1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.392 2006/09/18 14:49:23 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
14 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
26 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
29 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
34 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
37 PH/05 Applied Nico Efrurth's refactoring patch to tidy up spool_mbox.c.
39 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
41 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
43 PH/08 An error is now given if message_size_limit is specified negative.
49 SC/01 Use a glob alias rather than an array ref in eximstats generated
50 parser. This improves both readability and performance.
52 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
53 Don't display local sender or destination tables in eximstats unless
54 there is data to show.
55 Added average volumes into the eximstats top table text output.
57 SC/03 Collect data on the number of addresses (recipients) as well
58 as the number of messages in eximstats.
60 TF/01 Correct an error in the documentation for the redirect router. Exim
61 does (usually) call initgroups() when daemonizing.
63 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
64 with consistent privilege compared to when running as a daemon.
66 TF/03 Note in the spec that $authenticated_id is not set for local
67 submissions from trusted users.
69 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
70 Thanks to Dean Brooks <dean@iglou.com> for the patch.
72 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
73 by adding some example configuration directives to the default
74 configuration file. A little bit of work is required to uncomment the
75 directives and define how usernames and passwords are checked, but
76 there is now a framework to start from.
78 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
79 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
80 without this. I don't know how relevant this is to other LDAP libraries.
82 PH/02 Add the verb name to the "unknown ACL verb" error.
84 PH/03 Magnus Holmgren's patch for filter_prepend_home.
86 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
88 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
89 directory not expanded when it should be if an expanded home directory
90 was set for the address (which is overridden by the transport).
92 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
95 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
96 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
97 because it is too late at that time, and has no effect.
99 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
100 security issue with \' (bugzilla #107). I could not use the
101 PQescapeStringConn() function, because it needs a PGconn value as one of
104 PH/08 When testing addresses using -bt, indicate those final addresses that
105 are duplicates that would not cause an additional delivery. At least one
106 person was confused, thinking that -bt output corresponded to deliveries.
107 (Suppressing duplicates isn't a good idea as you lose the information
108 about possibly different redirections that led to the duplicates.)
110 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
111 systems where poll() doesn't work, in particular OS X.
113 PH/10 Added more information to debugging output for retry time not reached.
115 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
116 operations in malware.c.
118 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
121 PH/13 If write_rejectlog was set false when logging was sent to syslog with
122 syslog_duplication set false, log lines that would normally be written
123 both the the main log and to the reject log were not written to syslog at
126 PH/14 In the default configuration, change the use of "message" in ACL warn
127 statements to "add_header".
129 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
130 not followed by a command (e.g. "seen endif").
132 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
133 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
136 PH/17 Added extra conditions to the default value of delay_warning_condition
140 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
141 { match{$h_precedence:}{(?i)bulk|list|junk} } \
142 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
145 The Auto-Submitted: and various List- headers are standardised, whereas I
146 don't think Precedence: ever was.
148 PH/18 Refactored debugging code in route_finduser() to show more information,
149 in particular, the error code if getpwnam() issues one.
151 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
152 This is apparently needed in addition to the PH/07 change above to avoid
153 any possible encoding problems.
155 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
156 but not after initializing Perl.
158 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
159 output them only if debugging. By default they are written stderr,
160 apparently, which is not desirable.
162 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
165 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
168 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
170 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
171 authenticated or an ident call has been made. Suppress the default
172 values for $authenticated_id and $authenticated_sender (but permit -oMai
173 and -oMas) when testing with -bh.
175 PH/24 Re-jigged the order of the tests in the default configuration so that the
176 tests for valid domains and recipients precede the DNS black list and CSA
177 tests, on the grounds that those ones are more expensive.
179 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
180 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
181 This bug exists in every version of Exim that I still have, right back to
184 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
185 However, an attempt to turn on case-sensitivity in a regex key by
186 including (?-i) didn't work because the subject string was already
187 lowercased, and the effects were non-intuitive. It turns out that a
188 one-line patch can be used to allow (?-i) to work as expected.
194 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
195 other effects) broke the use of negated acl sub-conditions.
197 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
200 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
201 "Deny" causes Exim to reject the incoming connection with a 554 error.
202 Unfortunately, if there is a major crisis, such as a disk failure,
203 tcp-wrappers gives "deny", whereas what one would like would be some
204 kind of temporary error. A kludge has been added to help with this.
205 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
206 554 error is used if errno is still zero or contains ENOENT (which occurs
207 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
210 PH/03 Add -lutil to the default FreeBSD LIBS setting.
212 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
213 errors. Otherwise a message that provokes a temporary error (when other
214 messages do not) can cause a whole host to time out.
216 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
217 addresses were routed directly to files or pipes from a redirect router.
218 File deliveries just didn't batch; pipe deliveries might have suffered
221 PH/06 A failure to get a lock for a hints database would erroneously always say
222 "Failed to get write lock", even when it was really a read lock.
224 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
225 of 0600. This has been changed to use the value of the lockfile_mode
226 option (which defaults to 0600).
228 PH/08 Applied small patch from the Sieve maintainer.
230 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
231 folder from quota calculations, a direct delivery into this folder messed
232 up the contents of the maildirsize file. This was because the regex was
233 used only to exclude .Trash (or whatever) when the size of the mailbox
234 was calculated. There was no check that a delivery was happening into an
235 excluded directory. This bug has been fixed by ignoring all quota
236 processing for deliveries into excluded directories.
238 PH/10 Added the maildirfolder_create_regex option to appendfile.
244 PH/01 The code for finding all the local interface addresses on a FreeBSD
245 system running IPv6 was broken. This may well have applied to all BSD
246 systems, as well as to others that have similar system calls. The broken
247 code found IPv4 interfaces correctly, but gave incorrect values for the
248 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
249 that it would not match correctly against @[] and not recognize the IPv6
252 PH/02 The ipliteral router was not recognizing addresses of the form user@
253 [ipv6:....] because it didn't know about the "ipv6:" prefix.
255 PH/03 Added disable_ipv6.
257 PH/04 Changed $reply_address to use the raw form of the headers instead of the
258 decoded form, because it is most often used to construct To: headers
259 lines in autoreplies, and the decoded form may well be syntactically
260 invalid. However, $reply_address has leading white space removed, and all
261 newlines turned into spaces so that the autoreply transport does not
264 PH/05 If group was specified without a user on a router, and no group or user
265 was specified on a transport, the group from the router was ignored.
267 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
268 for visible compile-time settings that can be used to change these
269 numbers, for those that want even more. Backwards compatibility with old
270 spool files has been maintained. However, going back to a previous Exim
271 release will lost any variables that are in spool files.
273 PH/07 Two small changes when running in the test harness: increase delay when
274 passing a TCP/IP connection to a new process, in case the original
275 process has to generate a bounce, and remove special handling of
276 127.0.0.2 (sic), which is no longer necessary.
278 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
279 be the same on different OS.
281 PH/09 Moved a debug statement in filter processing to avoid a race problem when
284 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
285 whether --show-vars was specified or not
287 JJ/02 exipick: Added support for new ACL variable spool format introduced
290 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
291 syntactically invalid From: or Reply-to: line, and a filter used this to
292 generate an autoreply, and therefore failed to obtain an address for the
293 autoreply, Exim could try to deliver to a non-existent relative file
294 name, causing unrelated and misleading errors. What now happens is that
295 it logs this as a hard delivery error, but does not attempt to create a
298 PH/11 The exinext utility has a -C option for testing purposes, but although
299 the given file was scanned by exinext itself; it wasn't being passed on
300 when Exim was called.
302 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
303 an end-of-file indication when reading a command response.
305 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
306 compiled. In many other places in Exim, IPv6 addresses are always
307 recognized, so I have changed this. It also means that IPv4 domain
308 literals of the form [IPV4:n.n.n.n] are now always recognized.
310 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
311 used if the router is not running as root, for example, when verifying at
312 ACL time, or when using -bh. The debugging output from this situation was
313 non-existent - all you got was a failure to exec. I have made two
316 (a) Failures to set uid/gid, the current directory, or a process leader
317 in a subprocess such as that created by queryprogram now generate
318 suitable debugging ouput when -d is set.
320 (b) The queryprogram router detects when it is not running as root,
321 outputs suitable debugging information if -d is set, and then runs
322 the subprocess without attempting to change uid/gid.
324 PH/15 Minor change to Makefile for building test_host (undocumented testing
327 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
328 additional section of a DNS packet that returns MX or SRV records.
329 Instead, it always explicitly searches for A/AAAA records. This avoids
330 major problems that occur when a DNS server includes only records of one
331 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
332 fixed another bug: if SRV records were looked up and the corresponding
333 address records were *not* found in the additional section, the port
334 values from the SRV records were lost.
336 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
337 using the correct key (the original address) when searching the retry
338 rules in order to find which one to use for generating the retry hint.
340 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
341 adding the default one. Similarly, if it contains a Reply-To: header, the
342 errors_reply_to option, if set, is not used.
344 PH/19 When calculating a retry time, Exim used to measure the "time since
345 failure" by looking at the "first failed" field in the retry record. Now
346 it does not use this if it is later than than the arrival time of the
347 message. Instead it uses the arrival time. This makes for better
348 behaviour in cases where some deliveries succeed, thus re-setting the
349 "first failed" field. An example is a quota failure for a huge message
350 when small messages continue to be delivered. Without this change, the
351 "time since failure" will always be short, possible causing more frequent
352 delivery attempts for the huge message than are intended.
353 [Note: This change was subsequently modified - see PH/04 for 4.62.]
355 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
356 $1, $2, $3) because the numerical variables can be reset during some
357 expansion items (e.g. "match"), thereby losing the authentication data.
359 PH/21 Make -bV show the size of off_t variables so that the test suite can
360 decide whether to run tests for quotas > 2G.
362 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
363 mailbox_size, and mailbox_filecount in the appendfile transport. If a
364 filecount value is greater than 2G or if a quota value is greater than 2G
365 on a system where the size of off_t is not greater than 4, a panic error
368 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
369 never match. The debug and -bh output now contains an explicit error
370 message indicating a malformed IPv4 address or mask.
372 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
373 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
376 PH/25 Do not write to syslog when running in the test harness. The only
377 occasion when this arises is a failure to open the main or panic logs
378 (for which there is an explicit test).
380 PH/26 Added the /no_tell option to "control=freeze".
382 PH/27 If a host name lookup failed very early in a connection, for example, if
383 the IP address matched host_lookup and the reverse lookup yielded a name
384 that did not have a forward lookup, an error message of the form "no IP
385 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
386 could be logged. Now it outputs the IP address instead of "NULL".
388 PH/28 An enabling patch from MH: add new function child_open_exim2() which
389 allows the sender and the authenticated sender to be set when
390 submitting a message from within Exim. Since child_open_exim() is
391 documented for local_scan(), the new function should be too.
393 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
394 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
395 results in an empty string is now treated as unset.
397 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
399 PH/31 Added sender_verify_fail logging option.
401 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
402 needed by RFC 822 but not by RFC 2822 was commented out. I have now
403 tidied the source and removed it altogether.
405 PH/33 When a queue run was abandoned because the load average was too high, a
406 log line was always written; now it is written only if the queue_run log
407 selector is set. In addition, the log line for abandonment now contains
408 information about the queue run such as the pid. This is always present
409 in "start" and "stop" lines but was omitted from the "abandon" line.
411 PH/34 Omit spaces between a header name and the colon in the error message that
412 is given when verify = headers_syntax fails (if there are lots of them,
413 the message gets confusing).
415 PH/35 Change the default for dns_check_names_pattern to allow slashes within
416 names, as there are now some PTR records that contain slashes. This check
417 is only to protect against broken name servers that fall over on strange
418 characters, so the fact that it applies to all lookups doesn't matter.
420 PH/36 Now that the new test suite is complete, we can remove some of the
421 special code in Exim that was needed for the old test suite. For example,
422 sorting DNS records because real resolvers return them in an arbitrary
423 order. The new test suite's fake resolver always returns records in the
426 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
427 bounces) except when queue_only is set, to avoid logging races between
428 the different processes.
430 PH/38 Panic-die if .include specifies a non-absolute path.
432 PH/39 A tweak to the "H" retry rule from its user.
434 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
435 a label. They prevented compilation on older perls.
437 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
438 a warning to be raised on newish perls.
440 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
441 on queue. Changes to match documented behaviour of showing count of
442 messages matching specified criteria.
444 PH/40 Changed the default ident timeout from 30s to 5s.
446 PH/41 Added support for the use of login_cap features, on those BSD systems
447 that have them, for controlling the resources used by pipe deliveries.
449 PH/42 The content-scanning code uses fopen() to create files in which to put
450 message data. Previously it was not paying any attention to the mode of
451 the files. Exim runs with umask(0) because the rest of the code creates
452 files with open(), and sets the required mode explicitly. Thus, these
453 files were ending up world-writeable. This was not a big issue, because,
454 being within the spool directory, they were not world-accessible. I have
455 created a function called modefopen, which takes an additional mode
456 argument. It sets umask(777), creates the file, chmods it to the required
457 mode, then resets the umask. All the relevant calls to fopen() in the
458 content scanning code have been changed to use this function.
460 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
461 to 24 hours. This avoids potential overflow problems when processing G
462 and H retry rules. I suspect nobody ever tinkers with this value.
464 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
466 PH/45 When the plaintext authenticator is running as a client, the server's
467 challenges are checked to ensure they are valid base64 strings. By
468 default, the authentication attempt is cancelled if an invalid string is
469 received. Setting client_ignore_invalid_base64 true ignores these errors.
470 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
471 they are received. Thus, the responses can be made to depend on the
472 challenges. If an invalid string is ignored, an empty string is placed in
475 PH/46 Messages that are created by the autoreply transport now contains a
476 References: header, in accordance with RFCs 2822 and 3834.
478 PH/47 Added authenticated_sender_force to the smtp transport.
480 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
482 PH/49 Installed latest patch from the Sieve maintainer.
484 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
485 was also set, the appendfile transport was unnecessarily scanning a
486 directory of message files (e.g. for maildir delivery) to find the count
487 of files (along with the size), even though it did not need this
488 information. It now does the scan only if it needs to find either the
489 size of the count of files.
491 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
493 PH/52 Two bugs concerned with error handling when the smtp transport is
496 (i) Exim was not creating retry information for temporary errors given
497 for individual recipients after the DATA command when the smtp transport
498 was used in LMTP mode. This meant that they could be retried too
499 frequently, and not timed out correctly.
501 (ii) Exim was setting the flag that allows error details to be returned
502 for LMTP errors on RCPT commands, but not for LMTP errors for individual
503 recipients that were returned after the DATA command.
505 PH/53 This is related to PH/52, but is more general: for any failing address,
506 when detailed error information was permitted to be returned to the
507 sender, but the error was temporary, then after the final timeout, only
508 "retry timeout exceeded" was returned. Now it returns the full error as
509 well as "retry timeout exceeded".
511 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
512 do this, and (what is worse) MTAs that accept it.
514 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
515 will now be deprecated.
517 PH/56 New os.c-cygwin from the Cygwin maintainer.
519 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
520 formats (previously only available in exim formats via -bpr, -bpru,
521 and -bpra. Now also available in native and exiqgrep formats)
523 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
524 with very large, slow to parse queues
526 JJ/08 exipick: added ! as generic prefix to negate any criteria format
528 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
530 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
531 responses to authentication challenges, though it was showing the
532 challenges; (ii) I've removed the CR characters from the debug output for
535 PH/58 Allow for the insertion of a newline as well as a space when a string
536 is turned into more than one encoded-word during RFC 2047 encoding. The
537 Sieve code now uses this.
539 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
540 data_4xx, lost_connection, tls_required.
542 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
543 message was being sent as an SMTP response.
545 PH/61 Add -l and -k options to exicyclog.
547 PH/62 When verifying, if an address was redirected to one new address, so that
548 verification continued, and the new address failed or deferred after
549 having set something in $address_data, the value of $address_data was not
550 passed back to the ACL. This was different to the case when no
551 redirection occurred. The value is now passed back in both cases.
553 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
554 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
555 use login_cap.h, so on its own it isn't the distinguishing feature. The
556 new name refers directly to the setclassresources() function.
558 PH/65 Added configuration files for NetBSD3.
560 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
562 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
563 is preferred over IPv4.
565 PH/68 The bounce_return_message and bounce_return_body options were not being
566 honoured for bounces generated during the reception of non-SMTP messages.
567 In particular, this applied to messages rejected by the ACL. This bug has
568 been fixed. However, if bounce_return_message is true and bounce_return_
569 body is false, the headers that are returned for a non-SMTP message
570 include only those that have been read before the error was detected.
571 (In the case of an ACL rejection, they have all been read.)
573 PH/69 The HTML version of the specification is now built in a directory called
574 spec_html instead of spec.html, because the latter looks like a path with
575 a MIME-type, and this confuses some software.
577 PH/70 Catch two compiler warnings in sieve.c.
579 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
580 function verify_get_ident() calls ip_connect() to connect a socket, but
581 if the "connect()" function timed out, ip_connect() used to close the
582 socket. However, verify_get_ident() also closes the socket later, and in
583 between Exim writes to the log, which may get opened at this point. When
584 the socket was closed in ip_connect(), the log could get the same file
585 descriptor number as the socket. This naturally causes chaos. The fix is
586 not to close the socket in ip_connect(); the socket should be closed by
587 the function that creates it. There was only one place in the code where
588 this was missing, in the iplookup router, which I don't think anybody now
589 uses, but I've fixed it anyway.
591 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
592 well as to direct DNS lookups. Otherwise the handling of names in host
593 lists is inconsistent and therefore confusing.
599 PH/01 Two changes to the default runtime configuration:
601 (1) Move the checks for relay_from_hosts and authenticated clients from
602 after to before the (commented out) DNS black list checks.
604 (2) Add control=submission to the relay_from_hosts and authenticated
605 clients checks, on the grounds that messages accepted by these
606 statements are most likely to be submissions.
608 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
610 (1) Generate an error if the third argument for the ${prvs expansion is
613 (2) Treat a missing third argument of ${prvscheck as if it were an empty
616 (3) Reset the variables that are obtained from the first argument of
617 ${prvscheck and used in the second argument before leaving the code,
618 because their memory is reclaimed, so using them afterwards may do
621 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
622 one (it's much easier than Tom thought :-).
624 (5) Because of (4), we can now allow for the use of $prvscheck_result
625 inside the third argument.
627 PH/03 For some reason, the default setting of PATH when running a command from
628 a pipe transport was just "/usr/bin". I have changed it to
631 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
632 anything to be listed in the output from -bV.
634 PH/05 When a filter generated an autoreply, the entire To: header line was
635 quoted in the delivery log line, like this:
637 => >A.N.Other <ano@some.domain> <original@ddress> ...
639 This has been changed so that it extracts the operative address. There
640 may be more than one such address. If so, they are comma-separated, like
643 => >ano@some.domain,ona@other.domain <original@ddress> ...
645 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
646 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
647 not being looked up in the rDNS to get a host name, Exim was showing the
648 IP address twice in Received: lines, even though the IP addresses were
649 identical. For example:
651 Received: from [1.2.3.4] (helo=[1.2.3.4])
653 However, if the real host name was known, it was omitting the HELO data
654 if it matched the actual IP address. This has been tidied up so that it
655 doesn't show the same IP address twice.
657 PH/07 When both +timestamp and +memory debugging was on, the value given by
658 $tod_xxx expansions could be wrong, because the tod_stamp() function was
659 called by the debug printing, thereby overwriting the timestamp buffer.
660 Debugging no longer uses the tod_stamp() function when +timestamp is set.
662 PH/08 When the original message was included in an autoreply transport, it
663 always said "this is a copy of the message, including all the headers",
664 even if body_only or headers_only was set. It now gives an appropriate
667 PH/09 Applied a patch from the Sieve maintainer which:
669 o fixes some comments
670 o adds the (disabled) notify extension core
671 o adds some debug output for the result of if/elsif tests
672 o points to the current vacation draft in the documentation
673 and documents the missing references header update
677 o fixes a bug in processing the envelope test (when testing
678 multiple envelope elements, the last element determinted the
681 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
682 Electronic Mail") by including:
684 Auto-submitted: auto-generated
686 in the messages that it generates (bounce messages and others, such as
687 warnings). In the case of bounce messages for non-SMTP mesages, there was
688 also a typo: it was using "Auto_submitted" (underscore instead of
689 hyphen). Since every message generated by Exim is necessarily in response
690 to another message, thes have all been changed to:
692 Auto-Submitted: auto-replied
694 in accordance with these statements in the RFC:
696 The auto-replied keyword:
698 - SHOULD be used on messages sent in direct response to another
699 message by an automatic process,
701 - MUST NOT be used on manually-generated messages,
703 - MAY be used on Delivery Status Notifications (DSNs) and Message
704 Disposition Notifications (MDNs),
706 - MUST NOT be used on messages generated by automatic or periodic
707 processes, except for messages which are automatic responses to
710 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
711 to the default Received: header definition.
713 PH/12 Added log selector acl_warn_skipped (default on).
715 PH/13 After a successful wildlsearch lookup, discard the values of numeric
716 variables because (a) they are in the wrong storage pool and (b) even if
717 they were copied, it wouldn't work properly because of the caching.
719 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
720 checking when decoding. Apparently there are clients that generate
721 overlong encoded strings. Why am I not surprised?
723 PH/15 If the first argument of "${if match_address" was not empty, but did not
724 contain an "@" character, Exim crashed. Now it writes a panic log message
725 and treats the condition as false.
727 PH/16 In autoreply, treat an empty string for "once" the same as unset.
729 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
730 extension "envelope-auth". The code is finished and in agreement with
731 other implementations, but there is no documentation so far and in fact,
732 nobody wrote the draft yet. This extension is currently #undef'ed, thus
733 not changing the active code.
735 Print executed "if" and "elsif" statements when debugging is used. This
736 helps a great deal to understand what a filter does.
738 Document more things not specified clearly in RFC3028. I had all this
739 sorted out, when out of a sudden new issues came to my mind. Oops."
741 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
744 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
745 canonical form (as documented). However, after a host name lookup from
746 the IP address, check_host() was doing a simple string comparison with
747 addresses acquired from the DNS when checking that the found name did
748 have the original IP as one of its addresses. Since any found IPv6
749 addresses are likely to be in abbreviated form, the comparison could
750 fail. Luckily, there already exists a function for doing the comparison
751 by converting both addresses to binary, so now that is used instead of
754 PH/20 There was another similar case to PH/19, when a complete host name was
755 given in a host list; looking up its IP address could give an abbreviated
756 form, whereas the current host's name might or might not be abbreviated.
757 The same fix has been applied.
763 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
764 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
767 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
768 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
770 PH/03 Typo: missing ".o" in src/pcre/Makefile.
772 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
773 header line, restrict the check to what is listed in RFCs 2369 and 2929.
774 Also, for "Auto-Submitted", treat anything other than "no" as
775 non-personal, in accordance with RFC 3834. (Previously it treated
776 anything starting "auto-" as non-personal.)
778 TF/01 The control=submission/name=... option had a problem with syntax
779 errors if the name included a slash character. The /name= option
780 now slurps the rest of the string, so it can include any characters
781 but it must come last in the list of options (after /sender_retain
784 PH/05 Some modifications to the interface to the fake nameserver for the new
792 TK/01 Added the "success_on_redirect" address verification option. See
793 NewStuff for rationale and an example.
795 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
797 PH/02 Patch to exigrep to allow it to work on syslog lines.
799 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
800 fread() to skip over the body file's header line, because in Cygwin the
801 header line is locked and is inaccessible.
803 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
804 co-exist for some time) to make it clear that it is the Exim ID that is
805 referenced, not the Message-ID: header line.
807 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
808 string_format() function, because snprintf() does not exist on all
811 PH/06 The use of forbid_filter_existstest now also locks out the use of the
812 ${stat: expansion item.
814 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
815 protocol synchronization error", to keep the pedants happy.
817 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
818 well as for IRIX systems, when gcc is being used. See the host.c source
821 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
823 PH/10 Named domain lists were not working if used in a queue_smtp_domains
826 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
827 transport and to the smtp transport in LMTP mode.
829 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
831 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
832 run a filter in a subprocess. This could lead to confusion in subsequent
833 lookups in the parent process. There should also be a search_tidyup() at
834 the end of the subprocess.
836 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
837 only if the host matched helo_try_verify_hosts, which caused the
838 verification to occur when the EHLO/HELO command was issued. The ACL just
839 tested the remembered result. Now, if a previous verification attempt has
840 not happened, "verify = helo" does it there and then.
842 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
844 TK/03 Fix log output including CR from clamd.
846 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
847 exist provoked a memory error which could cause a segfault.
849 PH/15 Installed PCRE 6.2
851 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
853 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
854 of the problem. Specifically, suggested +O2 rather than +O1 for the
857 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
859 PH/20 If a delivery was routed to a non-standard port by means of an SRV
860 record, the port was not correctly logged when the outgoing_port log
861 selector was set (it logged the transort's default port).
863 PH/21 Added support for host-specific ports to manualroute, queryprogram,
864 fallback_hosts, and "hosts" in the smtp transport.
866 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
867 host errors such as "Connection refused".
869 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
870 authentication with radiusclient 0.4.9:
872 - Error returned from rc_read_config was caught wrongly
873 - Username/password not passed on to radius server due to wrong length.
875 The presumption is that some radiusclient API changes for 4.51/PH/17
876 were not taken care of correctly. The code is still untested by me (my
877 Linux distribution still has 0.3.2 of radiusclient), but it was
878 contributed by a Radius user.
880 PH/24 When doing a callout, the value of $domain wasn't set correctly when
881 expanding the "port" option of the smtp transport.
883 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
884 while reading a MIME header. Thanks to Tom Hughes for a patch.
886 PH/24 Include config.h inside local_scan.h so that configuration settings are
889 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
890 that in an ACL for RCPT (for example), you can examine exactly what was
893 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
894 commands, but it was not correctly comparing the address with the actual
895 client host address. Thus, it would show the EHLO address in Received:
896 header lines when this was not necessary.
898 PH/27 Added the % operator to ${eval:}.
900 PH/28 Exim tries to create and chdir to its spool directory when it starts;
901 it should be ignoring failures (because with -C, for example, it has lost
902 privilege). It wasn't ignoring creation failures other than "already
905 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
908 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
909 because an input line was too long, either on its own, or by virtue of
910 too many continuations, the temporary file was not being removed, and the
911 return code was incorrect.
913 PH/31 Missing "BOOL" in function definition in filtertest.c.
915 PH/32 Applied Sieve patches from the maintainer.
917 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
919 PH/33 Added "verify = not_blind".
921 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
922 Local/Makefile (with some defaults set). These are used in built scripts
923 such as exicyclog, but they have never been used in the exim_install
924 script (though there are many overriding facilities there). I have
925 arranged that the exim_install script now takes note of these two
928 PH/35 Installed configuration files for Dragonfly.
930 PH/36 When a locally submitted message by a trusted user did not contain a
931 From: header, and the sender address was obtained from -f or from an SMTP
932 MAIL command, and the trusted user did not use -F to supply a sender
933 name, $originator_name was incorrectly used when constructing a From:
934 header. Furthermore, $originator_name was used for submission mode
935 messages from external hosts without From: headers in a similar way,
936 which is clearly wrong.
938 PH/37 Added control=suppress_local_fixups.
940 PH/38 When log_selector = +received_sender was set, and the addition of the
941 sender made the log line's construction buffer exactly full, or one byte
942 less than full, an overflow happened when the terminating "\n" was
945 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
946 when the result of a list match is failure because a DNS lookup failed.
948 PH/40 RM_COMMAND is now used in the building process.
950 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
951 the "build-* directories that it finds.
953 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
954 address in a domain literal was a prefix of an interface address.
956 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
957 when verifying a sender address, unless rewrite_headers is false.
959 PH/44 Wrote a long comment about why errors_to addresses are verified as
960 recipients, not senders.
962 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
963 the ratelimit ACL was added.
965 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
967 PH/46 Added extra information about PostgreSQL errors to the error string.
969 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
970 avoiding the need to install special zones in a real server. This is
971 backwards compatible; if it can't find the fake resolver, it drops back.
972 Thus, both old and new test suites can be run.
974 TF/02 Added util/ratelimit.pl
976 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
977 clock is set back in time.
979 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
980 Candler <B.Candler@pobox.com>.
982 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
983 OK for addresses that are the result of redirections.
985 PH/48 A number of further additions for the benefit of the new test suite,
986 including a fake gethostbyname() that interfaces to the fake DNS resolver
989 TF/06 The fix for widen_domains has also been applied to qualify_single and
990 search_parents which are the other dnslookup options that can cause
993 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
996 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
998 TF/07 Exim produced the error message "an SRV record indicated no SMTP
999 service" if it encountered an MX record with an empty target hostname.
1000 The message is now "an MX or SRV record indicated no SMTP service".
1002 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1003 if the DNS of the sending site is misconfigured. This is quite a
1004 common situation. This change restores the behaviour of treating a
1005 helo verification defer as a failure.
1007 PH/51 If self=fail was set on a router, the bounce message did not include the
1008 actual error message.
1014 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1016 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1017 command itself ended in error, the underlying message about the transport
1018 filter timeout was being overwritten with the pipe command error. Now the
1019 underlying error message should be appended to the second error message.
1021 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1023 PH/02 Reduce the amount of output that "make" produces by default. Full output
1024 can still be requested.
1026 PH/03 The warning log line about a condition test deferring for a "warn" verb
1027 was being output only once per connection, rather than after each
1028 occurrence (because it was using the same function as for successful
1029 "warn" verbs). This seems wrong, so I have changed it.
1031 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1032 it should not have, which might have caused a crash in the right
1033 circumstances, but probably never did.
1035 PH/04 Installed a modified version of Tony Finch's patch to make submission
1036 mode fix the return path as well as the Sender: header line, and to
1037 add a /name= option so that you can make the user's friendly name appear
1040 TF/03 Added the control = fakedefer ACL modifier.
1042 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1043 Mark Lowes for thorough testing.
1045 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1047 TK/03 Merged latest SRS patch from Miles Wilton.
1049 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1050 with the definition in sysexits.h (which is #included earlier).
1051 Fortunately, Exim does not actually use EX_OK. The code used to try to
1052 preserve the sysexits.h value, by assumimg that macro definitions were
1053 scanned for macro replacements. I have been disabused of this notion,
1054 so now the code just undefines EX_OK before #including unistd.h.
1056 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1057 in the smtp transport. When a block could not be written in a single
1058 write() function, the timeout was being re-applied to each part-write.
1059 This seems wrong - if the receiver was accepting one byte at a time it
1060 would take for ever. The timeout is now adjusted when this happens. It
1061 doesn't have to be particularly precise.
1063 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1064 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1066 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1067 without a domain if the check to <postmaster@domain> fails.
1069 SC/01 Eximstats: added -xls and the ability to specify output files
1070 (patch written by Frank Heydlauf).
1072 SC/02 Eximstats: use FileHandles for outputing results.
1074 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1076 SC/04 Eximstats: fixed display of large numbers with -nvr option
1078 SC/05 Eximstats: fixed merging of reports with empty tables.
1080 SC/06 Eximstats: added the -include_original_destination flag
1082 SC/07 Eximstats: removed tabs and trailing whitespace.
1084 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1086 TK/06 MBOX spool code: Add real "From " MBOX separator line
1087 so the .eml file is really in mbox format (even though
1088 most programs do not really care). Patch from Alex Miller.
1090 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1091 The latter is generated from $received_to and is only set if the
1092 message has one envelope recipient. SA can use these headers,
1093 obviously out-of-the-box. Patch from Alex Miller.
1095 PH/08 The ${def test on a variable was returning false if the variable's
1096 value was "0", contrary to what the specification has always said!
1097 The result should be true unless the variable is empty.
1099 PH/09 The syntax error of a character other than { following "${if
1100 def:variable_name" (after optional whitespace) was not being diagnosed.
1101 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1102 accidental colon was present, for example, could give incorrect results.
1104 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1105 result is used (not including appendfile, where other changes are about
1108 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1109 This involved changing a lot of size variables from int to off_t. It
1110 should work with maildirs and everything.
1112 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1113 spamd dying while we are connected to it.
1115 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1118 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1119 function for BDB 4.3.
1121 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1123 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1124 the sources for PCRE have been re-arranged and more files are now
1127 PH/15 The code I had for printing potentially long long variables in PH/11
1128 above was not the best (it lost precision). The length of off_t variables
1129 is now inspected at build time, and an appropriate printing format (%ld
1130 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1131 to be "long long int" or "long int". This is needed for the internal
1132 formatting function string_vformat().
1134 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1135 the configuration file to be ":syslog", then the script "guesses" where
1136 the logs files are, rather than using the compiled in default. In our
1137 case the guess is not the same as the compiled default, so the script
1138 suddenly stopped working when I started to use syslog. The patch checks
1139 to see if log_file_path is "". If so, it attempts to read it from exim
1140 with no configuration file to get the compiled in version, before it
1141 falls back to the previous guessing code."
1143 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1144 implementing BATV in an Exim configuration. See NewStuff for the gory
1147 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1148 Makefile that are specific to HP-UX.
1150 PH/18 If the "use_postmaster" option was set for a recipient callout together
1151 with the "random" option, the postmaster address was used as the MAIL
1152 FROM address for the random test, but not for the subsequent recipient
1153 test. It is now used for both.
1155 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1156 patch removes a few documentation additions to RFC 3028, because the
1157 latest draft now contains them. It adds the new en;ascii-case comparator
1158 and a new error check for 8bit text in MIME parts. Comparator and
1159 require names are now matched exactly. I enabled the subaddress
1160 extension, but it is not well tested yet (read: it works for me)."
1162 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1163 rework some of the code of TK/09 above to avoid the hardwired use of
1164 "%lld" and "long long". Replaced the call to snprintf() with a call to
1167 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1168 records point to non-existent hosts", "retry timeout exceeded", and
1169 "retry time not reached for any host after a long failure period".
1171 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1172 experimental DomainKeys support:
1174 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1175 (2) On an error such as an illegally used "control", the wrong name for
1176 the control was given.
1178 These problems did NOT occur unless DomainKeys support was compiled.
1180 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1182 PH/24 Added ${if match_ip condition.
1184 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1185 databases so that it will be absolutely obvious if a crash occurs in the
1186 DB library. This is a regular occurrence (often caused by mis-matched
1189 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1190 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1191 picked up on a user's system that detects such things. There doesn't seem
1192 to be a gcc warning option for this - only an attribute that has to be
1193 put on the function's prototype. It seems that in Fedora Core 4 they have
1194 set this on a number of new functions. No doubt there will be more in due
1197 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1198 specify a transport. However, if an address that was verified by such a
1199 router was the subject of a callout, Exim crashed because it tried to
1200 read the rcpt_include_affixes from the non-existent transport. Now it
1201 just assumes that the setting of that option is false. This bug was
1202 introduced by 4.51/PH/31.
1204 PH/28 Changed -d+all to exclude +memory, because that information is very
1205 rarely of interest, but it makes the output a lot bigger. People tend to
1206 do -d+all out of habit.
1208 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1209 code in os-type was giving problems when libc.so lives in lib64, like on
1212 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1213 aren't the modern standard, and it seems that some systems' include files
1214 don't always have them. Exim was already checking for some of the newer
1215 ones like T_AAAA, and defining it itself. I've added checks for all the
1216 record types that Exim uses.
1218 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1219 not automatically generating a new one, as it is supposed to. This
1220 prevented TLS from working. If the file did exist, but contained invalid
1221 data, a new version was generated, as expected. It was only the case of a
1222 non-existent file that was broken.
1224 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1225 with a change in libdomainkeys > 0.64.
1227 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1228 from DNS. If the selector record carries the flag, it now has
1229 precedence over the domain-wide flag.
1231 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1233 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1234 the use of a port name that isn't defined in /etc/services) occurred, the
1235 message was deferred as in a normal delivery, and thus remained on the
1236 spool, instead of being failed because of the mua_wrapper setting. This
1237 is now fixed, and I tidied up some of the mua_wrapper messages at the
1240 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1241 the messages in a hash of arrays rather than using individual hashes.
1242 This is a bit cleaner and results in dramatic memory savings, albeit
1243 at a slight CPU cost.
1245 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1246 as requested by Marc Sherman.
1248 SC/10 Eximstats: added histograms for user specified patterns as requested
1251 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1253 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1254 fopen() in the content-scanning modules that did not already have it.
1260 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1261 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1263 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1265 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1267 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1268 to test Sieve filters that use "vacation".
1270 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1271 that changes the way the GnuTLS parameters are stored in the cache file.
1272 The new format can be generated externally. For backward compatibility,
1273 if the data in the cache doesn't make sense, Exim assumes it has read an
1274 old-format file, and it generates new data and writes a new file. This
1275 means that you can't go back to an older release without removing the
1278 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1279 work if there are any delivery delays because "one_time" forces the
1280 parent to be marked "delivered", so its unseen clone is never tried
1281 again. For this reason, Exim now forbids the simultaneous setting of
1284 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1285 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1286 entry if you want to know the details. The fix, however, neglected to
1287 consider the case when local delivery batching is involved. The test for
1288 "previously delivered" was not happening when checking to see if an
1289 address could be batched with a previous (undelivered) one; under
1290 certain circumstances this could lead to multiple deliveries to the same
1293 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1294 in its include files, and this causes problems building Exim.
1296 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1297 header_syntax) but Exim was just ignoring anything given after a slash.
1298 In particular, this caused confusion with an attempt to use "verify =
1299 reverse_host_lookup/defer_ok". An error is now given when options are
1300 supplied for verify items that do not have them. (Maybe reverse_host_
1301 lookup should have a defer_ok option, but that's a different point.)
1303 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1304 defined by RFC 821) to 2048, because there were problems with some AUTH
1305 commands, and RFC 1869 says the size should be increased for extended
1306 SMTP commands that take arguments.
1308 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1311 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1312 "unknown" error; now it says that the functionality isn't in the binary.
1314 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1315 an address' error message when a string expansion fails (syntax or
1316 whatever). Otherwise the password may appear in the log. Following change
1317 PH/42 below, there is no longer a chance of it appearing in a bounce
1320 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1322 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1323 of its addresses was ever tried. (Bugzilla bug #2.)
1325 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1326 the result incorrectly in the debug output. (It correctly added a newline
1327 to what was transported.)
1329 TF/01 Added $received_time.
1331 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1332 commented out examples of how to interface to a virus scanner and to
1333 SpamAssassin. Also added commented examples of av_scanner and
1334 spamd_address settings.
1336 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1337 and controls are allowed in which ACLs. There were a couple of minor
1338 errors. Some of the entries in the conditions table (which is a table of
1339 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1340 negation of where the condition IS allowed.
1342 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1344 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1345 header file does not have a version number, so I've had to invent a new
1346 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1347 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1348 radiusclient), but it was contributed by a Radius user.
1350 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1351 files or directories, for OpenSSL.
1353 PH/19 When an Exim process that is running as root has to create an Exim log
1354 file, it does so in a subprocess that runs as exim:exim so as to get the
1355 ownership right at creation (otherwise, other Exim processes might see
1356 the file with the wrong ownership). There was no test for failure of this
1357 fork() call, which would lead to the process getting stuck as it waited
1358 for a non-existent subprocess. Forks do occasionally fail when resources
1359 run out. I reviewed all the other calls to fork(); they all seem to check
1362 PH/20 When checking for unexpected SMTP input at connect time (before writing
1363 the banner), Exim was not dealing correctly with a non-positive return
1364 from the read() function. If the client had disconnected by this time,
1365 the result was a log entry for a synchronization error with an empty
1366 string after "input=" when read() returned zero. If read() returned -1
1367 (an event I could not check), uninitialized data bytes were printed.
1368 There were reports of junk text (parts of files, etc) appearing after
1371 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1373 PH/22 Added support for macro redefinition, and (re)definition in between
1374 driver and ACL definitions.
1376 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1377 forgetting to use the resulting value; it was using the unexpanded value.
1379 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1380 hadn't been configured. The fix is from Juergen Kreileder, who
1381 understands it better than I do:
1383 "Here's what I see happening with three configured cyrus_sasl
1384 authenticators configured (plain, login, cram-md5):
1386 On startup auth_cyrus_sasl_init() gets called for each of these.
1387 This means three calls to sasl_listmech() without a specified mech_list.
1388 => SASL tests which mechs of all available mechs actually work
1389 => three warnings about OTP not working
1390 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1392 With the patch, sasl_listmech() also gets called three times. But now
1393 SASL's mech_list option is set to the server_mech specified in the the
1394 authenticator. Or in other words, the answer from sasl_listmech()
1395 gets limited to just the mech you're testing for (which is different
1397 => the return list contains just 'plain' or 'login', 'cram-md5' or
1398 nothing depending on the value of ob->server_mech.
1400 I've just tested the patch: Authentication still works fine,
1401 unavailable mechs specified in the exim configuration are still
1402 caught, and the auth.log warnings about OTP are gone."
1404 PH/25 When debugging is enabled, the contents of the command line are added
1405 to the debugging output, even when log_selector=+arguments is not
1408 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1409 answer is "GNU", and only if the return is "GNU/something" is the answer
1412 PH/27 $acl_verify_message is now set immediately after the failure of a
1413 verification in an ACL, and so is available in subsequent modifiers. In
1414 particular, the message can be preserved by coding like this:
1416 warn !verify = sender
1417 set acl_m0 = $acl_verify_message
1419 Previously, $acl_verify_message was set only while expanding "message"
1420 and "log_message" when a very denied access.
1422 PH/28 Modified OS/os.c-Linux with
1424 -#ifndef OS_LOAD_AVERAGE
1425 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1427 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1428 nomenclature these days.)
1430 PH/29 Installed patch from the Sieve maintainer that adds the options
1431 sieve_useraddress and sieve_subaddress to the redirect router.
1433 PH/30 In these circumstances:
1434 . Two addresses routed to the same list of hosts;
1435 . First host does not offer TLS;
1436 . First host accepts first address;
1437 . First host gives temporary error to second address;
1438 . Second host offers TLS and a TLS session is established;
1439 . Second host accepts second address.
1440 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1441 and peerdn, if requested) that were in fact used only for the second
1444 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1445 attention to any local part prefix or suffix that was matched by the
1446 router that accepted the address. It now behaves in the same way as it
1447 does for delivery: the affixes are removed from the local part unless
1448 rcpt_include_affixes is set on the transport.
1450 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1451 timeout during the DATA phase of an incoming message.
1453 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1454 applied a patch sanctioned by the Sieve maintainer.
1456 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1457 the uid or gid is negative. A case of a negative gid caused this to be
1458 noticed. The fix allows for either to be negative.
1460 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1461 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1464 PH/36 The change PH/12 above was broken. Fixed it.
1466 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1467 the grounds that routing the same address twice would always produce the
1468 same answer. This might have been true once, but it is certainly no
1469 longer true now. Routing a child address may depend on the previous
1470 routing that produced that child. Some complicated redirection strategies
1471 went wrong when messages had multiple recipients, and made Exim's
1472 behaviour dependent on the order in which the addresses were given.
1474 I have moved the duplicate checking until after the routing is complete.
1475 Exim scans the addresses that are assigned to local and remote
1476 transports, and removes any duplicates. This means that more work will be
1477 done, as duplicates will always all be routed, but duplicates are
1478 presumably rare, so I don't expect this is of any significance.
1480 For deliveries to pipes, files, and autoreplies, the duplicate checking
1481 still happens during the routing process, since they are not going to be
1484 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1485 It corrects a timeout issue with spamd. This is Ian's comment: "The
1486 background is that sometimes spamd either never reads data from a
1487 connection it has accepted, or it never writes response data. The exiscan
1488 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1489 blindly assumes that writes won't block so it may never time out."
1491 PH/39 Allow G after quota size as well as K and M.
1493 PH/40 The value set for $authenticated_id in an authenticator may not contain
1494 binary zeroes or newlines because the value is written to log lines and
1495 to spool files. There was no check on this. Now the value is run through
1496 the string_printing() function so that such characters are converted to
1497 printable escape sequences.
1499 PH/41 $message_linecount is a new variable that contains the total number of
1500 lines in the message. Compare $body_linecount, which is the count for the
1503 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1504 bounce and delay warning messages, except in certain special cases, which
1507 (a) An SMTP error message from a remote host;
1508 (b) A message specified in a :fail: redirection;
1509 (c) A message specified in a "fail" command in a system filter;
1510 (d) A message specified in a FAIL return from the queryprogram router;
1511 (e) A message specified by the cannot_route_message router option.
1513 In these cases only, Exim does include the error details in bounce and
1514 warning messages. There are also a few cases where bland messages such
1515 as "unrouteable address" or "local delivery error" are given.
1517 PH/43 $value is now also set for the "else" part of a ${run expansion.
1519 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1520 being worked on, but at least Exim now implements the latest version to
1523 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1524 process to complete was treated as a delivery failure, a timeout while
1525 writing the message to the pipe was logged, but erroneously treated as a
1526 successful delivery. Such timeouts include transport filter timeouts. For
1527 consistency with the overall process timeout, these timeouts are now
1528 treated as errors, giving rise to delivery failures by default. However,
1529 there is now a new Boolean option for the pipe transport called
1530 timeout_defer, which, if set TRUE, converts the failures into defers for
1531 both kinds of timeout. A transport filter timeout is now identified in
1534 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1535 systems where "make" and "gmake" are different, calling "gmake" at top
1536 level broke things. I've arranged for the value of $(MAKE) to be passed
1537 from the Makefile to this script so that it can call the same version of
1541 A note about Exim versions 4.44 and 4.50
1542 ----------------------------------------
1544 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1545 changes of various kinds. As a consequence, a big documentation update was
1546 needed. This delayed the release for rather longer than seemed good, especially
1547 in the light of a couple of (minor) security issues. Therefore, the changes
1548 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1549 release. So 4.44 and 4.50 are in effect two different branches that both start
1552 I have left the 4.50 change log unchanged; it contains all the changes since
1553 4.43. The change log for 4.44 is below; many of its items are identical to
1554 those for 4.50. This seems to be the most sensible way to preserve the
1555 historical information.
1561 1. Minor wording change to the doc/README.SIEVE file.
1563 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1564 computation of the current number of files was incorrect.
1566 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1567 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1570 4. Give more explanation in the error message when the command for a transport
1571 filter fails to execute.
1573 5. There are several places where Exim runs a non-Exim command in a
1574 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1575 was being done only for the command run by the queryprogram router. It is
1576 now done for all such subprocesses. The other cases are: ${run, transport
1577 filters, and the commands run by the lmtp and pipe transports.
1579 6. Added CONFIGURE_GROUP build-time option.
1581 7. Some older OS have a limit of 256 on the maximum number of file
1582 descriptors. Exim was using setrlimit() to set 1000 as a large value
1583 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1584 systems. I've change it so that if it can't get 1000, it tries for 256.
1586 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1587 was an oversight, and furthermore, ever since the addition of extra
1588 controls (e.g. 4.43/32), the checks on when to allow different forms of
1589 "control" were broken. There should now be diagnostics for all cases when a
1590 control that does not make sense is encountered.
1592 9. Added the /retain_sender option to "control=submission".
1594 10. $recipients is now available in the predata ACL (oversight).
1596 11. Tidy the search cache before the fork to do a delivery from a message
1597 received from the command line. Otherwise the child will trigger a lookup
1598 failure and thereby defer the delivery if it tries to use (for example) a
1599 cached ldap connection that the parent has called unbind on.
1601 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1602 of $address_data from the recipient verification was clobbered by the
1603 sender verification.
1605 13. The value of address_data from a sender verification is now available in
1606 $sender_address_data in subsequent conditions in the ACL statement.
1608 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1610 15. Added a new option "connect=<time>" to callout options, to set a different
1613 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1614 was its contents. (It was OK if the option was not defined at all.)
1616 17. A "Completed" log line is now written for messages that are removed from
1617 the spool by the -Mrm option.
1619 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1620 information about exactly what failed.
1622 19. Added -dd to debug only the daemon process.
1624 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1625 handles timeouts, both on the server side and network timeouts. Renamed the
1626 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1628 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1629 It is now set to "smtps".
1631 22. $host_address is now set to the target address during the checking of
1632 ignore_target_hosts.
1634 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1635 being passed; this would have caused $sender_host_name to have been used if
1636 matching the list had actually called for a host name (not very likely,
1637 since this list is usually IP addresses). A host name is now passed as
1640 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1641 code that specifies a non-restarting handler (typically sigaction() in
1642 modern systems) in an attempt to fix a rare and obscure crash bug.
1644 25. Narrowed the window for a race in the daemon that could cause it to ignore
1645 SIGCHLD signals. This is not a major problem, because they are used only to
1646 wake it up if nothing else does.
1648 26. A malformed maildirsize file could cause Exim to calculate negative values
1649 for the mailbox size or file count. Odd effects could occur as a result.
1650 The maildirsize information is now recalculated if the size or filecount
1653 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1654 support for a long time. Removed HAVE_SYS_VFS_H.
1656 28. Installed the latest version of exipick from John Jetmore.
1658 29. In an address list, if the pattern was not a regular expression, an empty
1659 subject address (from a bounce message) matched only if the pattern was an
1660 empty string. Non-empty patterns were not even tested. This was the wrong
1661 because it is perfectly reasonable to use an empty address as part of a
1662 database query. An empty address is now tested by patterns that are
1663 lookups. However, all the other forms of pattern expect the subject to
1664 contain a local part and a domain, and therefore, for them, an empty
1665 address still always fails if the pattern is not itself empty.
1667 30. Exim went into a mad DNS loop when attempting to do a callout where the
1668 host was specified on an smtp transport, and looking it up yielded more
1669 than one IP address.
1671 31. Re-factored the code for checking spool and log partition space into a
1672 function that finds that data and another that does the check. The former
1673 is then used to implement four new variables: $spool_space, $log_space,
1674 $spool_inodes, and $log_inodes.
1676 32. The RFC2047 encoding function was originally intended for short strings
1677 such as real names; it was not keeping to the 75-character limit for
1678 encoded words that the RFC imposes. It now respects the limit, and
1679 generates multiple encoded words if necessary. To be on the safe side, I
1680 have increased the buffer size for the ${rfc2047: expansion operator from
1683 33. It is now permitted to omit both strings after an "if" condition; if the
1684 condition is true, the result is "true". As before, when the second string
1685 is omitted, a false condition yields an empty string. This makes it less
1686 cumbersome to write custom ACL and router conditions.
1688 34. Failure to deliver a bounce message always caused it to be frozen, even if
1689 there was an errors_to setting on the router. The errors_to setting is now
1692 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1693 canonical form (fully expanded) before being placed in
1694 $sender_host_address.
1696 36. The table in the code that translates DNS record types into text (T_A to
1697 "A" for instance) was missing entries for NS and CNAME. It is just possible
1698 that this could have caused confusion if both these types were looked up
1699 for the same domain, because the text type is used as part of Exim's
1700 per-process caching. But the chance of anyone hitting this buglet seems
1703 37. The dnsdb lookup has been extended in a number of ways.
1705 (1) There is a new type, "zns", which walks up the domain tree until it
1706 finds some nameserver records. It should be used with care.
1708 (2) There is a new type, "mxh", which is like "mx" except that it returns
1709 just the host names, not the priorities.
1711 (3) It is now possible to give a list of domains (or IP addresses) to be
1712 looked up. The behaviour when one of the lookups defers can be
1713 controlled by a keyword.
1715 (4) It is now possible to specify the separator character for use when
1716 multiple records are returned.
1718 38. The dnslists ACL condition has been extended: it is now possible to supply
1719 a list of IP addresses and/or domains to be looked up in a particular DNS
1722 39. Added log_selector=+queue_time_overall.
1724 40. When running the queue in the test harness, wait just a tad after forking a
1725 delivery process, to get repeatability of debugging output.
1727 41. Include certificate and key file names in error message when GnuTLS fails
1728 to set them up, because the GnuTLS error message doesn't include the name
1729 of the failing file when there is a problem reading it.
1731 42. Allow both -bf and -bF in the same test run.
1733 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1735 44. The "Exiscan patch" is now merged into the mainline Exim source.
1737 45. Sometimes the final signoff response after QUIT could fail to get
1738 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1739 < 0 before doing a fflush(). This bug looks as though it goes back to the
1740 introduction of TLS in release 3.20, but "sometimes" must have been rare
1741 because the tests only now provoked it.
1743 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1744 (this can affect the format of dates).
1746 47. exim_tidydb, when checking for the continued existence of a message for
1747 which it has found a message-specific retry record, was not finding
1748 messages that were in split spool directories. Consequently, it was
1749 deleting retry records that should have stayed in existence.
1751 48. Steve fixed some bugs in eximstats.
1753 49. The SPA authentication driver was not abandoning authentication and moving
1754 on to the next authenticator when an expansion was forced to fail,
1755 contradicting the general specification for all authenticators. Instead it
1756 was generating a temporary error. It now behaves as specified.
1758 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1759 (the order specifies the preference for clients). The order is now AES256,
1760 AES128, 3DES, ARCFOUR128.
1762 51. Small patch to Sieve code - explicitly set From: when generating an
1765 52. Exim crashed if a remote delivery caused a very long error message to be
1766 recorded - for instance if somebody sent an entire SpamAssassin report back
1767 as a large number of 550 error lines. This bug was coincidentally fixed by
1768 increasing the size of one of Exim's internal buffers (big_buffer) that
1769 happened as part of the Exiscan merge. However, to be on the safe side, I
1770 have made the code more robust (and fixed the comments that describe what
1773 53. Now that there can be additional text after "Completed" in log lines (if
1774 the queue_time_overall log selector is set), a one-byte patch to exigrep
1775 was needed to allow it to recognize "Completed" as not the last thing in
1778 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1779 patch that reportedly fixes this has been added. I am not expert enough to
1780 create a test for it. This is what the patch creator wrote:
1782 "I found a little strange behaviour of ldap code when working with
1783 Windows 2003 AD Domain, where users was placed in more than one
1784 Organization Units. When I tried to give exim partial DN, the exit code
1785 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1786 But simultaneously result of request was absolutely normal ldap result,
1787 so I produce this patch..."
1789 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1790 REFERENCE, so I have modified the code to exclude the patch when that macro
1793 55. Some experimental protocols are using DNS PTR records for new purposes. The
1794 keys for these records are domain names, not reversed IP addresses. The
1795 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1796 leaves it alone. Component reversal etc. now happens only for IP addresses.
1799 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1801 57. Double the size of the debug message buffer (to 2048) so that more of very
1802 long debug lines gets shown.
1804 58. The exicyclog utility now does better if the number of log files to keep
1805 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1807 59. Two changes related to the smtp_active_hostname option:
1809 (1) $smtp_active_hostname is now available as a variable.
1810 (2) The default for smtp_banner uses $smtp_active_hostname instead
1811 of $primary_hostname.
1813 60. The host_aton() function is supposed to be passed a string that is known
1814 to be a valid IP address. However, in the case of IPv6 addresses, it was
1815 not checking this. This is a hostage to fortune. Exim now panics and dies
1816 if the condition is not met. A case was found where this could be provoked
1817 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1818 components; fortuitously, this particular loophole had already been fixed
1819 by change 4.50/55 above.
1821 If there are any other similar loopholes, the new check in host_aton()
1822 itself should stop them being exploited. The report I received stated that
1823 data on the command line could provoke the exploit when Exim was running as
1824 exim, but did not say which command line option was involved. All I could
1825 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1826 running as the user.
1829 61. There was a buffer overflow vulnerability in the SPA authentication code
1830 (which came originally from the Samba project). I have added a test to the
1831 spa_base64_to_bits() function which I hope fixes it.
1834 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1835 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1837 63. The daemon start-up calls getloadavg() while still root for those OS that
1838 need the first call to be done as root, but it missed one case: when
1839 deliver_queue_load_max is set with deliver_drop_privilege. This is
1840 necessary for the benefit of the queue runner, because there is no re-exec
1841 when deliver_drop_privilege is set.
1843 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1844 This has been fixed.
1846 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1847 was in use, was not putting the data itself into the right store pool;
1848 consequently, it could be overwritten for a subsequent message in the same
1849 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1852 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1854 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1855 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1856 the function were treating the return as a boolean value, which happened to
1857 work because 0=false and not-0=true, but is not correct code.
1859 68. The host_aton() function was not handling scoped IPv6 addresses (those
1860 with, for example, "%eth0" on the end) correctly.
1862 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1863 negated items (that is, ~something) in unsigned ints. Some compilers
1864 apparently mutter when there is no cast.
1866 70. If an address verification called from an ACL failed, and did not produce a
1867 user-specific message (i.e. there was only a "system" message), nothing was
1868 put in $acl_verify_message. In this situation, it now puts the system
1871 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1872 session; change 4.31/43 added the unwanted input to the log line - except
1873 that it did not do this in the start of session case. It now does.
1875 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1876 This is wrong and can cause the other end to generate a synchronization
1877 error if it is another Exim or anything else that does the synchronization
1878 check. A QUIT command is no longer sent after a timeout.
1880 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1881 during host lookups.
1883 74. The defer_ok option of callout verification was not working if it was used
1884 when verifying addresses in header lines, that is, for this case:
1886 verify = header_sender/callout=defer_ok
1888 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1889 those file descriptors could be used for SMTP connections. If anything
1890 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1891 could be sent to the SMTP client, causing chaos. The daemon now opens
1892 stdin, stdout, and stderr to /dev/null when it puts itself into the
1895 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1896 log by default. The user can override this with suitable Perl magic.
1898 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1899 the log message when discard triggers, was not working for the DATA ACL or
1900 for the non-SMTP ACL.
1902 78. Error message wording change in sieve.c.
1904 79. If smtp_accept_max_per_host was set, the number of connections could be
1905 restricted to fewer than expected, because the daemon was trying to set up
1906 a new connection before checking whether the processes handling previous
1907 connections had finished. The check for completed processes is now done
1908 earlier. On busy systems, this bug wouldn't be noticed because something
1909 else would have woken the daemon, and it would have reaped the completed
1912 80. If a message was submitted locally by a user whose login name contained one
1913 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1914 It caused a spool format error. I have fixed the spool reading code. A
1915 related problem was that the "from" clause in the Received: line became
1916 illegal because of the space(s). It is now covered by ${quote_local_part.
1918 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1921 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1923 83. Patch from Sieve maintainer to fix unterminated string problem in
1924 "vacation" handling.
1926 84. Some minor changes to the Linux configuration files to help with other
1927 OS variants using glibc.
1929 85. One more patch for Sieve to update vacation handling to latest spec.
1932 ----------------------------------------------------
1933 See the note above about the 4.44 and 4.50 releases.
1934 ----------------------------------------------------
1940 1. Change 4.43/35 introduced a bug that caused file counts to be
1941 incorrectly computed when quota_filecount was set in an appendfile
1944 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1945 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1948 3. Give more explanation in the error message when the command for a transport
1949 filter fails to execute.
1951 4. There are several places where Exim runs a non-Exim command in a
1952 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1953 was being done only for the command run by the queryprogram router. It is
1954 now done for all such subprocesses. The other cases are: ${run, transport
1955 filters, and the commands run by the lmtp and pipe transports.
1957 5. Some older OS have a limit of 256 on the maximum number of file
1958 descriptors. Exim was using setrlimit() to set 1000 as a large value
1959 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1960 systems. I've change it so that if it can't get 1000, it tries for 256.
1962 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1963 was an oversight, and furthermore, ever since the addition of extra
1964 controls (e.g. 4.43/32), the checks on when to allow different forms of
1965 "control" were broken. There should now be diagnostics for all cases when a
1966 control that does not make sense is encountered.
1968 7. $recipients is now available in the predata ACL (oversight).
1970 8. Tidy the search cache before the fork to do a delivery from a message
1971 received from the command line. Otherwise the child will trigger a lookup
1972 failure and thereby defer the delivery if it tries to use (for example) a
1973 cached ldap connection that the parent has called unbind on.
1975 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1976 of $address_data from the recipient verification was clobbered by the
1977 sender verification.
1979 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1980 was its contents. (It was OK if the option was not defined at all.)
1982 11. A "Completed" log line is now written for messages that are removed from
1983 the spool by the -Mrm option.
1985 12. $host_address is now set to the target address during the checking of
1986 ignore_target_hosts.
1988 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1989 being passed; this would have caused $sender_host_name to have been used if
1990 matching the list had actually called for a host name (not very likely,
1991 since this list is usually IP addresses). A host name is now passed as
1994 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1995 code that specifies a non-restarting handler (typically sigaction() in
1996 modern systems) in an attempt to fix a rare and obscure crash bug.
1998 15. Narrowed the window for a race in the daemon that could cause it to ignore
1999 SIGCHLD signals. This is not a major problem, because they are used only to
2000 wake it up if nothing else does.
2002 16. A malformed maildirsize file could cause Exim to calculate negative values
2003 for the mailbox size or file count. Odd effects could occur as a result.
2004 The maildirsize information is now recalculated if the size or filecount
2007 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2008 support for a long time. Removed HAVE_SYS_VFS_H.
2010 18. Updated exipick to current release from John Jetmore.
2012 19. Allow an empty sender to be matched against a lookup in an address list.
2013 Previously the only cases considered were a regular expression, or an
2016 20. Exim went into a mad DNS lookup loop when doing a callout where the
2017 host was specified on the transport, if the DNS lookup yielded more than
2020 21. The RFC2047 encoding function was originally intended for short strings
2021 such as real names; it was not keeping to the 75-character limit for
2022 encoded words that the RFC imposes. It now respects the limit, and
2023 generates multiple encoded words if necessary. To be on the safe side, I
2024 have increased the buffer size for the ${rfc2047: expansion operator from
2027 22. Failure to deliver a bounce message always caused it to be frozen, even if
2028 there was an errors_to setting on the router. The errors_to setting is now
2031 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2032 canonical form (fully expanded) before being placed in
2033 $sender_host_address.
2035 24. Updated eximstats to version 1.33
2037 25. Include certificate and key file names in error message when GnuTLS fails
2038 to set them up, because the GnuTLS error message doesn't include the name
2039 of the failing file when there is a problem reading it.
2041 26. Expand error message when OpenSSL has problems setting up cert/key files.
2044 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2045 (this can affect the format of dates).
2047 28. exim_tidydb, when checking for the continued existence of a message for
2048 which it has found a message-specific retry record, was not finding
2049 messages that were in split spool directories. Consequently, it was
2050 deleting retry records that should have stayed in existence.
2052 29. eximstats updated to version 1.35
2053 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2054 1.35 - bugfix such that pie charts by volume are generated correctly
2056 30. The SPA authentication driver was not abandoning authentication and moving
2057 on to the next authenticator when an expansion was forced to fail,
2058 contradicting the general specification for all authenticators. Instead it
2059 was generating a temporary error. It now behaves as specified.
2061 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2062 (the order specifies the preference for clients). The order is now AES256,
2063 AES128, 3DES, ARCFOUR128.
2065 31. Small patch to Sieve code - explicitly set From: when generating an
2068 32. Exim crashed if a remote delivery caused a very long error message to be
2069 recorded - for instance if somebody sent an entire SpamAssassin report back
2070 as a large number of 550 error lines. This bug was coincidentally fixed by
2071 increasing the size of one of Exim's internal buffers (big_buffer) that
2072 happened as part of the Exiscan merge. However, to be on the safe side, I
2073 have made the code more robust (and fixed the comments that describe what
2076 33. Some experimental protocols are using DNS PTR records for new purposes. The
2077 keys for these records are domain names, not reversed IP addresses. The
2078 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2079 leaves it alone. Component reversal etc. now happens only for IP addresses.
2082 34. The host_aton() function is supposed to be passed a string that is known
2083 to be a valid IP address. However, in the case of IPv6 addresses, it was
2084 not checking this. This is a hostage to fortune. Exim now panics and dies
2085 if the condition is not met. A case was found where this could be provoked
2086 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2087 components; fortuitously, this particular loophole had already been fixed
2088 by change 4.50/55 or 4.44/33 above.
2090 If there are any other similar loopholes, the new check in host_aton()
2091 itself should stop them being exploited. The report I received stated that
2092 data on the command line could provoke the exploit when Exim was running as
2093 exim, but did not say which command line option was involved. All I could
2094 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2095 running as the user.
2098 35. There was a buffer overflow vulnerability in the SPA authentication code
2099 (which came originally from the Samba project). I have added a test to the
2100 spa_base64_to_bits() function which I hope fixes it.
2103 36. The daemon start-up calls getloadavg() while still root for those OS that
2104 need the first call to be done as root, but it missed one case: when
2105 deliver_queue_load_max is set with deliver_drop_privilege. This is
2106 necessary for the benefit of the queue runner, because there is no re-exec
2107 when deliver_drop_privilege is set.
2109 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2110 was in use, was not putting the data itself into the right store pool;
2111 consequently, it could be overwritten for a subsequent message in the same
2112 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2115 38. Sometimes the final signoff response after QUIT could fail to get
2116 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2117 < 0 before doing a fflush(). This bug looks as though it goes back to the
2118 introduction of TLS in release 3.20, but "sometimes" must have been rare
2119 because the tests only now provoked it.
2125 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2126 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2127 Now it returns an integer. A number of calls were still expecting a T/F
2128 return. Fortuitously, in all cases, the tests worked in OK situations,
2129 which is the norm. However, things would have gone wrong on any write
2130 failures on the smtp file descriptor. This function is used when sending
2131 messages over SMTP and also when doing verify callouts.
2133 2. When Exim is called to do synchronous delivery of a locally submitted
2134 message (the -odf or -odi options), it no longer closes stderr before doing
2137 3. Implemented the mua_wrapper option.
2139 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2141 5. Implemented the functions header_remove(), header_testname(),
2142 header_add_at_position(), and receive_remove_recipient(), and exported them
2145 6. If an ACL "warn" statement specified the addition of headers, Exim already
2146 inserted X-ACL-Warn: at the start if there was no header name. However, it
2147 was not making this test for the second and subsequent header lines if
2148 there were newlines in the string. This meant that an invalid header could
2149 be inserted if Exim was badly configured.
2151 7. Allow an ACL "warn" statement to add header lines at the start or after all
2152 the Received: headers, as well as at the end.
2154 8. Added the rcpt_4xx retry error code.
2156 9. Added postmaster_mailfrom=xxx to callout verification option.
2158 10. Added mailfrom=xxxx to the callout verification option, for verify=
2161 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2162 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2164 12. Inserted some casts to stop certain compilers warning when using pointer
2165 differences as field lengths or precisions in printf-type calls (mostly
2166 affecting debugging statements).
2168 13. Added optional readline() support for -be (dynamically loaded).
2170 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2171 same clock tick as a message's arrival, so that its received time was the
2172 same as the "first fail" time on the retry record, and that message
2173 remained on the queue past the ultimate address timeout, every queue runner
2174 would try a delivery (because it was past the ultimate address timeout) but
2175 after another failure, the ultimate address timeout, which should have then
2176 bounced the address, did not kick in. This was a "< instead of <=" error;
2177 in most cases the first failure would have been in the next clock tick
2178 after the received time, and all would be well.
2180 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2181 being recognized when the domain list was tested by the match_domain
2182 condition in an expansion string.
2184 16. Added the ${str2b64: operator.
2186 17. Exim was always calling setrlimit() to set a large limit for the number of
2187 processes, without checking whether the existing limit was already
2188 adequate. (It did check for the limit on file descriptors.) Furthermore,
2189 errors from getrlimit() and setrlimit() were being ignored. Now they are
2190 logged to the main and panic logs, but Exim does carry on, to try to do its
2191 job under whatever limits there are.
2193 18. Imported PCRE 5.0.
2195 19. Trivial typo in log message " temporarily refused connection" (the leading
2198 20. If the log selector return_path_on_delivery was set and an address was
2199 redirected to /dev/null, the delivery process crashed because it assumed
2200 that a return path would always be set for a "successful" delivery. In this
2201 case, the whole delivery is bypassed as an optimization, and therefore no
2204 21. Internal re-arrangement: the function for sending a challenge and reading
2205 a response while authentication was assuming a zero-terminated challenge
2206 string. It's now changed to take a pointer and a length, to allow for
2207 binary data in such strings.
2209 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2211 23. Exim was not respecting finduser_retries when seeking the login of the
2212 uid under which it was called; it was always trying 10 times. (The default
2213 setting of finduser_retries is zero.) Also, it was sleeping after the final
2214 failure, which is pointless.
2216 24. Implemented tls_on_connect_ports.
2218 25. Implemented acl_smtp_predata.
2220 26. If the domain in control=submission is set empty, Exim assumes that the
2221 authenticated id is a complete email address when it generates From: or
2222 Sender: header lines.
2224 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2225 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2226 chown and chgrp in /bin and hostname in /usr/bin.
2228 28. Exim was keeping the "process log" file open after each use, just as it
2229 does for the main log. This opens the possibility of it remaining open for
2230 long periods when the USR1 signal hits a daemon. Occasional processlog
2231 errors were reported, that could have been caused by this. Anyway, it seems
2232 much more sensible not to leave this file open at all, so that is what now
2235 29. The long-running daemon process does not normally write to the log once it
2236 has entered its main loop, and it closes the log before doing so. This is
2237 so that log files can straightforwardly be renamed and moved. However,
2238 there are a couple of unusual error situations where the daemon does write
2239 log entries, and I had neglected to close the log afterwards.
2241 30. The text of an SMTP error response that was received during a remote
2242 delivery was being truncated at 512 bytes. This is too short for some of
2243 the long messages that one sometimes sees. I've increased the limit to
2246 31. It is now possible to make retry rules that apply only when a message has a
2247 specific sender, in particular, an empty sender.
2249 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2250 it possible to be selective about when SMTP synchronization is enforced.
2252 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2254 32. Implemented hosts_connection_nolog.
2256 33. Added an ACL for QUIT.
2258 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2261 35. Added mailbox_size and mailbox_filecount to appendfile.
2263 36. Added control = no_multiline_responses to ACLs.
2265 37. There was a bug in the logic of the code that waits for the clock to tick
2266 in the case where the clock went backwards by a substantial amount such
2267 that the microsecond fraction of "now" was more than the microsecond
2268 fraction of "then" (but the whole seconds number was less).
2270 38. Added support for the libradius Radius client library this is found on
2271 FreeBSD (previously only the radiusclient library was supported).
2277 1. When certain lookups returned multiple values in the form name=value, the
2278 quoting of the values was not always being done properly. Specifically:
2279 (a) If the value started with a double quote, but contained no whitespace,
2281 (b) If the value contained whitespace other than a space character (i.e.
2282 tabs or newlines or carriage returns) it was not quoted.
2283 This fix has been applied to the mysql and pgsql lookups by writing a
2284 separate quoting function and calling it from the lookup code. The fix
2285 should probably also be applied to nisplus, ibase and oracle lookups, but
2286 since I cannot test any of those, I have not disturbed their existing code.
2288 2. A hit in the callout cache for a specific address caused a log line with no
2289 reason for rejecting RCPT. Now it says "Previous (cached) callout
2290 verification failure".
2292 3. There was an off-by-one bug in the queryprogram router. An over-long
2293 return line was truncated at 256 instead of 255 characters, thereby
2294 overflowing its buffer with the terminating zero. As well as fixing this, I
2295 have increased the buffer size to 1024 (and made a note to document this).
2297 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2298 when Exim is waiting for an SMTP response from a remote server, Exim
2299 restarts its select() call on the socket, thereby resetting its timeout.
2300 This is not a problem when such interrupts are rare. Somebody set up a cron
2301 job to run exiwhat every 2 minutes, which is less than the normal select()
2302 timeout (5 or 10 minutes). This meant that the select() timeout never
2303 kicked in because it was always reset. I have fixed this by comparing the
2304 time when an interrupt arrives with the time at the start of the first call
2305 to select(). If more time than the timeout has elapsed, the interrupt is
2306 treated as a timeout.
2308 5. Some internal re-factoring in preparation for the addition of Sieve
2309 extensions (by MH). In particular, the "personal" test is moved to a
2310 separate function, and given an option for scanning Cc: and Bcc: (which is
2311 not set for Exim filters).
2313 6. When Exim created an email address using the login of the caller as the
2314 local part (e.g. when creating a From: or Sender: header line), it was not
2315 quoting the local part when it contained special characters such as @.
2317 7. Installed new OpenBSD configuration files.
2319 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2320 try to make them clearer.
2322 9. Callout options, other than the timeout value, were being ignored when
2323 verifying sender addresses in header lines. For example, when using
2325 verify = header_sender/callout=no_cache
2327 the cache was (incorrectly) being used.
2329 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2330 only the Cygwin environment.
2332 11. When return_path_on_delivery was set as a log selector, if different remote
2333 addresses in the same message used different return paths and parallel
2334 remote delivery occurred, the wrong values would sometimes be logged.
2335 (Whenever a remote delivery process finished, the return path value from
2336 the most recently started remote delivery process was logged.)
2338 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2339 lines when AUTH and/or TLS are in use. This is the "received protocol"
2340 field. Exim used to use "asmtp" for authenticated SMTP, without any
2341 indication (in the protocol name) for TLS use. Now it follows the RFC and
2342 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2343 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2344 names appear in log lines as well as in Received: header lines.
2346 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2347 extensions, and comparison tests, and to fix some bugs.
2349 14. Changes to the "personal" filter test:
2351 (1) The test was buggy in that it was just doing the equivalent of
2352 "contains" tests on header lines. For example, if a user's address was
2353 anne@some.where, the "personal" test would incorrectly be true for
2355 To: susanne@some.where
2357 This test is now done by extracting each address from the header in turn,
2358 and checking the entire address. Other tests that are part of "personal"
2359 are now done using regular expressions (for example, to check local parts
2360 of addresses in From: header lines).
2362 (2) The list of non-personal local parts in From: addresses has been
2363 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2364 taken from the Sieve specification recommendations.
2366 (3) If the message contains any header line starting with "List-" it is
2367 treated as non-personal.
2369 (4) The test for "circular" in the Subject: header line has been removed
2370 because it now seems ill-conceived.
2372 15. Minor typos in src/EDITME comments corrected.
2374 16. Installed latest exipick from John Jetmore.
2376 17. If headers_add on a router specified a text string that was too long for
2377 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2378 of string_sprintf() is now avoided.
2380 18. $message_body_size was not set (it was always zero) when running the DATA
2381 ACL and the local_scan() function.
2383 19. For the "mail" command in an Exim filter, no default was being set for
2384 the once_repeat time, causing a random time value to be used if "once" was
2385 specified. (If the value happened to be <= 0, no repeat happened.) The
2386 default is now 0s, meaning "never repeat". The "vacation" command was OK
2387 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2388 (I found it when inspecting the code).
2390 20. There is now an overall timeout for performing a callout verification. It
2391 defaults to 4 times the callout timeout, which applies to individual SMTP
2392 commands during the callout. The overall timeout applies when there is more
2393 than one host that can be tried. The timeout is checked before trying the
2394 next host. This prevents very long delays if there are a large number of
2395 hosts and all are timing out (e.g. when the network connections are timing
2396 out). The value of the overall timeout can be changed by specifying an
2397 additional sub-option for "callout", called "maxwait". For example:
2399 verify = sender/callout=5s,maxwait=20s
2401 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2402 to the end before writing, but this should make it even safer).
2404 22. Exim was forgetting that it had advertised PIPELINING for the second and
2405 subsequent messages on an SMTP connection. It was also not resetting its
2406 memory on STARTTLS and an internal HELO.
2408 23. When Exim logs an SMTP synchronization error within a session, it now
2409 records whether PIPELINING has been advertised or not.
2411 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2412 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2413 rather than long int.
2415 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2417 26. Added the never_mail option to autoreply.
2423 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2424 crash if the getsockname() call failed; this can happen if a connection is
2425 closed very soon after it is established. The problem was simply in the
2426 order in which certain operations were done, causing Exim to try to write
2427 to the SMTP stream before it had set up the file descriptor. The bug has
2428 been fixed by making things happen in the correct order.
2434 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2435 before the connection was closed, thus losing the rejection response.
2437 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2438 some early Solaris releases, but causes trouble in current releases where
2439 socklen_t is defined.
2441 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2444 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2447 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2448 that releasing the top of it at the end releases what was used for sub-
2449 expansions (unless the block got too big). However, discard this block if
2450 the first thing is a variable or header, so that we can use its block when
2451 it is dynamic (useful for very large $message_headers, for example).
2453 6. Lookups now cache *every* query, not just the most recent. A new, separate
2454 store pool is used for this. It can be recovered when all lookup caches are
2455 flushed. Lookups now release memory at the end of their result strings.
2456 This has involved some general refactoring of the lookup sources.
2458 7. Some code has been added to the store_xxx() functions to reduce the amount
2459 of flapping under certain conditions.
2461 8. log_incoming_interface used to affect only the <= reception log lines. Now
2462 it causes the local interface and port to be added to several more SMTP log
2463 lines, for example "SMTP connection from", and rejection lines.
2465 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2467 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2469 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2470 could be overwritten at the end of the current message (or the start of a
2471 new message if it was set in a HELO ACL). The value is now preserved for
2472 the duration of the SMTP connection.
2474 12. If a transport had a headers_rewrite setting, and a matching header line
2475 contained an unqualified address, that address was qualified, even if it
2476 did not match any rewriting rules. The underlying bug was that the values
2477 of the flags that permit the existence of unqualified sender and recipient
2478 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2479 non-local messages, and by -bnq for local messages) were not being
2480 preserved with the message after it was received.
2482 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2483 "next input=" as part of the text comprising the host identity instead of
2484 the correct text. The code was using the same buffer for two different
2485 strings. However, depending on which order the printing function evaluated
2486 its arguments, the bug did not always show up. Under Linux, for example, my
2487 test suite worked just fine.
2489 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2490 doesn't work with some older versions of Perl. It has been changed to "my",
2491 which in any case is probably the better facility to use.
2493 15. A really picky compiler found some instances of statements for creating
2494 error messages that either had too many or two few arguments for the format
2497 16. The size of the buffer for calls to the DNS resolver has been increased
2498 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2499 for addresses that have a lot of PTR records. This alleviates a problem; it
2500 does not fully solve it.
2502 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2503 buffer now truncates the list and logs the incident, which is the same
2504 action as happens when Exim is looking up a host name and its aliases.
2505 Previously in this situation something unpredictable would happen;
2506 sometimes it was "internal error: store_reset failed".
2508 18. If a server dropped the connection unexpectedly when an Exim client was
2509 using GnuTLS and trying to read a response, the client delivery process
2510 crashed while trying to generate an error log message.
2512 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2513 string, for example:
2515 warn message = H1: something\nH2: something
2517 the text was added as a single header line from Exim's point of view
2518 though it ended up OK in the delivered message. However, searching for the
2519 second and subsequent header lines using $h_h2: did not work. This has been
2520 fixed. Similarly, if a system filter added multiple headers in this way,
2521 the routers could not see them.
2523 20. Expanded the error message when iplsearch is called with an invalid key to
2524 suggest using net-iplsearch in a host list.
2526 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2527 ACLs are no longer actually imposed (and a message to that effect is
2530 22. If a "gecos" field in a passwd entry contained escaped characters, in
2531 particular, if it contained a \" sequence, Exim got it wrong when building
2532 a From: or a Sender: header from that name. A second bug also caused
2533 incorrect handling when an unquoted " was present following a character
2534 that needed quoting.
2536 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2537 was not being matched caselessly.
2539 24. Arranged for all hyphens in the exim.8 source to be escaped with
2542 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2543 properly. Recipient callout cache records were still being keyed to include
2544 the sender, even when use_sender was set false. This led to far more
2545 callouts that were necessary. The sender is no longer included in the key
2546 when use_sender is false.
2548 26. Added "control = submission" modifier to ACLs.
2550 27. Added the ${base62d: operator to decode base 62 numbers.
2552 28. dnsdb lookups can now access SRV records.
2554 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2555 the configuration file.
2557 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2558 (-v) mode. This makes the output for a verbose queue run more intelligible.
2560 31. Added a use_postmaster feature to recipient callouts.
2562 32. Added the $body_zerocount variable, containing the number of binary zero
2563 bytes in the message body.
2565 33. The time of last modification of the "new" subdirectory is now used as the
2566 "mailbox time last read" when there is a quota error for a maildir
2569 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2571 35. Added +ignore_unknown as a special item in host lists.
2573 36. Code for decoding IPv6 addresses in host lists is now included, even if
2574 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2575 address was recognized as an IP address, but was then not correctly decoded
2576 into binary, causing unexpected and incorrect effects when compared with
2583 1. Very minor rewording of debugging text in manualroute to say "list of
2584 hosts" instead of "hostlist".
2586 2. If verify=header_syntax was set, and a header line with an unqualified
2587 address (no domain) and a large number of spaces between the end of the
2588 name and the colon was received, the reception process suffered a buffer
2589 overflow, and (when I tested it) crashed. This was caused by some obsolete
2590 code that should have been removed. The fix is to remove it!
2592 3. When running in the test harness, delay a bit after writing a bounce
2593 message to get a bit more predictability in the log output.
2595 4. Added a call to search_tidyup() just before forking a reception process. In
2596 theory, someone could use a lookup in the expansion of smtp_accept_max_
2597 per_host which, without the tidyup, could leave open a database connection.
2599 5. Added the variables $recipient_data and $sender_data which get set from a
2600 lookup success in an ACL "recipients" or "senders" condition, or a router
2601 "senders" option, similar to $domain_data and $local_part_data.
2603 6. Moved the writing of debug_print from before to after the "senders" test
2606 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2607 problems for message scanning, either using a data ACL, or using
2608 local_scan() because the Received: header was not generated till after they
2609 were called (in order to set the time as the time of reception completion).
2610 I have revised the way this works. The header is now generated after the
2611 body is received, but before the ACL or local_scan() are called. After they
2612 are run, the timestamp in the header is updated.
2618 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2619 before starting a queue runner without re-exec. This happened only when
2620 deliver_drop_privilege was set or when the Exim user was set to root. The
2621 effect of the bug was that timeouts during subsequent deliveries caused
2622 crashes instead of being properly handled. The handler is now left at its
2623 default (and expected) setting.
2625 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2626 message, again when deliver_drop_privilege is set or Exim is run as root.
2627 The bug described in (1) was not present in this case, but the tidying up
2628 of the other signals was missing. I have made the two cases consistent.
2630 3. The ignore_target_hosts setting on a manualroute router was being ignored
2631 for hosts that were looked up using the /MX notation.
2633 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2636 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2637 operated on the sender address. After changing the $sender_address to <>
2638 for the sender address verify, Exim was re-instated it as the original
2639 (before rewriting) address, but remembering that it had rewritten it, so it
2640 wasn't rewriting it again. This bug also had the effect of breaking the
2641 sender address verification caching when the sender address was rewritten.
2643 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2644 This has been changed so that if the ip literal address matches
2645 ignore_target_hosts, the router declines.
2647 7. Added expansion conditions match_domain, match_address, and match_local_
2648 part (NOT match_host).
2650 8. The placeholder for the Received: header didn't have a length field set.
2652 9. Added code to Exim itself and to exim_lock to test for a specific race
2653 condition that could lead to file corruption when using MBX delivery. The
2654 issue is with the lockfile that is created in /tmp. If this file is removed
2655 after a process has opened it but before that process has acquired a lock,
2656 there is the potential for a second process to recreate the file and also
2657 acquire a lock. This could lead to two Exim processes writing to the file
2658 at the same time. The added code performs the same test as UW imapd; it
2659 checks after acquiring the lock that its file descriptor still refers to
2660 the same named file.
2662 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2663 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2666 11. Added the smtp_active_hostname option. If used, this will typically be made
2667 to depend on the incoming interface address. Because $interface_address is
2668 not set up until the daemon has forked a reception process, error responses
2669 that can happen earlier (such as "too many connections") no longer contain
2672 12. If an expansion in a condition on a "warn" statement fails because a lookup
2673 defers, the "warn" statement is abandoned, and the next ACL statement is
2674 processed. Previously this caused the whole ACL to be aborted.
2676 13. Added the iplsearch lookup type.
2678 14. Added ident_timeout as a log selector.
2680 15. Added tls_certificate_verified as a log selector.
2682 16. Added a global option tls_require_ciphers (compare the smtp transport
2683 option of the same name). This controls incoming TLS connections.
2685 17. I finally figured out how to make tls_require_ciphers do a similar thing
2686 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2687 before starting the TLS session.
2689 18. Tabs are now shown as \t in -bP output.
2691 19. If the log selector return_path_on_delivery was set, Exim crashed when
2692 bouncing a message because it had too many Received: header lines.
2694 20. If two routers both had headers_remove settings, and the first one included
2695 a superfluous trailing colon, the final name in the first list and the
2696 first name in the second list were incorrectly joined into one item (with a
2697 colon in the middle).
2703 1. Added -C and -D options to the exinext utility, mainly to make it easier
2704 to include in the automated testing, but these could be helpful when
2705 multiple configurations are in use.
2707 2. The exinext utility was not formatting the output nicely when there was
2708 an alternate port involved in the retry record key, nor when there was a
2709 message id as well (for retries that were specific to a specific message
2710 and a specific host). It was also confused by IPv6 addresses, because of
2711 the additional colons they contain. I have fixed the IPv4 problem, and
2712 patched it up to do a reasonable job for IPv6.
2714 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2715 delivery, the log line now contains "pipelined" if PIPELINING was used.
2717 4. An SMTP transport process used to panic and die if the bind() call to set
2718 an explicit outgoing interface failed. This has been changed; it is now
2719 treated in the same way as a connect() failure.
2721 5. A reference to $sender_host_name in the part of a conditional expansion
2722 that was being skipped was still causing a DNS lookup. This no longer
2725 6. The def: expansion condition was not recognizing references to header lines
2726 that used bh_ and bheader_.
2728 7. Added the _cache feature to named lists.
2730 8. The code for checking quota_filecount in the appendfile transport was
2731 allowing one more file than it should have been.
2733 9. For compatibility with Sendmail, the command line option
2741 and sets the incoming protocol and host name (for trusted callers). The
2742 host name and its colon can be omitted when only the protocol is to be set.
2743 Note the Exim already has two private options, -pd and -ps, that refer to
2744 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2745 "s", but I don't think that's a major issue.
2747 10. A number of refactoring changes to the code, none of which should affect
2750 (a) The number of logging options was getting close to filling up the
2751 32-bit word that was used as a bit map. I have split them into two classes:
2752 those that are passed in the argument to log_write(), and those that are
2753 only ever tested independently outside of that function. These are now in
2754 separate 32-bit words, so there is plenty of room for expansion again.
2755 There is no change in the user interface or the logging behaviour.
2757 (b) When building, for example, log lines, the code previously used a
2758 macro that called string_cat() twice, in order to add two strings. This is
2759 not really sufficiently general. Furthermore, there was one instance where
2760 it was actually wrong because one of the argument was used twice, and in
2761 one call a function was used. (As it happened, calling the function twice
2762 did not affect the overall behaviour.) The macro has been replaced by a
2763 function that can join an arbitrary number of extra strings onto a growing
2766 (c) The code for expansion conditions now uses a table and a binary chop
2767 instead of a serial search (which was left over from when there were very
2768 few conditions). Also, it now recognizes conditions like "pam" even when
2769 the relevant support is not compiled in: a suitably worded error message is
2770 given if an attempt is made to use such a condition.
2772 11. Added ${time_interval:xxxxx}.
2774 12. A bug was causing one of the ddress fields not to be passed back correctly
2775 from remote delivery subprocesses. The field in question was not being
2776 subsequently used, so this caused to problems in practice.
2778 13. Added new log selectors queue_time and deliver_time.
2780 14. Might have fixed a bug in maildirsizefile handling that threw up
2781 "unexpected character" debug warnings, and recalculated the data
2782 unnecessarily. In any case, I expanded the warning message to give more
2785 15. Added the message "Restricted characters in address" to the statements in
2786 the default ACL that block characters like @ and % in local parts.
2788 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2789 Three changes have been made:
2791 (a) There was a serious bug; a negative response to MAIL caused the whole
2792 recipient domain to be cached as invalid, thereby blocking all messages
2793 to all local parts at the same domain, from all senders. This bug has
2794 been fixed. The domain is no longer cached after a negative response to
2795 MAIL if the sender used is not empty.
2797 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2800 (c) A new callout option, "use_sender" has been added for people who want
2801 the modified behaviour.
2807 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2810 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2811 indeed breaks things for older releases.
2813 3. Added additional logging to the case where there is a problem reading data
2814 from a filter that is running in a subprocess using a pipe, in order to
2815 try to track down a specific problem.
2817 4. Testing facility fudge: when running in the test harness and attempting
2818 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2819 getting "No route to host". Convert this to a timeout.
2821 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2824 6. Some OS don't have socklen_t but use size_t instead. This affects the
2825 fifth argument of getsockopt() amongst other things. This is now
2826 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2827 can be set for individual OS. I have set it for SunOS5, OSF1, and
2828 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2829 some earlier ones do not.
2831 7. Change 4.30/15 was not doing the test caselessly.
2833 8. The standard form for an IPv6 address literal was being rejected by address
2834 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2835 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2836 this, as well as the form without the "IPv6" on the front (but only when
2837 address literals are enabled, of course).
2839 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2841 10. Exim crashed if a message with an empty sender address specified by -f
2842 encountered a router with an errors_to setting. This could be provoked only
2843 by a command such as
2847 where an empty string was supplied; "<>" did not hit this bug.
2849 11. Installed PCRE release 4.5.
2851 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2852 remained set. It is now erased.
2854 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2855 times from message ids (which are base 36 rather than the normal 62).
2857 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2858 were being counted as actual protocol errors, and logged if the log
2859 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2860 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2861 and DATA following a set of rejected RCPTs do not count as protocol errors.
2862 In other words, Exim assumes they were pipelined, though this may not
2863 actually be the case. Of course, in all cases the client gets an
2864 appropriate error code.
2866 15. If a lookup fails in an ACL condition, a message about the failure may
2867 be available; it is used if testing the ACL cannot continue, because most
2868 such messages specify what the cause of the deferral is. However, some
2869 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2870 that caused an old message to be retained and used if a later statement
2871 caused a defer, replacing the real cause of the deferral.
2873 16. If an IP address had so many PTR records that the DNS lookup buffer
2874 was not large enough to hold them, Exim could crash while trying to process
2875 the truncated data. It now detects and logs this case.
2877 17. Further to 4.21/58, another change has been made: if (and only if) the
2878 first line of a message (the first header line) ends with CRLF, a bare LF
2879 in a subsequent header line has a space inserted after it, so as not to
2880 terminate the header.
2882 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2883 unnecessarily, used atoi() instead of strtol(), and didn't check the
2884 termination when getting file sizes from file names by regex.
2886 19. Completely re-implemented the support for maildirsize files, in the light
2887 of a number of problems with the previous contributed implementation
2888 (4.30/29). In particular:
2890 . If the quota is zero, the maildirsize file is maintained, but no quota is
2893 . If the maildir directory does not exist, it is created before any attempt
2894 to write a maildirsize file.
2896 . The quota value in the file is just a cache; if the quota is changed in
2897 the transport, the new value overrides.
2899 . A regular expression is available for excluding directories from the
2902 20. The autoreply transport checks the characters in options that define the
2903 message's headers; it allows continued headers, but it was checking with
2904 isspace() after an embedded newline instead of explicitly looking for a
2907 21. If all the "regular" hosts to which an address was routed had passed their
2908 expiry times, and had not reached their retry times, the address was
2909 bounced, even if fallback hosts were defined. Now Exim should go on to try
2912 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2913 equivalent code in the SMTP transport. Some hosts send humungous responses
2914 to HELO/EHLO, more than 1024 it seems.
2916 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2917 gives compiler warnings in some environments. I've now done it "properly",
2920 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2921 (because of problems with the built-in one) was declared to return uschar *
2922 instead of char *, causing compiler failure.
2924 25. Fixed a file descriptor leak when processing alias/forward files.
2926 26. Fixed a minor format string issue in dbfn.c.
2928 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2930 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2931 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2934 29. When a sender address is verified, it is cached, to save repeating the test
2935 when there is more than one recipient in a message. However, when the
2936 verification involves a callout, it is possible for different callout
2937 options to be set for different recipients. It is too complicated to keep
2938 track of this in the cache, so now Exim always runs a verification when a
2939 callout is required, relying on the callout cache for the optimization.
2940 The overhead is duplication of the address routing, but this should not be
2943 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2944 to be verified with callout=postmaster, and the main callout worked but the
2945 postmaster check failed, the verification correctly failed. However, if a
2946 subsequent RCPT command asked for sender verification *without* the
2947 postmaster check, incorrect caching caused this verification also to fail,
2950 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2951 it was not caching the DNS options (qualify_single, search_parents) that
2952 were used when the lookup failed. A subsequent lookup with different
2953 options therefore always gave the same answer, though there were cases
2954 where it should not have. (Example: a "domains = !$mx_any" option on a
2955 dnslookup router: the "domains" option is always processed without any
2956 widening, but the router might have qualify_single set.) Now Exim uses the
2957 cached value only when the same options are set.
2959 32. Added John Jetmore's "exipick" utility to the distribution.
2961 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2962 than a timeout (e.g. a certificate is required, and is not provided), an
2963 Exim server now closes the connection immediately. Previously it waited for
2964 the client to close - but if the client is SSL, it seems that they each
2965 wait for each other, leading to a delay before one of them times out.
2967 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2968 maintained 0.8.x compatibility because I don't think many are using it, and
2969 it is clearly obsolete.
2971 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2974 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2975 client certificate was expired. A simple patch fixes this, though I don't
2976 understand the full logic of why the verify callback is called multiple
2979 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2980 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2981 which causes problems with some clients (such as the Certicom SSL Plus
2982 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2983 disables the coutermeasure allowing Eudora to connect."
2985 38. Exim was not checking that a write() to a log file succeeded. This could
2986 lead to Bad Things if a log got too big, in particular if it hit a file
2987 size limit. Exim now panics and dies if it cannot write to a log file, just
2988 as it does if it cannot open a log file.
2990 39. Modified OS/Makefile-Linux so that it now contains
2992 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2994 The two -D definitions ensure that Exim is compiled with large file
2995 support, which makes it possible to handle log files that are bigger than
2998 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2999 instance) a domain was checked against a named list that involved a lookup,
3000 causing $domain_data to be set, then another domain was checked against the
3001 same list, then the first domain was re-checked, the value of $domain_data
3002 after the final check could be wrong. In particular, if the second check
3003 failed, it could be set empty. This bug probably also applied to
3006 41. The strip_trailing_dot option was not being applied to the address given
3007 with the -f command-line option.
3009 42. The code for reading a message's header from the spool was incrementing
3010 $received_count, but never initializing it. This meant that the value was
3011 incorrect (doubled) while delivering a message in the same process in which
3012 it was received. In the most common configuration of Exim, this never
3013 happens - a fresh exec is done - but it can happen when
3014 deliver_drop_privilege is set.
3016 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3017 it now includes up to 150 characters of the unexpected data in the log
3020 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3021 and building data strings. The size of both of these buffers was 10 000
3022 bytes - far larger than anybody would *ever* want, thought I. Needless to
3023 say, somebody hit the limit. I have increased the maximum line length to
3024 20 000 and the maximum data length of concatenated lines to 100 000. I have
3025 also fixed two bugs, because there was no checking on these buffers. Tsk,
3026 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3027 buffer is too small.
3029 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3030 lsearch lookups. Now it does.
3032 46. When parsing a route_list item in a manualroute router, a fixed-length
3033 buffer was used for the list of hosts. I made this 1024 bytes long,
3034 thinking that nobody would ever have a list of hosts that long. Wrong.
3035 Somebody had a whole pile of complicated expansion conditions, and the
3036 string was silently truncated, leading to an expansion error. It turns out
3037 that it is easier to change to an unlimited length (owing to other changes
3038 that have happened since this code was originally written) than to build
3039 structure for giving a limitation error. The length of the item that
3040 expands into the list of hosts is now unlimited.
3042 47. The lsearch lookup could not handle data where the length of text line was
3043 more than 4095 characters. Such lines were truncated, leading to shortened
3044 data being returned. It should now handle lines of any length.
3046 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3047 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3050 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3051 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3052 no longer happens in comment lines. A list of replacements is now placed
3053 at the head of all of the source files, except those whose only change is
3054 to replace PERL_COMMAND in the very first #! line.
3056 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3057 messages on the queue, with a bottom-up merge sort, using code contributed
3058 by Michael Haardt. This should make operations like -bp somewhat faster on
3059 large queues. It won't affect queue runners, except when queue_run_in_order
3062 51. Installed eximstats 1.31 in the distribution.
3064 52. Added support for SRV lookups to the dnslookup router.
3066 53. If an ACL referred to $message_body or $message_body_end, the value was not
3067 reset for any messages that followed in the same SMTP session.
3069 54. The store-handling optimization for building very long strings was not
3070 differentiating between the different store pools. I don't think this
3071 actually made any difference in practice, but I've tidied it.
3073 55. While running the routers to verify a sender address, $sender_address
3074 was still set to the sender address. This is wrong, because when routing to
3075 send a bounce to the sender, it would be empty. Therefore, I have changed
3076 it so that, while verifying a sender address, $sender_address is set to <>.
3077 (There is no change to what happens when verifying a recipient address.)
3079 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3080 target A or AAAA records (if not already returned) without resetting the
3081 qualify_single or search_parents options of the DNS resolver. These are
3082 inappropriate in this case because the targets of MX and SRV records must
3083 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3084 target that, when qualified, matched something in the local domain. These
3085 two options are now turned off when doing these lookups.
3087 57. It seems that at least some releases of Reiserfs (which does not have the
3088 concept of a fixed number of inodes) returns zero and not -1 for the
3089 number of available inodes. This interacted badly with check_spool_inodes,
3090 which assumed that -1 was the "no such thing" setting. What I have done is
3091 to check that the total number of inodes is greater than zero before doing
3092 the test of how many are available.
3094 58. When a "warn" ACL statement has a log_message modifier, the message is
3095 remembered, and not repeated. This is to avoid a lot of repetition when a
3096 message has many recipients that cause the same warning to be written.
3097 Howewer, Exim was preserving the list of already written lines for an
3098 entire SMTP session, which doesn't seem right. The memory is now reset if a
3099 new message is started.
3101 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3102 debugging output unless log_rewrite was also set.
3104 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3105 of (int)(handle) when we know that handle contains (void *)(-1).
3107 61. The Exim daemon panic-logs an error return when it closes the incoming
3108 connection. However "connection reset by peer" seems to be common, and
3109 isn't really an error worthy of noting specially, so that particular error
3112 62. When Exim is trying to find all the local interfaces, it used to panic and
3113 die if the ioctl to get the interface flags failed. However, it seems that
3114 on at least one OS (Solaris 9) it is possible to have an interface that is
3115 included in the list of interfaces, but for which you get a failure error
3116 for this call. This happens when the interface is not "plumbed" into a
3117 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3118 failure of the "get flags" call assumes that the interface is down.
3120 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3121 makes life easier for people who are doing arithmetic on fields extracted
3122 from dates, where you often get leading zeros that should not be
3123 interpreted as octal.
3125 64. Added qualify_domain to the redirect router, to override the global
3128 65. If a pathologically long header line contained very many addresses (the
3129 report of this problem mentioned 10 000) and each of them was rewritten,
3130 Exim could use up a very large amount of memory. (It kept on making new
3131 copies of the header line as it rewrote, and never released the old ones.)
3132 At the expense of a bit more processing, the header rewriting function has
3133 been changed so that it no longer eats memory in this way.
3135 66. The generation of the Received: header has been moved from the time that a
3136 message starts to be received, to the time that it finishes. The timestamp
3137 in the Received: header should now be very close to that of the <= log
3138 line. There are two side-effects of this change:
3140 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3141 logged header lines no longer include the local Received: line, because
3142 it has not yet been created. The same applies to a copy of the message
3143 that is returned to a non-SMTP sender when a message is rejected.
3145 (b) When a filter file is tested using -bf, no additional Received: header
3146 is added to the test message. After some thought, I decided that this
3149 This change does not affect the value of $received_for. It is still set
3150 after address rewriting, but before local_scan() is called.
3152 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3154 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3155 gave an unhelpful panic error message, and a defer error. I have managed to
3156 change this behaviour so that it now rejects any supplied certificate,
3157 which seems right, as the list of acceptable certificates is empty.
3159 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3160 gave an unhelpful defer error. I have not managed to make this reject any
3161 supplied certificates, but the error message it gives is "no certificate
3162 supplied", which is not helpful.
3164 70. exigrep's output now also includes lines that are not associated with any
3165 message, but which match the given pattern. Implemented by a patch from
3166 Martin Sluka, which also tidied up the Perl a bit.
3168 71. Recipient callout verification, like sender verification, was using <> in
3169 the MAIL FROM command. This isn't really the right thing, since the actual
3170 sender may affect whether the remote host accepts the recipient or not. I
3171 have changed it to use the actual sender in the callout; this means that
3172 the cache record is now keyed on a recipient/sender pair, not just the
3173 recipient address. There doesn't seem to be a real danger of callout loops,
3174 since a callout by the remote host to check the sender would use <>.
3175 [SEE ABOVE: changed after hitting problems.]
3177 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3178 temporary errors. However, in the case of such a code being given after
3179 the end of a data transmission (i.e. after ".") Exim was failing to write
3180 a retry record for the message. (Yes, there was some broken host that was
3181 actually sending 8xx at this point.)
3183 73. An unknown lookup type in a host list could cause Exim to panic-die when
3184 the list was checked. (An example that provoked this was putting <; in the
3185 middle of a list instead of at the start.) If this happened during a DATA
3186 ACL check, a -D file could be left lying around. This kind of configuration
3187 error no longer causes Exim to die; instead it causes a defer errror. The
3188 incident is still logged to the main and panic logs.
3190 74. Buglet left over from Exim 3 conversion. The message "too many messages
3191 in one connection" was written to the rejectlog but not the mainlog, except
3192 when address rewriting (yes!) was being logged.
3194 75. Added write_rejectlog option.
3196 76. When a system filter was run not as root (that is, when system_filter_user
3197 was set), the values of the $n variables were not being returned to the
3198 main process; thus, they were not subsequently available in the $sn
3201 77. Added +return_path_on_delivery log selector.
3203 78. A connection timeout was being treated differently from recipients deferred
3204 when testing hosts_max_try with a message that was older than the host's
3205 retry timeout. (The host should not be counted, thus allowing all hosts to
3206 be tried at least once before bouncing.) This may have been the cause of an
3207 occasionally reported bug whereby a message would remain on the queue
3208 longer than the retry timeout, but would be bounced if a delivery was
3209 forced. I say "may" because I never totally pinned down the problem;
3210 setting up timeout/retry tests is difficult. See also the next item.
3212 79. The ultimate address timeout was not being applied to errors that involved
3213 a combination of host plus message (for example, a timeout on a MAIL
3214 command). When an address resolved to a number of possible hosts, and they
3215 were not all tried for each delivery (e.g. because of hosts_max_try), a
3216 message could remain on the queue longer than the retry timeout.
3218 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3221 81. Fixed an obscure SMTP outgoing bug which required at least the following
3222 conditions: (a) there was another message waiting for the same server;
3223 (b) the server returned 5xx to all RCPT commands in the first message so
3224 that the message was not completed; (c) the server dropped the connection
3225 or gave a negative response to the RSET that Exim sends to abort the
3226 transaction. The observed case was a dropped connection after DATA that had
3227 been sent in pipelining mode. That is, the server had advertised PIPELINING
3228 but was not implementing it correctly. The effect of the bug was incorrect
3229 behaviour, such as trying another host, and this could lead to a crash.
3235 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3236 and daemon.c were passed as pointers to ints; they should have been
3237 pointers to socklen_t variables (which are typically unsigned ints).
3239 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3242 3. Fixed a really odd bug that affected only the testing scheme; patching a
3243 certain fixed string in the binary changed the value of another string that
3244 happened to be identical to the end of the original first string.
3246 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3247 name", it returns that address as the IP address. On some operating
3248 systems (e.g. Solaris), it also passes back the IP address string as the
3249 "host name". However, on others (e.g. Linux), it passes back an empty
3250 string. Exim wasn't checking for this, and was changing the host name to an
3251 empty string, assuming it had been canonicized.
3253 5. Although rare, it is permitted to have more than one PTR record for a given
3254 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3255 all the names associated with an address, because they do in Solaris.
3256 However, it seems that they do not in Linux for data that comes from the
3257 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3258 I found this out when I moved to a new Linux workstation and tried to run
3259 the Exim test suite.
3261 To get round this problem I have changed the code so that it now does its
3262 own call to the DNS to look up PTR records when searching for a host name.
3263 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3264 addresses that are only in /etc/hosts are still found.
3266 This behaviour is, however, controlled by an option called host_lookup_
3267 order, which defaults to "bydns:byaddr". If people want to use the other
3268 order, or indeed, just use one or the other means of lookup, they can
3269 specify it in this variable.
3271 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3272 some operating systems, this comes back from gethostbyaddr() as an empty
3273 string, and this is what Exim used to test for. However, it seems that in
3274 other systems, "." is yielded. Exim now tests for this case too.
3276 7. The values of check_spool_space and check_log_space are now held internally
3277 as a number of kilobytes instead of an absolute number of bytes. If a
3278 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3279 kilobyte. This means that much larger values can be stored.
3281 8. Exim monitor: an attempt to get the action menu when not actually pointing
3282 at a message produces an empty menu entitled "No message selected". This
3283 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3284 no entries in it ("Shell widget menu has zero width and/or height"). So I
3285 have added a single, blank menu entry in this case.
3287 9. Added ${quote_local_part.
3289 10. MIME decoding is now applied to the contents of Subject: header lines when
3292 11. Now that a reference to $sender_host_address automatically causes a reverse
3293 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3294 host lookup before query-style lookups in lists that might use this
3295 variable. This has therefore been abolished, and the "net-" prefix is no
3296 longer necessary for query-style lookups.
3298 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3299 have been a typo for LFLAGS, so it has been changed.
3301 13. The install script calls Exim with "-C /dev/null" in order to find the
3302 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3303 to be output. Howeve, since Exim outputs its version number before the
3304 error, it didn't break the script. It just looked ugly. I fixed this by
3305 always allowing "-C /dev/null" if the caller is root.
3307 14. Ignore overlarge ACL variable number when reading spool file - insurance
3308 against a later release with more variables having written the file.
3310 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3311 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3312 this, as well as the form without the "IPv6" on the front.
3314 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3315 OS/Makefile-Darwin file.
3317 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3318 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3320 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3321 variable. However, some OS use ints for inodes. I've added cast to long int
3322 to get rid of the compiler warning.
3324 19. I had forgotten to lock out "/../" in configuration file names when
3325 ALT_CONFIG_PREFIX was set.
3327 20. Routers used for verification do not need to specify transports. However,
3328 if such a router generated a host list, and callout was configured, Exim
3329 crashed, because it could not find a port number from the (non-existent)
3330 transport. It now assumes port 25 in this circumstance.
3332 21. Added the -t option to exigrep.
3334 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3335 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3336 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3337 them is set, LOOKUP_LSEARCH is forced.
3339 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3341 24. Added sender and host information to the "rejected by local_scan()" log
3342 line; previously there was no indication of these.
3344 25. Added .include_if_exists.
3346 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3347 Linux. It turns out that not all file systems support this. Apparently some
3348 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3349 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3350 not supported on the file descriptor, is now ignored when Exim is trying to
3351 sync a directory. This applies only to Linux.
3353 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3355 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3356 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3359 29. Added support for maildirsize files from supplied patch (modified a bit).
3361 30. The use of :fail: followed by an empty string could lead Exim to respond to
3362 sender verification failures with (e.g.):
3364 550 Verification failed for <xxx>
3365 550 Sender verify failed
3367 where the first response line was missing the '-' that indicates it is not
3368 the final line of the response.
3370 31. The loop for finding the name of the user that called Exim had a hardwired
3371 limit of 10; it now uses the value of finduser_retries, which is used for
3372 all other user lookups.
3374 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3377 33. Exim was neglecting to zero errno before one call of strtol() when
3378 expanding a string and expecting an integer value. On some systems this
3379 resulted in spurious "integer overflow" errors. Also, it was casting the
3380 result into an int without checking.
3382 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3383 did not work. The code looks as if it has *never* worked, though it appears
3384 to have been documented since at least releast 1.62. I have made it work.
3386 35. The "timeout_DNS" error in retry rules, also documented since at least
3387 1.62, also never worked. As it isn't clear exactly what this means, and
3388 clearly it isn't a major issue, I have abolished the feature by treating it
3389 as "timeout", and writing a warning to the main and panic logs.
3391 36. The display of retry rules for -brt wasn't always showing the error code
3394 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3395 timeout_connect_A, timeout_connect_MX.
3397 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3398 to the empty sender.
3400 39. The daemon was not analysing the content of -oX till after it had closed
3401 stderr and disconnected from the controlling terminal. This meant that any
3402 syntax errors were only noted on the panic log, and the return code from
3403 the command was 0. By re-arranging the code a little, I've made the
3404 decoding happen first, so such errors now appear on stderr, and the return
3405 code is 1. However, the actual setting up of the sockets still happens in
3406 the disconnected process, so errors there are still only recorded on the
3409 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3410 connections (as happens on some IP stacks) was logged at start up time as
3411 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3412 it from "IPv6 and IPv4", which means that two separate sockets are being
3415 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3416 says whether AF_INET or AF_INET6 was passed as an argument.
3418 42. Exiwhat output was messed up when time zones were included in log
3421 43. Exiwhat now gives more information about the daemon's listening ports,
3422 and whether -tls-on-connect was used.
3424 44. The "port" option of the smtp transport is now expanded.
3426 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3427 silently ignored. Now an error message is written to the main and panic
3430 46. There's a new ACL modifier called "logwrite" which writes to a log file
3431 as soon as it is encountered.
3433 47. Added $local_user_uid and $local_user_gid at routing time.
3435 48. Exim crashed when trying to verify a sender address that was being
3438 49. Exim was recognizing only a space character after ".include". It now also
3439 recognizes a tab character.
3441 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3442 extracting the relevant information from the specification. The man page no
3443 longer contains scrambled data for the -d option, and I've added a section
3444 at the front about calling Exim under different names.
3446 51. Added "extra_headers" argument to the "mail" command in filter files.
3448 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3451 53. Installed eximstats 1.29.
3453 54. Added transport_filter_timeout as a generic transport option.
3455 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3456 Cc: header lines. This was required by RFC 822, but it not required by RFC
3459 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3460 incoming messages that did not have them. Now it does so only if the
3461 message originates locally, that is, if there is no associated remote host
3462 address. When Resent- header lines are present, this applies to the Resent-
3463 lines rather than the non-Resent- lines.
3465 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3466 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3468 58. Messages for configuration errors now include the name of the main
3469 configuration file - useful now that there may be more than one file in a
3470 list (.included file names were always shown).
3472 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3473 for those rare installations that do not start the daemon as root or run it
3474 setuid root. I've cut out the call to initgroups() if the daemon is not
3477 60. The Exim user and group can now be bound into the binary as text strings
3478 that are looked up at the start of Exim's processing.
3480 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3482 62. Added $mailstore_basename variable.
3484 63. Installed patch to sieve.c from Michael Haardt.
3486 64. When Exim failed to open the panic log after failing to open the main log,
3487 the original message it was trying to log was written to stderr and debug
3488 output, but if they were not available (the usual case in production), it
3489 was lost. Now it is written to syslog before the two lines that record the
3490 failures to open the logs.
3492 65. Users' Exim filters run in subprocesses under the user's uid. It is
3493 possible for a "deliver" command or an alias in a "personal" command to
3494 provoke an address rewrite. If logging of address rewriting is configured,
3495 this fails because the process is not running as root or exim. There may be
3496 a better way of dealing with this, but for the moment (because 4.30 needs
3497 to be released), I have disabled address rewrite logging when running a
3498 filter in a non-root, non-exim process.
3504 1. The buildconfig auxiliary program wasn't quoting the value set for
3505 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3506 not defined. This bug was masked in 4.22 by the effect that was fixed in
3509 2. Some messages that were rejected after a message id was allocated were
3510 shown as "incomplete" by exigrep. It no longer does this for messages that
3511 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3513 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3514 have allow_domain_literals set, the ID did not get logged in the <= line.
3515 Domain literals are now always recognized in Message-ID: header lines.
3517 4. The first argument for a ${extract expansion item is the key name or field
3518 number. Leading and trailing spaces in this item were not being ignored,
3519 causing some misleading effects.
3521 5. When deliver_drop_privilege was set, single queue runner processes started
3522 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3523 same command in the process it spins off) were not dropping privilege.
3525 6. When the daemon running as "exim" started a queue runner, it always
3526 re-executed Exim in the spun-off process. This is a waste of effort when
3527 deliver_drop_privilege is set. The new process now just calls the
3528 queue-runner function directly.
3534 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3537 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3538 ignored. Though the use of -oP was forcing the writing of a pid file, it
3539 was always written to the default place.
3541 3. If the message "no IP address found for host xxxx" is generated during
3542 incoming verification, it is now followed by identification of the incoming
3543 connection (so you can more easily find what provoked it).
3545 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3547 5. Added some features to "harden" Exim a bit more against certain attacks:
3549 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3550 be put in Local/Makefile. This is like the never_users runtime option,
3551 but it cannot be overridden. The default setting is "root".
3553 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3554 prefix string with which any file named in a -C command line option
3557 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3558 is retained for -C and -D only if the caller of Exim is root. Without
3559 it, the exim user may also use -C and -D and retain privilege.
3561 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3562 command line option is disabled.
3564 6. Macro names set by the -D option must start with an upper case letter, just
3565 like macro names defined in the configuration file.
3567 7. Added "dereference=" facility to LDAP.
3569 8. Two instances of the typo "uknown" in the source files are fixed.
3571 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3572 the Configure-Makefile script screwed up while processing it.
3574 10. Incorporated PCRE 4.4.
3576 11. The SMTP synchronization check was not operating right at the start of an
3577 SMTP session. For example, it could not catch a HELO sent before the client
3578 waited for the greeting. There is now a check for outstanding input at the
3579 point when the greeting is written. Because of the duplex, asynchronous
3580 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3581 way, but not yet received, when the check is performed.
3583 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3584 on TCP/IP sockets, because this apparently causes some broken clients to
3587 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3588 unchanged) from the Cygwin maintainer.
3590 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3591 was supported instead of testing for mbx. Effectively a typo.
3593 15. The spa authenticator server code was not checking that the input it
3594 received was valid base64.
3596 16. The debug output line for the "set" modifier in ACLs was not showing the
3597 name of the variable that was being set.
3599 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3601 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3602 lookup on its own. Something else was needed to trigger the lookup. For
3603 example, a match in host_lookup or the need for a host name in a host list.
3604 Now, if $sender_host_name is referenced and the host name has not yet been
3605 looked up, a lookup is performed. If the lookup fails, the variable remains
3606 empty, and $host_lookup_failed is set to "1".
3608 19. Added "eqi" as a case-independent comparison operator.
3610 20. The saslauthd authentication condition could segfault if neither service
3611 nor realm was specified.
3613 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3614 error message that was logged was misleading, and incoming SMTP
3615 connections were dropped. The message is now more accurate, and temporary
3616 errors are given to SMTP connections.
3618 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3619 (except RSET) with a 421 error, until QUIT is received. However, it was
3620 failing to send a response to QUIT.
3622 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3623 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3624 ACL is now run after the helo_try_verify_hosts code.
3626 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3627 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3628 case-independent, so other case variants are also recognized). Apparently
3629 some systems use these upper case variants.
3631 25. If more than two messages were waiting for the same host, and a transport
3632 filter was specified for the transport, Exim sent two messages over the
3633 same TCP/IP connection, and then failed with "socket operation on non-
3634 socket" when it tried to send the third.
3636 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3638 27. The extern definition of crypt16() in expand.c was not being excluded when
3639 the OS had its own crypt16() function.
3641 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3642 as a preferred synonym for return_size_limit, both as an option and as an
3645 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3647 30. Changed the default configuration ACL to relax the local part checking rule
3648 for addresses that are not in any local domains. For these addresses,
3649 slashes and pipe symbols are allowed within local parts, but the sequence
3650 /../ is explicitly forbidden.
3652 31. SPA server authentication was not clearing the challenge buffer before
3655 32. log_message in a "warn" ACL statement was writing to the reject log as
3656 well as to the main log, which contradicts the documentation and doesn't
3657 seem right (because no rejection is happening). So I have stopped it.
3659 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3660 However, I am unable to do any testing of this.
3662 34. Fixed an infelicity in the appendfile transport. When checking directories
3663 for a mailbox, to see if any needed to be created, it was accidentally
3664 using path names with one or more superfluous leading slashes; tracing
3665 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3667 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3668 contents are added to the log line that is written for every discarded
3669 recipient. (Previously a log_message setting was ignored.)
3671 36. The ${quote: operator now quotes the string if it is empty.
3673 37. The install script runs exim in order to find its version number. If for
3674 some reason other than non-existence or emptiness, which it checks, it
3675 could not run './exim', it was installing it with an empty version number,
3676 i.e. as "exim-". This error state is now caught, and the installation is
3679 38. An argument was missing from the function that creates an error message
3680 when Exim fails to connect to the socket for saslauthd authentication.
3681 This could cause Exim to crash, or give a corrupted message.
3683 39. Added isip, isip4, and isip6 to ${if conditions.
3685 40. The ACL variables $acl_xx are now saved with the message, and can be
3686 accessed later in routers, transports, and filters.
3688 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3689 strings in the file are not string-expanded.
3691 42. If a MAIL command specified a SIZE value that was too large to fit into an
3692 int variable, the check against message_size_limit failed. Such values are
3693 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3694 day this will have to be increased, but I don't think I want to be around
3695 when emails are that large.
3702 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3703 iconv() is not standard in FreeBSD.
3705 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3706 IPv6 enabled. The observed symptom was a segmentation fault on return from
3707 the function os_common_find_running_interfaces() in src/os.c.
3709 3. In the check_special_case() function in daemon.c I had used "errno" as an
3710 argument name, which causes warnings on some systems. This was basically a
3711 typo, since it was named "eno" in the comments!
3713 4. The code that waits for the clock to tick (at a resolution of some fraction
3714 of a second) so as to ensure message-id uniqueness was always waiting for
3715 at least one whole tick, when it could have waited for less. [This is
3716 almost certainly not relevant at current processor speeds, where it is
3717 unlikely to ever wait at all. But we try to future-proof.]
3719 5. The function that sleeps for a time interval that includes fractions of a
3720 second contained a race. It did not block SIGALRM between setting the
3721 timer, and suspending (a couple of lines later). If the interval was short
3722 and the sigsuspend() was delayed until after it had expired, the suspension
3723 never ended. On busy systems this could lead to processes getting stuck for
3726 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3727 process, before it forks any delivery processes. The open lookup caching
3728 mechanism meant that the open file or database connection was passed into
3729 the delivery process. The problem was that delivery processes always tidy
3730 up cached lookup data. This could cause a problem for the next delivery
3731 process started by the queue runner, because the external queue runner
3732 process does not know about the closure. So the next delivery process
3733 still has data in the lookup cache. In the case of a file lookup, there was
3734 no problem because closing a file descriptor in a subprocess doesn't affect
3735 the parent. However, if the lookup was caching a connection to a database,
3736 the connection was closed, and the second delivery process was likely to
3737 see errors such as "PGSQL: query failed: server closed the connection
3738 unexpectedly". The problem has been fixed by closing all cached lookups
3739 in a queue runner before running a delivery process.
3741 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3742 seem to have the "const" qualifier which it has on other OS. I've
3745 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3746 *of the same type* (client or server) with the same public name that an
3747 error should be diagnosed.
3749 9. When Exim looked up a host name for an IP address, but failed to find the
3750 original IP address when looking up the host name (a safety check), it
3751 output the message "<ip address> does not match any IP for NULL", which was
3752 confusing, to say the least. The bug was that the host name should have
3753 appeared instead of "NULL".
3755 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3756 user that is built into the binary, and the -C or -D options is used, root
3757 privilege is dropped before the configuration file is read. In addition,
3758 logging is switched to stderr instead of the normal log files. If the
3759 configuration then re-defines the Exim user, the unprivileged environment
3760 is probably not what is expected, so Exim logs a panic warning message (but
3763 However, if deliver_drop_privilege is set, the unprivileged state may well
3764 be exactly what is intended, so the warning has been cut out in that case,
3765 and Exim is allowed to try to write to its normal log files.
3771 1. smtp_return_error_details was not giving details for temporary sender
3772 or receiver verification errors.
3774 2. Diagnose a configuration error if two authenticators have the same public
3777 3. Exim used not to create the message log file for a message until the first
3778 delivery attempt. This could be confusing when incoming messages were held
3779 for policy or load reasons. The message log file is now created at the time
3780 the message is received, and an initial "Received" line is written to it.
3782 4. The automatically generated man page for command line options had a minor
3783 bug that caused no ill effects; however, a more serious problem was that
3784 the procedure for building the man page automatically didn't always
3785 operate. Consequently, release 4.20 contains an out-of-date version. This
3786 shouldn't happen again.
3788 5. When building Exim with embedded Perl support, the script that builds the
3789 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3790 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3792 6. The freeze_tell option was not being used for messages that were frozen on
3793 arrival, either by an ACL or by local_scan().
3795 7. Added the smtp_incomplete_transaction log selector.
3797 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3798 was accepting AUTH without a new EHLO.
3800 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3801 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3804 10. Logging of TCP/IP connections (when configured) now happens in the main
3805 daemon process instead of the child process, so that the TCP/IP connection
3806 count is more accurate (but it can never be perfect).
3808 11. The use of "drop" in a nested ACL was not being handled correctly in the
3809 outer ACL. Now, if condition failure induced by the nested "drop" causes
3810 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3811 or "require"), the connection is dropped.
3813 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3814 that yield "discard" can now be used with an "accept" or a "discard" verb,
3815 but an error is generated for any others (because I can't see a useful way
3816 to define what should happen).
3818 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3819 are now processed in the same way as lines in the Exim configuration file.
3820 In particular, continuation lines are supported.
3822 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3824 15. Added -ti meaning -t -i.
3826 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3827 domains, and warn by logging if others are found.
3829 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3830 returned from GIFCONF when Exim is trying to find the list of interfaces on
3831 a host. The code in os.c has been modified to copy each ifreq to an aligned
3832 structure in all cases.
3834 Also, in some cases, the returned ifreq's were being copied to a 'struct
3835 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3836 means the last couple of bytes of an IPv6 address could be chopped if the
3837 ifreq contained only a normal sockaddr (14 bytes storage).
3839 18. Named domain lists were not supported in the hosts_treat_as_local option.
3840 An entry such as +xxxx was not recognized, and was treated as a literal
3843 19. Ensure that header lines added by a DATA ACL are included in the reject log
3844 if the ACL subsequently rejects the message.
3846 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3847 MD5 (which is deprecated).
3849 21. When testing a filter file using -bf, Exim was writing a message when it
3850 took the sender from a "From " line in the message, but it was not doing so
3851 when it took $return_path from a Return-Path: header line. It now does.
3853 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3854 with a valid header line field name (a series of printing characters
3855 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3857 23. Changed "disc" in the source to "disk" to conform to the documentation and
3858 the book and for uniformity.
3860 24. Ignore Sendmail's -Ooption=value command line item.
3862 25. When execve() failed while trying to run a command in a pipe transport,
3863 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3864 could be confused with a return value of 69 from the command itself. This
3865 has been changed to 127, the value the shell returns if it is asked to run
3866 a non-existent command. The wording for the related log line suggests a
3867 non-existent command as the problem.
3869 26. If received_header_text expands to an empty string, do not add a Received:
3870 header line to the message. (Well, it adds a token one on the spool, but
3871 marks it "old" so that it doesn't get used or transmitted.)
3873 27. Installed eximstats 1.28 (addition of -nt option).
3875 28. There was no check for failure on the call to getsockname() in the daemon
3876 code. This can fail if there is a shortage of resources on the system, with
3877 ENOMEM, for example. A temporary error is now given on failure.
3879 29. Contrary to the C standard, it seems that in some environments, the
3880 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3881 program. Exim now does this explicitly; it affects the formatting of
3882 timestamps using strftime().
3884 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3885 complaints. I've now initialized all the variables, to avoid this.
3887 32. Header lines added by a system filter were not being "seen" during
3888 transport-time rewrites.
3890 33. The info_callback() function passed to OpenSSL is set up with type void
3891 (*)(SSL *, int, int), as described somewhere. However, when calling the
3892 function (actually a macro) that sets it up, the type void(*)() is
3893 expected. I've put in a cast to prevent warnings from picky compilers.
3895 34. If a DNS black list lookup found a CNAME record, but there were no A
3896 records associated with the domain it pointed at, Exim crashed.
3898 35. If a DNS black list lookup returned more than one A record, Exim ignored
3899 all but the first. It now scans all returned addresses if a particular IP
3900 value is being sought. In this situation, the contents of the
3901 $dnslist_value variable are a list of all the addresses, separated by a
3904 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3905 to accept a host name and all its aliases if the forward lookup for any of
3906 them yielded the IP address of the incoming connection. Now it accepts only
3907 those names whose forward lookup yields the correct IP address. Any other
3908 names are discarded. This closes a loophole whereby a rogue DNS
3909 administrator could create reverse DNS records to break through a
3910 wildcarded host restriction in an ACL.
3912 37. If a user filter or a system filter that ran in a subprocess used any of
3913 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3914 the wrong values were passed to the pipe command ($thisaddress had the
3915 value of $0, $0 had the value of $1, etc). This bug was introduced by
3916 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3918 38. Improved the line breaking for long SMTP error messages from ACLs.
3919 Previously, if there was no break point between 40 and 75 characters, Exim
3920 left the rest of the message alone. Two changes have been made: (a) I've
3921 reduced the minimum length to 35 characters; (b) if it can't find a break
3922 point between 35 and 75 characters, it looks ahead and uses the first one
3923 that it finds. This may give the occasional overlong line, but at least the
3924 remaining text gets split now.
3926 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3927 file descriptors might be low, and that setting 1000 would always raise it.
3928 It turns out that in some environments, the limit is already over 1000 and
3929 that lowering it causes trouble. So now Exim takes care not to decrease it.
3931 40. When delivering a message, the value of $return_path is set to $sender_
3932 address at the start of routing (routers may change the value). By an
3933 oversight, this default was not being set up when an address was tested by
3934 -bt or -bv, which affected the outcome if any router or filter referred to
3937 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3938 log only when "message" or "log_message" are set. However, if one of the
3939 conditions was an address verification, or a call to a nested ACL, the
3940 messages generated by the underlying test were being passed through. This
3941 no longer happens. The underlying message is available in $acl_verify_
3942 message for both "message" and "log_message" expansions, so it can be
3943 passed through if needed.
3945 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3946 new expansion $bh_ to give the encoded byte string without charset
3947 translation. Translation happens only if iconv() is available; HAVE_ICONV
3948 indicates this at build time. HEADERS_CHARSET gives the charset to
3949 translate to; headers_charset can change it in the configuration, and
3950 "headers charset" can change it in an individual filter file.
3952 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3953 that creates RFC 2047 encoded "words" labels them as that charset instead
3954 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3955 expansion operator; (ii) when Exim creates a From: line for a local
3956 message; (iii) when a header line is rewritten to include a "phrase" part.
3958 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3959 buggy, causing it to skip the first lines of messages whose message ID
3960 ended in 'D'. This would not have bitten before Exim release 4.14, because
3961 message IDs were unlikely to end in 'D' before then. The effect was to have
3962 incorrect size information for certain domains.
3964 45. #include "config.h" was missing at the start of the crypt16.c module. This
3965 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3968 46. If there was a timeout during a "random" callout check, Exim treated it as
3969 a failure of the random address, and carried on sending RSET and the real
3970 address. If the delay was just some slowness somewhere, the response to the
3971 original RCPT would be taken as a response to RSET and so on, causing
3972 mayhem of various kinds.
3974 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3975 when I implemented it. It didn't allow for the fact that some option values
3976 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3977 the right test for positive values.
3979 48. Domain names in DNS records are case-independent. Exim always looks them up
3980 in lower case. Some resolvers return domain names in exactly the case they
3981 appear in the zone file, that is, they may contain uppercase letters. Not
3982 all resolvers do this - some return always lower case. Exim was treating a
3983 change of case by a resolver as a change of domain, similar to a widening
3984 of a domain abbreviation. This triggered its re-routing code and so it was
3985 trying to route what was effectively the same domain again. This normally
3986 caused routing to fail (because the router wouldn't handle the domain
3987 twice). Now Exim checks for this case specially, and just changes the
3988 casing of the domain that it ultimately uses when it transmits the message
3991 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3994 50. If a filter generated a file delivery with a non-absolute name (possible if
3995 no home directory exists for the router), the forbid_file option was not
3998 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3999 the existing equality matching.
4001 52. Exim was using ints instead of ino_t variables in some places where it was
4002 dealing with inode numbers.
4004 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4005 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4006 called TMPDIR, and if it finds it is different, it changes its value.
4008 54. The smtp_printf() function is now made available to local_scan() so
4009 additional output lines can be written before returning. There is also an
4010 smtp_fflush() function to enable the detection of a dropped connection.
4011 The variables smtp_input and smtp_batched_input are exported to
4014 55. Changed the default runtime configuration: the message "Unknown user"
4015 has been removed from the ACL, and instead placed on the localuser router,
4016 using the cannot_route_message feature. This means that any verification
4017 failures that generate their own messages won't get overridden. Similarly,
4018 the "Unrouteable address" message that was in the ACL for unverifiable
4019 relay addresses has also been removed.
4021 56. Added hosts_avoid_esmtp to the smtp transport.
4023 57. The exicyclog script was not checking for the esoteric option
4024 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4025 will work only if exicyclog is run under the appropriate euid.
4027 58. Following a discussion on the list, the rules by which Exim recognises line
4028 endings on incoming messages have been changed. The -dropcr and drop_cr
4029 options are now no-ops, retained only for backwards compatibility. The
4030 following line terminators are recognized: LF CRLF CR. However, special
4031 processing applies to CR:
4033 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4034 nor a local message in the state where . is a terminator.
4036 (ii) If a bare CR is encountered in a header line, an extra space is added
4037 after the line terminator so as not to end the header. The reasoning
4038 behind this is that bare CRs in header lines are most likely either
4039 to be mistakes, or people trying to play silly games.
4041 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4042 was being incorrectly given as 18 bytes larger than it should have been.
4043 This is a VOB (very old bug).
4045 60. This may never have affected anything current, but just in case it has:
4046 When the local host is found other than at the start of a list of hosts,
4047 the local host, those with the same MX, and any that follow, are discarded.
4048 When the list in question was part of a longer list of hosts, the following
4049 hosts (not currently being processed) were also being discarded. This no
4050 longer happens. I'm not sure if this situation could ever has previously
4053 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4056 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4057 header. This is something that is recommended in a new Internet Draft, and
4058 is something that is documented as being done by Sendmail. There are two
4059 possible values. For messages generated by the autoreply transport, Exim
4062 Auto-Submitted: auto-replied
4064 whereas for all other generated messages (e.g. bounces) it adds
4066 Auto-Submitted: auto-generated
4068 63. The "personal" condition in filters now includes a test for the
4069 Auto-Submitted: header. If it contains the string "auto-" the message it
4070 not considered personal.
4072 64. Added rcpt_include_affixes as a generic transport option.
4074 65. Added queue_only_override (default true).
4076 66. Added the syslog_duplication option.
4078 67. If what should have been the first header line of a message consisted of
4079 a space followed by a colon, Exim was mis-interpreting it as a header line.
4080 It isn't of course - it is syntactically invalid and should therefore be
4081 treated as the start of the message body. The misbehaviour could have
4082 caused a number of strange effects, including loss of data in subsequent
4083 header lines, and spool format errors.
4085 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4086 client host had authenticated. This control can now be exercised by an ACL
4087 for more flexibility.
4089 69. By default, callouts do not happen when testing with -bh. There is now a
4090 variant, -bhc, which does actually run the callout code, including
4091 consulting and updating the callout cache.
4093 70. Added support for saslauthd authentication, courtesy of Alexander
4096 71. If statvfs() failed on the spool or log directories while checking their
4097 size for availability, Exim confusingly gave the error "space shortage".
4098 Furthermore, in debugging mode it crashed with a floating point exception.
4099 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4100 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4101 serious problem, Exim now writes to the main and panic logs when this
4102 happens, with details of the failure. It then refuses to accept the
4103 incoming message, giving the message "spool directory problem" or "log
4104 directory problem" with a 421 code for SMTP messages.
4106 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4107 0, 1, and 2 exist, because some OS complain for execs without them (see
4108 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4109 descriptors. However, the code omitted to check that the open succeeded,
4110 causing mysterious errors if for some reason the permissions on /dev/null
4111 got screwed. Now Exim writes a message to the main and panic logs, and
4112 bombs out if it can't open /dev/null.
4114 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4115 interact so that it is all more flexible. It is supposed to remain
4116 backwards compatible. Also added extra_local_interfaces.
4118 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4119 to bomb out with an assertion failure - to the client this appears as a
4120 connection drop. This problem occurs in the part of the code that was taken
4121 from the Samba project. Fortunately, the assertion is in a very simple
4122 function, so I have fixed this by reproducing the function inline in the
4123 one place where it is called, and arranging for authentication to fail
4124 instead of killing the process with assert().
4126 75. The SPA client code was not working when the server requested OEM rather
4127 than Unicode encoding.
4129 76. Added code to make require_files with a specific uid setting more usable in
4130 the case where statting the file as root fails - usually a non-root-mounted
4131 NFS file system. When this happens and the failure is EACCES, Exim now
4132 forks a subprocess and does the per-uid checking as the relevant uid.
4134 77. Added process_log_path.
4136 78. If log_file_path was not explicitly set, a setting of check_log_space or
4137 check_log_inodes was ignored.
4139 79. If a space check for the spool or log partitions fails, the incident is now
4140 logged. Of course, in the latter case the data may get lost...
4142 80. Added the %p formatting code to string_format() so that it can be used to
4143 print addresses in debug_print(). Adjusted all the address printing in the
4144 debugging in store.c to use %p rather than %d.
4146 81. There was a concern that a line of code in smtp_in.c could overflow a
4147 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4148 initially expressed, the concern was not well-founded, because trailing
4149 spaces are removed early. However, if the trailing spaces were followed by
4150 a NULL, they did not get removed, so the overflow was possible. Two fixes
4153 (a) I re-wrote the offending code in a cleaner fashion.
4154 (b) If an incoming SMTP command contains a NULL character, it is rejected
4157 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4158 runs initgroups(), so that if the Exim user is in any additional groups,
4159 they will be used during message reception.
4165 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob