented tls_on_connect_ports.
2212 25. Implemented acl_smtp_predata.
2214 26. If the domain in control=submission is set empty, Exim assumes that the
2215 authenticated id is a complete email address when it generates From: or
2216 Sender: header lines.
2218 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2219 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2220 chown and chgrp in /bin and hostname in /usr/bin.
2222 28. Exim was keeping the "process log" file open after each use, just as it
2223 does for the main log. This opens the possibility of it remaining open for
2224 long periods when the USR1 signal hits a daemon. Occasional processlog
2225 errors were reported, that could have been caused by this. Anyway, it seems
2226 much more sensible not to leave this file open at all, so that is what now
2229 29. The long-running daemon process does not normally write to the log once it
2230 has entered its main loop, and it closes the log before doing so. This is
2231 so that log files can straightforwardly be renamed and moved. However,
2232 there are a couple of unusual error situations where the daemon does write
2233 log entries, and I had neglected to close the log afterwards.
2235 30. The text of an SMTP error response that was received during a remote
2236 delivery was being truncated at 512 bytes. This is too short for some of
2237 the long messages that one sometimes sees. I've increased the limit to
2240 31. It is now possible to make retry rules that apply only when a message has a
2241 specific sender, in particular, an empty sender.
2243 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2244 it possible to be selective about when SMTP synchronization is enforced.
2246 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2248 32. Implemented hosts_connection_nolog.
2250 33. Added an ACL for QUIT.
2252 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2255 35. Added mailbox_size and mailbox_filecount to appendfile.
2257 36. Added control = no_multiline_responses to ACLs.
2259 37. There was a bug in the logic of the code that waits for the clock to tick
2260 in the case where the clock went backwards by a substantial amount such
2261 that the microsecond fraction of "now" was more than the microsecond
2262 fraction of "then" (but the whole seconds number was less).
2264 38. Added support for the libradius Radius client library this is found on
2265 FreeBSD (previously only the radiusclient library was supported).
2271 1. When certain lookups returned multiple values in the form name=value, the
2272 quoting of the values was not always being done properly. Specifically:
2273 (a) If the value started with a double quote, but contained no whitespace,
2275 (b) If the value contained whitespace other than a space character (i.e.
2276 tabs or newlines or carriage returns) it was not quoted.
2277 This fix has been applied to the mysql and pgsql lookups by writing a
2278 separate quoting function and calling it from the lookup code. The fix
2279 should probably also be applied to nisplus, ibase and oracle lookups, but
2280 since I cannot test any of those, I have not disturbed their existing code.
2282 2. A hit in the callout cache for a specific address caused a log line with no
2283 reason for rejecting RCPT. Now it says "Previous (cached) callout
2284 verification failure".
2286 3. There was an off-by-one bug in the queryprogram router. An over-long
2287 return line was truncated at 256 instead of 255 characters, thereby
2288 overflowing its buffer with the terminating zero. As well as fixing this, I
2289 have increased the buffer size to 1024 (and made a note to document this).
2291 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2292 when Exim is waiting for an SMTP response from a remote server, Exim
2293 restarts its select() call on the socket, thereby resetting its timeout.
2294 This is not a problem when such interrupts are rare. Somebody set up a cron
2295 job to run exiwhat every 2 minutes, which is less than the normal select()
2296 timeout (5 or 10 minutes). This meant that the select() timeout never
2297 kicked in because it was always reset. I have fixed this by comparing the
2298 time when an interrupt arrives with the time at the start of the first call
2299 to select(). If more time than the timeout has elapsed, the interrupt is
2300 treated as a timeout.
2302 5. Some internal re-factoring in preparation for the addition of Sieve
2303 extensions (by MH). In particular, the "personal" test is moved to a
2304 separate function, and given an option for scanning Cc: and Bcc: (which is
2305 not set for Exim filters).
2307 6. When Exim created an email address using the login of the caller as the
2308 local part (e.g. when creating a From: or Sender: header line), it was not
2309 quoting the local part when it contained special characters such as @.
2311 7. Installed new OpenBSD configuration files.
2313 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2314 try to make them clearer.
2316 9. Callout options, other than the timeout value, were being ignored when
2317 verifying sender addresses in header lines. For example, when using
2319 verify = header_sender/callout=no_cache
2321 the cache was (incorrectly) being used.
2323 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2324 only the Cygwin environment.
2326 11. When return_path_on_delivery was set as a log selector, if different remote
2327 addresses in the same message used different return paths and parallel
2328 remote delivery occurred, the wrong values would sometimes be logged.
2329 (Whenever a remote delivery process finished, the return path value from
2330 the most recently started remote delivery process was logged.)
2332 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2333 lines when AUTH and/or TLS are in use. This is the "received protocol"
2334 field. Exim used to use "asmtp" for authenticated SMTP, without any
2335 indication (in the protocol name) for TLS use. Now it follows the RFC and
2336 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2337 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2338 names appear in log lines as well as in Received: header lines.
2340 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2341 extensions, and comparison tests, and to fix some bugs.
2343 14. Changes to the "personal" filter test:
2345 (1) The test was buggy in that it was just doing the equivalent of
2346 "contains" tests on header lines. For example, if a user's address was
2347 anne@some.where, the "personal" test would incorrectly be true for
2349 To: susanne@some.where
2351 This test is now done by extracting each address from the header in turn,
2352 and checking the entire address. Other tests that are part of "personal"
2353 are now done using regular expressions (for example, to check local parts
2354 of addresses in From: header lines).
2356 (2) The list of non-personal local parts in From: addresses has been
2357 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2358 taken from the Sieve specification recommendations.
2360 (3) If the message contains any header line starting with "List-" it is
2361 treated as non-personal.
2363 (4) The test for "circular" in the Subject: header line has been removed
2364 because it now seems ill-conceived.
2366 15. Minor typos in src/EDITME comments corrected.
2368 16. Installed latest exipick from John Jetmore.
2370 17. If headers_add on a router specified a text string that was too long for
2371 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2372 of string_sprintf() is now avoided.
2374 18. $message_body_size was not set (it was always zero) when running the DATA
2375 ACL and the local_scan() function.
2377 19. For the "mail" command in an Exim filter, no default was being set for
2378 the once_repeat time, causing a random time value to be used if "once" was
2379 specified. (If the value happened to be <= 0, no repeat happened.) The
2380 default is now 0s, meaning "never repeat". The "vacation" command was OK
2381 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2382 (I found it when inspecting the code).
2384 20. There is now an overall timeout for performing a callout verification. It
2385 defaults to 4 times the callout timeout, which applies to individual SMTP
2386 commands during the callout. The overall timeout applies when there is more
2387 than one host that can be tried. The timeout is checked before trying the
2388 next host. This prevents very long delays if there are a large number of
2389 hosts and all are timing out (e.g. when the network connections are timing
2390 out). The value of the overall timeout can be changed by specifying an
2391 additional sub-option for "callout", called "maxwait". For example:
2393 verify = sender/callout=5s,maxwait=20s
2395 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2396 to the end before writing, but this should make it even safer).
2398 22. Exim was forgetting that it had advertised PIPELINING for the second and
2399 subsequent messages on an SMTP connection. It was also not resetting its
2400 memory on STARTTLS and an internal HELO.
2402 23. When Exim logs an SMTP synchronization error within a session, it now
2403 records whether PIPELINING has been advertised or not.
2405 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2406 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2407 rather than long int.
2409 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2411 26. Added the never_mail option to autoreply.
2417 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2418 crash if the getsockname() call failed; this can happen if a connection is
2419 closed very soon after it is established. The problem was simply in the
2420 order in which certain operations were done, causing Exim to try to write
2421 to the SMTP stream before it had set up the file descriptor. The bug has
2422 been fixed by making things happen in the correct order.
2428 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2429 before the connection was closed, thus losing the rejection response.
2431 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2432 some early Solaris releases, but causes trouble in current releases where
2433 socklen_t is defined.
2435 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2438 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2441 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2442 that releasing the top of it at the end releases what was used for sub-
2443 expansions (unless the block got too big). However, discard this block if
2444 the first thing is a variable or header, so that we can use its block when
2445 it is dynamic (useful for very large $message_headers, for example).
2447 6. Lookups now cache *every* query, not just the most recent. A new, separate
2448 store pool is used for this. It can be recovered when all lookup caches are
2449 flushed. Lookups now release memory at the end of their result strings.
2450 This has involved some general refactoring of the lookup sources.
2452 7. Some code has been added to the store_xxx() functions to reduce the amount
2453 of flapping under certain conditions.
2455 8. log_incoming_interface used to affect only the <= reception log lines. Now
2456 it causes the local interface and port to be added to several more SMTP log
2457 lines, for example "SMTP connection from", and rejection lines.
2459 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2461 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2463 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2464 could be overwritten at the end of the current message (or the start of a
2465 new message if it was set in a HELO ACL). The value is now preserved for
2466 the duration of the SMTP connection.
2468 12. If a transport had a headers_rewrite setting, and a matching header line
2469 contained an unqualified address, that address was qualified, even if it
2470 did not match any rewriting rules. The underlying bug was that the values
2471 of the flags that permit the existence of unqualified sender and recipient
2472 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2473 non-local messages, and by -bnq for local messages) were not being
2474 preserved with the message after it was received.
2476 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2477 "next input=" as part of the text comprising the host identity instead of
2478 the correct text. The code was using the same buffer for two different
2479 strings. However, depending on which order the printing function evaluated
2480 its arguments, the bug did not always show up. Under Linux, for example, my
2481 test suite worked just fine.
2483 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2484 doesn't work with some older versions of Perl. It has been changed to "my",
2485 which in any case is probably the better facility to use.
2487 15. A really picky compiler found some instances of statements for creating
2488 error messages that either had too many or two few arguments for the format
2491 16. The size of the buffer for calls to the DNS resolver has been increased
2492 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2493 for addresses that have a lot of PTR records. This alleviates a problem; it
2494 does not fully solve it.
2496 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2497 buffer now truncates the list and logs the incident, which is the same
2498 action as happens when Exim is looking up a host name and its aliases.
2499 Previously in this situation something unpredictable would happen;
2500 sometimes it was "internal error: store_reset failed".
2502 18. If a server dropped the connection unexpectedly when an Exim client was
2503 using GnuTLS and trying to read a response, the client delivery process
2504 crashed while trying to generate an error log message.
2506 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2507 string, for example:
2509 warn message = H1: something\nH2: something
2511 the text was added as a single header line from Exim's point of view
2512 though it ended up OK in the delivered message. However, searching for the
2513 second and subsequent header lines using $h_h2: did not work. This has been
2514 fixed. Similarly, if a system filter added multiple headers in this way,
2515 the routers could not see them.
2517 20. Expanded the error message when iplsearch is called with an invalid key to
2518 suggest using net-iplsearch in a host list.
2520 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2521 ACLs are no longer actually imposed (and a message to that effect is
2524 22. If a "gecos" field in a passwd entry contained escaped characters, in
2525 particular, if it contained a \" sequence, Exim got it wrong when building
2526 a From: or a Sender: header from that name. A second bug also caused
2527 incorrect handling when an unquoted " was present following a character
2528 that needed quoting.
2530 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2531 was not being matched caselessly.
2533 24. Arranged for all hyphens in the exim.8 source to be escaped with
2536 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2537 properly. Recipient callout cache records were still being keyed to include
2538 the sender, even when use_sender was set false. This led to far more
2539 callouts that were necessary. The sender is no longer included in the key
2540 when use_sender is false.
2542 26. Added "control = submission" modifier to ACLs.
2544 27. Added the ${base62d: operator to decode base 62 numbers.
2546 28. dnsdb lookups can now access SRV records.
2548 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2549 the configuration file.
2551 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2552 (-v) mode. This makes the output for a verbose queue run more intelligible.
2554 31. Added a use_postmaster feature to recipient callouts.
2556 32. Added the $body_zerocount variable, containing the number of binary zero
2557 bytes in the message body.
2559 33. The time of last modification of the "new" subdirectory is now used as the
2560 "mailbox time last read" when there is a quota error for a maildir
2563 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2565 35. Added +ignore_unknown as a special item in host lists.
2567 36. Code for decoding IPv6 addresses in host lists is now included, even if
2568 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2569 address was recognized as an IP address, but was then not correctly decoded
2570 into binary, causing unexpected and incorrect effects when compared with
2577 1. Very minor rewording of debugging text in manualroute to say "list of
2578 hosts" instead of "hostlist".
2580 2. If verify=header_syntax was set, and a header line with an unqualified
2581 address (no domain) and a large number of spaces between the end of the
2582 name and the colon was received, the reception process suffered a buffer
2583 overflow, and (when I tested it) crashed. This was caused by some obsolete
2584 code that should have been removed. The fix is to remove it!
2586 3. When running in the test harness, delay a bit after writing a bounce
2587 message to get a bit more predictability in the log output.
2589 4. Added a call to search_tidyup() just before forking a reception process. In
2590 theory, someone could use a lookup in the expansion of smtp_accept_max_
2591 per_host which, without the tidyup, could leave open a database connection.
2593 5. Added the variables $recipient_data and $sender_data which get set from a
2594 lookup success in an ACL "recipients" or "senders" condition, or a router
2595 "senders" option, similar to $domain_data and $local_part_data.
2597 6. Moved the writing of debug_print from before to after the "senders" test
2600 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2601 problems for message scanning, either using a data ACL, or using
2602 local_scan() because the Received: header was not generated till after they
2603 were called (in order to set the time as the time of reception completion).
2604 I have revised the way this works. The header is now generated after the
2605 body is received, but before the ACL or local_scan() are called. After they
2606 are run, the timestamp in the header is updated.
2612 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2613 before starting a queue runner without re-exec. This happened only when
2614 deliver_drop_privilege was set or when the Exim user was set to root. The
2615 effect of the bug was that timeouts during subsequent deliveries caused
2616 crashes instead of being properly handled. The handler is now left at its
2617 default (and expected) setting.
2619 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2620 message, again when deliver_drop_privilege is set or Exim is run as root.
2621 The bug described in (1) was not present in this case, but the tidying up
2622 of the other signals was missing. I have made the two cases consistent.
2624 3. The ignore_target_hosts setting on a manualroute router was being ignored
2625 for hosts that were looked up using the /MX notation.
2627 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2630 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2631 operated on the sender address. After changing the $sender_address to <>
2632 for the sender address verify, Exim was re-instated it as the original
2633 (before rewriting) address, but remembering that it had rewritten it, so it
2634 wasn't rewriting it again. This bug also had the effect of breaking the
2635 sender address verification caching when the sender address was rewritten.
2637 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2638 This has been changed so that if the ip literal address matches
2639 ignore_target_hosts, the router declines.
2641 7. Added expansion conditions match_domain, match_address, and match_local_
2642 part (NOT match_host).
2644 8. The placeholder for the Received: header didn't have a length field set.
2646 9. Added code to Exim itself and to exim_lock to test for a specific race
2647 condition that could lead to file corruption when using MBX delivery. The
2648 issue is with the lockfile that is created in /tmp. If this file is removed
2649 after a process has opened it but before that process has acquired a lock,
2650 there is the potential for a second process to recreate the file and also
2651 acquire a lock. This could lead to two Exim processes writing to the file
2652 at the same time. The added code performs the same test as UW imapd; it
2653 checks after acquiring the lock that its file descriptor still refers to
2654 the same named file.
2656 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2657 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2660 11. Added the smtp_active_hostname option. If used, this will typically be made
2661 to depend on the incoming interface address. Because $interface_address is
2662 not set up until the daemon has forked a reception process, error responses
2663 that can happen earlier (such as "too many connections") no longer contain
2666 12. If an expansion in a condition on a "warn" statement fails because a lookup
2667 defers, the "warn" statement is abandoned, and the next ACL statement is
2668 processed. Previously this caused the whole ACL to be aborted.
2670 13. Added the iplsearch lookup type.
2672 14. Added ident_timeout as a log selector.
2674 15. Added tls_certificate_verified as a log selector.
2676 16. Added a global option tls_require_ciphers (compare the smtp transport
2677 option of the same name). This controls incoming TLS connections.
2679 17. I finally figured out how to make tls_require_ciphers do a similar thing
2680 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2681 before starting the TLS session.
2683 18. Tabs are now shown as \t in -bP output.
2685 19. If the log selector return_path_on_delivery was set, Exim crashed when
2686 bouncing a message because it had too many Received: header lines.
2688 20. If two routers both had headers_remove settings, and the first one included
2689 a superfluous trailing colon, the final name in the first list and the
2690 first name in the second list were incorrectly joined into one item (with a
2691 colon in the middle).
2697 1. Added -C and -D options to the exinext utility, mainly to make it easier
2698 to include in the automated testing, but these could be helpful when
2699 multiple configurations are in use.
2701 2. The exinext utility was not formatting the output nicely when there was
2702 an alternate port involved in the retry record key, nor when there was a
2703 message id as well (for retries that were specific to a specific message
2704 and a specific host). It was also confused by IPv6 addresses, because of
2705 the additional colons they contain. I have fixed the IPv4 problem, and
2706 patched it up to do a reasonable job for IPv6.
2708 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2709 delivery, the log line now contains "pipelined" if PIPELINING was used.
2711 4. An SMTP transport process used to panic and die if the bind() call to set
2712 an explicit outgoing interface failed. This has been changed; it is now
2713 treated in the same way as a connect() failure.
2715 5. A reference to $sender_host_name in the part of a conditional expansion
2716 that was being skipped was still causing a DNS lookup. This no longer
2719 6. The def: expansion condition was not recognizing references to header lines
2720 that used bh_ and bheader_.
2722 7. Added the _cache feature to named lists.
2724 8. The code for checking quota_filecount in the appendfile transport was
2725 allowing one more file than it should have been.
2727 9. For compatibility with Sendmail, the command line option
2735 and sets the incoming protocol and host name (for trusted callers). The
2736 host name and its colon can be omitted when only the protocol is to be set.
2737 Note the Exim already has two private options, -pd and -ps, that refer to
2738 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2739 "s", but I don't think that's a major issue.
2741 10. A number of refactoring changes to the code, none of which should affect
2744 (a) The number of logging options was getting close to filling up the
2745 32-bit word that was used as a bit map. I have split them into two classes:
2746 those that are passed in the argument to log_write(), and those that are
2747 only ever tested independently outside of that function. These are now in
2748 separate 32-bit words, so there is plenty of room for expansion again.
2749 There is no change in the user interface or the logging behaviour.
2751 (b) When building, for example, log lines, the code previously used a
2752 macro that called string_cat() twice, in order to add two strings. This is
2753 not really sufficiently general. Furthermore, there was one instance where
2754 it was actually wrong because one of the argument was used twice, and in
2755 one call a function was used. (As it happened, calling the function twice
2756 did not affect the overall behaviour.) The macro has been replaced by a
2757 function that can join an arbitrary number of extra strings onto a growing
2760 (c) The code for expansion conditions now uses a table and a binary chop
2761 instead of a serial search (which was left over from when there were very
2762 few conditions). Also, it now recognizes conditions like "pam" even when
2763 the relevant support is not compiled in: a suitably worded error message is
2764 given if an attempt is made to use such a condition.
2766 11. Added ${time_interval:xxxxx}.
2768 12. A bug was causing one of the ddress fields not to be passed back correctly
2769 from remote delivery subprocesses. The field in question was not being
2770 subsequently used, so this caused to problems in practice.
2772 13. Added new log selectors queue_time and deliver_time.
2774 14. Might have fixed a bug in maildirsizefile handling that threw up
2775 "unexpected character" debug warnings, and recalculated the data
2776 unnecessarily. In any case, I expanded the warning message to give more
2779 15. Added the message "Restricted characters in address" to the statements in
2780 the default ACL that block characters like @ and % in local parts.
2782 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2783 Three changes have been made:
2785 (a) There was a serious bug; a negative response to MAIL caused the whole
2786 recipient domain to be cached as invalid, thereby blocking all messages
2787 to all local parts at the same domain, from all senders. This bug has
2788 been fixed. The domain is no longer cached after a negative response to
2789 MAIL if the sender used is not empty.
2791 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2794 (c) A new callout option, "use_sender" has been added for people who want
2795 the modified behaviour.
2801 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2804 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2805 indeed breaks things for older releases.
2807 3. Added additional logging to the case where there is a problem reading data
2808 from a filter that is running in a subprocess using a pipe, in order to
2809 try to track down a specific problem.
2811 4. Testing facility fudge: when running in the test harness and attempting
2812 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2813 getting "No route to host". Convert this to a timeout.
2815 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2818 6. Some OS don't have socklen_t but use size_t instead. This affects the
2819 fifth argument of getsockopt() amongst other things. This is now
2820 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2821 can be set for individual OS. I have set it for SunOS5, OSF1, and
2822 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2823 some earlier ones do not.
2825 7. Change 4.30/15 was not doing the test caselessly.
2827 8. The standard form for an IPv6 address literal was being rejected by address
2828 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2829 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2830 this, as well as the form without the "IPv6" on the front (but only when
2831 address literals are enabled, of course).
2833 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2835 10. Exim crashed if a message with an empty sender address specified by -f
2836 encountered a router with an errors_to setting. This could be provoked only
2837 by a command such as
2841 where an empty string was supplied; "<>" did not hit this bug.
2843 11. Installed PCRE release 4.5.
2845 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2846 remained set. It is now erased.
2848 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2849 times from message ids (which are base 36 rather than the normal 62).
2851 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2852 were being counted as actual protocol errors, and logged if the log
2853 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2854 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2855 and DATA following a set of rejected RCPTs do not count as protocol errors.
2856 In other words, Exim assumes they were pipelined, though this may not
2857 actually be the case. Of course, in all cases the client gets an
2858 appropriate error code.
2860 15. If a lookup fails in an ACL condition, a message about the failure may
2861 be available; it is used if testing the ACL cannot continue, because most
2862 such messages specify what the cause of the deferral is. However, some
2863 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2864 that caused an old message to be retained and used if a later statement
2865 caused a defer, replacing the real cause of the deferral.
2867 16. If an IP address had so many PTR records that the DNS lookup buffer
2868 was not large enough to hold them, Exim could crash while trying to process
2869 the truncated data. It now detects and logs this case.
2871 17. Further to 4.21/58, another change has been made: if (and only if) the
2872 first line of a message (the first header line) ends with CRLF, a bare LF
2873 in a subsequent header line has a space inserted after it, so as not to
2874 terminate the header.
2876 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2877 unnecessarily, used atoi() instead of strtol(), and didn't check the
2878 termination when getting file sizes from file names by regex.
2880 19. Completely re-implemented the support for maildirsize files, in the light
2881 of a number of problems with the previous contributed implementation
2882 (4.30/29). In particular:
2884 . If the quota is zero, the maildirsize file is maintained, but no quota is
2887 . If the maildir directory does not exist, it is created before any attempt
2888 to write a maildirsize file.
2890 . The quota value in the file is just a cache; if the quota is changed in
2891 the transport, the new value overrides.
2893 . A regular expression is available for excluding directories from the
2896 20. The autoreply transport checks the characters in options that define the
2897 message's headers; it allows continued headers, but it was checking with
2898 isspace() after an embedded newline instead of explicitly looking for a
2901 21. If all the "regular" hosts to which an address was routed had passed their
2902 expiry times, and had not reached their retry times, the address was
2903 bounced, even if fallback hosts were defined. Now Exim should go on to try
2906 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2907 equivalent code in the SMTP transport. Some hosts send humungous responses
2908 to HELO/EHLO, more than 1024 it seems.
2910 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2911 gives compiler warnings in some environments. I've now done it "properly",
2914 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2915 (because of problems with the built-in one) was declared to return uschar *
2916 instead of char *, causing compiler failure.
2918 25. Fixed a file descriptor leak when processing alias/forward files.
2920 26. Fixed a minor format string issue in dbfn.c.
2922 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2924 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2925 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2928 29. When a sender address is verified, it is cached, to save repeating the test
2929 when there is more than one recipient in a message. However, when the
2930 verification involves a callout, it is possible for different callout
2931 options to be set for different recipients. It is too complicated to keep
2932 track of this in the cache, so now Exim always runs a verification when a
2933 callout is required, relying on the callout cache for the optimization.
2934 The overhead is duplication of the address routing, but this should not be
2937 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2938 to be verified with callout=postmaster, and the main callout worked but the
2939 postmaster check failed, the verification correctly failed. However, if a
2940 subsequent RCPT command asked for sender verification *without* the
2941 postmaster check, incorrect caching caused this verification also to fail,
2944 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2945 it was not caching the DNS options (qualify_single, search_parents) that
2946 were used when the lookup failed. A subsequent lookup with different
2947 options therefore always gave the same answer, though there were cases
2948 where it should not have. (Example: a "domains = !$mx_any" option on a
2949 dnslookup router: the "domains" option is always processed without any
2950 widening, but the router might have qualify_single set.) Now Exim uses the
2951 cached value only when the same options are set.
2953 32. Added John Jetmore's "exipick" utility to the distribution.
2955 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2956 than a timeout (e.g. a certificate is required, and is not provided), an
2957 Exim server now closes the connection immediately. Previously it waited for
2958 the client to close - but if the client is SSL, it seems that they each
2959 wait for each other, leading to a delay before one of them times out.
2961 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2962 maintained 0.8.x compatibility because I don't think many are using it, and
2963 it is clearly obsolete.
2965 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2968 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2969 client certificate was expired. A simple patch fixes this, though I don't
2970 understand the full logic of why the verify callback is called multiple
2973 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2974 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2975 which causes problems with some clients (such as the Certicom SSL Plus
2976 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2977 disables the coutermeasure allowing Eudora to connect."
2979 38. Exim was not checking that a write() to a log file succeeded. This could
2980 lead to Bad Things if a log got too big, in particular if it hit a file
2981 size limit. Exim now panics and dies if it cannot write to a log file, just
2982 as it does if it cannot open a log file.
2984 39. Modified OS/Makefile-Linux so that it now contains
2986 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2988 The two -D definitions ensure that Exim is compiled with large file
2989 support, which makes it possible to handle log files that are bigger than
2992 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2993 instance) a domain was checked against a named list that involved a lookup,
2994 causing $domain_data to be set, then another domain was checked against the
2995 same list, then the first domain was re-checked, the value of $domain_data
2996 after the final check could be wrong. In particular, if the second check
2997 failed, it could be set empty. This bug probably also applied to
3000 41. The strip_trailing_dot option was not being applied to the address given
3001 with the -f command-line option.
3003 42. The code for reading a message's header from the spool was incrementing
3004 $received_count, but never initializing it. This meant that the value was
3005 incorrect (doubled) while delivering a message in the same process in which
3006 it was received. In the most common configuration of Exim, this never
3007 happens - a fresh exec is done - but it can happen when
3008 deliver_drop_privilege is set.
3010 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3011 it now includes up to 150 characters of the unexpected data in the log
3014 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3015 and building data strings. The size of both of these buffers was 10 000
3016 bytes - far larger than anybody would *ever* want, thought I. Needless to
3017 say, somebody hit the limit. I have increased the maximum line length to
3018 20 000 and the maximum data length of concatenated lines to 100 000. I have
3019 also fixed two bugs, because there was no checking on these buffers. Tsk,
3020 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3021 buffer is too small.
3023 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3024 lsearch lookups. Now it does.
3026 46. When parsing a route_list item in a manualroute router, a fixed-length
3027 buffer was used for the list of hosts. I made this 1024 bytes long,
3028 thinking that nobody would ever have a list of hosts that long. Wrong.
3029 Somebody had a whole pile of complicated expansion conditions, and the
3030 string was silently truncated, leading to an expansion error. It turns out
3031 that it is easier to change to an unlimited length (owing to other changes
3032 that have happened since this code was originally written) than to build
3033 structure for giving a limitation error. The length of the item that
3034 expands into the list of hosts is now unlimited.
3036 47. The lsearch lookup could not handle data where the length of text line was
3037 more than 4095 characters. Such lines were truncated, leading to shortened
3038 data being returned. It should now handle lines of any length.
3040 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3041 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3044 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3045 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3046 no longer happens in comment lines. A list of replacements is now placed
3047 at the head of all of the source files, except those whose only change is
3048 to replace PERL_COMMAND in the very first #! line.
3050 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3051 messages on the queue, with a bottom-up merge sort, using code contributed
3052 by Michael Haardt. This should make operations like -bp somewhat faster on
3053 large queues. It won't affect queue runners, except when queue_run_in_order
3056 51. Installed eximstats 1.31 in the distribution.
3058 52. Added support for SRV lookups to the dnslookup router.
3060 53. If an ACL referred to $message_body or $message_body_end, the value was not
3061 reset for any messages that followed in the same SMTP session.
3063 54. The store-handling optimization for building very long strings was not
3064 differentiating between the different store pools. I don't think this
3065 actually made any difference in practice, but I've tidied it.
3067 55. While running the routers to verify a sender address, $sender_address
3068 was still set to the sender address. This is wrong, because when routing to
3069 send a bounce to the sender, it would be empty. Therefore, I have changed
3070 it so that, while verifying a sender address, $sender_address is set to <>.
3071 (There is no change to what happens when verifying a recipient address.)
3073 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3074 target A or AAAA records (if not already returned) without resetting the
3075 qualify_single or search_parents options of the DNS resolver. These are
3076 inappropriate in this case because the targets of MX and SRV records must
3077 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3078 target that, when qualified, matched something in the local domain. These
3079 two options are now turned off when doing these lookups.
3081 57. It seems that at least some releases of Reiserfs (which does not have the
3082 concept of a fixed number of inodes) returns zero and not -1 for the
3083 number of available inodes. This interacted badly with check_spool_inodes,
3084 which assumed that -1 was the "no such thing" setting. What I have done is
3085 to check that the total number of inodes is greater than zero before doing
3086 the test of how many are available.
3088 58. When a "warn" ACL statement has a log_message modifier, the message is
3089 remembered, and not repeated. This is to avoid a lot of repetition when a
3090 message has many recipients that cause the same warning to be written.
3091 Howewer, Exim was preserving the list of already written lines for an
3092 entire SMTP session, which doesn't seem right. The memory is now reset if a
3093 new message is started.
3095 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3096 debugging output unless log_rewrite was also set.
3098 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3099 of (int)(handle) when we know that handle contains (void *)(-1).
3101 61. The Exim daemon panic-logs an error return when it closes the incoming
3102 connection. However "connection reset by peer" seems to be common, and
3103 isn't really an error worthy of noting specially, so that particular error
3106 62. When Exim is trying to find all the local interfaces, it used to panic and
3107 die if the ioctl to get the interface flags failed. However, it seems that
3108 on at least one OS (Solaris 9) it is possible to have an interface that is
3109 included in the list of interfaces, but for which you get a failure error
3110 for this call. This happens when the interface is not "plumbed" into a
3111 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3112 failure of the "get flags" call assumes that the interface is down.
3114 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3115 makes life easier for people who are doing arithmetic on fields extracted
3116 from dates, where you often get leading zeros that should not be
3117 interpreted as octal.
3119 64. Added qualify_domain to the redirect router, to override the global
3122 65. If a pathologically long header line contained very many addresses (the
3123 report of this problem mentioned 10 000) and each of them was rewritten,
3124 Exim could use up a very large amount of memory. (It kept on making new
3125 copies of the header line as it rewrote, and never released the old ones.)
3126 At the expense of a bit more processing, the header rewriting function has
3127 been changed so that it no longer eats memory in this way.
3129 66. The generation of the Received: header has been moved from the time that a
3130 message starts to be received, to the time that it finishes. The timestamp
3131 in the Received: header should now be very close to that of the <= log
3132 line. There are two side-effects of this change:
3134 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3135 logged header lines no longer include the local Received: line, because
3136 it has not yet been created. The same applies to a copy of the message
3137 that is returned to a non-SMTP sender when a message is rejected.
3139 (b) When a filter file is tested using -bf, no additional Received: header
3140 is added to the test message. After some thought, I decided that this
3143 This change does not affect the value of $received_for. It is still set
3144 after address rewriting, but before local_scan() is called.
3146 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3148 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3149 gave an unhelpful panic error message, and a defer error. I have managed to
3150 change this behaviour so that it now rejects any supplied certificate,
3151 which seems right, as the list of acceptable certificates is empty.
3153 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3154 gave an unhelpful defer error. I have not managed to make this reject any
3155 supplied certificates, but the error message it gives is "no certificate
3156 supplied", which is not helpful.
3158 70. exigrep's output now also includes lines that are not associated with any
3159 message, but which match the given pattern. Implemented by a patch from
3160 Martin Sluka, which also tidied up the Perl a bit.
3162 71. Recipient callout verification, like sender verification, was using <> in
3163 the MAIL FROM command. This isn't really the right thing, since the actual
3164 sender may affect whether the remote host accepts the recipient or not. I
3165 have changed it to use the actual sender in the callout; this means that
3166 the cache record is now keyed on a recipient/sender pair, not just the
3167 recipient address. There doesn't seem to be a real danger of callout loops,
3168 since a callout by the remote host to check the sender would use <>.
3169 [SEE ABOVE: changed after hitting problems.]
3171 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3172 temporary errors. However, in the case of such a code being given after
3173 the end of a data transmission (i.e. after ".") Exim was failing to write
3174 a retry record for the message. (Yes, there was some broken host that was
3175 actually sending 8xx at this point.)
3177 73. An unknown lookup type in a host list could cause Exim to panic-die when
3178 the list was checked. (An example that provoked this was putting <; in the
3179 middle of a list instead of at the start.) If this happened during a DATA
3180 ACL check, a -D file could be left lying around. This kind of configuration
3181 error no longer causes Exim to die; instead it causes a defer errror. The
3182 incident is still logged to the main and panic logs.
3184 74. Buglet left over from Exim 3 conversion. The message "too many messages
3185 in one connection" was written to the rejectlog but not the mainlog, except
3186 when address rewriting (yes!) was being logged.
3188 75. Added write_rejectlog option.
3190 76. When a system filter was run not as root (that is, when system_filter_user
3191 was set), the values of the $n variables were not being returned to the
3192 main process; thus, they were not subsequently available in the $sn
3195 77. Added +return_path_on_delivery log selector.
3197 78. A connection timeout was being treated differently from recipients deferred
3198 when testing hosts_max_try with a message that was older than the host's
3199 retry timeout. (The host should not be counted, thus allowing all hosts to
3200 be tried at least once before bouncing.) This may have been the cause of an
3201 occasionally reported bug whereby a message would remain on the queue
3202 longer than the retry timeout, but would be bounced if a delivery was
3203 forced. I say "may" because I never totally pinned down the problem;
3204 setting up timeout/retry tests is difficult. See also the next item.
3206 79. The ultimate address timeout was not being applied to errors that involved
3207 a combination of host plus message (for example, a timeout on a MAIL
3208 command). When an address resolved to a number of possible hosts, and they
3209 were not all tried for each delivery (e.g. because of hosts_max_try), a
3210 message could remain on the queue longer than the retry timeout.
3212 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3215 81. Fixed an obscure SMTP outgoing bug which required at least the following
3216 conditions: (a) there was another message waiting for the same server;
3217 (b) the server returned 5xx to all RCPT commands in the first message so
3218 that the message was not completed; (c) the server dropped the connection
3219 or gave a negative response to the RSET that Exim sends to abort the
3220 transaction. The observed case was a dropped connection after DATA that had
3221 been sent in pipelining mode. That is, the server had advertised PIPELINING
3222 but was not implementing it correctly. The effect of the bug was incorrect
3223 behaviour, such as trying another host, and this could lead to a crash.
3229 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3230 and daemon.c were passed as pointers to ints; they should have been
3231 pointers to socklen_t variables (which are typically unsigned ints).
3233 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3236 3. Fixed a really odd bug that affected only the testing scheme; patching a
3237 certain fixed string in the binary changed the value of another string that
3238 happened to be identical to the end of the original first string.
3240 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3241 name", it returns that address as the IP address. On some operating
3242 systems (e.g. Solaris), it also passes back the IP address string as the
3243 "host name". However, on others (e.g. Linux), it passes back an empty
3244 string. Exim wasn't checking for this, and was changing the host name to an
3245 empty string, assuming it had been canonicized.
3247 5. Although rare, it is permitted to have more than one PTR record for a given
3248 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3249 all the names associated with an address, because they do in Solaris.
3250 However, it seems that they do not in Linux for data that comes from the
3251 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3252 I found this out when I moved to a new Linux workstation and tried to run
3253 the Exim test suite.
3255 To get round this problem I have changed the code so that it now does its
3256 own call to the DNS to look up PTR records when searching for a host name.
3257 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3258 addresses that are only in /etc/hosts are still found.
3260 This behaviour is, however, controlled by an option called host_lookup_
3261 order, which defaults to "bydns:byaddr". If people want to use the other
3262 order, or indeed, just use one or the other means of lookup, they can
3263 specify it in this variable.
3265 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3266 some operating systems, this comes back from gethostbyaddr() as an empty
3267 string, and this is what Exim used to test for. However, it seems that in
3268 other systems, "." is yielded. Exim now tests for this case too.
3270 7. The values of check_spool_space and check_log_space are now held internally
3271 as a number of kilobytes instead of an absolute number of bytes. If a
3272 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3273 kilobyte. This means that much larger values can be stored.
3275 8. Exim monitor: an attempt to get the action menu when not actually pointing
3276 at a message produces an empty menu entitled "No message selected". This
3277 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3278 no entries in it ("Shell widget menu has zero width and/or height"). So I
3279 have added a single, blank menu entry in this case.
3281 9. Added ${quote_local_part.
3283 10. MIME decoding is now applied to the contents of Subject: header lines when
3286 11. Now that a reference to $sender_host_address automatically causes a reverse
3287 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3288 host lookup before query-style lookups in lists that might use this
3289 variable. This has therefore been abolished, and the "net-" prefix is no
3290 longer necessary for query-style lookups.
3292 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3293 have been a typo for LFLAGS, so it has been changed.
3295 13. The install script calls Exim with "-C /dev/null" in order to find the
3296 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3297 to be output. Howeve, since Exim outputs its version number before the
3298 error, it didn't break the script. It just looked ugly. I fixed this by
3299 always allowing "-C /dev/null" if the caller is root.
3301 14. Ignore overlarge ACL variable number when reading spool file - insurance
3302 against a later release with more variables having written the file.
3304 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3305 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3306 this, as well as the form without the "IPv6" on the front.
3308 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3309 OS/Makefile-Darwin file.
3311 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3312 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3314 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3315 variable. However, some OS use ints for inodes. I've added cast to long int
3316 to get rid of the compiler warning.
3318 19. I had forgotten to lock out "/../" in configuration file names when
3319 ALT_CONFIG_PREFIX was set.
3321 20. Routers used for verification do not need to specify transports. However,
3322 if such a router generated a host list, and callout was configured, Exim
3323 crashed, because it could not find a port number from the (non-existent)
3324 transport. It now assumes port 25 in this circumstance.
3326 21. Added the -t option to exigrep.
3328 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3329 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3330 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3331 them is set, LOOKUP_LSEARCH is forced.
3333 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3335 24. Added sender and host information to the "rejected by local_scan()" log
3336 line; previously there was no indication of these.
3338 25. Added .include_if_exists.
3340 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3341 Linux. It turns out that not all file systems support this. Apparently some
3342 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3343 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3344 not supported on the file descriptor, is now ignored when Exim is trying to
3345 sync a directory. This applies only to Linux.
3347 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3349 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3350 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3353 29. Added support for maildirsize files from supplied patch (modified a bit).
3355 30. The use of :fail: followed by an empty string could lead Exim to respond to
3356 sender verification failures with (e.g.):
3358 550 Verification failed for <xxx>
3359 550 Sender verify failed
3361 where the first response line was missing the '-' that indicates it is not
3362 the final line of the response.
3364 31. The loop for finding the name of the user that called Exim had a hardwired
3365 limit of 10; it now uses the value of finduser_retries, which is used for
3366 all other user lookups.
3368 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3371 33. Exim was neglecting to zero errno before one call of strtol() when
3372 expanding a string and expecting an integer value. On some systems this
3373 resulted in spurious "integer overflow" errors. Also, it was casting the
3374 result into an int without checking.
3376 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3377 did not work. The code looks as if it has *never* worked, though it appears
3378 to have been documented since at least releast 1.62. I have made it work.
3380 35. The "timeout_DNS" error in retry rules, also documented since at least
3381 1.62, also never worked. As it isn't clear exactly what this means, and
3382 clearly it isn't a major issue, I have abolished the feature by treating it
3383 as "timeout", and writing a warning to the main and panic logs.
3385 36. The display of retry rules for -brt wasn't always showing the error code
3388 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3389 timeout_connect_A, timeout_connect_MX.
3391 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3392 to the empty sender.
3394 39. The daemon was not analysing the content of -oX till after it had closed
3395 stderr and disconnected from the controlling terminal. This meant that any
3396 syntax errors were only noted on the panic log, and the return code from
3397 the command was 0. By re-arranging the code a little, I've made the
3398 decoding happen first, so such errors now appear on stderr, and the return
3399 code is 1. However, the actual setting up of the sockets still happens in
3400 the disconnected process, so errors there are still only recorded on the
3403 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3404 connections (as happens on some IP stacks) was logged at start up time as
3405 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3406 it from "IPv6 and IPv4", which means that two separate sockets are being
3409 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3410 says whether AF_INET or AF_INET6 was passed as an argument.
3412 42. Exiwhat output was messed up when time zones were included in log
3415 43. Exiwhat now gives more information about the daemon's listening ports,
3416 and whether -tls-on-connect was used.
3418 44. The "port" option of the smtp transport is now expanded.
3420 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3421 silently ignored. Now an error message is written to the main and panic
3424 46. There's a new ACL modifier called "logwrite" which writes to a log file
3425 as soon as it is encountered.
3427 47. Added $local_user_uid and $local_user_gid at routing time.
3429 48. Exim crashed when trying to verify a sender address that was being
3432 49. Exim was recognizing only a space character after ".include". It now also
3433 recognizes a tab character.
3435 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3436 extracting the relevant information from the specification. The man page no
3437 longer contains scrambled data for the -d option, and I've added a section
3438 at the front about calling Exim under different names.
3440 51. Added "extra_headers" argument to the "mail" command in filter files.
3442 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3445 53. Installed eximstats 1.29.
3447 54. Added transport_filter_timeout as a generic transport option.
3449 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3450 Cc: header lines. This was required by RFC 822, but it not required by RFC
3453 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3454 incoming messages that did not have them. Now it does so only if the
3455 message originates locally, that is, if there is no associated remote host
3456 address. When Resent- header lines are present, this applies to the Resent-
3457 lines rather than the non-Resent- lines.
3459 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3460 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3462 58. Messages for configuration errors now include the name of the main
3463 configuration file - useful now that there may be more than one file in a
3464 list (.included file names were always shown).
3466 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3467 for those rare installations that do not start the daemon as root or run it
3468 setuid root. I've cut out the call to initgroups() if the daemon is not
3471 60. The Exim user and group can now be bound into the binary as text strings
3472 that are looked up at the start of Exim's processing.
3474 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3476 62. Added $mailstore_basename variable.
3478 63. Installed patch to sieve.c from Michael Haardt.
3480 64. When Exim failed to open the panic log after failing to open the main log,
3481 the original message it was trying to log was written to stderr and debug
3482 output, but if they were not available (the usual case in production), it
3483 was lost. Now it is written to syslog before the two lines that record the
3484 failures to open the logs.
3486 65. Users' Exim filters run in subprocesses under the user's uid. It is
3487 possible for a "deliver" command or an alias in a "personal" command to
3488 provoke an address rewrite. If logging of address rewriting is configured,
3489 this fails because the process is not running as root or exim. There may be
3490 a better way of dealing with this, but for the moment (because 4.30 needs
3491 to be released), I have disabled address rewrite logging when running a
3492 filter in a non-root, non-exim process.
3498 1. The buildconfig auxiliary program wasn't quoting the value set for
3499 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3500 not defined. This bug was masked in 4.22 by the effect that was fixed in
3503 2. Some messages that were rejected after a message id was allocated were
3504 shown as "incomplete" by exigrep. It no longer does this for messages that
3505 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3507 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3508 have allow_domain_literals set, the ID did not get logged in the <= line.
3509 Domain literals are now always recognized in Message-ID: header lines.
3511 4. The first argument for a ${extract expansion item is the key name or field
3512 number. Leading and trailing spaces in this item were not being ignored,
3513 causing some misleading effects.
3515 5. When deliver_drop_privilege was set, single queue runner processes started
3516 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3517 same command in the process it spins off) were not dropping privilege.
3519 6. When the daemon running as "exim" started a queue runner, it always
3520 re-executed Exim in the spun-off process. This is a waste of effort when
3521 deliver_drop_privilege is set. The new process now just calls the
3522 queue-runner function directly.
3528 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3531 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3532 ignored. Though the use of -oP was forcing the writing of a pid file, it
3533 was always written to the default place.
3535 3. If the message "no IP address found for host xxxx" is generated during
3536 incoming verification, it is now followed by identification of the incoming
3537 connection (so you can more easily find what provoked it).
3539 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3541 5. Added some features to "harden" Exim a bit more against certain attacks:
3543 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3544 be put in Local/Makefile. This is like the never_users runtime option,
3545 but it cannot be overridden. The default setting is "root".
3547 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3548 prefix string with which any file named in a -C command line option
3551 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3552 is retained for -C and -D only if the caller of Exim is root. Without
3553 it, the exim user may also use -C and -D and retain privilege.
3555 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3556 command line option is disabled.
3558 6. Macro names set by the -D option must start with an upper case letter, just
3559 like macro names defined in the configuration file.
3561 7. Added "dereference=" facility to LDAP.
3563 8. Two instances of the typo "uknown" in the source files are fixed.
3565 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3566 the Configure-Makefile script screwed up while processing it.
3568 10. Incorporated PCRE 4.4.
3570 11. The SMTP synchronization check was not operating right at the start of an
3571 SMTP session. For example, it could not catch a HELO sent before the client
3572 waited for the greeting. There is now a check for outstanding input at the
3573 point when the greeting is written. Because of the duplex, asynchronous
3574 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3575 way, but not yet received, when the check is performed.
3577 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3578 on TCP/IP sockets, because this apparently causes some broken clients to
3581 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3582 unchanged) from the Cygwin maintainer.
3584 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3585 was supported instead of testing for mbx. Effectively a typo.
3587 15. The spa authenticator server code was not checking that the input it
3588 received was valid base64.
3590 16. The debug output line for the "set" modifier in ACLs was not showing the
3591 name of the variable that was being set.
3593 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3595 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3596 lookup on its own. Something else was needed to trigger the lookup. For
3597 example, a match in host_lookup or the need for a host name in a host list.
3598 Now, if $sender_host_name is referenced and the host name has not yet been
3599 looked up, a lookup is performed. If the lookup fails, the variable remains
3600 empty, and $host_lookup_failed is set to "1".
3602 19. Added "eqi" as a case-independent comparison operator.
3604 20. The saslauthd authentication condition could segfault if neither service
3605 nor realm was specified.
3607 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3608 error message that was logged was misleading, and incoming SMTP
3609 connections were dropped. The message is now more accurate, and temporary
3610 errors are given to SMTP connections.
3612 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3613 (except RSET) with a 421 error, until QUIT is received. However, it was
3614 failing to send a response to QUIT.
3616 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3617 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3618 ACL is now run after the helo_try_verify_hosts code.
3620 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3621 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3622 case-independent, so other case variants are also recognized). Apparently
3623 some systems use these upper case variants.
3625 25. If more than two messages were waiting for the same host, and a transport
3626 filter was specified for the transport, Exim sent two messages over the
3627 same TCP/IP connection, and then failed with "socket operation on non-
3628 socket" when it tried to send the third.
3630 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3632 27. The extern definition of crypt16() in expand.c was not being excluded when
3633 the OS had its own crypt16() function.
3635 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3636 as a preferred synonym for return_size_limit, both as an option and as an
3639 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3641 30. Changed the default configuration ACL to relax the local part checking rule
3642 for addresses that are not in any local domains. For these addresses,
3643 slashes and pipe symbols are allowed within local parts, but the sequence
3644 /../ is explicitly forbidden.
3646 31. SPA server authentication was not clearing the challenge buffer before
3649 32. log_message in a "warn" ACL statement was writing to the reject log as
3650 well as to the main log, which contradicts the documentation and doesn't
3651 seem right (because no rejection is happening). So I have stopped it.
3653 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3654 However, I am unable to do any testing of this.
3656 34. Fixed an infelicity in the appendfile transport. When checking directories
3657 for a mailbox, to see if any needed to be created, it was accidentally
3658 using path names with one or more superfluous leading slashes; tracing
3659 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3661 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3662 contents are added to the log line that is written for every discarded
3663 recipient. (Previously a log_message setting was ignored.)
3665 36. The ${quote: operator now quotes the string if it is empty.
3667 37. The install script runs exim in order to find its version number. If for
3668 some reason other than non-existence or emptiness, which it checks, it
3669 could not run './exim', it was installing it with an empty version number,
3670 i.e. as "exim-". This error state is now caught, and the installation is
3673 38. An argument was missing from the function that creates an error message
3674 when Exim fails to connect to the socket for saslauthd authentication.
3675 This could cause Exim to crash, or give a corrupted message.
3677 39. Added isip, isip4, and isip6 to ${if conditions.
3679 40. The ACL variables $acl_xx are now saved with the message, and can be
3680 accessed later in routers, transports, and filters.
3682 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3683 strings in the file are not string-expanded.
3685 42. If a MAIL command specified a SIZE value that was too large to fit into an
3686 int variable, the check against message_size_limit failed. Such values are
3687 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3688 day this will have to be increased, but I don't think I want to be around
3689 when emails are that large.
3696 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3697 iconv() is not standard in FreeBSD.
3699 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3700 IPv6 enabled. The observed symptom was a segmentation fault on return from
3701 the function os_common_find_running_interfaces() in src/os.c.
3703 3. In the check_special_case() function in daemon.c I had used "errno" as an
3704 argument name, which causes warnings on some systems. This was basically a
3705 typo, since it was named "eno" in the comments!
3707 4. The code that waits for the clock to tick (at a resolution of some fraction
3708 of a second) so as to ensure message-id uniqueness was always waiting for
3709 at least one whole tick, when it could have waited for less. [This is
3710 almost certainly not relevant at current processor speeds, where it is
3711 unlikely to ever wait at all. But we try to future-proof.]
3713 5. The function that sleeps for a time interval that includes fractions of a
3714 second contained a race. It did not block SIGALRM between setting the
3715 timer, and suspending (a couple of lines later). If the interval was short
3716 and the sigsuspend() was delayed until after it had expired, the suspension
3717 never ended. On busy systems this could lead to processes getting stuck for
3720 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3721 process, before it forks any delivery processes. The open lookup caching
3722 mechanism meant that the open file or database connection was passed into
3723 the delivery process. The problem was that delivery processes always tidy
3724 up cached lookup data. This could cause a problem for the next delivery
3725 process started by the queue runner, because the external queue runner
3726 process does not know about the closure. So the next delivery process
3727 still has data in the lookup cache. In the case of a file lookup, there was
3728 no problem because closing a file descriptor in a subprocess doesn't affect
3729 the parent. However, if the lookup was caching a connection to a database,
3730 the connection was closed, and the second delivery process was likely to
3731 see errors such as "PGSQL: query failed: server closed the connection
3732 unexpectedly". The problem has been fixed by closing all cached lookups
3733 in a queue runner before running a delivery process.
3735 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3736 seem to have the "const" qualifier which it has on other OS. I've
3739 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3740 *of the same type* (client or server) with the same public name that an
3741 error should be diagnosed.
3743 9. When Exim looked up a host name for an IP address, but failed to find the
3744 original IP address when looking up the host name (a safety check), it
3745 output the message "<ip address> does not match any IP for NULL", which was
3746 confusing, to say the least. The bug was that the host name should have
3747 appeared instead of "NULL".
3749 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3750 user that is built into the binary, and the -C or -D options is used, root
3751 privilege is dropped before the configuration file is read. In addition,
3752 logging is switched to stderr instead of the normal log files. If the
3753 configuration then re-defines the Exim user, the unprivileged environment
3754 is probably not what is expected, so Exim logs a panic warning message (but
3757 However, if deliver_drop_privilege is set, the unprivileged state may well
3758 be exactly what is intended, so the warning has been cut out in that case,
3759 and Exim is allowed to try to write to its normal log files.
3765 1. smtp_return_error_details was not giving details for temporary sender
3766 or receiver verification errors.
3768 2. Diagnose a configuration error if two authenticators have the same public
3771 3. Exim used not to create the message log file for a message until the first
3772 delivery attempt. This could be confusing when incoming messages were held
3773 for policy or load reasons. The message log file is now created at the time
3774 the message is received, and an initial "Received" line is written to it.
3776 4. The automatically generated man page for command line options had a minor
3777 bug that caused no ill effects; however, a more serious problem was that
3778 the procedure for building the man page automatically didn't always
3779 operate. Consequently, release 4.20 contains an out-of-date version. This
3780 shouldn't happen again.
3782 5. When building Exim with embedded Perl support, the script that builds the
3783 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3784 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3786 6. The freeze_tell option was not being used for messages that were frozen on
3787 arrival, either by an ACL or by local_scan().
3789 7. Added the smtp_incomplete_transaction log selector.
3791 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3792 was accepting AUTH without a new EHLO.
3794 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3795 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3798 10. Logging of TCP/IP connections (when configured) now happens in the main
3799 daemon process instead of the child process, so that the TCP/IP connection
3800 count is more accurate (but it can never be perfect).
3802 11. The use of "drop" in a nested ACL was not being handled correctly in the
3803 outer ACL. Now, if condition failure induced by the nested "drop" causes
3804 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3805 or "require"), the connection is dropped.
3807 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3808 that yield "discard" can now be used with an "accept" or a "discard" verb,
3809 but an error is generated for any others (because I can't see a useful way
3810 to define what should happen).
3812 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3813 are now processed in the same way as lines in the Exim configuration file.
3814 In particular, continuation lines are supported.
3816 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3818 15. Added -ti meaning -t -i.
3820 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3821 domains, and warn by logging if others are found.
3823 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3824 returned from GIFCONF when Exim is trying to find the list of interfaces on
3825 a host. The code in os.c has been modified to copy each ifreq to an aligned
3826 structure in all cases.
3828 Also, in some cases, the returned ifreq's were being copied to a 'struct
3829 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3830 means the last couple of bytes of an IPv6 address could be chopped if the
3831 ifreq contained only a normal sockaddr (14 bytes storage).
3833 18. Named domain lists were not supported in the hosts_treat_as_local option.
3834 An entry such as +xxxx was not recognized, and was treated as a literal
3837 19. Ensure that header lines added by a DATA ACL are included in the reject log
3838 if the ACL subsequently rejects the message.
3840 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3841 MD5 (which is deprecated).
3843 21. When testing a filter file using -bf, Exim was writing a message when it
3844 took the sender from a "From " line in the message, but it was not doing so
3845 when it took $return_path from a Return-Path: header line. It now does.
3847 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3848 with a valid header line field name (a series of printing characters
3849 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3851 23. Changed "disc" in the source to "disk" to conform to the documentation and
3852 the book and for uniformity.
3854 24. Ignore Sendmail's -Ooption=value command line item.
3856 25. When execve() failed while trying to run a command in a pipe transport,
3857 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3858 could be confused with a return value of 69 from the command itself. This
3859 has been changed to 127, the value the shell returns if it is asked to run
3860 a non-existent command. The wording for the related log line suggests a
3861 non-existent command as the problem.
3863 26. If received_header_text expands to an empty string, do not add a Received:
3864 header line to the message. (Well, it adds a token one on the spool, but
3865 marks it "old" so that it doesn't get used or transmitted.)
3867 27. Installed eximstats 1.28 (addition of -nt option).
3869 28. There was no check for failure on the call to getsockname() in the daemon
3870 code. This can fail if there is a shortage of resources on the system, with
3871 ENOMEM, for example. A temporary error is now given on failure.
3873 29. Contrary to the C standard, it seems that in some environments, the
3874 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3875 program. Exim now does this explicitly; it affects the formatting of
3876 timestamps using strftime().
3878 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3879 complaints. I've now initialized all the variables, to avoid this.
3881 32. Header lines added by a system filter were not being "seen" during
3882 transport-time rewrites.
3884 33. The info_callback() function passed to OpenSSL is set up with type void
3885 (*)(SSL *, int, int), as described somewhere. However, when calling the
3886 function (actually a macro) that sets it up, the type void(*)() is
3887 expected. I've put in a cast to prevent warnings from picky compilers.
3889 34. If a DNS black list lookup found a CNAME record, but there were no A
3890 records associated with the domain it pointed at, Exim crashed.
3892 35. If a DNS black list lookup returned more than one A record, Exim ignored
3893 all but the first. It now scans all returned addresses if a particular IP
3894 value is being sought. In this situation, the contents of the
3895 $dnslist_value variable are a list of all the addresses, separated by a
3898 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3899 to accept a host name and all its aliases if the forward lookup for any of
3900 them yielded the IP address of the incoming connection. Now it accepts only
3901 those names whose forward lookup yields the correct IP address. Any other
3902 names are discarded. This closes a loophole whereby a rogue DNS
3903 administrator could create reverse DNS records to break through a
3904 wildcarded host restriction in an ACL.
3906 37. If a user filter or a system filter that ran in a subprocess used any of
3907 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3908 the wrong values were passed to the pipe command ($thisaddress had the
3909 value of $0, $0 had the value of $1, etc). This bug was introduced by
3910 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3912 38. Improved the line breaking for long SMTP error messages from ACLs.
3913 Previously, if there was no break point between 40 and 75 characters, Exim
3914 left the rest of the message alone. Two changes have been made: (a) I've
3915 reduced the minimum length to 35 characters; (b) if it can't find a break
3916 point between 35 and 75 characters, it looks ahead and uses the first one
3917 that it finds. This may give the occasional overlong line, but at least the
3918 remaining text gets split now.
3920 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3921 file descriptors might be low, and that setting 1000 would always raise it.
3922 It turns out that in some environments, the limit is already over 1000 and
3923 that lowering it causes trouble. So now Exim takes care not to decrease it.
3925 40. When delivering a message, the value of $return_path is set to $sender_
3926 address at the start of routing (routers may change the value). By an
3927 oversight, this default was not being set up when an address was tested by
3928 -bt or -bv, which affected the outcome if any router or filter referred to
3931 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3932 log only when "message" or "log_message" are set. However, if one of the
3933 conditions was an address verification, or a call to a nested ACL, the
3934 messages generated by the underlying test were being passed through. This
3935 no longer happens. The underlying message is available in $acl_verify_
3936 message for both "message" and "log_message" expansions, so it can be
3937 passed through if needed.
3939 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3940 new expansion $bh_ to give the encoded byte string without charset
3941 translation. Translation happens only if iconv() is available; HAVE_ICONV
3942 indicates this at build time. HEADERS_CHARSET gives the charset to
3943 translate to; headers_charset can change it in the configuration, and
3944 "headers charset" can change it in an individual filter file.
3946 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3947 that creates RFC 2047 encoded "words" labels them as that charset instead
3948 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3949 expansion operator; (ii) when Exim creates a From: line for a local
3950 message; (iii) when a header line is rewritten to include a "phrase" part.
3952 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3953 buggy, causing it to skip the first lines of messages whose message ID
3954 ended in 'D'. This would not have bitten before Exim release 4.14, because
3955 message IDs were unlikely to end in 'D' before then. The effect was to have
3956 incorrect size information for certain domains.
3958 45. #include "config.h" was missing at the start of the crypt16.c module. This
3959 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3962 46. If there was a timeout during a "random" callout check, Exim treated it as
3963 a failure of the random address, and carried on sending RSET and the real
3964 address. If the delay was just some slowness somewhere, the response to the
3965 original RCPT would be taken as a response to RSET and so on, causing
3966 mayhem of various kinds.
3968 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3969 when I implemented it. It didn't allow for the fact that some option values
3970 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3971 the right test for positive values.
3973 48. Domain names in DNS records are case-independent. Exim always looks them up
3974 in lower case. Some resolvers return domain names in exactly the case they
3975 appear in the zone file, that is, they may contain uppercase letters. Not
3976 all resolvers do this - some return always lower case. Exim was treating a
3977 change of case by a resolver as a change of domain, similar to a widening
3978 of a domain abbreviation. This triggered its re-routing code and so it was
3979 trying to route what was effectively the same domain again. This normally
3980 caused routing to fail (because the router wouldn't handle the domain
3981 twice). Now Exim checks for this case specially, and just changes the
3982 casing of the domain that it ultimately uses when it transmits the message
3985 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3988 50. If a filter generated a file delivery with a non-absolute name (possible if
3989 no home directory exists for the router), the forbid_file option was not
3992 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3993 the existing equality matching.
3995 52. Exim was using ints instead of ino_t variables in some places where it was
3996 dealing with inode numbers.
3998 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3999 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4000 called TMPDIR, and if it finds it is different, it changes its value.
4002 54. The smtp_printf() function is now made available to local_scan() so
4003 additional output lines can be written before returning. There is also an
4004 smtp_fflush() function to enable the detection of a dropped connection.
4005 The variables smtp_input and smtp_batched_input are exported to
4008 55. Changed the default runtime configuration: the message "Unknown user"
4009 has been removed from the ACL, and instead placed on the localuser router,
4010 using the cannot_route_message feature. This means that any verification
4011 failures that generate their own messages won't get overridden. Similarly,
4012 the "Unrouteable address" message that was in the ACL for unverifiable
4013 relay addresses has also been removed.
4015 56. Added hosts_avoid_esmtp to the smtp transport.
4017 57. The exicyclog script was not checking for the esoteric option
4018 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4019 will work only if exicyclog is run under the appropriate euid.
4021 58. Following a discussion on the list, the rules by which Exim recognises line
4022 endings on incoming messages have been changed. The -dropcr and drop_cr
4023 options are now no-ops, retained only for backwards compatibility. The
4024 following line terminators are recognized: LF CRLF CR. However, special
4025 processing applies to CR:
4027 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4028 nor a local message in the state where . is a terminator.
4030 (ii) If a bare CR is encountered in a header line, an extra space is added
4031 after the line terminator so as not to end the header. The reasoning
4032 behind this is that bare CRs in header lines are most likely either
4033 to be mistakes, or people trying to play silly games.
4035 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4036 was being incorrectly given as 18 bytes larger than it should have been.
4037 This is a VOB (very old bug).
4039 60. This may never have affected anything current, but just in case it has:
4040 When the local host is found other than at the start of a list of hosts,
4041 the local host, those with the same MX, and any that follow, are discarded.
4042 When the list in question was part of a longer list of hosts, the following
4043 hosts (not currently being processed) were also being discarded. This no
4044 longer happens. I'm not sure if this situation could ever has previously
4047 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4050 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4051 header. This is something that is recommended in a new Internet Draft, and
4052 is something that is documented as being done by Sendmail. There are two
4053 possible values. For messages generated by the autoreply transport, Exim
4056 Auto-Submitted: auto-replied
4058 whereas for all other generated messages (e.g. bounces) it adds
4060 Auto-Submitted: auto-generated
4062 63. The "personal" condition in filters now includes a test for the
4063 Auto-Submitted: header. If it contains the string "auto-" the message it
4064 not considered personal.
4066 64. Added rcpt_include_affixes as a generic transport option.
4068 65. Added queue_only_override (default true).
4070 66. Added the syslog_duplication option.
4072 67. If what should have been the first header line of a message consisted of
4073 a space followed by a colon, Exim was mis-interpreting it as a header line.
4074 It isn't of course - it is syntactically invalid and should therefore be
4075 treated as the start of the message body. The misbehaviour could have
4076 caused a number of strange effects, including loss of data in subsequent
4077 header lines, and spool format errors.
4079 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4080 client host had authenticated. This control can now be exercised by an ACL
4081 for more flexibility.
4083 69. By default, callouts do not happen when testing with -bh. There is now a
4084 variant, -bhc, which does actually run the callout code, including
4085 consulting and updating the callout cache.
4087 70. Added support for saslauthd authentication, courtesy of Alexander
4090 71. If statvfs() failed on the spool or log directories while checking their
4091 size for availability, Exim confusingly gave the error "space shortage".
4092 Furthermore, in debugging mode it crashed with a floating point exception.
4093 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4094 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4095 serious problem, Exim now writes to the main and panic logs when this
4096 happens, with details of the failure. It then refuses to accept the
4097 incoming message, giving the message "spool directory problem" or "log
4098 directory problem" with a 421 code for SMTP messages.
4100 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4101 0, 1, and 2 exist, because some OS complain for execs without them (see
4102 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4103 descriptors. However, the code omitted to check that the open succeeded,
4104 causing mysterious errors if for some reason the permissions on /dev/null
4105 got screwed. Now Exim writes a message to the main and panic logs, and
4106 bombs out if it can't open /dev/null.
4108 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4109 interact so that it is all more flexible. It is supposed to remain
4110 backwards compatible. Also added extra_local_interfaces.
4112 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4113 to bomb out with an assertion failure - to the client this appears as a
4114 connection drop. This problem occurs in the part of the code that was taken
4115 from the Samba project. Fortunately, the assertion is in a very simple
4116 function, so I have fixed this by reproducing the function inline in the
4117 one place where it is called, and arranging for authentication to fail
4118 instead of killing the process with assert().
4120 75. The SPA client code was not working when the server requested OEM rather
4121 than Unicode encoding.
4123 76. Added code to make require_files with a specific uid setting more usable in
4124 the case where statting the file as root fails - usually a non-root-mounted
4125 NFS file system. When this happens and the failure is EACCES, Exim now
4126 forks a subprocess and does the per-uid checking as the relevant uid.
4128 77. Added process_log_path.
4130 78. If log_file_path was not explicitly set, a setting of check_log_space or
4131 check_log_inodes was ignored.
4133 79. If a space check for the spool or log partitions fails, the incident is now
4134 logged. Of course, in the latter case the data may get lost...
4136 80. Added the %p formatting code to string_format() so that it can be used to
4137 print addresses in debug_print(). Adjusted all the address printing in the
4138 debugging in store.c to use %p rather than %d.
4140 81. There was a concern that a line of code in smtp_in.c could overflow a
4141 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4142 initially expressed, the concern was not well-founded, because trailing
4143 spaces are removed early. However, if the trailing spaces were followed by
4144 a NULL, they did not get removed, so the overflow was possible. Two fixes
4147 (a) I re-wrote the offending code in a cleaner fashion.
4148 (b) If an incoming SMTP command contains a NULL character, it is rejected
4151 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4152 runs initgroups(), so that if the Exim user is in any additional groups,
4153 they will be used during message reception.
4159 The change log for 4.20 and earlier releases has been archived.