1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.391 2006/09/18 11:06:20 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
14 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
26 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
29 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
34 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
37 PH/05 Applied Nico Efrurth's refactoring patch to tidy up spool_mbox.c.
39 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
45 SC/01 Use a glob alias rather than an array ref in eximstats generated
46 parser. This improves both readability and performance.
48 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
49 Don't display local sender or destination tables in eximstats unless
50 there is data to show.
51 Added average volumes into the eximstats top table text output.
53 SC/03 Collect data on the number of addresses (recipients) as well
54 as the number of messages in eximstats.
56 TF/01 Correct an error in the documentation for the redirect router. Exim
57 does (usually) call initgroups() when daemonizing.
59 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
60 with consistent privilege compared to when running as a daemon.
62 TF/03 Note in the spec that $authenticated_id is not set for local
63 submissions from trusted users.
65 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
66 Thanks to Dean Brooks <dean@iglou.com> for the patch.
68 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
69 by adding some example configuration directives to the default
70 configuration file. A little bit of work is required to uncomment the
71 directives and define how usernames and passwords are checked, but
72 there is now a framework to start from.
74 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
75 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
76 without this. I don't know how relevant this is to other LDAP libraries.
78 PH/02 Add the verb name to the "unknown ACL verb" error.
80 PH/03 Magnus Holmgren's patch for filter_prepend_home.
82 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
84 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
85 directory not expanded when it should be if an expanded home directory
86 was set for the address (which is overridden by the transport).
88 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
91 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
92 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
93 because it is too late at that time, and has no effect.
95 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
96 security issue with \' (bugzilla #107). I could not use the
97 PQescapeStringConn() function, because it needs a PGconn value as one of
100 PH/08 When testing addresses using -bt, indicate those final addresses that
101 are duplicates that would not cause an additional delivery. At least one
102 person was confused, thinking that -bt output corresponded to deliveries.
103 (Suppressing duplicates isn't a good idea as you lose the information
104 about possibly different redirections that led to the duplicates.)
106 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
107 systems where poll() doesn't work, in particular OS X.
109 PH/10 Added more information to debugging output for retry time not reached.
111 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
112 operations in malware.c.
114 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
117 PH/13 If write_rejectlog was set false when logging was sent to syslog with
118 syslog_duplication set false, log lines that would normally be written
119 both the the main log and to the reject log were not written to syslog at
122 PH/14 In the default configuration, change the use of "message" in ACL warn
123 statements to "add_header".
125 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
126 not followed by a command (e.g. "seen endif").
128 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
129 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
132 PH/17 Added extra conditions to the default value of delay_warning_condition
136 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
137 { match{$h_precedence:}{(?i)bulk|list|junk} } \
138 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
141 The Auto-Submitted: and various List- headers are standardised, whereas I
142 don't think Precedence: ever was.
144 PH/18 Refactored debugging code in route_finduser() to show more information,
145 in particular, the error code if getpwnam() issues one.
147 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
148 This is apparently needed in addition to the PH/07 change above to avoid
149 any possible encoding problems.
151 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
152 but not after initializing Perl.
154 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
155 output them only if debugging. By default they are written stderr,
156 apparently, which is not desirable.
158 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
161 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
164 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
166 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
167 authenticated or an ident call has been made. Suppress the default
168 values for $authenticated_id and $authenticated_sender (but permit -oMai
169 and -oMas) when testing with -bh.
171 PH/24 Re-jigged the order of the tests in the default configuration so that the
172 tests for valid domains and recipients precede the DNS black list and CSA
173 tests, on the grounds that those ones are more expensive.
175 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
176 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
177 This bug exists in every version of Exim that I still have, right back to
180 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
181 However, an attempt to turn on case-sensitivity in a regex key by
182 including (?-i) didn't work because the subject string was already
183 lowercased, and the effects were non-intuitive. It turns out that a
184 one-line patch can be used to allow (?-i) to work as expected.
190 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
191 other effects) broke the use of negated acl sub-conditions.
193 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
196 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
197 "Deny" causes Exim to reject the incoming connection with a 554 error.
198 Unfortunately, if there is a major crisis, such as a disk failure,
199 tcp-wrappers gives "deny", whereas what one would like would be some
200 kind of temporary error. A kludge has been added to help with this.
201 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
202 554 error is used if errno is still zero or contains ENOENT (which occurs
203 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
206 PH/03 Add -lutil to the default FreeBSD LIBS setting.
208 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
209 errors. Otherwise a message that provokes a temporary error (when other
210 messages do not) can cause a whole host to time out.
212 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
213 addresses were routed directly to files or pipes from a redirect router.
214 File deliveries just didn't batch; pipe deliveries might have suffered
217 PH/06 A failure to get a lock for a hints database would erroneously always say
218 "Failed to get write lock", even when it was really a read lock.
220 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
221 of 0600. This has been changed to use the value of the lockfile_mode
222 option (which defaults to 0600).
224 PH/08 Applied small patch from the Sieve maintainer.
226 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
227 folder from quota calculations, a direct delivery into this folder messed
228 up the contents of the maildirsize file. This was because the regex was
229 used only to exclude .Trash (or whatever) when the size of the mailbox
230 was calculated. There was no check that a delivery was happening into an
231 excluded directory. This bug has been fixed by ignoring all quota
232 processing for deliveries into excluded directories.
234 PH/10 Added the maildirfolder_create_regex option to appendfile.
240 PH/01 The code for finding all the local interface addresses on a FreeBSD
241 system running IPv6 was broken. This may well have applied to all BSD
242 systems, as well as to others that have similar system calls. The broken
243 code found IPv4 interfaces correctly, but gave incorrect values for the
244 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
245 that it would not match correctly against @[] and not recognize the IPv6
248 PH/02 The ipliteral router was not recognizing addresses of the form user@
249 [ipv6:....] because it didn't know about the "ipv6:" prefix.
251 PH/03 Added disable_ipv6.
253 PH/04 Changed $reply_address to use the raw form of the headers instead of the
254 decoded form, because it is most often used to construct To: headers
255 lines in autoreplies, and the decoded form may well be syntactically
256 invalid. However, $reply_address has leading white space removed, and all
257 newlines turned into spaces so that the autoreply transport does not
260 PH/05 If group was specified without a user on a router, and no group or user
261 was specified on a transport, the group from the router was ignored.
263 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
264 for visible compile-time settings that can be used to change these
265 numbers, for those that want even more. Backwards compatibility with old
266 spool files has been maintained. However, going back to a previous Exim
267 release will lost any variables that are in spool files.
269 PH/07 Two small changes when running in the test harness: increase delay when
270 passing a TCP/IP connection to a new process, in case the original
271 process has to generate a bounce, and remove special handling of
272 127.0.0.2 (sic), which is no longer necessary.
274 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
275 be the same on different OS.
277 PH/09 Moved a debug statement in filter processing to avoid a race problem when
280 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
281 whether --show-vars was specified or not
283 JJ/02 exipick: Added support for new ACL variable spool format introduced
286 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
287 syntactically invalid From: or Reply-to: line, and a filter used this to
288 generate an autoreply, and therefore failed to obtain an address for the
289 autoreply, Exim could try to deliver to a non-existent relative file
290 name, causing unrelated and misleading errors. What now happens is that
291 it logs this as a hard delivery error, but does not attempt to create a
294 PH/11 The exinext utility has a -C option for testing purposes, but although
295 the given file was scanned by exinext itself; it wasn't being passed on
296 when Exim was called.
298 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
299 an end-of-file indication when reading a command response.
301 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
302 compiled. In many other places in Exim, IPv6 addresses are always
303 recognized, so I have changed this. It also means that IPv4 domain
304 literals of the form [IPV4:n.n.n.n] are now always recognized.
306 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
307 used if the router is not running as root, for example, when verifying at
308 ACL time, or when using -bh. The debugging output from this situation was
309 non-existent - all you got was a failure to exec. I have made two
312 (a) Failures to set uid/gid, the current directory, or a process leader
313 in a subprocess such as that created by queryprogram now generate
314 suitable debugging ouput when -d is set.
316 (b) The queryprogram router detects when it is not running as root,
317 outputs suitable debugging information if -d is set, and then runs
318 the subprocess without attempting to change uid/gid.
320 PH/15 Minor change to Makefile for building test_host (undocumented testing
323 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
324 additional section of a DNS packet that returns MX or SRV records.
325 Instead, it always explicitly searches for A/AAAA records. This avoids
326 major problems that occur when a DNS server includes only records of one
327 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
328 fixed another bug: if SRV records were looked up and the corresponding
329 address records were *not* found in the additional section, the port
330 values from the SRV records were lost.
332 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
333 using the correct key (the original address) when searching the retry
334 rules in order to find which one to use for generating the retry hint.
336 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
337 adding the default one. Similarly, if it contains a Reply-To: header, the
338 errors_reply_to option, if set, is not used.
340 PH/19 When calculating a retry time, Exim used to measure the "time since
341 failure" by looking at the "first failed" field in the retry record. Now
342 it does not use this if it is later than than the arrival time of the
343 message. Instead it uses the arrival time. This makes for better
344 behaviour in cases where some deliveries succeed, thus re-setting the
345 "first failed" field. An example is a quota failure for a huge message
346 when small messages continue to be delivered. Without this change, the
347 "time since failure" will always be short, possible causing more frequent
348 delivery attempts for the huge message than are intended.
349 [Note: This change was subsequently modified - see PH/04 for 4.62.]
351 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
352 $1, $2, $3) because the numerical variables can be reset during some
353 expansion items (e.g. "match"), thereby losing the authentication data.
355 PH/21 Make -bV show the size of off_t variables so that the test suite can
356 decide whether to run tests for quotas > 2G.
358 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
359 mailbox_size, and mailbox_filecount in the appendfile transport. If a
360 filecount value is greater than 2G or if a quota value is greater than 2G
361 on a system where the size of off_t is not greater than 4, a panic error
364 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
365 never match. The debug and -bh output now contains an explicit error
366 message indicating a malformed IPv4 address or mask.
368 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
369 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
372 PH/25 Do not write to syslog when running in the test harness. The only
373 occasion when this arises is a failure to open the main or panic logs
374 (for which there is an explicit test).
376 PH/26 Added the /no_tell option to "control=freeze".
378 PH/27 If a host name lookup failed very early in a connection, for example, if
379 the IP address matched host_lookup and the reverse lookup yielded a name
380 that did not have a forward lookup, an error message of the form "no IP
381 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
382 could be logged. Now it outputs the IP address instead of "NULL".
384 PH/28 An enabling patch from MH: add new function child_open_exim2() which
385 allows the sender and the authenticated sender to be set when
386 submitting a message from within Exim. Since child_open_exim() is
387 documented for local_scan(), the new function should be too.
389 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
390 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
391 results in an empty string is now treated as unset.
393 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
395 PH/31 Added sender_verify_fail logging option.
397 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
398 needed by RFC 822 but not by RFC 2822 was commented out. I have now
399 tidied the source and removed it altogether.
401 PH/33 When a queue run was abandoned because the load average was too high, a
402 log line was always written; now it is written only if the queue_run log
403 selector is set. In addition, the log line for abandonment now contains
404 information about the queue run such as the pid. This is always present
405 in "start" and "stop" lines but was omitted from the "abandon" line.
407 PH/34 Omit spaces between a header name and the colon in the error message that
408 is given when verify = headers_syntax fails (if there are lots of them,
409 the message gets confusing).
411 PH/35 Change the default for dns_check_names_pattern to allow slashes within
412 names, as there are now some PTR records that contain slashes. This check
413 is only to protect against broken name servers that fall over on strange
414 characters, so the fact that it applies to all lookups doesn't matter.
416 PH/36 Now that the new test suite is complete, we can remove some of the
417 special code in Exim that was needed for the old test suite. For example,
418 sorting DNS records because real resolvers return them in an arbitrary
419 order. The new test suite's fake resolver always returns records in the
422 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
423 bounces) except when queue_only is set, to avoid logging races between
424 the different processes.
426 PH/38 Panic-die if .include specifies a non-absolute path.
428 PH/39 A tweak to the "H" retry rule from its user.
430 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
431 a label. They prevented compilation on older perls.
433 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
434 a warning to be raised on newish perls.
436 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
437 on queue. Changes to match documented behaviour of showing count of
438 messages matching specified criteria.
440 PH/40 Changed the default ident timeout from 30s to 5s.
442 PH/41 Added support for the use of login_cap features, on those BSD systems
443 that have them, for controlling the resources used by pipe deliveries.
445 PH/42 The content-scanning code uses fopen() to create files in which to put
446 message data. Previously it was not paying any attention to the mode of
447 the files. Exim runs with umask(0) because the rest of the code creates
448 files with open(), and sets the required mode explicitly. Thus, these
449 files were ending up world-writeable. This was not a big issue, because,
450 being within the spool directory, they were not world-accessible. I have
451 created a function called modefopen, which takes an additional mode
452 argument. It sets umask(777), creates the file, chmods it to the required
453 mode, then resets the umask. All the relevant calls to fopen() in the
454 content scanning code have been changed to use this function.
456 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
457 to 24 hours. This avoids potential overflow problems when processing G
458 and H retry rules. I suspect nobody ever tinkers with this value.
460 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
462 PH/45 When the plaintext authenticator is running as a client, the server's
463 challenges are checked to ensure they are valid base64 strings. By
464 default, the authentication attempt is cancelled if an invalid string is
465 received. Setting client_ignore_invalid_base64 true ignores these errors.
466 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
467 they are received. Thus, the responses can be made to depend on the
468 challenges. If an invalid string is ignored, an empty string is placed in
471 PH/46 Messages that are created by the autoreply transport now contains a
472 References: header, in accordance with RFCs 2822 and 3834.
474 PH/47 Added authenticated_sender_force to the smtp transport.
476 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
478 PH/49 Installed latest patch from the Sieve maintainer.
480 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
481 was also set, the appendfile transport was unnecessarily scanning a
482 directory of message files (e.g. for maildir delivery) to find the count
483 of files (along with the size), even though it did not need this
484 information. It now does the scan only if it needs to find either the
485 size of the count of files.
487 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
489 PH/52 Two bugs concerned with error handling when the smtp transport is
492 (i) Exim was not creating retry information for temporary errors given
493 for individual recipients after the DATA command when the smtp transport
494 was used in LMTP mode. This meant that they could be retried too
495 frequently, and not timed out correctly.
497 (ii) Exim was setting the flag that allows error details to be returned
498 for LMTP errors on RCPT commands, but not for LMTP errors for individual
499 recipients that were returned after the DATA command.
501 PH/53 This is related to PH/52, but is more general: for any failing address,
502 when detailed error information was permitted to be returned to the
503 sender, but the error was temporary, then after the final timeout, only
504 "retry timeout exceeded" was returned. Now it returns the full error as
505 well as "retry timeout exceeded".
507 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
508 do this, and (what is worse) MTAs that accept it.
510 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
511 will now be deprecated.
513 PH/56 New os.c-cygwin from the Cygwin maintainer.
515 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
516 formats (previously only available in exim formats via -bpr, -bpru,
517 and -bpra. Now also available in native and exiqgrep formats)
519 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
520 with very large, slow to parse queues
522 JJ/08 exipick: added ! as generic prefix to negate any criteria format
524 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
526 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
527 responses to authentication challenges, though it was showing the
528 challenges; (ii) I've removed the CR characters from the debug output for
531 PH/58 Allow for the insertion of a newline as well as a space when a string
532 is turned into more than one encoded-word during RFC 2047 encoding. The
533 Sieve code now uses this.
535 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
536 data_4xx, lost_connection, tls_required.
538 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
539 message was being sent as an SMTP response.
541 PH/61 Add -l and -k options to exicyclog.
543 PH/62 When verifying, if an address was redirected to one new address, so that
544 verification continued, and the new address failed or deferred after
545 having set something in $address_data, the value of $address_data was not
546 passed back to the ACL. This was different to the case when no
547 redirection occurred. The value is now passed back in both cases.
549 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
550 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
551 use login_cap.h, so on its own it isn't the distinguishing feature. The
552 new name refers directly to the setclassresources() function.
554 PH/65 Added configuration files for NetBSD3.
556 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
558 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
559 is preferred over IPv4.
561 PH/68 The bounce_return_message and bounce_return_body options were not being
562 honoured for bounces generated during the reception of non-SMTP messages.
563 In particular, this applied to messages rejected by the ACL. This bug has
564 been fixed. However, if bounce_return_message is true and bounce_return_
565 body is false, the headers that are returned for a non-SMTP message
566 include only those that have been read before the error was detected.
567 (In the case of an ACL rejection, they have all been read.)
569 PH/69 The HTML version of the specification is now built in a directory called
570 spec_html instead of spec.html, because the latter looks like a path with
571 a MIME-type, and this confuses some software.
573 PH/70 Catch two compiler warnings in sieve.c.
575 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
576 function verify_get_ident() calls ip_connect() to connect a socket, but
577 if the "connect()" function timed out, ip_connect() used to close the
578 socket. However, verify_get_ident() also closes the socket later, and in
579 between Exim writes to the log, which may get opened at this point. When
580 the socket was closed in ip_connect(), the log could get the same file
581 descriptor number as the socket. This naturally causes chaos. The fix is
582 not to close the socket in ip_connect(); the socket should be closed by
583 the function that creates it. There was only one place in the code where
584 this was missing, in the iplookup router, which I don't think anybody now
585 uses, but I've fixed it anyway.
587 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
588 well as to direct DNS lookups. Otherwise the handling of names in host
589 lists is inconsistent and therefore confusing.
595 PH/01 Two changes to the default runtime configuration:
597 (1) Move the checks for relay_from_hosts and authenticated clients from
598 after to before the (commented out) DNS black list checks.
600 (2) Add control=submission to the relay_from_hosts and authenticated
601 clients checks, on the grounds that messages accepted by these
602 statements are most likely to be submissions.
604 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
606 (1) Generate an error if the third argument for the ${prvs expansion is
609 (2) Treat a missing third argument of ${prvscheck as if it were an empty
612 (3) Reset the variables that are obtained from the first argument of
613 ${prvscheck and used in the second argument before leaving the code,
614 because their memory is reclaimed, so using them afterwards may do
617 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
618 one (it's much easier than Tom thought :-).
620 (5) Because of (4), we can now allow for the use of $prvscheck_result
621 inside the third argument.
623 PH/03 For some reason, the default setting of PATH when running a command from
624 a pipe transport was just "/usr/bin". I have changed it to
627 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
628 anything to be listed in the output from -bV.
630 PH/05 When a filter generated an autoreply, the entire To: header line was
631 quoted in the delivery log line, like this:
633 => >A.N.Other <ano@some.domain> <original@ddress> ...
635 This has been changed so that it extracts the operative address. There
636 may be more than one such address. If so, they are comma-separated, like
639 => >ano@some.domain,ona@other.domain <original@ddress> ...
641 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
642 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
643 not being looked up in the rDNS to get a host name, Exim was showing the
644 IP address twice in Received: lines, even though the IP addresses were
645 identical. For example:
647 Received: from [1.2.3.4] (helo=[1.2.3.4])
649 However, if the real host name was known, it was omitting the HELO data
650 if it matched the actual IP address. This has been tidied up so that it
651 doesn't show the same IP address twice.
653 PH/07 When both +timestamp and +memory debugging was on, the value given by
654 $tod_xxx expansions could be wrong, because the tod_stamp() function was
655 called by the debug printing, thereby overwriting the timestamp buffer.
656 Debugging no longer uses the tod_stamp() function when +timestamp is set.
658 PH/08 When the original message was included in an autoreply transport, it
659 always said "this is a copy of the message, including all the headers",
660 even if body_only or headers_only was set. It now gives an appropriate
663 PH/09 Applied a patch from the Sieve maintainer which:
665 o fixes some comments
666 o adds the (disabled) notify extension core
667 o adds some debug output for the result of if/elsif tests
668 o points to the current vacation draft in the documentation
669 and documents the missing references header update
673 o fixes a bug in processing the envelope test (when testing
674 multiple envelope elements, the last element determinted the
677 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
678 Electronic Mail") by including:
680 Auto-submitted: auto-generated
682 in the messages that it generates (bounce messages and others, such as
683 warnings). In the case of bounce messages for non-SMTP mesages, there was
684 also a typo: it was using "Auto_submitted" (underscore instead of
685 hyphen). Since every message generated by Exim is necessarily in response
686 to another message, thes have all been changed to:
688 Auto-Submitted: auto-replied
690 in accordance with these statements in the RFC:
692 The auto-replied keyword:
694 - SHOULD be used on messages sent in direct response to another
695 message by an automatic process,
697 - MUST NOT be used on manually-generated messages,
699 - MAY be used on Delivery Status Notifications (DSNs) and Message
700 Disposition Notifications (MDNs),
702 - MUST NOT be used on messages generated by automatic or periodic
703 processes, except for messages which are automatic responses to
706 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
707 to the default Received: header definition.
709 PH/12 Added log selector acl_warn_skipped (default on).
711 PH/13 After a successful wildlsearch lookup, discard the values of numeric
712 variables because (a) they are in the wrong storage pool and (b) even if
713 they were copied, it wouldn't work properly because of the caching.
715 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
716 checking when decoding. Apparently there are clients that generate
717 overlong encoded strings. Why am I not surprised?
719 PH/15 If the first argument of "${if match_address" was not empty, but did not
720 contain an "@" character, Exim crashed. Now it writes a panic log message
721 and treats the condition as false.
723 PH/16 In autoreply, treat an empty string for "once" the same as unset.
725 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
726 extension "envelope-auth". The code is finished and in agreement with
727 other implementations, but there is no documentation so far and in fact,
728 nobody wrote the draft yet. This extension is currently #undef'ed, thus
729 not changing the active code.
731 Print executed "if" and "elsif" statements when debugging is used. This
732 helps a great deal to understand what a filter does.
734 Document more things not specified clearly in RFC3028. I had all this
735 sorted out, when out of a sudden new issues came to my mind. Oops."
737 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
740 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
741 canonical form (as documented). However, after a host name lookup from
742 the IP address, check_host() was doing a simple string comparison with
743 addresses acquired from the DNS when checking that the found name did
744 have the original IP as one of its addresses. Since any found IPv6
745 addresses are likely to be in abbreviated form, the comparison could
746 fail. Luckily, there already exists a function for doing the comparison
747 by converting both addresses to binary, so now that is used instead of
750 PH/20 There was another similar case to PH/19, when a complete host name was
751 given in a host list; looking up its IP address could give an abbreviated
752 form, whereas the current host's name might or might not be abbreviated.
753 The same fix has been applied.
759 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
760 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
763 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
764 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
766 PH/03 Typo: missing ".o" in src/pcre/Makefile.
768 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
769 header line, restrict the check to what is listed in RFCs 2369 and 2929.
770 Also, for "Auto-Submitted", treat anything other than "no" as
771 non-personal, in accordance with RFC 3834. (Previously it treated
772 anything starting "auto-" as non-personal.)
774 TF/01 The control=submission/name=... option had a problem with syntax
775 errors if the name included a slash character. The /name= option
776 now slurps the rest of the string, so it can include any characters
777 but it must come last in the list of options (after /sender_retain
780 PH/05 Some modifications to the interface to the fake nameserver for the new
788 TK/01 Added the "success_on_redirect" address verification option. See
789 NewStuff for rationale and an example.
791 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
793 PH/02 Patch to exigrep to allow it to work on syslog lines.
795 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
796 fread() to skip over the body file's header line, because in Cygwin the
797 header line is locked and is inaccessible.
799 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
800 co-exist for some time) to make it clear that it is the Exim ID that is
801 referenced, not the Message-ID: header line.
803 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
804 string_format() function, because snprintf() does not exist on all
807 PH/06 The use of forbid_filter_existstest now also locks out the use of the
808 ${stat: expansion item.
810 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
811 protocol synchronization error", to keep the pedants happy.
813 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
814 well as for IRIX systems, when gcc is being used. See the host.c source
817 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
819 PH/10 Named domain lists were not working if used in a queue_smtp_domains
822 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
823 transport and to the smtp transport in LMTP mode.
825 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
827 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
828 run a filter in a subprocess. This could lead to confusion in subsequent
829 lookups in the parent process. There should also be a search_tidyup() at
830 the end of the subprocess.
832 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
833 only if the host matched helo_try_verify_hosts, which caused the
834 verification to occur when the EHLO/HELO command was issued. The ACL just
835 tested the remembered result. Now, if a previous verification attempt has
836 not happened, "verify = helo" does it there and then.
838 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
840 TK/03 Fix log output including CR from clamd.
842 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
843 exist provoked a memory error which could cause a segfault.
845 PH/15 Installed PCRE 6.2
847 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
849 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
850 of the problem. Specifically, suggested +O2 rather than +O1 for the
853 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
855 PH/20 If a delivery was routed to a non-standard port by means of an SRV
856 record, the port was not correctly logged when the outgoing_port log
857 selector was set (it logged the transort's default port).
859 PH/21 Added support for host-specific ports to manualroute, queryprogram,
860 fallback_hosts, and "hosts" in the smtp transport.
862 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
863 host errors such as "Connection refused".
865 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
866 authentication with radiusclient 0.4.9:
868 - Error returned from rc_read_config was caught wrongly
869 - Username/password not passed on to radius server due to wrong length.
871 The presumption is that some radiusclient API changes for 4.51/PH/17
872 were not taken care of correctly. The code is still untested by me (my
873 Linux distribution still has 0.3.2 of radiusclient), but it was
874 contributed by a Radius user.
876 PH/24 When doing a callout, the value of $domain wasn't set correctly when
877 expanding the "port" option of the smtp transport.
879 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
880 while reading a MIME header. Thanks to Tom Hughes for a patch.
882 PH/24 Include config.h inside local_scan.h so that configuration settings are
885 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
886 that in an ACL for RCPT (for example), you can examine exactly what was
889 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
890 commands, but it was not correctly comparing the address with the actual
891 client host address. Thus, it would show the EHLO address in Received:
892 header lines when this was not necessary.
894 PH/27 Added the % operator to ${eval:}.
896 PH/28 Exim tries to create and chdir to its spool directory when it starts;
897 it should be ignoring failures (because with -C, for example, it has lost
898 privilege). It wasn't ignoring creation failures other than "already
901 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
904 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
905 because an input line was too long, either on its own, or by virtue of
906 too many continuations, the temporary file was not being removed, and the
907 return code was incorrect.
909 PH/31 Missing "BOOL" in function definition in filtertest.c.
911 PH/32 Applied Sieve patches from the maintainer.
913 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
915 PH/33 Added "verify = not_blind".
917 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
918 Local/Makefile (with some defaults set). These are used in built scripts
919 such as exicyclog, but they have never been used in the exim_install
920 script (though there are many overriding facilities there). I have
921 arranged that the exim_install script now takes note of these two
924 PH/35 Installed configuration files for Dragonfly.
926 PH/36 When a locally submitted message by a trusted user did not contain a
927 From: header, and the sender address was obtained from -f or from an SMTP
928 MAIL command, and the trusted user did not use -F to supply a sender
929 name, $originator_name was incorrectly used when constructing a From:
930 header. Furthermore, $originator_name was used for submission mode
931 messages from external hosts without From: headers in a similar way,
932 which is clearly wrong.
934 PH/37 Added control=suppress_local_fixups.
936 PH/38 When log_selector = +received_sender was set, and the addition of the
937 sender made the log line's construction buffer exactly full, or one byte
938 less than full, an overflow happened when the terminating "\n" was
941 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
942 when the result of a list match is failure because a DNS lookup failed.
944 PH/40 RM_COMMAND is now used in the building process.
946 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
947 the "build-* directories that it finds.
949 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
950 address in a domain literal was a prefix of an interface address.
952 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
953 when verifying a sender address, unless rewrite_headers is false.
955 PH/44 Wrote a long comment about why errors_to addresses are verified as
956 recipients, not senders.
958 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
959 the ratelimit ACL was added.
961 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
963 PH/46 Added extra information about PostgreSQL errors to the error string.
965 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
966 avoiding the need to install special zones in a real server. This is
967 backwards compatible; if it can't find the fake resolver, it drops back.
968 Thus, both old and new test suites can be run.
970 TF/02 Added util/ratelimit.pl
972 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
973 clock is set back in time.
975 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
976 Candler <B.Candler@pobox.com>.
978 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
979 OK for addresses that are the result of redirections.
981 PH/48 A number of further additions for the benefit of the new test suite,
982 including a fake gethostbyname() that interfaces to the fake DNS resolver
985 TF/06 The fix for widen_domains has also been applied to qualify_single and
986 search_parents which are the other dnslookup options that can cause
989 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
992 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
994 TF/07 Exim produced the error message "an SRV record indicated no SMTP
995 service" if it encountered an MX record with an empty target hostname.
996 The message is now "an MX or SRV record indicated no SMTP service".
998 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
999 if the DNS of the sending site is misconfigured. This is quite a
1000 common situation. This change restores the behaviour of treating a
1001 helo verification defer as a failure.
1003 PH/51 If self=fail was set on a router, the bounce message did not include the
1004 actual error message.
1010 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1012 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1013 command itself ended in error, the underlying message about the transport
1014 filter timeout was being overwritten with the pipe command error. Now the
1015 underlying error message should be appended to the second error message.
1017 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1019 PH/02 Reduce the amount of output that "make" produces by default. Full output
1020 can still be requested.
1022 PH/03 The warning log line about a condition test deferring for a "warn" verb
1023 was being output only once per connection, rather than after each
1024 occurrence (because it was using the same function as for successful
1025 "warn" verbs). This seems wrong, so I have changed it.
1027 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1028 it should not have, which might have caused a crash in the right
1029 circumstances, but probably never did.
1031 PH/04 Installed a modified version of Tony Finch's patch to make submission
1032 mode fix the return path as well as the Sender: header line, and to
1033 add a /name= option so that you can make the user's friendly name appear
1036 TF/03 Added the control = fakedefer ACL modifier.
1038 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1039 Mark Lowes for thorough testing.
1041 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1043 TK/03 Merged latest SRS patch from Miles Wilton.
1045 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1046 with the definition in sysexits.h (which is #included earlier).
1047 Fortunately, Exim does not actually use EX_OK. The code used to try to
1048 preserve the sysexits.h value, by assumimg that macro definitions were
1049 scanned for macro replacements. I have been disabused of this notion,
1050 so now the code just undefines EX_OK before #including unistd.h.
1052 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1053 in the smtp transport. When a block could not be written in a single
1054 write() function, the timeout was being re-applied to each part-write.
1055 This seems wrong - if the receiver was accepting one byte at a time it
1056 would take for ever. The timeout is now adjusted when this happens. It
1057 doesn't have to be particularly precise.
1059 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1060 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1062 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1063 without a domain if the check to <postmaster@domain> fails.
1065 SC/01 Eximstats: added -xls and the ability to specify output files
1066 (patch written by Frank Heydlauf).
1068 SC/02 Eximstats: use FileHandles for outputing results.
1070 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1072 SC/04 Eximstats: fixed display of large numbers with -nvr option
1074 SC/05 Eximstats: fixed merging of reports with empty tables.
1076 SC/06 Eximstats: added the -include_original_destination flag
1078 SC/07 Eximstats: removed tabs and trailing whitespace.
1080 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1082 TK/06 MBOX spool code: Add real "From " MBOX separator line
1083 so the .eml file is really in mbox format (even though
1084 most programs do not really care). Patch from Alex Miller.
1086 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1087 The latter is generated from $received_to and is only set if the
1088 message has one envelope recipient. SA can use these headers,
1089 obviously out-of-the-box. Patch from Alex Miller.
1091 PH/08 The ${def test on a variable was returning false if the variable's
1092 value was "0", contrary to what the specification has always said!
1093 The result should be true unless the variable is empty.
1095 PH/09 The syntax error of a character other than { following "${if
1096 def:variable_name" (after optional whitespace) was not being diagnosed.
1097 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1098 accidental colon was present, for example, could give incorrect results.
1100 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1101 result is used (not including appendfile, where other changes are about
1104 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1105 This involved changing a lot of size variables from int to off_t. It
1106 should work with maildirs and everything.
1108 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1109 spamd dying while we are connected to it.
1111 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1114 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1115 function for BDB 4.3.
1117 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1119 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1120 the sources for PCRE have been re-arranged and more files are now
1123 PH/15 The code I had for printing potentially long long variables in PH/11
1124 above was not the best (it lost precision). The length of off_t variables
1125 is now inspected at build time, and an appropriate printing format (%ld
1126 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1127 to be "long long int" or "long int". This is needed for the internal
1128 formatting function string_vformat().
1130 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1131 the configuration file to be ":syslog", then the script "guesses" where
1132 the logs files are, rather than using the compiled in default. In our
1133 case the guess is not the same as the compiled default, so the script
1134 suddenly stopped working when I started to use syslog. The patch checks
1135 to see if log_file_path is "". If so, it attempts to read it from exim
1136 with no configuration file to get the compiled in version, before it
1137 falls back to the previous guessing code."
1139 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1140 implementing BATV in an Exim configuration. See NewStuff for the gory
1143 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1144 Makefile that are specific to HP-UX.
1146 PH/18 If the "use_postmaster" option was set for a recipient callout together
1147 with the "random" option, the postmaster address was used as the MAIL
1148 FROM address for the random test, but not for the subsequent recipient
1149 test. It is now used for both.
1151 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1152 patch removes a few documentation additions to RFC 3028, because the
1153 latest draft now contains them. It adds the new en;ascii-case comparator
1154 and a new error check for 8bit text in MIME parts. Comparator and
1155 require names are now matched exactly. I enabled the subaddress
1156 extension, but it is not well tested yet (read: it works for me)."
1158 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1159 rework some of the code of TK/09 above to avoid the hardwired use of
1160 "%lld" and "long long". Replaced the call to snprintf() with a call to
1163 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1164 records point to non-existent hosts", "retry timeout exceeded", and
1165 "retry time not reached for any host after a long failure period".
1167 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1168 experimental DomainKeys support:
1170 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1171 (2) On an error such as an illegally used "control", the wrong name for
1172 the control was given.
1174 These problems did NOT occur unless DomainKeys support was compiled.
1176 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1178 PH/24 Added ${if match_ip condition.
1180 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1181 databases so that it will be absolutely obvious if a crash occurs in the
1182 DB library. This is a regular occurrence (often caused by mis-matched
1185 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1186 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1187 picked up on a user's system that detects such things. There doesn't seem
1188 to be a gcc warning option for this - only an attribute that has to be
1189 put on the function's prototype. It seems that in Fedora Core 4 they have
1190 set this on a number of new functions. No doubt there will be more in due
1193 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1194 specify a transport. However, if an address that was verified by such a
1195 router was the subject of a callout, Exim crashed because it tried to
1196 read the rcpt_include_affixes from the non-existent transport. Now it
1197 just assumes that the setting of that option is false. This bug was
1198 introduced by 4.51/PH/31.
1200 PH/28 Changed -d+all to exclude +memory, because that information is very
1201 rarely of interest, but it makes the output a lot bigger. People tend to
1202 do -d+all out of habit.
1204 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1205 code in os-type was giving problems when libc.so lives in lib64, like on
1208 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1209 aren't the modern standard, and it seems that some systems' include files
1210 don't always have them. Exim was already checking for some of the newer
1211 ones like T_AAAA, and defining it itself. I've added checks for all the
1212 record types that Exim uses.
1214 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1215 not automatically generating a new one, as it is supposed to. This
1216 prevented TLS from working. If the file did exist, but contained invalid
1217 data, a new version was generated, as expected. It was only the case of a
1218 non-existent file that was broken.
1220 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1221 with a change in libdomainkeys > 0.64.
1223 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1224 from DNS. If the selector record carries the flag, it now has
1225 precedence over the domain-wide flag.
1227 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1229 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1230 the use of a port name that isn't defined in /etc/services) occurred, the
1231 message was deferred as in a normal delivery, and thus remained on the
1232 spool, instead of being failed because of the mua_wrapper setting. This
1233 is now fixed, and I tidied up some of the mua_wrapper messages at the
1236 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1237 the messages in a hash of arrays rather than using individual hashes.
1238 This is a bit cleaner and results in dramatic memory savings, albeit
1239 at a slight CPU cost.
1241 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1242 as requested by Marc Sherman.
1244 SC/10 Eximstats: added histograms for user specified patterns as requested
1247 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1249 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1250 fopen() in the content-scanning modules that did not already have it.
1256 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1257 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1259 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1261 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1263 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1264 to test Sieve filters that use "vacation".
1266 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1267 that changes the way the GnuTLS parameters are stored in the cache file.
1268 The new format can be generated externally. For backward compatibility,
1269 if the data in the cache doesn't make sense, Exim assumes it has read an
1270 old-format file, and it generates new data and writes a new file. This
1271 means that you can't go back to an older release without removing the
1274 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1275 work if there are any delivery delays because "one_time" forces the
1276 parent to be marked "delivered", so its unseen clone is never tried
1277 again. For this reason, Exim now forbids the simultaneous setting of
1280 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1281 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1282 entry if you want to know the details. The fix, however, neglected to
1283 consider the case when local delivery batching is involved. The test for
1284 "previously delivered" was not happening when checking to see if an
1285 address could be batched with a previous (undelivered) one; under
1286 certain circumstances this could lead to multiple deliveries to the same
1289 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1290 in its include files, and this causes problems building Exim.
1292 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1293 header_syntax) but Exim was just ignoring anything given after a slash.
1294 In particular, this caused confusion with an attempt to use "verify =
1295 reverse_host_lookup/defer_ok". An error is now given when options are
1296 supplied for verify items that do not have them. (Maybe reverse_host_
1297 lookup should have a defer_ok option, but that's a different point.)
1299 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1300 defined by RFC 821) to 2048, because there were problems with some AUTH
1301 commands, and RFC 1869 says the size should be increased for extended
1302 SMTP commands that take arguments.
1304 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1307 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1308 "unknown" error; now it says that the functionality isn't in the binary.
1310 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1311 an address' error message when a string expansion fails (syntax or
1312 whatever). Otherwise the password may appear in the log. Following change
1313 PH/42 below, there is no longer a chance of it appearing in a bounce
1316 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1318 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1319 of its addresses was ever tried. (Bugzilla bug #2.)
1321 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1322 the result incorrectly in the debug output. (It correctly added a newline
1323 to what was transported.)
1325 TF/01 Added $received_time.
1327 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1328 commented out examples of how to interface to a virus scanner and to
1329 SpamAssassin. Also added commented examples of av_scanner and
1330 spamd_address settings.
1332 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1333 and controls are allowed in which ACLs. There were a couple of minor
1334 errors. Some of the entries in the conditions table (which is a table of
1335 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1336 negation of where the condition IS allowed.
1338 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1340 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1341 header file does not have a version number, so I've had to invent a new
1342 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1343 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1344 radiusclient), but it was contributed by a Radius user.
1346 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1347 files or directories, for OpenSSL.
1349 PH/19 When an Exim process that is running as root has to create an Exim log
1350 file, it does so in a subprocess that runs as exim:exim so as to get the
1351 ownership right at creation (otherwise, other Exim processes might see
1352 the file with the wrong ownership). There was no test for failure of this
1353 fork() call, which would lead to the process getting stuck as it waited
1354 for a non-existent subprocess. Forks do occasionally fail when resources
1355 run out. I reviewed all the other calls to fork(); they all seem to check
1358 PH/20 When checking for unexpected SMTP input at connect time (before writing
1359 the banner), Exim was not dealing correctly with a non-positive return
1360 from the read() function. If the client had disconnected by this time,
1361 the result was a log entry for a synchronization error with an empty
1362 string after "input=" when read() returned zero. If read() returned -1
1363 (an event I could not check), uninitialized data bytes were printed.
1364 There were reports of junk text (parts of files, etc) appearing after
1367 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1369 PH/22 Added support for macro redefinition, and (re)definition in between
1370 driver and ACL definitions.
1372 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1373 forgetting to use the resulting value; it was using the unexpanded value.
1375 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1376 hadn't been configured. The fix is from Juergen Kreileder, who
1377 understands it better than I do:
1379 "Here's what I see happening with three configured cyrus_sasl
1380 authenticators configured (plain, login, cram-md5):
1382 On startup auth_cyrus_sasl_init() gets called for each of these.
1383 This means three calls to sasl_listmech() without a specified mech_list.
1384 => SASL tests which mechs of all available mechs actually work
1385 => three warnings about OTP not working
1386 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1388 With the patch, sasl_listmech() also gets called three times. But now
1389 SASL's mech_list option is set to the server_mech specified in the the
1390 authenticator. Or in other words, the answer from sasl_listmech()
1391 gets limited to just the mech you're testing for (which is different
1393 => the return list contains just 'plain' or 'login', 'cram-md5' or
1394 nothing depending on the value of ob->server_mech.
1396 I've just tested the patch: Authentication still works fine,
1397 unavailable mechs specified in the exim configuration are still
1398 caught, and the auth.log warnings about OTP are gone."
1400 PH/25 When debugging is enabled, the contents of the command line are added
1401 to the debugging output, even when log_selector=+arguments is not
1404 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1405 answer is "GNU", and only if the return is "GNU/something" is the answer
1408 PH/27 $acl_verify_message is now set immediately after the failure of a
1409 verification in an ACL, and so is available in subsequent modifiers. In
1410 particular, the message can be preserved by coding like this:
1412 warn !verify = sender
1413 set acl_m0 = $acl_verify_message
1415 Previously, $acl_verify_message was set only while expanding "message"
1416 and "log_message" when a very denied access.
1418 PH/28 Modified OS/os.c-Linux with
1420 -#ifndef OS_LOAD_AVERAGE
1421 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1423 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1424 nomenclature these days.)
1426 PH/29 Installed patch from the Sieve maintainer that adds the options
1427 sieve_useraddress and sieve_subaddress to the redirect router.
1429 PH/30 In these circumstances:
1430 . Two addresses routed to the same list of hosts;
1431 . First host does not offer TLS;
1432 . First host accepts first address;
1433 . First host gives temporary error to second address;
1434 . Second host offers TLS and a TLS session is established;
1435 . Second host accepts second address.
1436 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1437 and peerdn, if requested) that were in fact used only for the second
1440 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1441 attention to any local part prefix or suffix that was matched by the
1442 router that accepted the address. It now behaves in the same way as it
1443 does for delivery: the affixes are removed from the local part unless
1444 rcpt_include_affixes is set on the transport.
1446 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1447 timeout during the DATA phase of an incoming message.
1449 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1450 applied a patch sanctioned by the Sieve maintainer.
1452 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1453 the uid or gid is negative. A case of a negative gid caused this to be
1454 noticed. The fix allows for either to be negative.
1456 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1457 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1460 PH/36 The change PH/12 above was broken. Fixed it.
1462 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1463 the grounds that routing the same address twice would always produce the
1464 same answer. This might have been true once, but it is certainly no
1465 longer true now. Routing a child address may depend on the previous
1466 routing that produced that child. Some complicated redirection strategies
1467 went wrong when messages had multiple recipients, and made Exim's
1468 behaviour dependent on the order in which the addresses were given.
1470 I have moved the duplicate checking until after the routing is complete.
1471 Exim scans the addresses that are assigned to local and remote
1472 transports, and removes any duplicates. This means that more work will be
1473 done, as duplicates will always all be routed, but duplicates are
1474 presumably rare, so I don't expect this is of any significance.
1476 For deliveries to pipes, files, and autoreplies, the duplicate checking
1477 still happens during the routing process, since they are not going to be
1480 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1481 It corrects a timeout issue with spamd. This is Ian's comment: "The
1482 background is that sometimes spamd either never reads data from a
1483 connection it has accepted, or it never writes response data. The exiscan
1484 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1485 blindly assumes that writes won't block so it may never time out."
1487 PH/39 Allow G after quota size as well as K and M.
1489 PH/40 The value set for $authenticated_id in an authenticator may not contain
1490 binary zeroes or newlines because the value is written to log lines and
1491 to spool files. There was no check on this. Now the value is run through
1492 the string_printing() function so that such characters are converted to
1493 printable escape sequences.
1495 PH/41 $message_linecount is a new variable that contains the total number of
1496 lines in the message. Compare $body_linecount, which is the count for the
1499 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1500 bounce and delay warning messages, except in certain special cases, which
1503 (a) An SMTP error message from a remote host;
1504 (b) A message specified in a :fail: redirection;
1505 (c) A message specified in a "fail" command in a system filter;
1506 (d) A message specified in a FAIL return from the queryprogram router;
1507 (e) A message specified by the cannot_route_message router option.
1509 In these cases only, Exim does include the error details in bounce and
1510 warning messages. There are also a few cases where bland messages such
1511 as "unrouteable address" or "local delivery error" are given.
1513 PH/43 $value is now also set for the "else" part of a ${run expansion.
1515 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1516 being worked on, but at least Exim now implements the latest version to
1519 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1520 process to complete was treated as a delivery failure, a timeout while
1521 writing the message to the pipe was logged, but erroneously treated as a
1522 successful delivery. Such timeouts include transport filter timeouts. For
1523 consistency with the overall process timeout, these timeouts are now
1524 treated as errors, giving rise to delivery failures by default. However,
1525 there is now a new Boolean option for the pipe transport called
1526 timeout_defer, which, if set TRUE, converts the failures into defers for
1527 both kinds of timeout. A transport filter timeout is now identified in
1530 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1531 systems where "make" and "gmake" are different, calling "gmake" at top
1532 level broke things. I've arranged for the value of $(MAKE) to be passed
1533 from the Makefile to this script so that it can call the same version of
1537 A note about Exim versions 4.44 and 4.50
1538 ----------------------------------------
1540 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1541 changes of various kinds. As a consequence, a big documentation update was
1542 needed. This delayed the release for rather longer than seemed good, especially
1543 in the light of a couple of (minor) security issues. Therefore, the changes
1544 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1545 release. So 4.44 and 4.50 are in effect two different branches that both start
1548 I have left the 4.50 change log unchanged; it contains all the changes since
1549 4.43. The change log for 4.44 is below; many of its items are identical to
1550 those for 4.50. This seems to be the most sensible way to preserve the
1551 historical information.
1557 1. Minor wording change to the doc/README.SIEVE file.
1559 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1560 computation of the current number of files was incorrect.
1562 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1563 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1566 4. Give more explanation in the error message when the command for a transport
1567 filter fails to execute.
1569 5. There are several places where Exim runs a non-Exim command in a
1570 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1571 was being done only for the command run by the queryprogram router. It is
1572 now done for all such subprocesses. The other cases are: ${run, transport
1573 filters, and the commands run by the lmtp and pipe transports.
1575 6. Added CONFIGURE_GROUP build-time option.
1577 7. Some older OS have a limit of 256 on the maximum number of file
1578 descriptors. Exim was using setrlimit() to set 1000 as a large value
1579 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1580 systems. I've change it so that if it can't get 1000, it tries for 256.
1582 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1583 was an oversight, and furthermore, ever since the addition of extra
1584 controls (e.g. 4.43/32), the checks on when to allow different forms of
1585 "control" were broken. There should now be diagnostics for all cases when a
1586 control that does not make sense is encountered.
1588 9. Added the /retain_sender option to "control=submission".
1590 10. $recipients is now available in the predata ACL (oversight).
1592 11. Tidy the search cache before the fork to do a delivery from a message
1593 received from the command line. Otherwise the child will trigger a lookup
1594 failure and thereby defer the delivery if it tries to use (for example) a
1595 cached ldap connection that the parent has called unbind on.
1597 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1598 of $address_data from the recipient verification was clobbered by the
1599 sender verification.
1601 13. The value of address_data from a sender verification is now available in
1602 $sender_address_data in subsequent conditions in the ACL statement.
1604 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1606 15. Added a new option "connect=<time>" to callout options, to set a different
1609 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1610 was its contents. (It was OK if the option was not defined at all.)
1612 17. A "Completed" log line is now written for messages that are removed from
1613 the spool by the -Mrm option.
1615 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1616 information about exactly what failed.
1618 19. Added -dd to debug only the daemon process.
1620 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1621 handles timeouts, both on the server side and network timeouts. Renamed the
1622 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1624 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1625 It is now set to "smtps".
1627 22. $host_address is now set to the target address during the checking of
1628 ignore_target_hosts.
1630 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1631 being passed; this would have caused $sender_host_name to have been used if
1632 matching the list had actually called for a host name (not very likely,
1633 since this list is usually IP addresses). A host name is now passed as
1636 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1637 code that specifies a non-restarting handler (typically sigaction() in
1638 modern systems) in an attempt to fix a rare and obscure crash bug.
1640 25. Narrowed the window for a race in the daemon that could cause it to ignore
1641 SIGCHLD signals. This is not a major problem, because they are used only to
1642 wake it up if nothing else does.
1644 26. A malformed maildirsize file could cause Exim to calculate negative values
1645 for the mailbox size or file count. Odd effects could occur as a result.
1646 The maildirsize information is now recalculated if the size or filecount
1649 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1650 support for a long time. Removed HAVE_SYS_VFS_H.
1652 28. Installed the latest version of exipick from John Jetmore.
1654 29. In an address list, if the pattern was not a regular expression, an empty
1655 subject address (from a bounce message) matched only if the pattern was an
1656 empty string. Non-empty patterns were not even tested. This was the wrong
1657 because it is perfectly reasonable to use an empty address as part of a
1658 database query. An empty address is now tested by patterns that are
1659 lookups. However, all the other forms of pattern expect the subject to
1660 contain a local part and a domain, and therefore, for them, an empty
1661 address still always fails if the pattern is not itself empty.
1663 30. Exim went into a mad DNS loop when attempting to do a callout where the
1664 host was specified on an smtp transport, and looking it up yielded more
1665 than one IP address.
1667 31. Re-factored the code for checking spool and log partition space into a
1668 function that finds that data and another that does the check. The former
1669 is then used to implement four new variables: $spool_space, $log_space,
1670 $spool_inodes, and $log_inodes.
1672 32. The RFC2047 encoding function was originally intended for short strings
1673 such as real names; it was not keeping to the 75-character limit for
1674 encoded words that the RFC imposes. It now respects the limit, and
1675 generates multiple encoded words if necessary. To be on the safe side, I
1676 have increased the buffer size for the ${rfc2047: expansion operator from
1679 33. It is now permitted to omit both strings after an "if" condition; if the
1680 condition is true, the result is "true". As before, when the second string
1681 is omitted, a false condition yields an empty string. This makes it less
1682 cumbersome to write custom ACL and router conditions.
1684 34. Failure to deliver a bounce message always caused it to be frozen, even if
1685 there was an errors_to setting on the router. The errors_to setting is now
1688 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1689 canonical form (fully expanded) before being placed in
1690 $sender_host_address.
1692 36. The table in the code that translates DNS record types into text (T_A to
1693 "A" for instance) was missing entries for NS and CNAME. It is just possible
1694 that this could have caused confusion if both these types were looked up
1695 for the same domain, because the text type is used as part of Exim's
1696 per-process caching. But the chance of anyone hitting this buglet seems
1699 37. The dnsdb lookup has been extended in a number of ways.
1701 (1) There is a new type, "zns", which walks up the domain tree until it
1702 finds some nameserver records. It should be used with care.
1704 (2) There is a new type, "mxh", which is like "mx" except that it returns
1705 just the host names, not the priorities.
1707 (3) It is now possible to give a list of domains (or IP addresses) to be
1708 looked up. The behaviour when one of the lookups defers can be
1709 controlled by a keyword.
1711 (4) It is now possible to specify the separator character for use when
1712 multiple records are returned.
1714 38. The dnslists ACL condition has been extended: it is now possible to supply
1715 a list of IP addresses and/or domains to be looked up in a particular DNS
1718 39. Added log_selector=+queue_time_overall.
1720 40. When running the queue in the test harness, wait just a tad after forking a
1721 delivery process, to get repeatability of debugging output.
1723 41. Include certificate and key file names in error message when GnuTLS fails
1724 to set them up, because the GnuTLS error message doesn't include the name
1725 of the failing file when there is a problem reading it.
1727 42. Allow both -bf and -bF in the same test run.
1729 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1731 44. The "Exiscan patch" is now merged into the mainline Exim source.
1733 45. Sometimes the final signoff response after QUIT could fail to get
1734 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1735 < 0 before doing a fflush(). This bug looks as though it goes back to the
1736 introduction of TLS in release 3.20, but "sometimes" must have been rare
1737 because the tests only now provoked it.
1739 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1740 (this can affect the format of dates).
1742 47. exim_tidydb, when checking for the continued existence of a message for
1743 which it has found a message-specific retry record, was not finding
1744 messages that were in split spool directories. Consequently, it was
1745 deleting retry records that should have stayed in existence.
1747 48. Steve fixed some bugs in eximstats.
1749 49. The SPA authentication driver was not abandoning authentication and moving
1750 on to the next authenticator when an expansion was forced to fail,
1751 contradicting the general specification for all authenticators. Instead it
1752 was generating a temporary error. It now behaves as specified.
1754 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1755 (the order specifies the preference for clients). The order is now AES256,
1756 AES128, 3DES, ARCFOUR128.
1758 51. Small patch to Sieve code - explicitly set From: when generating an
1761 52. Exim crashed if a remote delivery caused a very long error message to be
1762 recorded - for instance if somebody sent an entire SpamAssassin report back
1763 as a large number of 550 error lines. This bug was coincidentally fixed by
1764 increasing the size of one of Exim's internal buffers (big_buffer) that
1765 happened as part of the Exiscan merge. However, to be on the safe side, I
1766 have made the code more robust (and fixed the comments that describe what
1769 53. Now that there can be additional text after "Completed" in log lines (if
1770 the queue_time_overall log selector is set), a one-byte patch to exigrep
1771 was needed to allow it to recognize "Completed" as not the last thing in
1774 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1775 patch that reportedly fixes this has been added. I am not expert enough to
1776 create a test for it. This is what the patch creator wrote:
1778 "I found a little strange behaviour of ldap code when working with
1779 Windows 2003 AD Domain, where users was placed in more than one
1780 Organization Units. When I tried to give exim partial DN, the exit code
1781 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1782 But simultaneously result of request was absolutely normal ldap result,
1783 so I produce this patch..."
1785 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1786 REFERENCE, so I have modified the code to exclude the patch when that macro
1789 55. Some experimental protocols are using DNS PTR records for new purposes. The
1790 keys for these records are domain names, not reversed IP addresses. The
1791 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1792 leaves it alone. Component reversal etc. now happens only for IP addresses.
1795 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1797 57. Double the size of the debug message buffer (to 2048) so that more of very
1798 long debug lines gets shown.
1800 58. The exicyclog utility now does better if the number of log files to keep
1801 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1803 59. Two changes related to the smtp_active_hostname option:
1805 (1) $smtp_active_hostname is now available as a variable.
1806 (2) The default for smtp_banner uses $smtp_active_hostname instead
1807 of $primary_hostname.
1809 60. The host_aton() function is supposed to be passed a string that is known
1810 to be a valid IP address. However, in the case of IPv6 addresses, it was
1811 not checking this. This is a hostage to fortune. Exim now panics and dies
1812 if the condition is not met. A case was found where this could be provoked
1813 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1814 components; fortuitously, this particular loophole had already been fixed
1815 by change 4.50/55 above.
1817 If there are any other similar loopholes, the new check in host_aton()
1818 itself should stop them being exploited. The report I received stated that
1819 data on the command line could provoke the exploit when Exim was running as
1820 exim, but did not say which command line option was involved. All I could
1821 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1822 running as the user.
1825 61. There was a buffer overflow vulnerability in the SPA authentication code
1826 (which came originally from the Samba project). I have added a test to the
1827 spa_base64_to_bits() function which I hope fixes it.
1830 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1831 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1833 63. The daemon start-up calls getloadavg() while still root for those OS that
1834 need the first call to be done as root, but it missed one case: when
1835 deliver_queue_load_max is set with deliver_drop_privilege. This is
1836 necessary for the benefit of the queue runner, because there is no re-exec
1837 when deliver_drop_privilege is set.
1839 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1840 This has been fixed.
1842 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1843 was in use, was not putting the data itself into the right store pool;
1844 consequently, it could be overwritten for a subsequent message in the same
1845 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1848 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1850 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1851 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1852 the function were treating the return as a boolean value, which happened to
1853 work because 0=false and not-0=true, but is not correct code.
1855 68. The host_aton() function was not handling scoped IPv6 addresses (those
1856 with, for example, "%eth0" on the end) correctly.
1858 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1859 negated items (that is, ~something) in unsigned ints. Some compilers
1860 apparently mutter when there is no cast.
1862 70. If an address verification called from an ACL failed, and did not produce a
1863 user-specific message (i.e. there was only a "system" message), nothing was
1864 put in $acl_verify_message. In this situation, it now puts the system
1867 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1868 session; change 4.31/43 added the unwanted input to the log line - except
1869 that it did not do this in the start of session case. It now does.
1871 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1872 This is wrong and can cause the other end to generate a synchronization
1873 error if it is another Exim or anything else that does the synchronization
1874 check. A QUIT command is no longer sent after a timeout.
1876 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1877 during host lookups.
1879 74. The defer_ok option of callout verification was not working if it was used
1880 when verifying addresses in header lines, that is, for this case:
1882 verify = header_sender/callout=defer_ok
1884 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1885 those file descriptors could be used for SMTP connections. If anything
1886 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1887 could be sent to the SMTP client, causing chaos. The daemon now opens
1888 stdin, stdout, and stderr to /dev/null when it puts itself into the
1891 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1892 log by default. The user can override this with suitable Perl magic.
1894 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1895 the log message when discard triggers, was not working for the DATA ACL or
1896 for the non-SMTP ACL.
1898 78. Error message wording change in sieve.c.
1900 79. If smtp_accept_max_per_host was set, the number of connections could be
1901 restricted to fewer than expected, because the daemon was trying to set up
1902 a new connection before checking whether the processes handling previous
1903 connections had finished. The check for completed processes is now done
1904 earlier. On busy systems, this bug wouldn't be noticed because something
1905 else would have woken the daemon, and it would have reaped the completed
1908 80. If a message was submitted locally by a user whose login name contained one
1909 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1910 It caused a spool format error. I have fixed the spool reading code. A
1911 related problem was that the "from" clause in the Received: line became
1912 illegal because of the space(s). It is now covered by ${quote_local_part.
1914 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1917 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1919 83. Patch from Sieve maintainer to fix unterminated string problem in
1920 "vacation" handling.
1922 84. Some minor changes to the Linux configuration files to help with other
1923 OS variants using glibc.
1925 85. One more patch for Sieve to update vacation handling to latest spec.
1928 ----------------------------------------------------
1929 See the note above about the 4.44 and 4.50 releases.
1930 ----------------------------------------------------
1936 1. Change 4.43/35 introduced a bug that caused file counts to be
1937 incorrectly computed when quota_filecount was set in an appendfile
1940 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1941 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1944 3. Give more explanation in the error message when the command for a transport
1945 filter fails to execute.
1947 4. There are several places where Exim runs a non-Exim command in a
1948 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1949 was being done only for the command run by the queryprogram router. It is
1950 now done for all such subprocesses. The other cases are: ${run, transport
1951 filters, and the commands run by the lmtp and pipe transports.
1953 5. Some older OS have a limit of 256 on the maximum number of file
1954 descriptors. Exim was using setrlimit() to set 1000 as a large value
1955 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1956 systems. I've change it so that if it can't get 1000, it tries for 256.
1958 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1959 was an oversight, and furthermore, ever since the addition of extra
1960 controls (e.g. 4.43/32), the checks on when to allow different forms of
1961 "control" were broken. There should now be diagnostics for all cases when a
1962 control that does not make sense is encountered.
1964 7. $recipients is now available in the predata ACL (oversight).
1966 8. Tidy the search cache before the fork to do a delivery from a message
1967 received from the command line. Otherwise the child will trigger a lookup
1968 failure and thereby defer the delivery if it tries to use (for example) a
1969 cached ldap connection that the parent has called unbind on.
1971 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1972 of $address_data from the recipient verification was clobbered by the
1973 sender verification.
1975 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1976 was its contents. (It was OK if the option was not defined at all.)
1978 11. A "Completed" log line is now written for messages that are removed from
1979 the spool by the -Mrm option.
1981 12. $host_address is now set to the target address during the checking of
1982 ignore_target_hosts.
1984 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1985 being passed; this would have caused $sender_host_name to have been used if
1986 matching the list had actually called for a host name (not very likely,
1987 since this list is usually IP addresses). A host name is now passed as
1990 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1991 code that specifies a non-restarting handler (typically sigaction() in
1992 modern systems) in an attempt to fix a rare and obscure crash bug.
1994 15. Narrowed the window for a race in the daemon that could cause it to ignore
1995 SIGCHLD signals. This is not a major problem, because they are used only to
1996 wake it up if nothing else does.
1998 16. A malformed maildirsize file could cause Exim to calculate negative values
1999 for the mailbox size or file count. Odd effects could occur as a result.
2000 The maildirsize information is now recalculated if the size or filecount
2003 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2004 support for a long time. Removed HAVE_SYS_VFS_H.
2006 18. Updated exipick to current release from John Jetmore.
2008 19. Allow an empty sender to be matched against a lookup in an address list.
2009 Previously the only cases considered were a regular expression, or an
2012 20. Exim went into a mad DNS lookup loop when doing a callout where the
2013 host was specified on the transport, if the DNS lookup yielded more than
2016 21. The RFC2047 encoding function was originally intended for short strings
2017 such as real names; it was not keeping to the 75-character limit for
2018 encoded words that the RFC imposes. It now respects the limit, and
2019 generates multiple encoded words if necessary. To be on the safe side, I
2020 have increased the buffer size for the ${rfc2047: expansion operator from
2023 22. Failure to deliver a bounce message always caused it to be frozen, even if
2024 there was an errors_to setting on the router. The errors_to setting is now
2027 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2028 canonical form (fully expanded) before being placed in
2029 $sender_host_address.
2031 24. Updated eximstats to version 1.33
2033 25. Include certificate and key file names in error message when GnuTLS fails
2034 to set them up, because the GnuTLS error message doesn't include the name
2035 of the failing file when there is a problem reading it.
2037 26. Expand error message when OpenSSL has problems setting up cert/key files.
2040 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2041 (this can affect the format of dates).
2043 28. exim_tidydb, when checking for the continued existence of a message for
2044 which it has found a message-specific retry record, was not finding
2045 messages that were in split spool directories. Consequently, it was
2046 deleting retry records that should have stayed in existence.
2048 29. eximstats updated to version 1.35
2049 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2050 1.35 - bugfix such that pie charts by volume are generated correctly
2052 30. The SPA authentication driver was not abandoning authentication and moving
2053 on to the next authenticator when an expansion was forced to fail,
2054 contradicting the general specification for all authenticators. Instead it
2055 was generating a temporary error. It now behaves as specified.
2057 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2058 (the order specifies the preference for clients). The order is now AES256,
2059 AES128, 3DES, ARCFOUR128.
2061 31. Small patch to Sieve code - explicitly set From: when generating an
2064 32. Exim crashed if a remote delivery caused a very long error message to be
2065 recorded - for instance if somebody sent an entire SpamAssassin report back
2066 as a large number of 550 error lines. This bug was coincidentally fixed by
2067 increasing the size of one of Exim's internal buffers (big_buffer) that
2068 happened as part of the Exiscan merge. However, to be on the safe side, I
2069 have made the code more robust (and fixed the comments that describe what
2072 33. Some experimental protocols are using DNS PTR records for new purposes. The
2073 keys for these records are domain names, not reversed IP addresses. The
2074 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2075 leaves it alone. Component reversal etc. now happens only for IP addresses.
2078 34. The host_aton() function is supposed to be passed a string that is known
2079 to be a valid IP address. However, in the case of IPv6 addresses, it was
2080 not checking this. This is a hostage to fortune. Exim now panics and dies
2081 if the condition is not met. A case was found where this could be provoked
2082 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2083 components; fortuitously, this particular loophole had already been fixed
2084 by change 4.50/55 or 4.44/33 above.
2086 If there are any other similar loopholes, the new check in host_aton()
2087 itself should stop them being exploited. The report I received stated that
2088 data on the command line could provoke the exploit when Exim was running as
2089 exim, but did not say which command line option was involved. All I could
2090 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2091 running as the user.
2094 35. There was a buffer overflow vulnerability in the SPA authentication code
2095 (which came originally from the Samba project). I have added a test to the
2096 spa_base64_to_bits() function which I hope fixes it.
2099 36. The daemon start-up calls getloadavg() while still root for those OS that
2100 need the first call to be done as root, but it missed one case: when
2101 deliver_queue_load_max is set with deliver_drop_privilege. This is
2102 necessary for the benefit of the queue runner, because there is no re-exec
2103 when deliver_drop_privilege is set.
2105 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2106 was in use, was not putting the data itself into the right store pool;
2107 consequently, it could be overwritten for a subsequent message in the same
2108 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2111 38. Sometimes the final signoff response after QUIT could fail to get
2112 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2113 < 0 before doing a fflush(). This bug looks as though it goes back to the
2114 introduction of TLS in release 3.20, but "sometimes" must have been rare
2115 because the tests only now provoked it.
2121 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2122 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2123 Now it returns an integer. A number of calls were still expecting a T/F
2124 return. Fortuitously, in all cases, the tests worked in OK situations,
2125 which is the norm. However, things would have gone wrong on any write
2126 failures on the smtp file descriptor. This function is used when sending
2127 messages over SMTP and also when doing verify callouts.
2129 2. When Exim is called to do synchronous delivery of a locally submitted
2130 message (the -odf or -odi options), it no longer closes stderr before doing
2133 3. Implemented the mua_wrapper option.
2135 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2137 5. Implemented the functions header_remove(), header_testname(),
2138 header_add_at_position(), and receive_remove_recipient(), and exported them
2141 6. If an ACL "warn" statement specified the addition of headers, Exim already
2142 inserted X-ACL-Warn: at the start if there was no header name. However, it
2143 was not making this test for the second and subsequent header lines if
2144 there were newlines in the string. This meant that an invalid header could
2145 be inserted if Exim was badly configured.
2147 7. Allow an ACL "warn" statement to add header lines at the start or after all
2148 the Received: headers, as well as at the end.
2150 8. Added the rcpt_4xx retry error code.
2152 9. Added postmaster_mailfrom=xxx to callout verification option.
2154 10. Added mailfrom=xxxx to the callout verification option, for verify=
2157 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2158 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2160 12. Inserted some casts to stop certain compilers warning when using pointer
2161 differences as field lengths or precisions in printf-type calls (mostly
2162 affecting debugging statements).
2164 13. Added optional readline() support for -be (dynamically loaded).
2166 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2167 same clock tick as a message's arrival, so that its received time was the
2168 same as the "first fail" time on the retry record, and that message
2169 remained on the queue past the ultimate address timeout, every queue runner
2170 would try a delivery (because it was past the ultimate address timeout) but
2171 after another failure, the ultimate address timeout, which should have then
2172 bounced the address, did not kick in. This was a "< instead of <=" error;
2173 in most cases the first failure would have been in the next clock tick
2174 after the received time, and all would be well.
2176 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2177 being recognized when the domain list was tested by the match_domain
2178 condition in an expansion string.
2180 16. Added the ${str2b64: operator.
2182 17. Exim was always calling setrlimit() to set a large limit for the number of
2183 processes, without checking whether the existing limit was already
2184 adequate. (It did check for the limit on file descriptors.) Furthermore,
2185 errors from getrlimit() and setrlimit() were being ignored. Now they are
2186 logged to the main and panic logs, but Exim does carry on, to try to do its
2187 job under whatever limits there are.
2189 18. Imported PCRE 5.0.
2191 19. Trivial typo in log message " temporarily refused connection" (the leading
2194 20. If the log selector return_path_on_delivery was set and an address was
2195 redirected to /dev/null, the delivery process crashed because it assumed
2196 that a return path would always be set for a "successful" delivery. In this
2197 case, the whole delivery is bypassed as an optimization, and therefore no
2200 21. Internal re-arrangement: the function for sending a challenge and reading
2201 a response while authentication was assuming a zero-terminated challenge
2202 string. It's now changed to take a pointer and a length, to allow for
2203 binary data in such strings.
2205 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2207 23. Exim was not respecting finduser_retries when seeking the login of the
2208 uid under which it was called; it was always trying 10 times. (The default
2209 setting of finduser_retries is zero.) Also, it was sleeping after the final
2210 failure, which is pointless.
2212 24. Implemented tls_on_connect_ports.
2214 25. Implemented acl_smtp_predata.
2216 26. If the domain in control=submission is set empty, Exim assumes that the
2217 authenticated id is a complete email address when it generates From: or
2218 Sender: header lines.
2220 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2221 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2222 chown and chgrp in /bin and hostname in /usr/bin.
2224 28. Exim was keeping the "process log" file open after each use, just as it
2225 does for the main log. This opens the possibility of it remaining open for
2226 long periods when the USR1 signal hits a daemon. Occasional processlog
2227 errors were reported, that could have been caused by this. Anyway, it seems
2228 much more sensible not to leave this file open at all, so that is what now
2231 29. The long-running daemon process does not normally write to the log once it
2232 has entered its main loop, and it closes the log before doing so. This is
2233 so that log files can straightforwardly be renamed and moved. However,
2234 there are a couple of unusual error situations where the daemon does write
2235 log entries, and I had neglected to close the log afterwards.
2237 30. The text of an SMTP error response that was received during a remote
2238 delivery was being truncated at 512 bytes. This is too short for some of
2239 the long messages that one sometimes sees. I've increased the limit to
2242 31. It is now possible to make retry rules that apply only when a message has a
2243 specific sender, in particular, an empty sender.
2245 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2246 it possible to be selective about when SMTP synchronization is enforced.
2248 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2250 32. Implemented hosts_connection_nolog.
2252 33. Added an ACL for QUIT.
2254 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2257 35. Added mailbox_size and mailbox_filecount to appendfile.
2259 36. Added control = no_multiline_responses to ACLs.
2261 37. There was a bug in the logic of the code that waits for the clock to tick
2262 in the case where the clock went backwards by a substantial amount such
2263 that the microsecond fraction of "now" was more than the microsecond
2264 fraction of "then" (but the whole seconds number was less).
2266 38. Added support for the libradius Radius client library this is found on
2267 FreeBSD (previously only the radiusclient library was supported).
2273 1. When certain lookups returned multiple values in the form name=value, the
2274 quoting of the values was not always being done properly. Specifically:
2275 (a) If the value started with a double quote, but contained no whitespace,
2277 (b) If the value contained whitespace other than a space character (i.e.
2278 tabs or newlines or carriage returns) it was not quoted.
2279 This fix has been applied to the mysql and pgsql lookups by writing a
2280 separate quoting function and calling it from the lookup code. The fix
2281 should probably also be applied to nisplus, ibase and oracle lookups, but
2282 since I cannot test any of those, I have not disturbed their existing code.
2284 2. A hit in the callout cache for a specific address caused a log line with no
2285 reason for rejecting RCPT. Now it says "Previous (cached) callout
2286 verification failure".
2288 3. There was an off-by-one bug in the queryprogram router. An over-long
2289 return line was truncated at 256 instead of 255 characters, thereby
2290 overflowing its buffer with the terminating zero. As well as fixing this, I
2291 have increased the buffer size to 1024 (and made a note to document this).
2293 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2294 when Exim is waiting for an SMTP response from a remote server, Exim
2295 restarts its select() call on the socket, thereby resetting its timeout.
2296 This is not a problem when such interrupts are rare. Somebody set up a cron
2297 job to run exiwhat every 2 minutes, which is less than the normal select()
2298 timeout (5 or 10 minutes). This meant that the select() timeout never
2299 kicked in because it was always reset. I have fixed this by comparing the
2300 time when an interrupt arrives with the time at the start of the first call
2301 to select(). If more time than the timeout has elapsed, the interrupt is
2302 treated as a timeout.
2304 5. Some internal re-factoring in preparation for the addition of Sieve
2305 extensions (by MH). In particular, the "personal" test is moved to a
2306 separate function, and given an option for scanning Cc: and Bcc: (which is
2307 not set for Exim filters).
2309 6. When Exim created an email address using the login of the caller as the
2310 local part (e.g. when creating a From: or Sender: header line), it was not
2311 quoting the local part when it contained special characters such as @.
2313 7. Installed new OpenBSD configuration files.
2315 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2316 try to make them clearer.
2318 9. Callout options, other than the timeout value, were being ignored when
2319 verifying sender addresses in header lines. For example, when using
2321 verify = header_sender/callout=no_cache
2323 the cache was (incorrectly) being used.
2325 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2326 only the Cygwin environment.
2328 11. When return_path_on_delivery was set as a log selector, if different remote
2329 addresses in the same message used different return paths and parallel
2330 remote delivery occurred, the wrong values would sometimes be logged.
2331 (Whenever a remote delivery process finished, the return path value from
2332 the most recently started remote delivery process was logged.)
2334 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2335 lines when AUTH and/or TLS are in use. This is the "received protocol"
2336 field. Exim used to use "asmtp" for authenticated SMTP, without any
2337 indication (in the protocol name) for TLS use. Now it follows the RFC and
2338 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2339 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2340 names appear in log lines as well as in Received: header lines.
2342 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2343 extensions, and comparison tests, and to fix some bugs.
2345 14. Changes to the "personal" filter test:
2347 (1) The test was buggy in that it was just doing the equivalent of
2348 "contains" tests on header lines. For example, if a user's address was
2349 anne@some.where, the "personal" test would incorrectly be true for
2351 To: susanne@some.where
2353 This test is now done by extracting each address from the header in turn,
2354 and checking the entire address. Other tests that are part of "personal"
2355 are now done using regular expressions (for example, to check local parts
2356 of addresses in From: header lines).
2358 (2) The list of non-personal local parts in From: addresses has been
2359 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2360 taken from the Sieve specification recommendations.
2362 (3) If the message contains any header line starting with "List-" it is
2363 treated as non-personal.
2365 (4) The test for "circular" in the Subject: header line has been removed
2366 because it now seems ill-conceived.
2368 15. Minor typos in src/EDITME comments corrected.
2370 16. Installed latest exipick from John Jetmore.
2372 17. If headers_add on a router specified a text string that was too long for
2373 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2374 of string_sprintf() is now avoided.
2376 18. $message_body_size was not set (it was always zero) when running the DATA
2377 ACL and the local_scan() function.
2379 19. For the "mail" command in an Exim filter, no default was being set for
2380 the once_repeat time, causing a random time value to be used if "once" was
2381 specified. (If the value happened to be <= 0, no repeat happened.) The
2382 default is now 0s, meaning "never repeat". The "vacation" command was OK
2383 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2384 (I found it when inspecting the code).
2386 20. There is now an overall timeout for performing a callout verification. It
2387 defaults to 4 times the callout timeout, which applies to individual SMTP
2388 commands during the callout. The overall timeout applies when there is more
2389 than one host that can be tried. The timeout is checked before trying the
2390 next host. This prevents very long delays if there are a large number of
2391 hosts and all are timing out (e.g. when the network connections are timing
2392 out). The value of the overall timeout can be changed by specifying an
2393 additional sub-option for "callout", called "maxwait". For example:
2395 verify = sender/callout=5s,maxwait=20s
2397 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2398 to the end before writing, but this should make it even safer).
2400 22. Exim was forgetting that it had advertised PIPELINING for the second and
2401 subsequent messages on an SMTP connection. It was also not resetting its
2402 memory on STARTTLS and an internal HELO.
2404 23. When Exim logs an SMTP synchronization error within a session, it now
2405 records whether PIPELINING has been advertised or not.
2407 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2408 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2409 rather than long int.
2411 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2413 26. Added the never_mail option to autoreply.
2419 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2420 crash if the getsockname() call failed; this can happen if a connection is
2421 closed very soon after it is established. The problem was simply in the
2422 order in which certain operations were done, causing Exim to try to write
2423 to the SMTP stream before it had set up the file descriptor. The bug has
2424 been fixed by making things happen in the correct order.
2430 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2431 before the connection was closed, thus losing the rejection response.
2433 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2434 some early Solaris releases, but causes trouble in current releases where
2435 socklen_t is defined.
2437 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2440 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2443 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2444 that releasing the top of it at the end releases what was used for sub-
2445 expansions (unless the block got too big). However, discard this block if
2446 the first thing is a variable or header, so that we can use its block when
2447 it is dynamic (useful for very large $message_headers, for example).
2449 6. Lookups now cache *every* query, not just the most recent. A new, separate
2450 store pool is used for this. It can be recovered when all lookup caches are
2451 flushed. Lookups now release memory at the end of their result strings.
2452 This has involved some general refactoring of the lookup sources.
2454 7. Some code has been added to the store_xxx() functions to reduce the amount
2455 of flapping under certain conditions.
2457 8. log_incoming_interface used to affect only the <= reception log lines. Now
2458 it causes the local interface and port to be added to several more SMTP log
2459 lines, for example "SMTP connection from", and rejection lines.
2461 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2463 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2465 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2466 could be overwritten at the end of the current message (or the start of a
2467 new message if it was set in a HELO ACL). The value is now preserved for
2468 the duration of the SMTP connection.
2470 12. If a transport had a headers_rewrite setting, and a matching header line
2471 contained an unqualified address, that address was qualified, even if it
2472 did not match any rewriting rules. The underlying bug was that the values
2473 of the flags that permit the existence of unqualified sender and recipient
2474 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2475 non-local messages, and by -bnq for local messages) were not being
2476 preserved with the message after it was received.
2478 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2479 "next input=" as part of the text comprising the host identity instead of
2480 the correct text. The code was using the same buffer for two different
2481 strings. However, depending on which order the printing function evaluated
2482 its arguments, the bug did not always show up. Under Linux, for example, my
2483 test suite worked just fine.
2485 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2486 doesn't work with some older versions of Perl. It has been changed to "my",
2487 which in any case is probably the better facility to use.
2489 15. A really picky compiler found some instances of statements for creating
2490 error messages that either had too many or two few arguments for the format
2493 16. The size of the buffer for calls to the DNS resolver has been increased
2494 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2495 for addresses that have a lot of PTR records. This alleviates a problem; it
2496 does not fully solve it.
2498 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2499 buffer now truncates the list and logs the incident, which is the same
2500 action as happens when Exim is looking up a host name and its aliases.
2501 Previously in this situation something unpredictable would happen;
2502 sometimes it was "internal error: store_reset failed".
2504 18. If a server dropped the connection unexpectedly when an Exim client was
2505 using GnuTLS and trying to read a response, the client delivery process
2506 crashed while trying to generate an error log message.
2508 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2509 string, for example:
2511 warn message = H1: something\nH2: something
2513 the text was added as a single header line from Exim's point of view
2514 though it ended up OK in the delivered message. However, searching for the
2515 second and subsequent header lines using $h_h2: did not work. This has been
2516 fixed. Similarly, if a system filter added multiple headers in this way,
2517 the routers could not see them.
2519 20. Expanded the error message when iplsearch is called with an invalid key to
2520 suggest using net-iplsearch in a host list.
2522 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2523 ACLs are no longer actually imposed (and a message to that effect is
2526 22. If a "gecos" field in a passwd entry contained escaped characters, in
2527 particular, if it contained a \" sequence, Exim got it wrong when building
2528 a From: or a Sender: header from that name. A second bug also caused
2529 incorrect handling when an unquoted " was present following a character
2530 that needed quoting.
2532 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2533 was not being matched caselessly.
2535 24. Arranged for all hyphens in the exim.8 source to be escaped with
2538 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2539 properly. Recipient callout cache records were still being keyed to include
2540 the sender, even when use_sender was set false. This led to far more
2541 callouts that were necessary. The sender is no longer included in the key
2542 when use_sender is false.
2544 26. Added "control = submission" modifier to ACLs.
2546 27. Added the ${base62d: operator to decode base 62 numbers.
2548 28. dnsdb lookups can now access SRV records.
2550 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2551 the configuration file.
2553 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2554 (-v) mode. This makes the output for a verbose queue run more intelligible.
2556 31. Added a use_postmaster feature to recipient callouts.
2558 32. Added the $body_zerocount variable, containing the number of binary zero
2559 bytes in the message body.
2561 33. The time of last modification of the "new" subdirectory is now used as the
2562 "mailbox time last read" when there is a quota error for a maildir
2565 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2567 35. Added +ignore_unknown as a special item in host lists.
2569 36. Code for decoding IPv6 addresses in host lists is now included, even if
2570 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2571 address was recognized as an IP address, but was then not correctly decoded
2572 into binary, causing unexpected and incorrect effects when compared with
2579 1. Very minor rewording of debugging text in manualroute to say "list of
2580 hosts" instead of "hostlist".
2582 2. If verify=header_syntax was set, and a header line with an unqualified
2583 address (no domain) and a large number of spaces between the end of the
2584 name and the colon was received, the reception process suffered a buffer
2585 overflow, and (when I tested it) crashed. This was caused by some obsolete
2586 code that should have been removed. The fix is to remove it!
2588 3. When running in the test harness, delay a bit after writing a bounce
2589 message to get a bit more predictability in the log output.
2591 4. Added a call to search_tidyup() just before forking a reception process. In
2592 theory, someone could use a lookup in the expansion of smtp_accept_max_
2593 per_host which, without the tidyup, could leave open a database connection.
2595 5. Added the variables $recipient_data and $sender_data which get set from a
2596 lookup success in an ACL "recipients" or "senders" condition, or a router
2597 "senders" option, similar to $domain_data and $local_part_data.
2599 6. Moved the writing of debug_print from before to after the "senders" test
2602 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2603 problems for message scanning, either using a data ACL, or using
2604 local_scan() because the Received: header was not generated till after they
2605 were called (in order to set the time as the time of reception completion).
2606 I have revised the way this works. The header is now generated after the
2607 body is received, but before the ACL or local_scan() are called. After they
2608 are run, the timestamp in the header is updated.
2614 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2615 before starting a queue runner without re-exec. This happened only when
2616 deliver_drop_privilege was set or when the Exim user was set to root. The
2617 effect of the bug was that timeouts during subsequent deliveries caused
2618 crashes instead of being properly handled. The handler is now left at its
2619 default (and expected) setting.
2621 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2622 message, again when deliver_drop_privilege is set or Exim is run as root.
2623 The bug described in (1) was not present in this case, but the tidying up
2624 of the other signals was missing. I have made the two cases consistent.
2626 3. The ignore_target_hosts setting on a manualroute router was being ignored
2627 for hosts that were looked up using the /MX notation.
2629 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2632 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2633 operated on the sender address. After changing the $sender_address to <>
2634 for the sender address verify, Exim was re-instated it as the original
2635 (before rewriting) address, but remembering that it had rewritten it, so it
2636 wasn't rewriting it again. This bug also had the effect of breaking the
2637 sender address verification caching when the sender address was rewritten.
2639 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2640 This has been changed so that if the ip literal address matches
2641 ignore_target_hosts, the router declines.
2643 7. Added expansion conditions match_domain, match_address, and match_local_
2644 part (NOT match_host).
2646 8. The placeholder for the Received: header didn't have a length field set.
2648 9. Added code to Exim itself and to exim_lock to test for a specific race
2649 condition that could lead to file corruption when using MBX delivery. The
2650 issue is with the lockfile that is created in /tmp. If this file is removed
2651 after a process has opened it but before that process has acquired a lock,
2652 there is the potential for a second process to recreate the file and also
2653 acquire a lock. This could lead to two Exim processes writing to the file
2654 at the same time. The added code performs the same test as UW imapd; it
2655 checks after acquiring the lock that its file descriptor still refers to
2656 the same named file.
2658 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2659 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2662 11. Added the smtp_active_hostname option. If used, this will typically be made
2663 to depend on the incoming interface address. Because $interface_address is
2664 not set up until the daemon has forked a reception process, error responses
2665 that can happen earlier (such as "too many connections") no longer contain
2668 12. If an expansion in a condition on a "warn" statement fails because a lookup
2669 defers, the "warn" statement is abandoned, and the next ACL statement is
2670 processed. Previously this caused the whole ACL to be aborted.
2672 13. Added the iplsearch lookup type.
2674 14. Added ident_timeout as a log selector.
2676 15. Added tls_certificate_verified as a log selector.
2678 16. Added a global option tls_require_ciphers (compare the smtp transport
2679 option of the same name). This controls incoming TLS connections.
2681 17. I finally figured out how to make tls_require_ciphers do a similar thing
2682 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2683 before starting the TLS session.
2685 18. Tabs are now shown as \t in -bP output.
2687 19. If the log selector return_path_on_delivery was set, Exim crashed when
2688 bouncing a message because it had too many Received: header lines.
2690 20. If two routers both had headers_remove settings, and the first one included
2691 a superfluous trailing colon, the final name in the first list and the
2692 first name in the second list were incorrectly joined into one item (with a
2693 colon in the middle).
2699 1. Added -C and -D options to the exinext utility, mainly to make it easier
2700 to include in the automated testing, but these could be helpful when
2701 multiple configurations are in use.
2703 2. The exinext utility was not formatting the output nicely when there was
2704 an alternate port involved in the retry record key, nor when there was a
2705 message id as well (for retries that were specific to a specific message
2706 and a specific host). It was also confused by IPv6 addresses, because of
2707 the additional colons they contain. I have fixed the IPv4 problem, and
2708 patched it up to do a reasonable job for IPv6.
2710 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2711 delivery, the log line now contains "pipelined" if PIPELINING was used.
2713 4. An SMTP transport process used to panic and die if the bind() call to set
2714 an explicit outgoing interface failed. This has been changed; it is now
2715 treated in the same way as a connect() failure.
2717 5. A reference to $sender_host_name in the part of a conditional expansion
2718 that was being skipped was still causing a DNS lookup. This no longer
2721 6. The def: expansion condition was not recognizing references to header lines
2722 that used bh_ and bheader_.
2724 7. Added the _cache feature to named lists.
2726 8. The code for checking quota_filecount in the appendfile transport was
2727 allowing one more file than it should have been.
2729 9. For compatibility with Sendmail, the command line option
2737 and sets the incoming protocol and host name (for trusted callers). The
2738 host name and its colon can be omitted when only the protocol is to be set.
2739 Note the Exim already has two private options, -pd and -ps, that refer to
2740 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2741 "s", but I don't think that's a major issue.
2743 10. A number of refactoring changes to the code, none of which should affect
2746 (a) The number of logging options was getting close to filling up the
2747 32-bit word that was used as a bit map. I have split them into two classes:
2748 those that are passed in the argument to log_write(), and those that are
2749 only ever tested independently outside of that function. These are now in
2750 separate 32-bit words, so there is plenty of room for expansion again.
2751 There is no change in the user interface or the logging behaviour.
2753 (b) When building, for example, log lines, the code previously used a
2754 macro that called string_cat() twice, in order to add two strings. This is
2755 not really sufficiently general. Furthermore, there was one instance where
2756 it was actually wrong because one of the argument was used twice, and in
2757 one call a function was used. (As it happened, calling the function twice
2758 did not affect the overall behaviour.) The macro has been replaced by a
2759 function that can join an arbitrary number of extra strings onto a growing
2762 (c) The code for expansion conditions now uses a table and a binary chop
2763 instead of a serial search (which was left over from when there were very
2764 few conditions). Also, it now recognizes conditions like "pam" even when
2765 the relevant support is not compiled in: a suitably worded error message is
2766 given if an attempt is made to use such a condition.
2768 11. Added ${time_interval:xxxxx}.
2770 12. A bug was causing one of the ddress fields not to be passed back correctly
2771 from remote delivery subprocesses. The field in question was not being
2772 subsequently used, so this caused to problems in practice.
2774 13. Added new log selectors queue_time and deliver_time.
2776 14. Might have fixed a bug in maildirsizefile handling that threw up
2777 "unexpected character" debug warnings, and recalculated the data
2778 unnecessarily. In any case, I expanded the warning message to give more
2781 15. Added the message "Restricted characters in address" to the statements in
2782 the default ACL that block characters like @ and % in local parts.
2784 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2785 Three changes have been made:
2787 (a) There was a serious bug; a negative response to MAIL caused the whole
2788 recipient domain to be cached as invalid, thereby blocking all messages
2789 to all local parts at the same domain, from all senders. This bug has
2790 been fixed. The domain is no longer cached after a negative response to
2791 MAIL if the sender used is not empty.
2793 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2796 (c) A new callout option, "use_sender" has been added for people who want
2797 the modified behaviour.
2803 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2806 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2807 indeed breaks things for older releases.
2809 3. Added additional logging to the case where there is a problem reading data
2810 from a filter that is running in a subprocess using a pipe, in order to
2811 try to track down a specific problem.
2813 4. Testing facility fudge: when running in the test harness and attempting
2814 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2815 getting "No route to host". Convert this to a timeout.
2817 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2820 6. Some OS don't have socklen_t but use size_t instead. This affects the
2821 fifth argument of getsockopt() amongst other things. This is now
2822 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2823 can be set for individual OS. I have set it for SunOS5, OSF1, and
2824 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2825 some earlier ones do not.
2827 7. Change 4.30/15 was not doing the test caselessly.
2829 8. The standard form for an IPv6 address literal was being rejected by address
2830 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2831 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2832 this, as well as the form without the "IPv6" on the front (but only when
2833 address literals are enabled, of course).
2835 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2837 10. Exim crashed if a message with an empty sender address specified by -f
2838 encountered a router with an errors_to setting. This could be provoked only
2839 by a command such as
2843 where an empty string was supplied; "<>" did not hit this bug.
2845 11. Installed PCRE release 4.5.
2847 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2848 remained set. It is now erased.
2850 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2851 times from message ids (which are base 36 rather than the normal 62).
2853 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2854 were being counted as actual protocol errors, and logged if the log
2855 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2856 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2857 and DATA following a set of rejected RCPTs do not count as protocol errors.
2858 In other words, Exim assumes they were pipelined, though this may not
2859 actually be the case. Of course, in all cases the client gets an
2860 appropriate error code.
2862 15. If a lookup fails in an ACL condition, a message about the failure may
2863 be available; it is used if testing the ACL cannot continue, because most
2864 such messages specify what the cause of the deferral is. However, some
2865 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2866 that caused an old message to be retained and used if a later statement
2867 caused a defer, replacing the real cause of the deferral.
2869 16. If an IP address had so many PTR records that the DNS lookup buffer
2870 was not large enough to hold them, Exim could crash while trying to process
2871 the truncated data. It now detects and logs this case.
2873 17. Further to 4.21/58, another change has been made: if (and only if) the
2874 first line of a message (the first header line) ends with CRLF, a bare LF
2875 in a subsequent header line has a space inserted after it, so as not to
2876 terminate the header.
2878 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2879 unnecessarily, used atoi() instead of strtol(), and didn't check the
2880 termination when getting file sizes from file names by regex.
2882 19. Completely re-implemented the support for maildirsize files, in the light
2883 of a number of problems with the previous contributed implementation
2884 (4.30/29). In particular:
2886 . If the quota is zero, the maildirsize file is maintained, but no quota is
2889 . If the maildir directory does not exist, it is created before any attempt
2890 to write a maildirsize file.
2892 . The quota value in the file is just a cache; if the quota is changed in
2893 the transport, the new value overrides.
2895 . A regular expression is available for excluding directories from the
2898 20. The autoreply transport checks the characters in options that define the
2899 message's headers; it allows continued headers, but it was checking with
2900 isspace() after an embedded newline instead of explicitly looking for a
2903 21. If all the "regular" hosts to which an address was routed had passed their
2904 expiry times, and had not reached their retry times, the address was
2905 bounced, even if fallback hosts were defined. Now Exim should go on to try
2908 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2909 equivalent code in the SMTP transport. Some hosts send humungous responses
2910 to HELO/EHLO, more than 1024 it seems.
2912 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2913 gives compiler warnings in some environments. I've now done it "properly",
2916 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2917 (because of problems with the built-in one) was declared to return uschar *
2918 instead of char *, causing compiler failure.
2920 25. Fixed a file descriptor leak when processing alias/forward files.
2922 26. Fixed a minor format string issue in dbfn.c.
2924 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2926 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2927 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2930 29. When a sender address is verified, it is cached, to save repeating the test
2931 when there is more than one recipient in a message. However, when the
2932 verification involves a callout, it is possible for different callout
2933 options to be set for different recipients. It is too complicated to keep
2934 track of this in the cache, so now Exim always runs a verification when a
2935 callout is required, relying on the callout cache for the optimization.
2936 The overhead is duplication of the address routing, but this should not be
2939 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2940 to be verified with callout=postmaster, and the main callout worked but the
2941 postmaster check failed, the verification correctly failed. However, if a
2942 subsequent RCPT command asked for sender verification *without* the
2943 postmaster check, incorrect caching caused this verification also to fail,
2946 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2947 it was not caching the DNS options (qualify_single, search_parents) that
2948 were used when the lookup failed. A subsequent lookup with different
2949 options therefore always gave the same answer, though there were cases
2950 where it should not have. (Example: a "domains = !$mx_any" option on a
2951 dnslookup router: the "domains" option is always processed without any
2952 widening, but the router might have qualify_single set.) Now Exim uses the
2953 cached value only when the same options are set.
2955 32. Added John Jetmore's "exipick" utility to the distribution.
2957 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2958 than a timeout (e.g. a certificate is required, and is not provided), an
2959 Exim server now closes the connection immediately. Previously it waited for
2960 the client to close - but if the client is SSL, it seems that they each
2961 wait for each other, leading to a delay before one of them times out.
2963 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2964 maintained 0.8.x compatibility because I don't think many are using it, and
2965 it is clearly obsolete.
2967 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2970 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2971 client certificate was expired. A simple patch fixes this, though I don't
2972 understand the full logic of why the verify callback is called multiple
2975 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2976 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2977 which causes problems with some clients (such as the Certicom SSL Plus
2978 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2979 disables the coutermeasure allowing Eudora to connect."
2981 38. Exim was not checking that a write() to a log file succeeded. This could
2982 lead to Bad Things if a log got too big, in particular if it hit a file
2983 size limit. Exim now panics and dies if it cannot write to a log file, just
2984 as it does if it cannot open a log file.
2986 39. Modified OS/Makefile-Linux so that it now contains
2988 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2990 The two -D definitions ensure that Exim is compiled with large file
2991 support, which makes it possible to handle log files that are bigger than
2994 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2995 instance) a domain was checked against a named list that involved a lookup,
2996 causing $domain_data to be set, then another domain was checked against the
2997 same list, then the first domain was re-checked, the value of $domain_data
2998 after the final check could be wrong. In particular, if the second check
2999 failed, it could be set empty. This bug probably also applied to
3002 41. The strip_trailing_dot option was not being applied to the address given
3003 with the -f command-line option.
3005 42. The code for reading a message's header from the spool was incrementing
3006 $received_count, but never initializing it. This meant that the value was
3007 incorrect (doubled) while delivering a message in the same process in which
3008 it was received. In the most common configuration of Exim, this never
3009 happens - a fresh exec is done - but it can happen when
3010 deliver_drop_privilege is set.
3012 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3013 it now includes up to 150 characters of the unexpected data in the log
3016 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3017 and building data strings. The size of both of these buffers was 10 000
3018 bytes - far larger than anybody would *ever* want, thought I. Needless to
3019 say, somebody hit the limit. I have increased the maximum line length to
3020 20 000 and the maximum data length of concatenated lines to 100 000. I have
3021 also fixed two bugs, because there was no checking on these buffers. Tsk,
3022 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3023 buffer is too small.
3025 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3026 lsearch lookups. Now it does.
3028 46. When parsing a route_list item in a manualroute router, a fixed-length
3029 buffer was used for the list of hosts. I made this 1024 bytes long,
3030 thinking that nobody would ever have a list of hosts that long. Wrong.
3031 Somebody had a whole pile of complicated expansion conditions, and the
3032 string was silently truncated, leading to an expansion error. It turns out
3033 that it is easier to change to an unlimited length (owing to other changes
3034 that have happened since this code was originally written) than to build
3035 structure for giving a limitation error. The length of the item that
3036 expands into the list of hosts is now unlimited.
3038 47. The lsearch lookup could not handle data where the length of text line was
3039 more than 4095 characters. Such lines were truncated, leading to shortened
3040 data being returned. It should now handle lines of any length.
3042 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3043 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3046 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3047 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3048 no longer happens in comment lines. A list of replacements is now placed
3049 at the head of all of the source files, except those whose only change is
3050 to replace PERL_COMMAND in the very first #! line.
3052 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3053 messages on the queue, with a bottom-up merge sort, using code contributed
3054 by Michael Haardt. This should make operations like -bp somewhat faster on
3055 large queues. It won't affect queue runners, except when queue_run_in_order
3058 51. Installed eximstats 1.31 in the distribution.
3060 52. Added support for SRV lookups to the dnslookup router.
3062 53. If an ACL referred to $message_body or $message_body_end, the value was not
3063 reset for any messages that followed in the same SMTP session.
3065 54. The store-handling optimization for building very long strings was not
3066 differentiating between the different store pools. I don't think this
3067 actually made any difference in practice, but I've tidied it.
3069 55. While running the routers to verify a sender address, $sender_address
3070 was still set to the sender address. This is wrong, because when routing to
3071 send a bounce to the sender, it would be empty. Therefore, I have changed
3072 it so that, while verifying a sender address, $sender_address is set to <>.
3073 (There is no change to what happens when verifying a recipient address.)
3075 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3076 target A or AAAA records (if not already returned) without resetting the
3077 qualify_single or search_parents options of the DNS resolver. These are
3078 inappropriate in this case because the targets of MX and SRV records must
3079 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3080 target that, when qualified, matched something in the local domain. These
3081 two options are now turned off when doing these lookups.
3083 57. It seems that at least some releases of Reiserfs (which does not have the
3084 concept of a fixed number of inodes) returns zero and not -1 for the
3085 number of available inodes. This interacted badly with check_spool_inodes,
3086 which assumed that -1 was the "no such thing" setting. What I have done is
3087 to check that the total number of inodes is greater than zero before doing
3088 the test of how many are available.
3090 58. When a "warn" ACL statement has a log_message modifier, the message is
3091 remembered, and not repeated. This is to avoid a lot of repetition when a
3092 message has many recipients that cause the same warning to be written.
3093 Howewer, Exim was preserving the list of already written lines for an
3094 entire SMTP session, which doesn't seem right. The memory is now reset if a
3095 new message is started.
3097 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3098 debugging output unless log_rewrite was also set.
3100 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3101 of (int)(handle) when we know that handle contains (void *)(-1).
3103 61. The Exim daemon panic-logs an error return when it closes the incoming
3104 connection. However "connection reset by peer" seems to be common, and
3105 isn't really an error worthy of noting specially, so that particular error
3108 62. When Exim is trying to find all the local interfaces, it used to panic and
3109 die if the ioctl to get the interface flags failed. However, it seems that
3110 on at least one OS (Solaris 9) it is possible to have an interface that is
3111 included in the list of interfaces, but for which you get a failure error
3112 for this call. This happens when the interface is not "plumbed" into a
3113 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3114 failure of the "get flags" call assumes that the interface is down.
3116 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3117 makes life easier for people who are doing arithmetic on fields extracted
3118 from dates, where you often get leading zeros that should not be
3119 interpreted as octal.
3121 64. Added qualify_domain to the redirect router, to override the global
3124 65. If a pathologically long header line contained very many addresses (the
3125 report of this problem mentioned 10 000) and each of them was rewritten,
3126 Exim could use up a very large amount of memory. (It kept on making new
3127 copies of the header line as it rewrote, and never released the old ones.)
3128 At the expense of a bit more processing, the header rewriting function has
3129 been changed so that it no longer eats memory in this way.
3131 66. The generation of the Received: header has been moved from the time that a
3132 message starts to be received, to the time that it finishes. The timestamp
3133 in the Received: header should now be very close to that of the <= log
3134 line. There are two side-effects of this change:
3136 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3137 logged header lines no longer include the local Received: line, because
3138 it has not yet been created. The same applies to a copy of the message
3139 that is returned to a non-SMTP sender when a message is rejected.
3141 (b) When a filter file is tested using -bf, no additional Received: header
3142 is added to the test message. After some thought, I decided that this
3145 This change does not affect the value of $received_for. It is still set
3146 after address rewriting, but before local_scan() is called.
3148 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3150 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3151 gave an unhelpful panic error message, and a defer error. I have managed to
3152 change this behaviour so that it now rejects any supplied certificate,
3153 which seems right, as the list of acceptable certificates is empty.
3155 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3156 gave an unhelpful defer error. I have not managed to make this reject any
3157 supplied certificates, but the error message it gives is "no certificate
3158 supplied", which is not helpful.
3160 70. exigrep's output now also includes lines that are not associated with any
3161 message, but which match the given pattern. Implemented by a patch from
3162 Martin Sluka, which also tidied up the Perl a bit.
3164 71. Recipient callout verification, like sender verification, was using <> in
3165 the MAIL FROM command. This isn't really the right thing, since the actual
3166 sender may affect whether the remote host accepts the recipient or not. I
3167 have changed it to use the actual sender in the callout; this means that
3168 the cache record is now keyed on a recipient/sender pair, not just the
3169 recipient address. There doesn't seem to be a real danger of callout loops,
3170 since a callout by the remote host to check the sender would use <>.
3171 [SEE ABOVE: changed after hitting problems.]
3173 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3174 temporary errors. However, in the case of such a code being given after
3175 the end of a data transmission (i.e. after ".") Exim was failing to write
3176 a retry record for the message. (Yes, there was some broken host that was
3177 actually sending 8xx at this point.)
3179 73. An unknown lookup type in a host list could cause Exim to panic-die when
3180 the list was checked. (An example that provoked this was putting <; in the
3181 middle of a list instead of at the start.) If this happened during a DATA
3182 ACL check, a -D file could be left lying around. This kind of configuration
3183 error no longer causes Exim to die; instead it causes a defer errror. The
3184 incident is still logged to the main and panic logs.
3186 74. Buglet left over from Exim 3 conversion. The message "too many messages
3187 in one connection" was written to the rejectlog but not the mainlog, except
3188 when address rewriting (yes!) was being logged.
3190 75. Added write_rejectlog option.
3192 76. When a system filter was run not as root (that is, when system_filter_user
3193 was set), the values of the $n variables were not being returned to the
3194 main process; thus, they were not subsequently available in the $sn
3197 77. Added +return_path_on_delivery log selector.
3199 78. A connection timeout was being treated differently from recipients deferred
3200 when testing hosts_max_try with a message that was older than the host's
3201 retry timeout. (The host should not be counted, thus allowing all hosts to
3202 be tried at least once before bouncing.) This may have been the cause of an
3203 occasionally reported bug whereby a message would remain on the queue
3204 longer than the retry timeout, but would be bounced if a delivery was
3205 forced. I say "may" because I never totally pinned down the problem;
3206 setting up timeout/retry tests is difficult. See also the next item.
3208 79. The ultimate address timeout was not being applied to errors that involved
3209 a combination of host plus message (for example, a timeout on a MAIL
3210 command). When an address resolved to a number of possible hosts, and they
3211 were not all tried for each delivery (e.g. because of hosts_max_try), a
3212 message could remain on the queue longer than the retry timeout.
3214 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3217 81. Fixed an obscure SMTP outgoing bug which required at least the following
3218 conditions: (a) there was another message waiting for the same server;
3219 (b) the server returned 5xx to all RCPT commands in the first message so
3220 that the message was not completed; (c) the server dropped the connection
3221 or gave a negative response to the RSET that Exim sends to abort the
3222 transaction. The observed case was a dropped connection after DATA that had
3223 been sent in pipelining mode. That is, the server had advertised PIPELINING
3224 but was not implementing it correctly. The effect of the bug was incorrect
3225 behaviour, such as trying another host, and this could lead to a crash.
3231 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3232 and daemon.c were passed as pointers to ints; they should have been
3233 pointers to socklen_t variables (which are typically unsigned ints).
3235 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3238 3. Fixed a really odd bug that affected only the testing scheme; patching a
3239 certain fixed string in the binary changed the value of another string that
3240 happened to be identical to the end of the original first string.
3242 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3243 name", it returns that address as the IP address. On some operating
3244 systems (e.g. Solaris), it also passes back the IP address string as the
3245 "host name". However, on others (e.g. Linux), it passes back an empty
3246 string. Exim wasn't checking for this, and was changing the host name to an
3247 empty string, assuming it had been canonicized.
3249 5. Although rare, it is permitted to have more than one PTR record for a given
3250 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3251 all the names associated with an address, because they do in Solaris.
3252 However, it seems that they do not in Linux for data that comes from the
3253 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3254 I found this out when I moved to a new Linux workstation and tried to run
3255 the Exim test suite.
3257 To get round this problem I have changed the code so that it now does its
3258 own call to the DNS to look up PTR records when searching for a host name.
3259 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3260 addresses that are only in /etc/hosts are still found.
3262 This behaviour is, however, controlled by an option called host_lookup_
3263 order, which defaults to "bydns:byaddr". If people want to use the other
3264 order, or indeed, just use one or the other means of lookup, they can
3265 specify it in this variable.
3267 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3268 some operating systems, this comes back from gethostbyaddr() as an empty
3269 string, and this is what Exim used to test for. However, it seems that in
3270 other systems, "." is yielded. Exim now tests for this case too.
3272 7. The values of check_spool_space and check_log_space are now held internally
3273 as a number of kilobytes instead of an absolute number of bytes. If a
3274 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3275 kilobyte. This means that much larger values can be stored.
3277 8. Exim monitor: an attempt to get the action menu when not actually pointing
3278 at a message produces an empty menu entitled "No message selected". This
3279 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3280 no entries in it ("Shell widget menu has zero width and/or height"). So I
3281 have added a single, blank menu entry in this case.
3283 9. Added ${quote_local_part.
3285 10. MIME decoding is now applied to the contents of Subject: header lines when
3288 11. Now that a reference to $sender_host_address automatically causes a reverse
3289 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3290 host lookup before query-style lookups in lists that might use this
3291 variable. This has therefore been abolished, and the "net-" prefix is no
3292 longer necessary for query-style lookups.
3294 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3295 have been a typo for LFLAGS, so it has been changed.
3297 13. The install script calls Exim with "-C /dev/null" in order to find the
3298 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3299 to be output. Howeve, since Exim outputs its version number before the
3300 error, it didn't break the script. It just looked ugly. I fixed this by
3301 always allowing "-C /dev/null" if the caller is root.
3303 14. Ignore overlarge ACL variable number when reading spool file - insurance
3304 against a later release with more variables having written the file.
3306 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3307 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3308 this, as well as the form without the "IPv6" on the front.
3310 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3311 OS/Makefile-Darwin file.
3313 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3314 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3316 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3317 variable. However, some OS use ints for inodes. I've added cast to long int
3318 to get rid of the compiler warning.
3320 19. I had forgotten to lock out "/../" in configuration file names when
3321 ALT_CONFIG_PREFIX was set.
3323 20. Routers used for verification do not need to specify transports. However,
3324 if such a router generated a host list, and callout was configured, Exim
3325 crashed, because it could not find a port number from the (non-existent)
3326 transport. It now assumes port 25 in this circumstance.
3328 21. Added the -t option to exigrep.
3330 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3331 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3332 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3333 them is set, LOOKUP_LSEARCH is forced.
3335 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3337 24. Added sender and host information to the "rejected by local_scan()" log
3338 line; previously there was no indication of these.
3340 25. Added .include_if_exists.
3342 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3343 Linux. It turns out that not all file systems support this. Apparently some
3344 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3345 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3346 not supported on the file descriptor, is now ignored when Exim is trying to
3347 sync a directory. This applies only to Linux.
3349 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3351 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3352 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3355 29. Added support for maildirsize files from supplied patch (modified a bit).
3357 30. The use of :fail: followed by an empty string could lead Exim to respond to
3358 sender verification failures with (e.g.):
3360 550 Verification failed for <xxx>
3361 550 Sender verify failed
3363 where the first response line was missing the '-' that indicates it is not
3364 the final line of the response.
3366 31. The loop for finding the name of the user that called Exim had a hardwired
3367 limit of 10; it now uses the value of finduser_retries, which is used for
3368 all other user lookups.
3370 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3373 33. Exim was neglecting to zero errno before one call of strtol() when
3374 expanding a string and expecting an integer value. On some systems this
3375 resulted in spurious "integer overflow" errors. Also, it was casting the
3376 result into an int without checking.
3378 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3379 did not work. The code looks as if it has *never* worked, though it appears
3380 to have been documented since at least releast 1.62. I have made it work.
3382 35. The "timeout_DNS" error in retry rules, also documented since at least
3383 1.62, also never worked. As it isn't clear exactly what this means, and
3384 clearly it isn't a major issue, I have abolished the feature by treating it
3385 as "timeout", and writing a warning to the main and panic logs.
3387 36. The display of retry rules for -brt wasn't always showing the error code
3390 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3391 timeout_connect_A, timeout_connect_MX.
3393 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3394 to the empty sender.
3396 39. The daemon was not analysing the content of -oX till after it had closed
3397 stderr and disconnected from the controlling terminal. This meant that any
3398 syntax errors were only noted on the panic log, and the return code from
3399 the command was 0. By re-arranging the code a little, I've made the
3400 decoding happen first, so such errors now appear on stderr, and the return
3401 code is 1. However, the actual setting up of the sockets still happens in
3402 the disconnected process, so errors there are still only recorded on the
3405 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3406 connections (as happens on some IP stacks) was logged at start up time as
3407 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3408 it from "IPv6 and IPv4", which means that two separate sockets are being
3411 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3412 says whether AF_INET or AF_INET6 was passed as an argument.
3414 42. Exiwhat output was messed up when time zones were included in log
3417 43. Exiwhat now gives more information about the daemon's listening ports,
3418 and whether -tls-on-connect was used.
3420 44. The "port" option of the smtp transport is now expanded.
3422 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3423 silently ignored. Now an error message is written to the main and panic
3426 46. There's a new ACL modifier called "logwrite" which writes to a log file
3427 as soon as it is encountered.
3429 47. Added $local_user_uid and $local_user_gid at routing time.
3431 48. Exim crashed when trying to verify a sender address that was being
3434 49. Exim was recognizing only a space character after ".include". It now also
3435 recognizes a tab character.
3437 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3438 extracting the relevant information from the specification. The man page no
3439 longer contains scrambled data for the -d option, and I've added a section
3440 at the front about calling Exim under different names.
3442 51. Added "extra_headers" argument to the "mail" command in filter files.
3444 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3447 53. Installed eximstats 1.29.
3449 54. Added transport_filter_timeout as a generic transport option.
3451 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3452 Cc: header lines. This was required by RFC 822, but it not required by RFC
3455 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3456 incoming messages that did not have them. Now it does so only if the
3457 message originates locally, that is, if there is no associated remote host
3458 address. When Resent- header lines are present, this applies to the Resent-
3459 lines rather than the non-Resent- lines.
3461 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3462 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3464 58. Messages for configuration errors now include the name of the main
3465 configuration file - useful now that there may be more than one file in a
3466 list (.included file names were always shown).
3468 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3469 for those rare installations that do not start the daemon as root or run it
3470 setuid root. I've cut out the call to initgroups() if the daemon is not
3473 60. The Exim user and group can now be bound into the binary as text strings
3474 that are looked up at the start of Exim's processing.
3476 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3478 62. Added $mailstore_basename variable.
3480 63. Installed patch to sieve.c from Michael Haardt.
3482 64. When Exim failed to open the panic log after failing to open the main log,
3483 the original message it was trying to log was written to stderr and debug
3484 output, but if they were not available (the usual case in production), it
3485 was lost. Now it is written to syslog before the two lines that record the
3486 failures to open the logs.
3488 65. Users' Exim filters run in subprocesses under the user's uid. It is
3489 possible for a "deliver" command or an alias in a "personal" command to
3490 provoke an address rewrite. If logging of address rewriting is configured,
3491 this fails because the process is not running as root or exim. There may be
3492 a better way of dealing with this, but for the moment (because 4.30 needs
3493 to be released), I have disabled address rewrite logging when running a
3494 filter in a non-root, non-exim process.
3500 1. The buildconfig auxiliary program wasn't quoting the value set for
3501 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3502 not defined. This bug was masked in 4.22 by the effect that was fixed in
3505 2. Some messages that were rejected after a message id was allocated were
3506 shown as "incomplete" by exigrep. It no longer does this for messages that
3507 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3509 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3510 have allow_domain_literals set, the ID did not get logged in the <= line.
3511 Domain literals are now always recognized in Message-ID: header lines.
3513 4. The first argument for a ${extract expansion item is the key name or field
3514 number. Leading and trailing spaces in this item were not being ignored,
3515 causing some misleading effects.
3517 5. When deliver_drop_privilege was set, single queue runner processes started
3518 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3519 same command in the process it spins off) were not dropping privilege.
3521 6. When the daemon running as "exim" started a queue runner, it always
3522 re-executed Exim in the spun-off process. This is a waste of effort when
3523 deliver_drop_privilege is set. The new process now just calls the
3524 queue-runner function directly.
3530 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3533 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3534 ignored. Though the use of -oP was forcing the writing of a pid file, it
3535 was always written to the default place.
3537 3. If the message "no IP address found for host xxxx" is generated during
3538 incoming verification, it is now followed by identification of the incoming
3539 connection (so you can more easily find what provoked it).
3541 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3543 5. Added some features to "harden" Exim a bit more against certain attacks:
3545 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3546 be put in Local/Makefile. This is like the never_users runtime option,
3547 but it cannot be overridden. The default setting is "root".
3549 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3550 prefix string with which any file named in a -C command line option
3553 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3554 is retained for -C and -D only if the caller of Exim is root. Without
3555 it, the exim user may also use -C and -D and retain privilege.
3557 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3558 command line option is disabled.
3560 6. Macro names set by the -D option must start with an upper case letter, just
3561 like macro names defined in the configuration file.
3563 7. Added "dereference=" facility to LDAP.
3565 8. Two instances of the typo "uknown" in the source files are fixed.
3567 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3568 the Configure-Makefile script screwed up while processing it.
3570 10. Incorporated PCRE 4.4.
3572 11. The SMTP synchronization check was not operating right at the start of an
3573 SMTP session. For example, it could not catch a HELO sent before the client
3574 waited for the greeting. There is now a check for outstanding input at the
3575 point when the greeting is written. Because of the duplex, asynchronous
3576 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3577 way, but not yet received, when the check is performed.
3579 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3580 on TCP/IP sockets, because this apparently causes some broken clients to
3583 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3584 unchanged) from the Cygwin maintainer.
3586 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3587 was supported instead of testing for mbx. Effectively a typo.
3589 15. The spa authenticator server code was not checking that the input it
3590 received was valid base64.
3592 16. The debug output line for the "set" modifier in ACLs was not showing the
3593 name of the variable that was being set.
3595 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3597 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3598 lookup on its own. Something else was needed to trigger the lookup. For
3599 example, a match in host_lookup or the need for a host name in a host list.
3600 Now, if $sender_host_name is referenced and the host name has not yet been
3601 looked up, a lookup is performed. If the lookup fails, the variable remains
3602 empty, and $host_lookup_failed is set to "1".
3604 19. Added "eqi" as a case-independent comparison operator.
3606 20. The saslauthd authentication condition could segfault if neither service
3607 nor realm was specified.
3609 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3610 error message that was logged was misleading, and incoming SMTP
3611 connections were dropped. The message is now more accurate, and temporary
3612 errors are given to SMTP connections.
3614 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3615 (except RSET) with a 421 error, until QUIT is received. However, it was
3616 failing to send a response to QUIT.
3618 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3619 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3620 ACL is now run after the helo_try_verify_hosts code.
3622 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3623 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3624 case-independent, so other case variants are also recognized). Apparently
3625 some systems use these upper case variants.
3627 25. If more than two messages were waiting for the same host, and a transport
3628 filter was specified for the transport, Exim sent two messages over the
3629 same TCP/IP connection, and then failed with "socket operation on non-
3630 socket" when it tried to send the third.
3632 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3634 27. The extern definition of crypt16() in expand.c was not being excluded when
3635 the OS had its own crypt16() function.
3637 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3638 as a preferred synonym for return_size_limit, both as an option and as an
3641 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3643 30. Changed the default configuration ACL to relax the local part checking rule
3644 for addresses that are not in any local domains. For these addresses,
3645 slashes and pipe symbols are allowed within local parts, but the sequence
3646 /../ is explicitly forbidden.
3648 31. SPA server authentication was not clearing the challenge buffer before
3651 32. log_message in a "warn" ACL statement was writing to the reject log as
3652 well as to the main log, which contradicts the documentation and doesn't
3653 seem right (because no rejection is happening). So I have stopped it.
3655 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3656 However, I am unable to do any testing of this.
3658 34. Fixed an infelicity in the appendfile transport. When checking directories
3659 for a mailbox, to see if any needed to be created, it was accidentally
3660 using path names with one or more superfluous leading slashes; tracing
3661 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3663 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3664 contents are added to the log line that is written for every discarded
3665 recipient. (Previously a log_message setting was ignored.)
3667 36. The ${quote: operator now quotes the string if it is empty.
3669 37. The install script runs exim in order to find its version number. If for
3670 some reason other than non-existence or emptiness, which it checks, it
3671 could not run './exim', it was installing it with an empty version number,
3672 i.e. as "exim-". This error state is now caught, and the installation is
3675 38. An argument was missing from the function that creates an error message
3676 when Exim fails to connect to the socket for saslauthd authentication.
3677 This could cause Exim to crash, or give a corrupted message.
3679 39. Added isip, isip4, and isip6 to ${if conditions.
3681 40. The ACL variables $acl_xx are now saved with the message, and can be
3682 accessed later in routers, transports, and filters.
3684 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3685 strings in the file are not string-expanded.
3687 42. If a MAIL command specified a SIZE value that was too large to fit into an
3688 int variable, the check against message_size_limit failed. Such values are
3689 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3690 day this will have to be increased, but I don't think I want to be around
3691 when emails are that large.
3698 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3699 iconv() is not standard in FreeBSD.
3701 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3702 IPv6 enabled. The observed symptom was a segmentation fault on return from
3703 the function os_common_find_running_interfaces() in src/os.c.
3705 3. In the check_special_case() function in daemon.c I had used "errno" as an
3706 argument name, which causes warnings on some systems. This was basically a
3707 typo, since it was named "eno" in the comments!
3709 4. The code that waits for the clock to tick (at a resolution of some fraction
3710 of a second) so as to ensure message-id uniqueness was always waiting for
3711 at least one whole tick, when it could have waited for less. [This is
3712 almost certainly not relevant at current processor speeds, where it is
3713 unlikely to ever wait at all. But we try to future-proof.]
3715 5. The function that sleeps for a time interval that includes fractions of a
3716 second contained a race. It did not block SIGALRM between setting the
3717 timer, and suspending (a couple of lines later). If the interval was short
3718 and the sigsuspend() was delayed until after it had expired, the suspension
3719 never ended. On busy systems this could lead to processes getting stuck for
3722 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3723 process, before it forks any delivery processes. The open lookup caching
3724 mechanism meant that the open file or database connection was passed into
3725 the delivery process. The problem was that delivery processes always tidy
3726 up cached lookup data. This could cause a problem for the next delivery
3727 process started by the queue runner, because the external queue runner
3728 process does not know about the closure. So the next delivery process
3729 still has data in the lookup cache. In the case of a file lookup, there was
3730 no problem because closing a file descriptor in a subprocess doesn't affect
3731 the parent. However, if the lookup was caching a connection to a database,
3732 the connection was closed, and the second delivery process was likely to
3733 see errors such as "PGSQL: query failed: server closed the connection
3734 unexpectedly". The problem has been fixed by closing all cached lookups
3735 in a queue runner before running a delivery process.
3737 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3738 seem to have the "const" qualifier which it has on other OS. I've
3741 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3742 *of the same type* (client or server) with the same public name that an
3743 error should be diagnosed.
3745 9. When Exim looked up a host name for an IP address, but failed to find the
3746 original IP address when looking up the host name (a safety check), it
3747 output the message "<ip address> does not match any IP for NULL", which was
3748 confusing, to say the least. The bug was that the host name should have
3749 appeared instead of "NULL".
3751 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3752 user that is built into the binary, and the -C or -D options is used, root
3753 privilege is dropped before the configuration file is read. In addition,
3754 logging is switched to stderr instead of the normal log files. If the
3755 configuration then re-defines the Exim user, the unprivileged environment
3756 is probably not what is expected, so Exim logs a panic warning message (but
3759 However, if deliver_drop_privilege is set, the unprivileged state may well
3760 be exactly what is intended, so the warning has been cut out in that case,
3761 and Exim is allowed to try to write to its normal log files.
3767 1. smtp_return_error_details was not giving details for temporary sender
3768 or receiver verification errors.
3770 2. Diagnose a configuration error if two authenticators have the same public
3773 3. Exim used not to create the message log file for a message until the first
3774 delivery attempt. This could be confusing when incoming messages were held
3775 for policy or load reasons. The message log file is now created at the time
3776 the message is received, and an initial "Received" line is written to it.
3778 4. The automatically generated man page for command line options had a minor
3779 bug that caused no ill effects; however, a more serious problem was that
3780 the procedure for building the man page automatically didn't always
3781 operate. Consequently, release 4.20 contains an out-of-date version. This
3782 shouldn't happen again.
3784 5. When building Exim with embedded Perl support, the script that builds the
3785 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3786 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3788 6. The freeze_tell option was not being used for messages that were frozen on
3789 arrival, either by an ACL or by local_scan().
3791 7. Added the smtp_incomplete_transaction log selector.
3793 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3794 was accepting AUTH without a new EHLO.
3796 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3797 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3800 10. Logging of TCP/IP connections (when configured) now happens in the main
3801 daemon process instead of the child process, so that the TCP/IP connection
3802 count is more accurate (but it can never be perfect).
3804 11. The use of "drop" in a nested ACL was not being handled correctly in the
3805 outer ACL. Now, if condition failure induced by the nested "drop" causes
3806 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3807 or "require"), the connection is dropped.
3809 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3810 that yield "discard" can now be used with an "accept" or a "discard" verb,
3811 but an error is generated for any others (because I can't see a useful way
3812 to define what should happen).
3814 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3815 are now processed in the same way as lines in the Exim configuration file.
3816 In particular, continuation lines are supported.
3818 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3820 15. Added -ti meaning -t -i.
3822 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3823 domains, and warn by logging if others are found.
3825 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3826 returned from GIFCONF when Exim is trying to find the list of interfaces on
3827 a host. The code in os.c has been modified to copy each ifreq to an aligned
3828 structure in all cases.
3830 Also, in some cases, the returned ifreq's were being copied to a 'struct
3831 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3832 means the last couple of bytes of an IPv6 address could be chopped if the
3833 ifreq contained only a normal sockaddr (14 bytes storage).
3835 18. Named domain lists were not supported in the hosts_treat_as_local option.
3836 An entry such as +xxxx was not recognized, and was treated as a literal
3839 19. Ensure that header lines added by a DATA ACL are included in the reject log
3840 if the ACL subsequently rejects the message.
3842 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3843 MD5 (which is deprecated).
3845 21. When testing a filter file using -bf, Exim was writing a message when it
3846 took the sender from a "From " line in the message, but it was not doing so
3847 when it took $return_path from a Return-Path: header line. It now does.
3849 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3850 with a valid header line field name (a series of printing characters
3851 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3853 23. Changed "disc" in the source to "disk" to conform to the documentation and
3854 the book and for uniformity.
3856 24. Ignore Sendmail's -Ooption=value command line item.
3858 25. When execve() failed while trying to run a command in a pipe transport,
3859 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3860 could be confused with a return value of 69 from the command itself. This
3861 has been changed to 127, the value the shell returns if it is asked to run
3862 a non-existent command. The wording for the related log line suggests a
3863 non-existent command as the problem.
3865 26. If received_header_text expands to an empty string, do not add a Received:
3866 header line to the message. (Well, it adds a token one on the spool, but
3867 marks it "old" so that it doesn't get used or transmitted.)
3869 27. Installed eximstats 1.28 (addition of -nt option).
3871 28. There was no check for failure on the call to getsockname() in the daemon
3872 code. This can fail if there is a shortage of resources on the system, with
3873 ENOMEM, for example. A temporary error is now given on failure.
3875 29. Contrary to the C standard, it seems that in some environments, the
3876 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3877 program. Exim now does this explicitly; it affects the formatting of
3878 timestamps using strftime().
3880 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3881 complaints. I've now initialized all the variables, to avoid this.
3883 32. Header lines added by a system filter were not being "seen" during
3884 transport-time rewrites.
3886 33. The info_callback() function passed to OpenSSL is set up with type void
3887 (*)(SSL *, int, int), as described somewhere. However, when calling the
3888 function (actually a macro) that sets it up, the type void(*)() is
3889 expected. I've put in a cast to prevent warnings from picky compilers.
3891 34. If a DNS black list lookup found a CNAME record, but there were no A
3892 records associated with the domain it pointed at, Exim crashed.
3894 35. If a DNS black list lookup returned more than one A record, Exim ignored
3895 all but the first. It now scans all returned addresses if a particular IP
3896 value is being sought. In this situation, the contents of the
3897 $dnslist_value variable are a list of all the addresses, separated by a
3900 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3901 to accept a host name and all its aliases if the forward lookup for any of
3902 them yielded the IP address of the incoming connection. Now it accepts only
3903 those names whose forward lookup yields the correct IP address. Any other
3904 names are discarded. This closes a loophole whereby a rogue DNS
3905 administrator could create reverse DNS records to break through a
3906 wildcarded host restriction in an ACL.
3908 37. If a user filter or a system filter that ran in a subprocess used any of
3909 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3910 the wrong values were passed to the pipe command ($thisaddress had the
3911 value of $0, $0 had the value of $1, etc). This bug was introduced by
3912 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3914 38. Improved the line breaking for long SMTP error messages from ACLs.
3915 Previously, if there was no break point between 40 and 75 characters, Exim
3916 left the rest of the message alone. Two changes have been made: (a) I've
3917 reduced the minimum length to 35 characters; (b) if it can't find a break
3918 point between 35 and 75 characters, it looks ahead and uses the first one
3919 that it finds. This may give the occasional overlong line, but at least the
3920 remaining text gets split now.
3922 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3923 file descriptors might be low, and that setting 1000 would always raise it.
3924 It turns out that in some environments, the limit is already over 1000 and
3925 that lowering it causes trouble. So now Exim takes care not to decrease it.
3927 40. When delivering a message, the value of $return_path is set to $sender_
3928 address at the start of routing (routers may change the value). By an
3929 oversight, this default was not being set up when an address was tested by
3930 -bt or -bv, which affected the outcome if any router or filter referred to
3933 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3934 log only when "message" or "log_message" are set. However, if one of the
3935 conditions was an address verification, or a call to a nested ACL, the
3936 messages generated by the underlying test were being passed through. This
3937 no longer happens. The underlying message is available in $acl_verify_
3938 message for both "message" and "log_message" expansions, so it can be
3939 passed through if needed.
3941 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3942 new expansion $bh_ to give the encoded byte string without charset
3943 translation. Translation happens only if iconv() is available; HAVE_ICONV
3944 indicates this at build time. HEADERS_CHARSET gives the charset to
3945 translate to; headers_charset can change it in the configuration, and
3946 "headers charset" can change it in an individual filter file.
3948 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3949 that creates RFC 2047 encoded "words" labels them as that charset instead
3950 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3951 expansion operator; (ii) when Exim creates a From: line for a local
3952 message; (iii) when a header line is rewritten to include a "phrase" part.
3954 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3955 buggy, causing it to skip the first lines of messages whose message ID
3956 ended in 'D'. This would not have bitten before Exim release 4.14, because
3957 message IDs were unlikely to end in 'D' before then. The effect was to have
3958 incorrect size information for certain domains.
3960 45. #include "config.h" was missing at the start of the crypt16.c module. This
3961 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3964 46. If there was a timeout during a "random" callout check, Exim treated it as
3965 a failure of the random address, and carried on sending RSET and the real
3966 address. If the delay was just some slowness somewhere, the response to the
3967 original RCPT would be taken as a response to RSET and so on, causing
3968 mayhem of various kinds.
3970 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3971 when I implemented it. It didn't allow for the fact that some option values
3972 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3973 the right test for positive values.
3975 48. Domain names in DNS records are case-independent. Exim always looks them up
3976 in lower case. Some resolvers return domain names in exactly the case they
3977 appear in the zone file, that is, they may contain uppercase letters. Not
3978 all resolvers do this - some return always lower case. Exim was treating a
3979 change of case by a resolver as a change of domain, similar to a widening
3980 of a domain abbreviation. This triggered its re-routing code and so it was
3981 trying to route what was effectively the same domain again. This normally
3982 caused routing to fail (because the router wouldn't handle the domain
3983 twice). Now Exim checks for this case specially, and just changes the
3984 casing of the domain that it ultimately uses when it transmits the message
3987 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3990 50. If a filter generated a file delivery with a non-absolute name (possible if
3991 no home directory exists for the router), the forbid_file option was not
3994 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3995 the existing equality matching.
3997 52. Exim was using ints instead of ino_t variables in some places where it was
3998 dealing with inode numbers.
4000 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4001 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4002 called TMPDIR, and if it finds it is different, it changes its value.
4004 54. The smtp_printf() function is now made available to local_scan() so
4005 additional output lines can be written before returning. There is also an
4006 smtp_fflush() function to enable the detection of a dropped connection.
4007 The variables smtp_input and smtp_batched_input are exported to
4010 55. Changed the default runtime configuration: the message "Unknown user"
4011 has been removed from the ACL, and instead placed on the localuser router,
4012 using the cannot_route_message feature. This means that any verification
4013 failures that generate their own messages won't get overridden. Similarly,
4014 the "Unrouteable address" message that was in the ACL for unverifiable
4015 relay addresses has also been removed.
4017 56. Added hosts_avoid_esmtp to the smtp transport.
4019 57. The exicyclog script was not checking for the esoteric option
4020 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4021 will work only if exicyclog is run under the appropriate euid.
4023 58. Following a discussion on the list, the rules by which Exim recognises line
4024 endings on incoming messages have been changed. The -dropcr and drop_cr
4025 options are now no-ops, retained only for backwards compatibility. The
4026 following line terminators are recognized: LF CRLF CR. However, special
4027 processing applies to CR:
4029 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4030 nor a local message in the state where . is a terminator.
4032 (ii) If a bare CR is encountered in a header line, an extra space is added
4033 after the line terminator so as not to end the header. The reasoning
4034 behind this is that bare CRs in header lines are most likely either
4035 to be mistakes, or people trying to play silly games.
4037 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4038 was being incorrectly given as 18 bytes larger than it should have been.
4039 This is a VOB (very old bug).
4041 60. This may never have affected anything current, but just in case it has:
4042 When the local host is found other than at the start of a list of hosts,
4043 the local host, those with the same MX, and any that follow, are discarded.
4044 When the list in question was part of a longer list of hosts, the following
4045 hosts (not currently being processed) were also being discarded. This no
4046 longer happens. I'm not sure if this situation could ever has previously
4049 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4052 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4053 header. This is something that is recommended in a new Internet Draft, and
4054 is something that is documented as being done by Sendmail. There are two
4055 possible values. For messages generated by the autoreply transport, Exim
4058 Auto-Submitted: auto-replied
4060 whereas for all other generated messages (e.g. bounces) it adds
4062 Auto-Submitted: auto-generated
4064 63. The "personal" condition in filters now includes a test for the
4065 Auto-Submitted: header. If it contains the string "auto-" the message it
4066 not considered personal.
4068 64. Added rcpt_include_affixes as a generic transport option.
4070 65. Added queue_only_override (default true).
4072 66. Added the syslog_duplication option.
4074 67. If what should have been the first header line of a message consisted of
4075 a space followed by a colon, Exim was mis-interpreting it as a header line.
4076 It isn't of course - it is syntactically invalid and should therefore be
4077 treated as the start of the message body. The misbehaviour could have
4078 caused a number of strange effects, including loss of data in subsequent
4079 header lines, and spool format errors.
4081 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4082 client host had authenticated. This control can now be exercised by an ACL
4083 for more flexibility.
4085 69. By default, callouts do not happen when testing with -bh. There is now a
4086 variant, -bhc, which does actually run the callout code, including
4087 consulting and updating the callout cache.
4089 70. Added support for saslauthd authentication, courtesy of Alexander
4092 71. If statvfs() failed on the spool or log directories while checking their
4093 size for availability, Exim confusingly gave the error "space shortage".
4094 Furthermore, in debugging mode it crashed with a floating point exception.
4095 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4096 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4097 serious problem, Exim now writes to the main and panic logs when this
4098 happens, with details of the failure. It then refuses to accept the
4099 incoming message, giving the message "spool directory problem" or "log
4100 directory problem" with a 421 code for SMTP messages.
4102 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4103 0, 1, and 2 exist, because some OS complain for execs without them (see
4104 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4105 descriptors. However, the code omitted to check that the open succeeded,
4106 causing mysterious errors if for some reason the permissions on /dev/null
4107 got screwed. Now Exim writes a message to the main and panic logs, and
4108 bombs out if it can't open /dev/null.
4110 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4111 interact so that it is all more flexible. It is supposed to remain
4112 backwards compatible. Also added extra_local_interfaces.
4114 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4115 to bomb out with an assertion failure - to the client this appears as a
4116 connection drop. This problem occurs in the part of the code that was taken
4117 from the Samba project. Fortunately, the assertion is in a very simple
4118 function, so I have fixed this by reproducing the function inline in the
4119 one place where it is called, and arranging for authentication to fail
4120 instead of killing the process with assert().
4122 75. The SPA client code was not working when the server requested OEM rather
4123 than Unicode encoding.
4125 76. Added code to make require_files with a specific uid setting more usable in
4126 the case where statting the file as root fails - usually a non-root-mounted
4127 NFS file system. When this happens and the failure is EACCES, Exim now
4128 forks a subprocess and does the per-uid checking as the relevant uid.
4130 77. Added process_log_path.
4132 78. If log_file_path was not explicitly set, a setting of check_log_space or
4133 check_log_inodes was ignored.
4135 79. If a space check for the spool or log partitions fails, the incident is now
4136 logged. Of course, in the latter case the data may get lost...
4138 80. Added the %p formatting code to string_format() so that it can be used to
4139 print addresses in debug_print(). Adjusted all the address printing in the
4140 debugging in store.c to use %p rather than %d.
4142 81. There was a concern that a line of code in smtp_in.c could overflow a
4143 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4144 initially expressed, the concern was not well-founded, because trailing
4145 spaces are removed early. However, if the trailing spaces were followed by
4146 a NULL, they did not get removed, so the overflow was possible. Two fixes
4149 (a) I re-wrote the offending code in a cleaner fashion.
4150 (b) If an incoming SMTP command contains a NULL character, it is rejected
4153 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4154 runs initgroups(), so that if the Exim user is in any additional groups,
4155 they will be used during message reception.
4161 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob