1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.389 2006/09/05 15:34:40 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
14 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
26 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
29 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
34 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
41 SC/01 Use a glob alias rather than an array ref in eximstats generated
42 parser. This improves both readability and performance.
44 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
45 Don't display local sender or destination tables in eximstats unless
46 there is data to show.
47 Added average volumes into the eximstats top table text output.
49 SC/03 Collect data on the number of addresses (recipients) as well
50 as the number of messages in eximstats.
52 TF/01 Correct an error in the documentation for the redirect router. Exim
53 does (usually) call initgroups() when daemonizing.
55 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
56 with consistent privilege compared to when running as a daemon.
58 TF/03 Note in the spec that $authenticated_id is not set for local
59 submissions from trusted users.
61 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
62 Thanks to Dean Brooks <dean@iglou.com> for the patch.
64 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
65 by adding some example configuration directives to the default
66 configuration file. A little bit of work is required to uncomment the
67 directives and define how usernames and passwords are checked, but
68 there is now a framework to start from.
70 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
71 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
72 without this. I don't know how relevant this is to other LDAP libraries.
74 PH/02 Add the verb name to the "unknown ACL verb" error.
76 PH/03 Magnus Holmgren's patch for filter_prepend_home.
78 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
80 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
81 directory not expanded when it should be if an expanded home directory
82 was set for the address (which is overridden by the transport).
84 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
87 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
88 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
89 because it is too late at that time, and has no effect.
91 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
92 security issue with \' (bugzilla #107). I could not use the
93 PQescapeStringConn() function, because it needs a PGconn value as one of
96 PH/08 When testing addresses using -bt, indicate those final addresses that
97 are duplicates that would not cause an additional delivery. At least one
98 person was confused, thinking that -bt output corresponded to deliveries.
99 (Suppressing duplicates isn't a good idea as you lose the information
100 about possibly different redirections that led to the duplicates.)
102 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
103 systems where poll() doesn't work, in particular OS X.
105 PH/10 Added more information to debugging output for retry time not reached.
107 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
108 operations in malware.c.
110 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
113 PH/13 If write_rejectlog was set false when logging was sent to syslog with
114 syslog_duplication set false, log lines that would normally be written
115 both the the main log and to the reject log were not written to syslog at
118 PH/14 In the default configuration, change the use of "message" in ACL warn
119 statements to "add_header".
121 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
122 not followed by a command (e.g. "seen endif").
124 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
125 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
128 PH/17 Added extra conditions to the default value of delay_warning_condition
132 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
133 { match{$h_precedence:}{(?i)bulk|list|junk} } \
134 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
137 The Auto-Submitted: and various List- headers are standardised, whereas I
138 don't think Precedence: ever was.
140 PH/18 Refactored debugging code in route_finduser() to show more information,
141 in particular, the error code if getpwnam() issues one.
143 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
144 This is apparently needed in addition to the PH/07 change above to avoid
145 any possible encoding problems.
147 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
148 but not after initializing Perl.
150 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
151 output them only if debugging. By default they are written stderr,
152 apparently, which is not desirable.
154 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
157 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
160 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
162 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
163 authenticated or an ident call has been made. Suppress the default
164 values for $authenticated_id and $authenticated_sender (but permit -oMai
165 and -oMas) when testing with -bh.
167 PH/24 Re-jigged the order of the tests in the default configuration so that the
168 tests for valid domains and recipients precede the DNS black list and CSA
169 tests, on the grounds that those ones are more expensive.
171 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
172 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
173 This bug exists in every version of Exim that I still have, right back to
176 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
177 However, an attempt to turn on case-sensitivity in a regex key by
178 including (?-i) didn't work because the subject string was already
179 lowercased, and the effects were non-intuitive. It turns out that a
180 one-line patch can be used to allow (?-i) to work as expected.
186 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
187 other effects) broke the use of negated acl sub-conditions.
189 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
192 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
193 "Deny" causes Exim to reject the incoming connection with a 554 error.
194 Unfortunately, if there is a major crisis, such as a disk failure,
195 tcp-wrappers gives "deny", whereas what one would like would be some
196 kind of temporary error. A kludge has been added to help with this.
197 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
198 554 error is used if errno is still zero or contains ENOENT (which occurs
199 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
202 PH/03 Add -lutil to the default FreeBSD LIBS setting.
204 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
205 errors. Otherwise a message that provokes a temporary error (when other
206 messages do not) can cause a whole host to time out.
208 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
209 addresses were routed directly to files or pipes from a redirect router.
210 File deliveries just didn't batch; pipe deliveries might have suffered
213 PH/06 A failure to get a lock for a hints database would erroneously always say
214 "Failed to get write lock", even when it was really a read lock.
216 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
217 of 0600. This has been changed to use the value of the lockfile_mode
218 option (which defaults to 0600).
220 PH/08 Applied small patch from the Sieve maintainer.
222 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
223 folder from quota calculations, a direct delivery into this folder messed
224 up the contents of the maildirsize file. This was because the regex was
225 used only to exclude .Trash (or whatever) when the size of the mailbox
226 was calculated. There was no check that a delivery was happening into an
227 excluded directory. This bug has been fixed by ignoring all quota
228 processing for deliveries into excluded directories.
230 PH/10 Added the maildirfolder_create_regex option to appendfile.
236 PH/01 The code for finding all the local interface addresses on a FreeBSD
237 system running IPv6 was broken. This may well have applied to all BSD
238 systems, as well as to others that have similar system calls. The broken
239 code found IPv4 interfaces correctly, but gave incorrect values for the
240 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
241 that it would not match correctly against @[] and not recognize the IPv6
244 PH/02 The ipliteral router was not recognizing addresses of the form user@
245 [ipv6:....] because it didn't know about the "ipv6:" prefix.
247 PH/03 Added disable_ipv6.
249 PH/04 Changed $reply_address to use the raw form of the headers instead of the
250 decoded form, because it is most often used to construct To: headers
251 lines in autoreplies, and the decoded form may well be syntactically
252 invalid. However, $reply_address has leading white space removed, and all
253 newlines turned into spaces so that the autoreply transport does not
256 PH/05 If group was specified without a user on a router, and no group or user
257 was specified on a transport, the group from the router was ignored.
259 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
260 for visible compile-time settings that can be used to change these
261 numbers, for those that want even more. Backwards compatibility with old
262 spool files has been maintained. However, going back to a previous Exim
263 release will lost any variables that are in spool files.
265 PH/07 Two small changes when running in the test harness: increase delay when
266 passing a TCP/IP connection to a new process, in case the original
267 process has to generate a bounce, and remove special handling of
268 127.0.0.2 (sic), which is no longer necessary.
270 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
271 be the same on different OS.
273 PH/09 Moved a debug statement in filter processing to avoid a race problem when
276 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
277 whether --show-vars was specified or not
279 JJ/02 exipick: Added support for new ACL variable spool format introduced
282 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
283 syntactically invalid From: or Reply-to: line, and a filter used this to
284 generate an autoreply, and therefore failed to obtain an address for the
285 autoreply, Exim could try to deliver to a non-existent relative file
286 name, causing unrelated and misleading errors. What now happens is that
287 it logs this as a hard delivery error, but does not attempt to create a
290 PH/11 The exinext utility has a -C option for testing purposes, but although
291 the given file was scanned by exinext itself; it wasn't being passed on
292 when Exim was called.
294 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
295 an end-of-file indication when reading a command response.
297 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
298 compiled. In many other places in Exim, IPv6 addresses are always
299 recognized, so I have changed this. It also means that IPv4 domain
300 literals of the form [IPV4:n.n.n.n] are now always recognized.
302 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
303 used if the router is not running as root, for example, when verifying at
304 ACL time, or when using -bh. The debugging output from this situation was
305 non-existent - all you got was a failure to exec. I have made two
308 (a) Failures to set uid/gid, the current directory, or a process leader
309 in a subprocess such as that created by queryprogram now generate
310 suitable debugging ouput when -d is set.
312 (b) The queryprogram router detects when it is not running as root,
313 outputs suitable debugging information if -d is set, and then runs
314 the subprocess without attempting to change uid/gid.
316 PH/15 Minor change to Makefile for building test_host (undocumented testing
319 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
320 additional section of a DNS packet that returns MX or SRV records.
321 Instead, it always explicitly searches for A/AAAA records. This avoids
322 major problems that occur when a DNS server includes only records of one
323 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
324 fixed another bug: if SRV records were looked up and the corresponding
325 address records were *not* found in the additional section, the port
326 values from the SRV records were lost.
328 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
329 using the correct key (the original address) when searching the retry
330 rules in order to find which one to use for generating the retry hint.
332 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
333 adding the default one. Similarly, if it contains a Reply-To: header, the
334 errors_reply_to option, if set, is not used.
336 PH/19 When calculating a retry time, Exim used to measure the "time since
337 failure" by looking at the "first failed" field in the retry record. Now
338 it does not use this if it is later than than the arrival time of the
339 message. Instead it uses the arrival time. This makes for better
340 behaviour in cases where some deliveries succeed, thus re-setting the
341 "first failed" field. An example is a quota failure for a huge message
342 when small messages continue to be delivered. Without this change, the
343 "time since failure" will always be short, possible causing more frequent
344 delivery attempts for the huge message than are intended.
345 [Note: This change was subsequently modified - see PH/04 for 4.62.]
347 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
348 $1, $2, $3) because the numerical variables can be reset during some
349 expansion items (e.g. "match"), thereby losing the authentication data.
351 PH/21 Make -bV show the size of off_t variables so that the test suite can
352 decide whether to run tests for quotas > 2G.
354 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
355 mailbox_size, and mailbox_filecount in the appendfile transport. If a
356 filecount value is greater than 2G or if a quota value is greater than 2G
357 on a system where the size of off_t is not greater than 4, a panic error
360 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
361 never match. The debug and -bh output now contains an explicit error
362 message indicating a malformed IPv4 address or mask.
364 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
365 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
368 PH/25 Do not write to syslog when running in the test harness. The only
369 occasion when this arises is a failure to open the main or panic logs
370 (for which there is an explicit test).
372 PH/26 Added the /no_tell option to "control=freeze".
374 PH/27 If a host name lookup failed very early in a connection, for example, if
375 the IP address matched host_lookup and the reverse lookup yielded a name
376 that did not have a forward lookup, an error message of the form "no IP
377 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
378 could be logged. Now it outputs the IP address instead of "NULL".
380 PH/28 An enabling patch from MH: add new function child_open_exim2() which
381 allows the sender and the authenticated sender to be set when
382 submitting a message from within Exim. Since child_open_exim() is
383 documented for local_scan(), the new function should be too.
385 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
386 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
387 results in an empty string is now treated as unset.
389 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
391 PH/31 Added sender_verify_fail logging option.
393 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
394 needed by RFC 822 but not by RFC 2822 was commented out. I have now
395 tidied the source and removed it altogether.
397 PH/33 When a queue run was abandoned because the load average was too high, a
398 log line was always written; now it is written only if the queue_run log
399 selector is set. In addition, the log line for abandonment now contains
400 information about the queue run such as the pid. This is always present
401 in "start" and "stop" lines but was omitted from the "abandon" line.
403 PH/34 Omit spaces between a header name and the colon in the error message that
404 is given when verify = headers_syntax fails (if there are lots of them,
405 the message gets confusing).
407 PH/35 Change the default for dns_check_names_pattern to allow slashes within
408 names, as there are now some PTR records that contain slashes. This check
409 is only to protect against broken name servers that fall over on strange
410 characters, so the fact that it applies to all lookups doesn't matter.
412 PH/36 Now that the new test suite is complete, we can remove some of the
413 special code in Exim that was needed for the old test suite. For example,
414 sorting DNS records because real resolvers return them in an arbitrary
415 order. The new test suite's fake resolver always returns records in the
418 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
419 bounces) except when queue_only is set, to avoid logging races between
420 the different processes.
422 PH/38 Panic-die if .include specifies a non-absolute path.
424 PH/39 A tweak to the "H" retry rule from its user.
426 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
427 a label. They prevented compilation on older perls.
429 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
430 a warning to be raised on newish perls.
432 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
433 on queue. Changes to match documented behaviour of showing count of
434 messages matching specified criteria.
436 PH/40 Changed the default ident timeout from 30s to 5s.
438 PH/41 Added support for the use of login_cap features, on those BSD systems
439 that have them, for controlling the resources used by pipe deliveries.
441 PH/42 The content-scanning code uses fopen() to create files in which to put
442 message data. Previously it was not paying any attention to the mode of
443 the files. Exim runs with umask(0) because the rest of the code creates
444 files with open(), and sets the required mode explicitly. Thus, these
445 files were ending up world-writeable. This was not a big issue, because,
446 being within the spool directory, they were not world-accessible. I have
447 created a function called modefopen, which takes an additional mode
448 argument. It sets umask(777), creates the file, chmods it to the required
449 mode, then resets the umask. All the relevant calls to fopen() in the
450 content scanning code have been changed to use this function.
452 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
453 to 24 hours. This avoids potential overflow problems when processing G
454 and H retry rules. I suspect nobody ever tinkers with this value.
456 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
458 PH/45 When the plaintext authenticator is running as a client, the server's
459 challenges are checked to ensure they are valid base64 strings. By
460 default, the authentication attempt is cancelled if an invalid string is
461 received. Setting client_ignore_invalid_base64 true ignores these errors.
462 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
463 they are received. Thus, the responses can be made to depend on the
464 challenges. If an invalid string is ignored, an empty string is placed in
467 PH/46 Messages that are created by the autoreply transport now contains a
468 References: header, in accordance with RFCs 2822 and 3834.
470 PH/47 Added authenticated_sender_force to the smtp transport.
472 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
474 PH/49 Installed latest patch from the Sieve maintainer.
476 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
477 was also set, the appendfile transport was unnecessarily scanning a
478 directory of message files (e.g. for maildir delivery) to find the count
479 of files (along with the size), even though it did not need this
480 information. It now does the scan only if it needs to find either the
481 size of the count of files.
483 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
485 PH/52 Two bugs concerned with error handling when the smtp transport is
488 (i) Exim was not creating retry information for temporary errors given
489 for individual recipients after the DATA command when the smtp transport
490 was used in LMTP mode. This meant that they could be retried too
491 frequently, and not timed out correctly.
493 (ii) Exim was setting the flag that allows error details to be returned
494 for LMTP errors on RCPT commands, but not for LMTP errors for individual
495 recipients that were returned after the DATA command.
497 PH/53 This is related to PH/52, but is more general: for any failing address,
498 when detailed error information was permitted to be returned to the
499 sender, but the error was temporary, then after the final timeout, only
500 "retry timeout exceeded" was returned. Now it returns the full error as
501 well as "retry timeout exceeded".
503 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
504 do this, and (what is worse) MTAs that accept it.
506 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
507 will now be deprecated.
509 PH/56 New os.c-cygwin from the Cygwin maintainer.
511 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
512 formats (previously only available in exim formats via -bpr, -bpru,
513 and -bpra. Now also available in native and exiqgrep formats)
515 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
516 with very large, slow to parse queues
518 JJ/08 exipick: added ! as generic prefix to negate any criteria format
520 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
522 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
523 responses to authentication challenges, though it was showing the
524 challenges; (ii) I've removed the CR characters from the debug output for
527 PH/58 Allow for the insertion of a newline as well as a space when a string
528 is turned into more than one encoded-word during RFC 2047 encoding. The
529 Sieve code now uses this.
531 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
532 data_4xx, lost_connection, tls_required.
534 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
535 message was being sent as an SMTP response.
537 PH/61 Add -l and -k options to exicyclog.
539 PH/62 When verifying, if an address was redirected to one new address, so that
540 verification continued, and the new address failed or deferred after
541 having set something in $address_data, the value of $address_data was not
542 passed back to the ACL. This was different to the case when no
543 redirection occurred. The value is now passed back in both cases.
545 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
546 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
547 use login_cap.h, so on its own it isn't the distinguishing feature. The
548 new name refers directly to the setclassresources() function.
550 PH/65 Added configuration files for NetBSD3.
552 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
554 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
555 is preferred over IPv4.
557 PH/68 The bounce_return_message and bounce_return_body options were not being
558 honoured for bounces generated during the reception of non-SMTP messages.
559 In particular, this applied to messages rejected by the ACL. This bug has
560 been fixed. However, if bounce_return_message is true and bounce_return_
561 body is false, the headers that are returned for a non-SMTP message
562 include only those that have been read before the error was detected.
563 (In the case of an ACL rejection, they have all been read.)
565 PH/69 The HTML version of the specification is now built in a directory called
566 spec_html instead of spec.html, because the latter looks like a path with
567 a MIME-type, and this confuses some software.
569 PH/70 Catch two compiler warnings in sieve.c.
571 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
572 function verify_get_ident() calls ip_connect() to connect a socket, but
573 if the "connect()" function timed out, ip_connect() used to close the
574 socket. However, verify_get_ident() also closes the socket later, and in
575 between Exim writes to the log, which may get opened at this point. When
576 the socket was closed in ip_connect(), the log could get the same file
577 descriptor number as the socket. This naturally causes chaos. The fix is
578 not to close the socket in ip_connect(); the socket should be closed by
579 the function that creates it. There was only one place in the code where
580 this was missing, in the iplookup router, which I don't think anybody now
581 uses, but I've fixed it anyway.
583 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
584 well as to direct DNS lookups. Otherwise the handling of names in host
585 lists is inconsistent and therefore confusing.
591 PH/01 Two changes to the default runtime configuration:
593 (1) Move the checks for relay_from_hosts and authenticated clients from
594 after to before the (commented out) DNS black list checks.
596 (2) Add control=submission to the relay_from_hosts and authenticated
597 clients checks, on the grounds that messages accepted by these
598 statements are most likely to be submissions.
600 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
602 (1) Generate an error if the third argument for the ${prvs expansion is
605 (2) Treat a missing third argument of ${prvscheck as if it were an empty
608 (3) Reset the variables that are obtained from the first argument of
609 ${prvscheck and used in the second argument before leaving the code,
610 because their memory is reclaimed, so using them afterwards may do
613 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
614 one (it's much easier than Tom thought :-).
616 (5) Because of (4), we can now allow for the use of $prvscheck_result
617 inside the third argument.
619 PH/03 For some reason, the default setting of PATH when running a command from
620 a pipe transport was just "/usr/bin". I have changed it to
623 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
624 anything to be listed in the output from -bV.
626 PH/05 When a filter generated an autoreply, the entire To: header line was
627 quoted in the delivery log line, like this:
629 => >A.N.Other <ano@some.domain> <original@ddress> ...
631 This has been changed so that it extracts the operative address. There
632 may be more than one such address. If so, they are comma-separated, like
635 => >ano@some.domain,ona@other.domain <original@ddress> ...
637 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
638 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
639 not being looked up in the rDNS to get a host name, Exim was showing the
640 IP address twice in Received: lines, even though the IP addresses were
641 identical. For example:
643 Received: from [1.2.3.4] (helo=[1.2.3.4])
645 However, if the real host name was known, it was omitting the HELO data
646 if it matched the actual IP address. This has been tidied up so that it
647 doesn't show the same IP address twice.
649 PH/07 When both +timestamp and +memory debugging was on, the value given by
650 $tod_xxx expansions could be wrong, because the tod_stamp() function was
651 called by the debug printing, thereby overwriting the timestamp buffer.
652 Debugging no longer uses the tod_stamp() function when +timestamp is set.
654 PH/08 When the original message was included in an autoreply transport, it
655 always said "this is a copy of the message, including all the headers",
656 even if body_only or headers_only was set. It now gives an appropriate
659 PH/09 Applied a patch from the Sieve maintainer which:
661 o fixes some comments
662 o adds the (disabled) notify extension core
663 o adds some debug output for the result of if/elsif tests
664 o points to the current vacation draft in the documentation
665 and documents the missing references header update
669 o fixes a bug in processing the envelope test (when testing
670 multiple envelope elements, the last element determinted the
673 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
674 Electronic Mail") by including:
676 Auto-submitted: auto-generated
678 in the messages that it generates (bounce messages and others, such as
679 warnings). In the case of bounce messages for non-SMTP mesages, there was
680 also a typo: it was using "Auto_submitted" (underscore instead of
681 hyphen). Since every message generated by Exim is necessarily in response
682 to another message, thes have all been changed to:
684 Auto-Submitted: auto-replied
686 in accordance with these statements in the RFC:
688 The auto-replied keyword:
690 - SHOULD be used on messages sent in direct response to another
691 message by an automatic process,
693 - MUST NOT be used on manually-generated messages,
695 - MAY be used on Delivery Status Notifications (DSNs) and Message
696 Disposition Notifications (MDNs),
698 - MUST NOT be used on messages generated by automatic or periodic
699 processes, except for messages which are automatic responses to
702 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
703 to the default Received: header definition.
705 PH/12 Added log selector acl_warn_skipped (default on).
707 PH/13 After a successful wildlsearch lookup, discard the values of numeric
708 variables because (a) they are in the wrong storage pool and (b) even if
709 they were copied, it wouldn't work properly because of the caching.
711 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
712 checking when decoding. Apparently there are clients that generate
713 overlong encoded strings. Why am I not surprised?
715 PH/15 If the first argument of "${if match_address" was not empty, but did not
716 contain an "@" character, Exim crashed. Now it writes a panic log message
717 and treats the condition as false.
719 PH/16 In autoreply, treat an empty string for "once" the same as unset.
721 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
722 extension "envelope-auth". The code is finished and in agreement with
723 other implementations, but there is no documentation so far and in fact,
724 nobody wrote the draft yet. This extension is currently #undef'ed, thus
725 not changing the active code.
727 Print executed "if" and "elsif" statements when debugging is used. This
728 helps a great deal to understand what a filter does.
730 Document more things not specified clearly in RFC3028. I had all this
731 sorted out, when out of a sudden new issues came to my mind. Oops."
733 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
736 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
737 canonical form (as documented). However, after a host name lookup from
738 the IP address, check_host() was doing a simple string comparison with
739 addresses acquired from the DNS when checking that the found name did
740 have the original IP as one of its addresses. Since any found IPv6
741 addresses are likely to be in abbreviated form, the comparison could
742 fail. Luckily, there already exists a function for doing the comparison
743 by converting both addresses to binary, so now that is used instead of
746 PH/20 There was another similar case to PH/19, when a complete host name was
747 given in a host list; looking up its IP address could give an abbreviated
748 form, whereas the current host's name might or might not be abbreviated.
749 The same fix has been applied.
755 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
756 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
759 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
760 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
762 PH/03 Typo: missing ".o" in src/pcre/Makefile.
764 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
765 header line, restrict the check to what is listed in RFCs 2369 and 2929.
766 Also, for "Auto-Submitted", treat anything other than "no" as
767 non-personal, in accordance with RFC 3834. (Previously it treated
768 anything starting "auto-" as non-personal.)
770 TF/01 The control=submission/name=... option had a problem with syntax
771 errors if the name included a slash character. The /name= option
772 now slurps the rest of the string, so it can include any characters
773 but it must come last in the list of options (after /sender_retain
776 PH/05 Some modifications to the interface to the fake nameserver for the new
784 TK/01 Added the "success_on_redirect" address verification option. See
785 NewStuff for rationale and an example.
787 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
789 PH/02 Patch to exigrep to allow it to work on syslog lines.
791 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
792 fread() to skip over the body file's header line, because in Cygwin the
793 header line is locked and is inaccessible.
795 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
796 co-exist for some time) to make it clear that it is the Exim ID that is
797 referenced, not the Message-ID: header line.
799 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
800 string_format() function, because snprintf() does not exist on all
803 PH/06 The use of forbid_filter_existstest now also locks out the use of the
804 ${stat: expansion item.
806 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
807 protocol synchronization error", to keep the pedants happy.
809 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
810 well as for IRIX systems, when gcc is being used. See the host.c source
813 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
815 PH/10 Named domain lists were not working if used in a queue_smtp_domains
818 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
819 transport and to the smtp transport in LMTP mode.
821 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
823 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
824 run a filter in a subprocess. This could lead to confusion in subsequent
825 lookups in the parent process. There should also be a search_tidyup() at
826 the end of the subprocess.
828 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
829 only if the host matched helo_try_verify_hosts, which caused the
830 verification to occur when the EHLO/HELO command was issued. The ACL just
831 tested the remembered result. Now, if a previous verification attempt has
832 not happened, "verify = helo" does it there and then.
834 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
836 TK/03 Fix log output including CR from clamd.
838 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
839 exist provoked a memory error which could cause a segfault.
841 PH/15 Installed PCRE 6.2
843 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
845 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
846 of the problem. Specifically, suggested +O2 rather than +O1 for the
849 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
851 PH/20 If a delivery was routed to a non-standard port by means of an SRV
852 record, the port was not correctly logged when the outgoing_port log
853 selector was set (it logged the transort's default port).
855 PH/21 Added support for host-specific ports to manualroute, queryprogram,
856 fallback_hosts, and "hosts" in the smtp transport.
858 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
859 host errors such as "Connection refused".
861 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
862 authentication with radiusclient 0.4.9:
864 - Error returned from rc_read_config was caught wrongly
865 - Username/password not passed on to radius server due to wrong length.
867 The presumption is that some radiusclient API changes for 4.51/PH/17
868 were not taken care of correctly. The code is still untested by me (my
869 Linux distribution still has 0.3.2 of radiusclient), but it was
870 contributed by a Radius user.
872 PH/24 When doing a callout, the value of $domain wasn't set correctly when
873 expanding the "port" option of the smtp transport.
875 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
876 while reading a MIME header. Thanks to Tom Hughes for a patch.
878 PH/24 Include config.h inside local_scan.h so that configuration settings are
881 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
882 that in an ACL for RCPT (for example), you can examine exactly what was
885 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
886 commands, but it was not correctly comparing the address with the actual
887 client host address. Thus, it would show the EHLO address in Received:
888 header lines when this was not necessary.
890 PH/27 Added the % operator to ${eval:}.
892 PH/28 Exim tries to create and chdir to its spool directory when it starts;
893 it should be ignoring failures (because with -C, for example, it has lost
894 privilege). It wasn't ignoring creation failures other than "already
897 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
900 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
901 because an input line was too long, either on its own, or by virtue of
902 too many continuations, the temporary file was not being removed, and the
903 return code was incorrect.
905 PH/31 Missing "BOOL" in function definition in filtertest.c.
907 PH/32 Applied Sieve patches from the maintainer.
909 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
911 PH/33 Added "verify = not_blind".
913 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
914 Local/Makefile (with some defaults set). These are used in built scripts
915 such as exicyclog, but they have never been used in the exim_install
916 script (though there are many overriding facilities there). I have
917 arranged that the exim_install script now takes note of these two
920 PH/35 Installed configuration files for Dragonfly.
922 PH/36 When a locally submitted message by a trusted user did not contain a
923 From: header, and the sender address was obtained from -f or from an SMTP
924 MAIL command, and the trusted user did not use -F to supply a sender
925 name, $originator_name was incorrectly used when constructing a From:
926 header. Furthermore, $originator_name was used for submission mode
927 messages from external hosts without From: headers in a similar way,
928 which is clearly wrong.
930 PH/37 Added control=suppress_local_fixups.
932 PH/38 When log_selector = +received_sender was set, and the addition of the
933 sender made the log line's construction buffer exactly full, or one byte
934 less than full, an overflow happened when the terminating "\n" was
937 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
938 when the result of a list match is failure because a DNS lookup failed.
940 PH/40 RM_COMMAND is now used in the building process.
942 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
943 the "build-* directories that it finds.
945 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
946 address in a domain literal was a prefix of an interface address.
948 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
949 when verifying a sender address, unless rewrite_headers is false.
951 PH/44 Wrote a long comment about why errors_to addresses are verified as
952 recipients, not senders.
954 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
955 the ratelimit ACL was added.
957 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
959 PH/46 Added extra information about PostgreSQL errors to the error string.
961 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
962 avoiding the need to install special zones in a real server. This is
963 backwards compatible; if it can't find the fake resolver, it drops back.
964 Thus, both old and new test suites can be run.
966 TF/02 Added util/ratelimit.pl
968 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
969 clock is set back in time.
971 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
972 Candler <B.Candler@pobox.com>.
974 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
975 OK for addresses that are the result of redirections.
977 PH/48 A number of further additions for the benefit of the new test suite,
978 including a fake gethostbyname() that interfaces to the fake DNS resolver
981 TF/06 The fix for widen_domains has also been applied to qualify_single and
982 search_parents which are the other dnslookup options that can cause
985 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
988 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
990 TF/07 Exim produced the error message "an SRV record indicated no SMTP
991 service" if it encountered an MX record with an empty target hostname.
992 The message is now "an MX or SRV record indicated no SMTP service".
994 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
995 if the DNS of the sending site is misconfigured. This is quite a
996 common situation. This change restores the behaviour of treating a
997 helo verification defer as a failure.
999 PH/51 If self=fail was set on a router, the bounce message did not include the
1000 actual error message.
1006 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1008 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1009 command itself ended in error, the underlying message about the transport
1010 filter timeout was being overwritten with the pipe command error. Now the
1011 underlying error message should be appended to the second error message.
1013 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1015 PH/02 Reduce the amount of output that "make" produces by default. Full output
1016 can still be requested.
1018 PH/03 The warning log line about a condition test deferring for a "warn" verb
1019 was being output only once per connection, rather than after each
1020 occurrence (because it was using the same function as for successful
1021 "warn" verbs). This seems wrong, so I have changed it.
1023 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1024 it should not have, which might have caused a crash in the right
1025 circumstances, but probably never did.
1027 PH/04 Installed a modified version of Tony Finch's patch to make submission
1028 mode fix the return path as well as the Sender: header line, and to
1029 add a /name= option so that you can make the user's friendly name appear
1032 TF/03 Added the control = fakedefer ACL modifier.
1034 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1035 Mark Lowes for thorough testing.
1037 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1039 TK/03 Merged latest SRS patch from Miles Wilton.
1041 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1042 with the definition in sysexits.h (which is #included earlier).
1043 Fortunately, Exim does not actually use EX_OK. The code used to try to
1044 preserve the sysexits.h value, by assumimg that macro definitions were
1045 scanned for macro replacements. I have been disabused of this notion,
1046 so now the code just undefines EX_OK before #including unistd.h.
1048 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1049 in the smtp transport. When a block could not be written in a single
1050 write() function, the timeout was being re-applied to each part-write.
1051 This seems wrong - if the receiver was accepting one byte at a time it
1052 would take for ever. The timeout is now adjusted when this happens. It
1053 doesn't have to be particularly precise.
1055 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1056 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1058 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1059 without a domain if the check to <postmaster@domain> fails.
1061 SC/01 Eximstats: added -xls and the ability to specify output files
1062 (patch written by Frank Heydlauf).
1064 SC/02 Eximstats: use FileHandles for outputing results.
1066 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1068 SC/04 Eximstats: fixed display of large numbers with -nvr option
1070 SC/05 Eximstats: fixed merging of reports with empty tables.
1072 SC/06 Eximstats: added the -include_original_destination flag
1074 SC/07 Eximstats: removed tabs and trailing whitespace.
1076 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1078 TK/06 MBOX spool code: Add real "From " MBOX separator line
1079 so the .eml file is really in mbox format (even though
1080 most programs do not really care). Patch from Alex Miller.
1082 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1083 The latter is generated from $received_to and is only set if the
1084 message has one envelope recipient. SA can use these headers,
1085 obviously out-of-the-box. Patch from Alex Miller.
1087 PH/08 The ${def test on a variable was returning false if the variable's
1088 value was "0", contrary to what the specification has always said!
1089 The result should be true unless the variable is empty.
1091 PH/09 The syntax error of a character other than { following "${if
1092 def:variable_name" (after optional whitespace) was not being diagnosed.
1093 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1094 accidental colon was present, for example, could give incorrect results.
1096 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1097 result is used (not including appendfile, where other changes are about
1100 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1101 This involved changing a lot of size variables from int to off_t. It
1102 should work with maildirs and everything.
1104 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1105 spamd dying while we are connected to it.
1107 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1110 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1111 function for BDB 4.3.
1113 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1115 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1116 the sources for PCRE have been re-arranged and more files are now
1119 PH/15 The code I had for printing potentially long long variables in PH/11
1120 above was not the best (it lost precision). The length of off_t variables
1121 is now inspected at build time, and an appropriate printing format (%ld
1122 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1123 to be "long long int" or "long int". This is needed for the internal
1124 formatting function string_vformat().
1126 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1127 the configuration file to be ":syslog", then the script "guesses" where
1128 the logs files are, rather than using the compiled in default. In our
1129 case the guess is not the same as the compiled default, so the script
1130 suddenly stopped working when I started to use syslog. The patch checks
1131 to see if log_file_path is "". If so, it attempts to read it from exim
1132 with no configuration file to get the compiled in version, before it
1133 falls back to the previous guessing code."
1135 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1136 implementing BATV in an Exim configuration. See NewStuff for the gory
1139 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1140 Makefile that are specific to HP-UX.
1142 PH/18 If the "use_postmaster" option was set for a recipient callout together
1143 with the "random" option, the postmaster address was used as the MAIL
1144 FROM address for the random test, but not for the subsequent recipient
1145 test. It is now used for both.
1147 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1148 patch removes a few documentation additions to RFC 3028, because the
1149 latest draft now contains them. It adds the new en;ascii-case comparator
1150 and a new error check for 8bit text in MIME parts. Comparator and
1151 require names are now matched exactly. I enabled the subaddress
1152 extension, but it is not well tested yet (read: it works for me)."
1154 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1155 rework some of the code of TK/09 above to avoid the hardwired use of
1156 "%lld" and "long long". Replaced the call to snprintf() with a call to
1159 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1160 records point to non-existent hosts", "retry timeout exceeded", and
1161 "retry time not reached for any host after a long failure period".
1163 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1164 experimental DomainKeys support:
1166 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1167 (2) On an error such as an illegally used "control", the wrong name for
1168 the control was given.
1170 These problems did NOT occur unless DomainKeys support was compiled.
1172 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1174 PH/24 Added ${if match_ip condition.
1176 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1177 databases so that it will be absolutely obvious if a crash occurs in the
1178 DB library. This is a regular occurrence (often caused by mis-matched
1181 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1182 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1183 picked up on a user's system that detects such things. There doesn't seem
1184 to be a gcc warning option for this - only an attribute that has to be
1185 put on the function's prototype. It seems that in Fedora Core 4 they have
1186 set this on a number of new functions. No doubt there will be more in due
1189 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1190 specify a transport. However, if an address that was verified by such a
1191 router was the subject of a callout, Exim crashed because it tried to
1192 read the rcpt_include_affixes from the non-existent transport. Now it
1193 just assumes that the setting of that option is false. This bug was
1194 introduced by 4.51/PH/31.
1196 PH/28 Changed -d+all to exclude +memory, because that information is very
1197 rarely of interest, but it makes the output a lot bigger. People tend to
1198 do -d+all out of habit.
1200 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1201 code in os-type was giving problems when libc.so lives in lib64, like on
1204 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1205 aren't the modern standard, and it seems that some systems' include files
1206 don't always have them. Exim was already checking for some of the newer
1207 ones like T_AAAA, and defining it itself. I've added checks for all the
1208 record types that Exim uses.
1210 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1211 not automatically generating a new one, as it is supposed to. This
1212 prevented TLS from working. If the file did exist, but contained invalid
1213 data, a new version was generated, as expected. It was only the case of a
1214 non-existent file that was broken.
1216 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1217 with a change in libdomainkeys > 0.64.
1219 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1220 from DNS. If the selector record carries the flag, it now has
1221 precedence over the domain-wide flag.
1223 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1225 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1226 the use of a port name that isn't defined in /etc/services) occurred, the
1227 message was deferred as in a normal delivery, and thus remained on the
1228 spool, instead of being failed because of the mua_wrapper setting. This
1229 is now fixed, and I tidied up some of the mua_wrapper messages at the
1232 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1233 the messages in a hash of arrays rather than using individual hashes.
1234 This is a bit cleaner and results in dramatic memory savings, albeit
1235 at a slight CPU cost.
1237 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1238 as requested by Marc Sherman.
1240 SC/10 Eximstats: added histograms for user specified patterns as requested
1243 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1245 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1246 fopen() in the content-scanning modules that did not already have it.
1252 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1253 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1255 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1257 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1259 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1260 to test Sieve filters that use "vacation".
1262 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1263 that changes the way the GnuTLS parameters are stored in the cache file.
1264 The new format can be generated externally. For backward compatibility,
1265 if the data in the cache doesn't make sense, Exim assumes it has read an
1266 old-format file, and it generates new data and writes a new file. This
1267 means that you can't go back to an older release without removing the
1270 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1271 work if there are any delivery delays because "one_time" forces the
1272 parent to be marked "delivered", so its unseen clone is never tried
1273 again. For this reason, Exim now forbids the simultaneous setting of
1276 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1277 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1278 entry if you want to know the details. The fix, however, neglected to
1279 consider the case when local delivery batching is involved. The test for
1280 "previously delivered" was not happening when checking to see if an
1281 address could be batched with a previous (undelivered) one; under
1282 certain circumstances this could lead to multiple deliveries to the same
1285 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1286 in its include files, and this causes problems building Exim.
1288 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1289 header_syntax) but Exim was just ignoring anything given after a slash.
1290 In particular, this caused confusion with an attempt to use "verify =
1291 reverse_host_lookup/defer_ok". An error is now given when options are
1292 supplied for verify items that do not have them. (Maybe reverse_host_
1293 lookup should have a defer_ok option, but that's a different point.)
1295 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1296 defined by RFC 821) to 2048, because there were problems with some AUTH
1297 commands, and RFC 1869 says the size should be increased for extended
1298 SMTP commands that take arguments.
1300 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1303 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1304 "unknown" error; now it says that the functionality isn't in the binary.
1306 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1307 an address' error message when a string expansion fails (syntax or
1308 whatever). Otherwise the password may appear in the log. Following change
1309 PH/42 below, there is no longer a chance of it appearing in a bounce
1312 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1314 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1315 of its addresses was ever tried. (Bugzilla bug #2.)
1317 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1318 the result incorrectly in the debug output. (It correctly added a newline
1319 to what was transported.)
1321 TF/01 Added $received_time.
1323 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1324 commented out examples of how to interface to a virus scanner and to
1325 SpamAssassin. Also added commented examples of av_scanner and
1326 spamd_address settings.
1328 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1329 and controls are allowed in which ACLs. There were a couple of minor
1330 errors. Some of the entries in the conditions table (which is a table of
1331 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1332 negation of where the condition IS allowed.
1334 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1336 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1337 header file does not have a version number, so I've had to invent a new
1338 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1339 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1340 radiusclient), but it was contributed by a Radius user.
1342 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1343 files or directories, for OpenSSL.
1345 PH/19 When an Exim process that is running as root has to create an Exim log
1346 file, it does so in a subprocess that runs as exim:exim so as to get the
1347 ownership right at creation (otherwise, other Exim processes might see
1348 the file with the wrong ownership). There was no test for failure of this
1349 fork() call, which would lead to the process getting stuck as it waited
1350 for a non-existent subprocess. Forks do occasionally fail when resources
1351 run out. I reviewed all the other calls to fork(); they all seem to check
1354 PH/20 When checking for unexpected SMTP input at connect time (before writing
1355 the banner), Exim was not dealing correctly with a non-positive return
1356 from the read() function. If the client had disconnected by this time,
1357 the result was a log entry for a synchronization error with an empty
1358 string after "input=" when read() returned zero. If read() returned -1
1359 (an event I could not check), uninitialized data bytes were printed.
1360 There were reports of junk text (parts of files, etc) appearing after
1363 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1365 PH/22 Added support for macro redefinition, and (re)definition in between
1366 driver and ACL definitions.
1368 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1369 forgetting to use the resulting value; it was using the unexpanded value.
1371 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1372 hadn't been configured. The fix is from Juergen Kreileder, who
1373 understands it better than I do:
1375 "Here's what I see happening with three configured cyrus_sasl
1376 authenticators configured (plain, login, cram-md5):
1378 On startup auth_cyrus_sasl_init() gets called for each of these.
1379 This means three calls to sasl_listmech() without a specified mech_list.
1380 => SASL tests which mechs of all available mechs actually work
1381 => three warnings about OTP not working
1382 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1384 With the patch, sasl_listmech() also gets called three times. But now
1385 SASL's mech_list option is set to the server_mech specified in the the
1386 authenticator. Or in other words, the answer from sasl_listmech()
1387 gets limited to just the mech you're testing for (which is different
1389 => the return list contains just 'plain' or 'login', 'cram-md5' or
1390 nothing depending on the value of ob->server_mech.
1392 I've just tested the patch: Authentication still works fine,
1393 unavailable mechs specified in the exim configuration are still
1394 caught, and the auth.log warnings about OTP are gone."
1396 PH/25 When debugging is enabled, the contents of the command line are added
1397 to the debugging output, even when log_selector=+arguments is not
1400 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1401 answer is "GNU", and only if the return is "GNU/something" is the answer
1404 PH/27 $acl_verify_message is now set immediately after the failure of a
1405 verification in an ACL, and so is available in subsequent modifiers. In
1406 particular, the message can be preserved by coding like this:
1408 warn !verify = sender
1409 set acl_m0 = $acl_verify_message
1411 Previously, $acl_verify_message was set only while expanding "message"
1412 and "log_message" when a very denied access.
1414 PH/28 Modified OS/os.c-Linux with
1416 -#ifndef OS_LOAD_AVERAGE
1417 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1419 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1420 nomenclature these days.)
1422 PH/29 Installed patch from the Sieve maintainer that adds the options
1423 sieve_useraddress and sieve_subaddress to the redirect router.
1425 PH/30 In these circumstances:
1426 . Two addresses routed to the same list of hosts;
1427 . First host does not offer TLS;
1428 . First host accepts first address;
1429 . First host gives temporary error to second address;
1430 . Second host offers TLS and a TLS session is established;
1431 . Second host accepts second address.
1432 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1433 and peerdn, if requested) that were in fact used only for the second
1436 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1437 attention to any local part prefix or suffix that was matched by the
1438 router that accepted the address. It now behaves in the same way as it
1439 does for delivery: the affixes are removed from the local part unless
1440 rcpt_include_affixes is set on the transport.
1442 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1443 timeout during the DATA phase of an incoming message.
1445 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1446 applied a patch sanctioned by the Sieve maintainer.
1448 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1449 the uid or gid is negative. A case of a negative gid caused this to be
1450 noticed. The fix allows for either to be negative.
1452 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1453 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1456 PH/36 The change PH/12 above was broken. Fixed it.
1458 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1459 the grounds that routing the same address twice would always produce the
1460 same answer. This might have been true once, but it is certainly no
1461 longer true now. Routing a child address may depend on the previous
1462 routing that produced that child. Some complicated redirection strategies
1463 went wrong when messages had multiple recipients, and made Exim's
1464 behaviour dependent on the order in which the addresses were given.
1466 I have moved the duplicate checking until after the routing is complete.
1467 Exim scans the addresses that are assigned to local and remote
1468 transports, and removes any duplicates. This means that more work will be
1469 done, as duplicates will always all be routed, but duplicates are
1470 presumably rare, so I don't expect this is of any significance.
1472 For deliveries to pipes, files, and autoreplies, the duplicate checking
1473 still happens during the routing process, since they are not going to be
1476 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1477 It corrects a timeout issue with spamd. This is Ian's comment: "The
1478 background is that sometimes spamd either never reads data from a
1479 connection it has accepted, or it never writes response data. The exiscan
1480 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1481 blindly assumes that writes won't block so it may never time out."
1483 PH/39 Allow G after quota size as well as K and M.
1485 PH/40 The value set for $authenticated_id in an authenticator may not contain
1486 binary zeroes or newlines because the value is written to log lines and
1487 to spool files. There was no check on this. Now the value is run through
1488 the string_printing() function so that such characters are converted to
1489 printable escape sequences.
1491 PH/41 $message_linecount is a new variable that contains the total number of
1492 lines in the message. Compare $body_linecount, which is the count for the
1495 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1496 bounce and delay warning messages, except in certain special cases, which
1499 (a) An SMTP error message from a remote host;
1500 (b) A message specified in a :fail: redirection;
1501 (c) A message specified in a "fail" command in a system filter;
1502 (d) A message specified in a FAIL return from the queryprogram router;
1503 (e) A message specified by the cannot_route_message router option.
1505 In these cases only, Exim does include the error details in bounce and
1506 warning messages. There are also a few cases where bland messages such
1507 as "unrouteable address" or "local delivery error" are given.
1509 PH/43 $value is now also set for the "else" part of a ${run expansion.
1511 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1512 being worked on, but at least Exim now implements the latest version to
1515 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1516 process to complete was treated as a delivery failure, a timeout while
1517 writing the message to the pipe was logged, but erroneously treated as a
1518 successful delivery. Such timeouts include transport filter timeouts. For
1519 consistency with the overall process timeout, these timeouts are now
1520 treated as errors, giving rise to delivery failures by default. However,
1521 there is now a new Boolean option for the pipe transport called
1522 timeout_defer, which, if set TRUE, converts the failures into defers for
1523 both kinds of timeout. A transport filter timeout is now identified in
1526 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1527 systems where "make" and "gmake" are different, calling "gmake" at top
1528 level broke things. I've arranged for the value of $(MAKE) to be passed
1529 from the Makefile to this script so that it can call the same version of
1533 A note about Exim versions 4.44 and 4.50
1534 ----------------------------------------
1536 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1537 changes of various kinds. As a consequence, a big documentation update was
1538 needed. This delayed the release for rather longer than seemed good, especially
1539 in the light of a couple of (minor) security issues. Therefore, the changes
1540 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1541 release. So 4.44 and 4.50 are in effect two different branches that both start
1544 I have left the 4.50 change log unchanged; it contains all the changes since
1545 4.43. The change log for 4.44 is below; many of its items are identical to
1546 those for 4.50. This seems to be the most sensible way to preserve the
1547 historical information.
1553 1. Minor wording change to the doc/README.SIEVE file.
1555 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1556 computation of the current number of files was incorrect.
1558 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1559 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1562 4. Give more explanation in the error message when the command for a transport
1563 filter fails to execute.
1565 5. There are several places where Exim runs a non-Exim command in a
1566 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1567 was being done only for the command run by the queryprogram router. It is
1568 now done for all such subprocesses. The other cases are: ${run, transport
1569 filters, and the commands run by the lmtp and pipe transports.
1571 6. Added CONFIGURE_GROUP build-time option.
1573 7. Some older OS have a limit of 256 on the maximum number of file
1574 descriptors. Exim was using setrlimit() to set 1000 as a large value
1575 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1576 systems. I've change it so that if it can't get 1000, it tries for 256.
1578 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1579 was an oversight, and furthermore, ever since the addition of extra
1580 controls (e.g. 4.43/32), the checks on when to allow different forms of
1581 "control" were broken. There should now be diagnostics for all cases when a
1582 control that does not make sense is encountered.
1584 9. Added the /retain_sender option to "control=submission".
1586 10. $recipients is now available in the predata ACL (oversight).
1588 11. Tidy the search cache before the fork to do a delivery from a message
1589 received from the command line. Otherwise the child will trigger a lookup
1590 failure and thereby defer the delivery if it tries to use (for example) a
1591 cached ldap connection that the parent has called unbind on.
1593 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1594 of $address_data from the recipient verification was clobbered by the
1595 sender verification.
1597 13. The value of address_data from a sender verification is now available in
1598 $sender_address_data in subsequent conditions in the ACL statement.
1600 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1602 15. Added a new option "connect=<time>" to callout options, to set a different
1605 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1606 was its contents. (It was OK if the option was not defined at all.)
1608 17. A "Completed" log line is now written for messages that are removed from
1609 the spool by the -Mrm option.
1611 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1612 information about exactly what failed.
1614 19. Added -dd to debug only the daemon process.
1616 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1617 handles timeouts, both on the server side and network timeouts. Renamed the
1618 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1620 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1621 It is now set to "smtps".
1623 22. $host_address is now set to the target address during the checking of
1624 ignore_target_hosts.
1626 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1627 being passed; this would have caused $sender_host_name to have been used if
1628 matching the list had actually called for a host name (not very likely,
1629 since this list is usually IP addresses). A host name is now passed as
1632 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1633 code that specifies a non-restarting handler (typically sigaction() in
1634 modern systems) in an attempt to fix a rare and obscure crash bug.
1636 25. Narrowed the window for a race in the daemon that could cause it to ignore
1637 SIGCHLD signals. This is not a major problem, because they are used only to
1638 wake it up if nothing else does.
1640 26. A malformed maildirsize file could cause Exim to calculate negative values
1641 for the mailbox size or file count. Odd effects could occur as a result.
1642 The maildirsize information is now recalculated if the size or filecount
1645 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1646 support for a long time. Removed HAVE_SYS_VFS_H.
1648 28. Installed the latest version of exipick from John Jetmore.
1650 29. In an address list, if the pattern was not a regular expression, an empty
1651 subject address (from a bounce message) matched only if the pattern was an
1652 empty string. Non-empty patterns were not even tested. This was the wrong
1653 because it is perfectly reasonable to use an empty address as part of a
1654 database query. An empty address is now tested by patterns that are
1655 lookups. However, all the other forms of pattern expect the subject to
1656 contain a local part and a domain, and therefore, for them, an empty
1657 address still always fails if the pattern is not itself empty.
1659 30. Exim went into a mad DNS loop when attempting to do a callout where the
1660 host was specified on an smtp transport, and looking it up yielded more
1661 than one IP address.
1663 31. Re-factored the code for checking spool and log partition space into a
1664 function that finds that data and another that does the check. The former
1665 is then used to implement four new variables: $spool_space, $log_space,
1666 $spool_inodes, and $log_inodes.
1668 32. The RFC2047 encoding function was originally intended for short strings
1669 such as real names; it was not keeping to the 75-character limit for
1670 encoded words that the RFC imposes. It now respects the limit, and
1671 generates multiple encoded words if necessary. To be on the safe side, I
1672 have increased the buffer size for the ${rfc2047: expansion operator from
1675 33. It is now permitted to omit both strings after an "if" condition; if the
1676 condition is true, the result is "true". As before, when the second string
1677 is omitted, a false condition yields an empty string. This makes it less
1678 cumbersome to write custom ACL and router conditions.
1680 34. Failure to deliver a bounce message always caused it to be frozen, even if
1681 there was an errors_to setting on the router. The errors_to setting is now
1684 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1685 canonical form (fully expanded) before being placed in
1686 $sender_host_address.
1688 36. The table in the code that translates DNS record types into text (T_A to
1689 "A" for instance) was missing entries for NS and CNAME. It is just possible
1690 that this could have caused confusion if both these types were looked up
1691 for the same domain, because the text type is used as part of Exim's
1692 per-process caching. But the chance of anyone hitting this buglet seems
1695 37. The dnsdb lookup has been extended in a number of ways.
1697 (1) There is a new type, "zns", which walks up the domain tree until it
1698 finds some nameserver records. It should be used with care.
1700 (2) There is a new type, "mxh", which is like "mx" except that it returns
1701 just the host names, not the priorities.
1703 (3) It is now possible to give a list of domains (or IP addresses) to be
1704 looked up. The behaviour when one of the lookups defers can be
1705 controlled by a keyword.
1707 (4) It is now possible to specify the separator character for use when
1708 multiple records are returned.
1710 38. The dnslists ACL condition has been extended: it is now possible to supply
1711 a list of IP addresses and/or domains to be looked up in a particular DNS
1714 39. Added log_selector=+queue_time_overall.
1716 40. When running the queue in the test harness, wait just a tad after forking a
1717 delivery process, to get repeatability of debugging output.
1719 41. Include certificate and key file names in error message when GnuTLS fails
1720 to set them up, because the GnuTLS error message doesn't include the name
1721 of the failing file when there is a problem reading it.
1723 42. Allow both -bf and -bF in the same test run.
1725 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1727 44. The "Exiscan patch" is now merged into the mainline Exim source.
1729 45. Sometimes the final signoff response after QUIT could fail to get
1730 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1731 < 0 before doing a fflush(). This bug looks as though it goes back to the
1732 introduction of TLS in release 3.20, but "sometimes" must have been rare
1733 because the tests only now provoked it.
1735 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1736 (this can affect the format of dates).
1738 47. exim_tidydb, when checking for the continued existence of a message for
1739 which it has found a message-specific retry record, was not finding
1740 messages that were in split spool directories. Consequently, it was
1741 deleting retry records that should have stayed in existence.
1743 48. Steve fixed some bugs in eximstats.
1745 49. The SPA authentication driver was not abandoning authentication and moving
1746 on to the next authenticator when an expansion was forced to fail,
1747 contradicting the general specification for all authenticators. Instead it
1748 was generating a temporary error. It now behaves as specified.
1750 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1751 (the order specifies the preference for clients). The order is now AES256,
1752 AES128, 3DES, ARCFOUR128.
1754 51. Small patch to Sieve code - explicitly set From: when generating an
1757 52. Exim crashed if a remote delivery caused a very long error message to be
1758 recorded - for instance if somebody sent an entire SpamAssassin report back
1759 as a large number of 550 error lines. This bug was coincidentally fixed by
1760 increasing the size of one of Exim's internal buffers (big_buffer) that
1761 happened as part of the Exiscan merge. However, to be on the safe side, I
1762 have made the code more robust (and fixed the comments that describe what
1765 53. Now that there can be additional text after "Completed" in log lines (if
1766 the queue_time_overall log selector is set), a one-byte patch to exigrep
1767 was needed to allow it to recognize "Completed" as not the last thing in
1770 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1771 patch that reportedly fixes this has been added. I am not expert enough to
1772 create a test for it. This is what the patch creator wrote:
1774 "I found a little strange behaviour of ldap code when working with
1775 Windows 2003 AD Domain, where users was placed in more than one
1776 Organization Units. When I tried to give exim partial DN, the exit code
1777 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1778 But simultaneously result of request was absolutely normal ldap result,
1779 so I produce this patch..."
1781 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1782 REFERENCE, so I have modified the code to exclude the patch when that macro
1785 55. Some experimental protocols are using DNS PTR records for new purposes. The
1786 keys for these records are domain names, not reversed IP addresses. The
1787 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1788 leaves it alone. Component reversal etc. now happens only for IP addresses.
1791 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1793 57. Double the size of the debug message buffer (to 2048) so that more of very
1794 long debug lines gets shown.
1796 58. The exicyclog utility now does better if the number of log files to keep
1797 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1799 59. Two changes related to the smtp_active_hostname option:
1801 (1) $smtp_active_hostname is now available as a variable.
1802 (2) The default for smtp_banner uses $smtp_active_hostname instead
1803 of $primary_hostname.
1805 60. The host_aton() function is supposed to be passed a string that is known
1806 to be a valid IP address. However, in the case of IPv6 addresses, it was
1807 not checking this. This is a hostage to fortune. Exim now panics and dies
1808 if the condition is not met. A case was found where this could be provoked
1809 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1810 components; fortuitously, this particular loophole had already been fixed
1811 by change 4.50/55 above.
1813 If there are any other similar loopholes, the new check in host_aton()
1814 itself should stop them being exploited. The report I received stated that
1815 data on the command line could provoke the exploit when Exim was running as
1816 exim, but did not say which command line option was involved. All I could
1817 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1818 running as the user.
1821 61. There was a buffer overflow vulnerability in the SPA authentication code
1822 (which came originally from the Samba project). I have added a test to the
1823 spa_base64_to_bits() function which I hope fixes it.
1826 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1827 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1829 63. The daemon start-up calls getloadavg() while still root for those OS that
1830 need the first call to be done as root, but it missed one case: when
1831 deliver_queue_load_max is set with deliver_drop_privilege. This is
1832 necessary for the benefit of the queue runner, because there is no re-exec
1833 when deliver_drop_privilege is set.
1835 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1836 This has been fixed.
1838 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1839 was in use, was not putting the data itself into the right store pool;
1840 consequently, it could be overwritten for a subsequent message in the same
1841 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1844 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1846 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1847 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1848 the function were treating the return as a boolean value, which happened to
1849 work because 0=false and not-0=true, but is not correct code.
1851 68. The host_aton() function was not handling scoped IPv6 addresses (those
1852 with, for example, "%eth0" on the end) correctly.
1854 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1855 negated items (that is, ~something) in unsigned ints. Some compilers
1856 apparently mutter when there is no cast.
1858 70. If an address verification called from an ACL failed, and did not produce a
1859 user-specific message (i.e. there was only a "system" message), nothing was
1860 put in $acl_verify_message. In this situation, it now puts the system
1863 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1864 session; change 4.31/43 added the unwanted input to the log line - except
1865 that it did not do this in the start of session case. It now does.
1867 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1868 This is wrong and can cause the other end to generate a synchronization
1869 error if it is another Exim or anything else that does the synchronization
1870 check. A QUIT command is no longer sent after a timeout.
1872 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1873 during host lookups.
1875 74. The defer_ok option of callout verification was not working if it was used
1876 when verifying addresses in header lines, that is, for this case:
1878 verify = header_sender/callout=defer_ok
1880 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1881 those file descriptors could be used for SMTP connections. If anything
1882 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1883 could be sent to the SMTP client, causing chaos. The daemon now opens
1884 stdin, stdout, and stderr to /dev/null when it puts itself into the
1887 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1888 log by default. The user can override this with suitable Perl magic.
1890 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1891 the log message when discard triggers, was not working for the DATA ACL or
1892 for the non-SMTP ACL.
1894 78. Error message wording change in sieve.c.
1896 79. If smtp_accept_max_per_host was set, the number of connections could be
1897 restricted to fewer than expected, because the daemon was trying to set up
1898 a new connection before checking whether the processes handling previous
1899 connections had finished. The check for completed processes is now done
1900 earlier. On busy systems, this bug wouldn't be noticed because something
1901 else would have woken the daemon, and it would have reaped the completed
1904 80. If a message was submitted locally by a user whose login name contained one
1905 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1906 It caused a spool format error. I have fixed the spool reading code. A
1907 related problem was that the "from" clause in the Received: line became
1908 illegal because of the space(s). It is now covered by ${quote_local_part.
1910 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1913 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1915 83. Patch from Sieve maintainer to fix unterminated string problem in
1916 "vacation" handling.
1918 84. Some minor changes to the Linux configuration files to help with other
1919 OS variants using glibc.
1921 85. One more patch for Sieve to update vacation handling to latest spec.
1924 ----------------------------------------------------
1925 See the note above about the 4.44 and 4.50 releases.
1926 ----------------------------------------------------
1932 1. Change 4.43/35 introduced a bug that caused file counts to be
1933 incorrectly computed when quota_filecount was set in an appendfile
1936 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1937 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1940 3. Give more explanation in the error message when the command for a transport
1941 filter fails to execute.
1943 4. There are several places where Exim runs a non-Exim command in a
1944 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1945 was being done only for the command run by the queryprogram router. It is
1946 now done for all such subprocesses. The other cases are: ${run, transport
1947 filters, and the commands run by the lmtp and pipe transports.
1949 5. Some older OS have a limit of 256 on the maximum number of file
1950 descriptors. Exim was using setrlimit() to set 1000 as a large value
1951 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1952 systems. I've change it so that if it can't get 1000, it tries for 256.
1954 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1955 was an oversight, and furthermore, ever since the addition of extra
1956 controls (e.g. 4.43/32), the checks on when to allow different forms of
1957 "control" were broken. There should now be diagnostics for all cases when a
1958 control that does not make sense is encountered.
1960 7. $recipients is now available in the predata ACL (oversight).
1962 8. Tidy the search cache before the fork to do a delivery from a message
1963 received from the command line. Otherwise the child will trigger a lookup
1964 failure and thereby defer the delivery if it tries to use (for example) a
1965 cached ldap connection that the parent has called unbind on.
1967 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1968 of $address_data from the recipient verification was clobbered by the
1969 sender verification.
1971 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1972 was its contents. (It was OK if the option was not defined at all.)
1974 11. A "Completed" log line is now written for messages that are removed from
1975 the spool by the -Mrm option.
1977 12. $host_address is now set to the target address during the checking of
1978 ignore_target_hosts.
1980 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1981 being passed; this would have caused $sender_host_name to have been used if
1982 matching the list had actually called for a host name (not very likely,
1983 since this list is usually IP addresses). A host name is now passed as
1986 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1987 code that specifies a non-restarting handler (typically sigaction() in
1988 modern systems) in an attempt to fix a rare and obscure crash bug.
1990 15. Narrowed the window for a race in the daemon that could cause it to ignore
1991 SIGCHLD signals. This is not a major problem, because they are used only to
1992 wake it up if nothing else does.
1994 16. A malformed maildirsize file could cause Exim to calculate negative values
1995 for the mailbox size or file count. Odd effects could occur as a result.
1996 The maildirsize information is now recalculated if the size or filecount
1999 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2000 support for a long time. Removed HAVE_SYS_VFS_H.
2002 18. Updated exipick to current release from John Jetmore.
2004 19. Allow an empty sender to be matched against a lookup in an address list.
2005 Previously the only cases considered were a regular expression, or an
2008 20. Exim went into a mad DNS lookup loop when doing a callout where the
2009 host was specified on the transport, if the DNS lookup yielded more than
2012 21. The RFC2047 encoding function was originally intended for short strings
2013 such as real names; it was not keeping to the 75-character limit for
2014 encoded words that the RFC imposes. It now respects the limit, and
2015 generates multiple encoded words if necessary. To be on the safe side, I
2016 have increased the buffer size for the ${rfc2047: expansion operator from
2019 22. Failure to deliver a bounce message always caused it to be frozen, even if
2020 there was an errors_to setting on the router. The errors_to setting is now
2023 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2024 canonical form (fully expanded) before being placed in
2025 $sender_host_address.
2027 24. Updated eximstats to version 1.33
2029 25. Include certificate and key file names in error message when GnuTLS fails
2030 to set them up, because the GnuTLS error message doesn't include the name
2031 of the failing file when there is a problem reading it.
2033 26. Expand error message when OpenSSL has problems setting up cert/key files.
2036 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2037 (this can affect the format of dates).
2039 28. exim_tidydb, when checking for the continued existence of a message for
2040 which it has found a message-specific retry record, was not finding
2041 messages that were in split spool directories. Consequently, it was
2042 deleting retry records that should have stayed in existence.
2044 29. eximstats updated to version 1.35
2045 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2046 1.35 - bugfix such that pie charts by volume are generated correctly
2048 30. The SPA authentication driver was not abandoning authentication and moving
2049 on to the next authenticator when an expansion was forced to fail,
2050 contradicting the general specification for all authenticators. Instead it
2051 was generating a temporary error. It now behaves as specified.
2053 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2054 (the order specifies the preference for clients). The order is now AES256,
2055 AES128, 3DES, ARCFOUR128.
2057 31. Small patch to Sieve code - explicitly set From: when generating an
2060 32. Exim crashed if a remote delivery caused a very long error message to be
2061 recorded - for instance if somebody sent an entire SpamAssassin report back
2062 as a large number of 550 error lines. This bug was coincidentally fixed by
2063 increasing the size of one of Exim's internal buffers (big_buffer) that
2064 happened as part of the Exiscan merge. However, to be on the safe side, I
2065 have made the code more robust (and fixed the comments that describe what
2068 33. Some experimental protocols are using DNS PTR records for new purposes. The
2069 keys for these records are domain names, not reversed IP addresses. The
2070 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2071 leaves it alone. Component reversal etc. now happens only for IP addresses.
2074 34. The host_aton() function is supposed to be passed a string that is known
2075 to be a valid IP address. However, in the case of IPv6 addresses, it was
2076 not checking this. This is a hostage to fortune. Exim now panics and dies
2077 if the condition is not met. A case was found where this could be provoked
2078 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2079 components; fortuitously, this particular loophole had already been fixed
2080 by change 4.50/55 or 4.44/33 above.
2082 If there are any other similar loopholes, the new check in host_aton()
2083 itself should stop them being exploited. The report I received stated that
2084 data on the command line could provoke the exploit when Exim was running as
2085 exim, but did not say which command line option was involved. All I could
2086 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2087 running as the user.
2090 35. There was a buffer overflow vulnerability in the SPA authentication code
2091 (which came originally from the Samba project). I have added a test to the
2092 spa_base64_to_bits() function which I hope fixes it.
2095 36. The daemon start-up calls getloadavg() while still root for those OS that
2096 need the first call to be done as root, but it missed one case: when
2097 deliver_queue_load_max is set with deliver_drop_privilege. This is
2098 necessary for the benefit of the queue runner, because there is no re-exec
2099 when deliver_drop_privilege is set.
2101 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2102 was in use, was not putting the data itself into the right store pool;
2103 consequently, it could be overwritten for a subsequent message in the same
2104 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2107 38. Sometimes the final signoff response after QUIT could fail to get
2108 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2109 < 0 before doing a fflush(). This bug looks as though it goes back to the
2110 introduction of TLS in release 3.20, but "sometimes" must have been rare
2111 because the tests only now provoked it.
2117 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2118 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2119 Now it returns an integer. A number of calls were still expecting a T/F
2120 return. Fortuitously, in all cases, the tests worked in OK situations,
2121 which is the norm. However, things would have gone wrong on any write
2122 failures on the smtp file descriptor. This function is used when sending
2123 messages over SMTP and also when doing verify callouts.
2125 2. When Exim is called to do synchronous delivery of a locally submitted
2126 message (the -odf or -odi options), it no longer closes stderr before doing
2129 3. Implemented the mua_wrapper option.
2131 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2133 5. Implemented the functions header_remove(), header_testname(),
2134 header_add_at_position(), and receive_remove_recipient(), and exported them
2137 6. If an ACL "warn" statement specified the addition of headers, Exim already
2138 inserted X-ACL-Warn: at the start if there was no header name. However, it
2139 was not making this test for the second and subsequent header lines if
2140 there were newlines in the string. This meant that an invalid header could
2141 be inserted if Exim was badly configured.
2143 7. Allow an ACL "warn" statement to add header lines at the start or after all
2144 the Received: headers, as well as at the end.
2146 8. Added the rcpt_4xx retry error code.
2148 9. Added postmaster_mailfrom=xxx to callout verification option.
2150 10. Added mailfrom=xxxx to the callout verification option, for verify=
2153 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2154 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2156 12. Inserted some casts to stop certain compilers warning when using pointer
2157 differences as field lengths or precisions in printf-type calls (mostly
2158 affecting debugging statements).
2160 13. Added optional readline() support for -be (dynamically loaded).
2162 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2163 same clock tick as a message's arrival, so that its received time was the
2164 same as the "first fail" time on the retry record, and that message
2165 remained on the queue past the ultimate address timeout, every queue runner
2166 would try a delivery (because it was past the ultimate address timeout) but
2167 after another failure, the ultimate address timeout, which should have then
2168 bounced the address, did not kick in. This was a "< instead of <=" error;
2169 in most cases the first failure would have been in the next clock tick
2170 after the received time, and all would be well.
2172 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2173 being recognized when the domain list was tested by the match_domain
2174 condition in an expansion string.
2176 16. Added the ${str2b64: operator.
2178 17. Exim was always calling setrlimit() to set a large limit for the number of
2179 processes, without checking whether the existing limit was already
2180 adequate. (It did check for the limit on file descriptors.) Furthermore,
2181 errors from getrlimit() and setrlimit() were being ignored. Now they are
2182 logged to the main and panic logs, but Exim does carry on, to try to do its
2183 job under whatever limits there are.
2185 18. Imported PCRE 5.0.
2187 19. Trivial typo in log message " temporarily refused connection" (the leading
2190 20. If the log selector return_path_on_delivery was set and an address was
2191 redirected to /dev/null, the delivery process crashed because it assumed
2192 that a return path would always be set for a "successful" delivery. In this
2193 case, the whole delivery is bypassed as an optimization, and therefore no
2196 21. Internal re-arrangement: the function for sending a challenge and reading
2197 a response while authentication was assuming a zero-terminated challenge
2198 string. It's now changed to take a pointer and a length, to allow for
2199 binary data in such strings.
2201 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2203 23. Exim was not respecting finduser_retries when seeking the login of the
2204 uid under which it was called; it was always trying 10 times. (The default
2205 setting of finduser_retries is zero.) Also, it was sleeping after the final
2206 failure, which is pointless.
2208 24. Implemented tls_on_connect_ports.
2210 25. Implemented acl_smtp_predata.
2212 26. If the domain in control=submission is set empty, Exim assumes that the
2213 authenticated id is a complete email address when it generates From: or
2214 Sender: header lines.
2216 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2217 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2218 chown and chgrp in /bin and hostname in /usr/bin.
2220 28. Exim was keeping the "process log" file open after each use, just as it
2221 does for the main log. This opens the possibility of it remaining open for
2222 long periods when the USR1 signal hits a daemon. Occasional processlog
2223 errors were reported, that could have been caused by this. Anyway, it seems
2224 much more sensible not to leave this file open at all, so that is what now
2227 29. The long-running daemon process does not normally write to the log once it
2228 has entered its main loop, and it closes the log before doing so. This is
2229 so that log files can straightforwardly be renamed and moved. However,
2230 there are a couple of unusual error situations where the daemon does write
2231 log entries, and I had neglected to close the log afterwards.
2233 30. The text of an SMTP error response that was received during a remote
2234 delivery was being truncated at 512 bytes. This is too short for some of
2235 the long messages that one sometimes sees. I've increased the limit to
2238 31. It is now possible to make retry rules that apply only when a message has a
2239 specific sender, in particular, an empty sender.
2241 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2242 it possible to be selective about when SMTP synchronization is enforced.
2244 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2246 32. Implemented hosts_connection_nolog.
2248 33. Added an ACL for QUIT.
2250 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2253 35. Added mailbox_size and mailbox_filecount to appendfile.
2255 36. Added control = no_multiline_responses to ACLs.
2257 37. There was a bug in the logic of the code that waits for the clock to tick
2258 in the case where the clock went backwards by a substantial amount such
2259 that the microsecond fraction of "now" was more than the microsecond
2260 fraction of "then" (but the whole seconds number was less).
2262 38. Added support for the libradius Radius client library this is found on
2263 FreeBSD (previously only the radiusclient library was supported).
2269 1. When certain lookups returned multiple values in the form name=value, the
2270 quoting of the values was not always being done properly. Specifically:
2271 (a) If the value started with a double quote, but contained no whitespace,
2273 (b) If the value contained whitespace other than a space character (i.e.
2274 tabs or newlines or carriage returns) it was not quoted.
2275 This fix has been applied to the mysql and pgsql lookups by writing a
2276 separate quoting function and calling it from the lookup code. The fix
2277 should probably also be applied to nisplus, ibase and oracle lookups, but
2278 since I cannot test any of those, I have not disturbed their existing code.
2280 2. A hit in the callout cache for a specific address caused a log line with no
2281 reason for rejecting RCPT. Now it says "Previous (cached) callout
2282 verification failure".
2284 3. There was an off-by-one bug in the queryprogram router. An over-long
2285 return line was truncated at 256 instead of 255 characters, thereby
2286 overflowing its buffer with the terminating zero. As well as fixing this, I
2287 have increased the buffer size to 1024 (and made a note to document this).
2289 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2290 when Exim is waiting for an SMTP response from a remote server, Exim
2291 restarts its select() call on the socket, thereby resetting its timeout.
2292 This is not a problem when such interrupts are rare. Somebody set up a cron
2293 job to run exiwhat every 2 minutes, which is less than the normal select()
2294 timeout (5 or 10 minutes). This meant that the select() timeout never
2295 kicked in because it was always reset. I have fixed this by comparing the
2296 time when an interrupt arrives with the time at the start of the first call
2297 to select(). If more time than the timeout has elapsed, the interrupt is
2298 treated as a timeout.
2300 5. Some internal re-factoring in preparation for the addition of Sieve
2301 extensions (by MH). In particular, the "personal" test is moved to a
2302 separate function, and given an option for scanning Cc: and Bcc: (which is
2303 not set for Exim filters).
2305 6. When Exim created an email address using the login of the caller as the
2306 local part (e.g. when creating a From: or Sender: header line), it was not
2307 quoting the local part when it contained special characters such as @.
2309 7. Installed new OpenBSD configuration files.
2311 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2312 try to make them clearer.
2314 9. Callout options, other than the timeout value, were being ignored when
2315 verifying sender addresses in header lines. For example, when using
2317 verify = header_sender/callout=no_cache
2319 the cache was (incorrectly) being used.
2321 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2322 only the Cygwin environment.
2324 11. When return_path_on_delivery was set as a log selector, if different remote
2325 addresses in the same message used different return paths and parallel
2326 remote delivery occurred, the wrong values would sometimes be logged.
2327 (Whenever a remote delivery process finished, the return path value from
2328 the most recently started remote delivery process was logged.)
2330 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2331 lines when AUTH and/or TLS are in use. This is the "received protocol"
2332 field. Exim used to use "asmtp" for authenticated SMTP, without any
2333 indication (in the protocol name) for TLS use. Now it follows the RFC and
2334 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2335 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2336 names appear in log lines as well as in Received: header lines.
2338 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2339 extensions, and comparison tests, and to fix some bugs.
2341 14. Changes to the "personal" filter test:
2343 (1) The test was buggy in that it was just doing the equivalent of
2344 "contains" tests on header lines. For example, if a user's address was
2345 anne@some.where, the "personal" test would incorrectly be true for
2347 To: susanne@some.where
2349 This test is now done by extracting each address from the header in turn,
2350 and checking the entire address. Other tests that are part of "personal"
2351 are now done using regular expressions (for example, to check local parts
2352 of addresses in From: header lines).
2354 (2) The list of non-personal local parts in From: addresses has been
2355 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2356 taken from the Sieve specification recommendations.
2358 (3) If the message contains any header line starting with "List-" it is
2359 treated as non-personal.
2361 (4) The test for "circular" in the Subject: header line has been removed
2362 because it now seems ill-conceived.
2364 15. Minor typos in src/EDITME comments corrected.
2366 16. Installed latest exipick from John Jetmore.
2368 17. If headers_add on a router specified a text string that was too long for
2369 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2370 of string_sprintf() is now avoided.
2372 18. $message_body_size was not set (it was always zero) when running the DATA
2373 ACL and the local_scan() function.
2375 19. For the "mail" command in an Exim filter, no default was being set for
2376 the once_repeat time, causing a random time value to be used if "once" was
2377 specified. (If the value happened to be <= 0, no repeat happened.) The
2378 default is now 0s, meaning "never repeat". The "vacation" command was OK
2379 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2380 (I found it when inspecting the code).
2382 20. There is now an overall timeout for performing a callout verification. It
2383 defaults to 4 times the callout timeout, which applies to individual SMTP
2384 commands during the callout. The overall timeout applies when there is more
2385 than one host that can be tried. The timeout is checked before trying the
2386 next host. This prevents very long delays if there are a large number of
2387 hosts and all are timing out (e.g. when the network connections are timing
2388 out). The value of the overall timeout can be changed by specifying an
2389 additional sub-option for "callout", called "maxwait". For example:
2391 verify = sender/callout=5s,maxwait=20s
2393 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2394 to the end before writing, but this should make it even safer).
2396 22. Exim was forgetting that it had advertised PIPELINING for the second and
2397 subsequent messages on an SMTP connection. It was also not resetting its
2398 memory on STARTTLS and an internal HELO.
2400 23. When Exim logs an SMTP synchronization error within a session, it now
2401 records whether PIPELINING has been advertised or not.
2403 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2404 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2405 rather than long int.
2407 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2409 26. Added the never_mail option to autoreply.
2415 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2416 crash if the getsockname() call failed; this can happen if a connection is
2417 closed very soon after it is established. The problem was simply in the
2418 order in which certain operations were done, causing Exim to try to write
2419 to the SMTP stream before it had set up the file descriptor. The bug has
2420 been fixed by making things happen in the correct order.
2426 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2427 before the connection was closed, thus losing the rejection response.
2429 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2430 some early Solaris releases, but causes trouble in current releases where
2431 socklen_t is defined.
2433 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2436 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2439 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2440 that releasing the top of it at the end releases what was used for sub-
2441 expansions (unless the block got too big). However, discard this block if
2442 the first thing is a variable or header, so that we can use its block when
2443 it is dynamic (useful for very large $message_headers, for example).
2445 6. Lookups now cache *every* query, not just the most recent. A new, separate
2446 store pool is used for this. It can be recovered when all lookup caches are
2447 flushed. Lookups now release memory at the end of their result strings.
2448 This has involved some general refactoring of the lookup sources.
2450 7. Some code has been added to the store_xxx() functions to reduce the amount
2451 of flapping under certain conditions.
2453 8. log_incoming_interface used to affect only the <= reception log lines. Now
2454 it causes the local interface and port to be added to several more SMTP log
2455 lines, for example "SMTP connection from", and rejection lines.
2457 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2459 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2461 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2462 could be overwritten at the end of the current message (or the start of a
2463 new message if it was set in a HELO ACL). The value is now preserved for
2464 the duration of the SMTP connection.
2466 12. If a transport had a headers_rewrite setting, and a matching header line
2467 contained an unqualified address, that address was qualified, even if it
2468 did not match any rewriting rules. The underlying bug was that the values
2469 of the flags that permit the existence of unqualified sender and recipient
2470 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2471 non-local messages, and by -bnq for local messages) were not being
2472 preserved with the message after it was received.
2474 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2475 "next input=" as part of the text comprising the host identity instead of
2476 the correct text. The code was using the same buffer for two different
2477 strings. However, depending on which order the printing function evaluated
2478 its arguments, the bug did not always show up. Under Linux, for example, my
2479 test suite worked just fine.
2481 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2482 doesn't work with some older versions of Perl. It has been changed to "my",
2483 which in any case is probably the better facility to use.
2485 15. A really picky compiler found some instances of statements for creating
2486 error messages that either had too many or two few arguments for the format
2489 16. The size of the buffer for calls to the DNS resolver has been increased
2490 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2491 for addresses that have a lot of PTR records. This alleviates a problem; it
2492 does not fully solve it.
2494 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2495 buffer now truncates the list and logs the incident, which is the same
2496 action as happens when Exim is looking up a host name and its aliases.
2497 Previously in this situation something unpredictable would happen;
2498 sometimes it was "internal error: store_reset failed".
2500 18. If a server dropped the connection unexpectedly when an Exim client was
2501 using GnuTLS and trying to read a response, the client delivery process
2502 crashed while trying to generate an error log message.
2504 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2505 string, for example:
2507 warn message = H1: something\nH2: something
2509 the text was added as a single header line from Exim's point of view
2510 though it ended up OK in the delivered message. However, searching for the
2511 second and subsequent header lines using $h_h2: did not work. This has been
2512 fixed. Similarly, if a system filter added multiple headers in this way,
2513 the routers could not see them.
2515 20. Expanded the error message when iplsearch is called with an invalid key to
2516 suggest using net-iplsearch in a host list.
2518 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2519 ACLs are no longer actually imposed (and a message to that effect is
2522 22. If a "gecos" field in a passwd entry contained escaped characters, in
2523 particular, if it contained a \" sequence, Exim got it wrong when building
2524 a From: or a Sender: header from that name. A second bug also caused
2525 incorrect handling when an unquoted " was present following a character
2526 that needed quoting.
2528 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2529 was not being matched caselessly.
2531 24. Arranged for all hyphens in the exim.8 source to be escaped with
2534 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2535 properly. Recipient callout cache records were still being keyed to include
2536 the sender, even when use_sender was set false. This led to far more
2537 callouts that were necessary. The sender is no longer included in the key
2538 when use_sender is false.
2540 26. Added "control = submission" modifier to ACLs.
2542 27. Added the ${base62d: operator to decode base 62 numbers.
2544 28. dnsdb lookups can now access SRV records.
2546 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2547 the configuration file.
2549 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2550 (-v) mode. This makes the output for a verbose queue run more intelligible.
2552 31. Added a use_postmaster feature to recipient callouts.
2554 32. Added the $body_zerocount variable, containing the number of binary zero
2555 bytes in the message body.
2557 33. The time of last modification of the "new" subdirectory is now used as the
2558 "mailbox time last read" when there is a quota error for a maildir
2561 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2563 35. Added +ignore_unknown as a special item in host lists.
2565 36. Code for decoding IPv6 addresses in host lists is now included, even if
2566 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2567 address was recognized as an IP address, but was then not correctly decoded
2568 into binary, causing unexpected and incorrect effects when compared with
2575 1. Very minor rewording of debugging text in manualroute to say "list of
2576 hosts" instead of "hostlist".
2578 2. If verify=header_syntax was set, and a header line with an unqualified
2579 address (no domain) and a large number of spaces between the end of the
2580 name and the colon was received, the reception process suffered a buffer
2581 overflow, and (when I tested it) crashed. This was caused by some obsolete
2582 code that should have been removed. The fix is to remove it!
2584 3. When running in the test harness, delay a bit after writing a bounce
2585 message to get a bit more predictability in the log output.
2587 4. Added a call to search_tidyup() just before forking a reception process. In
2588 theory, someone could use a lookup in the expansion of smtp_accept_max_
2589 per_host which, without the tidyup, could leave open a database connection.
2591 5. Added the variables $recipient_data and $sender_data which get set from a
2592 lookup success in an ACL "recipients" or "senders" condition, or a router
2593 "senders" option, similar to $domain_data and $local_part_data.
2595 6. Moved the writing of debug_print from before to after the "senders" test
2598 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2599 problems for message scanning, either using a data ACL, or using
2600 local_scan() because the Received: header was not generated till after they
2601 were called (in order to set the time as the time of reception completion).
2602 I have revised the way this works. The header is now generated after the
2603 body is received, but before the ACL or local_scan() are called. After they
2604 are run, the timestamp in the header is updated.
2610 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2611 before starting a queue runner without re-exec. This happened only when
2612 deliver_drop_privilege was set or when the Exim user was set to root. The
2613 effect of the bug was that timeouts during subsequent deliveries caused
2614 crashes instead of being properly handled. The handler is now left at its
2615 default (and expected) setting.
2617 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2618 message, again when deliver_drop_privilege is set or Exim is run as root.
2619 The bug described in (1) was not present in this case, but the tidying up
2620 of the other signals was missing. I have made the two cases consistent.
2622 3. The ignore_target_hosts setting on a manualroute router was being ignored
2623 for hosts that were looked up using the /MX notation.
2625 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2628 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2629 operated on the sender address. After changing the $sender_address to <>
2630 for the sender address verify, Exim was re-instated it as the original
2631 (before rewriting) address, but remembering that it had rewritten it, so it
2632 wasn't rewriting it again. This bug also had the effect of breaking the
2633 sender address verification caching when the sender address was rewritten.
2635 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2636 This has been changed so that if the ip literal address matches
2637 ignore_target_hosts, the router declines.
2639 7. Added expansion conditions match_domain, match_address, and match_local_
2640 part (NOT match_host).
2642 8. The placeholder for the Received: header didn't have a length field set.
2644 9. Added code to Exim itself and to exim_lock to test for a specific race
2645 condition that could lead to file corruption when using MBX delivery. The
2646 issue is with the lockfile that is created in /tmp. If this file is removed
2647 after a process has opened it but before that process has acquired a lock,
2648 there is the potential for a second process to recreate the file and also
2649 acquire a lock. This could lead to two Exim processes writing to the file
2650 at the same time. The added code performs the same test as UW imapd; it
2651 checks after acquiring the lock that its file descriptor still refers to
2652 the same named file.
2654 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2655 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2658 11. Added the smtp_active_hostname option. If used, this will typically be made
2659 to depend on the incoming interface address. Because $interface_address is
2660 not set up until the daemon has forked a reception process, error responses
2661 that can happen earlier (such as "too many connections") no longer contain
2664 12. If an expansion in a condition on a "warn" statement fails because a lookup
2665 defers, the "warn" statement is abandoned, and the next ACL statement is
2666 processed. Previously this caused the whole ACL to be aborted.
2668 13. Added the iplsearch lookup type.
2670 14. Added ident_timeout as a log selector.
2672 15. Added tls_certificate_verified as a log selector.
2674 16. Added a global option tls_require_ciphers (compare the smtp transport
2675 option of the same name). This controls incoming TLS connections.
2677 17. I finally figured out how to make tls_require_ciphers do a similar thing
2678 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2679 before starting the TLS session.
2681 18. Tabs are now shown as \t in -bP output.
2683 19. If the log selector return_path_on_delivery was set, Exim crashed when
2684 bouncing a message because it had too many Received: header lines.
2686 20. If two routers both had headers_remove settings, and the first one included
2687 a superfluous trailing colon, the final name in the first list and the
2688 first name in the second list were incorrectly joined into one item (with a
2689 colon in the middle).
2695 1. Added -C and -D options to the exinext utility, mainly to make it easier
2696 to include in the automated testing, but these could be helpful when
2697 multiple configurations are in use.
2699 2. The exinext utility was not formatting the output nicely when there was
2700 an alternate port involved in the retry record key, nor when there was a
2701 message id as well (for retries that were specific to a specific message
2702 and a specific host). It was also confused by IPv6 addresses, because of
2703 the additional colons they contain. I have fixed the IPv4 problem, and
2704 patched it up to do a reasonable job for IPv6.
2706 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2707 delivery, the log line now contains "pipelined" if PIPELINING was used.
2709 4. An SMTP transport process used to panic and die if the bind() call to set
2710 an explicit outgoing interface failed. This has been changed; it is now
2711 treated in the same way as a connect() failure.
2713 5. A reference to $sender_host_name in the part of a conditional expansion
2714 that was being skipped was still causing a DNS lookup. This no longer
2717 6. The def: expansion condition was not recognizing references to header lines
2718 that used bh_ and bheader_.
2720 7. Added the _cache feature to named lists.
2722 8. The code for checking quota_filecount in the appendfile transport was
2723 allowing one more file than it should have been.
2725 9. For compatibility with Sendmail, the command line option
2733 and sets the incoming protocol and host name (for trusted callers). The
2734 host name and its colon can be omitted when only the protocol is to be set.
2735 Note the Exim already has two private options, -pd and -ps, that refer to
2736 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2737 "s", but I don't think that's a major issue.
2739 10. A number of refactoring changes to the code, none of which should affect
2742 (a) The number of logging options was getting close to filling up the
2743 32-bit word that was used as a bit map. I have split them into two classes:
2744 those that are passed in the argument to log_write(), and those that are
2745 only ever tested independently outside of that function. These are now in
2746 separate 32-bit words, so there is plenty of room for expansion again.
2747 There is no change in the user interface or the logging behaviour.
2749 (b) When building, for example, log lines, the code previously used a
2750 macro that called string_cat() twice, in order to add two strings. This is
2751 not really sufficiently general. Furthermore, there was one instance where
2752 it was actually wrong because one of the argument was used twice, and in
2753 one call a function was used. (As it happened, calling the function twice
2754 did not affect the overall behaviour.) The macro has been replaced by a
2755 function that can join an arbitrary number of extra strings onto a growing
2758 (c) The code for expansion conditions now uses a table and a binary chop
2759 instead of a serial search (which was left over from when there were very
2760 few conditions). Also, it now recognizes conditions like "pam" even when
2761 the relevant support is not compiled in: a suitably worded error message is
2762 given if an attempt is made to use such a condition.
2764 11. Added ${time_interval:xxxxx}.
2766 12. A bug was causing one of the ddress fields not to be passed back correctly
2767 from remote delivery subprocesses. The field in question was not being
2768 subsequently used, so this caused to problems in practice.
2770 13. Added new log selectors queue_time and deliver_time.
2772 14. Might have fixed a bug in maildirsizefile handling that threw up
2773 "unexpected character" debug warnings, and recalculated the data
2774 unnecessarily. In any case, I expanded the warning message to give more
2777 15. Added the message "Restricted characters in address" to the statements in
2778 the default ACL that block characters like @ and % in local parts.
2780 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2781 Three changes have been made:
2783 (a) There was a serious bug; a negative response to MAIL caused the whole
2784 recipient domain to be cached as invalid, thereby blocking all messages
2785 to all local parts at the same domain, from all senders. This bug has
2786 been fixed. The domain is no longer cached after a negative response to
2787 MAIL if the sender used is not empty.
2789 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2792 (c) A new callout option, "use_sender" has been added for people who want
2793 the modified behaviour.
2799 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2802 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2803 indeed breaks things for older releases.
2805 3. Added additional logging to the case where there is a problem reading data
2806 from a filter that is running in a subprocess using a pipe, in order to
2807 try to track down a specific problem.
2809 4. Testing facility fudge: when running in the test harness and attempting
2810 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2811 getting "No route to host". Convert this to a timeout.
2813 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2816 6. Some OS don't have socklen_t but use size_t instead. This affects the
2817 fifth argument of getsockopt() amongst other things. This is now
2818 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2819 can be set for individual OS. I have set it for SunOS5, OSF1, and
2820 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2821 some earlier ones do not.
2823 7. Change 4.30/15 was not doing the test caselessly.
2825 8. The standard form for an IPv6 address literal was being rejected by address
2826 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2827 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2828 this, as well as the form without the "IPv6" on the front (but only when
2829 address literals are enabled, of course).
2831 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2833 10. Exim crashed if a message with an empty sender address specified by -f
2834 encountered a router with an errors_to setting. This could be provoked only
2835 by a command such as
2839 where an empty string was supplied; "<>" did not hit this bug.
2841 11. Installed PCRE release 4.5.
2843 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2844 remained set. It is now erased.
2846 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2847 times from message ids (which are base 36 rather than the normal 62).
2849 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2850 were being counted as actual protocol errors, and logged if the log
2851 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2852 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2853 and DATA following a set of rejected RCPTs do not count as protocol errors.
2854 In other words, Exim assumes they were pipelined, though this may not
2855 actually be the case. Of course, in all cases the client gets an
2856 appropriate error code.
2858 15. If a lookup fails in an ACL condition, a message about the failure may
2859 be available; it is used if testing the ACL cannot continue, because most
2860 such messages specify what the cause of the deferral is. However, some
2861 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2862 that caused an old message to be retained and used if a later statement
2863 caused a defer, replacing the real cause of the deferral.
2865 16. If an IP address had so many PTR records that the DNS lookup buffer
2866 was not large enough to hold them, Exim could crash while trying to process
2867 the truncated data. It now detects and logs this case.
2869 17. Further to 4.21/58, another change has been made: if (and only if) the
2870 first line of a message (the first header line) ends with CRLF, a bare LF
2871 in a subsequent header line has a space inserted after it, so as not to
2872 terminate the header.
2874 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2875 unnecessarily, used atoi() instead of strtol(), and didn't check the
2876 termination when getting file sizes from file names by regex.
2878 19. Completely re-implemented the support for maildirsize files, in the light
2879 of a number of problems with the previous contributed implementation
2880 (4.30/29). In particular:
2882 . If the quota is zero, the maildirsize file is maintained, but no quota is
2885 . If the maildir directory does not exist, it is created before any attempt
2886 to write a maildirsize file.
2888 . The quota value in the file is just a cache; if the quota is changed in
2889 the transport, the new value overrides.
2891 . A regular expression is available for excluding directories from the
2894 20. The autoreply transport checks the characters in options that define the
2895 message's headers; it allows continued headers, but it was checking with
2896 isspace() after an embedded newline instead of explicitly looking for a
2899 21. If all the "regular" hosts to which an address was routed had passed their
2900 expiry times, and had not reached their retry times, the address was
2901 bounced, even if fallback hosts were defined. Now Exim should go on to try
2904 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2905 equivalent code in the SMTP transport. Some hosts send humungous responses
2906 to HELO/EHLO, more than 1024 it seems.
2908 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2909 gives compiler warnings in some environments. I've now done it "properly",
2912 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2913 (because of problems with the built-in one) was declared to return uschar *
2914 instead of char *, causing compiler failure.
2916 25. Fixed a file descriptor leak when processing alias/forward files.
2918 26. Fixed a minor format string issue in dbfn.c.
2920 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2922 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2923 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2926 29. When a sender address is verified, it is cached, to save repeating the test
2927 when there is more than one recipient in a message. However, when the
2928 verification involves a callout, it is possible for different callout
2929 options to be set for different recipients. It is too complicated to keep
2930 track of this in the cache, so now Exim always runs a verification when a
2931 callout is required, relying on the callout cache for the optimization.
2932 The overhead is duplication of the address routing, but this should not be
2935 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2936 to be verified with callout=postmaster, and the main callout worked but the
2937 postmaster check failed, the verification correctly failed. However, if a
2938 subsequent RCPT command asked for sender verification *without* the
2939 postmaster check, incorrect caching caused this verification also to fail,
2942 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2943 it was not caching the DNS options (qualify_single, search_parents) that
2944 were used when the lookup failed. A subsequent lookup with different
2945 options therefore always gave the same answer, though there were cases
2946 where it should not have. (Example: a "domains = !$mx_any" option on a
2947 dnslookup router: the "domains" option is always processed without any
2948 widening, but the router might have qualify_single set.) Now Exim uses the
2949 cached value only when the same options are set.
2951 32. Added John Jetmore's "exipick" utility to the distribution.
2953 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2954 than a timeout (e.g. a certificate is required, and is not provided), an
2955 Exim server now closes the connection immediately. Previously it waited for
2956 the client to close - but if the client is SSL, it seems that they each
2957 wait for each other, leading to a delay before one of them times out.
2959 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2960 maintained 0.8.x compatibility because I don't think many are using it, and
2961 it is clearly obsolete.
2963 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2966 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2967 client certificate was expired. A simple patch fixes this, though I don't
2968 understand the full logic of why the verify callback is called multiple
2971 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2972 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2973 which causes problems with some clients (such as the Certicom SSL Plus
2974 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2975 disables the coutermeasure allowing Eudora to connect."
2977 38. Exim was not checking that a write() to a log file succeeded. This could
2978 lead to Bad Things if a log got too big, in particular if it hit a file
2979 size limit. Exim now panics and dies if it cannot write to a log file, just
2980 as it does if it cannot open a log file.
2982 39. Modified OS/Makefile-Linux so that it now contains
2984 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2986 The two -D definitions ensure that Exim is compiled with large file
2987 support, which makes it possible to handle log files that are bigger than
2990 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2991 instance) a domain was checked against a named list that involved a lookup,
2992 causing $domain_data to be set, then another domain was checked against the
2993 same list, then the first domain was re-checked, the value of $domain_data
2994 after the final check could be wrong. In particular, if the second check
2995 failed, it could be set empty. This bug probably also applied to
2998 41. The strip_trailing_dot option was not being applied to the address given
2999 with the -f command-line option.
3001 42. The code for reading a message's header from the spool was incrementing
3002 $received_count, but never initializing it. This meant that the value was
3003 incorrect (doubled) while delivering a message in the same process in which
3004 it was received. In the most common configuration of Exim, this never
3005 happens - a fresh exec is done - but it can happen when
3006 deliver_drop_privilege is set.
3008 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3009 it now includes up to 150 characters of the unexpected data in the log
3012 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3013 and building data strings. The size of both of these buffers was 10 000
3014 bytes - far larger than anybody would *ever* want, thought I. Needless to
3015 say, somebody hit the limit. I have increased the maximum line length to
3016 20 000 and the maximum data length of concatenated lines to 100 000. I have
3017 also fixed two bugs, because there was no checking on these buffers. Tsk,
3018 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3019 buffer is too small.
3021 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3022 lsearch lookups. Now it does.
3024 46. When parsing a route_list item in a manualroute router, a fixed-length
3025 buffer was used for the list of hosts. I made this 1024 bytes long,
3026 thinking that nobody would ever have a list of hosts that long. Wrong.
3027 Somebody had a whole pile of complicated expansion conditions, and the
3028 string was silently truncated, leading to an expansion error. It turns out
3029 that it is easier to change to an unlimited length (owing to other changes
3030 that have happened since this code was originally written) than to build
3031 structure for giving a limitation error. The length of the item that
3032 expands into the list of hosts is now unlimited.
3034 47. The lsearch lookup could not handle data where the length of text line was
3035 more than 4095 characters. Such lines were truncated, leading to shortened
3036 data being returned. It should now handle lines of any length.
3038 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3039 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3042 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3043 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3044 no longer happens in comment lines. A list of replacements is now placed
3045 at the head of all of the source files, except those whose only change is
3046 to replace PERL_COMMAND in the very first #! line.
3048 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3049 messages on the queue, with a bottom-up merge sort, using code contributed
3050 by Michael Haardt. This should make operations like -bp somewhat faster on
3051 large queues. It won't affect queue runners, except when queue_run_in_order
3054 51. Installed eximstats 1.31 in the distribution.
3056 52. Added support for SRV lookups to the dnslookup router.
3058 53. If an ACL referred to $message_body or $message_body_end, the value was not
3059 reset for any messages that followed in the same SMTP session.
3061 54. The store-handling optimization for building very long strings was not
3062 differentiating between the different store pools. I don't think this
3063 actually made any difference in practice, but I've tidied it.
3065 55. While running the routers to verify a sender address, $sender_address
3066 was still set to the sender address. This is wrong, because when routing to
3067 send a bounce to the sender, it would be empty. Therefore, I have changed
3068 it so that, while verifying a sender address, $sender_address is set to <>.
3069 (There is no change to what happens when verifying a recipient address.)
3071 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3072 target A or AAAA records (if not already returned) without resetting the
3073 qualify_single or search_parents options of the DNS resolver. These are
3074 inappropriate in this case because the targets of MX and SRV records must
3075 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3076 target that, when qualified, matched something in the local domain. These
3077 two options are now turned off when doing these lookups.
3079 57. It seems that at least some releases of Reiserfs (which does not have the
3080 concept of a fixed number of inodes) returns zero and not -1 for the
3081 number of available inodes. This interacted badly with check_spool_inodes,
3082 which assumed that -1 was the "no such thing" setting. What I have done is
3083 to check that the total number of inodes is greater than zero before doing
3084 the test of how many are available.
3086 58. When a "warn" ACL statement has a log_message modifier, the message is
3087 remembered, and not repeated. This is to avoid a lot of repetition when a
3088 message has many recipients that cause the same warning to be written.
3089 Howewer, Exim was preserving the list of already written lines for an
3090 entire SMTP session, which doesn't seem right. The memory is now reset if a
3091 new message is started.
3093 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3094 debugging output unless log_rewrite was also set.
3096 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3097 of (int)(handle) when we know that handle contains (void *)(-1).
3099 61. The Exim daemon panic-logs an error return when it closes the incoming
3100 connection. However "connection reset by peer" seems to be common, and
3101 isn't really an error worthy of noting specially, so that particular error
3104 62. When Exim is trying to find all the local interfaces, it used to panic and
3105 die if the ioctl to get the interface flags failed. However, it seems that
3106 on at least one OS (Solaris 9) it is possible to have an interface that is
3107 included in the list of interfaces, but for which you get a failure error
3108 for this call. This happens when the interface is not "plumbed" into a
3109 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3110 failure of the "get flags" call assumes that the interface is down.
3112 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3113 makes life easier for people who are doing arithmetic on fields extracted
3114 from dates, where you often get leading zeros that should not be
3115 interpreted as octal.
3117 64. Added qualify_domain to the redirect router, to override the global
3120 65. If a pathologically long header line contained very many addresses (the
3121 report of this problem mentioned 10 000) and each of them was rewritten,
3122 Exim could use up a very large amount of memory. (It kept on making new
3123 copies of the header line as it rewrote, and never released the old ones.)
3124 At the expense of a bit more processing, the header rewriting function has
3125 been changed so that it no longer eats memory in this way.
3127 66. The generation of the Received: header has been moved from the time that a
3128 message starts to be received, to the time that it finishes. The timestamp
3129 in the Received: header should now be very close to that of the <= log
3130 line. There are two side-effects of this change:
3132 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3133 logged header lines no longer include the local Received: line, because
3134 it has not yet been created. The same applies to a copy of the message
3135 that is returned to a non-SMTP sender when a message is rejected.
3137 (b) When a filter file is tested using -bf, no additional Received: header
3138 is added to the test message. After some thought, I decided that this
3141 This change does not affect the value of $received_for. It is still set
3142 after address rewriting, but before local_scan() is called.
3144 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3146 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3147 gave an unhelpful panic error message, and a defer error. I have managed to
3148 change this behaviour so that it now rejects any supplied certificate,
3149 which seems right, as the list of acceptable certificates is empty.
3151 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3152 gave an unhelpful defer error. I have not managed to make this reject any
3153 supplied certificates, but the error message it gives is "no certificate
3154 supplied", which is not helpful.
3156 70. exigrep's output now also includes lines that are not associated with any
3157 message, but which match the given pattern. Implemented by a patch from
3158 Martin Sluka, which also tidied up the Perl a bit.
3160 71. Recipient callout verification, like sender verification, was using <> in
3161 the MAIL FROM command. This isn't really the right thing, since the actual
3162 sender may affect whether the remote host accepts the recipient or not. I
3163 have changed it to use the actual sender in the callout; this means that
3164 the cache record is now keyed on a recipient/sender pair, not just the
3165 recipient address. There doesn't seem to be a real danger of callout loops,
3166 since a callout by the remote host to check the sender would use <>.
3167 [SEE ABOVE: changed after hitting problems.]
3169 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3170 temporary errors. However, in the case of such a code being given after
3171 the end of a data transmission (i.e. after ".") Exim was failing to write
3172 a retry record for the message. (Yes, there was some broken host that was
3173 actually sending 8xx at this point.)
3175 73. An unknown lookup type in a host list could cause Exim to panic-die when
3176 the list was checked. (An example that provoked this was putting <; in the
3177 middle of a list instead of at the start.) If this happened during a DATA
3178 ACL check, a -D file could be left lying around. This kind of configuration
3179 error no longer causes Exim to die; instead it causes a defer errror. The
3180 incident is still logged to the main and panic logs.
3182 74. Buglet left over from Exim 3 conversion. The message "too many messages
3183 in one connection" was written to the rejectlog but not the mainlog, except
3184 when address rewriting (yes!) was being logged.
3186 75. Added write_rejectlog option.
3188 76. When a system filter was run not as root (that is, when system_filter_user
3189 was set), the values of the $n variables were not being returned to the
3190 main process; thus, they were not subsequently available in the $sn
3193 77. Added +return_path_on_delivery log selector.
3195 78. A connection timeout was being treated differently from recipients deferred
3196 when testing hosts_max_try with a message that was older than the host's
3197 retry timeout. (The host should not be counted, thus allowing all hosts to
3198 be tried at least once before bouncing.) This may have been the cause of an
3199 occasionally reported bug whereby a message would remain on the queue
3200 longer than the retry timeout, but would be bounced if a delivery was
3201 forced. I say "may" because I never totally pinned down the problem;
3202 setting up timeout/retry tests is difficult. See also the next item.
3204 79. The ultimate address timeout was not being applied to errors that involved
3205 a combination of host plus message (for example, a timeout on a MAIL
3206 command). When an address resolved to a number of possible hosts, and they
3207 were not all tried for each delivery (e.g. because of hosts_max_try), a
3208 message could remain on the queue longer than the retry timeout.
3210 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3213 81. Fixed an obscure SMTP outgoing bug which required at least the following
3214 conditions: (a) there was another message waiting for the same server;
3215 (b) the server returned 5xx to all RCPT commands in the first message so
3216 that the message was not completed; (c) the server dropped the connection
3217 or gave a negative response to the RSET that Exim sends to abort the
3218 transaction. The observed case was a dropped connection after DATA that had
3219 been sent in pipelining mode. That is, the server had advertised PIPELINING
3220 but was not implementing it correctly. The effect of the bug was incorrect
3221 behaviour, such as trying another host, and this could lead to a crash.
3227 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3228 and daemon.c were passed as pointers to ints; they should have been
3229 pointers to socklen_t variables (which are typically unsigned ints).
3231 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3234 3. Fixed a really odd bug that affected only the testing scheme; patching a
3235 certain fixed string in the binary changed the value of another string that
3236 happened to be identical to the end of the original first string.
3238 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3239 name", it returns that address as the IP address. On some operating
3240 systems (e.g. Solaris), it also passes back the IP address string as the
3241 "host name". However, on others (e.g. Linux), it passes back an empty
3242 string. Exim wasn't checking for this, and was changing the host name to an
3243 empty string, assuming it had been canonicized.
3245 5. Although rare, it is permitted to have more than one PTR record for a given
3246 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3247 all the names associated with an address, because they do in Solaris.
3248 However, it seems that they do not in Linux for data that comes from the
3249 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3250 I found this out when I moved to a new Linux workstation and tried to run
3251 the Exim test suite.
3253 To get round this problem I have changed the code so that it now does its
3254 own call to the DNS to look up PTR records when searching for a host name.
3255 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3256 addresses that are only in /etc/hosts are still found.
3258 This behaviour is, however, controlled by an option called host_lookup_
3259 order, which defaults to "bydns:byaddr". If people want to use the other
3260 order, or indeed, just use one or the other means of lookup, they can
3261 specify it in this variable.
3263 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3264 some operating systems, this comes back from gethostbyaddr() as an empty
3265 string, and this is what Exim used to test for. However, it seems that in
3266 other systems, "." is yielded. Exim now tests for this case too.
3268 7. The values of check_spool_space and check_log_space are now held internally
3269 as a number of kilobytes instead of an absolute number of bytes. If a
3270 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3271 kilobyte. This means that much larger values can be stored.
3273 8. Exim monitor: an attempt to get the action menu when not actually pointing
3274 at a message produces an empty menu entitled "No message selected". This
3275 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3276 no entries in it ("Shell widget menu has zero width and/or height"). So I
3277 have added a single, blank menu entry in this case.
3279 9. Added ${quote_local_part.
3281 10. MIME decoding is now applied to the contents of Subject: header lines when
3284 11. Now that a reference to $sender_host_address automatically causes a reverse
3285 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3286 host lookup before query-style lookups in lists that might use this
3287 variable. This has therefore been abolished, and the "net-" prefix is no
3288 longer necessary for query-style lookups.
3290 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3291 have been a typo for LFLAGS, so it has been changed.
3293 13. The install script calls Exim with "-C /dev/null" in order to find the
3294 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3295 to be output. Howeve, since Exim outputs its version number before the
3296 error, it didn't break the script. It just looked ugly. I fixed this by
3297 always allowing "-C /dev/null" if the caller is root.
3299 14. Ignore overlarge ACL variable number when reading spool file - insurance
3300 against a later release with more variables having written the file.
3302 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3303 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3304 this, as well as the form without the "IPv6" on the front.
3306 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3307 OS/Makefile-Darwin file.
3309 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3310 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3312 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3313 variable. However, some OS use ints for inodes. I've added cast to long int
3314 to get rid of the compiler warning.
3316 19. I had forgotten to lock out "/../" in configuration file names when
3317 ALT_CONFIG_PREFIX was set.
3319 20. Routers used for verification do not need to specify transports. However,
3320 if such a router generated a host list, and callout was configured, Exim
3321 crashed, because it could not find a port number from the (non-existent)
3322 transport. It now assumes port 25 in this circumstance.
3324 21. Added the -t option to exigrep.
3326 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3327 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3328 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3329 them is set, LOOKUP_LSEARCH is forced.
3331 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3333 24. Added sender and host information to the "rejected by local_scan()" log
3334 line; previously there was no indication of these.
3336 25. Added .include_if_exists.
3338 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3339 Linux. It turns out that not all file systems support this. Apparently some
3340 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3341 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3342 not supported on the file descriptor, is now ignored when Exim is trying to
3343 sync a directory. This applies only to Linux.
3345 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3347 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3348 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3351 29. Added support for maildirsize files from supplied patch (modified a bit).
3353 30. The use of :fail: followed by an empty string could lead Exim to respond to
3354 sender verification failures with (e.g.):
3356 550 Verification failed for <xxx>
3357 550 Sender verify failed
3359 where the first response line was missing the '-' that indicates it is not
3360 the final line of the response.
3362 31. The loop for finding the name of the user that called Exim had a hardwired
3363 limit of 10; it now uses the value of finduser_retries, which is used for
3364 all other user lookups.
3366 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3369 33. Exim was neglecting to zero errno before one call of strtol() when
3370 expanding a string and expecting an integer value. On some systems this
3371 resulted in spurious "integer overflow" errors. Also, it was casting the
3372 result into an int without checking.
3374 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3375 did not work. The code looks as if it has *never* worked, though it appears
3376 to have been documented since at least releast 1.62. I have made it work.
3378 35. The "timeout_DNS" error in retry rules, also documented since at least
3379 1.62, also never worked. As it isn't clear exactly what this means, and
3380 clearly it isn't a major issue, I have abolished the feature by treating it
3381 as "timeout", and writing a warning to the main and panic logs.
3383 36. The display of retry rules for -brt wasn't always showing the error code
3386 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3387 timeout_connect_A, timeout_connect_MX.
3389 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3390 to the empty sender.
3392 39. The daemon was not analysing the content of -oX till after it had closed
3393 stderr and disconnected from the controlling terminal. This meant that any
3394 syntax errors were only noted on the panic log, and the return code from
3395 the command was 0. By re-arranging the code a little, I've made the
3396 decoding happen first, so such errors now appear on stderr, and the return
3397 code is 1. However, the actual setting up of the sockets still happens in
3398 the disconnected process, so errors there are still only recorded on the
3401 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3402 connections (as happens on some IP stacks) was logged at start up time as
3403 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3404 it from "IPv6 and IPv4", which means that two separate sockets are being
3407 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3408 says whether AF_INET or AF_INET6 was passed as an argument.
3410 42. Exiwhat output was messed up when time zones were included in log
3413 43. Exiwhat now gives more information about the daemon's listening ports,
3414 and whether -tls-on-connect was used.
3416 44. The "port" option of the smtp transport is now expanded.
3418 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3419 silently ignored. Now an error message is written to the main and panic
3422 46. There's a new ACL modifier called "logwrite" which writes to a log file
3423 as soon as it is encountered.
3425 47. Added $local_user_uid and $local_user_gid at routing time.
3427 48. Exim crashed when trying to verify a sender address that was being
3430 49. Exim was recognizing only a space character after ".include". It now also
3431 recognizes a tab character.
3433 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3434 extracting the relevant information from the specification. The man page no
3435 longer contains scrambled data for the -d option, and I've added a section
3436 at the front about calling Exim under different names.
3438 51. Added "extra_headers" argument to the "mail" command in filter files.
3440 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3443 53. Installed eximstats 1.29.
3445 54. Added transport_filter_timeout as a generic transport option.
3447 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3448 Cc: header lines. This was required by RFC 822, but it not required by RFC
3451 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3452 incoming messages that did not have them. Now it does so only if the
3453 message originates locally, that is, if there is no associated remote host
3454 address. When Resent- header lines are present, this applies to the Resent-
3455 lines rather than the non-Resent- lines.
3457 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3458 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3460 58. Messages for configuration errors now include the name of the main
3461 configuration file - useful now that there may be more than one file in a
3462 list (.included file names were always shown).
3464 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3465 for those rare installations that do not start the daemon as root or run it
3466 setuid root. I've cut out the call to initgroups() if the daemon is not
3469 60. The Exim user and group can now be bound into the binary as text strings
3470 that are looked up at the start of Exim's processing.
3472 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3474 62. Added $mailstore_basename variable.
3476 63. Installed patch to sieve.c from Michael Haardt.
3478 64. When Exim failed to open the panic log after failing to open the main log,
3479 the original message it was trying to log was written to stderr and debug
3480 output, but if they were not available (the usual case in production), it
3481 was lost. Now it is written to syslog before the two lines that record the
3482 failures to open the logs.
3484 65. Users' Exim filters run in subprocesses under the user's uid. It is
3485 possible for a "deliver" command or an alias in a "personal" command to
3486 provoke an address rewrite. If logging of address rewriting is configured,
3487 this fails because the process is not running as root or exim. There may be
3488 a better way of dealing with this, but for the moment (because 4.30 needs
3489 to be released), I have disabled address rewrite logging when running a
3490 filter in a non-root, non-exim process.
3496 1. The buildconfig auxiliary program wasn't quoting the value set for
3497 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3498 not defined. This bug was masked in 4.22 by the effect that was fixed in
3501 2. Some messages that were rejected after a message id was allocated were
3502 shown as "incomplete" by exigrep. It no longer does this for messages that
3503 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3505 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3506 have allow_domain_literals set, the ID did not get logged in the <= line.
3507 Domain literals are now always recognized in Message-ID: header lines.
3509 4. The first argument for a ${extract expansion item is the key name or field
3510 number. Leading and trailing spaces in this item were not being ignored,
3511 causing some misleading effects.
3513 5. When deliver_drop_privilege was set, single queue runner processes started
3514 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3515 same command in the process it spins off) were not dropping privilege.
3517 6. When the daemon running as "exim" started a queue runner, it always
3518 re-executed Exim in the spun-off process. This is a waste of effort when
3519 deliver_drop_privilege is set. The new process now just calls the
3520 queue-runner function directly.
3526 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3529 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3530 ignored. Though the use of -oP was forcing the writing of a pid file, it
3531 was always written to the default place.
3533 3. If the message "no IP address found for host xxxx" is generated during
3534 incoming verification, it is now followed by identification of the incoming
3535 connection (so you can more easily find what provoked it).
3537 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3539 5. Added some features to "harden" Exim a bit more against certain attacks:
3541 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3542 be put in Local/Makefile. This is like the never_users runtime option,
3543 but it cannot be overridden. The default setting is "root".
3545 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3546 prefix string with which any file named in a -C command line option
3549 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3550 is retained for -C and -D only if the caller of Exim is root. Without
3551 it, the exim user may also use -C and -D and retain privilege.
3553 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3554 command line option is disabled.
3556 6. Macro names set by the -D option must start with an upper case letter, just
3557 like macro names defined in the configuration file.
3559 7. Added "dereference=" facility to LDAP.
3561 8. Two instances of the typo "uknown" in the source files are fixed.
3563 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3564 the Configure-Makefile script screwed up while processing it.
3566 10. Incorporated PCRE 4.4.
3568 11. The SMTP synchronization check was not operating right at the start of an
3569 SMTP session. For example, it could not catch a HELO sent before the client
3570 waited for the greeting. There is now a check for outstanding input at the
3571 point when the greeting is written. Because of the duplex, asynchronous
3572 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3573 way, but not yet received, when the check is performed.
3575 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3576 on TCP/IP sockets, because this apparently causes some broken clients to
3579 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3580 unchanged) from the Cygwin maintainer.
3582 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3583 was supported instead of testing for mbx. Effectively a typo.
3585 15. The spa authenticator server code was not checking that the input it
3586 received was valid base64.
3588 16. The debug output line for the "set" modifier in ACLs was not showing the
3589 name of the variable that was being set.
3591 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3593 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3594 lookup on its own. Something else was needed to trigger the lookup. For
3595 example, a match in host_lookup or the need for a host name in a host list.
3596 Now, if $sender_host_name is referenced and the host name has not yet been
3597 looked up, a lookup is performed. If the lookup fails, the variable remains
3598 empty, and $host_lookup_failed is set to "1".
3600 19. Added "eqi" as a case-independent comparison operator.
3602 20. The saslauthd authentication condition could segfault if neither service
3603 nor realm was specified.
3605 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3606 error message that was logged was misleading, and incoming SMTP
3607 connections were dropped. The message is now more accurate, and temporary
3608 errors are given to SMTP connections.
3610 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3611 (except RSET) with a 421 error, until QUIT is received. However, it was
3612 failing to send a response to QUIT.
3614 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3615 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3616 ACL is now run after the helo_try_verify_hosts code.
3618 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3619 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3620 case-independent, so other case variants are also recognized). Apparently
3621 some systems use these upper case variants.
3623 25. If more than two messages were waiting for the same host, and a transport
3624 filter was specified for the transport, Exim sent two messages over the
3625 same TCP/IP connection, and then failed with "socket operation on non-
3626 socket" when it tried to send the third.
3628 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3630 27. The extern definition of crypt16() in expand.c was not being excluded when
3631 the OS had its own crypt16() function.
3633 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3634 as a preferred synonym for return_size_limit, both as an option and as an
3637 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3639 30. Changed the default configuration ACL to relax the local part checking rule
3640 for addresses that are not in any local domains. For these addresses,
3641 slashes and pipe symbols are allowed within local parts, but the sequence
3642 /../ is explicitly forbidden.
3644 31. SPA server authentication was not clearing the challenge buffer before
3647 32. log_message in a "warn" ACL statement was writing to the reject log as
3648 well as to the main log, which contradicts the documentation and doesn't
3649 seem right (because no rejection is happening). So I have stopped it.
3651 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3652 However, I am unable to do any testing of this.
3654 34. Fixed an infelicity in the appendfile transport. When checking directories
3655 for a mailbox, to see if any needed to be created, it was accidentally
3656 using path names with one or more superfluous leading slashes; tracing
3657 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3659 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3660 contents are added to the log line that is written for every discarded
3661 recipient. (Previously a log_message setting was ignored.)
3663 36. The ${quote: operator now quotes the string if it is empty.
3665 37. The install script runs exim in order to find its version number. If for
3666 some reason other than non-existence or emptiness, which it checks, it
3667 could not run './exim', it was installing it with an empty version number,
3668 i.e. as "exim-". This error state is now caught, and the installation is
3671 38. An argument was missing from the function that creates an error message
3672 when Exim fails to connect to the socket for saslauthd authentication.
3673 This could cause Exim to crash, or give a corrupted message.
3675 39. Added isip, isip4, and isip6 to ${if conditions.
3677 40. The ACL variables $acl_xx are now saved with the message, and can be
3678 accessed later in routers, transports, and filters.
3680 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3681 strings in the file are not string-expanded.
3683 42. If a MAIL command specified a SIZE value that was too large to fit into an
3684 int variable, the check against message_size_limit failed. Such values are
3685 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3686 day this will have to be increased, but I don't think I want to be around
3687 when emails are that large.
3694 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3695 iconv() is not standard in FreeBSD.
3697 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3698 IPv6 enabled. The observed symptom was a segmentation fault on return from
3699 the function os_common_find_running_interfaces() in src/os.c.
3701 3. In the check_special_case() function in daemon.c I had used "errno" as an
3702 argument name, which causes warnings on some systems. This was basically a
3703 typo, since it was named "eno" in the comments!
3705 4. The code that waits for the clock to tick (at a resolution of some fraction
3706 of a second) so as to ensure message-id uniqueness was always waiting for
3707 at least one whole tick, when it could have waited for less. [This is
3708 almost certainly not relevant at current processor speeds, where it is
3709 unlikely to ever wait at all. But we try to future-proof.]
3711 5. The function that sleeps for a time interval that includes fractions of a
3712 second contained a race. It did not block SIGALRM between setting the
3713 timer, and suspending (a couple of lines later). If the interval was short
3714 and the sigsuspend() was delayed until after it had expired, the suspension
3715 never ended. On busy systems this could lead to processes getting stuck for
3718 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3719 process, before it forks any delivery processes. The open lookup caching
3720 mechanism meant that the open file or database connection was passed into
3721 the delivery process. The problem was that delivery processes always tidy
3722 up cached lookup data. This could cause a problem for the next delivery
3723 process started by the queue runner, because the external queue runner
3724 process does not know about the closure. So the next delivery process
3725 still has data in the lookup cache. In the case of a file lookup, there was
3726 no problem because closing a file descriptor in a subprocess doesn't affect
3727 the parent. However, if the lookup was caching a connection to a database,
3728 the connection was closed, and the second delivery process was likely to
3729 see errors such as "PGSQL: query failed: server closed the connection
3730 unexpectedly". The problem has been fixed by closing all cached lookups
3731 in a queue runner before running a delivery process.
3733 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3734 seem to have the "const" qualifier which it has on other OS. I've
3737 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3738 *of the same type* (client or server) with the same public name that an
3739 error should be diagnosed.
3741 9. When Exim looked up a host name for an IP address, but failed to find the
3742 original IP address when looking up the host name (a safety check), it
3743 output the message "<ip address> does not match any IP for NULL", which was
3744 confusing, to say the least. The bug was that the host name should have
3745 appeared instead of "NULL".
3747 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3748 user that is built into the binary, and the -C or -D options is used, root
3749 privilege is dropped before the configuration file is read. In addition,
3750 logging is switched to stderr instead of the normal log files. If the
3751 configuration then re-defines the Exim user, the unprivileged environment
3752 is probably not what is expected, so Exim logs a panic warning message (but
3755 However, if deliver_drop_privilege is set, the unprivileged state may well
3756 be exactly what is intended, so the warning has been cut out in that case,
3757 and Exim is allowed to try to write to its normal log files.
3763 1. smtp_return_error_details was not giving details for temporary sender
3764 or receiver verification errors.
3766 2. Diagnose a configuration error if two authenticators have the same public
3769 3. Exim used not to create the message log file for a message until the first
3770 delivery attempt. This could be confusing when incoming messages were held
3771 for policy or load reasons. The message log file is now created at the time
3772 the message is received, and an initial "Received" line is written to it.
3774 4. The automatically generated man page for command line options had a minor
3775 bug that caused no ill effects; however, a more serious problem was that
3776 the procedure for building the man page automatically didn't always
3777 operate. Consequently, release 4.20 contains an out-of-date version. This
3778 shouldn't happen again.
3780 5. When building Exim with embedded Perl support, the script that builds the
3781 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3782 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3784 6. The freeze_tell option was not being used for messages that were frozen on
3785 arrival, either by an ACL or by local_scan().
3787 7. Added the smtp_incomplete_transaction log selector.
3789 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3790 was accepting AUTH without a new EHLO.
3792 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3793 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3796 10. Logging of TCP/IP connections (when configured) now happens in the main
3797 daemon process instead of the child process, so that the TCP/IP connection
3798 count is more accurate (but it can never be perfect).
3800 11. The use of "drop" in a nested ACL was not being handled correctly in the
3801 outer ACL. Now, if condition failure induced by the nested "drop" causes
3802 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3803 or "require"), the connection is dropped.
3805 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3806 that yield "discard" can now be used with an "accept" or a "discard" verb,
3807 but an error is generated for any others (because I can't see a useful way
3808 to define what should happen).
3810 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3811 are now processed in the same way as lines in the Exim configuration file.
3812 In particular, continuation lines are supported.
3814 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3816 15. Added -ti meaning -t -i.
3818 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3819 domains, and warn by logging if others are found.
3821 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3822 returned from GIFCONF when Exim is trying to find the list of interfaces on
3823 a host. The code in os.c has been modified to copy each ifreq to an aligned
3824 structure in all cases.
3826 Also, in some cases, the returned ifreq's were being copied to a 'struct
3827 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3828 means the last couple of bytes of an IPv6 address could be chopped if the
3829 ifreq contained only a normal sockaddr (14 bytes storage).
3831 18. Named domain lists were not supported in the hosts_treat_as_local option.
3832 An entry such as +xxxx was not recognized, and was treated as a literal
3835 19. Ensure that header lines added by a DATA ACL are included in the reject log
3836 if the ACL subsequently rejects the message.
3838 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3839 MD5 (which is deprecated).
3841 21. When testing a filter file using -bf, Exim was writing a message when it
3842 took the sender from a "From " line in the message, but it was not doing so
3843 when it took $return_path from a Return-Path: header line. It now does.
3845 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3846 with a valid header line field name (a series of printing characters
3847 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3849 23. Changed "disc" in the source to "disk" to conform to the documentation and
3850 the book and for uniformity.
3852 24. Ignore Sendmail's -Ooption=value command line item.
3854 25. When execve() failed while trying to run a command in a pipe transport,
3855 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3856 could be confused with a return value of 69 from the command itself. This
3857 has been changed to 127, the value the shell returns if it is asked to run
3858 a non-existent command. The wording for the related log line suggests a
3859 non-existent command as the problem.
3861 26. If received_header_text expands to an empty string, do not add a Received:
3862 header line to the message. (Well, it adds a token one on the spool, but
3863 marks it "old" so that it doesn't get used or transmitted.)
3865 27. Installed eximstats 1.28 (addition of -nt option).
3867 28. There was no check for failure on the call to getsockname() in the daemon
3868 code. This can fail if there is a shortage of resources on the system, with
3869 ENOMEM, for example. A temporary error is now given on failure.
3871 29. Contrary to the C standard, it seems that in some environments, the
3872 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3873 program. Exim now does this explicitly; it affects the formatting of
3874 timestamps using strftime().
3876 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3877 complaints. I've now initialized all the variables, to avoid this.
3879 32. Header lines added by a system filter were not being "seen" during
3880 transport-time rewrites.
3882 33. The info_callback() function passed to OpenSSL is set up with type void
3883 (*)(SSL *, int, int), as described somewhere. However, when calling the
3884 function (actually a macro) that sets it up, the type void(*)() is
3885 expected. I've put in a cast to prevent warnings from picky compilers.
3887 34. If a DNS black list lookup found a CNAME record, but there were no A
3888 records associated with the domain it pointed at, Exim crashed.
3890 35. If a DNS black list lookup returned more than one A record, Exim ignored
3891 all but the first. It now scans all returned addresses if a particular IP
3892 value is being sought. In this situation, the contents of the
3893 $dnslist_value variable are a list of all the addresses, separated by a
3896 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3897 to accept a host name and all its aliases if the forward lookup for any of
3898 them yielded the IP address of the incoming connection. Now it accepts only
3899 those names whose forward lookup yields the correct IP address. Any other
3900 names are discarded. This closes a loophole whereby a rogue DNS
3901 administrator could create reverse DNS records to break through a
3902 wildcarded host restriction in an ACL.
3904 37. If a user filter or a system filter that ran in a subprocess used any of
3905 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3906 the wrong values were passed to the pipe command ($thisaddress had the
3907 value of $0, $0 had the value of $1, etc). This bug was introduced by
3908 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3910 38. Improved the line breaking for long SMTP error messages from ACLs.
3911 Previously, if there was no break point between 40 and 75 characters, Exim
3912 left the rest of the message alone. Two changes have been made: (a) I've
3913 reduced the minimum length to 35 characters; (b) if it can't find a break
3914 point between 35 and 75 characters, it looks ahead and uses the first one
3915 that it finds. This may give the occasional overlong line, but at least the
3916 remaining text gets split now.
3918 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3919 file descriptors might be low, and that setting 1000 would always raise it.
3920 It turns out that in some environments, the limit is already over 1000 and
3921 that lowering it causes trouble. So now Exim takes care not to decrease it.
3923 40. When delivering a message, the value of $return_path is set to $sender_
3924 address at the start of routing (routers may change the value). By an
3925 oversight, this default was not being set up when an address was tested by
3926 -bt or -bv, which affected the outcome if any router or filter referred to
3929 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3930 log only when "message" or "log_message" are set. However, if one of the
3931 conditions was an address verification, or a call to a nested ACL, the
3932 messages generated by the underlying test were being passed through. This
3933 no longer happens. The underlying message is available in $acl_verify_
3934 message for both "message" and "log_message" expansions, so it can be
3935 passed through if needed.
3937 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3938 new expansion $bh_ to give the encoded byte string without charset
3939 translation. Translation happens only if iconv() is available; HAVE_ICONV
3940 indicates this at build time. HEADERS_CHARSET gives the charset to
3941 translate to; headers_charset can change it in the configuration, and
3942 "headers charset" can change it in an individual filter file.
3944 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3945 that creates RFC 2047 encoded "words" labels them as that charset instead
3946 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3947 expansion operator; (ii) when Exim creates a From: line for a local
3948 message; (iii) when a header line is rewritten to include a "phrase" part.
3950 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3951 buggy, causing it to skip the first lines of messages whose message ID
3952 ended in 'D'. This would not have bitten before Exim release 4.14, because
3953 message IDs were unlikely to end in 'D' before then. The effect was to have
3954 incorrect size information for certain domains.
3956 45. #include "config.h" was missing at the start of the crypt16.c module. This
3957 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3960 46. If there was a timeout during a "random" callout check, Exim treated it as
3961 a failure of the random address, and carried on sending RSET and the real
3962 address. If the delay was just some slowness somewhere, the response to the
3963 original RCPT would be taken as a response to RSET and so on, causing
3964 mayhem of various kinds.
3966 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3967 when I implemented it. It didn't allow for the fact that some option values
3968 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3969 the right test for positive values.
3971 48. Domain names in DNS records are case-independent. Exim always looks them up
3972 in lower case. Some resolvers return domain names in exactly the case they
3973 appear in the zone file, that is, they may contain uppercase letters. Not
3974 all resolvers do this - some return always lower case. Exim was treating a
3975 change of case by a resolver as a change of domain, similar to a widening
3976 of a domain abbreviation. This triggered its re-routing code and so it was
3977 trying to route what was effectively the same domain again. This normally
3978 caused routing to fail (because the router wouldn't handle the domain
3979 twice). Now Exim checks for this case specially, and just changes the
3980 casing of the domain that it ultimately uses when it transmits the message
3983 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3986 50. If a filter generated a file delivery with a non-absolute name (possible if
3987 no home directory exists for the router), the forbid_file option was not
3990 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3991 the existing equality matching.
3993 52. Exim was using ints instead of ino_t variables in some places where it was
3994 dealing with inode numbers.
3996 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3997 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3998 called TMPDIR, and if it finds it is different, it changes its value.
4000 54. The smtp_printf() function is now made available to local_scan() so
4001 additional output lines can be written before returning. There is also an
4002 smtp_fflush() function to enable the detection of a dropped connection.
4003 The variables smtp_input and smtp_batched_input are exported to
4006 55. Changed the default runtime configuration: the message "Unknown user"
4007 has been removed from the ACL, and instead placed on the localuser router,
4008 using the cannot_route_message feature. This means that any verification
4009 failures that generate their own messages won't get overridden. Similarly,
4010 the "Unrouteable address" message that was in the ACL for unverifiable
4011 relay addresses has also been removed.
4013 56. Added hosts_avoid_esmtp to the smtp transport.
4015 57. The exicyclog script was not checking for the esoteric option
4016 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4017 will work only if exicyclog is run under the appropriate euid.
4019 58. Following a discussion on the list, the rules by which Exim recognises line
4020 endings on incoming messages have been changed. The -dropcr and drop_cr
4021 options are now no-ops, retained only for backwards compatibility. The
4022 following line terminators are recognized: LF CRLF CR. However, special
4023 processing applies to CR:
4025 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4026 nor a local message in the state where . is a terminator.
4028 (ii) If a bare CR is encountered in a header line, an extra space is added
4029 after the line terminator so as not to end the header. The reasoning
4030 behind this is that bare CRs in header lines are most likely either
4031 to be mistakes, or people trying to play silly games.
4033 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4034 was being incorrectly given as 18 bytes larger than it should have been.
4035 This is a VOB (very old bug).
4037 60. This may never have affected anything current, but just in case it has:
4038 When the local host is found other than at the start of a list of hosts,
4039 the local host, those with the same MX, and any that follow, are discarded.
4040 When the list in question was part of a longer list of hosts, the following
4041 hosts (not currently being processed) were also being discarded. This no
4042 longer happens. I'm not sure if this situation could ever has previously
4045 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4048 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4049 header. This is something that is recommended in a new Internet Draft, and
4050 is something that is documented as being done by Sendmail. There are two
4051 possible values. For messages generated by the autoreply transport, Exim
4054 Auto-Submitted: auto-replied
4056 whereas for all other generated messages (e.g. bounces) it adds
4058 Auto-Submitted: auto-generated
4060 63. The "personal" condition in filters now includes a test for the
4061 Auto-Submitted: header. If it contains the string "auto-" the message it
4062 not considered personal.
4064 64. Added rcpt_include_affixes as a generic transport option.
4066 65. Added queue_only_override (default true).
4068 66. Added the syslog_duplication option.
4070 67. If what should have been the first header line of a message consisted of
4071 a space followed by a colon, Exim was mis-interpreting it as a header line.
4072 It isn't of course - it is syntactically invalid and should therefore be
4073 treated as the start of the message body. The misbehaviour could have
4074 caused a number of strange effects, including loss of data in subsequent
4075 header lines, and spool format errors.
4077 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4078 client host had authenticated. This control can now be exercised by an ACL
4079 for more flexibility.
4081 69. By default, callouts do not happen when testing with -bh. There is now a
4082 variant, -bhc, which does actually run the callout code, including
4083 consulting and updating the callout cache.
4085 70. Added support for saslauthd authentication, courtesy of Alexander
4088 71. If statvfs() failed on the spool or log directories while checking their
4089 size for availability, Exim confusingly gave the error "space shortage".
4090 Furthermore, in debugging mode it crashed with a floating point exception.
4091 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4092 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4093 serious problem, Exim now writes to the main and panic logs when this
4094 happens, with details of the failure. It then refuses to accept the
4095 incoming message, giving the message "spool directory problem" or "log
4096 directory problem" with a 421 code for SMTP messages.
4098 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4099 0, 1, and 2 exist, because some OS complain for execs without them (see
4100 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4101 descriptors. However, the code omitted to check that the open succeeded,
4102 causing mysterious errors if for some reason the permissions on /dev/null
4103 got screwed. Now Exim writes a message to the main and panic logs, and
4104 bombs out if it can't open /dev/null.
4106 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4107 interact so that it is all more flexible. It is supposed to remain
4108 backwards compatible. Also added extra_local_interfaces.
4110 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4111 to bomb out with an assertion failure - to the client this appears as a
4112 connection drop. This problem occurs in the part of the code that was taken
4113 from the Samba project. Fortunately, the assertion is in a very simple
4114 function, so I have fixed this by reproducing the function inline in the
4115 one place where it is called, and arranging for authentication to fail
4116 instead of killing the process with assert().
4118 75. The SPA client code was not working when the server requested OEM rather
4119 than Unicode encoding.
4121 76. Added code to make require_files with a specific uid setting more usable in
4122 the case where statting the file as root fails - usually a non-root-mounted
4123 NFS file system. When this happens and the failure is EACCES, Exim now
4124 forks a subprocess and does the per-uid checking as the relevant uid.
4126 77. Added process_log_path.
4128 78. If log_file_path was not explicitly set, a setting of check_log_space or
4129 check_log_inodes was ignored.
4131 79. If a space check for the spool or log partitions fails, the incident is now
4132 logged. Of course, in the latter case the data may get lost...
4134 80. Added the %p formatting code to string_format() so that it can be used to
4135 print addresses in debug_print(). Adjusted all the address printing in the
4136 debugging in store.c to use %p rather than %d.
4138 81. There was a concern that a line of code in smtp_in.c could overflow a
4139 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4140 initially expressed, the concern was not well-founded, because trailing
4141 spaces are removed early. However, if the trailing spaces were followed by
4142 a NULL, they did not get removed, so the overflow was possible. Two fixes
4145 (a) I re-wrote the offending code in a cleaner fashion.
4146 (b) If an incoming SMTP command contains a NULL character, it is rejected
4149 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4150 runs initgroups(), so that if the Exim user is in any additional groups,
4151 they will be used during message reception.
4157 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob