1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.388 2006/09/05 14:14:32 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.
38 SC/01 Use a glob alias rather than an array ref in eximstats generated
39 parser. This improves both readability and performance.
41 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
42 Don't display local sender or destination tables in eximstats unless
43 there is data to show.
44 Added average volumes into the eximstats top table text output.
46 SC/03 Collect data on the number of addresses (recipients) as well
47 as the number of messages in eximstats.
49 TF/01 Correct an error in the documentation for the redirect router. Exim
50 does (usually) call initgroups() when daemonizing.
52 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
53 with consistent privilege compared to when running as a daemon.
55 TF/03 Note in the spec that $authenticated_id is not set for local
56 submissions from trusted users.
58 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
59 Thanks to Dean Brooks <dean@iglou.com> for the patch.
61 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
62 by adding some example configuration directives to the default
63 configuration file. A little bit of work is required to uncomment the
64 directives and define how usernames and passwords are checked, but
65 there is now a framework to start from.
67 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
68 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
69 without this. I don't know how relevant this is to other LDAP libraries.
71 PH/02 Add the verb name to the "unknown ACL verb" error.
73 PH/03 Magnus Holmgren's patch for filter_prepend_home.
75 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
77 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
78 directory not expanded when it should be if an expanded home directory
79 was set for the address (which is overridden by the transport).
81 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
84 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
85 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
86 because it is too late at that time, and has no effect.
88 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
89 security issue with \' (bugzilla #107). I could not use the
90 PQescapeStringConn() function, because it needs a PGconn value as one of
93 PH/08 When testing addresses using -bt, indicate those final addresses that
94 are duplicates that would not cause an additional delivery. At least one
95 person was confused, thinking that -bt output corresponded to deliveries.
96 (Suppressing duplicates isn't a good idea as you lose the information
97 about possibly different redirections that led to the duplicates.)
99 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
100 systems where poll() doesn't work, in particular OS X.
102 PH/10 Added more information to debugging output for retry time not reached.
104 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
105 operations in malware.c.
107 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
110 PH/13 If write_rejectlog was set false when logging was sent to syslog with
111 syslog_duplication set false, log lines that would normally be written
112 both the the main log and to the reject log were not written to syslog at
115 PH/14 In the default configuration, change the use of "message" in ACL warn
116 statements to "add_header".
118 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
119 not followed by a command (e.g. "seen endif").
121 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
122 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
125 PH/17 Added extra conditions to the default value of delay_warning_condition
129 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
130 { match{$h_precedence:}{(?i)bulk|list|junk} } \
131 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
134 The Auto-Submitted: and various List- headers are standardised, whereas I
135 don't think Precedence: ever was.
137 PH/18 Refactored debugging code in route_finduser() to show more information,
138 in particular, the error code if getpwnam() issues one.
140 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
141 This is apparently needed in addition to the PH/07 change above to avoid
142 any possible encoding problems.
144 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
145 but not after initializing Perl.
147 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
148 output them only if debugging. By default they are written stderr,
149 apparently, which is not desirable.
151 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
154 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
157 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
159 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
160 authenticated or an ident call has been made. Suppress the default
161 values for $authenticated_id and $authenticated_sender (but permit -oMai
162 and -oMas) when testing with -bh.
164 PH/24 Re-jigged the order of the tests in the default configuration so that the
165 tests for valid domains and recipients precede the DNS black list and CSA
166 tests, on the grounds that those ones are more expensive.
168 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
169 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
170 This bug exists in every version of Exim that I still have, right back to
173 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
174 However, an attempt to turn on case-sensitivity in a regex key by
175 including (?-i) didn't work because the subject string was already
176 lowercased, and the effects were non-intuitive. It turns out that a
177 one-line patch can be used to allow (?-i) to work as expected.
183 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
184 other effects) broke the use of negated acl sub-conditions.
186 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
189 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
190 "Deny" causes Exim to reject the incoming connection with a 554 error.
191 Unfortunately, if there is a major crisis, such as a disk failure,
192 tcp-wrappers gives "deny", whereas what one would like would be some
193 kind of temporary error. A kludge has been added to help with this.
194 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
195 554 error is used if errno is still zero or contains ENOENT (which occurs
196 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
199 PH/03 Add -lutil to the default FreeBSD LIBS setting.
201 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
202 errors. Otherwise a message that provokes a temporary error (when other
203 messages do not) can cause a whole host to time out.
205 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
206 addresses were routed directly to files or pipes from a redirect router.
207 File deliveries just didn't batch; pipe deliveries might have suffered
210 PH/06 A failure to get a lock for a hints database would erroneously always say
211 "Failed to get write lock", even when it was really a read lock.
213 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
214 of 0600. This has been changed to use the value of the lockfile_mode
215 option (which defaults to 0600).
217 PH/08 Applied small patch from the Sieve maintainer.
219 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
220 folder from quota calculations, a direct delivery into this folder messed
221 up the contents of the maildirsize file. This was because the regex was
222 used only to exclude .Trash (or whatever) when the size of the mailbox
223 was calculated. There was no check that a delivery was happening into an
224 excluded directory. This bug has been fixed by ignoring all quota
225 processing for deliveries into excluded directories.
227 PH/10 Added the maildirfolder_create_regex option to appendfile.
233 PH/01 The code for finding all the local interface addresses on a FreeBSD
234 system running IPv6 was broken. This may well have applied to all BSD
235 systems, as well as to others that have similar system calls. The broken
236 code found IPv4 interfaces correctly, but gave incorrect values for the
237 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
238 that it would not match correctly against @[] and not recognize the IPv6
241 PH/02 The ipliteral router was not recognizing addresses of the form user@
242 [ipv6:....] because it didn't know about the "ipv6:" prefix.
244 PH/03 Added disable_ipv6.
246 PH/04 Changed $reply_address to use the raw form of the headers instead of the
247 decoded form, because it is most often used to construct To: headers
248 lines in autoreplies, and the decoded form may well be syntactically
249 invalid. However, $reply_address has leading white space removed, and all
250 newlines turned into spaces so that the autoreply transport does not
253 PH/05 If group was specified without a user on a router, and no group or user
254 was specified on a transport, the group from the router was ignored.
256 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
257 for visible compile-time settings that can be used to change these
258 numbers, for those that want even more. Backwards compatibility with old
259 spool files has been maintained. However, going back to a previous Exim
260 release will lost any variables that are in spool files.
262 PH/07 Two small changes when running in the test harness: increase delay when
263 passing a TCP/IP connection to a new process, in case the original
264 process has to generate a bounce, and remove special handling of
265 127.0.0.2 (sic), which is no longer necessary.
267 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
268 be the same on different OS.
270 PH/09 Moved a debug statement in filter processing to avoid a race problem when
273 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
274 whether --show-vars was specified or not
276 JJ/02 exipick: Added support for new ACL variable spool format introduced
279 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
280 syntactically invalid From: or Reply-to: line, and a filter used this to
281 generate an autoreply, and therefore failed to obtain an address for the
282 autoreply, Exim could try to deliver to a non-existent relative file
283 name, causing unrelated and misleading errors. What now happens is that
284 it logs this as a hard delivery error, but does not attempt to create a
287 PH/11 The exinext utility has a -C option for testing purposes, but although
288 the given file was scanned by exinext itself; it wasn't being passed on
289 when Exim was called.
291 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
292 an end-of-file indication when reading a command response.
294 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
295 compiled. In many other places in Exim, IPv6 addresses are always
296 recognized, so I have changed this. It also means that IPv4 domain
297 literals of the form [IPV4:n.n.n.n] are now always recognized.
299 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
300 used if the router is not running as root, for example, when verifying at
301 ACL time, or when using -bh. The debugging output from this situation was
302 non-existent - all you got was a failure to exec. I have made two
305 (a) Failures to set uid/gid, the current directory, or a process leader
306 in a subprocess such as that created by queryprogram now generate
307 suitable debugging ouput when -d is set.
309 (b) The queryprogram router detects when it is not running as root,
310 outputs suitable debugging information if -d is set, and then runs
311 the subprocess without attempting to change uid/gid.
313 PH/15 Minor change to Makefile for building test_host (undocumented testing
316 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
317 additional section of a DNS packet that returns MX or SRV records.
318 Instead, it always explicitly searches for A/AAAA records. This avoids
319 major problems that occur when a DNS server includes only records of one
320 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
321 fixed another bug: if SRV records were looked up and the corresponding
322 address records were *not* found in the additional section, the port
323 values from the SRV records were lost.
325 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
326 using the correct key (the original address) when searching the retry
327 rules in order to find which one to use for generating the retry hint.
329 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
330 adding the default one. Similarly, if it contains a Reply-To: header, the
331 errors_reply_to option, if set, is not used.
333 PH/19 When calculating a retry time, Exim used to measure the "time since
334 failure" by looking at the "first failed" field in the retry record. Now
335 it does not use this if it is later than than the arrival time of the
336 message. Instead it uses the arrival time. This makes for better
337 behaviour in cases where some deliveries succeed, thus re-setting the
338 "first failed" field. An example is a quota failure for a huge message
339 when small messages continue to be delivered. Without this change, the
340 "time since failure" will always be short, possible causing more frequent
341 delivery attempts for the huge message than are intended.
342 [Note: This change was subsequently modified - see PH/04 for 4.62.]
344 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
345 $1, $2, $3) because the numerical variables can be reset during some
346 expansion items (e.g. "match"), thereby losing the authentication data.
348 PH/21 Make -bV show the size of off_t variables so that the test suite can
349 decide whether to run tests for quotas > 2G.
351 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
352 mailbox_size, and mailbox_filecount in the appendfile transport. If a
353 filecount value is greater than 2G or if a quota value is greater than 2G
354 on a system where the size of off_t is not greater than 4, a panic error
357 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
358 never match. The debug and -bh output now contains an explicit error
359 message indicating a malformed IPv4 address or mask.
361 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
362 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
365 PH/25 Do not write to syslog when running in the test harness. The only
366 occasion when this arises is a failure to open the main or panic logs
367 (for which there is an explicit test).
369 PH/26 Added the /no_tell option to "control=freeze".
371 PH/27 If a host name lookup failed very early in a connection, for example, if
372 the IP address matched host_lookup and the reverse lookup yielded a name
373 that did not have a forward lookup, an error message of the form "no IP
374 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
375 could be logged. Now it outputs the IP address instead of "NULL".
377 PH/28 An enabling patch from MH: add new function child_open_exim2() which
378 allows the sender and the authenticated sender to be set when
379 submitting a message from within Exim. Since child_open_exim() is
380 documented for local_scan(), the new function should be too.
382 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
383 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
384 results in an empty string is now treated as unset.
386 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
388 PH/31 Added sender_verify_fail logging option.
390 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
391 needed by RFC 822 but not by RFC 2822 was commented out. I have now
392 tidied the source and removed it altogether.
394 PH/33 When a queue run was abandoned because the load average was too high, a
395 log line was always written; now it is written only if the queue_run log
396 selector is set. In addition, the log line for abandonment now contains
397 information about the queue run such as the pid. This is always present
398 in "start" and "stop" lines but was omitted from the "abandon" line.
400 PH/34 Omit spaces between a header name and the colon in the error message that
401 is given when verify = headers_syntax fails (if there are lots of them,
402 the message gets confusing).
404 PH/35 Change the default for dns_check_names_pattern to allow slashes within
405 names, as there are now some PTR records that contain slashes. This check
406 is only to protect against broken name servers that fall over on strange
407 characters, so the fact that it applies to all lookups doesn't matter.
409 PH/36 Now that the new test suite is complete, we can remove some of the
410 special code in Exim that was needed for the old test suite. For example,
411 sorting DNS records because real resolvers return them in an arbitrary
412 order. The new test suite's fake resolver always returns records in the
415 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
416 bounces) except when queue_only is set, to avoid logging races between
417 the different processes.
419 PH/38 Panic-die if .include specifies a non-absolute path.
421 PH/39 A tweak to the "H" retry rule from its user.
423 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
424 a label. They prevented compilation on older perls.
426 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
427 a warning to be raised on newish perls.
429 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
430 on queue. Changes to match documented behaviour of showing count of
431 messages matching specified criteria.
433 PH/40 Changed the default ident timeout from 30s to 5s.
435 PH/41 Added support for the use of login_cap features, on those BSD systems
436 that have them, for controlling the resources used by pipe deliveries.
438 PH/42 The content-scanning code uses fopen() to create files in which to put
439 message data. Previously it was not paying any attention to the mode of
440 the files. Exim runs with umask(0) because the rest of the code creates
441 files with open(), and sets the required mode explicitly. Thus, these
442 files were ending up world-writeable. This was not a big issue, because,
443 being within the spool directory, they were not world-accessible. I have
444 created a function called modefopen, which takes an additional mode
445 argument. It sets umask(777), creates the file, chmods it to the required
446 mode, then resets the umask. All the relevant calls to fopen() in the
447 content scanning code have been changed to use this function.
449 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
450 to 24 hours. This avoids potential overflow problems when processing G
451 and H retry rules. I suspect nobody ever tinkers with this value.
453 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
455 PH/45 When the plaintext authenticator is running as a client, the server's
456 challenges are checked to ensure they are valid base64 strings. By
457 default, the authentication attempt is cancelled if an invalid string is
458 received. Setting client_ignore_invalid_base64 true ignores these errors.
459 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
460 they are received. Thus, the responses can be made to depend on the
461 challenges. If an invalid string is ignored, an empty string is placed in
464 PH/46 Messages that are created by the autoreply transport now contains a
465 References: header, in accordance with RFCs 2822 and 3834.
467 PH/47 Added authenticated_sender_force to the smtp transport.
469 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
471 PH/49 Installed latest patch from the Sieve maintainer.
473 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
474 was also set, the appendfile transport was unnecessarily scanning a
475 directory of message files (e.g. for maildir delivery) to find the count
476 of files (along with the size), even though it did not need this
477 information. It now does the scan only if it needs to find either the
478 size of the count of files.
480 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
482 PH/52 Two bugs concerned with error handling when the smtp transport is
485 (i) Exim was not creating retry information for temporary errors given
486 for individual recipients after the DATA command when the smtp transport
487 was used in LMTP mode. This meant that they could be retried too
488 frequently, and not timed out correctly.
490 (ii) Exim was setting the flag that allows error details to be returned
491 for LMTP errors on RCPT commands, but not for LMTP errors for individual
492 recipients that were returned after the DATA command.
494 PH/53 This is related to PH/52, but is more general: for any failing address,
495 when detailed error information was permitted to be returned to the
496 sender, but the error was temporary, then after the final timeout, only
497 "retry timeout exceeded" was returned. Now it returns the full error as
498 well as "retry timeout exceeded".
500 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
501 do this, and (what is worse) MTAs that accept it.
503 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
504 will now be deprecated.
506 PH/56 New os.c-cygwin from the Cygwin maintainer.
508 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
509 formats (previously only available in exim formats via -bpr, -bpru,
510 and -bpra. Now also available in native and exiqgrep formats)
512 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
513 with very large, slow to parse queues
515 JJ/08 exipick: added ! as generic prefix to negate any criteria format
517 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
519 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
520 responses to authentication challenges, though it was showing the
521 challenges; (ii) I've removed the CR characters from the debug output for
524 PH/58 Allow for the insertion of a newline as well as a space when a string
525 is turned into more than one encoded-word during RFC 2047 encoding. The
526 Sieve code now uses this.
528 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
529 data_4xx, lost_connection, tls_required.
531 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
532 message was being sent as an SMTP response.
534 PH/61 Add -l and -k options to exicyclog.
536 PH/62 When verifying, if an address was redirected to one new address, so that
537 verification continued, and the new address failed or deferred after
538 having set something in $address_data, the value of $address_data was not
539 passed back to the ACL. This was different to the case when no
540 redirection occurred. The value is now passed back in both cases.
542 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
543 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
544 use login_cap.h, so on its own it isn't the distinguishing feature. The
545 new name refers directly to the setclassresources() function.
547 PH/65 Added configuration files for NetBSD3.
549 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
551 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
552 is preferred over IPv4.
554 PH/68 The bounce_return_message and bounce_return_body options were not being
555 honoured for bounces generated during the reception of non-SMTP messages.
556 In particular, this applied to messages rejected by the ACL. This bug has
557 been fixed. However, if bounce_return_message is true and bounce_return_
558 body is false, the headers that are returned for a non-SMTP message
559 include only those that have been read before the error was detected.
560 (In the case of an ACL rejection, they have all been read.)
562 PH/69 The HTML version of the specification is now built in a directory called
563 spec_html instead of spec.html, because the latter looks like a path with
564 a MIME-type, and this confuses some software.
566 PH/70 Catch two compiler warnings in sieve.c.
568 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
569 function verify_get_ident() calls ip_connect() to connect a socket, but
570 if the "connect()" function timed out, ip_connect() used to close the
571 socket. However, verify_get_ident() also closes the socket later, and in
572 between Exim writes to the log, which may get opened at this point. When
573 the socket was closed in ip_connect(), the log could get the same file
574 descriptor number as the socket. This naturally causes chaos. The fix is
575 not to close the socket in ip_connect(); the socket should be closed by
576 the function that creates it. There was only one place in the code where
577 this was missing, in the iplookup router, which I don't think anybody now
578 uses, but I've fixed it anyway.
580 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
581 well as to direct DNS lookups. Otherwise the handling of names in host
582 lists is inconsistent and therefore confusing.
588 PH/01 Two changes to the default runtime configuration:
590 (1) Move the checks for relay_from_hosts and authenticated clients from
591 after to before the (commented out) DNS black list checks.
593 (2) Add control=submission to the relay_from_hosts and authenticated
594 clients checks, on the grounds that messages accepted by these
595 statements are most likely to be submissions.
597 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
599 (1) Generate an error if the third argument for the ${prvs expansion is
602 (2) Treat a missing third argument of ${prvscheck as if it were an empty
605 (3) Reset the variables that are obtained from the first argument of
606 ${prvscheck and used in the second argument before leaving the code,
607 because their memory is reclaimed, so using them afterwards may do
610 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
611 one (it's much easier than Tom thought :-).
613 (5) Because of (4), we can now allow for the use of $prvscheck_result
614 inside the third argument.
616 PH/03 For some reason, the default setting of PATH when running a command from
617 a pipe transport was just "/usr/bin". I have changed it to
620 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
621 anything to be listed in the output from -bV.
623 PH/05 When a filter generated an autoreply, the entire To: header line was
624 quoted in the delivery log line, like this:
626 => >A.N.Other <ano@some.domain> <original@ddress> ...
628 This has been changed so that it extracts the operative address. There
629 may be more than one such address. If so, they are comma-separated, like
632 => >ano@some.domain,ona@other.domain <original@ddress> ...
634 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
635 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
636 not being looked up in the rDNS to get a host name, Exim was showing the
637 IP address twice in Received: lines, even though the IP addresses were
638 identical. For example:
640 Received: from [1.2.3.4] (helo=[1.2.3.4])
642 However, if the real host name was known, it was omitting the HELO data
643 if it matched the actual IP address. This has been tidied up so that it
644 doesn't show the same IP address twice.
646 PH/07 When both +timestamp and +memory debugging was on, the value given by
647 $tod_xxx expansions could be wrong, because the tod_stamp() function was
648 called by the debug printing, thereby overwriting the timestamp buffer.
649 Debugging no longer uses the tod_stamp() function when +timestamp is set.
651 PH/08 When the original message was included in an autoreply transport, it
652 always said "this is a copy of the message, including all the headers",
653 even if body_only or headers_only was set. It now gives an appropriate
656 PH/09 Applied a patch from the Sieve maintainer which:
658 o fixes some comments
659 o adds the (disabled) notify extension core
660 o adds some debug output for the result of if/elsif tests
661 o points to the current vacation draft in the documentation
662 and documents the missing references header update
666 o fixes a bug in processing the envelope test (when testing
667 multiple envelope elements, the last element determinted the
670 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
671 Electronic Mail") by including:
673 Auto-submitted: auto-generated
675 in the messages that it generates (bounce messages and others, such as
676 warnings). In the case of bounce messages for non-SMTP mesages, there was
677 also a typo: it was using "Auto_submitted" (underscore instead of
678 hyphen). Since every message generated by Exim is necessarily in response
679 to another message, thes have all been changed to:
681 Auto-Submitted: auto-replied
683 in accordance with these statements in the RFC:
685 The auto-replied keyword:
687 - SHOULD be used on messages sent in direct response to another
688 message by an automatic process,
690 - MUST NOT be used on manually-generated messages,
692 - MAY be used on Delivery Status Notifications (DSNs) and Message
693 Disposition Notifications (MDNs),
695 - MUST NOT be used on messages generated by automatic or periodic
696 processes, except for messages which are automatic responses to
699 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
700 to the default Received: header definition.
702 PH/12 Added log selector acl_warn_skipped (default on).
704 PH/13 After a successful wildlsearch lookup, discard the values of numeric
705 variables because (a) they are in the wrong storage pool and (b) even if
706 they were copied, it wouldn't work properly because of the caching.
708 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
709 checking when decoding. Apparently there are clients that generate
710 overlong encoded strings. Why am I not surprised?
712 PH/15 If the first argument of "${if match_address" was not empty, but did not
713 contain an "@" character, Exim crashed. Now it writes a panic log message
714 and treats the condition as false.
716 PH/16 In autoreply, treat an empty string for "once" the same as unset.
718 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
719 extension "envelope-auth". The code is finished and in agreement with
720 other implementations, but there is no documentation so far and in fact,
721 nobody wrote the draft yet. This extension is currently #undef'ed, thus
722 not changing the active code.
724 Print executed "if" and "elsif" statements when debugging is used. This
725 helps a great deal to understand what a filter does.
727 Document more things not specified clearly in RFC3028. I had all this
728 sorted out, when out of a sudden new issues came to my mind. Oops."
730 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
733 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
734 canonical form (as documented). However, after a host name lookup from
735 the IP address, check_host() was doing a simple string comparison with
736 addresses acquired from the DNS when checking that the found name did
737 have the original IP as one of its addresses. Since any found IPv6
738 addresses are likely to be in abbreviated form, the comparison could
739 fail. Luckily, there already exists a function for doing the comparison
740 by converting both addresses to binary, so now that is used instead of
743 PH/20 There was another similar case to PH/19, when a complete host name was
744 given in a host list; looking up its IP address could give an abbreviated
745 form, whereas the current host's name might or might not be abbreviated.
746 The same fix has been applied.
752 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
753 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
756 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
757 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
759 PH/03 Typo: missing ".o" in src/pcre/Makefile.
761 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
762 header line, restrict the check to what is listed in RFCs 2369 and 2929.
763 Also, for "Auto-Submitted", treat anything other than "no" as
764 non-personal, in accordance with RFC 3834. (Previously it treated
765 anything starting "auto-" as non-personal.)
767 TF/01 The control=submission/name=... option had a problem with syntax
768 errors if the name included a slash character. The /name= option
769 now slurps the rest of the string, so it can include any characters
770 but it must come last in the list of options (after /sender_retain
773 PH/05 Some modifications to the interface to the fake nameserver for the new
781 TK/01 Added the "success_on_redirect" address verification option. See
782 NewStuff for rationale and an example.
784 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
786 PH/02 Patch to exigrep to allow it to work on syslog lines.
788 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
789 fread() to skip over the body file's header line, because in Cygwin the
790 header line is locked and is inaccessible.
792 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
793 co-exist for some time) to make it clear that it is the Exim ID that is
794 referenced, not the Message-ID: header line.
796 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
797 string_format() function, because snprintf() does not exist on all
800 PH/06 The use of forbid_filter_existstest now also locks out the use of the
801 ${stat: expansion item.
803 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
804 protocol synchronization error", to keep the pedants happy.
806 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
807 well as for IRIX systems, when gcc is being used. See the host.c source
810 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
812 PH/10 Named domain lists were not working if used in a queue_smtp_domains
815 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
816 transport and to the smtp transport in LMTP mode.
818 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
820 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
821 run a filter in a subprocess. This could lead to confusion in subsequent
822 lookups in the parent process. There should also be a search_tidyup() at
823 the end of the subprocess.
825 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
826 only if the host matched helo_try_verify_hosts, which caused the
827 verification to occur when the EHLO/HELO command was issued. The ACL just
828 tested the remembered result. Now, if a previous verification attempt has
829 not happened, "verify = helo" does it there and then.
831 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
833 TK/03 Fix log output including CR from clamd.
835 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
836 exist provoked a memory error which could cause a segfault.
838 PH/15 Installed PCRE 6.2
840 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
842 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
843 of the problem. Specifically, suggested +O2 rather than +O1 for the
846 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
848 PH/20 If a delivery was routed to a non-standard port by means of an SRV
849 record, the port was not correctly logged when the outgoing_port log
850 selector was set (it logged the transort's default port).
852 PH/21 Added support for host-specific ports to manualroute, queryprogram,
853 fallback_hosts, and "hosts" in the smtp transport.
855 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
856 host errors such as "Connection refused".
858 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
859 authentication with radiusclient 0.4.9:
861 - Error returned from rc_read_config was caught wrongly
862 - Username/password not passed on to radius server due to wrong length.
864 The presumption is that some radiusclient API changes for 4.51/PH/17
865 were not taken care of correctly. The code is still untested by me (my
866 Linux distribution still has 0.3.2 of radiusclient), but it was
867 contributed by a Radius user.
869 PH/24 When doing a callout, the value of $domain wasn't set correctly when
870 expanding the "port" option of the smtp transport.
872 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
873 while reading a MIME header. Thanks to Tom Hughes for a patch.
875 PH/24 Include config.h inside local_scan.h so that configuration settings are
878 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
879 that in an ACL for RCPT (for example), you can examine exactly what was
882 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
883 commands, but it was not correctly comparing the address with the actual
884 client host address. Thus, it would show the EHLO address in Received:
885 header lines when this was not necessary.
887 PH/27 Added the % operator to ${eval:}.
889 PH/28 Exim tries to create and chdir to its spool directory when it starts;
890 it should be ignoring failures (because with -C, for example, it has lost
891 privilege). It wasn't ignoring creation failures other than "already
894 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
897 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
898 because an input line was too long, either on its own, or by virtue of
899 too many continuations, the temporary file was not being removed, and the
900 return code was incorrect.
902 PH/31 Missing "BOOL" in function definition in filtertest.c.
904 PH/32 Applied Sieve patches from the maintainer.
906 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
908 PH/33 Added "verify = not_blind".
910 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
911 Local/Makefile (with some defaults set). These are used in built scripts
912 such as exicyclog, but they have never been used in the exim_install
913 script (though there are many overriding facilities there). I have
914 arranged that the exim_install script now takes note of these two
917 PH/35 Installed configuration files for Dragonfly.
919 PH/36 When a locally submitted message by a trusted user did not contain a
920 From: header, and the sender address was obtained from -f or from an SMTP
921 MAIL command, and the trusted user did not use -F to supply a sender
922 name, $originator_name was incorrectly used when constructing a From:
923 header. Furthermore, $originator_name was used for submission mode
924 messages from external hosts without From: headers in a similar way,
925 which is clearly wrong.
927 PH/37 Added control=suppress_local_fixups.
929 PH/38 When log_selector = +received_sender was set, and the addition of the
930 sender made the log line's construction buffer exactly full, or one byte
931 less than full, an overflow happened when the terminating "\n" was
934 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
935 when the result of a list match is failure because a DNS lookup failed.
937 PH/40 RM_COMMAND is now used in the building process.
939 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
940 the "build-* directories that it finds.
942 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
943 address in a domain literal was a prefix of an interface address.
945 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
946 when verifying a sender address, unless rewrite_headers is false.
948 PH/44 Wrote a long comment about why errors_to addresses are verified as
949 recipients, not senders.
951 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
952 the ratelimit ACL was added.
954 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
956 PH/46 Added extra information about PostgreSQL errors to the error string.
958 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
959 avoiding the need to install special zones in a real server. This is
960 backwards compatible; if it can't find the fake resolver, it drops back.
961 Thus, both old and new test suites can be run.
963 TF/02 Added util/ratelimit.pl
965 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
966 clock is set back in time.
968 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
969 Candler <B.Candler@pobox.com>.
971 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
972 OK for addresses that are the result of redirections.
974 PH/48 A number of further additions for the benefit of the new test suite,
975 including a fake gethostbyname() that interfaces to the fake DNS resolver
978 TF/06 The fix for widen_domains has also been applied to qualify_single and
979 search_parents which are the other dnslookup options that can cause
982 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
985 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
987 TF/07 Exim produced the error message "an SRV record indicated no SMTP
988 service" if it encountered an MX record with an empty target hostname.
989 The message is now "an MX or SRV record indicated no SMTP service".
991 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
992 if the DNS of the sending site is misconfigured. This is quite a
993 common situation. This change restores the behaviour of treating a
994 helo verification defer as a failure.
996 PH/51 If self=fail was set on a router, the bounce message did not include the
997 actual error message.
1003 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1005 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1006 command itself ended in error, the underlying message about the transport
1007 filter timeout was being overwritten with the pipe command error. Now the
1008 underlying error message should be appended to the second error message.
1010 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1012 PH/02 Reduce the amount of output that "make" produces by default. Full output
1013 can still be requested.
1015 PH/03 The warning log line about a condition test deferring for a "warn" verb
1016 was being output only once per connection, rather than after each
1017 occurrence (because it was using the same function as for successful
1018 "warn" verbs). This seems wrong, so I have changed it.
1020 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1021 it should not have, which might have caused a crash in the right
1022 circumstances, but probably never did.
1024 PH/04 Installed a modified version of Tony Finch's patch to make submission
1025 mode fix the return path as well as the Sender: header line, and to
1026 add a /name= option so that you can make the user's friendly name appear
1029 TF/03 Added the control = fakedefer ACL modifier.
1031 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1032 Mark Lowes for thorough testing.
1034 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1036 TK/03 Merged latest SRS patch from Miles Wilton.
1038 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1039 with the definition in sysexits.h (which is #included earlier).
1040 Fortunately, Exim does not actually use EX_OK. The code used to try to
1041 preserve the sysexits.h value, by assumimg that macro definitions were
1042 scanned for macro replacements. I have been disabused of this notion,
1043 so now the code just undefines EX_OK before #including unistd.h.
1045 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1046 in the smtp transport. When a block could not be written in a single
1047 write() function, the timeout was being re-applied to each part-write.
1048 This seems wrong - if the receiver was accepting one byte at a time it
1049 would take for ever. The timeout is now adjusted when this happens. It
1050 doesn't have to be particularly precise.
1052 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1053 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1055 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1056 without a domain if the check to <postmaster@domain> fails.
1058 SC/01 Eximstats: added -xls and the ability to specify output files
1059 (patch written by Frank Heydlauf).
1061 SC/02 Eximstats: use FileHandles for outputing results.
1063 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1065 SC/04 Eximstats: fixed display of large numbers with -nvr option
1067 SC/05 Eximstats: fixed merging of reports with empty tables.
1069 SC/06 Eximstats: added the -include_original_destination flag
1071 SC/07 Eximstats: removed tabs and trailing whitespace.
1073 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1075 TK/06 MBOX spool code: Add real "From " MBOX separator line
1076 so the .eml file is really in mbox format (even though
1077 most programs do not really care). Patch from Alex Miller.
1079 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1080 The latter is generated from $received_to and is only set if the
1081 message has one envelope recipient. SA can use these headers,
1082 obviously out-of-the-box. Patch from Alex Miller.
1084 PH/08 The ${def test on a variable was returning false if the variable's
1085 value was "0", contrary to what the specification has always said!
1086 The result should be true unless the variable is empty.
1088 PH/09 The syntax error of a character other than { following "${if
1089 def:variable_name" (after optional whitespace) was not being diagnosed.
1090 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1091 accidental colon was present, for example, could give incorrect results.
1093 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1094 result is used (not including appendfile, where other changes are about
1097 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1098 This involved changing a lot of size variables from int to off_t. It
1099 should work with maildirs and everything.
1101 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1102 spamd dying while we are connected to it.
1104 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1107 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1108 function for BDB 4.3.
1110 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1112 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1113 the sources for PCRE have been re-arranged and more files are now
1116 PH/15 The code I had for printing potentially long long variables in PH/11
1117 above was not the best (it lost precision). The length of off_t variables
1118 is now inspected at build time, and an appropriate printing format (%ld
1119 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1120 to be "long long int" or "long int". This is needed for the internal
1121 formatting function string_vformat().
1123 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1124 the configuration file to be ":syslog", then the script "guesses" where
1125 the logs files are, rather than using the compiled in default. In our
1126 case the guess is not the same as the compiled default, so the script
1127 suddenly stopped working when I started to use syslog. The patch checks
1128 to see if log_file_path is "". If so, it attempts to read it from exim
1129 with no configuration file to get the compiled in version, before it
1130 falls back to the previous guessing code."
1132 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1133 implementing BATV in an Exim configuration. See NewStuff for the gory
1136 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1137 Makefile that are specific to HP-UX.
1139 PH/18 If the "use_postmaster" option was set for a recipient callout together
1140 with the "random" option, the postmaster address was used as the MAIL
1141 FROM address for the random test, but not for the subsequent recipient
1142 test. It is now used for both.
1144 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1145 patch removes a few documentation additions to RFC 3028, because the
1146 latest draft now contains them. It adds the new en;ascii-case comparator
1147 and a new error check for 8bit text in MIME parts. Comparator and
1148 require names are now matched exactly. I enabled the subaddress
1149 extension, but it is not well tested yet (read: it works for me)."
1151 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1152 rework some of the code of TK/09 above to avoid the hardwired use of
1153 "%lld" and "long long". Replaced the call to snprintf() with a call to
1156 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1157 records point to non-existent hosts", "retry timeout exceeded", and
1158 "retry time not reached for any host after a long failure period".
1160 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1161 experimental DomainKeys support:
1163 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1164 (2) On an error such as an illegally used "control", the wrong name for
1165 the control was given.
1167 These problems did NOT occur unless DomainKeys support was compiled.
1169 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1171 PH/24 Added ${if match_ip condition.
1173 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1174 databases so that it will be absolutely obvious if a crash occurs in the
1175 DB library. This is a regular occurrence (often caused by mis-matched
1178 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1179 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1180 picked up on a user's system that detects such things. There doesn't seem
1181 to be a gcc warning option for this - only an attribute that has to be
1182 put on the function's prototype. It seems that in Fedora Core 4 they have
1183 set this on a number of new functions. No doubt there will be more in due
1186 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1187 specify a transport. However, if an address that was verified by such a
1188 router was the subject of a callout, Exim crashed because it tried to
1189 read the rcpt_include_affixes from the non-existent transport. Now it
1190 just assumes that the setting of that option is false. This bug was
1191 introduced by 4.51/PH/31.
1193 PH/28 Changed -d+all to exclude +memory, because that information is very
1194 rarely of interest, but it makes the output a lot bigger. People tend to
1195 do -d+all out of habit.
1197 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1198 code in os-type was giving problems when libc.so lives in lib64, like on
1201 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1202 aren't the modern standard, and it seems that some systems' include files
1203 don't always have them. Exim was already checking for some of the newer
1204 ones like T_AAAA, and defining it itself. I've added checks for all the
1205 record types that Exim uses.
1207 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1208 not automatically generating a new one, as it is supposed to. This
1209 prevented TLS from working. If the file did exist, but contained invalid
1210 data, a new version was generated, as expected. It was only the case of a
1211 non-existent file that was broken.
1213 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1214 with a change in libdomainkeys > 0.64.
1216 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1217 from DNS. If the selector record carries the flag, it now has
1218 precedence over the domain-wide flag.
1220 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1222 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1223 the use of a port name that isn't defined in /etc/services) occurred, the
1224 message was deferred as in a normal delivery, and thus remained on the
1225 spool, instead of being failed because of the mua_wrapper setting. This
1226 is now fixed, and I tidied up some of the mua_wrapper messages at the
1229 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1230 the messages in a hash of arrays rather than using individual hashes.
1231 This is a bit cleaner and results in dramatic memory savings, albeit
1232 at a slight CPU cost.
1234 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1235 as requested by Marc Sherman.
1237 SC/10 Eximstats: added histograms for user specified patterns as requested
1240 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1242 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1243 fopen() in the content-scanning modules that did not already have it.
1249 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1250 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1252 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1254 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1256 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1257 to test Sieve filters that use "vacation".
1259 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1260 that changes the way the GnuTLS parameters are stored in the cache file.
1261 The new format can be generated externally. For backward compatibility,
1262 if the data in the cache doesn't make sense, Exim assumes it has read an
1263 old-format file, and it generates new data and writes a new file. This
1264 means that you can't go back to an older release without removing the
1267 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1268 work if there are any delivery delays because "one_time" forces the
1269 parent to be marked "delivered", so its unseen clone is never tried
1270 again. For this reason, Exim now forbids the simultaneous setting of
1273 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1274 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1275 entry if you want to know the details. The fix, however, neglected to
1276 consider the case when local delivery batching is involved. The test for
1277 "previously delivered" was not happening when checking to see if an
1278 address could be batched with a previous (undelivered) one; under
1279 certain circumstances this could lead to multiple deliveries to the same
1282 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1283 in its include files, and this causes problems building Exim.
1285 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1286 header_syntax) but Exim was just ignoring anything given after a slash.
1287 In particular, this caused confusion with an attempt to use "verify =
1288 reverse_host_lookup/defer_ok". An error is now given when options are
1289 supplied for verify items that do not have them. (Maybe reverse_host_
1290 lookup should have a defer_ok option, but that's a different point.)
1292 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1293 defined by RFC 821) to 2048, because there were problems with some AUTH
1294 commands, and RFC 1869 says the size should be increased for extended
1295 SMTP commands that take arguments.
1297 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1300 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1301 "unknown" error; now it says that the functionality isn't in the binary.
1303 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1304 an address' error message when a string expansion fails (syntax or
1305 whatever). Otherwise the password may appear in the log. Following change
1306 PH/42 below, there is no longer a chance of it appearing in a bounce
1309 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1311 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1312 of its addresses was ever tried. (Bugzilla bug #2.)
1314 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1315 the result incorrectly in the debug output. (It correctly added a newline
1316 to what was transported.)
1318 TF/01 Added $received_time.
1320 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1321 commented out examples of how to interface to a virus scanner and to
1322 SpamAssassin. Also added commented examples of av_scanner and
1323 spamd_address settings.
1325 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1326 and controls are allowed in which ACLs. There were a couple of minor
1327 errors. Some of the entries in the conditions table (which is a table of
1328 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1329 negation of where the condition IS allowed.
1331 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1333 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1334 header file does not have a version number, so I've had to invent a new
1335 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1336 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1337 radiusclient), but it was contributed by a Radius user.
1339 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1340 files or directories, for OpenSSL.
1342 PH/19 When an Exim process that is running as root has to create an Exim log
1343 file, it does so in a subprocess that runs as exim:exim so as to get the
1344 ownership right at creation (otherwise, other Exim processes might see
1345 the file with the wrong ownership). There was no test for failure of this
1346 fork() call, which would lead to the process getting stuck as it waited
1347 for a non-existent subprocess. Forks do occasionally fail when resources
1348 run out. I reviewed all the other calls to fork(); they all seem to check
1351 PH/20 When checking for unexpected SMTP input at connect time (before writing
1352 the banner), Exim was not dealing correctly with a non-positive return
1353 from the read() function. If the client had disconnected by this time,
1354 the result was a log entry for a synchronization error with an empty
1355 string after "input=" when read() returned zero. If read() returned -1
1356 (an event I could not check), uninitialized data bytes were printed.
1357 There were reports of junk text (parts of files, etc) appearing after
1360 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1362 PH/22 Added support for macro redefinition, and (re)definition in between
1363 driver and ACL definitions.
1365 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1366 forgetting to use the resulting value; it was using the unexpanded value.
1368 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1369 hadn't been configured. The fix is from Juergen Kreileder, who
1370 understands it better than I do:
1372 "Here's what I see happening with three configured cyrus_sasl
1373 authenticators configured (plain, login, cram-md5):
1375 On startup auth_cyrus_sasl_init() gets called for each of these.
1376 This means three calls to sasl_listmech() without a specified mech_list.
1377 => SASL tests which mechs of all available mechs actually work
1378 => three warnings about OTP not working
1379 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1381 With the patch, sasl_listmech() also gets called three times. But now
1382 SASL's mech_list option is set to the server_mech specified in the the
1383 authenticator. Or in other words, the answer from sasl_listmech()
1384 gets limited to just the mech you're testing for (which is different
1386 => the return list contains just 'plain' or 'login', 'cram-md5' or
1387 nothing depending on the value of ob->server_mech.
1389 I've just tested the patch: Authentication still works fine,
1390 unavailable mechs specified in the exim configuration are still
1391 caught, and the auth.log warnings about OTP are gone."
1393 PH/25 When debugging is enabled, the contents of the command line are added
1394 to the debugging output, even when log_selector=+arguments is not
1397 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1398 answer is "GNU", and only if the return is "GNU/something" is the answer
1401 PH/27 $acl_verify_message is now set immediately after the failure of a
1402 verification in an ACL, and so is available in subsequent modifiers. In
1403 particular, the message can be preserved by coding like this:
1405 warn !verify = sender
1406 set acl_m0 = $acl_verify_message
1408 Previously, $acl_verify_message was set only while expanding "message"
1409 and "log_message" when a very denied access.
1411 PH/28 Modified OS/os.c-Linux with
1413 -#ifndef OS_LOAD_AVERAGE
1414 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1416 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1417 nomenclature these days.)
1419 PH/29 Installed patch from the Sieve maintainer that adds the options
1420 sieve_useraddress and sieve_subaddress to the redirect router.
1422 PH/30 In these circumstances:
1423 . Two addresses routed to the same list of hosts;
1424 . First host does not offer TLS;
1425 . First host accepts first address;
1426 . First host gives temporary error to second address;
1427 . Second host offers TLS and a TLS session is established;
1428 . Second host accepts second address.
1429 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1430 and peerdn, if requested) that were in fact used only for the second
1433 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1434 attention to any local part prefix or suffix that was matched by the
1435 router that accepted the address. It now behaves in the same way as it
1436 does for delivery: the affixes are removed from the local part unless
1437 rcpt_include_affixes is set on the transport.
1439 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1440 timeout during the DATA phase of an incoming message.
1442 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1443 applied a patch sanctioned by the Sieve maintainer.
1445 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1446 the uid or gid is negative. A case of a negative gid caused this to be
1447 noticed. The fix allows for either to be negative.
1449 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1450 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1453 PH/36 The change PH/12 above was broken. Fixed it.
1455 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1456 the grounds that routing the same address twice would always produce the
1457 same answer. This might have been true once, but it is certainly no
1458 longer true now. Routing a child address may depend on the previous
1459 routing that produced that child. Some complicated redirection strategies
1460 went wrong when messages had multiple recipients, and made Exim's
1461 behaviour dependent on the order in which the addresses were given.
1463 I have moved the duplicate checking until after the routing is complete.
1464 Exim scans the addresses that are assigned to local and remote
1465 transports, and removes any duplicates. This means that more work will be
1466 done, as duplicates will always all be routed, but duplicates are
1467 presumably rare, so I don't expect this is of any significance.
1469 For deliveries to pipes, files, and autoreplies, the duplicate checking
1470 still happens during the routing process, since they are not going to be
1473 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1474 It corrects a timeout issue with spamd. This is Ian's comment: "The
1475 background is that sometimes spamd either never reads data from a
1476 connection it has accepted, or it never writes response data. The exiscan
1477 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1478 blindly assumes that writes won't block so it may never time out."
1480 PH/39 Allow G after quota size as well as K and M.
1482 PH/40 The value set for $authenticated_id in an authenticator may not contain
1483 binary zeroes or newlines because the value is written to log lines and
1484 to spool files. There was no check on this. Now the value is run through
1485 the string_printing() function so that such characters are converted to
1486 printable escape sequences.
1488 PH/41 $message_linecount is a new variable that contains the total number of
1489 lines in the message. Compare $body_linecount, which is the count for the
1492 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1493 bounce and delay warning messages, except in certain special cases, which
1496 (a) An SMTP error message from a remote host;
1497 (b) A message specified in a :fail: redirection;
1498 (c) A message specified in a "fail" command in a system filter;
1499 (d) A message specified in a FAIL return from the queryprogram router;
1500 (e) A message specified by the cannot_route_message router option.
1502 In these cases only, Exim does include the error details in bounce and
1503 warning messages. There are also a few cases where bland messages such
1504 as "unrouteable address" or "local delivery error" are given.
1506 PH/43 $value is now also set for the "else" part of a ${run expansion.
1508 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1509 being worked on, but at least Exim now implements the latest version to
1512 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1513 process to complete was treated as a delivery failure, a timeout while
1514 writing the message to the pipe was logged, but erroneously treated as a
1515 successful delivery. Such timeouts include transport filter timeouts. For
1516 consistency with the overall process timeout, these timeouts are now
1517 treated as errors, giving rise to delivery failures by default. However,
1518 there is now a new Boolean option for the pipe transport called
1519 timeout_defer, which, if set TRUE, converts the failures into defers for
1520 both kinds of timeout. A transport filter timeout is now identified in
1523 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1524 systems where "make" and "gmake" are different, calling "gmake" at top
1525 level broke things. I've arranged for the value of $(MAKE) to be passed
1526 from the Makefile to this script so that it can call the same version of
1530 A note about Exim versions 4.44 and 4.50
1531 ----------------------------------------
1533 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1534 changes of various kinds. As a consequence, a big documentation update was
1535 needed. This delayed the release for rather longer than seemed good, especially
1536 in the light of a couple of (minor) security issues. Therefore, the changes
1537 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1538 release. So 4.44 and 4.50 are in effect two different branches that both start
1541 I have left the 4.50 change log unchanged; it contains all the changes since
1542 4.43. The change log for 4.44 is below; many of its items are identical to
1543 those for 4.50. This seems to be the most sensible way to preserve the
1544 historical information.
1550 1. Minor wording change to the doc/README.SIEVE file.
1552 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1553 computation of the current number of files was incorrect.
1555 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1556 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1559 4. Give more explanation in the error message when the command for a transport
1560 filter fails to execute.
1562 5. There are several places where Exim runs a non-Exim command in a
1563 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1564 was being done only for the command run by the queryprogram router. It is
1565 now done for all such subprocesses. The other cases are: ${run, transport
1566 filters, and the commands run by the lmtp and pipe transports.
1568 6. Added CONFIGURE_GROUP build-time option.
1570 7. Some older OS have a limit of 256 on the maximum number of file
1571 descriptors. Exim was using setrlimit() to set 1000 as a large value
1572 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1573 systems. I've change it so that if it can't get 1000, it tries for 256.
1575 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1576 was an oversight, and furthermore, ever since the addition of extra
1577 controls (e.g. 4.43/32), the checks on when to allow different forms of
1578 "control" were broken. There should now be diagnostics for all cases when a
1579 control that does not make sense is encountered.
1581 9. Added the /retain_sender option to "control=submission".
1583 10. $recipients is now available in the predata ACL (oversight).
1585 11. Tidy the search cache before the fork to do a delivery from a message
1586 received from the command line. Otherwise the child will trigger a lookup
1587 failure and thereby defer the delivery if it tries to use (for example) a
1588 cached ldap connection that the parent has called unbind on.
1590 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1591 of $address_data from the recipient verification was clobbered by the
1592 sender verification.
1594 13. The value of address_data from a sender verification is now available in
1595 $sender_address_data in subsequent conditions in the ACL statement.
1597 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1599 15. Added a new option "connect=<time>" to callout options, to set a different
1602 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1603 was its contents. (It was OK if the option was not defined at all.)
1605 17. A "Completed" log line is now written for messages that are removed from
1606 the spool by the -Mrm option.
1608 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1609 information about exactly what failed.
1611 19. Added -dd to debug only the daemon process.
1613 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1614 handles timeouts, both on the server side and network timeouts. Renamed the
1615 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1617 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1618 It is now set to "smtps".
1620 22. $host_address is now set to the target address during the checking of
1621 ignore_target_hosts.
1623 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1624 being passed; this would have caused $sender_host_name to have been used if
1625 matching the list had actually called for a host name (not very likely,
1626 since this list is usually IP addresses). A host name is now passed as
1629 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1630 code that specifies a non-restarting handler (typically sigaction() in
1631 modern systems) in an attempt to fix a rare and obscure crash bug.
1633 25. Narrowed the window for a race in the daemon that could cause it to ignore
1634 SIGCHLD signals. This is not a major problem, because they are used only to
1635 wake it up if nothing else does.
1637 26. A malformed maildirsize file could cause Exim to calculate negative values
1638 for the mailbox size or file count. Odd effects could occur as a result.
1639 The maildirsize information is now recalculated if the size or filecount
1642 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1643 support for a long time. Removed HAVE_SYS_VFS_H.
1645 28. Installed the latest version of exipick from John Jetmore.
1647 29. In an address list, if the pattern was not a regular expression, an empty
1648 subject address (from a bounce message) matched only if the pattern was an
1649 empty string. Non-empty patterns were not even tested. This was the wrong
1650 because it is perfectly reasonable to use an empty address as part of a
1651 database query. An empty address is now tested by patterns that are
1652 lookups. However, all the other forms of pattern expect the subject to
1653 contain a local part and a domain, and therefore, for them, an empty
1654 address still always fails if the pattern is not itself empty.
1656 30. Exim went into a mad DNS loop when attempting to do a callout where the
1657 host was specified on an smtp transport, and looking it up yielded more
1658 than one IP address.
1660 31. Re-factored the code for checking spool and log partition space into a
1661 function that finds that data and another that does the check. The former
1662 is then used to implement four new variables: $spool_space, $log_space,
1663 $spool_inodes, and $log_inodes.
1665 32. The RFC2047 encoding function was originally intended for short strings
1666 such as real names; it was not keeping to the 75-character limit for
1667 encoded words that the RFC imposes. It now respects the limit, and
1668 generates multiple encoded words if necessary. To be on the safe side, I
1669 have increased the buffer size for the ${rfc2047: expansion operator from
1672 33. It is now permitted to omit both strings after an "if" condition; if the
1673 condition is true, the result is "true". As before, when the second string
1674 is omitted, a false condition yields an empty string. This makes it less
1675 cumbersome to write custom ACL and router conditions.
1677 34. Failure to deliver a bounce message always caused it to be frozen, even if
1678 there was an errors_to setting on the router. The errors_to setting is now
1681 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1682 canonical form (fully expanded) before being placed in
1683 $sender_host_address.
1685 36. The table in the code that translates DNS record types into text (T_A to
1686 "A" for instance) was missing entries for NS and CNAME. It is just possible
1687 that this could have caused confusion if both these types were looked up
1688 for the same domain, because the text type is used as part of Exim's
1689 per-process caching. But the chance of anyone hitting this buglet seems
1692 37. The dnsdb lookup has been extended in a number of ways.
1694 (1) There is a new type, "zns", which walks up the domain tree until it
1695 finds some nameserver records. It should be used with care.
1697 (2) There is a new type, "mxh", which is like "mx" except that it returns
1698 just the host names, not the priorities.
1700 (3) It is now possible to give a list of domains (or IP addresses) to be
1701 looked up. The behaviour when one of the lookups defers can be
1702 controlled by a keyword.
1704 (4) It is now possible to specify the separator character for use when
1705 multiple records are returned.
1707 38. The dnslists ACL condition has been extended: it is now possible to supply
1708 a list of IP addresses and/or domains to be looked up in a particular DNS
1711 39. Added log_selector=+queue_time_overall.
1713 40. When running the queue in the test harness, wait just a tad after forking a
1714 delivery process, to get repeatability of debugging output.
1716 41. Include certificate and key file names in error message when GnuTLS fails
1717 to set them up, because the GnuTLS error message doesn't include the name
1718 of the failing file when there is a problem reading it.
1720 42. Allow both -bf and -bF in the same test run.
1722 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1724 44. The "Exiscan patch" is now merged into the mainline Exim source.
1726 45. Sometimes the final signoff response after QUIT could fail to get
1727 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1728 < 0 before doing a fflush(). This bug looks as though it goes back to the
1729 introduction of TLS in release 3.20, but "sometimes" must have been rare
1730 because the tests only now provoked it.
1732 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1733 (this can affect the format of dates).
1735 47. exim_tidydb, when checking for the continued existence of a message for
1736 which it has found a message-specific retry record, was not finding
1737 messages that were in split spool directories. Consequently, it was
1738 deleting retry records that should have stayed in existence.
1740 48. Steve fixed some bugs in eximstats.
1742 49. The SPA authentication driver was not abandoning authentication and moving
1743 on to the next authenticator when an expansion was forced to fail,
1744 contradicting the general specification for all authenticators. Instead it
1745 was generating a temporary error. It now behaves as specified.
1747 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1748 (the order specifies the preference for clients). The order is now AES256,
1749 AES128, 3DES, ARCFOUR128.
1751 51. Small patch to Sieve code - explicitly set From: when generating an
1754 52. Exim crashed if a remote delivery caused a very long error message to be
1755 recorded - for instance if somebody sent an entire SpamAssassin report back
1756 as a large number of 550 error lines. This bug was coincidentally fixed by
1757 increasing the size of one of Exim's internal buffers (big_buffer) that
1758 happened as part of the Exiscan merge. However, to be on the safe side, I
1759 have made the code more robust (and fixed the comments that describe what
1762 53. Now that there can be additional text after "Completed" in log lines (if
1763 the queue_time_overall log selector is set), a one-byte patch to exigrep
1764 was needed to allow it to recognize "Completed" as not the last thing in
1767 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1768 patch that reportedly fixes this has been added. I am not expert enough to
1769 create a test for it. This is what the patch creator wrote:
1771 "I found a little strange behaviour of ldap code when working with
1772 Windows 2003 AD Domain, where users was placed in more than one
1773 Organization Units. When I tried to give exim partial DN, the exit code
1774 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1775 But simultaneously result of request was absolutely normal ldap result,
1776 so I produce this patch..."
1778 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1779 REFERENCE, so I have modified the code to exclude the patch when that macro
1782 55. Some experimental protocols are using DNS PTR records for new purposes. The
1783 keys for these records are domain names, not reversed IP addresses. The
1784 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1785 leaves it alone. Component reversal etc. now happens only for IP addresses.
1788 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1790 57. Double the size of the debug message buffer (to 2048) so that more of very
1791 long debug lines gets shown.
1793 58. The exicyclog utility now does better if the number of log files to keep
1794 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1796 59. Two changes related to the smtp_active_hostname option:
1798 (1) $smtp_active_hostname is now available as a variable.
1799 (2) The default for smtp_banner uses $smtp_active_hostname instead
1800 of $primary_hostname.
1802 60. The host_aton() function is supposed to be passed a string that is known
1803 to be a valid IP address. However, in the case of IPv6 addresses, it was
1804 not checking this. This is a hostage to fortune. Exim now panics and dies
1805 if the condition is not met. A case was found where this could be provoked
1806 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1807 components; fortuitously, this particular loophole had already been fixed
1808 by change 4.50/55 above.
1810 If there are any other similar loopholes, the new check in host_aton()
1811 itself should stop them being exploited. The report I received stated that
1812 data on the command line could provoke the exploit when Exim was running as
1813 exim, but did not say which command line option was involved. All I could
1814 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1815 running as the user.
1818 61. There was a buffer overflow vulnerability in the SPA authentication code
1819 (which came originally from the Samba project). I have added a test to the
1820 spa_base64_to_bits() function which I hope fixes it.
1823 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1824 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1826 63. The daemon start-up calls getloadavg() while still root for those OS that
1827 need the first call to be done as root, but it missed one case: when
1828 deliver_queue_load_max is set with deliver_drop_privilege. This is
1829 necessary for the benefit of the queue runner, because there is no re-exec
1830 when deliver_drop_privilege is set.
1832 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1833 This has been fixed.
1835 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1836 was in use, was not putting the data itself into the right store pool;
1837 consequently, it could be overwritten for a subsequent message in the same
1838 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1841 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1843 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1844 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1845 the function were treating the return as a boolean value, which happened to
1846 work because 0=false and not-0=true, but is not correct code.
1848 68. The host_aton() function was not handling scoped IPv6 addresses (those
1849 with, for example, "%eth0" on the end) correctly.
1851 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1852 negated items (that is, ~something) in unsigned ints. Some compilers
1853 apparently mutter when there is no cast.
1855 70. If an address verification called from an ACL failed, and did not produce a
1856 user-specific message (i.e. there was only a "system" message), nothing was
1857 put in $acl_verify_message. In this situation, it now puts the system
1860 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1861 session; change 4.31/43 added the unwanted input to the log line - except
1862 that it did not do this in the start of session case. It now does.
1864 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1865 This is wrong and can cause the other end to generate a synchronization
1866 error if it is another Exim or anything else that does the synchronization
1867 check. A QUIT command is no longer sent after a timeout.
1869 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1870 during host lookups.
1872 74. The defer_ok option of callout verification was not working if it was used
1873 when verifying addresses in header lines, that is, for this case:
1875 verify = header_sender/callout=defer_ok
1877 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1878 those file descriptors could be used for SMTP connections. If anything
1879 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1880 could be sent to the SMTP client, causing chaos. The daemon now opens
1881 stdin, stdout, and stderr to /dev/null when it puts itself into the
1884 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1885 log by default. The user can override this with suitable Perl magic.
1887 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1888 the log message when discard triggers, was not working for the DATA ACL or
1889 for the non-SMTP ACL.
1891 78. Error message wording change in sieve.c.
1893 79. If smtp_accept_max_per_host was set, the number of connections could be
1894 restricted to fewer than expected, because the daemon was trying to set up
1895 a new connection before checking whether the processes handling previous
1896 connections had finished. The check for completed processes is now done
1897 earlier. On busy systems, this bug wouldn't be noticed because something
1898 else would have woken the daemon, and it would have reaped the completed
1901 80. If a message was submitted locally by a user whose login name contained one
1902 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1903 It caused a spool format error. I have fixed the spool reading code. A
1904 related problem was that the "from" clause in the Received: line became
1905 illegal because of the space(s). It is now covered by ${quote_local_part.
1907 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1910 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1912 83. Patch from Sieve maintainer to fix unterminated string problem in
1913 "vacation" handling.
1915 84. Some minor changes to the Linux configuration files to help with other
1916 OS variants using glibc.
1918 85. One more patch for Sieve to update vacation handling to latest spec.
1921 ----------------------------------------------------
1922 See the note above about the 4.44 and 4.50 releases.
1923 ----------------------------------------------------
1929 1. Change 4.43/35 introduced a bug that caused file counts to be
1930 incorrectly computed when quota_filecount was set in an appendfile
1933 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1934 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1937 3. Give more explanation in the error message when the command for a transport
1938 filter fails to execute.
1940 4. There are several places where Exim runs a non-Exim command in a
1941 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1942 was being done only for the command run by the queryprogram router. It is
1943 now done for all such subprocesses. The other cases are: ${run, transport
1944 filters, and the commands run by the lmtp and pipe transports.
1946 5. Some older OS have a limit of 256 on the maximum number of file
1947 descriptors. Exim was using setrlimit() to set 1000 as a large value
1948 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1949 systems. I've change it so that if it can't get 1000, it tries for 256.
1951 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1952 was an oversight, and furthermore, ever since the addition of extra
1953 controls (e.g. 4.43/32), the checks on when to allow different forms of
1954 "control" were broken. There should now be diagnostics for all cases when a
1955 control that does not make sense is encountered.
1957 7. $recipients is now available in the predata ACL (oversight).
1959 8. Tidy the search cache before the fork to do a delivery from a message
1960 received from the command line. Otherwise the child will trigger a lookup
1961 failure and thereby defer the delivery if it tries to use (for example) a
1962 cached ldap connection that the parent has called unbind on.
1964 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1965 of $address_data from the recipient verification was clobbered by the
1966 sender verification.
1968 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1969 was its contents. (It was OK if the option was not defined at all.)
1971 11. A "Completed" log line is now written for messages that are removed from
1972 the spool by the -Mrm option.
1974 12. $host_address is now set to the target address during the checking of
1975 ignore_target_hosts.
1977 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1978 being passed; this would have caused $sender_host_name to have been used if
1979 matching the list had actually called for a host name (not very likely,
1980 since this list is usually IP addresses). A host name is now passed as
1983 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1984 code that specifies a non-restarting handler (typically sigaction() in
1985 modern systems) in an attempt to fix a rare and obscure crash bug.
1987 15. Narrowed the window for a race in the daemon that could cause it to ignore
1988 SIGCHLD signals. This is not a major problem, because they are used only to
1989 wake it up if nothing else does.
1991 16. A malformed maildirsize file could cause Exim to calculate negative values
1992 for the mailbox size or file count. Odd effects could occur as a result.
1993 The maildirsize information is now recalculated if the size or filecount
1996 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1997 support for a long time. Removed HAVE_SYS_VFS_H.
1999 18. Updated exipick to current release from John Jetmore.
2001 19. Allow an empty sender to be matched against a lookup in an address list.
2002 Previously the only cases considered were a regular expression, or an
2005 20. Exim went into a mad DNS lookup loop when doing a callout where the
2006 host was specified on the transport, if the DNS lookup yielded more than
2009 21. The RFC2047 encoding function was originally intended for short strings
2010 such as real names; it was not keeping to the 75-character limit for
2011 encoded words that the RFC imposes. It now respects the limit, and
2012 generates multiple encoded words if necessary. To be on the safe side, I
2013 have increased the buffer size for the ${rfc2047: expansion operator from
2016 22. Failure to deliver a bounce message always caused it to be frozen, even if
2017 there was an errors_to setting on the router. The errors_to setting is now
2020 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2021 canonical form (fully expanded) before being placed in
2022 $sender_host_address.
2024 24. Updated eximstats to version 1.33
2026 25. Include certificate and key file names in error message when GnuTLS fails
2027 to set them up, because the GnuTLS error message doesn't include the name
2028 of the failing file when there is a problem reading it.
2030 26. Expand error message when OpenSSL has problems setting up cert/key files.
2033 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2034 (this can affect the format of dates).
2036 28. exim_tidydb, when checking for the continued existence of a message for
2037 which it has found a message-specific retry record, was not finding
2038 messages that were in split spool directories. Consequently, it was
2039 deleting retry records that should have stayed in existence.
2041 29. eximstats updated to version 1.35
2042 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2043 1.35 - bugfix such that pie charts by volume are generated correctly
2045 30. The SPA authentication driver was not abandoning authentication and moving
2046 on to the next authenticator when an expansion was forced to fail,
2047 contradicting the general specification for all authenticators. Instead it
2048 was generating a temporary error. It now behaves as specified.
2050 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2051 (the order specifies the preference for clients). The order is now AES256,
2052 AES128, 3DES, ARCFOUR128.
2054 31. Small patch to Sieve code - explicitly set From: when generating an
2057 32. Exim crashed if a remote delivery caused a very long error message to be
2058 recorded - for instance if somebody sent an entire SpamAssassin report back
2059 as a large number of 550 error lines. This bug was coincidentally fixed by
2060 increasing the size of one of Exim's internal buffers (big_buffer) that
2061 happened as part of the Exiscan merge. However, to be on the safe side, I
2062 have made the code more robust (and fixed the comments that describe what
2065 33. Some experimental protocols are using DNS PTR records for new purposes. The
2066 keys for these records are domain names, not reversed IP addresses. The
2067 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2068 leaves it alone. Component reversal etc. now happens only for IP addresses.
2071 34. The host_aton() function is supposed to be passed a string that is known
2072 to be a valid IP address. However, in the case of IPv6 addresses, it was
2073 not checking this. This is a hostage to fortune. Exim now panics and dies
2074 if the condition is not met. A case was found where this could be provoked
2075 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2076 components; fortuitously, this particular loophole had already been fixed
2077 by change 4.50/55 or 4.44/33 above.
2079 If there are any other similar loopholes, the new check in host_aton()
2080 itself should stop them being exploited. The report I received stated that
2081 data on the command line could provoke the exploit when Exim was running as
2082 exim, but did not say which command line option was involved. All I could
2083 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2084 running as the user.
2087 35. There was a buffer overflow vulnerability in the SPA authentication code
2088 (which came originally from the Samba project). I have added a test to the
2089 spa_base64_to_bits() function which I hope fixes it.
2092 36. The daemon start-up calls getloadavg() while still root for those OS that
2093 need the first call to be done as root, but it missed one case: when
2094 deliver_queue_load_max is set with deliver_drop_privilege. This is
2095 necessary for the benefit of the queue runner, because there is no re-exec
2096 when deliver_drop_privilege is set.
2098 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2099 was in use, was not putting the data itself into the right store pool;
2100 consequently, it could be overwritten for a subsequent message in the same
2101 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2104 38. Sometimes the final signoff response after QUIT could fail to get
2105 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2106 < 0 before doing a fflush(). This bug looks as though it goes back to the
2107 introduction of TLS in release 3.20, but "sometimes" must have been rare
2108 because the tests only now provoked it.
2114 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2115 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2116 Now it returns an integer. A number of calls were still expecting a T/F
2117 return. Fortuitously, in all cases, the tests worked in OK situations,
2118 which is the norm. However, things would have gone wrong on any write
2119 failures on the smtp file descriptor. This function is used when sending
2120 messages over SMTP and also when doing verify callouts.
2122 2. When Exim is called to do synchronous delivery of a locally submitted
2123 message (the -odf or -odi options), it no longer closes stderr before doing
2126 3. Implemented the mua_wrapper option.
2128 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2130 5. Implemented the functions header_remove(), header_testname(),
2131 header_add_at_position(), and receive_remove_recipient(), and exported them
2134 6. If an ACL "warn" statement specified the addition of headers, Exim already
2135 inserted X-ACL-Warn: at the start if there was no header name. However, it
2136 was not making this test for the second and subsequent header lines if
2137 there were newlines in the string. This meant that an invalid header could
2138 be inserted if Exim was badly configured.
2140 7. Allow an ACL "warn" statement to add header lines at the start or after all
2141 the Received: headers, as well as at the end.
2143 8. Added the rcpt_4xx retry error code.
2145 9. Added postmaster_mailfrom=xxx to callout verification option.
2147 10. Added mailfrom=xxxx to the callout verification option, for verify=
2150 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2151 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2153 12. Inserted some casts to stop certain compilers warning when using pointer
2154 differences as field lengths or precisions in printf-type calls (mostly
2155 affecting debugging statements).
2157 13. Added optional readline() support for -be (dynamically loaded).
2159 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2160 same clock tick as a message's arrival, so that its received time was the
2161 same as the "first fail" time on the retry record, and that message
2162 remained on the queue past the ultimate address timeout, every queue runner
2163 would try a delivery (because it was past the ultimate address timeout) but
2164 after another failure, the ultimate address timeout, which should have then
2165 bounced the address, did not kick in. This was a "< instead of <=" error;
2166 in most cases the first failure would have been in the next clock tick
2167 after the received time, and all would be well.
2169 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2170 being recognized when the domain list was tested by the match_domain
2171 condition in an expansion string.
2173 16. Added the ${str2b64: operator.
2175 17. Exim was always calling setrlimit() to set a large limit for the number of
2176 processes, without checking whether the existing limit was already
2177 adequate. (It did check for the limit on file descriptors.) Furthermore,
2178 errors from getrlimit() and setrlimit() were being ignored. Now they are
2179 logged to the main and panic logs, but Exim does carry on, to try to do its
2180 job under whatever limits there are.
2182 18. Imported PCRE 5.0.
2184 19. Trivial typo in log message " temporarily refused connection" (the leading
2187 20. If the log selector return_path_on_delivery was set and an address was
2188 redirected to /dev/null, the delivery process crashed because it assumed
2189 that a return path would always be set for a "successful" delivery. In this
2190 case, the whole delivery is bypassed as an optimization, and therefore no
2193 21. Internal re-arrangement: the function for sending a challenge and reading
2194 a response while authentication was assuming a zero-terminated challenge
2195 string. It's now changed to take a pointer and a length, to allow for
2196 binary data in such strings.
2198 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2200 23. Exim was not respecting finduser_retries when seeking the login of the
2201 uid under which it was called; it was always trying 10 times. (The default
2202 setting of finduser_retries is zero.) Also, it was sleeping after the final
2203 failure, which is pointless.
2205 24. Implemented tls_on_connect_ports.
2207 25. Implemented acl_smtp_predata.
2209 26. If the domain in control=submission is set empty, Exim assumes that the
2210 authenticated id is a complete email address when it generates From: or
2211 Sender: header lines.
2213 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2214 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2215 chown and chgrp in /bin and hostname in /usr/bin.
2217 28. Exim was keeping the "process log" file open after each use, just as it
2218 does for the main log. This opens the possibility of it remaining open for
2219 long periods when the USR1 signal hits a daemon. Occasional processlog
2220 errors were reported, that could have been caused by this. Anyway, it seems
2221 much more sensible not to leave this file open at all, so that is what now
2224 29. The long-running daemon process does not normally write to the log once it
2225 has entered its main loop, and it closes the log before doing so. This is
2226 so that log files can straightforwardly be renamed and moved. However,
2227 there are a couple of unusual error situations where the daemon does write
2228 log entries, and I had neglected to close the log afterwards.
2230 30. The text of an SMTP error response that was received during a remote
2231 delivery was being truncated at 512 bytes. This is too short for some of
2232 the long messages that one sometimes sees. I've increased the limit to
2235 31. It is now possible to make retry rules that apply only when a message has a
2236 specific sender, in particular, an empty sender.
2238 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2239 it possible to be selective about when SMTP synchronization is enforced.
2241 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2243 32. Implemented hosts_connection_nolog.
2245 33. Added an ACL for QUIT.
2247 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2250 35. Added mailbox_size and mailbox_filecount to appendfile.
2252 36. Added control = no_multiline_responses to ACLs.
2254 37. There was a bug in the logic of the code that waits for the clock to tick
2255 in the case where the clock went backwards by a substantial amount such
2256 that the microsecond fraction of "now" was more than the microsecond
2257 fraction of "then" (but the whole seconds number was less).
2259 38. Added support for the libradius Radius client library this is found on
2260 FreeBSD (previously only the radiusclient library was supported).
2266 1. When certain lookups returned multiple values in the form name=value, the
2267 quoting of the values was not always being done properly. Specifically:
2268 (a) If the value started with a double quote, but contained no whitespace,
2270 (b) If the value contained whitespace other than a space character (i.e.
2271 tabs or newlines or carriage returns) it was not quoted.
2272 This fix has been applied to the mysql and pgsql lookups by writing a
2273 separate quoting function and calling it from the lookup code. The fix
2274 should probably also be applied to nisplus, ibase and oracle lookups, but
2275 since I cannot test any of those, I have not disturbed their existing code.
2277 2. A hit in the callout cache for a specific address caused a log line with no
2278 reason for rejecting RCPT. Now it says "Previous (cached) callout
2279 verification failure".
2281 3. There was an off-by-one bug in the queryprogram router. An over-long
2282 return line was truncated at 256 instead of 255 characters, thereby
2283 overflowing its buffer with the terminating zero. As well as fixing this, I
2284 have increased the buffer size to 1024 (and made a note to document this).
2286 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2287 when Exim is waiting for an SMTP response from a remote server, Exim
2288 restarts its select() call on the socket, thereby resetting its timeout.
2289 This is not a problem when such interrupts are rare. Somebody set up a cron
2290 job to run exiwhat every 2 minutes, which is less than the normal select()
2291 timeout (5 or 10 minutes). This meant that the select() timeout never
2292 kicked in because it was always reset. I have fixed this by comparing the
2293 time when an interrupt arrives with the time at the start of the first call
2294 to select(). If more time than the timeout has elapsed, the interrupt is
2295 treated as a timeout.
2297 5. Some internal re-factoring in preparation for the addition of Sieve
2298 extensions (by MH). In particular, the "personal" test is moved to a
2299 separate function, and given an option for scanning Cc: and Bcc: (which is
2300 not set for Exim filters).
2302 6. When Exim created an email address using the login of the caller as the
2303 local part (e.g. when creating a From: or Sender: header line), it was not
2304 quoting the local part when it contained special characters such as @.
2306 7. Installed new OpenBSD configuration files.
2308 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2309 try to make them clearer.
2311 9. Callout options, other than the timeout value, were being ignored when
2312 verifying sender addresses in header lines. For example, when using
2314 verify = header_sender/callout=no_cache
2316 the cache was (incorrectly) being used.
2318 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2319 only the Cygwin environment.
2321 11. When return_path_on_delivery was set as a log selector, if different remote
2322 addresses in the same message used different return paths and parallel
2323 remote delivery occurred, the wrong values would sometimes be logged.
2324 (Whenever a remote delivery process finished, the return path value from
2325 the most recently started remote delivery process was logged.)
2327 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2328 lines when AUTH and/or TLS are in use. This is the "received protocol"
2329 field. Exim used to use "asmtp" for authenticated SMTP, without any
2330 indication (in the protocol name) for TLS use. Now it follows the RFC and
2331 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2332 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2333 names appear in log lines as well as in Received: header lines.
2335 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2336 extensions, and comparison tests, and to fix some bugs.
2338 14. Changes to the "personal" filter test:
2340 (1) The test was buggy in that it was just doing the equivalent of
2341 "contains" tests on header lines. For example, if a user's address was
2342 anne@some.where, the "personal" test would incorrectly be true for
2344 To: susanne@some.where
2346 This test is now done by extracting each address from the header in turn,
2347 and checking the entire address. Other tests that are part of "personal"
2348 are now done using regular expressions (for example, to check local parts
2349 of addresses in From: header lines).
2351 (2) The list of non-personal local parts in From: addresses has been
2352 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2353 taken from the Sieve specification recommendations.
2355 (3) If the message contains any header line starting with "List-" it is
2356 treated as non-personal.
2358 (4) The test for "circular" in the Subject: header line has been removed
2359 because it now seems ill-conceived.
2361 15. Minor typos in src/EDITME comments corrected.
2363 16. Installed latest exipick from John Jetmore.
2365 17. If headers_add on a router specified a text string that was too long for
2366 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2367 of string_sprintf() is now avoided.
2369 18. $message_body_size was not set (it was always zero) when running the DATA
2370 ACL and the local_scan() function.
2372 19. For the "mail" command in an Exim filter, no default was being set for
2373 the once_repeat time, causing a random time value to be used if "once" was
2374 specified. (If the value happened to be <= 0, no repeat happened.) The
2375 default is now 0s, meaning "never repeat". The "vacation" command was OK
2376 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2377 (I found it when inspecting the code).
2379 20. There is now an overall timeout for performing a callout verification. It
2380 defaults to 4 times the callout timeout, which applies to individual SMTP
2381 commands during the callout. The overall timeout applies when there is more
2382 than one host that can be tried. The timeout is checked before trying the
2383 next host. This prevents very long delays if there are a large number of
2384 hosts and all are timing out (e.g. when the network connections are timing
2385 out). The value of the overall timeout can be changed by specifying an
2386 additional sub-option for "callout", called "maxwait". For example:
2388 verify = sender/callout=5s,maxwait=20s
2390 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2391 to the end before writing, but this should make it even safer).
2393 22. Exim was forgetting that it had advertised PIPELINING for the second and
2394 subsequent messages on an SMTP connection. It was also not resetting its
2395 memory on STARTTLS and an internal HELO.
2397 23. When Exim logs an SMTP synchronization error within a session, it now
2398 records whether PIPELINING has been advertised or not.
2400 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2401 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2402 rather than long int.
2404 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2406 26. Added the never_mail option to autoreply.
2412 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2413 crash if the getsockname() call failed; this can happen if a connection is
2414 closed very soon after it is established. The problem was simply in the
2415 order in which certain operations were done, causing Exim to try to write
2416 to the SMTP stream before it had set up the file descriptor. The bug has
2417 been fixed by making things happen in the correct order.
2423 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2424 before the connection was closed, thus losing the rejection response.
2426 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2427 some early Solaris releases, but causes trouble in current releases where
2428 socklen_t is defined.
2430 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2433 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2436 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2437 that releasing the top of it at the end releases what was used for sub-
2438 expansions (unless the block got too big). However, discard this block if
2439 the first thing is a variable or header, so that we can use its block when
2440 it is dynamic (useful for very large $message_headers, for example).
2442 6. Lookups now cache *every* query, not just the most recent. A new, separate
2443 store pool is used for this. It can be recovered when all lookup caches are
2444 flushed. Lookups now release memory at the end of their result strings.
2445 This has involved some general refactoring of the lookup sources.
2447 7. Some code has been added to the store_xxx() functions to reduce the amount
2448 of flapping under certain conditions.
2450 8. log_incoming_interface used to affect only the <= reception log lines. Now
2451 it causes the local interface and port to be added to several more SMTP log
2452 lines, for example "SMTP connection from", and rejection lines.
2454 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2456 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2458 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2459 could be overwritten at the end of the current message (or the start of a
2460 new message if it was set in a HELO ACL). The value is now preserved for
2461 the duration of the SMTP connection.
2463 12. If a transport had a headers_rewrite setting, and a matching header line
2464 contained an unqualified address, that address was qualified, even if it
2465 did not match any rewriting rules. The underlying bug was that the values
2466 of the flags that permit the existence of unqualified sender and recipient
2467 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2468 non-local messages, and by -bnq for local messages) were not being
2469 preserved with the message after it was received.
2471 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2472 "next input=" as part of the text comprising the host identity instead of
2473 the correct text. The code was using the same buffer for two different
2474 strings. However, depending on which order the printing function evaluated
2475 its arguments, the bug did not always show up. Under Linux, for example, my
2476 test suite worked just fine.
2478 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2479 doesn't work with some older versions of Perl. It has been changed to "my",
2480 which in any case is probably the better facility to use.
2482 15. A really picky compiler found some instances of statements for creating
2483 error messages that either had too many or two few arguments for the format
2486 16. The size of the buffer for calls to the DNS resolver has been increased
2487 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2488 for addresses that have a lot of PTR records. This alleviates a problem; it
2489 does not fully solve it.
2491 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2492 buffer now truncates the list and logs the incident, which is the same
2493 action as happens when Exim is looking up a host name and its aliases.
2494 Previously in this situation something unpredictable would happen;
2495 sometimes it was "internal error: store_reset failed".
2497 18. If a server dropped the connection unexpectedly when an Exim client was
2498 using GnuTLS and trying to read a response, the client delivery process
2499 crashed while trying to generate an error log message.
2501 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2502 string, for example:
2504 warn message = H1: something\nH2: something
2506 the text was added as a single header line from Exim's point of view
2507 though it ended up OK in the delivered message. However, searching for the
2508 second and subsequent header lines using $h_h2: did not work. This has been
2509 fixed. Similarly, if a system filter added multiple headers in this way,
2510 the routers could not see them.
2512 20. Expanded the error message when iplsearch is called with an invalid key to
2513 suggest using net-iplsearch in a host list.
2515 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2516 ACLs are no longer actually imposed (and a message to that effect is
2519 22. If a "gecos" field in a passwd entry contained escaped characters, in
2520 particular, if it contained a \" sequence, Exim got it wrong when building
2521 a From: or a Sender: header from that name. A second bug also caused
2522 incorrect handling when an unquoted " was present following a character
2523 that needed quoting.
2525 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2526 was not being matched caselessly.
2528 24. Arranged for all hyphens in the exim.8 source to be escaped with
2531 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2532 properly. Recipient callout cache records were still being keyed to include
2533 the sender, even when use_sender was set false. This led to far more
2534 callouts that were necessary. The sender is no longer included in the key
2535 when use_sender is false.
2537 26. Added "control = submission" modifier to ACLs.
2539 27. Added the ${base62d: operator to decode base 62 numbers.
2541 28. dnsdb lookups can now access SRV records.
2543 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2544 the configuration file.
2546 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2547 (-v) mode. This makes the output for a verbose queue run more intelligible.
2549 31. Added a use_postmaster feature to recipient callouts.
2551 32. Added the $body_zerocount variable, containing the number of binary zero
2552 bytes in the message body.
2554 33. The time of last modification of the "new" subdirectory is now used as the
2555 "mailbox time last read" when there is a quota error for a maildir
2558 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2560 35. Added +ignore_unknown as a special item in host lists.
2562 36. Code for decoding IPv6 addresses in host lists is now included, even if
2563 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2564 address was recognized as an IP address, but was then not correctly decoded
2565 into binary, causing unexpected and incorrect effects when compared with
2572 1. Very minor rewording of debugging text in manualroute to say "list of
2573 hosts" instead of "hostlist".
2575 2. If verify=header_syntax was set, and a header line with an unqualified
2576 address (no domain) and a large number of spaces between the end of the
2577 name and the colon was received, the reception process suffered a buffer
2578 overflow, and (when I tested it) crashed. This was caused by some obsolete
2579 code that should have been removed. The fix is to remove it!
2581 3. When running in the test harness, delay a bit after writing a bounce
2582 message to get a bit more predictability in the log output.
2584 4. Added a call to search_tidyup() just before forking a reception process. In
2585 theory, someone could use a lookup in the expansion of smtp_accept_max_
2586 per_host which, without the tidyup, could leave open a database connection.
2588 5. Added the variables $recipient_data and $sender_data which get set from a
2589 lookup success in an ACL "recipients" or "senders" condition, or a router
2590 "senders" option, similar to $domain_data and $local_part_data.
2592 6. Moved the writing of debug_print from before to after the "senders" test
2595 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2596 problems for message scanning, either using a data ACL, or using
2597 local_scan() because the Received: header was not generated till after they
2598 were called (in order to set the time as the time of reception completion).
2599 I have revised the way this works. The header is now generated after the
2600 body is received, but before the ACL or local_scan() are called. After they
2601 are run, the timestamp in the header is updated.
2607 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2608 before starting a queue runner without re-exec. This happened only when
2609 deliver_drop_privilege was set or when the Exim user was set to root. The
2610 effect of the bug was that timeouts during subsequent deliveries caused
2611 crashes instead of being properly handled. The handler is now left at its
2612 default (and expected) setting.
2614 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2615 message, again when deliver_drop_privilege is set or Exim is run as root.
2616 The bug described in (1) was not present in this case, but the tidying up
2617 of the other signals was missing. I have made the two cases consistent.
2619 3. The ignore_target_hosts setting on a manualroute router was being ignored
2620 for hosts that were looked up using the /MX notation.
2622 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2625 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2626 operated on the sender address. After changing the $sender_address to <>
2627 for the sender address verify, Exim was re-instated it as the original
2628 (before rewriting) address, but remembering that it had rewritten it, so it
2629 wasn't rewriting it again. This bug also had the effect of breaking the
2630 sender address verification caching when the sender address was rewritten.
2632 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2633 This has been changed so that if the ip literal address matches
2634 ignore_target_hosts, the router declines.
2636 7. Added expansion conditions match_domain, match_address, and match_local_
2637 part (NOT match_host).
2639 8. The placeholder for the Received: header didn't have a length field set.
2641 9. Added code to Exim itself and to exim_lock to test for a specific race
2642 condition that could lead to file corruption when using MBX delivery. The
2643 issue is with the lockfile that is created in /tmp. If this file is removed
2644 after a process has opened it but before that process has acquired a lock,
2645 there is the potential for a second process to recreate the file and also
2646 acquire a lock. This could lead to two Exim processes writing to the file
2647 at the same time. The added code performs the same test as UW imapd; it
2648 checks after acquiring the lock that its file descriptor still refers to
2649 the same named file.
2651 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2652 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2655 11. Added the smtp_active_hostname option. If used, this will typically be made
2656 to depend on the incoming interface address. Because $interface_address is
2657 not set up until the daemon has forked a reception process, error responses
2658 that can happen earlier (such as "too many connections") no longer contain
2661 12. If an expansion in a condition on a "warn" statement fails because a lookup
2662 defers, the "warn" statement is abandoned, and the next ACL statement is
2663 processed. Previously this caused the whole ACL to be aborted.
2665 13. Added the iplsearch lookup type.
2667 14. Added ident_timeout as a log selector.
2669 15. Added tls_certificate_verified as a log selector.
2671 16. Added a global option tls_require_ciphers (compare the smtp transport
2672 option of the same name). This controls incoming TLS connections.
2674 17. I finally figured out how to make tls_require_ciphers do a similar thing
2675 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2676 before starting the TLS session.
2678 18. Tabs are now shown as \t in -bP output.
2680 19. If the log selector return_path_on_delivery was set, Exim crashed when
2681 bouncing a message because it had too many Received: header lines.
2683 20. If two routers both had headers_remove settings, and the first one included
2684 a superfluous trailing colon, the final name in the first list and the
2685 first name in the second list were incorrectly joined into one item (with a
2686 colon in the middle).
2692 1. Added -C and -D options to the exinext utility, mainly to make it easier
2693 to include in the automated testing, but these could be helpful when
2694 multiple configurations are in use.
2696 2. The exinext utility was not formatting the output nicely when there was
2697 an alternate port involved in the retry record key, nor when there was a
2698 message id as well (for retries that were specific to a specific message
2699 and a specific host). It was also confused by IPv6 addresses, because of
2700 the additional colons they contain. I have fixed the IPv4 problem, and
2701 patched it up to do a reasonable job for IPv6.
2703 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2704 delivery, the log line now contains "pipelined" if PIPELINING was used.
2706 4. An SMTP transport process used to panic and die if the bind() call to set
2707 an explicit outgoing interface failed. This has been changed; it is now
2708 treated in the same way as a connect() failure.
2710 5. A reference to $sender_host_name in the part of a conditional expansion
2711 that was being skipped was still causing a DNS lookup. This no longer
2714 6. The def: expansion condition was not recognizing references to header lines
2715 that used bh_ and bheader_.
2717 7. Added the _cache feature to named lists.
2719 8. The code for checking quota_filecount in the appendfile transport was
2720 allowing one more file than it should have been.
2722 9. For compatibility with Sendmail, the command line option
2730 and sets the incoming protocol and host name (for trusted callers). The
2731 host name and its colon can be omitted when only the protocol is to be set.
2732 Note the Exim already has two private options, -pd and -ps, that refer to
2733 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2734 "s", but I don't think that's a major issue.
2736 10. A number of refactoring changes to the code, none of which should affect
2739 (a) The number of logging options was getting close to filling up the
2740 32-bit word that was used as a bit map. I have split them into two classes:
2741 those that are passed in the argument to log_write(), and those that are
2742 only ever tested independently outside of that function. These are now in
2743 separate 32-bit words, so there is plenty of room for expansion again.
2744 There is no change in the user interface or the logging behaviour.
2746 (b) When building, for example, log lines, the code previously used a
2747 macro that called string_cat() twice, in order to add two strings. This is
2748 not really sufficiently general. Furthermore, there was one instance where
2749 it was actually wrong because one of the argument was used twice, and in
2750 one call a function was used. (As it happened, calling the function twice
2751 did not affect the overall behaviour.) The macro has been replaced by a
2752 function that can join an arbitrary number of extra strings onto a growing
2755 (c) The code for expansion conditions now uses a table and a binary chop
2756 instead of a serial search (which was left over from when there were very
2757 few conditions). Also, it now recognizes conditions like "pam" even when
2758 the relevant support is not compiled in: a suitably worded error message is
2759 given if an attempt is made to use such a condition.
2761 11. Added ${time_interval:xxxxx}.
2763 12. A bug was causing one of the ddress fields not to be passed back correctly
2764 from remote delivery subprocesses. The field in question was not being
2765 subsequently used, so this caused to problems in practice.
2767 13. Added new log selectors queue_time and deliver_time.
2769 14. Might have fixed a bug in maildirsizefile handling that threw up
2770 "unexpected character" debug warnings, and recalculated the data
2771 unnecessarily. In any case, I expanded the warning message to give more
2774 15. Added the message "Restricted characters in address" to the statements in
2775 the default ACL that block characters like @ and % in local parts.
2777 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2778 Three changes have been made:
2780 (a) There was a serious bug; a negative response to MAIL caused the whole
2781 recipient domain to be cached as invalid, thereby blocking all messages
2782 to all local parts at the same domain, from all senders. This bug has
2783 been fixed. The domain is no longer cached after a negative response to
2784 MAIL if the sender used is not empty.
2786 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2789 (c) A new callout option, "use_sender" has been added for people who want
2790 the modified behaviour.
2796 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2799 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2800 indeed breaks things for older releases.
2802 3. Added additional logging to the case where there is a problem reading data
2803 from a filter that is running in a subprocess using a pipe, in order to
2804 try to track down a specific problem.
2806 4. Testing facility fudge: when running in the test harness and attempting
2807 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2808 getting "No route to host". Convert this to a timeout.
2810 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2813 6. Some OS don't have socklen_t but use size_t instead. This affects the
2814 fifth argument of getsockopt() amongst other things. This is now
2815 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2816 can be set for individual OS. I have set it for SunOS5, OSF1, and
2817 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2818 some earlier ones do not.
2820 7. Change 4.30/15 was not doing the test caselessly.
2822 8. The standard form for an IPv6 address literal was being rejected by address
2823 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2824 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2825 this, as well as the form without the "IPv6" on the front (but only when
2826 address literals are enabled, of course).
2828 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2830 10. Exim crashed if a message with an empty sender address specified by -f
2831 encountered a router with an errors_to setting. This could be provoked only
2832 by a command such as
2836 where an empty string was supplied; "<>" did not hit this bug.
2838 11. Installed PCRE release 4.5.
2840 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2841 remained set. It is now erased.
2843 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2844 times from message ids (which are base 36 rather than the normal 62).
2846 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2847 were being counted as actual protocol errors, and logged if the log
2848 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2849 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2850 and DATA following a set of rejected RCPTs do not count as protocol errors.
2851 In other words, Exim assumes they were pipelined, though this may not
2852 actually be the case. Of course, in all cases the client gets an
2853 appropriate error code.
2855 15. If a lookup fails in an ACL condition, a message about the failure may
2856 be available; it is used if testing the ACL cannot continue, because most
2857 such messages specify what the cause of the deferral is. However, some
2858 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2859 that caused an old message to be retained and used if a later statement
2860 caused a defer, replacing the real cause of the deferral.
2862 16. If an IP address had so many PTR records that the DNS lookup buffer
2863 was not large enough to hold them, Exim could crash while trying to process
2864 the truncated data. It now detects and logs this case.
2866 17. Further to 4.21/58, another change has been made: if (and only if) the
2867 first line of a message (the first header line) ends with CRLF, a bare LF
2868 in a subsequent header line has a space inserted after it, so as not to
2869 terminate the header.
2871 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2872 unnecessarily, used atoi() instead of strtol(), and didn't check the
2873 termination when getting file sizes from file names by regex.
2875 19. Completely re-implemented the support for maildirsize files, in the light
2876 of a number of problems with the previous contributed implementation
2877 (4.30/29). In particular:
2879 . If the quota is zero, the maildirsize file is maintained, but no quota is
2882 . If the maildir directory does not exist, it is created before any attempt
2883 to write a maildirsize file.
2885 . The quota value in the file is just a cache; if the quota is changed in
2886 the transport, the new value overrides.
2888 . A regular expression is available for excluding directories from the
2891 20. The autoreply transport checks the characters in options that define the
2892 message's headers; it allows continued headers, but it was checking with
2893 isspace() after an embedded newline instead of explicitly looking for a
2896 21. If all the "regular" hosts to which an address was routed had passed their
2897 expiry times, and had not reached their retry times, the address was
2898 bounced, even if fallback hosts were defined. Now Exim should go on to try
2901 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2902 equivalent code in the SMTP transport. Some hosts send humungous responses
2903 to HELO/EHLO, more than 1024 it seems.
2905 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2906 gives compiler warnings in some environments. I've now done it "properly",
2909 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2910 (because of problems with the built-in one) was declared to return uschar *
2911 instead of char *, causing compiler failure.
2913 25. Fixed a file descriptor leak when processing alias/forward files.
2915 26. Fixed a minor format string issue in dbfn.c.
2917 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2919 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2920 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2923 29. When a sender address is verified, it is cached, to save repeating the test
2924 when there is more than one recipient in a message. However, when the
2925 verification involves a callout, it is possible for different callout
2926 options to be set for different recipients. It is too complicated to keep
2927 track of this in the cache, so now Exim always runs a verification when a
2928 callout is required, relying on the callout cache for the optimization.
2929 The overhead is duplication of the address routing, but this should not be
2932 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2933 to be verified with callout=postmaster, and the main callout worked but the
2934 postmaster check failed, the verification correctly failed. However, if a
2935 subsequent RCPT command asked for sender verification *without* the
2936 postmaster check, incorrect caching caused this verification also to fail,
2939 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2940 it was not caching the DNS options (qualify_single, search_parents) that
2941 were used when the lookup failed. A subsequent lookup with different
2942 options therefore always gave the same answer, though there were cases
2943 where it should not have. (Example: a "domains = !$mx_any" option on a
2944 dnslookup router: the "domains" option is always processed without any
2945 widening, but the router might have qualify_single set.) Now Exim uses the
2946 cached value only when the same options are set.
2948 32. Added John Jetmore's "exipick" utility to the distribution.
2950 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2951 than a timeout (e.g. a certificate is required, and is not provided), an
2952 Exim server now closes the connection immediately. Previously it waited for
2953 the client to close - but if the client is SSL, it seems that they each
2954 wait for each other, leading to a delay before one of them times out.
2956 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2957 maintained 0.8.x compatibility because I don't think many are using it, and
2958 it is clearly obsolete.
2960 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2963 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2964 client certificate was expired. A simple patch fixes this, though I don't
2965 understand the full logic of why the verify callback is called multiple
2968 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2969 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2970 which causes problems with some clients (such as the Certicom SSL Plus
2971 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2972 disables the coutermeasure allowing Eudora to connect."
2974 38. Exim was not checking that a write() to a log file succeeded. This could
2975 lead to Bad Things if a log got too big, in particular if it hit a file
2976 size limit. Exim now panics and dies if it cannot write to a log file, just
2977 as it does if it cannot open a log file.
2979 39. Modified OS/Makefile-Linux so that it now contains
2981 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2983 The two -D definitions ensure that Exim is compiled with large file
2984 support, which makes it possible to handle log files that are bigger than
2987 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2988 instance) a domain was checked against a named list that involved a lookup,
2989 causing $domain_data to be set, then another domain was checked against the
2990 same list, then the first domain was re-checked, the value of $domain_data
2991 after the final check could be wrong. In particular, if the second check
2992 failed, it could be set empty. This bug probably also applied to
2995 41. The strip_trailing_dot option was not being applied to the address given
2996 with the -f command-line option.
2998 42. The code for reading a message's header from the spool was incrementing
2999 $received_count, but never initializing it. This meant that the value was
3000 incorrect (doubled) while delivering a message in the same process in which
3001 it was received. In the most common configuration of Exim, this never
3002 happens - a fresh exec is done - but it can happen when
3003 deliver_drop_privilege is set.
3005 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3006 it now includes up to 150 characters of the unexpected data in the log
3009 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3010 and building data strings. The size of both of these buffers was 10 000
3011 bytes - far larger than anybody would *ever* want, thought I. Needless to
3012 say, somebody hit the limit. I have increased the maximum line length to
3013 20 000 and the maximum data length of concatenated lines to 100 000. I have
3014 also fixed two bugs, because there was no checking on these buffers. Tsk,
3015 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3016 buffer is too small.
3018 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3019 lsearch lookups. Now it does.
3021 46. When parsing a route_list item in a manualroute router, a fixed-length
3022 buffer was used for the list of hosts. I made this 1024 bytes long,
3023 thinking that nobody would ever have a list of hosts that long. Wrong.
3024 Somebody had a whole pile of complicated expansion conditions, and the
3025 string was silently truncated, leading to an expansion error. It turns out
3026 that it is easier to change to an unlimited length (owing to other changes
3027 that have happened since this code was originally written) than to build
3028 structure for giving a limitation error. The length of the item that
3029 expands into the list of hosts is now unlimited.
3031 47. The lsearch lookup could not handle data where the length of text line was
3032 more than 4095 characters. Such lines were truncated, leading to shortened
3033 data being returned. It should now handle lines of any length.
3035 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3036 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3039 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3040 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3041 no longer happens in comment lines. A list of replacements is now placed
3042 at the head of all of the source files, except those whose only change is
3043 to replace PERL_COMMAND in the very first #! line.
3045 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3046 messages on the queue, with a bottom-up merge sort, using code contributed
3047 by Michael Haardt. This should make operations like -bp somewhat faster on
3048 large queues. It won't affect queue runners, except when queue_run_in_order
3051 51. Installed eximstats 1.31 in the distribution.
3053 52. Added support for SRV lookups to the dnslookup router.
3055 53. If an ACL referred to $message_body or $message_body_end, the value was not
3056 reset for any messages that followed in the same SMTP session.
3058 54. The store-handling optimization for building very long strings was not
3059 differentiating between the different store pools. I don't think this
3060 actually made any difference in practice, but I've tidied it.
3062 55. While running the routers to verify a sender address, $sender_address
3063 was still set to the sender address. This is wrong, because when routing to
3064 send a bounce to the sender, it would be empty. Therefore, I have changed
3065 it so that, while verifying a sender address, $sender_address is set to <>.
3066 (There is no change to what happens when verifying a recipient address.)
3068 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3069 target A or AAAA records (if not already returned) without resetting the
3070 qualify_single or search_parents options of the DNS resolver. These are
3071 inappropriate in this case because the targets of MX and SRV records must
3072 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3073 target that, when qualified, matched something in the local domain. These
3074 two options are now turned off when doing these lookups.
3076 57. It seems that at least some releases of Reiserfs (which does not have the
3077 concept of a fixed number of inodes) returns zero and not -1 for the
3078 number of available inodes. This interacted badly with check_spool_inodes,
3079 which assumed that -1 was the "no such thing" setting. What I have done is
3080 to check that the total number of inodes is greater than zero before doing
3081 the test of how many are available.
3083 58. When a "warn" ACL statement has a log_message modifier, the message is
3084 remembered, and not repeated. This is to avoid a lot of repetition when a
3085 message has many recipients that cause the same warning to be written.
3086 Howewer, Exim was preserving the list of already written lines for an
3087 entire SMTP session, which doesn't seem right. The memory is now reset if a
3088 new message is started.
3090 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3091 debugging output unless log_rewrite was also set.
3093 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3094 of (int)(handle) when we know that handle contains (void *)(-1).
3096 61. The Exim daemon panic-logs an error return when it closes the incoming
3097 connection. However "connection reset by peer" seems to be common, and
3098 isn't really an error worthy of noting specially, so that particular error
3101 62. When Exim is trying to find all the local interfaces, it used to panic and
3102 die if the ioctl to get the interface flags failed. However, it seems that
3103 on at least one OS (Solaris 9) it is possible to have an interface that is
3104 included in the list of interfaces, but for which you get a failure error
3105 for this call. This happens when the interface is not "plumbed" into a
3106 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3107 failure of the "get flags" call assumes that the interface is down.
3109 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3110 makes life easier for people who are doing arithmetic on fields extracted
3111 from dates, where you often get leading zeros that should not be
3112 interpreted as octal.
3114 64. Added qualify_domain to the redirect router, to override the global
3117 65. If a pathologically long header line contained very many addresses (the
3118 report of this problem mentioned 10 000) and each of them was rewritten,
3119 Exim could use up a very large amount of memory. (It kept on making new
3120 copies of the header line as it rewrote, and never released the old ones.)
3121 At the expense of a bit more processing, the header rewriting function has
3122 been changed so that it no longer eats memory in this way.
3124 66. The generation of the Received: header has been moved from the time that a
3125 message starts to be received, to the time that it finishes. The timestamp
3126 in the Received: header should now be very close to that of the <= log
3127 line. There are two side-effects of this change:
3129 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3130 logged header lines no longer include the local Received: line, because
3131 it has not yet been created. The same applies to a copy of the message
3132 that is returned to a non-SMTP sender when a message is rejected.
3134 (b) When a filter file is tested using -bf, no additional Received: header
3135 is added to the test message. After some thought, I decided that this
3138 This change does not affect the value of $received_for. It is still set
3139 after address rewriting, but before local_scan() is called.
3141 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3143 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3144 gave an unhelpful panic error message, and a defer error. I have managed to
3145 change this behaviour so that it now rejects any supplied certificate,
3146 which seems right, as the list of acceptable certificates is empty.
3148 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3149 gave an unhelpful defer error. I have not managed to make this reject any
3150 supplied certificates, but the error message it gives is "no certificate
3151 supplied", which is not helpful.
3153 70. exigrep's output now also includes lines that are not associated with any
3154 message, but which match the given pattern. Implemented by a patch from
3155 Martin Sluka, which also tidied up the Perl a bit.
3157 71. Recipient callout verification, like sender verification, was using <> in
3158 the MAIL FROM command. This isn't really the right thing, since the actual
3159 sender may affect whether the remote host accepts the recipient or not. I
3160 have changed it to use the actual sender in the callout; this means that
3161 the cache record is now keyed on a recipient/sender pair, not just the
3162 recipient address. There doesn't seem to be a real danger of callout loops,
3163 since a callout by the remote host to check the sender would use <>.
3164 [SEE ABOVE: changed after hitting problems.]
3166 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3167 temporary errors. However, in the case of such a code being given after
3168 the end of a data transmission (i.e. after ".") Exim was failing to write
3169 a retry record for the message. (Yes, there was some broken host that was
3170 actually sending 8xx at this point.)
3172 73. An unknown lookup type in a host list could cause Exim to panic-die when
3173 the list was checked. (An example that provoked this was putting <; in the
3174 middle of a list instead of at the start.) If this happened during a DATA
3175 ACL check, a -D file could be left lying around. This kind of configuration
3176 error no longer causes Exim to die; instead it causes a defer errror. The
3177 incident is still logged to the main and panic logs.
3179 74. Buglet left over from Exim 3 conversion. The message "too many messages
3180 in one connection" was written to the rejectlog but not the mainlog, except
3181 when address rewriting (yes!) was being logged.
3183 75. Added write_rejectlog option.
3185 76. When a system filter was run not as root (that is, when system_filter_user
3186 was set), the values of the $n variables were not being returned to the
3187 main process; thus, they were not subsequently available in the $sn
3190 77. Added +return_path_on_delivery log selector.
3192 78. A connection timeout was being treated differently from recipients deferred
3193 when testing hosts_max_try with a message that was older than the host's
3194 retry timeout. (The host should not be counted, thus allowing all hosts to
3195 be tried at least once before bouncing.) This may have been the cause of an
3196 occasionally reported bug whereby a message would remain on the queue
3197 longer than the retry timeout, but would be bounced if a delivery was
3198 forced. I say "may" because I never totally pinned down the problem;
3199 setting up timeout/retry tests is difficult. See also the next item.
3201 79. The ultimate address timeout was not being applied to errors that involved
3202 a combination of host plus message (for example, a timeout on a MAIL
3203 command). When an address resolved to a number of possible hosts, and they
3204 were not all tried for each delivery (e.g. because of hosts_max_try), a
3205 message could remain on the queue longer than the retry timeout.
3207 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3210 81. Fixed an obscure SMTP outgoing bug which required at least the following
3211 conditions: (a) there was another message waiting for the same server;
3212 (b) the server returned 5xx to all RCPT commands in the first message so
3213 that the message was not completed; (c) the server dropped the connection
3214 or gave a negative response to the RSET that Exim sends to abort the
3215 transaction. The observed case was a dropped connection after DATA that had
3216 been sent in pipelining mode. That is, the server had advertised PIPELINING
3217 but was not implementing it correctly. The effect of the bug was incorrect
3218 behaviour, such as trying another host, and this could lead to a crash.
3224 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3225 and daemon.c were passed as pointers to ints; they should have been
3226 pointers to socklen_t variables (which are typically unsigned ints).
3228 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3231 3. Fixed a really odd bug that affected only the testing scheme; patching a
3232 certain fixed string in the binary changed the value of another string that
3233 happened to be identical to the end of the original first string.
3235 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3236 name", it returns that address as the IP address. On some operating
3237 systems (e.g. Solaris), it also passes back the IP address string as the
3238 "host name". However, on others (e.g. Linux), it passes back an empty
3239 string. Exim wasn't checking for this, and was changing the host name to an
3240 empty string, assuming it had been canonicized.
3242 5. Although rare, it is permitted to have more than one PTR record for a given
3243 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3244 all the names associated with an address, because they do in Solaris.
3245 However, it seems that they do not in Linux for data that comes from the
3246 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3247 I found this out when I moved to a new Linux workstation and tried to run
3248 the Exim test suite.
3250 To get round this problem I have changed the code so that it now does its
3251 own call to the DNS to look up PTR records when searching for a host name.
3252 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3253 addresses that are only in /etc/hosts are still found.
3255 This behaviour is, however, controlled by an option called host_lookup_
3256 order, which defaults to "bydns:byaddr". If people want to use the other
3257 order, or indeed, just use one or the other means of lookup, they can
3258 specify it in this variable.
3260 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3261 some operating systems, this comes back from gethostbyaddr() as an empty
3262 string, and this is what Exim used to test for. However, it seems that in
3263 other systems, "." is yielded. Exim now tests for this case too.
3265 7. The values of check_spool_space and check_log_space are now held internally
3266 as a number of kilobytes instead of an absolute number of bytes. If a
3267 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3268 kilobyte. This means that much larger values can be stored.
3270 8. Exim monitor: an attempt to get the action menu when not actually pointing
3271 at a message produces an empty menu entitled "No message selected". This
3272 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3273 no entries in it ("Shell widget menu has zero width and/or height"). So I
3274 have added a single, blank menu entry in this case.
3276 9. Added ${quote_local_part.
3278 10. MIME decoding is now applied to the contents of Subject: header lines when
3281 11. Now that a reference to $sender_host_address automatically causes a reverse
3282 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3283 host lookup before query-style lookups in lists that might use this
3284 variable. This has therefore been abolished, and the "net-" prefix is no
3285 longer necessary for query-style lookups.
3287 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3288 have been a typo for LFLAGS, so it has been changed.
3290 13. The install script calls Exim with "-C /dev/null" in order to find the
3291 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3292 to be output. Howeve, since Exim outputs its version number before the
3293 error, it didn't break the script. It just looked ugly. I fixed this by
3294 always allowing "-C /dev/null" if the caller is root.
3296 14. Ignore overlarge ACL variable number when reading spool file - insurance
3297 against a later release with more variables having written the file.
3299 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3300 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3301 this, as well as the form without the "IPv6" on the front.
3303 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3304 OS/Makefile-Darwin file.
3306 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3307 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3309 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3310 variable. However, some OS use ints for inodes. I've added cast to long int
3311 to get rid of the compiler warning.
3313 19. I had forgotten to lock out "/../" in configuration file names when
3314 ALT_CONFIG_PREFIX was set.
3316 20. Routers used for verification do not need to specify transports. However,
3317 if such a router generated a host list, and callout was configured, Exim
3318 crashed, because it could not find a port number from the (non-existent)
3319 transport. It now assumes port 25 in this circumstance.
3321 21. Added the -t option to exigrep.
3323 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3324 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3325 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3326 them is set, LOOKUP_LSEARCH is forced.
3328 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3330 24. Added sender and host information to the "rejected by local_scan()" log
3331 line; previously there was no indication of these.
3333 25. Added .include_if_exists.
3335 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3336 Linux. It turns out that not all file systems support this. Apparently some
3337 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3338 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3339 not supported on the file descriptor, is now ignored when Exim is trying to
3340 sync a directory. This applies only to Linux.
3342 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3344 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3345 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3348 29. Added support for maildirsize files from supplied patch (modified a bit).
3350 30. The use of :fail: followed by an empty string could lead Exim to respond to
3351 sender verification failures with (e.g.):
3353 550 Verification failed for <xxx>
3354 550 Sender verify failed
3356 where the first response line was missing the '-' that indicates it is not
3357 the final line of the response.
3359 31. The loop for finding the name of the user that called Exim had a hardwired
3360 limit of 10; it now uses the value of finduser_retries, which is used for
3361 all other user lookups.
3363 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3366 33. Exim was neglecting to zero errno before one call of strtol() when
3367 expanding a string and expecting an integer value. On some systems this
3368 resulted in spurious "integer overflow" errors. Also, it was casting the
3369 result into an int without checking.
3371 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3372 did not work. The code looks as if it has *never* worked, though it appears
3373 to have been documented since at least releast 1.62. I have made it work.
3375 35. The "timeout_DNS" error in retry rules, also documented since at least
3376 1.62, also never worked. As it isn't clear exactly what this means, and
3377 clearly it isn't a major issue, I have abolished the feature by treating it
3378 as "timeout", and writing a warning to the main and panic logs.
3380 36. The display of retry rules for -brt wasn't always showing the error code
3383 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3384 timeout_connect_A, timeout_connect_MX.
3386 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3387 to the empty sender.
3389 39. The daemon was not analysing the content of -oX till after it had closed
3390 stderr and disconnected from the controlling terminal. This meant that any
3391 syntax errors were only noted on the panic log, and the return code from
3392 the command was 0. By re-arranging the code a little, I've made the
3393 decoding happen first, so such errors now appear on stderr, and the return
3394 code is 1. However, the actual setting up of the sockets still happens in
3395 the disconnected process, so errors there are still only recorded on the
3398 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3399 connections (as happens on some IP stacks) was logged at start up time as
3400 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3401 it from "IPv6 and IPv4", which means that two separate sockets are being
3404 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3405 says whether AF_INET or AF_INET6 was passed as an argument.
3407 42. Exiwhat output was messed up when time zones were included in log
3410 43. Exiwhat now gives more information about the daemon's listening ports,
3411 and whether -tls-on-connect was used.
3413 44. The "port" option of the smtp transport is now expanded.
3415 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3416 silently ignored. Now an error message is written to the main and panic
3419 46. There's a new ACL modifier called "logwrite" which writes to a log file
3420 as soon as it is encountered.
3422 47. Added $local_user_uid and $local_user_gid at routing time.
3424 48. Exim crashed when trying to verify a sender address that was being
3427 49. Exim was recognizing only a space character after ".include". It now also
3428 recognizes a tab character.
3430 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3431 extracting the relevant information from the specification. The man page no
3432 longer contains scrambled data for the -d option, and I've added a section
3433 at the front about calling Exim under different names.
3435 51. Added "extra_headers" argument to the "mail" command in filter files.
3437 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3440 53. Installed eximstats 1.29.
3442 54. Added transport_filter_timeout as a generic transport option.
3444 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3445 Cc: header lines. This was required by RFC 822, but it not required by RFC
3448 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3449 incoming messages that did not have them. Now it does so only if the
3450 message originates locally, that is, if there is no associated remote host
3451 address. When Resent- header lines are present, this applies to the Resent-
3452 lines rather than the non-Resent- lines.
3454 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3455 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3457 58. Messages for configuration errors now include the name of the main
3458 configuration file - useful now that there may be more than one file in a
3459 list (.included file names were always shown).
3461 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3462 for those rare installations that do not start the daemon as root or run it
3463 setuid root. I've cut out the call to initgroups() if the daemon is not
3466 60. The Exim user and group can now be bound into the binary as text strings
3467 that are looked up at the start of Exim's processing.
3469 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3471 62. Added $mailstore_basename variable.
3473 63. Installed patch to sieve.c from Michael Haardt.
3475 64. When Exim failed to open the panic log after failing to open the main log,
3476 the original message it was trying to log was written to stderr and debug
3477 output, but if they were not available (the usual case in production), it
3478 was lost. Now it is written to syslog before the two lines that record the
3479 failures to open the logs.
3481 65. Users' Exim filters run in subprocesses under the user's uid. It is
3482 possible for a "deliver" command or an alias in a "personal" command to
3483 provoke an address rewrite. If logging of address rewriting is configured,
3484 this fails because the process is not running as root or exim. There may be
3485 a better way of dealing with this, but for the moment (because 4.30 needs
3486 to be released), I have disabled address rewrite logging when running a
3487 filter in a non-root, non-exim process.
3493 1. The buildconfig auxiliary program wasn't quoting the value set for
3494 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3495 not defined. This bug was masked in 4.22 by the effect that was fixed in
3498 2. Some messages that were rejected after a message id was allocated were
3499 shown as "incomplete" by exigrep. It no longer does this for messages that
3500 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3502 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3503 have allow_domain_literals set, the ID did not get logged in the <= line.
3504 Domain literals are now always recognized in Message-ID: header lines.
3506 4. The first argument for a ${extract expansion item is the key name or field
3507 number. Leading and trailing spaces in this item were not being ignored,
3508 causing some misleading effects.
3510 5. When deliver_drop_privilege was set, single queue runner processes started
3511 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3512 same command in the process it spins off) were not dropping privilege.
3514 6. When the daemon running as "exim" started a queue runner, it always
3515 re-executed Exim in the spun-off process. This is a waste of effort when
3516 deliver_drop_privilege is set. The new process now just calls the
3517 queue-runner function directly.
3523 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3526 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3527 ignored. Though the use of -oP was forcing the writing of a pid file, it
3528 was always written to the default place.
3530 3. If the message "no IP address found for host xxxx" is generated during
3531 incoming verification, it is now followed by identification of the incoming
3532 connection (so you can more easily find what provoked it).
3534 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3536 5. Added some features to "harden" Exim a bit more against certain attacks:
3538 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3539 be put in Local/Makefile. This is like the never_users runtime option,
3540 but it cannot be overridden. The default setting is "root".
3542 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3543 prefix string with which any file named in a -C command line option
3546 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3547 is retained for -C and -D only if the caller of Exim is root. Without
3548 it, the exim user may also use -C and -D and retain privilege.
3550 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3551 command line option is disabled.
3553 6. Macro names set by the -D option must start with an upper case letter, just
3554 like macro names defined in the configuration file.
3556 7. Added "dereference=" facility to LDAP.
3558 8. Two instances of the typo "uknown" in the source files are fixed.
3560 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3561 the Configure-Makefile script screwed up while processing it.
3563 10. Incorporated PCRE 4.4.
3565 11. The SMTP synchronization check was not operating right at the start of an
3566 SMTP session. For example, it could not catch a HELO sent before the client
3567 waited for the greeting. There is now a check for outstanding input at the
3568 point when the greeting is written. Because of the duplex, asynchronous
3569 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3570 way, but not yet received, when the check is performed.
3572 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3573 on TCP/IP sockets, because this apparently causes some broken clients to
3576 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3577 unchanged) from the Cygwin maintainer.
3579 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3580 was supported instead of testing for mbx. Effectively a typo.
3582 15. The spa authenticator server code was not checking that the input it
3583 received was valid base64.
3585 16. The debug output line for the "set" modifier in ACLs was not showing the
3586 name of the variable that was being set.
3588 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3590 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3591 lookup on its own. Something else was needed to trigger the lookup. For
3592 example, a match in host_lookup or the need for a host name in a host list.
3593 Now, if $sender_host_name is referenced and the host name has not yet been
3594 looked up, a lookup is performed. If the lookup fails, the variable remains
3595 empty, and $host_lookup_failed is set to "1".
3597 19. Added "eqi" as a case-independent comparison operator.
3599 20. The saslauthd authentication condition could segfault if neither service
3600 nor realm was specified.
3602 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3603 error message that was logged was misleading, and incoming SMTP
3604 connections were dropped. The message is now more accurate, and temporary
3605 errors are given to SMTP connections.
3607 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3608 (except RSET) with a 421 error, until QUIT is received. However, it was
3609 failing to send a response to QUIT.
3611 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3612 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3613 ACL is now run after the helo_try_verify_hosts code.
3615 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3616 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3617 case-independent, so other case variants are also recognized). Apparently
3618 some systems use these upper case variants.
3620 25. If more than two messages were waiting for the same host, and a transport
3621 filter was specified for the transport, Exim sent two messages over the
3622 same TCP/IP connection, and then failed with "socket operation on non-
3623 socket" when it tried to send the third.
3625 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3627 27. The extern definition of crypt16() in expand.c was not being excluded when
3628 the OS had its own crypt16() function.
3630 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3631 as a preferred synonym for return_size_limit, both as an option and as an
3634 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3636 30. Changed the default configuration ACL to relax the local part checking rule
3637 for addresses that are not in any local domains. For these addresses,
3638 slashes and pipe symbols are allowed within local parts, but the sequence
3639 /../ is explicitly forbidden.
3641 31. SPA server authentication was not clearing the challenge buffer before
3644 32. log_message in a "warn" ACL statement was writing to the reject log as
3645 well as to the main log, which contradicts the documentation and doesn't
3646 seem right (because no rejection is happening). So I have stopped it.
3648 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3649 However, I am unable to do any testing of this.
3651 34. Fixed an infelicity in the appendfile transport. When checking directories
3652 for a mailbox, to see if any needed to be created, it was accidentally
3653 using path names with one or more superfluous leading slashes; tracing
3654 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3656 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3657 contents are added to the log line that is written for every discarded
3658 recipient. (Previously a log_message setting was ignored.)
3660 36. The ${quote: operator now quotes the string if it is empty.
3662 37. The install script runs exim in order to find its version number. If for
3663 some reason other than non-existence or emptiness, which it checks, it
3664 could not run './exim', it was installing it with an empty version number,
3665 i.e. as "exim-". This error state is now caught, and the installation is
3668 38. An argument was missing from the function that creates an error message
3669 when Exim fails to connect to the socket for saslauthd authentication.
3670 This could cause Exim to crash, or give a corrupted message.
3672 39. Added isip, isip4, and isip6 to ${if conditions.
3674 40. The ACL variables $acl_xx are now saved with the message, and can be
3675 accessed later in routers, transports, and filters.
3677 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3678 strings in the file are not string-expanded.
3680 42. If a MAIL command specified a SIZE value that was too large to fit into an
3681 int variable, the check against message_size_limit failed. Such values are
3682 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3683 day this will have to be increased, but I don't think I want to be around
3684 when emails are that large.
3691 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3692 iconv() is not standard in FreeBSD.
3694 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3695 IPv6 enabled. The observed symptom was a segmentation fault on return from
3696 the function os_common_find_running_interfaces() in src/os.c.
3698 3. In the check_special_case() function in daemon.c I had used "errno" as an
3699 argument name, which causes warnings on some systems. This was basically a
3700 typo, since it was named "eno" in the comments!
3702 4. The code that waits for the clock to tick (at a resolution of some fraction
3703 of a second) so as to ensure message-id uniqueness was always waiting for
3704 at least one whole tick, when it could have waited for less. [This is
3705 almost certainly not relevant at current processor speeds, where it is
3706 unlikely to ever wait at all. But we try to future-proof.]
3708 5. The function that sleeps for a time interval that includes fractions of a
3709 second contained a race. It did not block SIGALRM between setting the
3710 timer, and suspending (a couple of lines later). If the interval was short
3711 and the sigsuspend() was delayed until after it had expired, the suspension
3712 never ended. On busy systems this could lead to processes getting stuck for
3715 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3716 process, before it forks any delivery processes. The open lookup caching
3717 mechanism meant that the open file or database connection was passed into
3718 the delivery process. The problem was that delivery processes always tidy
3719 up cached lookup data. This could cause a problem for the next delivery
3720 process started by the queue runner, because the external queue runner
3721 process does not know about the closure. So the next delivery process
3722 still has data in the lookup cache. In the case of a file lookup, there was
3723 no problem because closing a file descriptor in a subprocess doesn't affect
3724 the parent. However, if the lookup was caching a connection to a database,
3725 the connection was closed, and the second delivery process was likely to
3726 see errors such as "PGSQL: query failed: server closed the connection
3727 unexpectedly". The problem has been fixed by closing all cached lookups
3728 in a queue runner before running a delivery process.
3730 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3731 seem to have the "const" qualifier which it has on other OS. I've
3734 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3735 *of the same type* (client or server) with the same public name that an
3736 error should be diagnosed.
3738 9. When Exim looked up a host name for an IP address, but failed to find the
3739 original IP address when looking up the host name (a safety check), it
3740 output the message "<ip address> does not match any IP for NULL", which was
3741 confusing, to say the least. The bug was that the host name should have
3742 appeared instead of "NULL".
3744 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3745 user that is built into the binary, and the -C or -D options is used, root
3746 privilege is dropped before the configuration file is read. In addition,
3747 logging is switched to stderr instead of the normal log files. If the
3748 configuration then re-defines the Exim user, the unprivileged environment
3749 is probably not what is expected, so Exim logs a panic warning message (but
3752 However, if deliver_drop_privilege is set, the unprivileged state may well
3753 be exactly what is intended, so the warning has been cut out in that case,
3754 and Exim is allowed to try to write to its normal log files.
3760 1. smtp_return_error_details was not giving details for temporary sender
3761 or receiver verification errors.
3763 2. Diagnose a configuration error if two authenticators have the same public
3766 3. Exim used not to create the message log file for a message until the first
3767 delivery attempt. This could be confusing when incoming messages were held
3768 for policy or load reasons. The message log file is now created at the time
3769 the message is received, and an initial "Received" line is written to it.
3771 4. The automatically generated man page for command line options had a minor
3772 bug that caused no ill effects; however, a more serious problem was that
3773 the procedure for building the man page automatically didn't always
3774 operate. Consequently, release 4.20 contains an out-of-date version. This
3775 shouldn't happen again.
3777 5. When building Exim with embedded Perl support, the script that builds the
3778 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3779 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3781 6. The freeze_tell option was not being used for messages that were frozen on
3782 arrival, either by an ACL or by local_scan().
3784 7. Added the smtp_incomplete_transaction log selector.
3786 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3787 was accepting AUTH without a new EHLO.
3789 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3790 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3793 10. Logging of TCP/IP connections (when configured) now happens in the main
3794 daemon process instead of the child process, so that the TCP/IP connection
3795 count is more accurate (but it can never be perfect).
3797 11. The use of "drop" in a nested ACL was not being handled correctly in the
3798 outer ACL. Now, if condition failure induced by the nested "drop" causes
3799 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3800 or "require"), the connection is dropped.
3802 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3803 that yield "discard" can now be used with an "accept" or a "discard" verb,
3804 but an error is generated for any others (because I can't see a useful way
3805 to define what should happen).
3807 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3808 are now processed in the same way as lines in the Exim configuration file.
3809 In particular, continuation lines are supported.
3811 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3813 15. Added -ti meaning -t -i.
3815 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3816 domains, and warn by logging if others are found.
3818 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3819 returned from GIFCONF when Exim is trying to find the list of interfaces on
3820 a host. The code in os.c has been modified to copy each ifreq to an aligned
3821 structure in all cases.
3823 Also, in some cases, the returned ifreq's were being copied to a 'struct
3824 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3825 means the last couple of bytes of an IPv6 address could be chopped if the
3826 ifreq contained only a normal sockaddr (14 bytes storage).
3828 18. Named domain lists were not supported in the hosts_treat_as_local option.
3829 An entry such as +xxxx was not recognized, and was treated as a literal
3832 19. Ensure that header lines added by a DATA ACL are included in the reject log
3833 if the ACL subsequently rejects the message.
3835 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3836 MD5 (which is deprecated).
3838 21. When testing a filter file using -bf, Exim was writing a message when it
3839 took the sender from a "From " line in the message, but it was not doing so
3840 when it took $return_path from a Return-Path: header line. It now does.
3842 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3843 with a valid header line field name (a series of printing characters
3844 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3846 23. Changed "disc" in the source to "disk" to conform to the documentation and
3847 the book and for uniformity.
3849 24. Ignore Sendmail's -Ooption=value command line item.
3851 25. When execve() failed while trying to run a command in a pipe transport,
3852 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3853 could be confused with a return value of 69 from the command itself. This
3854 has been changed to 127, the value the shell returns if it is asked to run
3855 a non-existent command. The wording for the related log line suggests a
3856 non-existent command as the problem.
3858 26. If received_header_text expands to an empty string, do not add a Received:
3859 header line to the message. (Well, it adds a token one on the spool, but
3860 marks it "old" so that it doesn't get used or transmitted.)
3862 27. Installed eximstats 1.28 (addition of -nt option).
3864 28. There was no check for failure on the call to getsockname() in the daemon
3865 code. This can fail if there is a shortage of resources on the system, with
3866 ENOMEM, for example. A temporary error is now given on failure.
3868 29. Contrary to the C standard, it seems that in some environments, the
3869 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3870 program. Exim now does this explicitly; it affects the formatting of
3871 timestamps using strftime().
3873 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3874 complaints. I've now initialized all the variables, to avoid this.
3876 32. Header lines added by a system filter were not being "seen" during
3877 transport-time rewrites.
3879 33. The info_callback() function passed to OpenSSL is set up with type void
3880 (*)(SSL *, int, int), as described somewhere. However, when calling the
3881 function (actually a macro) that sets it up, the type void(*)() is
3882 expected. I've put in a cast to prevent warnings from picky compilers.
3884 34. If a DNS black list lookup found a CNAME record, but there were no A
3885 records associated with the domain it pointed at, Exim crashed.
3887 35. If a DNS black list lookup returned more than one A record, Exim ignored
3888 all but the first. It now scans all returned addresses if a particular IP
3889 value is being sought. In this situation, the contents of the
3890 $dnslist_value variable are a list of all the addresses, separated by a
3893 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3894 to accept a host name and all its aliases if the forward lookup for any of
3895 them yielded the IP address of the incoming connection. Now it accepts only
3896 those names whose forward lookup yields the correct IP address. Any other
3897 names are discarded. This closes a loophole whereby a rogue DNS
3898 administrator could create reverse DNS records to break through a
3899 wildcarded host restriction in an ACL.
3901 37. If a user filter or a system filter that ran in a subprocess used any of
3902 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3903 the wrong values were passed to the pipe command ($thisaddress had the
3904 value of $0, $0 had the value of $1, etc). This bug was introduced by
3905 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3907 38. Improved the line breaking for long SMTP error messages from ACLs.
3908 Previously, if there was no break point between 40 and 75 characters, Exim
3909 left the rest of the message alone. Two changes have been made: (a) I've
3910 reduced the minimum length to 35 characters; (b) if it can't find a break
3911 point between 35 and 75 characters, it looks ahead and uses the first one
3912 that it finds. This may give the occasional overlong line, but at least the
3913 remaining text gets split now.
3915 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3916 file descriptors might be low, and that setting 1000 would always raise it.
3917 It turns out that in some environments, the limit is already over 1000 and
3918 that lowering it causes trouble. So now Exim takes care not to decrease it.
3920 40. When delivering a message, the value of $return_path is set to $sender_
3921 address at the start of routing (routers may change the value). By an
3922 oversight, this default was not being set up when an address was tested by
3923 -bt or -bv, which affected the outcome if any router or filter referred to
3926 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3927 log only when "message" or "log_message" are set. However, if one of the
3928 conditions was an address verification, or a call to a nested ACL, the
3929 messages generated by the underlying test were being passed through. This
3930 no longer happens. The underlying message is available in $acl_verify_
3931 message for both "message" and "log_message" expansions, so it can be
3932 passed through if needed.
3934 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3935 new expansion $bh_ to give the encoded byte string without charset
3936 translation. Translation happens only if iconv() is available; HAVE_ICONV
3937 indicates this at build time. HEADERS_CHARSET gives the charset to
3938 translate to; headers_charset can change it in the configuration, and
3939 "headers charset" can change it in an individual filter file.
3941 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3942 that creates RFC 2047 encoded "words" labels them as that charset instead
3943 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3944 expansion operator; (ii) when Exim creates a From: line for a local
3945 message; (iii) when a header line is rewritten to include a "phrase" part.
3947 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3948 buggy, causing it to skip the first lines of messages whose message ID
3949 ended in 'D'. This would not have bitten before Exim release 4.14, because
3950 message IDs were unlikely to end in 'D' before then. The effect was to have
3951 incorrect size information for certain domains.
3953 45. #include "config.h" was missing at the start of the crypt16.c module. This
3954 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3957 46. If there was a timeout during a "random" callout check, Exim treated it as
3958 a failure of the random address, and carried on sending RSET and the real
3959 address. If the delay was just some slowness somewhere, the response to the
3960 original RCPT would be taken as a response to RSET and so on, causing
3961 mayhem of various kinds.
3963 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3964 when I implemented it. It didn't allow for the fact that some option values
3965 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3966 the right test for positive values.
3968 48. Domain names in DNS records are case-independent. Exim always looks them up
3969 in lower case. Some resolvers return domain names in exactly the case they
3970 appear in the zone file, that is, they may contain uppercase letters. Not
3971 all resolvers do this - some return always lower case. Exim was treating a
3972 change of case by a resolver as a change of domain, similar to a widening
3973 of a domain abbreviation. This triggered its re-routing code and so it was
3974 trying to route what was effectively the same domain again. This normally
3975 caused routing to fail (because the router wouldn't handle the domain
3976 twice). Now Exim checks for this case specially, and just changes the
3977 casing of the domain that it ultimately uses when it transmits the message
3980 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3983 50. If a filter generated a file delivery with a non-absolute name (possible if
3984 no home directory exists for the router), the forbid_file option was not
3987 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3988 the existing equality matching.
3990 52. Exim was using ints instead of ino_t variables in some places where it was
3991 dealing with inode numbers.
3993 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3994 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3995 called TMPDIR, and if it finds it is different, it changes its value.
3997 54. The smtp_printf() function is now made available to local_scan() so
3998 additional output lines can be written before returning. There is also an
3999 smtp_fflush() function to enable the detection of a dropped connection.
4000 The variables smtp_input and smtp_batched_input are exported to
4003 55. Changed the default runtime configuration: the message "Unknown user"
4004 has been removed from the ACL, and instead placed on the localuser router,
4005 using the cannot_route_message feature. This means that any verification
4006 failures that generate their own messages won't get overridden. Similarly,
4007 the "Unrouteable address" message that was in the ACL for unverifiable
4008 relay addresses has also been removed.
4010 56. Added hosts_avoid_esmtp to the smtp transport.
4012 57. The exicyclog script was not checking for the esoteric option
4013 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4014 will work only if exicyclog is run under the appropriate euid.
4016 58. Following a discussion on the list, the rules by which Exim recognises line
4017 endings on incoming messages have been changed. The -dropcr and drop_cr
4018 options are now no-ops, retained only for backwards compatibility. The
4019 following line terminators are recognized: LF CRLF CR. However, special
4020 processing applies to CR:
4022 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4023 nor a local message in the state where . is a terminator.
4025 (ii) If a bare CR is encountered in a header line, an extra space is added
4026 after the line terminator so as not to end the header. The reasoning
4027 behind this is that bare CRs in header lines are most likely either
4028 to be mistakes, or people trying to play silly games.
4030 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4031 was being incorrectly given as 18 bytes larger than it should have been.
4032 This is a VOB (very old bug).
4034 60. This may never have affected anything current, but just in case it has:
4035 When the local host is found other than at the start of a list of hosts,
4036 the local host, those with the same MX, and any that follow, are discarded.
4037 When the list in question was part of a longer list of hosts, the following
4038 hosts (not currently being processed) were also being discarded. This no
4039 longer happens. I'm not sure if this situation could ever has previously
4042 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4045 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4046 header. This is something that is recommended in a new Internet Draft, and
4047 is something that is documented as being done by Sendmail. There are two
4048 possible values. For messages generated by the autoreply transport, Exim
4051 Auto-Submitted: auto-replied
4053 whereas for all other generated messages (e.g. bounces) it adds
4055 Auto-Submitted: auto-generated
4057 63. The "personal" condition in filters now includes a test for the
4058 Auto-Submitted: header. If it contains the string "auto-" the message it
4059 not considered personal.
4061 64. Added rcpt_include_affixes as a generic transport option.
4063 65. Added queue_only_override (default true).
4065 66. Added the syslog_duplication option.
4067 67. If what should have been the first header line of a message consisted of
4068 a space followed by a colon, Exim was mis-interpreting it as a header line.
4069 It isn't of course - it is syntactically invalid and should therefore be
4070 treated as the start of the message body. The misbehaviour could have
4071 caused a number of strange effects, including loss of data in subsequent
4072 header lines, and spool format errors.
4074 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4075 client host had authenticated. This control can now be exercised by an ACL
4076 for more flexibility.
4078 69. By default, callouts do not happen when testing with -bh. There is now a
4079 variant, -bhc, which does actually run the callout code, including
4080 consulting and updating the callout cache.
4082 70. Added support for saslauthd authentication, courtesy of Alexander
4085 71. If statvfs() failed on the spool or log directories while checking their
4086 size for availability, Exim confusingly gave the error "space shortage".
4087 Furthermore, in debugging mode it crashed with a floating point exception.
4088 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4089 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4090 serious problem, Exim now writes to the main and panic logs when this
4091 happens, with details of the failure. It then refuses to accept the
4092 incoming message, giving the message "spool directory problem" or "log
4093 directory problem" with a 421 code for SMTP messages.
4095 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4096 0, 1, and 2 exist, because some OS complain for execs without them (see
4097 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4098 descriptors. However, the code omitted to check that the open succeeded,
4099 causing mysterious errors if for some reason the permissions on /dev/null
4100 got screwed. Now Exim writes a message to the main and panic logs, and
4101 bombs out if it can't open /dev/null.
4103 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4104 interact so that it is all more flexible. It is supposed to remain
4105 backwards compatible. Also added extra_local_interfaces.
4107 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4108 to bomb out with an assertion failure - to the client this appears as a
4109 connection drop. This problem occurs in the part of the code that was taken
4110 from the Samba project. Fortunately, the assertion is in a very simple
4111 function, so I have fixed this by reproducing the function inline in the
4112 one place where it is called, and arranging for authentication to fail
4113 instead of killing the process with assert().
4115 75. The SPA client code was not working when the server requested OEM rather
4116 than Unicode encoding.
4118 76. Added code to make require_files with a specific uid setting more usable in
4119 the case where statting the file as root fails - usually a non-root-mounted
4120 NFS file system. When this happens and the failure is EACCES, Exim now
4121 forks a subprocess and does the per-uid checking as the relevant uid.
4123 77. Added process_log_path.
4125 78. If log_file_path was not explicitly set, a setting of check_log_space or
4126 check_log_inodes was ignored.
4128 79. If a space check for the spool or log partitions fails, the incident is now
4129 logged. Of course, in the latter case the data may get lost...
4131 80. Added the %p formatting code to string_format() so that it can be used to
4132 print addresses in debug_print(). Adjusted all the address printing in the
4133 debugging in store.c to use %p rather than %d.
4135 81. There was a concern that a line of code in smtp_in.c could overflow a
4136 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4137 initially expressed, the concern was not well-founded, because trailing
4138 spaces are removed early. However, if the trailing spaces were followed by
4139 a NULL, they did not get removed, so the overflow was possible. Two fixes
4142 (a) I re-wrote the offending code in a cleaner fashion.
4143 (b) If an incoming SMTP command contains a NULL character, it is rejected
4146 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4147 runs initgroups(), so that if the Exim user is in any additional groups,
4148 they will be used during message reception.
4154 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob