1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.400 2006/10/02 13:38:17 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
14 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
26 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
29 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
34 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
37 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
38 later the small subsequent patch to fix an introduced bug.
40 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
42 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
44 PH/08 An error is now given if message_size_limit is specified negative.
46 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
47 to be given (somewhat) arbitrary names.
49 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
52 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
53 miscellaneous code fixes
55 PH/10 Added the log_reject_target ACL modifier to specify where to log
58 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
59 hostname. This is wrong, because it relates to the incoming message (and
60 probably the interface on which it is arriving) and not to the outgoing
61 callout (which could be using a different interface). This has been
62 changed to use the value of the helo_data option from the smtp transport
63 instead - this is what is used when a message is actually being sent. If
64 there is no remote transport (possible with a router that sets up host
65 addresses), $smtp_active_hostname is used.
67 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
68 tweaks were necessary in order to get it to work:
69 (a) The code assumed that strncpy() returns a negative number on buffer
70 overflow, which isn't the case. Replaced with Exim's string_format()
72 (b) There were several signed/unsigned issues. I just did the minimum
73 hacking in of casts. There is scope for a larger refactoring.
74 (c) The code used strcasecmp() which is not a standard C function.
75 Replaced with Exim's strcmpic() function.
76 (d) The code set only $1; it now sets $auth1 as well.
77 (e) A simple test gave the error "authentication client didn't specify
78 service in request". It would seem that Dovecot has changed its
79 interface. Fortunately there's a specification; I followed it and
80 changed what the client sends and it appears to be working now.
86 SC/01 Use a glob alias rather than an array ref in eximstats generated
87 parser. This improves both readability and performance.
89 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
90 Don't display local sender or destination tables in eximstats unless
91 there is data to show.
92 Added average volumes into the eximstats top table text output.
94 SC/03 Collect data on the number of addresses (recipients) as well
95 as the number of messages in eximstats.
97 TF/01 Correct an error in the documentation for the redirect router. Exim
98 does (usually) call initgroups() when daemonizing.
100 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
101 with consistent privilege compared to when running as a daemon.
103 TF/03 Note in the spec that $authenticated_id is not set for local
104 submissions from trusted users.
106 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
107 Thanks to Dean Brooks <dean@iglou.com> for the patch.
109 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
110 by adding some example configuration directives to the default
111 configuration file. A little bit of work is required to uncomment the
112 directives and define how usernames and passwords are checked, but
113 there is now a framework to start from.
115 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
116 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
117 without this. I don't know how relevant this is to other LDAP libraries.
119 PH/02 Add the verb name to the "unknown ACL verb" error.
121 PH/03 Magnus Holmgren's patch for filter_prepend_home.
123 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
125 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
126 directory not expanded when it should be if an expanded home directory
127 was set for the address (which is overridden by the transport).
129 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
132 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
133 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
134 because it is too late at that time, and has no effect.
136 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
137 security issue with \' (bugzilla #107). I could not use the
138 PQescapeStringConn() function, because it needs a PGconn value as one of
141 PH/08 When testing addresses using -bt, indicate those final addresses that
142 are duplicates that would not cause an additional delivery. At least one
143 person was confused, thinking that -bt output corresponded to deliveries.
144 (Suppressing duplicates isn't a good idea as you lose the information
145 about possibly different redirections that led to the duplicates.)
147 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
148 systems where poll() doesn't work, in particular OS X.
150 PH/10 Added more information to debugging output for retry time not reached.
152 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
153 operations in malware.c.
155 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
158 PH/13 If write_rejectlog was set false when logging was sent to syslog with
159 syslog_duplication set false, log lines that would normally be written
160 both the the main log and to the reject log were not written to syslog at
163 PH/14 In the default configuration, change the use of "message" in ACL warn
164 statements to "add_header".
166 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
167 not followed by a command (e.g. "seen endif").
169 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
170 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
173 PH/17 Added extra conditions to the default value of delay_warning_condition
177 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
178 { match{$h_precedence:}{(?i)bulk|list|junk} } \
179 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
182 The Auto-Submitted: and various List- headers are standardised, whereas I
183 don't think Precedence: ever was.
185 PH/18 Refactored debugging code in route_finduser() to show more information,
186 in particular, the error code if getpwnam() issues one.
188 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
189 This is apparently needed in addition to the PH/07 change above to avoid
190 any possible encoding problems.
192 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
193 but not after initializing Perl.
195 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
196 output them only if debugging. By default they are written stderr,
197 apparently, which is not desirable.
199 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
202 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
205 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
207 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
208 authenticated or an ident call has been made. Suppress the default
209 values for $authenticated_id and $authenticated_sender (but permit -oMai
210 and -oMas) when testing with -bh.
212 PH/24 Re-jigged the order of the tests in the default configuration so that the
213 tests for valid domains and recipients precede the DNS black list and CSA
214 tests, on the grounds that those ones are more expensive.
216 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
217 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
218 This bug exists in every version of Exim that I still have, right back to
221 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
222 However, an attempt to turn on case-sensitivity in a regex key by
223 including (?-i) didn't work because the subject string was already
224 lowercased, and the effects were non-intuitive. It turns out that a
225 one-line patch can be used to allow (?-i) to work as expected.
231 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
232 other effects) broke the use of negated acl sub-conditions.
234 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
237 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
238 "Deny" causes Exim to reject the incoming connection with a 554 error.
239 Unfortunately, if there is a major crisis, such as a disk failure,
240 tcp-wrappers gives "deny", whereas what one would like would be some
241 kind of temporary error. A kludge has been added to help with this.
242 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
243 554 error is used if errno is still zero or contains ENOENT (which occurs
244 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
247 PH/03 Add -lutil to the default FreeBSD LIBS setting.
249 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
250 errors. Otherwise a message that provokes a temporary error (when other
251 messages do not) can cause a whole host to time out.
253 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
254 addresses were routed directly to files or pipes from a redirect router.
255 File deliveries just didn't batch; pipe deliveries might have suffered
258 PH/06 A failure to get a lock for a hints database would erroneously always say
259 "Failed to get write lock", even when it was really a read lock.
261 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
262 of 0600. This has been changed to use the value of the lockfile_mode
263 option (which defaults to 0600).
265 PH/08 Applied small patch from the Sieve maintainer.
267 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
268 folder from quota calculations, a direct delivery into this folder messed
269 up the contents of the maildirsize file. This was because the regex was
270 used only to exclude .Trash (or whatever) when the size of the mailbox
271 was calculated. There was no check that a delivery was happening into an
272 excluded directory. This bug has been fixed by ignoring all quota
273 processing for deliveries into excluded directories.
275 PH/10 Added the maildirfolder_create_regex option to appendfile.
281 PH/01 The code for finding all the local interface addresses on a FreeBSD
282 system running IPv6 was broken. This may well have applied to all BSD
283 systems, as well as to others that have similar system calls. The broken
284 code found IPv4 interfaces correctly, but gave incorrect values for the
285 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
286 that it would not match correctly against @[] and not recognize the IPv6
289 PH/02 The ipliteral router was not recognizing addresses of the form user@
290 [ipv6:....] because it didn't know about the "ipv6:" prefix.
292 PH/03 Added disable_ipv6.
294 PH/04 Changed $reply_address to use the raw form of the headers instead of the
295 decoded form, because it is most often used to construct To: headers
296 lines in autoreplies, and the decoded form may well be syntactically
297 invalid. However, $reply_address has leading white space removed, and all
298 newlines turned into spaces so that the autoreply transport does not
301 PH/05 If group was specified without a user on a router, and no group or user
302 was specified on a transport, the group from the router was ignored.
304 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
305 for visible compile-time settings that can be used to change these
306 numbers, for those that want even more. Backwards compatibility with old
307 spool files has been maintained. However, going back to a previous Exim
308 release will lost any variables that are in spool files.
310 PH/07 Two small changes when running in the test harness: increase delay when
311 passing a TCP/IP connection to a new process, in case the original
312 process has to generate a bounce, and remove special handling of
313 127.0.0.2 (sic), which is no longer necessary.
315 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
316 be the same on different OS.
318 PH/09 Moved a debug statement in filter processing to avoid a race problem when
321 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
322 whether --show-vars was specified or not
324 JJ/02 exipick: Added support for new ACL variable spool format introduced
327 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
328 syntactically invalid From: or Reply-to: line, and a filter used this to
329 generate an autoreply, and therefore failed to obtain an address for the
330 autoreply, Exim could try to deliver to a non-existent relative file
331 name, causing unrelated and misleading errors. What now happens is that
332 it logs this as a hard delivery error, but does not attempt to create a
335 PH/11 The exinext utility has a -C option for testing purposes, but although
336 the given file was scanned by exinext itself; it wasn't being passed on
337 when Exim was called.
339 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
340 an end-of-file indication when reading a command response.
342 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
343 compiled. In many other places in Exim, IPv6 addresses are always
344 recognized, so I have changed this. It also means that IPv4 domain
345 literals of the form [IPV4:n.n.n.n] are now always recognized.
347 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
348 used if the router is not running as root, for example, when verifying at
349 ACL time, or when using -bh. The debugging output from this situation was
350 non-existent - all you got was a failure to exec. I have made two
353 (a) Failures to set uid/gid, the current directory, or a process leader
354 in a subprocess such as that created by queryprogram now generate
355 suitable debugging ouput when -d is set.
357 (b) The queryprogram router detects when it is not running as root,
358 outputs suitable debugging information if -d is set, and then runs
359 the subprocess without attempting to change uid/gid.
361 PH/15 Minor change to Makefile for building test_host (undocumented testing
364 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
365 additional section of a DNS packet that returns MX or SRV records.
366 Instead, it always explicitly searches for A/AAAA records. This avoids
367 major problems that occur when a DNS server includes only records of one
368 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
369 fixed another bug: if SRV records were looked up and the corresponding
370 address records were *not* found in the additional section, the port
371 values from the SRV records were lost.
373 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
374 using the correct key (the original address) when searching the retry
375 rules in order to find which one to use for generating the retry hint.
377 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
378 adding the default one. Similarly, if it contains a Reply-To: header, the
379 errors_reply_to option, if set, is not used.
381 PH/19 When calculating a retry time, Exim used to measure the "time since
382 failure" by looking at the "first failed" field in the retry record. Now
383 it does not use this if it is later than than the arrival time of the
384 message. Instead it uses the arrival time. This makes for better
385 behaviour in cases where some deliveries succeed, thus re-setting the
386 "first failed" field. An example is a quota failure for a huge message
387 when small messages continue to be delivered. Without this change, the
388 "time since failure" will always be short, possible causing more frequent
389 delivery attempts for the huge message than are intended.
390 [Note: This change was subsequently modified - see PH/04 for 4.62.]
392 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
393 $1, $2, $3) because the numerical variables can be reset during some
394 expansion items (e.g. "match"), thereby losing the authentication data.
396 PH/21 Make -bV show the size of off_t variables so that the test suite can
397 decide whether to run tests for quotas > 2G.
399 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
400 mailbox_size, and mailbox_filecount in the appendfile transport. If a
401 filecount value is greater than 2G or if a quota value is greater than 2G
402 on a system where the size of off_t is not greater than 4, a panic error
405 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
406 never match. The debug and -bh output now contains an explicit error
407 message indicating a malformed IPv4 address or mask.
409 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
410 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
413 PH/25 Do not write to syslog when running in the test harness. The only
414 occasion when this arises is a failure to open the main or panic logs
415 (for which there is an explicit test).
417 PH/26 Added the /no_tell option to "control=freeze".
419 PH/27 If a host name lookup failed very early in a connection, for example, if
420 the IP address matched host_lookup and the reverse lookup yielded a name
421 that did not have a forward lookup, an error message of the form "no IP
422 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
423 could be logged. Now it outputs the IP address instead of "NULL".
425 PH/28 An enabling patch from MH: add new function child_open_exim2() which
426 allows the sender and the authenticated sender to be set when
427 submitting a message from within Exim. Since child_open_exim() is
428 documented for local_scan(), the new function should be too.
430 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
431 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
432 results in an empty string is now treated as unset.
434 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
436 PH/31 Added sender_verify_fail logging option.
438 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
439 needed by RFC 822 but not by RFC 2822 was commented out. I have now
440 tidied the source and removed it altogether.
442 PH/33 When a queue run was abandoned because the load average was too high, a
443 log line was always written; now it is written only if the queue_run log
444 selector is set. In addition, the log line for abandonment now contains
445 information about the queue run such as the pid. This is always present
446 in "start" and "stop" lines but was omitted from the "abandon" line.
448 PH/34 Omit spaces between a header name and the colon in the error message that
449 is given when verify = headers_syntax fails (if there are lots of them,
450 the message gets confusing).
452 PH/35 Change the default for dns_check_names_pattern to allow slashes within
453 names, as there are now some PTR records that contain slashes. This check
454 is only to protect against broken name servers that fall over on strange
455 characters, so the fact that it applies to all lookups doesn't matter.
457 PH/36 Now that the new test suite is complete, we can remove some of the
458 special code in Exim that was needed for the old test suite. For example,
459 sorting DNS records because real resolvers return them in an arbitrary
460 order. The new test suite's fake resolver always returns records in the
463 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
464 bounces) except when queue_only is set, to avoid logging races between
465 the different processes.
467 PH/38 Panic-die if .include specifies a non-absolute path.
469 PH/39 A tweak to the "H" retry rule from its user.
471 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
472 a label. They prevented compilation on older perls.
474 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
475 a warning to be raised on newish perls.
477 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
478 on queue. Changes to match documented behaviour of showing count of
479 messages matching specified criteria.
481 PH/40 Changed the default ident timeout from 30s to 5s.
483 PH/41 Added support for the use of login_cap features, on those BSD systems
484 that have them, for controlling the resources used by pipe deliveries.
486 PH/42 The content-scanning code uses fopen() to create files in which to put
487 message data. Previously it was not paying any attention to the mode of
488 the files. Exim runs with umask(0) because the rest of the code creates
489 files with open(), and sets the required mode explicitly. Thus, these
490 files were ending up world-writeable. This was not a big issue, because,
491 being within the spool directory, they were not world-accessible. I have
492 created a function called modefopen, which takes an additional mode
493 argument. It sets umask(777), creates the file, chmods it to the required
494 mode, then resets the umask. All the relevant calls to fopen() in the
495 content scanning code have been changed to use this function.
497 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
498 to 24 hours. This avoids potential overflow problems when processing G
499 and H retry rules. I suspect nobody ever tinkers with this value.
501 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
503 PH/45 When the plaintext authenticator is running as a client, the server's
504 challenges are checked to ensure they are valid base64 strings. By
505 default, the authentication attempt is cancelled if an invalid string is
506 received. Setting client_ignore_invalid_base64 true ignores these errors.
507 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
508 they are received. Thus, the responses can be made to depend on the
509 challenges. If an invalid string is ignored, an empty string is placed in
512 PH/46 Messages that are created by the autoreply transport now contains a
513 References: header, in accordance with RFCs 2822 and 3834.
515 PH/47 Added authenticated_sender_force to the smtp transport.
517 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
519 PH/49 Installed latest patch from the Sieve maintainer.
521 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
522 was also set, the appendfile transport was unnecessarily scanning a
523 directory of message files (e.g. for maildir delivery) to find the count
524 of files (along with the size), even though it did not need this
525 information. It now does the scan only if it needs to find either the
526 size of the count of files.
528 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
530 PH/52 Two bugs concerned with error handling when the smtp transport is
533 (i) Exim was not creating retry information for temporary errors given
534 for individual recipients after the DATA command when the smtp transport
535 was used in LMTP mode. This meant that they could be retried too
536 frequently, and not timed out correctly.
538 (ii) Exim was setting the flag that allows error details to be returned
539 for LMTP errors on RCPT commands, but not for LMTP errors for individual
540 recipients that were returned after the DATA command.
542 PH/53 This is related to PH/52, but is more general: for any failing address,
543 when detailed error information was permitted to be returned to the
544 sender, but the error was temporary, then after the final timeout, only
545 "retry timeout exceeded" was returned. Now it returns the full error as
546 well as "retry timeout exceeded".
548 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
549 do this, and (what is worse) MTAs that accept it.
551 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
552 will now be deprecated.
554 PH/56 New os.c-cygwin from the Cygwin maintainer.
556 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
557 formats (previously only available in exim formats via -bpr, -bpru,
558 and -bpra. Now also available in native and exiqgrep formats)
560 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
561 with very large, slow to parse queues
563 JJ/08 exipick: added ! as generic prefix to negate any criteria format
565 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
567 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
568 responses to authentication challenges, though it was showing the
569 challenges; (ii) I've removed the CR characters from the debug output for
572 PH/58 Allow for the insertion of a newline as well as a space when a string
573 is turned into more than one encoded-word during RFC 2047 encoding. The
574 Sieve code now uses this.
576 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
577 data_4xx, lost_connection, tls_required.
579 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
580 message was being sent as an SMTP response.
582 PH/61 Add -l and -k options to exicyclog.
584 PH/62 When verifying, if an address was redirected to one new address, so that
585 verification continued, and the new address failed or deferred after
586 having set something in $address_data, the value of $address_data was not
587 passed back to the ACL. This was different to the case when no
588 redirection occurred. The value is now passed back in both cases.
590 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
591 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
592 use login_cap.h, so on its own it isn't the distinguishing feature. The
593 new name refers directly to the setclassresources() function.
595 PH/65 Added configuration files for NetBSD3.
597 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
599 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
600 is preferred over IPv4.
602 PH/68 The bounce_return_message and bounce_return_body options were not being
603 honoured for bounces generated during the reception of non-SMTP messages.
604 In particular, this applied to messages rejected by the ACL. This bug has
605 been fixed. However, if bounce_return_message is true and bounce_return_
606 body is false, the headers that are returned for a non-SMTP message
607 include only those that have been read before the error was detected.
608 (In the case of an ACL rejection, they have all been read.)
610 PH/69 The HTML version of the specification is now built in a directory called
611 spec_html instead of spec.html, because the latter looks like a path with
612 a MIME-type, and this confuses some software.
614 PH/70 Catch two compiler warnings in sieve.c.
616 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
617 function verify_get_ident() calls ip_connect() to connect a socket, but
618 if the "connect()" function timed out, ip_connect() used to close the
619 socket. However, verify_get_ident() also closes the socket later, and in
620 between Exim writes to the log, which may get opened at this point. When
621 the socket was closed in ip_connect(), the log could get the same file
622 descriptor number as the socket. This naturally causes chaos. The fix is
623 not to close the socket in ip_connect(); the socket should be closed by
624 the function that creates it. There was only one place in the code where
625 this was missing, in the iplookup router, which I don't think anybody now
626 uses, but I've fixed it anyway.
628 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
629 well as to direct DNS lookups. Otherwise the handling of names in host
630 lists is inconsistent and therefore confusing.
636 PH/01 Two changes to the default runtime configuration:
638 (1) Move the checks for relay_from_hosts and authenticated clients from
639 after to before the (commented out) DNS black list checks.
641 (2) Add control=submission to the relay_from_hosts and authenticated
642 clients checks, on the grounds that messages accepted by these
643 statements are most likely to be submissions.
645 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
647 (1) Generate an error if the third argument for the ${prvs expansion is
650 (2) Treat a missing third argument of ${prvscheck as if it were an empty
653 (3) Reset the variables that are obtained from the first argument of
654 ${prvscheck and used in the second argument before leaving the code,
655 because their memory is reclaimed, so using them afterwards may do
658 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
659 one (it's much easier than Tom thought :-).
661 (5) Because of (4), we can now allow for the use of $prvscheck_result
662 inside the third argument.
664 PH/03 For some reason, the default setting of PATH when running a command from
665 a pipe transport was just "/usr/bin". I have changed it to
668 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
669 anything to be listed in the output from -bV.
671 PH/05 When a filter generated an autoreply, the entire To: header line was
672 quoted in the delivery log line, like this:
674 => >A.N.Other <ano@some.domain> <original@ddress> ...
676 This has been changed so that it extracts the operative address. There
677 may be more than one such address. If so, they are comma-separated, like
680 => >ano@some.domain,ona@other.domain <original@ddress> ...
682 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
683 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
684 not being looked up in the rDNS to get a host name, Exim was showing the
685 IP address twice in Received: lines, even though the IP addresses were
686 identical. For example:
688 Received: from [1.2.3.4] (helo=[1.2.3.4])
690 However, if the real host name was known, it was omitting the HELO data
691 if it matched the actual IP address. This has been tidied up so that it
692 doesn't show the same IP address twice.
694 PH/07 When both +timestamp and +memory debugging was on, the value given by
695 $tod_xxx expansions could be wrong, because the tod_stamp() function was
696 called by the debug printing, thereby overwriting the timestamp buffer.
697 Debugging no longer uses the tod_stamp() function when +timestamp is set.
699 PH/08 When the original message was included in an autoreply transport, it
700 always said "this is a copy of the message, including all the headers",
701 even if body_only or headers_only was set. It now gives an appropriate
704 PH/09 Applied a patch from the Sieve maintainer which:
706 o fixes some comments
707 o adds the (disabled) notify extension core
708 o adds some debug output for the result of if/elsif tests
709 o points to the current vacation draft in the documentation
710 and documents the missing references header update
714 o fixes a bug in processing the envelope test (when testing
715 multiple envelope elements, the last element determinted the
718 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
719 Electronic Mail") by including:
721 Auto-submitted: auto-generated
723 in the messages that it generates (bounce messages and others, such as
724 warnings). In the case of bounce messages for non-SMTP mesages, there was
725 also a typo: it was using "Auto_submitted" (underscore instead of
726 hyphen). Since every message generated by Exim is necessarily in response
727 to another message, thes have all been changed to:
729 Auto-Submitted: auto-replied
731 in accordance with these statements in the RFC:
733 The auto-replied keyword:
735 - SHOULD be used on messages sent in direct response to another
736 message by an automatic process,
738 - MUST NOT be used on manually-generated messages,
740 - MAY be used on Delivery Status Notifications (DSNs) and Message
741 Disposition Notifications (MDNs),
743 - MUST NOT be used on messages generated by automatic or periodic
744 processes, except for messages which are automatic responses to
747 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
748 to the default Received: header definition.
750 PH/12 Added log selector acl_warn_skipped (default on).
752 PH/13 After a successful wildlsearch lookup, discard the values of numeric
753 variables because (a) they are in the wrong storage pool and (b) even if
754 they were copied, it wouldn't work properly because of the caching.
756 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
757 checking when decoding. Apparently there are clients that generate
758 overlong encoded strings. Why am I not surprised?
760 PH/15 If the first argument of "${if match_address" was not empty, but did not
761 contain an "@" character, Exim crashed. Now it writes a panic log message
762 and treats the condition as false.
764 PH/16 In autoreply, treat an empty string for "once" the same as unset.
766 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
767 extension "envelope-auth". The code is finished and in agreement with
768 other implementations, but there is no documentation so far and in fact,
769 nobody wrote the draft yet. This extension is currently #undef'ed, thus
770 not changing the active code.
772 Print executed "if" and "elsif" statements when debugging is used. This
773 helps a great deal to understand what a filter does.
775 Document more things not specified clearly in RFC3028. I had all this
776 sorted out, when out of a sudden new issues came to my mind. Oops."
778 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
781 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
782 canonical form (as documented). However, after a host name lookup from
783 the IP address, check_host() was doing a simple string comparison with
784 addresses acquired from the DNS when checking that the found name did
785 have the original IP as one of its addresses. Since any found IPv6
786 addresses are likely to be in abbreviated form, the comparison could
787 fail. Luckily, there already exists a function for doing the comparison
788 by converting both addresses to binary, so now that is used instead of
791 PH/20 There was another similar case to PH/19, when a complete host name was
792 given in a host list; looking up its IP address could give an abbreviated
793 form, whereas the current host's name might or might not be abbreviated.
794 The same fix has been applied.
800 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
801 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
804 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
805 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
807 PH/03 Typo: missing ".o" in src/pcre/Makefile.
809 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
810 header line, restrict the check to what is listed in RFCs 2369 and 2929.
811 Also, for "Auto-Submitted", treat anything other than "no" as
812 non-personal, in accordance with RFC 3834. (Previously it treated
813 anything starting "auto-" as non-personal.)
815 TF/01 The control=submission/name=... option had a problem with syntax
816 errors if the name included a slash character. The /name= option
817 now slurps the rest of the string, so it can include any characters
818 but it must come last in the list of options (after /sender_retain
821 PH/05 Some modifications to the interface to the fake nameserver for the new
829 TK/01 Added the "success_on_redirect" address verification option. See
830 NewStuff for rationale and an example.
832 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
834 PH/02 Patch to exigrep to allow it to work on syslog lines.
836 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
837 fread() to skip over the body file's header line, because in Cygwin the
838 header line is locked and is inaccessible.
840 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
841 co-exist for some time) to make it clear that it is the Exim ID that is
842 referenced, not the Message-ID: header line.
844 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
845 string_format() function, because snprintf() does not exist on all
848 PH/06 The use of forbid_filter_existstest now also locks out the use of the
849 ${stat: expansion item.
851 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
852 protocol synchronization error", to keep the pedants happy.
854 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
855 well as for IRIX systems, when gcc is being used. See the host.c source
858 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
860 PH/10 Named domain lists were not working if used in a queue_smtp_domains
863 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
864 transport and to the smtp transport in LMTP mode.
866 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
868 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
869 run a filter in a subprocess. This could lead to confusion in subsequent
870 lookups in the parent process. There should also be a search_tidyup() at
871 the end of the subprocess.
873 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
874 only if the host matched helo_try_verify_hosts, which caused the
875 verification to occur when the EHLO/HELO command was issued. The ACL just
876 tested the remembered result. Now, if a previous verification attempt has
877 not happened, "verify = helo" does it there and then.
879 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
881 TK/03 Fix log output including CR from clamd.
883 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
884 exist provoked a memory error which could cause a segfault.
886 PH/15 Installed PCRE 6.2
888 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
890 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
891 of the problem. Specifically, suggested +O2 rather than +O1 for the
894 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
896 PH/20 If a delivery was routed to a non-standard port by means of an SRV
897 record, the port was not correctly logged when the outgoing_port log
898 selector was set (it logged the transort's default port).
900 PH/21 Added support for host-specific ports to manualroute, queryprogram,
901 fallback_hosts, and "hosts" in the smtp transport.
903 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
904 host errors such as "Connection refused".
906 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
907 authentication with radiusclient 0.4.9:
909 - Error returned from rc_read_config was caught wrongly
910 - Username/password not passed on to radius server due to wrong length.
912 The presumption is that some radiusclient API changes for 4.51/PH/17
913 were not taken care of correctly. The code is still untested by me (my
914 Linux distribution still has 0.3.2 of radiusclient), but it was
915 contributed by a Radius user.
917 PH/24 When doing a callout, the value of $domain wasn't set correctly when
918 expanding the "port" option of the smtp transport.
920 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
921 while reading a MIME header. Thanks to Tom Hughes for a patch.
923 PH/24 Include config.h inside local_scan.h so that configuration settings are
926 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
927 that in an ACL for RCPT (for example), you can examine exactly what was
930 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
931 commands, but it was not correctly comparing the address with the actual
932 client host address. Thus, it would show the EHLO address in Received:
933 header lines when this was not necessary.
935 PH/27 Added the % operator to ${eval:}.
937 PH/28 Exim tries to create and chdir to its spool directory when it starts;
938 it should be ignoring failures (because with -C, for example, it has lost
939 privilege). It wasn't ignoring creation failures other than "already
942 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
945 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
946 because an input line was too long, either on its own, or by virtue of
947 too many continuations, the temporary file was not being removed, and the
948 return code was incorrect.
950 PH/31 Missing "BOOL" in function definition in filtertest.c.
952 PH/32 Applied Sieve patches from the maintainer.
954 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
956 PH/33 Added "verify = not_blind".
958 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
959 Local/Makefile (with some defaults set). These are used in built scripts
960 such as exicyclog, but they have never been used in the exim_install
961 script (though there are many overriding facilities there). I have
962 arranged that the exim_install script now takes note of these two
965 PH/35 Installed configuration files for Dragonfly.
967 PH/36 When a locally submitted message by a trusted user did not contain a
968 From: header, and the sender address was obtained from -f or from an SMTP
969 MAIL command, and the trusted user did not use -F to supply a sender
970 name, $originator_name was incorrectly used when constructing a From:
971 header. Furthermore, $originator_name was used for submission mode
972 messages from external hosts without From: headers in a similar way,
973 which is clearly wrong.
975 PH/37 Added control=suppress_local_fixups.
977 PH/38 When log_selector = +received_sender was set, and the addition of the
978 sender made the log line's construction buffer exactly full, or one byte
979 less than full, an overflow happened when the terminating "\n" was
982 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
983 when the result of a list match is failure because a DNS lookup failed.
985 PH/40 RM_COMMAND is now used in the building process.
987 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
988 the "build-* directories that it finds.
990 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
991 address in a domain literal was a prefix of an interface address.
993 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
994 when verifying a sender address, unless rewrite_headers is false.
996 PH/44 Wrote a long comment about why errors_to addresses are verified as
997 recipients, not senders.
999 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1000 the ratelimit ACL was added.
1002 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1004 PH/46 Added extra information about PostgreSQL errors to the error string.
1006 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1007 avoiding the need to install special zones in a real server. This is
1008 backwards compatible; if it can't find the fake resolver, it drops back.
1009 Thus, both old and new test suites can be run.
1011 TF/02 Added util/ratelimit.pl
1013 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1014 clock is set back in time.
1016 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1017 Candler <B.Candler@pobox.com>.
1019 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1020 OK for addresses that are the result of redirections.
1022 PH/48 A number of further additions for the benefit of the new test suite,
1023 including a fake gethostbyname() that interfaces to the fake DNS resolver
1026 TF/06 The fix for widen_domains has also been applied to qualify_single and
1027 search_parents which are the other dnslookup options that can cause
1030 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1033 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1035 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1036 service" if it encountered an MX record with an empty target hostname.
1037 The message is now "an MX or SRV record indicated no SMTP service".
1039 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1040 if the DNS of the sending site is misconfigured. This is quite a
1041 common situation. This change restores the behaviour of treating a
1042 helo verification defer as a failure.
1044 PH/51 If self=fail was set on a router, the bounce message did not include the
1045 actual error message.
1051 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1053 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1054 command itself ended in error, the underlying message about the transport
1055 filter timeout was being overwritten with the pipe command error. Now the
1056 underlying error message should be appended to the second error message.
1058 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1060 PH/02 Reduce the amount of output that "make" produces by default. Full output
1061 can still be requested.
1063 PH/03 The warning log line about a condition test deferring for a "warn" verb
1064 was being output only once per connection, rather than after each
1065 occurrence (because it was using the same function as for successful
1066 "warn" verbs). This seems wrong, so I have changed it.
1068 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1069 it should not have, which might have caused a crash in the right
1070 circumstances, but probably never did.
1072 PH/04 Installed a modified version of Tony Finch's patch to make submission
1073 mode fix the return path as well as the Sender: header line, and to
1074 add a /name= option so that you can make the user's friendly name appear
1077 TF/03 Added the control = fakedefer ACL modifier.
1079 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1080 Mark Lowes for thorough testing.
1082 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1084 TK/03 Merged latest SRS patch from Miles Wilton.
1086 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1087 with the definition in sysexits.h (which is #included earlier).
1088 Fortunately, Exim does not actually use EX_OK. The code used to try to
1089 preserve the sysexits.h value, by assumimg that macro definitions were
1090 scanned for macro replacements. I have been disabused of this notion,
1091 so now the code just undefines EX_OK before #including unistd.h.
1093 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1094 in the smtp transport. When a block could not be written in a single
1095 write() function, the timeout was being re-applied to each part-write.
1096 This seems wrong - if the receiver was accepting one byte at a time it
1097 would take for ever. The timeout is now adjusted when this happens. It
1098 doesn't have to be particularly precise.
1100 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1101 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1103 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1104 without a domain if the check to <postmaster@domain> fails.
1106 SC/01 Eximstats: added -xls and the ability to specify output files
1107 (patch written by Frank Heydlauf).
1109 SC/02 Eximstats: use FileHandles for outputing results.
1111 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1113 SC/04 Eximstats: fixed display of large numbers with -nvr option
1115 SC/05 Eximstats: fixed merging of reports with empty tables.
1117 SC/06 Eximstats: added the -include_original_destination flag
1119 SC/07 Eximstats: removed tabs and trailing whitespace.
1121 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1123 TK/06 MBOX spool code: Add real "From " MBOX separator line
1124 so the .eml file is really in mbox format (even though
1125 most programs do not really care). Patch from Alex Miller.
1127 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1128 The latter is generated from $received_to and is only set if the
1129 message has one envelope recipient. SA can use these headers,
1130 obviously out-of-the-box. Patch from Alex Miller.
1132 PH/08 The ${def test on a variable was returning false if the variable's
1133 value was "0", contrary to what the specification has always said!
1134 The result should be true unless the variable is empty.
1136 PH/09 The syntax error of a character other than { following "${if
1137 def:variable_name" (after optional whitespace) was not being diagnosed.
1138 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1139 accidental colon was present, for example, could give incorrect results.
1141 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1142 result is used (not including appendfile, where other changes are about
1145 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1146 This involved changing a lot of size variables from int to off_t. It
1147 should work with maildirs and everything.
1149 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1150 spamd dying while we are connected to it.
1152 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1155 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1156 function for BDB 4.3.
1158 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1160 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1161 the sources for PCRE have been re-arranged and more files are now
1164 PH/15 The code I had for printing potentially long long variables in PH/11
1165 above was not the best (it lost precision). The length of off_t variables
1166 is now inspected at build time, and an appropriate printing format (%ld
1167 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1168 to be "long long int" or "long int". This is needed for the internal
1169 formatting function string_vformat().
1171 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1172 the configuration file to be ":syslog", then the script "guesses" where
1173 the logs files are, rather than using the compiled in default. In our
1174 case the guess is not the same as the compiled default, so the script
1175 suddenly stopped working when I started to use syslog. The patch checks
1176 to see if log_file_path is "". If so, it attempts to read it from exim
1177 with no configuration file to get the compiled in version, before it
1178 falls back to the previous guessing code."
1180 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1181 implementing BATV in an Exim configuration. See NewStuff for the gory
1184 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1185 Makefile that are specific to HP-UX.
1187 PH/18 If the "use_postmaster" option was set for a recipient callout together
1188 with the "random" option, the postmaster address was used as the MAIL
1189 FROM address for the random test, but not for the subsequent recipient
1190 test. It is now used for both.
1192 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1193 patch removes a few documentation additions to RFC 3028, because the
1194 latest draft now contains them. It adds the new en;ascii-case comparator
1195 and a new error check for 8bit text in MIME parts. Comparator and
1196 require names are now matched exactly. I enabled the subaddress
1197 extension, but it is not well tested yet (read: it works for me)."
1199 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1200 rework some of the code of TK/09 above to avoid the hardwired use of
1201 "%lld" and "long long". Replaced the call to snprintf() with a call to
1204 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1205 records point to non-existent hosts", "retry timeout exceeded", and
1206 "retry time not reached for any host after a long failure period".
1208 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1209 experimental DomainKeys support:
1211 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1212 (2) On an error such as an illegally used "control", the wrong name for
1213 the control was given.
1215 These problems did NOT occur unless DomainKeys support was compiled.
1217 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1219 PH/24 Added ${if match_ip condition.
1221 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1222 databases so that it will be absolutely obvious if a crash occurs in the
1223 DB library. This is a regular occurrence (often caused by mis-matched
1226 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1227 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1228 picked up on a user's system that detects such things. There doesn't seem
1229 to be a gcc warning option for this - only an attribute that has to be
1230 put on the function's prototype. It seems that in Fedora Core 4 they have
1231 set this on a number of new functions. No doubt there will be more in due
1234 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1235 specify a transport. However, if an address that was verified by such a
1236 router was the subject of a callout, Exim crashed because it tried to
1237 read the rcpt_include_affixes from the non-existent transport. Now it
1238 just assumes that the setting of that option is false. This bug was
1239 introduced by 4.51/PH/31.
1241 PH/28 Changed -d+all to exclude +memory, because that information is very
1242 rarely of interest, but it makes the output a lot bigger. People tend to
1243 do -d+all out of habit.
1245 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1246 code in os-type was giving problems when libc.so lives in lib64, like on
1249 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1250 aren't the modern standard, and it seems that some systems' include files
1251 don't always have them. Exim was already checking for some of the newer
1252 ones like T_AAAA, and defining it itself. I've added checks for all the
1253 record types that Exim uses.
1255 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1256 not automatically generating a new one, as it is supposed to. This
1257 prevented TLS from working. If the file did exist, but contained invalid
1258 data, a new version was generated, as expected. It was only the case of a
1259 non-existent file that was broken.
1261 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1262 with a change in libdomainkeys > 0.64.
1264 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1265 from DNS. If the selector record carries the flag, it now has
1266 precedence over the domain-wide flag.
1268 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1270 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1271 the use of a port name that isn't defined in /etc/services) occurred, the
1272 message was deferred as in a normal delivery, and thus remained on the
1273 spool, instead of being failed because of the mua_wrapper setting. This
1274 is now fixed, and I tidied up some of the mua_wrapper messages at the
1277 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1278 the messages in a hash of arrays rather than using individual hashes.
1279 This is a bit cleaner and results in dramatic memory savings, albeit
1280 at a slight CPU cost.
1282 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1283 as requested by Marc Sherman.
1285 SC/10 Eximstats: added histograms for user specified patterns as requested
1288 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1290 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1291 fopen() in the content-scanning modules that did not already have it.
1297 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1298 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1300 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1302 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1304 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1305 to test Sieve filters that use "vacation".
1307 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1308 that changes the way the GnuTLS parameters are stored in the cache file.
1309 The new format can be generated externally. For backward compatibility,
1310 if the data in the cache doesn't make sense, Exim assumes it has read an
1311 old-format file, and it generates new data and writes a new file. This
1312 means that you can't go back to an older release without removing the
1315 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1316 work if there are any delivery delays because "one_time" forces the
1317 parent to be marked "delivered", so its unseen clone is never tried
1318 again. For this reason, Exim now forbids the simultaneous setting of
1321 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1322 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1323 entry if you want to know the details. The fix, however, neglected to
1324 consider the case when local delivery batching is involved. The test for
1325 "previously delivered" was not happening when checking to see if an
1326 address could be batched with a previous (undelivered) one; under
1327 certain circumstances this could lead to multiple deliveries to the same
1330 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1331 in its include files, and this causes problems building Exim.
1333 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1334 header_syntax) but Exim was just ignoring anything given after a slash.
1335 In particular, this caused confusion with an attempt to use "verify =
1336 reverse_host_lookup/defer_ok". An error is now given when options are
1337 supplied for verify items that do not have them. (Maybe reverse_host_
1338 lookup should have a defer_ok option, but that's a different point.)
1340 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1341 defined by RFC 821) to 2048, because there were problems with some AUTH
1342 commands, and RFC 1869 says the size should be increased for extended
1343 SMTP commands that take arguments.
1345 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1348 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1349 "unknown" error; now it says that the functionality isn't in the binary.
1351 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1352 an address' error message when a string expansion fails (syntax or
1353 whatever). Otherwise the password may appear in the log. Following change
1354 PH/42 below, there is no longer a chance of it appearing in a bounce
1357 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1359 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1360 of its addresses was ever tried. (Bugzilla bug #2.)
1362 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1363 the result incorrectly in the debug output. (It correctly added a newline
1364 to what was transported.)
1366 TF/01 Added $received_time.
1368 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1369 commented out examples of how to interface to a virus scanner and to
1370 SpamAssassin. Also added commented examples of av_scanner and
1371 spamd_address settings.
1373 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1374 and controls are allowed in which ACLs. There were a couple of minor
1375 errors. Some of the entries in the conditions table (which is a table of
1376 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1377 negation of where the condition IS allowed.
1379 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1381 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1382 header file does not have a version number, so I've had to invent a new
1383 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1384 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1385 radiusclient), but it was contributed by a Radius user.
1387 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1388 files or directories, for OpenSSL.
1390 PH/19 When an Exim process that is running as root has to create an Exim log
1391 file, it does so in a subprocess that runs as exim:exim so as to get the
1392 ownership right at creation (otherwise, other Exim processes might see
1393 the file with the wrong ownership). There was no test for failure of this
1394 fork() call, which would lead to the process getting stuck as it waited
1395 for a non-existent subprocess. Forks do occasionally fail when resources
1396 run out. I reviewed all the other calls to fork(); they all seem to check
1399 PH/20 When checking for unexpected SMTP input at connect time (before writing
1400 the banner), Exim was not dealing correctly with a non-positive return
1401 from the read() function. If the client had disconnected by this time,
1402 the result was a log entry for a synchronization error with an empty
1403 string after "input=" when read() returned zero. If read() returned -1
1404 (an event I could not check), uninitialized data bytes were printed.
1405 There were reports of junk text (parts of files, etc) appearing after
1408 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1410 PH/22 Added support for macro redefinition, and (re)definition in between
1411 driver and ACL definitions.
1413 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1414 forgetting to use the resulting value; it was using the unexpanded value.
1416 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1417 hadn't been configured. The fix is from Juergen Kreileder, who
1418 understands it better than I do:
1420 "Here's what I see happening with three configured cyrus_sasl
1421 authenticators configured (plain, login, cram-md5):
1423 On startup auth_cyrus_sasl_init() gets called for each of these.
1424 This means three calls to sasl_listmech() without a specified mech_list.
1425 => SASL tests which mechs of all available mechs actually work
1426 => three warnings about OTP not working
1427 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1429 With the patch, sasl_listmech() also gets called three times. But now
1430 SASL's mech_list option is set to the server_mech specified in the the
1431 authenticator. Or in other words, the answer from sasl_listmech()
1432 gets limited to just the mech you're testing for (which is different
1434 => the return list contains just 'plain' or 'login', 'cram-md5' or
1435 nothing depending on the value of ob->server_mech.
1437 I've just tested the patch: Authentication still works fine,
1438 unavailable mechs specified in the exim configuration are still
1439 caught, and the auth.log warnings about OTP are gone."
1441 PH/25 When debugging is enabled, the contents of the command line are added
1442 to the debugging output, even when log_selector=+arguments is not
1445 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1446 answer is "GNU", and only if the return is "GNU/something" is the answer
1449 PH/27 $acl_verify_message is now set immediately after the failure of a
1450 verification in an ACL, and so is available in subsequent modifiers. In
1451 particular, the message can be preserved by coding like this:
1453 warn !verify = sender
1454 set acl_m0 = $acl_verify_message
1456 Previously, $acl_verify_message was set only while expanding "message"
1457 and "log_message" when a very denied access.
1459 PH/28 Modified OS/os.c-Linux with
1461 -#ifndef OS_LOAD_AVERAGE
1462 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1464 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1465 nomenclature these days.)
1467 PH/29 Installed patch from the Sieve maintainer that adds the options
1468 sieve_useraddress and sieve_subaddress to the redirect router.
1470 PH/30 In these circumstances:
1471 . Two addresses routed to the same list of hosts;
1472 . First host does not offer TLS;
1473 . First host accepts first address;
1474 . First host gives temporary error to second address;
1475 . Second host offers TLS and a TLS session is established;
1476 . Second host accepts second address.
1477 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1478 and peerdn, if requested) that were in fact used only for the second
1481 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1482 attention to any local part prefix or suffix that was matched by the
1483 router that accepted the address. It now behaves in the same way as it
1484 does for delivery: the affixes are removed from the local part unless
1485 rcpt_include_affixes is set on the transport.
1487 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1488 timeout during the DATA phase of an incoming message.
1490 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1491 applied a patch sanctioned by the Sieve maintainer.
1493 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1494 the uid or gid is negative. A case of a negative gid caused this to be
1495 noticed. The fix allows for either to be negative.
1497 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1498 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1501 PH/36 The change PH/12 above was broken. Fixed it.
1503 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1504 the grounds that routing the same address twice would always produce the
1505 same answer. This might have been true once, but it is certainly no
1506 longer true now. Routing a child address may depend on the previous
1507 routing that produced that child. Some complicated redirection strategies
1508 went wrong when messages had multiple recipients, and made Exim's
1509 behaviour dependent on the order in which the addresses were given.
1511 I have moved the duplicate checking until after the routing is complete.
1512 Exim scans the addresses that are assigned to local and remote
1513 transports, and removes any duplicates. This means that more work will be
1514 done, as duplicates will always all be routed, but duplicates are
1515 presumably rare, so I don't expect this is of any significance.
1517 For deliveries to pipes, files, and autoreplies, the duplicate checking
1518 still happens during the routing process, since they are not going to be
1521 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1522 It corrects a timeout issue with spamd. This is Ian's comment: "The
1523 background is that sometimes spamd either never reads data from a
1524 connection it has accepted, or it never writes response data. The exiscan
1525 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1526 blindly assumes that writes won't block so it may never time out."
1528 PH/39 Allow G after quota size as well as K and M.
1530 PH/40 The value set for $authenticated_id in an authenticator may not contain
1531 binary zeroes or newlines because the value is written to log lines and
1532 to spool files. There was no check on this. Now the value is run through
1533 the string_printing() function so that such characters are converted to
1534 printable escape sequences.
1536 PH/41 $message_linecount is a new variable that contains the total number of
1537 lines in the message. Compare $body_linecount, which is the count for the
1540 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1541 bounce and delay warning messages, except in certain special cases, which
1544 (a) An SMTP error message from a remote host;
1545 (b) A message specified in a :fail: redirection;
1546 (c) A message specified in a "fail" command in a system filter;
1547 (d) A message specified in a FAIL return from the queryprogram router;
1548 (e) A message specified by the cannot_route_message router option.
1550 In these cases only, Exim does include the error details in bounce and
1551 warning messages. There are also a few cases where bland messages such
1552 as "unrouteable address" or "local delivery error" are given.
1554 PH/43 $value is now also set for the "else" part of a ${run expansion.
1556 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1557 being worked on, but at least Exim now implements the latest version to
1560 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1561 process to complete was treated as a delivery failure, a timeout while
1562 writing the message to the pipe was logged, but erroneously treated as a
1563 successful delivery. Such timeouts include transport filter timeouts. For
1564 consistency with the overall process timeout, these timeouts are now
1565 treated as errors, giving rise to delivery failures by default. However,
1566 there is now a new Boolean option for the pipe transport called
1567 timeout_defer, which, if set TRUE, converts the failures into defers for
1568 both kinds of timeout. A transport filter timeout is now identified in
1571 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1572 systems where "make" and "gmake" are different, calling "gmake" at top
1573 level broke things. I've arranged for the value of $(MAKE) to be passed
1574 from the Makefile to this script so that it can call the same version of
1578 A note about Exim versions 4.44 and 4.50
1579 ----------------------------------------
1581 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1582 changes of various kinds. As a consequence, a big documentation update was
1583 needed. This delayed the release for rather longer than seemed good, especially
1584 in the light of a couple of (minor) security issues. Therefore, the changes
1585 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1586 release. So 4.44 and 4.50 are in effect two different branches that both start
1589 I have left the 4.50 change log unchanged; it contains all the changes since
1590 4.43. The change log for 4.44 is below; many of its items are identical to
1591 those for 4.50. This seems to be the most sensible way to preserve the
1592 historical information.
1598 1. Minor wording change to the doc/README.SIEVE file.
1600 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1601 computation of the current number of files was incorrect.
1603 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1604 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1607 4. Give more explanation in the error message when the command for a transport
1608 filter fails to execute.
1610 5. There are several places where Exim runs a non-Exim command in a
1611 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1612 was being done only for the command run by the queryprogram router. It is
1613 now done for all such subprocesses. The other cases are: ${run, transport
1614 filters, and the commands run by the lmtp and pipe transports.
1616 6. Added CONFIGURE_GROUP build-time option.
1618 7. Some older OS have a limit of 256 on the maximum number of file
1619 descriptors. Exim was using setrlimit() to set 1000 as a large value
1620 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1621 systems. I've change it so that if it can't get 1000, it tries for 256.
1623 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1624 was an oversight, and furthermore, ever since the addition of extra
1625 controls (e.g. 4.43/32), the checks on when to allow different forms of
1626 "control" were broken. There should now be diagnostics for all cases when a
1627 control that does not make sense is encountered.
1629 9. Added the /retain_sender option to "control=submission".
1631 10. $recipients is now available in the predata ACL (oversight).
1633 11. Tidy the search cache before the fork to do a delivery from a message
1634 received from the command line. Otherwise the child will trigger a lookup
1635 failure and thereby defer the delivery if it tries to use (for example) a
1636 cached ldap connection that the parent has called unbind on.
1638 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1639 of $address_data from the recipient verification was clobbered by the
1640 sender verification.
1642 13. The value of address_data from a sender verification is now available in
1643 $sender_address_data in subsequent conditions in the ACL statement.
1645 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1647 15. Added a new option "connect=<time>" to callout options, to set a different
1650 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1651 was its contents. (It was OK if the option was not defined at all.)
1653 17. A "Completed" log line is now written for messages that are removed from
1654 the spool by the -Mrm option.
1656 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1657 information about exactly what failed.
1659 19. Added -dd to debug only the daemon process.
1661 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1662 handles timeouts, both on the server side and network timeouts. Renamed the
1663 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1665 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1666 It is now set to "smtps".
1668 22. $host_address is now set to the target address during the checking of
1669 ignore_target_hosts.
1671 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1672 being passed; this would have caused $sender_host_name to have been used if
1673 matching the list had actually called for a host name (not very likely,
1674 since this list is usually IP addresses). A host name is now passed as
1677 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1678 code that specifies a non-restarting handler (typically sigaction() in
1679 modern systems) in an attempt to fix a rare and obscure crash bug.
1681 25. Narrowed the window for a race in the daemon that could cause it to ignore
1682 SIGCHLD signals. This is not a major problem, because they are used only to
1683 wake it up if nothing else does.
1685 26. A malformed maildirsize file could cause Exim to calculate negative values
1686 for the mailbox size or file count. Odd effects could occur as a result.
1687 The maildirsize information is now recalculated if the size or filecount
1690 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1691 support for a long time. Removed HAVE_SYS_VFS_H.
1693 28. Installed the latest version of exipick from John Jetmore.
1695 29. In an address list, if the pattern was not a regular expression, an empty
1696 subject address (from a bounce message) matched only if the pattern was an
1697 empty string. Non-empty patterns were not even tested. This was the wrong
1698 because it is perfectly reasonable to use an empty address as part of a
1699 database query. An empty address is now tested by patterns that are
1700 lookups. However, all the other forms of pattern expect the subject to
1701 contain a local part and a domain, and therefore, for them, an empty
1702 address still always fails if the pattern is not itself empty.
1704 30. Exim went into a mad DNS loop when attempting to do a callout where the
1705 host was specified on an smtp transport, and looking it up yielded more
1706 than one IP address.
1708 31. Re-factored the code for checking spool and log partition space into a
1709 function that finds that data and another that does the check. The former
1710 is then used to implement four new variables: $spool_space, $log_space,
1711 $spool_inodes, and $log_inodes.
1713 32. The RFC2047 encoding function was originally intended for short strings
1714 such as real names; it was not keeping to the 75-character limit for
1715 encoded words that the RFC imposes. It now respects the limit, and
1716 generates multiple encoded words if necessary. To be on the safe side, I
1717 have increased the buffer size for the ${rfc2047: expansion operator from
1720 33. It is now permitted to omit both strings after an "if" condition; if the
1721 condition is true, the result is "true". As before, when the second string
1722 is omitted, a false condition yields an empty string. This makes it less
1723 cumbersome to write custom ACL and router conditions.
1725 34. Failure to deliver a bounce message always caused it to be frozen, even if
1726 there was an errors_to setting on the router. The errors_to setting is now
1729 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1730 canonical form (fully expanded) before being placed in
1731 $sender_host_address.
1733 36. The table in the code that translates DNS record types into text (T_A to
1734 "A" for instance) was missing entries for NS and CNAME. It is just possible
1735 that this could have caused confusion if both these types were looked up
1736 for the same domain, because the text type is used as part of Exim's
1737 per-process caching. But the chance of anyone hitting this buglet seems
1740 37. The dnsdb lookup has been extended in a number of ways.
1742 (1) There is a new type, "zns", which walks up the domain tree until it
1743 finds some nameserver records. It should be used with care.
1745 (2) There is a new type, "mxh", which is like "mx" except that it returns
1746 just the host names, not the priorities.
1748 (3) It is now possible to give a list of domains (or IP addresses) to be
1749 looked up. The behaviour when one of the lookups defers can be
1750 controlled by a keyword.
1752 (4) It is now possible to specify the separator character for use when
1753 multiple records are returned.
1755 38. The dnslists ACL condition has been extended: it is now possible to supply
1756 a list of IP addresses and/or domains to be looked up in a particular DNS
1759 39. Added log_selector=+queue_time_overall.
1761 40. When running the queue in the test harness, wait just a tad after forking a
1762 delivery process, to get repeatability of debugging output.
1764 41. Include certificate and key file names in error message when GnuTLS fails
1765 to set them up, because the GnuTLS error message doesn't include the name
1766 of the failing file when there is a problem reading it.
1768 42. Allow both -bf and -bF in the same test run.
1770 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1772 44. The "Exiscan patch" is now merged into the mainline Exim source.
1774 45. Sometimes the final signoff response after QUIT could fail to get
1775 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1776 < 0 before doing a fflush(). This bug looks as though it goes back to the
1777 introduction of TLS in release 3.20, but "sometimes" must have been rare
1778 because the tests only now provoked it.
1780 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1781 (this can affect the format of dates).
1783 47. exim_tidydb, when checking for the continued existence of a message for
1784 which it has found a message-specific retry record, was not finding
1785 messages that were in split spool directories. Consequently, it was
1786 deleting retry records that should have stayed in existence.
1788 48. Steve fixed some bugs in eximstats.
1790 49. The SPA authentication driver was not abandoning authentication and moving
1791 on to the next authenticator when an expansion was forced to fail,
1792 contradicting the general specification for all authenticators. Instead it
1793 was generating a temporary error. It now behaves as specified.
1795 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1796 (the order specifies the preference for clients). The order is now AES256,
1797 AES128, 3DES, ARCFOUR128.
1799 51. Small patch to Sieve code - explicitly set From: when generating an
1802 52. Exim crashed if a remote delivery caused a very long error message to be
1803 recorded - for instance if somebody sent an entire SpamAssassin report back
1804 as a large number of 550 error lines. This bug was coincidentally fixed by
1805 increasing the size of one of Exim's internal buffers (big_buffer) that
1806 happened as part of the Exiscan merge. However, to be on the safe side, I
1807 have made the code more robust (and fixed the comments that describe what
1810 53. Now that there can be additional text after "Completed" in log lines (if
1811 the queue_time_overall log selector is set), a one-byte patch to exigrep
1812 was needed to allow it to recognize "Completed" as not the last thing in
1815 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1816 patch that reportedly fixes this has been added. I am not expert enough to
1817 create a test for it. This is what the patch creator wrote:
1819 "I found a little strange behaviour of ldap code when working with
1820 Windows 2003 AD Domain, where users was placed in more than one
1821 Organization Units. When I tried to give exim partial DN, the exit code
1822 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1823 But simultaneously result of request was absolutely normal ldap result,
1824 so I produce this patch..."
1826 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1827 REFERENCE, so I have modified the code to exclude the patch when that macro
1830 55. Some experimental protocols are using DNS PTR records for new purposes. The
1831 keys for these records are domain names, not reversed IP addresses. The
1832 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1833 leaves it alone. Component reversal etc. now happens only for IP addresses.
1836 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1838 57. Double the size of the debug message buffer (to 2048) so that more of very
1839 long debug lines gets shown.
1841 58. The exicyclog utility now does better if the number of log files to keep
1842 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1844 59. Two changes related to the smtp_active_hostname option:
1846 (1) $smtp_active_hostname is now available as a variable.
1847 (2) The default for smtp_banner uses $smtp_active_hostname instead
1848 of $primary_hostname.
1850 60. The host_aton() function is supposed to be passed a string that is known
1851 to be a valid IP address. However, in the case of IPv6 addresses, it was
1852 not checking this. This is a hostage to fortune. Exim now panics and dies
1853 if the condition is not met. A case was found where this could be provoked
1854 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1855 components; fortuitously, this particular loophole had already been fixed
1856 by change 4.50/55 above.
1858 If there are any other similar loopholes, the new check in host_aton()
1859 itself should stop them being exploited. The report I received stated that
1860 data on the command line could provoke the exploit when Exim was running as
1861 exim, but did not say which command line option was involved. All I could
1862 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1863 running as the user.
1866 61. There was a buffer overflow vulnerability in the SPA authentication code
1867 (which came originally from the Samba project). I have added a test to the
1868 spa_base64_to_bits() function which I hope fixes it.
1871 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1872 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1874 63. The daemon start-up calls getloadavg() while still root for those OS that
1875 need the first call to be done as root, but it missed one case: when
1876 deliver_queue_load_max is set with deliver_drop_privilege. This is
1877 necessary for the benefit of the queue runner, because there is no re-exec
1878 when deliver_drop_privilege is set.
1880 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1881 This has been fixed.
1883 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1884 was in use, was not putting the data itself into the right store pool;
1885 consequently, it could be overwritten for a subsequent message in the same
1886 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1889 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1891 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1892 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1893 the function were treating the return as a boolean value, which happened to
1894 work because 0=false and not-0=true, but is not correct code.
1896 68. The host_aton() function was not handling scoped IPv6 addresses (those
1897 with, for example, "%eth0" on the end) correctly.
1899 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1900 negated items (that is, ~something) in unsigned ints. Some compilers
1901 apparently mutter when there is no cast.
1903 70. If an address verification called from an ACL failed, and did not produce a
1904 user-specific message (i.e. there was only a "system" message), nothing was
1905 put in $acl_verify_message. In this situation, it now puts the system
1908 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1909 session; change 4.31/43 added the unwanted input to the log line - except
1910 that it did not do this in the start of session case. It now does.
1912 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1913 This is wrong and can cause the other end to generate a synchronization
1914 error if it is another Exim or anything else that does the synchronization
1915 check. A QUIT command is no longer sent after a timeout.
1917 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1918 during host lookups.
1920 74. The defer_ok option of callout verification was not working if it was used
1921 when verifying addresses in header lines, that is, for this case:
1923 verify = header_sender/callout=defer_ok
1925 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1926 those file descriptors could be used for SMTP connections. If anything
1927 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1928 could be sent to the SMTP client, causing chaos. The daemon now opens
1929 stdin, stdout, and stderr to /dev/null when it puts itself into the
1932 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1933 log by default. The user can override this with suitable Perl magic.
1935 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1936 the log message when discard triggers, was not working for the DATA ACL or
1937 for the non-SMTP ACL.
1939 78. Error message wording change in sieve.c.
1941 79. If smtp_accept_max_per_host was set, the number of connections could be
1942 restricted to fewer than expected, because the daemon was trying to set up
1943 a new connection before checking whether the processes handling previous
1944 connections had finished. The check for completed processes is now done
1945 earlier. On busy systems, this bug wouldn't be noticed because something
1946 else would have woken the daemon, and it would have reaped the completed
1949 80. If a message was submitted locally by a user whose login name contained one
1950 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1951 It caused a spool format error. I have fixed the spool reading code. A
1952 related problem was that the "from" clause in the Received: line became
1953 illegal because of the space(s). It is now covered by ${quote_local_part.
1955 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1958 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1960 83. Patch from Sieve maintainer to fix unterminated string problem in
1961 "vacation" handling.
1963 84. Some minor changes to the Linux configuration files to help with other
1964 OS variants using glibc.
1966 85. One more patch for Sieve to update vacation handling to latest spec.
1969 ----------------------------------------------------
1970 See the note above about the 4.44 and 4.50 releases.
1971 ----------------------------------------------------
1977 1. Change 4.43/35 introduced a bug that caused file counts to be
1978 incorrectly computed when quota_filecount was set in an appendfile
1981 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1982 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1985 3. Give more explanation in the error message when the command for a transport
1986 filter fails to execute.
1988 4. There are several places where Exim runs a non-Exim command in a
1989 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1990 was being done only for the command run by the queryprogram router. It is
1991 now done for all such subprocesses. The other cases are: ${run, transport
1992 filters, and the commands run by the lmtp and pipe transports.
1994 5. Some older OS have a limit of 256 on the maximum number of file
1995 descriptors. Exim was using setrlimit() to set 1000 as a large value
1996 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1997 systems. I've change it so that if it can't get 1000, it tries for 256.
1999 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2000 was an oversight, and furthermore, ever since the addition of extra
2001 controls (e.g. 4.43/32), the checks on when to allow different forms of
2002 "control" were broken. There should now be diagnostics for all cases when a
2003 control that does not make sense is encountered.
2005 7. $recipients is now available in the predata ACL (oversight).
2007 8. Tidy the search cache before the fork to do a delivery from a message
2008 received from the command line. Otherwise the child will trigger a lookup
2009 failure and thereby defer the delivery if it tries to use (for example) a
2010 cached ldap connection that the parent has called unbind on.
2012 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2013 of $address_data from the recipient verification was clobbered by the
2014 sender verification.
2016 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2017 was its contents. (It was OK if the option was not defined at all.)
2019 11. A "Completed" log line is now written for messages that are removed from
2020 the spool by the -Mrm option.
2022 12. $host_address is now set to the target address during the checking of
2023 ignore_target_hosts.
2025 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2026 being passed; this would have caused $sender_host_name to have been used if
2027 matching the list had actually called for a host name (not very likely,
2028 since this list is usually IP addresses). A host name is now passed as
2031 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2032 code that specifies a non-restarting handler (typically sigaction() in
2033 modern systems) in an attempt to fix a rare and obscure crash bug.
2035 15. Narrowed the window for a race in the daemon that could cause it to ignore
2036 SIGCHLD signals. This is not a major problem, because they are used only to
2037 wake it up if nothing else does.
2039 16. A malformed maildirsize file could cause Exim to calculate negative values
2040 for the mailbox size or file count. Odd effects could occur as a result.
2041 The maildirsize information is now recalculated if the size or filecount
2044 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2045 support for a long time. Removed HAVE_SYS_VFS_H.
2047 18. Updated exipick to current release from John Jetmore.
2049 19. Allow an empty sender to be matched against a lookup in an address list.
2050 Previously the only cases considered were a regular expression, or an
2053 20. Exim went into a mad DNS lookup loop when doing a callout where the
2054 host was specified on the transport, if the DNS lookup yielded more than
2057 21. The RFC2047 encoding function was originally intended for short strings
2058 such as real names; it was not keeping to the 75-character limit for
2059 encoded words that the RFC imposes. It now respects the limit, and
2060 generates multiple encoded words if necessary. To be on the safe side, I
2061 have increased the buffer size for the ${rfc2047: expansion operator from
2064 22. Failure to deliver a bounce message always caused it to be frozen, even if
2065 there was an errors_to setting on the router. The errors_to setting is now
2068 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2069 canonical form (fully expanded) before being placed in
2070 $sender_host_address.
2072 24. Updated eximstats to version 1.33
2074 25. Include certificate and key file names in error message when GnuTLS fails
2075 to set them up, because the GnuTLS error message doesn't include the name
2076 of the failing file when there is a problem reading it.
2078 26. Expand error message when OpenSSL has problems setting up cert/key files.
2081 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2082 (this can affect the format of dates).
2084 28. exim_tidydb, when checking for the continued existence of a message for
2085 which it has found a message-specific retry record, was not finding
2086 messages that were in split spool directories. Consequently, it was
2087 deleting retry records that should have stayed in existence.
2089 29. eximstats updated to version 1.35
2090 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2091 1.35 - bugfix such that pie charts by volume are generated correctly
2093 30. The SPA authentication driver was not abandoning authentication and moving
2094 on to the next authenticator when an expansion was forced to fail,
2095 contradicting the general specification for all authenticators. Instead it
2096 was generating a temporary error. It now behaves as specified.
2098 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2099 (the order specifies the preference for clients). The order is now AES256,
2100 AES128, 3DES, ARCFOUR128.
2102 31. Small patch to Sieve code - explicitly set From: when generating an
2105 32. Exim crashed if a remote delivery caused a very long error message to be
2106 recorded - for instance if somebody sent an entire SpamAssassin report back
2107 as a large number of 550 error lines. This bug was coincidentally fixed by
2108 increasing the size of one of Exim's internal buffers (big_buffer) that
2109 happened as part of the Exiscan merge. However, to be on the safe side, I
2110 have made the code more robust (and fixed the comments that describe what
2113 33. Some experimental protocols are using DNS PTR records for new purposes. The
2114 keys for these records are domain names, not reversed IP addresses. The
2115 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2116 leaves it alone. Component reversal etc. now happens only for IP addresses.
2119 34. The host_aton() function is supposed to be passed a string that is known
2120 to be a valid IP address. However, in the case of IPv6 addresses, it was
2121 not checking this. This is a hostage to fortune. Exim now panics and dies
2122 if the condition is not met. A case was found where this could be provoked
2123 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2124 components; fortuitously, this particular loophole had already been fixed
2125 by change 4.50/55 or 4.44/33 above.
2127 If there are any other similar loopholes, the new check in host_aton()
2128 itself should stop them being exploited. The report I received stated that
2129 data on the command line could provoke the exploit when Exim was running as
2130 exim, but did not say which command line option was involved. All I could
2131 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2132 running as the user.
2135 35. There was a buffer overflow vulnerability in the SPA authentication code
2136 (which came originally from the Samba project). I have added a test to the
2137 spa_base64_to_bits() function which I hope fixes it.
2140 36. The daemon start-up calls getloadavg() while still root for those OS that
2141 need the first call to be done as root, but it missed one case: when
2142 deliver_queue_load_max is set with deliver_drop_privilege. This is
2143 necessary for the benefit of the queue runner, because there is no re-exec
2144 when deliver_drop_privilege is set.
2146 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2147 was in use, was not putting the data itself into the right store pool;
2148 consequently, it could be overwritten for a subsequent message in the same
2149 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2152 38. Sometimes the final signoff response after QUIT could fail to get
2153 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2154 < 0 before doing a fflush(). This bug looks as though it goes back to the
2155 introduction of TLS in release 3.20, but "sometimes" must have been rare
2156 because the tests only now provoked it.
2162 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2163 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2164 Now it returns an integer. A number of calls were still expecting a T/F
2165 return. Fortuitously, in all cases, the tests worked in OK situations,
2166 which is the norm. However, things would have gone wrong on any write
2167 failures on the smtp file descriptor. This function is used when sending
2168 messages over SMTP and also when doing verify callouts.
2170 2. When Exim is called to do synchronous delivery of a locally submitted
2171 message (the -odf or -odi options), it no longer closes stderr before doing
2174 3. Implemented the mua_wrapper option.
2176 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2178 5. Implemented the functions header_remove(), header_testname(),
2179 header_add_at_position(), and receive_remove_recipient(), and exported them
2182 6. If an ACL "warn" statement specified the addition of headers, Exim already
2183 inserted X-ACL-Warn: at the start if there was no header name. However, it
2184 was not making this test for the second and subsequent header lines if
2185 there were newlines in the string. This meant that an invalid header could
2186 be inserted if Exim was badly configured.
2188 7. Allow an ACL "warn" statement to add header lines at the start or after all
2189 the Received: headers, as well as at the end.
2191 8. Added the rcpt_4xx retry error code.
2193 9. Added postmaster_mailfrom=xxx to callout verification option.
2195 10. Added mailfrom=xxxx to the callout verification option, for verify=
2198 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2199 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2201 12. Inserted some casts to stop certain compilers warning when using pointer
2202 differences as field lengths or precisions in printf-type calls (mostly
2203 affecting debugging statements).
2205 13. Added optional readline() support for -be (dynamically loaded).
2207 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2208 same clock tick as a message's arrival, so that its received time was the
2209 same as the "first fail" time on the retry record, and that message
2210 remained on the queue past the ultimate address timeout, every queue runner
2211 would try a delivery (because it was past the ultimate address timeout) but
2212 after another failure, the ultimate address timeout, which should have then
2213 bounced the address, did not kick in. This was a "< instead of <=" error;
2214 in most cases the first failure would have been in the next clock tick
2215 after the received time, and all would be well.
2217 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2218 being recognized when the domain list was tested by the match_domain
2219 condition in an expansion string.
2221 16. Added the ${str2b64: operator.
2223 17. Exim was always calling setrlimit() to set a large limit for the number of
2224 processes, without checking whether the existing limit was already
2225 adequate. (It did check for the limit on file descriptors.) Furthermore,
2226 errors from getrlimit() and setrlimit() were being ignored. Now they are
2227 logged to the main and panic logs, but Exim does carry on, to try to do its
2228 job under whatever limits there are.
2230 18. Imported PCRE 5.0.
2232 19. Trivial typo in log message " temporarily refused connection" (the leading
2235 20. If the log selector return_path_on_delivery was set and an address was
2236 redirected to /dev/null, the delivery process crashed because it assumed
2237 that a return path would always be set for a "successful" delivery. In this
2238 case, the whole delivery is bypassed as an optimization, and therefore no
2241 21. Internal re-arrangement: the function for sending a challenge and reading
2242 a response while authentication was assuming a zero-terminated challenge
2243 string. It's now changed to take a pointer and a length, to allow for
2244 binary data in such strings.
2246 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2248 23. Exim was not respecting finduser_retries when seeking the login of the
2249 uid under which it was called; it was always trying 10 times. (The default
2250 setting of finduser_retries is zero.) Also, it was sleeping after the final
2251 failure, which is pointless.
2253 24. Implemented tls_on_connect_ports.
2255 25. Implemented acl_smtp_predata.
2257 26. If the domain in control=submission is set empty, Exim assumes that the
2258 authenticated id is a complete email address when it generates From: or
2259 Sender: header lines.
2261 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2262 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2263 chown and chgrp in /bin and hostname in /usr/bin.
2265 28. Exim was keeping the "process log" file open after each use, just as it
2266 does for the main log. This opens the possibility of it remaining open for
2267 long periods when the USR1 signal hits a daemon. Occasional processlog
2268 errors were reported, that could have been caused by this. Anyway, it seems
2269 much more sensible not to leave this file open at all, so that is what now
2272 29. The long-running daemon process does not normally write to the log once it
2273 has entered its main loop, and it closes the log before doing so. This is
2274 so that log files can straightforwardly be renamed and moved. However,
2275 there are a couple of unusual error situations where the daemon does write
2276 log entries, and I had neglected to close the log afterwards.
2278 30. The text of an SMTP error response that was received during a remote
2279 delivery was being truncated at 512 bytes. This is too short for some of
2280 the long messages that one sometimes sees. I've increased the limit to
2283 31. It is now possible to make retry rules that apply only when a message has a
2284 specific sender, in particular, an empty sender.
2286 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2287 it possible to be selective about when SMTP synchronization is enforced.
2289 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2291 32. Implemented hosts_connection_nolog.
2293 33. Added an ACL for QUIT.
2295 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2298 35. Added mailbox_size and mailbox_filecount to appendfile.
2300 36. Added control = no_multiline_responses to ACLs.
2302 37. There was a bug in the logic of the code that waits for the clock to tick
2303 in the case where the clock went backwards by a substantial amount such
2304 that the microsecond fraction of "now" was more than the microsecond
2305 fraction of "then" (but the whole seconds number was less).
2307 38. Added support for the libradius Radius client library this is found on
2308 FreeBSD (previously only the radiusclient library was supported).
2314 1. When certain lookups returned multiple values in the form name=value, the
2315 quoting of the values was not always being done properly. Specifically:
2316 (a) If the value started with a double quote, but contained no whitespace,
2318 (b) If the value contained whitespace other than a space character (i.e.
2319 tabs or newlines or carriage returns) it was not quoted.
2320 This fix has been applied to the mysql and pgsql lookups by writing a
2321 separate quoting function and calling it from the lookup code. The fix
2322 should probably also be applied to nisplus, ibase and oracle lookups, but
2323 since I cannot test any of those, I have not disturbed their existing code.
2325 2. A hit in the callout cache for a specific address caused a log line with no
2326 reason for rejecting RCPT. Now it says "Previous (cached) callout
2327 verification failure".
2329 3. There was an off-by-one bug in the queryprogram router. An over-long
2330 return line was truncated at 256 instead of 255 characters, thereby
2331 overflowing its buffer with the terminating zero. As well as fixing this, I
2332 have increased the buffer size to 1024 (and made a note to document this).
2334 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2335 when Exim is waiting for an SMTP response from a remote server, Exim
2336 restarts its select() call on the socket, thereby resetting its timeout.
2337 This is not a problem when such interrupts are rare. Somebody set up a cron
2338 job to run exiwhat every 2 minutes, which is less than the normal select()
2339 timeout (5 or 10 minutes). This meant that the select() timeout never
2340 kicked in because it was always reset. I have fixed this by comparing the
2341 time when an interrupt arrives with the time at the start of the first call
2342 to select(). If more time than the timeout has elapsed, the interrupt is
2343 treated as a timeout.
2345 5. Some internal re-factoring in preparation for the addition of Sieve
2346 extensions (by MH). In particular, the "personal" test is moved to a
2347 separate function, and given an option for scanning Cc: and Bcc: (which is
2348 not set for Exim filters).
2350 6. When Exim created an email address using the login of the caller as the
2351 local part (e.g. when creating a From: or Sender: header line), it was not
2352 quoting the local part when it contained special characters such as @.
2354 7. Installed new OpenBSD configuration files.
2356 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2357 try to make them clearer.
2359 9. Callout options, other than the timeout value, were being ignored when
2360 verifying sender addresses in header lines. For example, when using
2362 verify = header_sender/callout=no_cache
2364 the cache was (incorrectly) being used.
2366 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2367 only the Cygwin environment.
2369 11. When return_path_on_delivery was set as a log selector, if different remote
2370 addresses in the same message used different return paths and parallel
2371 remote delivery occurred, the wrong values would sometimes be logged.
2372 (Whenever a remote delivery process finished, the return path value from
2373 the most recently started remote delivery process was logged.)
2375 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2376 lines when AUTH and/or TLS are in use. This is the "received protocol"
2377 field. Exim used to use "asmtp" for authenticated SMTP, without any
2378 indication (in the protocol name) for TLS use. Now it follows the RFC and
2379 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2380 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2381 names appear in log lines as well as in Received: header lines.
2383 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2384 extensions, and comparison tests, and to fix some bugs.
2386 14. Changes to the "personal" filter test:
2388 (1) The test was buggy in that it was just doing the equivalent of
2389 "contains" tests on header lines. For example, if a user's address was
2390 anne@some.where, the "personal" test would incorrectly be true for
2392 To: susanne@some.where
2394 This test is now done by extracting each address from the header in turn,
2395 and checking the entire address. Other tests that are part of "personal"
2396 are now done using regular expressions (for example, to check local parts
2397 of addresses in From: header lines).
2399 (2) The list of non-personal local parts in From: addresses has been
2400 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2401 taken from the Sieve specification recommendations.
2403 (3) If the message contains any header line starting with "List-" it is
2404 treated as non-personal.
2406 (4) The test for "circular" in the Subject: header line has been removed
2407 because it now seems ill-conceived.
2409 15. Minor typos in src/EDITME comments corrected.
2411 16. Installed latest exipick from John Jetmore.
2413 17. If headers_add on a router specified a text string that was too long for
2414 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2415 of string_sprintf() is now avoided.
2417 18. $message_body_size was not set (it was always zero) when running the DATA
2418 ACL and the local_scan() function.
2420 19. For the "mail" command in an Exim filter, no default was being set for
2421 the once_repeat time, causing a random time value to be used if "once" was
2422 specified. (If the value happened to be <= 0, no repeat happened.) The
2423 default is now 0s, meaning "never repeat". The "vacation" command was OK
2424 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2425 (I found it when inspecting the code).
2427 20. There is now an overall timeout for performing a callout verification. It
2428 defaults to 4 times the callout timeout, which applies to individual SMTP
2429 commands during the callout. The overall timeout applies when there is more
2430 than one host that can be tried. The timeout is checked before trying the
2431 next host. This prevents very long delays if there are a large number of
2432 hosts and all are timing out (e.g. when the network connections are timing
2433 out). The value of the overall timeout can be changed by specifying an
2434 additional sub-option for "callout", called "maxwait". For example:
2436 verify = sender/callout=5s,maxwait=20s
2438 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2439 to the end before writing, but this should make it even safer).
2441 22. Exim was forgetting that it had advertised PIPELINING for the second and
2442 subsequent messages on an SMTP connection. It was also not resetting its
2443 memory on STARTTLS and an internal HELO.
2445 23. When Exim logs an SMTP synchronization error within a session, it now
2446 records whether PIPELINING has been advertised or not.
2448 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2449 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2450 rather than long int.
2452 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2454 26. Added the never_mail option to autoreply.
2460 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2461 crash if the getsockname() call failed; this can happen if a connection is
2462 closed very soon after it is established. The problem was simply in the
2463 order in which certain operations were done, causing Exim to try to write
2464 to the SMTP stream before it had set up the file descriptor. The bug has
2465 been fixed by making things happen in the correct order.
2471 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2472 before the connection was closed, thus losing the rejection response.
2474 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2475 some early Solaris releases, but causes trouble in current releases where
2476 socklen_t is defined.
2478 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2481 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2484 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2485 that releasing the top of it at the end releases what was used for sub-
2486 expansions (unless the block got too big). However, discard this block if
2487 the first thing is a variable or header, so that we can use its block when
2488 it is dynamic (useful for very large $message_headers, for example).
2490 6. Lookups now cache *every* query, not just the most recent. A new, separate
2491 store pool is used for this. It can be recovered when all lookup caches are
2492 flushed. Lookups now release memory at the end of their result strings.
2493 This has involved some general refactoring of the lookup sources.
2495 7. Some code has been added to the store_xxx() functions to reduce the amount
2496 of flapping under certain conditions.
2498 8. log_incoming_interface used to affect only the <= reception log lines. Now
2499 it causes the local interface and port to be added to several more SMTP log
2500 lines, for example "SMTP connection from", and rejection lines.
2502 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2504 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2506 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2507 could be overwritten at the end of the current message (or the start of a
2508 new message if it was set in a HELO ACL). The value is now preserved for
2509 the duration of the SMTP connection.
2511 12. If a transport had a headers_rewrite setting, and a matching header line
2512 contained an unqualified address, that address was qualified, even if it
2513 did not match any rewriting rules. The underlying bug was that the values
2514 of the flags that permit the existence of unqualified sender and recipient
2515 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2516 non-local messages, and by -bnq for local messages) were not being
2517 preserved with the message after it was received.
2519 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2520 "next input=" as part of the text comprising the host identity instead of
2521 the correct text. The code was using the same buffer for two different
2522 strings. However, depending on which order the printing function evaluated
2523 its arguments, the bug did not always show up. Under Linux, for example, my
2524 test suite worked just fine.
2526 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2527 doesn't work with some older versions of Perl. It has been changed to "my",
2528 which in any case is probably the better facility to use.
2530 15. A really picky compiler found some instances of statements for creating
2531 error messages that either had too many or two few arguments for the format
2534 16. The size of the buffer for calls to the DNS resolver has been increased
2535 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2536 for addresses that have a lot of PTR records. This alleviates a problem; it
2537 does not fully solve it.
2539 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2540 buffer now truncates the list and logs the incident, which is the same
2541 action as happens when Exim is looking up a host name and its aliases.
2542 Previously in this situation something unpredictable would happen;
2543 sometimes it was "internal error: store_reset failed".
2545 18. If a server dropped the connection unexpectedly when an Exim client was
2546 using GnuTLS and trying to read a response, the client delivery process
2547 crashed while trying to generate an error log message.
2549 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2550 string, for example:
2552 warn message = H1: something\nH2: something
2554 the text was added as a single header line from Exim's point of view
2555 though it ended up OK in the delivered message. However, searching for the
2556 second and subsequent header lines using $h_h2: did not work. This has been
2557 fixed. Similarly, if a system filter added multiple headers in this way,
2558 the routers could not see them.
2560 20. Expanded the error message when iplsearch is called with an invalid key to
2561 suggest using net-iplsearch in a host list.
2563 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2564 ACLs are no longer actually imposed (and a message to that effect is
2567 22. If a "gecos" field in a passwd entry contained escaped characters, in
2568 particular, if it contained a \" sequence, Exim got it wrong when building
2569 a From: or a Sender: header from that name. A second bug also caused
2570 incorrect handling when an unquoted " was present following a character
2571 that needed quoting.
2573 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2574 was not being matched caselessly.
2576 24. Arranged for all hyphens in the exim.8 source to be escaped with
2579 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2580 properly. Recipient callout cache records were still being keyed to include
2581 the sender, even when use_sender was set false. This led to far more
2582 callouts that were necessary. The sender is no longer included in the key
2583 when use_sender is false.
2585 26. Added "control = submission" modifier to ACLs.
2587 27. Added the ${base62d: operator to decode base 62 numbers.
2589 28. dnsdb lookups can now access SRV records.
2591 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2592 the configuration file.
2594 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2595 (-v) mode. This makes the output for a verbose queue run more intelligible.
2597 31. Added a use_postmaster feature to recipient callouts.
2599 32. Added the $body_zerocount variable, containing the number of binary zero
2600 bytes in the message body.
2602 33. The time of last modification of the "new" subdirectory is now used as the
2603 "mailbox time last read" when there is a quota error for a maildir
2606 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2608 35. Added +ignore_unknown as a special item in host lists.
2610 36. Code for decoding IPv6 addresses in host lists is now included, even if
2611 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2612 address was recognized as an IP address, but was then not correctly decoded
2613 into binary, causing unexpected and incorrect effects when compared with
2620 1. Very minor rewording of debugging text in manualroute to say "list of
2621 hosts" instead of "hostlist".
2623 2. If verify=header_syntax was set, and a header line with an unqualified
2624 address (no domain) and a large number of spaces between the end of the
2625 name and the colon was received, the reception process suffered a buffer
2626 overflow, and (when I tested it) crashed. This was caused by some obsolete
2627 code that should have been removed. The fix is to remove it!
2629 3. When running in the test harness, delay a bit after writing a bounce
2630 message to get a bit more predictability in the log output.
2632 4. Added a call to search_tidyup() just before forking a reception process. In
2633 theory, someone could use a lookup in the expansion of smtp_accept_max_
2634 per_host which, without the tidyup, could leave open a database connection.
2636 5. Added the variables $recipient_data and $sender_data which get set from a
2637 lookup success in an ACL "recipients" or "senders" condition, or a router
2638 "senders" option, similar to $domain_data and $local_part_data.
2640 6. Moved the writing of debug_print from before to after the "senders" test
2643 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2644 problems for message scanning, either using a data ACL, or using
2645 local_scan() because the Received: header was not generated till after they
2646 were called (in order to set the time as the time of reception completion).
2647 I have revised the way this works. The header is now generated after the
2648 body is received, but before the ACL or local_scan() are called. After they
2649 are run, the timestamp in the header is updated.
2655 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2656 before starting a queue runner without re-exec. This happened only when
2657 deliver_drop_privilege was set or when the Exim user was set to root. The
2658 effect of the bug was that timeouts during subsequent deliveries caused
2659 crashes instead of being properly handled. The handler is now left at its
2660 default (and expected) setting.
2662 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2663 message, again when deliver_drop_privilege is set or Exim is run as root.
2664 The bug described in (1) was not present in this case, but the tidying up
2665 of the other signals was missing. I have made the two cases consistent.
2667 3. The ignore_target_hosts setting on a manualroute router was being ignored
2668 for hosts that were looked up using the /MX notation.
2670 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2673 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2674 operated on the sender address. After changing the $sender_address to <>
2675 for the sender address verify, Exim was re-instated it as the original
2676 (before rewriting) address, but remembering that it had rewritten it, so it
2677 wasn't rewriting it again. This bug also had the effect of breaking the
2678 sender address verification caching when the sender address was rewritten.
2680 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2681 This has been changed so that if the ip literal address matches
2682 ignore_target_hosts, the router declines.
2684 7. Added expansion conditions match_domain, match_address, and match_local_
2685 part (NOT match_host).
2687 8. The placeholder for the Received: header didn't have a length field set.
2689 9. Added code to Exim itself and to exim_lock to test for a specific race
2690 condition that could lead to file corruption when using MBX delivery. The
2691 issue is with the lockfile that is created in /tmp. If this file is removed
2692 after a process has opened it but before that process has acquired a lock,
2693 there is the potential for a second process to recreate the file and also
2694 acquire a lock. This could lead to two Exim processes writing to the file
2695 at the same time. The added code performs the same test as UW imapd; it
2696 checks after acquiring the lock that its file descriptor still refers to
2697 the same named file.
2699 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2700 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2703 11. Added the smtp_active_hostname option. If used, this will typically be made
2704 to depend on the incoming interface address. Because $interface_address is
2705 not set up until the daemon has forked a reception process, error responses
2706 that can happen earlier (such as "too many connections") no longer contain
2709 12. If an expansion in a condition on a "warn" statement fails because a lookup
2710 defers, the "warn" statement is abandoned, and the next ACL statement is
2711 processed. Previously this caused the whole ACL to be aborted.
2713 13. Added the iplsearch lookup type.
2715 14. Added ident_timeout as a log selector.
2717 15. Added tls_certificate_verified as a log selector.
2719 16. Added a global option tls_require_ciphers (compare the smtp transport
2720 option of the same name). This controls incoming TLS connections.
2722 17. I finally figured out how to make tls_require_ciphers do a similar thing
2723 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2724 before starting the TLS session.
2726 18. Tabs are now shown as \t in -bP output.
2728 19. If the log selector return_path_on_delivery was set, Exim crashed when
2729 bouncing a message because it had too many Received: header lines.
2731 20. If two routers both had headers_remove settings, and the first one included
2732 a superfluous trailing colon, the final name in the first list and the
2733 first name in the second list were incorrectly joined into one item (with a
2734 colon in the middle).
2740 1. Added -C and -D options to the exinext utility, mainly to make it easier
2741 to include in the automated testing, but these could be helpful when
2742 multiple configurations are in use.
2744 2. The exinext utility was not formatting the output nicely when there was
2745 an alternate port involved in the retry record key, nor when there was a
2746 message id as well (for retries that were specific to a specific message
2747 and a specific host). It was also confused by IPv6 addresses, because of
2748 the additional colons they contain. I have fixed the IPv4 problem, and
2749 patched it up to do a reasonable job for IPv6.
2751 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2752 delivery, the log line now contains "pipelined" if PIPELINING was used.
2754 4. An SMTP transport process used to panic and die if the bind() call to set
2755 an explicit outgoing interface failed. This has been changed; it is now
2756 treated in the same way as a connect() failure.
2758 5. A reference to $sender_host_name in the part of a conditional expansion
2759 that was being skipped was still causing a DNS lookup. This no longer
2762 6. The def: expansion condition was not recognizing references to header lines
2763 that used bh_ and bheader_.
2765 7. Added the _cache feature to named lists.
2767 8. The code for checking quota_filecount in the appendfile transport was
2768 allowing one more file than it should have been.
2770 9. For compatibility with Sendmail, the command line option
2778 and sets the incoming protocol and host name (for trusted callers). The
2779 host name and its colon can be omitted when only the protocol is to be set.
2780 Note the Exim already has two private options, -pd and -ps, that refer to
2781 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2782 "s", but I don't think that's a major issue.
2784 10. A number of refactoring changes to the code, none of which should affect
2787 (a) The number of logging options was getting close to filling up the
2788 32-bit word that was used as a bit map. I have split them into two classes:
2789 those that are passed in the argument to log_write(), and those that are
2790 only ever tested independently outside of that function. These are now in
2791 separate 32-bit words, so there is plenty of room for expansion again.
2792 There is no change in the user interface or the logging behaviour.
2794 (b) When building, for example, log lines, the code previously used a
2795 macro that called string_cat() twice, in order to add two strings. This is
2796 not really sufficiently general. Furthermore, there was one instance where
2797 it was actually wrong because one of the argument was used twice, and in
2798 one call a function was used. (As it happened, calling the function twice
2799 did not affect the overall behaviour.) The macro has been replaced by a
2800 function that can join an arbitrary number of extra strings onto a growing
2803 (c) The code for expansion conditions now uses a table and a binary chop
2804 instead of a serial search (which was left over from when there were very
2805 few conditions). Also, it now recognizes conditions like "pam" even when
2806 the relevant support is not compiled in: a suitably worded error message is
2807 given if an attempt is made to use such a condition.
2809 11. Added ${time_interval:xxxxx}.
2811 12. A bug was causing one of the ddress fields not to be passed back correctly
2812 from remote delivery subprocesses. The field in question was not being
2813 subsequently used, so this caused to problems in practice.
2815 13. Added new log selectors queue_time and deliver_time.
2817 14. Might have fixed a bug in maildirsizefile handling that threw up
2818 "unexpected character" debug warnings, and recalculated the data
2819 unnecessarily. In any case, I expanded the warning message to give more
2822 15. Added the message "Restricted characters in address" to the statements in
2823 the default ACL that block characters like @ and % in local parts.
2825 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2826 Three changes have been made:
2828 (a) There was a serious bug; a negative response to MAIL caused the whole
2829 recipient domain to be cached as invalid, thereby blocking all messages
2830 to all local parts at the same domain, from all senders. This bug has
2831 been fixed. The domain is no longer cached after a negative response to
2832 MAIL if the sender used is not empty.
2834 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2837 (c) A new callout option, "use_sender" has been added for people who want
2838 the modified behaviour.
2844 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2847 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2848 indeed breaks things for older releases.
2850 3. Added additional logging to the case where there is a problem reading data
2851 from a filter that is running in a subprocess using a pipe, in order to
2852 try to track down a specific problem.
2854 4. Testing facility fudge: when running in the test harness and attempting
2855 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2856 getting "No route to host". Convert this to a timeout.
2858 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2861 6. Some OS don't have socklen_t but use size_t instead. This affects the
2862 fifth argument of getsockopt() amongst other things. This is now
2863 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2864 can be set for individual OS. I have set it for SunOS5, OSF1, and
2865 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2866 some earlier ones do not.
2868 7. Change 4.30/15 was not doing the test caselessly.
2870 8. The standard form for an IPv6 address literal was being rejected by address
2871 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2872 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2873 this, as well as the form without the "IPv6" on the front (but only when
2874 address literals are enabled, of course).
2876 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2878 10. Exim crashed if a message with an empty sender address specified by -f
2879 encountered a router with an errors_to setting. This could be provoked only
2880 by a command such as
2884 where an empty string was supplied; "<>" did not hit this bug.
2886 11. Installed PCRE release 4.5.
2888 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2889 remained set. It is now erased.
2891 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2892 times from message ids (which are base 36 rather than the normal 62).
2894 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2895 were being counted as actual protocol errors, and logged if the log
2896 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2897 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2898 and DATA following a set of rejected RCPTs do not count as protocol errors.
2899 In other words, Exim assumes they were pipelined, though this may not
2900 actually be the case. Of course, in all cases the client gets an
2901 appropriate error code.
2903 15. If a lookup fails in an ACL condition, a message about the failure may
2904 be available; it is used if testing the ACL cannot continue, because most
2905 such messages specify what the cause of the deferral is. However, some
2906 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2907 that caused an old message to be retained and used if a later statement
2908 caused a defer, replacing the real cause of the deferral.
2910 16. If an IP address had so many PTR records that the DNS lookup buffer
2911 was not large enough to hold them, Exim could crash while trying to process
2912 the truncated data. It now detects and logs this case.
2914 17. Further to 4.21/58, another change has been made: if (and only if) the
2915 first line of a message (the first header line) ends with CRLF, a bare LF
2916 in a subsequent header line has a space inserted after it, so as not to
2917 terminate the header.
2919 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2920 unnecessarily, used atoi() instead of strtol(), and didn't check the
2921 termination when getting file sizes from file names by regex.
2923 19. Completely re-implemented the support for maildirsize files, in the light
2924 of a number of problems with the previous contributed implementation
2925 (4.30/29). In particular:
2927 . If the quota is zero, the maildirsize file is maintained, but no quota is
2930 . If the maildir directory does not exist, it is created before any attempt
2931 to write a maildirsize file.
2933 . The quota value in the file is just a cache; if the quota is changed in
2934 the transport, the new value overrides.
2936 . A regular expression is available for excluding directories from the
2939 20. The autoreply transport checks the characters in options that define the
2940 message's headers; it allows continued headers, but it was checking with
2941 isspace() after an embedded newline instead of explicitly looking for a
2944 21. If all the "regular" hosts to which an address was routed had passed their
2945 expiry times, and had not reached their retry times, the address was
2946 bounced, even if fallback hosts were defined. Now Exim should go on to try
2949 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2950 equivalent code in the SMTP transport. Some hosts send humungous responses
2951 to HELO/EHLO, more than 1024 it seems.
2953 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2954 gives compiler warnings in some environments. I've now done it "properly",
2957 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2958 (because of problems with the built-in one) was declared to return uschar *
2959 instead of char *, causing compiler failure.
2961 25. Fixed a file descriptor leak when processing alias/forward files.
2963 26. Fixed a minor format string issue in dbfn.c.
2965 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2967 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2968 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2971 29. When a sender address is verified, it is cached, to save repeating the test
2972 when there is more than one recipient in a message. However, when the
2973 verification involves a callout, it is possible for different callout
2974 options to be set for different recipients. It is too complicated to keep
2975 track of this in the cache, so now Exim always runs a verification when a
2976 callout is required, relying on the callout cache for the optimization.
2977 The overhead is duplication of the address routing, but this should not be
2980 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2981 to be verified with callout=postmaster, and the main callout worked but the
2982 postmaster check failed, the verification correctly failed. However, if a
2983 subsequent RCPT command asked for sender verification *without* the
2984 postmaster check, incorrect caching caused this verification also to fail,
2987 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2988 it was not caching the DNS options (qualify_single, search_parents) that
2989 were used when the lookup failed. A subsequent lookup with different
2990 options therefore always gave the same answer, though there were cases
2991 where it should not have. (Example: a "domains = !$mx_any" option on a
2992 dnslookup router: the "domains" option is always processed without any
2993 widening, but the router might have qualify_single set.) Now Exim uses the
2994 cached value only when the same options are set.
2996 32. Added John Jetmore's "exipick" utility to the distribution.
2998 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2999 than a timeout (e.g. a certificate is required, and is not provided), an
3000 Exim server now closes the connection immediately. Previously it waited for
3001 the client to close - but if the client is SSL, it seems that they each
3002 wait for each other, leading to a delay before one of them times out.
3004 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3005 maintained 0.8.x compatibility because I don't think many are using it, and
3006 it is clearly obsolete.
3008 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3011 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3012 client certificate was expired. A simple patch fixes this, though I don't
3013 understand the full logic of why the verify callback is called multiple
3016 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3017 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3018 which causes problems with some clients (such as the Certicom SSL Plus
3019 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3020 disables the coutermeasure allowing Eudora to connect."
3022 38. Exim was not checking that a write() to a log file succeeded. This could
3023 lead to Bad Things if a log got too big, in particular if it hit a file
3024 size limit. Exim now panics and dies if it cannot write to a log file, just
3025 as it does if it cannot open a log file.
3027 39. Modified OS/Makefile-Linux so that it now contains
3029 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3031 The two -D definitions ensure that Exim is compiled with large file
3032 support, which makes it possible to handle log files that are bigger than
3035 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3036 instance) a domain was checked against a named list that involved a lookup,
3037 causing $domain_data to be set, then another domain was checked against the
3038 same list, then the first domain was re-checked, the value of $domain_data
3039 after the final check could be wrong. In particular, if the second check
3040 failed, it could be set empty. This bug probably also applied to
3043 41. The strip_trailing_dot option was not being applied to the address given
3044 with the -f command-line option.
3046 42. The code for reading a message's header from the spool was incrementing
3047 $received_count, but never initializing it. This meant that the value was
3048 incorrect (doubled) while delivering a message in the same process in which
3049 it was received. In the most common configuration of Exim, this never
3050 happens - a fresh exec is done - but it can happen when
3051 deliver_drop_privilege is set.
3053 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3054 it now includes up to 150 characters of the unexpected data in the log
3057 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3058 and building data strings. The size of both of these buffers was 10 000
3059 bytes - far larger than anybody would *ever* want, thought I. Needless to
3060 say, somebody hit the limit. I have increased the maximum line length to
3061 20 000 and the maximum data length of concatenated lines to 100 000. I have
3062 also fixed two bugs, because there was no checking on these buffers. Tsk,
3063 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3064 buffer is too small.
3066 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3067 lsearch lookups. Now it does.
3069 46. When parsing a route_list item in a manualroute router, a fixed-length
3070 buffer was used for the list of hosts. I made this 1024 bytes long,
3071 thinking that nobody would ever have a list of hosts that long. Wrong.
3072 Somebody had a whole pile of complicated expansion conditions, and the
3073 string was silently truncated, leading to an expansion error. It turns out
3074 that it is easier to change to an unlimited length (owing to other changes
3075 that have happened since this code was originally written) than to build
3076 structure for giving a limitation error. The length of the item that
3077 expands into the list of hosts is now unlimited.
3079 47. The lsearch lookup could not handle data where the length of text line was
3080 more than 4095 characters. Such lines were truncated, leading to shortened
3081 data being returned. It should now handle lines of any length.
3083 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3084 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3087 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3088 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3089 no longer happens in comment lines. A list of replacements is now placed
3090 at the head of all of the source files, except those whose only change is
3091 to replace PERL_COMMAND in the very first #! line.
3093 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3094 messages on the queue, with a bottom-up merge sort, using code contributed
3095 by Michael Haardt. This should make operations like -bp somewhat faster on
3096 large queues. It won't affect queue runners, except when queue_run_in_order
3099 51. Installed eximstats 1.31 in the distribution.
3101 52. Added support for SRV lookups to the dnslookup router.
3103 53. If an ACL referred to $message_body or $message_body_end, the value was not
3104 reset for any messages that followed in the same SMTP session.
3106 54. The store-handling optimization for building very long strings was not
3107 differentiating between the different store pools. I don't think this
3108 actually made any difference in practice, but I've tidied it.
3110 55. While running the routers to verify a sender address, $sender_address
3111 was still set to the sender address. This is wrong, because when routing to
3112 send a bounce to the sender, it would be empty. Therefore, I have changed
3113 it so that, while verifying a sender address, $sender_address is set to <>.
3114 (There is no change to what happens when verifying a recipient address.)
3116 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3117 target A or AAAA records (if not already returned) without resetting the
3118 qualify_single or search_parents options of the DNS resolver. These are
3119 inappropriate in this case because the targets of MX and SRV records must
3120 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3121 target that, when qualified, matched something in the local domain. These
3122 two options are now turned off when doing these lookups.
3124 57. It seems that at least some releases of Reiserfs (which does not have the
3125 concept of a fixed number of inodes) returns zero and not -1 for the
3126 number of available inodes. This interacted badly with check_spool_inodes,
3127 which assumed that -1 was the "no such thing" setting. What I have done is
3128 to check that the total number of inodes is greater than zero before doing
3129 the test of how many are available.
3131 58. When a "warn" ACL statement has a log_message modifier, the message is
3132 remembered, and not repeated. This is to avoid a lot of repetition when a
3133 message has many recipients that cause the same warning to be written.
3134 Howewer, Exim was preserving the list of already written lines for an
3135 entire SMTP session, which doesn't seem right. The memory is now reset if a
3136 new message is started.
3138 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3139 debugging output unless log_rewrite was also set.
3141 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3142 of (int)(handle) when we know that handle contains (void *)(-1).
3144 61. The Exim daemon panic-logs an error return when it closes the incoming
3145 connection. However "connection reset by peer" seems to be common, and
3146 isn't really an error worthy of noting specially, so that particular error
3149 62. When Exim is trying to find all the local interfaces, it used to panic and
3150 die if the ioctl to get the interface flags failed. However, it seems that
3151 on at least one OS (Solaris 9) it is possible to have an interface that is
3152 included in the list of interfaces, but for which you get a failure error
3153 for this call. This happens when the interface is not "plumbed" into a
3154 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3155 failure of the "get flags" call assumes that the interface is down.
3157 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3158 makes life easier for people who are doing arithmetic on fields extracted
3159 from dates, where you often get leading zeros that should not be
3160 interpreted as octal.
3162 64. Added qualify_domain to the redirect router, to override the global
3165 65. If a pathologically long header line contained very many addresses (the
3166 report of this problem mentioned 10 000) and each of them was rewritten,
3167 Exim could use up a very large amount of memory. (It kept on making new
3168 copies of the header line as it rewrote, and never released the old ones.)
3169 At the expense of a bit more processing, the header rewriting function has
3170 been changed so that it no longer eats memory in this way.
3172 66. The generation of the Received: header has been moved from the time that a
3173 message starts to be received, to the time that it finishes. The timestamp
3174 in the Received: header should now be very close to that of the <= log
3175 line. There are two side-effects of this change:
3177 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3178 logged header lines no longer include the local Received: line, because
3179 it has not yet been created. The same applies to a copy of the message
3180 that is returned to a non-SMTP sender when a message is rejected.
3182 (b) When a filter file is tested using -bf, no additional Received: header
3183 is added to the test message. After some thought, I decided that this
3186 This change does not affect the value of $received_for. It is still set
3187 after address rewriting, but before local_scan() is called.
3189 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3191 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3192 gave an unhelpful panic error message, and a defer error. I have managed to
3193 change this behaviour so that it now rejects any supplied certificate,
3194 which seems right, as the list of acceptable certificates is empty.
3196 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3197 gave an unhelpful defer error. I have not managed to make this reject any
3198 supplied certificates, but the error message it gives is "no certificate
3199 supplied", which is not helpful.
3201 70. exigrep's output now also includes lines that are not associated with any
3202 message, but which match the given pattern. Implemented by a patch from
3203 Martin Sluka, which also tidied up the Perl a bit.
3205 71. Recipient callout verification, like sender verification, was using <> in
3206 the MAIL FROM command. This isn't really the right thing, since the actual
3207 sender may affect whether the remote host accepts the recipient or not. I
3208 have changed it to use the actual sender in the callout; this means that
3209 the cache record is now keyed on a recipient/sender pair, not just the
3210 recipient address. There doesn't seem to be a real danger of callout loops,
3211 since a callout by the remote host to check the sender would use <>.
3212 [SEE ABOVE: changed after hitting problems.]
3214 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3215 temporary errors. However, in the case of such a code being given after
3216 the end of a data transmission (i.e. after ".") Exim was failing to write
3217 a retry record for the message. (Yes, there was some broken host that was
3218 actually sending 8xx at this point.)
3220 73. An unknown lookup type in a host list could cause Exim to panic-die when
3221 the list was checked. (An example that provoked this was putting <; in the
3222 middle of a list instead of at the start.) If this happened during a DATA
3223 ACL check, a -D file could be left lying around. This kind of configuration
3224 error no longer causes Exim to die; instead it causes a defer errror. The
3225 incident is still logged to the main and panic logs.
3227 74. Buglet left over from Exim 3 conversion. The message "too many messages
3228 in one connection" was written to the rejectlog but not the mainlog, except
3229 when address rewriting (yes!) was being logged.
3231 75. Added write_rejectlog option.
3233 76. When a system filter was run not as root (that is, when system_filter_user
3234 was set), the values of the $n variables were not being returned to the
3235 main process; thus, they were not subsequently available in the $sn
3238 77. Added +return_path_on_delivery log selector.
3240 78. A connection timeout was being treated differently from recipients deferred
3241 when testing hosts_max_try with a message that was older than the host's
3242 retry timeout. (The host should not be counted, thus allowing all hosts to
3243 be tried at least once before bouncing.) This may have been the cause of an
3244 occasionally reported bug whereby a message would remain on the queue
3245 longer than the retry timeout, but would be bounced if a delivery was
3246 forced. I say "may" because I never totally pinned down the problem;
3247 setting up timeout/retry tests is difficult. See also the next item.
3249 79. The ultimate address timeout was not being applied to errors that involved
3250 a combination of host plus message (for example, a timeout on a MAIL
3251 command). When an address resolved to a number of possible hosts, and they
3252 were not all tried for each delivery (e.g. because of hosts_max_try), a
3253 message could remain on the queue longer than the retry timeout.
3255 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3258 81. Fixed an obscure SMTP outgoing bug which required at least the following
3259 conditions: (a) there was another message waiting for the same server;
3260 (b) the server returned 5xx to all RCPT commands in the first message so
3261 that the message was not completed; (c) the server dropped the connection
3262 or gave a negative response to the RSET that Exim sends to abort the
3263 transaction. The observed case was a dropped connection after DATA that had
3264 been sent in pipelining mode. That is, the server had advertised PIPELINING
3265 but was not implementing it correctly. The effect of the bug was incorrect
3266 behaviour, such as trying another host, and this could lead to a crash.
3272 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3273 and daemon.c were passed as pointers to ints; they should have been
3274 pointers to socklen_t variables (which are typically unsigned ints).
3276 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3279 3. Fixed a really odd bug that affected only the testing scheme; patching a
3280 certain fixed string in the binary changed the value of another string that
3281 happened to be identical to the end of the original first string.
3283 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3284 name", it returns that address as the IP address. On some operating
3285 systems (e.g. Solaris), it also passes back the IP address string as the
3286 "host name". However, on others (e.g. Linux), it passes back an empty
3287 string. Exim wasn't checking for this, and was changing the host name to an
3288 empty string, assuming it had been canonicized.
3290 5. Although rare, it is permitted to have more than one PTR record for a given
3291 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3292 all the names associated with an address, because they do in Solaris.
3293 However, it seems that they do not in Linux for data that comes from the
3294 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3295 I found this out when I moved to a new Linux workstation and tried to run
3296 the Exim test suite.
3298 To get round this problem I have changed the code so that it now does its
3299 own call to the DNS to look up PTR records when searching for a host name.
3300 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3301 addresses that are only in /etc/hosts are still found.
3303 This behaviour is, however, controlled by an option called host_lookup_
3304 order, which defaults to "bydns:byaddr". If people want to use the other
3305 order, or indeed, just use one or the other means of lookup, they can
3306 specify it in this variable.
3308 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3309 some operating systems, this comes back from gethostbyaddr() as an empty
3310 string, and this is what Exim used to test for. However, it seems that in
3311 other systems, "." is yielded. Exim now tests for this case too.
3313 7. The values of check_spool_space and check_log_space are now held internally
3314 as a number of kilobytes instead of an absolute number of bytes. If a
3315 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3316 kilobyte. This means that much larger values can be stored.
3318 8. Exim monitor: an attempt to get the action menu when not actually pointing
3319 at a message produces an empty menu entitled "No message selected". This
3320 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3321 no entries in it ("Shell widget menu has zero width and/or height"). So I
3322 have added a single, blank menu entry in this case.
3324 9. Added ${quote_local_part.
3326 10. MIME decoding is now applied to the contents of Subject: header lines when
3329 11. Now that a reference to $sender_host_address automatically causes a reverse
3330 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3331 host lookup before query-style lookups in lists that might use this
3332 variable. This has therefore been abolished, and the "net-" prefix is no
3333 longer necessary for query-style lookups.
3335 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3336 have been a typo for LFLAGS, so it has been changed.
3338 13. The install script calls Exim with "-C /dev/null" in order to find the
3339 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3340 to be output. Howeve, since Exim outputs its version number before the
3341 error, it didn't break the script. It just looked ugly. I fixed this by
3342 always allowing "-C /dev/null" if the caller is root.
3344 14. Ignore overlarge ACL variable number when reading spool file - insurance
3345 against a later release with more variables having written the file.
3347 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3348 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3349 this, as well as the form without the "IPv6" on the front.
3351 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3352 OS/Makefile-Darwin file.
3354 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3355 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3357 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3358 variable. However, some OS use ints for inodes. I've added cast to long int
3359 to get rid of the compiler warning.
3361 19. I had forgotten to lock out "/../" in configuration file names when
3362 ALT_CONFIG_PREFIX was set.
3364 20. Routers used for verification do not need to specify transports. However,
3365 if such a router generated a host list, and callout was configured, Exim
3366 crashed, because it could not find a port number from the (non-existent)
3367 transport. It now assumes port 25 in this circumstance.
3369 21. Added the -t option to exigrep.
3371 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3372 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3373 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3374 them is set, LOOKUP_LSEARCH is forced.
3376 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3378 24. Added sender and host information to the "rejected by local_scan()" log
3379 line; previously there was no indication of these.
3381 25. Added .include_if_exists.
3383 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3384 Linux. It turns out that not all file systems support this. Apparently some
3385 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3386 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3387 not supported on the file descriptor, is now ignored when Exim is trying to
3388 sync a directory. This applies only to Linux.
3390 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3392 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3393 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3396 29. Added support for maildirsize files from supplied patch (modified a bit).
3398 30. The use of :fail: followed by an empty string could lead Exim to respond to
3399 sender verification failures with (e.g.):
3401 550 Verification failed for <xxx>
3402 550 Sender verify failed
3404 where the first response line was missing the '-' that indicates it is not
3405 the final line of the response.
3407 31. The loop for finding the name of the user that called Exim had a hardwired
3408 limit of 10; it now uses the value of finduser_retries, which is used for
3409 all other user lookups.
3411 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3414 33. Exim was neglecting to zero errno before one call of strtol() when
3415 expanding a string and expecting an integer value. On some systems this
3416 resulted in spurious "integer overflow" errors. Also, it was casting the
3417 result into an int without checking.
3419 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3420 did not work. The code looks as if it has *never* worked, though it appears
3421 to have been documented since at least releast 1.62. I have made it work.
3423 35. The "timeout_DNS" error in retry rules, also documented since at least
3424 1.62, also never worked. As it isn't clear exactly what this means, and
3425 clearly it isn't a major issue, I have abolished the feature by treating it
3426 as "timeout", and writing a warning to the main and panic logs.
3428 36. The display of retry rules for -brt wasn't always showing the error code
3431 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3432 timeout_connect_A, timeout_connect_MX.
3434 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3435 to the empty sender.
3437 39. The daemon was not analysing the content of -oX till after it had closed
3438 stderr and disconnected from the controlling terminal. This meant that any
3439 syntax errors were only noted on the panic log, and the return code from
3440 the command was 0. By re-arranging the code a little, I've made the
3441 decoding happen first, so such errors now appear on stderr, and the return
3442 code is 1. However, the actual setting up of the sockets still happens in
3443 the disconnected process, so errors there are still only recorded on the
3446 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3447 connections (as happens on some IP stacks) was logged at start up time as
3448 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3449 it from "IPv6 and IPv4", which means that two separate sockets are being
3452 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3453 says whether AF_INET or AF_INET6 was passed as an argument.
3455 42. Exiwhat output was messed up when time zones were included in log
3458 43. Exiwhat now gives more information about the daemon's listening ports,
3459 and whether -tls-on-connect was used.
3461 44. The "port" option of the smtp transport is now expanded.
3463 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3464 silently ignored. Now an error message is written to the main and panic
3467 46. There's a new ACL modifier called "logwrite" which writes to a log file
3468 as soon as it is encountered.
3470 47. Added $local_user_uid and $local_user_gid at routing time.
3472 48. Exim crashed when trying to verify a sender address that was being
3475 49. Exim was recognizing only a space character after ".include". It now also
3476 recognizes a tab character.
3478 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3479 extracting the relevant information from the specification. The man page no
3480 longer contains scrambled data for the -d option, and I've added a section
3481 at the front about calling Exim under different names.
3483 51. Added "extra_headers" argument to the "mail" command in filter files.
3485 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3488 53. Installed eximstats 1.29.
3490 54. Added transport_filter_timeout as a generic transport option.
3492 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3493 Cc: header lines. This was required by RFC 822, but it not required by RFC
3496 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3497 incoming messages that did not have them. Now it does so only if the
3498 message originates locally, that is, if there is no associated remote host
3499 address. When Resent- header lines are present, this applies to the Resent-
3500 lines rather than the non-Resent- lines.
3502 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3503 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3505 58. Messages for configuration errors now include the name of the main
3506 configuration file - useful now that there may be more than one file in a
3507 list (.included file names were always shown).
3509 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3510 for those rare installations that do not start the daemon as root or run it
3511 setuid root. I've cut out the call to initgroups() if the daemon is not
3514 60. The Exim user and group can now be bound into the binary as text strings
3515 that are looked up at the start of Exim's processing.
3517 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3519 62. Added $mailstore_basename variable.
3521 63. Installed patch to sieve.c from Michael Haardt.
3523 64. When Exim failed to open the panic log after failing to open the main log,
3524 the original message it was trying to log was written to stderr and debug
3525 output, but if they were not available (the usual case in production), it
3526 was lost. Now it is written to syslog before the two lines that record the
3527 failures to open the logs.
3529 65. Users' Exim filters run in subprocesses under the user's uid. It is
3530 possible for a "deliver" command or an alias in a "personal" command to
3531 provoke an address rewrite. If logging of address rewriting is configured,
3532 this fails because the process is not running as root or exim. There may be
3533 a better way of dealing with this, but for the moment (because 4.30 needs
3534 to be released), I have disabled address rewrite logging when running a
3535 filter in a non-root, non-exim process.
3541 1. The buildconfig auxiliary program wasn't quoting the value set for
3542 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3543 not defined. This bug was masked in 4.22 by the effect that was fixed in
3546 2. Some messages that were rejected after a message id was allocated were
3547 shown as "incomplete" by exigrep. It no longer does this for messages that
3548 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3550 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3551 have allow_domain_literals set, the ID did not get logged in the <= line.
3552 Domain literals are now always recognized in Message-ID: header lines.
3554 4. The first argument for a ${extract expansion item is the key name or field
3555 number. Leading and trailing spaces in this item were not being ignored,
3556 causing some misleading effects.
3558 5. When deliver_drop_privilege was set, single queue runner processes started
3559 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3560 same command in the process it spins off) were not dropping privilege.
3562 6. When the daemon running as "exim" started a queue runner, it always
3563 re-executed Exim in the spun-off process. This is a waste of effort when
3564 deliver_drop_privilege is set. The new process now just calls the
3565 queue-runner function directly.
3571 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3574 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3575 ignored. Though the use of -oP was forcing the writing of a pid file, it
3576 was always written to the default place.
3578 3. If the message "no IP address found for host xxxx" is generated during
3579 incoming verification, it is now followed by identification of the incoming
3580 connection (so you can more easily find what provoked it).
3582 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3584 5. Added some features to "harden" Exim a bit more against certain attacks:
3586 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3587 be put in Local/Makefile. This is like the never_users runtime option,
3588 but it cannot be overridden. The default setting is "root".
3590 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3591 prefix string with which any file named in a -C command line option
3594 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3595 is retained for -C and -D only if the caller of Exim is root. Without
3596 it, the exim user may also use -C and -D and retain privilege.
3598 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3599 command line option is disabled.
3601 6. Macro names set by the -D option must start with an upper case letter, just
3602 like macro names defined in the configuration file.
3604 7. Added "dereference=" facility to LDAP.
3606 8. Two instances of the typo "uknown" in the source files are fixed.
3608 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3609 the Configure-Makefile script screwed up while processing it.
3611 10. Incorporated PCRE 4.4.
3613 11. The SMTP synchronization check was not operating right at the start of an
3614 SMTP session. For example, it could not catch a HELO sent before the client
3615 waited for the greeting. There is now a check for outstanding input at the
3616 point when the greeting is written. Because of the duplex, asynchronous
3617 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3618 way, but not yet received, when the check is performed.
3620 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3621 on TCP/IP sockets, because this apparently causes some broken clients to
3624 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3625 unchanged) from the Cygwin maintainer.
3627 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3628 was supported instead of testing for mbx. Effectively a typo.
3630 15. The spa authenticator server code was not checking that the input it
3631 received was valid base64.
3633 16. The debug output line for the "set" modifier in ACLs was not showing the
3634 name of the variable that was being set.
3636 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3638 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3639 lookup on its own. Something else was needed to trigger the lookup. For
3640 example, a match in host_lookup or the need for a host name in a host list.
3641 Now, if $sender_host_name is referenced and the host name has not yet been
3642 looked up, a lookup is performed. If the lookup fails, the variable remains
3643 empty, and $host_lookup_failed is set to "1".
3645 19. Added "eqi" as a case-independent comparison operator.
3647 20. The saslauthd authentication condition could segfault if neither service
3648 nor realm was specified.
3650 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3651 error message that was logged was misleading, and incoming SMTP
3652 connections were dropped. The message is now more accurate, and temporary
3653 errors are given to SMTP connections.
3655 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3656 (except RSET) with a 421 error, until QUIT is received. However, it was
3657 failing to send a response to QUIT.
3659 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3660 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3661 ACL is now run after the helo_try_verify_hosts code.
3663 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3664 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3665 case-independent, so other case variants are also recognized). Apparently
3666 some systems use these upper case variants.
3668 25. If more than two messages were waiting for the same host, and a transport
3669 filter was specified for the transport, Exim sent two messages over the
3670 same TCP/IP connection, and then failed with "socket operation on non-
3671 socket" when it tried to send the third.
3673 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3675 27. The extern definition of crypt16() in expand.c was not being excluded when
3676 the OS had its own crypt16() function.
3678 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3679 as a preferred synonym for return_size_limit, both as an option and as an
3682 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3684 30. Changed the default configuration ACL to relax the local part checking rule
3685 for addresses that are not in any local domains. For these addresses,
3686 slashes and pipe symbols are allowed within local parts, but the sequence
3687 /../ is explicitly forbidden.
3689 31. SPA server authentication was not clearing the challenge buffer before
3692 32. log_message in a "warn" ACL statement was writing to the reject log as
3693 well as to the main log, which contradicts the documentation and doesn't
3694 seem right (because no rejection is happening). So I have stopped it.
3696 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3697 However, I am unable to do any testing of this.
3699 34. Fixed an infelicity in the appendfile transport. When checking directories
3700 for a mailbox, to see if any needed to be created, it was accidentally
3701 using path names with one or more superfluous leading slashes; tracing
3702 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3704 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3705 contents are added to the log line that is written for every discarded
3706 recipient. (Previously a log_message setting was ignored.)
3708 36. The ${quote: operator now quotes the string if it is empty.
3710 37. The install script runs exim in order to find its version number. If for
3711 some reason other than non-existence or emptiness, which it checks, it
3712 could not run './exim', it was installing it with an empty version number,
3713 i.e. as "exim-". This error state is now caught, and the installation is
3716 38. An argument was missing from the function that creates an error message
3717 when Exim fails to connect to the socket for saslauthd authentication.
3718 This could cause Exim to crash, or give a corrupted message.
3720 39. Added isip, isip4, and isip6 to ${if conditions.
3722 40. The ACL variables $acl_xx are now saved with the message, and can be
3723 accessed later in routers, transports, and filters.
3725 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3726 strings in the file are not string-expanded.
3728 42. If a MAIL command specified a SIZE value that was too large to fit into an
3729 int variable, the check against message_size_limit failed. Such values are
3730 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3731 day this will have to be increased, but I don't think I want to be around
3732 when emails are that large.
3739 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3740 iconv() is not standard in FreeBSD.
3742 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3743 IPv6 enabled. The observed symptom was a segmentation fault on return from
3744 the function os_common_find_running_interfaces() in src/os.c.
3746 3. In the check_special_case() function in daemon.c I had used "errno" as an
3747 argument name, which causes warnings on some systems. This was basically a
3748 typo, since it was named "eno" in the comments!
3750 4. The code that waits for the clock to tick (at a resolution of some fraction
3751 of a second) so as to ensure message-id uniqueness was always waiting for
3752 at least one whole tick, when it could have waited for less. [This is
3753 almost certainly not relevant at current processor speeds, where it is
3754 unlikely to ever wait at all. But we try to future-proof.]
3756 5. The function that sleeps for a time interval that includes fractions of a
3757 second contained a race. It did not block SIGALRM between setting the
3758 timer, and suspending (a couple of lines later). If the interval was short
3759 and the sigsuspend() was delayed until after it had expired, the suspension
3760 never ended. On busy systems this could lead to processes getting stuck for
3763 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3764 process, before it forks any delivery processes. The open lookup caching
3765 mechanism meant that the open file or database connection was passed into
3766 the delivery process. The problem was that delivery processes always tidy
3767 up cached lookup data. This could cause a problem for the next delivery
3768 process started by the queue runner, because the external queue runner
3769 process does not know about the closure. So the next delivery process
3770 still has data in the lookup cache. In the case of a file lookup, there was
3771 no problem because closing a file descriptor in a subprocess doesn't affect
3772 the parent. However, if the lookup was caching a connection to a database,
3773 the connection was closed, and the second delivery process was likely to
3774 see errors such as "PGSQL: query failed: server closed the connection
3775 unexpectedly". The problem has been fixed by closing all cached lookups
3776 in a queue runner before running a delivery process.
3778 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3779 seem to have the "const" qualifier which it has on other OS. I've
3782 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3783 *of the same type* (client or server) with the same public name that an
3784 error should be diagnosed.
3786 9. When Exim looked up a host name for an IP address, but failed to find the
3787 original IP address when looking up the host name (a safety check), it
3788 output the message "<ip address> does not match any IP for NULL", which was
3789 confusing, to say the least. The bug was that the host name should have
3790 appeared instead of "NULL".
3792 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3793 user that is built into the binary, and the -C or -D options is used, root
3794 privilege is dropped before the configuration file is read. In addition,
3795 logging is switched to stderr instead of the normal log files. If the
3796 configuration then re-defines the Exim user, the unprivileged environment
3797 is probably not what is expected, so Exim logs a panic warning message (but
3800 However, if deliver_drop_privilege is set, the unprivileged state may well
3801 be exactly what is intended, so the warning has been cut out in that case,
3802 and Exim is allowed to try to write to its normal log files.
3808 1. smtp_return_error_details was not giving details for temporary sender
3809 or receiver verification errors.
3811 2. Diagnose a configuration error if two authenticators have the same public
3814 3. Exim used not to create the message log file for a message until the first
3815 delivery attempt. This could be confusing when incoming messages were held
3816 for policy or load reasons. The message log file is now created at the time
3817 the message is received, and an initial "Received" line is written to it.
3819 4. The automatically generated man page for command line options had a minor
3820 bug that caused no ill effects; however, a more serious problem was that
3821 the procedure for building the man page automatically didn't always
3822 operate. Consequently, release 4.20 contains an out-of-date version. This
3823 shouldn't happen again.
3825 5. When building Exim with embedded Perl support, the script that builds the
3826 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3827 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3829 6. The freeze_tell option was not being used for messages that were frozen on
3830 arrival, either by an ACL or by local_scan().
3832 7. Added the smtp_incomplete_transaction log selector.
3834 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3835 was accepting AUTH without a new EHLO.
3837 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3838 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3841 10. Logging of TCP/IP connections (when configured) now happens in the main
3842 daemon process instead of the child process, so that the TCP/IP connection
3843 count is more accurate (but it can never be perfect).
3845 11. The use of "drop" in a nested ACL was not being handled correctly in the
3846 outer ACL. Now, if condition failure induced by the nested "drop" causes
3847 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3848 or "require"), the connection is dropped.
3850 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3851 that yield "discard" can now be used with an "accept" or a "discard" verb,
3852 but an error is generated for any others (because I can't see a useful way
3853 to define what should happen).
3855 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3856 are now processed in the same way as lines in the Exim configuration file.
3857 In particular, continuation lines are supported.
3859 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3861 15. Added -ti meaning -t -i.
3863 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3864 domains, and warn by logging if others are found.
3866 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3867 returned from GIFCONF when Exim is trying to find the list of interfaces on
3868 a host. The code in os.c has been modified to copy each ifreq to an aligned
3869 structure in all cases.
3871 Also, in some cases, the returned ifreq's were being copied to a 'struct
3872 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3873 means the last couple of bytes of an IPv6 address could be chopped if the
3874 ifreq contained only a normal sockaddr (14 bytes storage).
3876 18. Named domain lists were not supported in the hosts_treat_as_local option.
3877 An entry such as +xxxx was not recognized, and was treated as a literal
3880 19. Ensure that header lines added by a DATA ACL are included in the reject log
3881 if the ACL subsequently rejects the message.
3883 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3884 MD5 (which is deprecated).
3886 21. When testing a filter file using -bf, Exim was writing a message when it
3887 took the sender from a "From " line in the message, but it was not doing so
3888 when it took $return_path from a Return-Path: header line. It now does.
3890 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3891 with a valid header line field name (a series of printing characters
3892 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3894 23. Changed "disc" in the source to "disk" to conform to the documentation and
3895 the book and for uniformity.
3897 24. Ignore Sendmail's -Ooption=value command line item.
3899 25. When execve() failed while trying to run a command in a pipe transport,
3900 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3901 could be confused with a return value of 69 from the command itself. This
3902 has been changed to 127, the value the shell returns if it is asked to run
3903 a non-existent command. The wording for the related log line suggests a
3904 non-existent command as the problem.
3906 26. If received_header_text expands to an empty string, do not add a Received:
3907 header line to the message. (Well, it adds a token one on the spool, but
3908 marks it "old" so that it doesn't get used or transmitted.)
3910 27. Installed eximstats 1.28 (addition of -nt option).
3912 28. There was no check for failure on the call to getsockname() in the daemon
3913 code. This can fail if there is a shortage of resources on the system, with
3914 ENOMEM, for example. A temporary error is now given on failure.
3916 29. Contrary to the C standard, it seems that in some environments, the
3917 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3918 program. Exim now does this explicitly; it affects the formatting of
3919 timestamps using strftime().
3921 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3922 complaints. I've now initialized all the variables, to avoid this.
3924 32. Header lines added by a system filter were not being "seen" during
3925 transport-time rewrites.
3927 33. The info_callback() function passed to OpenSSL is set up with type void
3928 (*)(SSL *, int, int), as described somewhere. However, when calling the
3929 function (actually a macro) that sets it up, the type void(*)() is
3930 expected. I've put in a cast to prevent warnings from picky compilers.
3932 34. If a DNS black list lookup found a CNAME record, but there were no A
3933 records associated with the domain it pointed at, Exim crashed.
3935 35. If a DNS black list lookup returned more than one A record, Exim ignored
3936 all but the first. It now scans all returned addresses if a particular IP
3937 value is being sought. In this situation, the contents of the
3938 $dnslist_value variable are a list of all the addresses, separated by a
3941 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3942 to accept a host name and all its aliases if the forward lookup for any of
3943 them yielded the IP address of the incoming connection. Now it accepts only
3944 those names whose forward lookup yields the correct IP address. Any other
3945 names are discarded. This closes a loophole whereby a rogue DNS
3946 administrator could create reverse DNS records to break through a
3947 wildcarded host restriction in an ACL.
3949 37. If a user filter or a system filter that ran in a subprocess used any of
3950 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3951 the wrong values were passed to the pipe command ($thisaddress had the
3952 value of $0, $0 had the value of $1, etc). This bug was introduced by
3953 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3955 38. Improved the line breaking for long SMTP error messages from ACLs.
3956 Previously, if there was no break point between 40 and 75 characters, Exim
3957 left the rest of the message alone. Two changes have been made: (a) I've
3958 reduced the minimum length to 35 characters; (b) if it can't find a break
3959 point between 35 and 75 characters, it looks ahead and uses the first one
3960 that it finds. This may give the occasional overlong line, but at least the
3961 remaining text gets split now.
3963 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3964 file descriptors might be low, and that setting 1000 would always raise it.
3965 It turns out that in some environments, the limit is already over 1000 and
3966 that lowering it causes trouble. So now Exim takes care not to decrease it.
3968 40. When delivering a message, the value of $return_path is set to $sender_
3969 address at the start of routing (routers may change the value). By an
3970 oversight, this default was not being set up when an address was tested by
3971 -bt or -bv, which affected the outcome if any router or filter referred to
3974 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3975 log only when "message" or "log_message" are set. However, if one of the
3976 conditions was an address verification, or a call to a nested ACL, the
3977 messages generated by the underlying test were being passed through. This
3978 no longer happens. The underlying message is available in $acl_verify_
3979 message for both "message" and "log_message" expansions, so it can be
3980 passed through if needed.
3982 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3983 new expansion $bh_ to give the encoded byte string without charset
3984 translation. Translation happens only if iconv() is available; HAVE_ICONV
3985 indicates this at build time. HEADERS_CHARSET gives the charset to
3986 translate to; headers_charset can change it in the configuration, and
3987 "headers charset" can change it in an individual filter file.
3989 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3990 that creates RFC 2047 encoded "words" labels them as that charset instead
3991 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3992 expansion operator; (ii) when Exim creates a From: line for a local
3993 message; (iii) when a header line is rewritten to include a "phrase" part.
3995 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3996 buggy, causing it to skip the first lines of messages whose message ID
3997 ended in 'D'. This would not have bitten before Exim release 4.14, because
3998 message IDs were unlikely to end in 'D' before then. The effect was to have
3999 incorrect size information for certain domains.
4001 45. #include "config.h" was missing at the start of the crypt16.c module. This
4002 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4005 46. If there was a timeout during a "random" callout check, Exim treated it as
4006 a failure of the random address, and carried on sending RSET and the real
4007 address. If the delay was just some slowness somewhere, the response to the
4008 original RCPT would be taken as a response to RSET and so on, causing
4009 mayhem of various kinds.
4011 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4012 when I implemented it. It didn't allow for the fact that some option values
4013 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4014 the right test for positive values.
4016 48. Domain names in DNS records are case-independent. Exim always looks them up
4017 in lower case. Some resolvers return domain names in exactly the case they
4018 appear in the zone file, that is, they may contain uppercase letters. Not
4019 all resolvers do this - some return always lower case. Exim was treating a
4020 change of case by a resolver as a change of domain, similar to a widening
4021 of a domain abbreviation. This triggered its re-routing code and so it was
4022 trying to route what was effectively the same domain again. This normally
4023 caused routing to fail (because the router wouldn't handle the domain
4024 twice). Now Exim checks for this case specially, and just changes the
4025 casing of the domain that it ultimately uses when it transmits the message
4028 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4031 50. If a filter generated a file delivery with a non-absolute name (possible if
4032 no home directory exists for the router), the forbid_file option was not
4035 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4036 the existing equality matching.
4038 52. Exim was using ints instead of ino_t variables in some places where it was
4039 dealing with inode numbers.
4041 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4042 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4043 called TMPDIR, and if it finds it is different, it changes its value.
4045 54. The smtp_printf() function is now made available to local_scan() so
4046 additional output lines can be written before returning. There is also an
4047 smtp_fflush() function to enable the detection of a dropped connection.
4048 The variables smtp_input and smtp_batched_input are exported to
4051 55. Changed the default runtime configuration: the message "Unknown user"
4052 has been removed from the ACL, and instead placed on the localuser router,
4053 using the cannot_route_message feature. This means that any verification
4054 failures that generate their own messages won't get overridden. Similarly,
4055 the "Unrouteable address" message that was in the ACL for unverifiable
4056 relay addresses has also been removed.
4058 56. Added hosts_avoid_esmtp to the smtp transport.
4060 57. The exicyclog script was not checking for the esoteric option
4061 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4062 will work only if exicyclog is run under the appropriate euid.
4064 58. Following a discussion on the list, the rules by which Exim recognises line
4065 endings on incoming messages have been changed. The -dropcr and drop_cr
4066 options are now no-ops, retained only for backwards compatibility. The
4067 following line terminators are recognized: LF CRLF CR. However, special
4068 processing applies to CR:
4070 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4071 nor a local message in the state where . is a terminator.
4073 (ii) If a bare CR is encountered in a header line, an extra space is added
4074 after the line terminator so as not to end the header. The reasoning
4075 behind this is that bare CRs in header lines are most likely either
4076 to be mistakes, or people trying to play silly games.
4078 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4079 was being incorrectly given as 18 bytes larger than it should have been.
4080 This is a VOB (very old bug).
4082 60. This may never have affected anything current, but just in case it has:
4083 When the local host is found other than at the start of a list of hosts,
4084 the local host, those with the same MX, and any that follow, are discarded.
4085 When the list in question was part of a longer list of hosts, the following
4086 hosts (not currently being processed) were also being discarded. This no
4087 longer happens. I'm not sure if this situation could ever has previously
4090 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4093 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4094 header. This is something that is recommended in a new Internet Draft, and
4095 is something that is documented as being done by Sendmail. There are two
4096 possible values. For messages generated by the autoreply transport, Exim
4099 Auto-Submitted: auto-replied
4101 whereas for all other generated messages (e.g. bounces) it adds
4103 Auto-Submitted: auto-generated
4105 63. The "personal" condition in filters now includes a test for the
4106 Auto-Submitted: header. If it contains the string "auto-" the message it
4107 not considered personal.
4109 64. Added rcpt_include_affixes as a generic transport option.
4111 65. Added queue_only_override (default true).
4113 66. Added the syslog_duplication option.
4115 67. If what should have been the first header line of a message consisted of
4116 a space followed by a colon, Exim was mis-interpreting it as a header line.
4117 It isn't of course - it is syntactically invalid and should therefore be
4118 treated as the start of the message body. The misbehaviour could have
4119 caused a number of strange effects, including loss of data in subsequent
4120 header lines, and spool format errors.
4122 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4123 client host had authenticated. This control can now be exercised by an ACL
4124 for more flexibility.
4126 69. By default, callouts do not happen when testing with -bh. There is now a
4127 variant, -bhc, which does actually run the callout code, including
4128 consulting and updating the callout cache.
4130 70. Added support for saslauthd authentication, courtesy of Alexander
4133 71. If statvfs() failed on the spool or log directories while checking their
4134 size for availability, Exim confusingly gave the error "space shortage".
4135 Furthermore, in debugging mode it crashed with a floating point exception.
4136 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4137 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4138 serious problem, Exim now writes to the main and panic logs when this
4139 happens, with details of the failure. It then refuses to accept the
4140 incoming message, giving the message "spool directory problem" or "log
4141 directory problem" with a 421 code for SMTP messages.
4143 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4144 0, 1, and 2 exist, because some OS complain for execs without them (see
4145 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4146 descriptors. However, the code omitted to check that the open succeeded,
4147 causing mysterious errors if for some reason the permissions on /dev/null
4148 got screwed. Now Exim writes a message to the main and panic logs, and
4149 bombs out if it can't open /dev/null.
4151 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4152 interact so that it is all more flexible. It is supposed to remain
4153 backwards compatible. Also added extra_local_interfaces.
4155 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4156 to bomb out with an assertion failure - to the client this appears as a
4157 connection drop. This problem occurs in the part of the code that was taken
4158 from the Samba project. Fortunately, the assertion is in a very simple
4159 function, so I have fixed this by reproducing the function inline in the
4160 one place where it is called, and arranging for authentication to fail
4161 instead of killing the process with assert().
4163 75. The SPA client code was not working when the server requested OEM rather
4164 than Unicode encoding.
4166 76. Added code to make require_files with a specific uid setting more usable in
4167 the case where statting the file as root fails - usually a non-root-mounted
4168 NFS file system. When this happens and the failure is EACCES, Exim now
4169 forks a subprocess and does the per-uid checking as the relevant uid.
4171 77. Added process_log_path.
4173 78. If log_file_path was not explicitly set, a setting of check_log_space or
4174 check_log_inodes was ignored.
4176 79. If a space check for the spool or log partitions fails, the incident is now
4177 logged. Of course, in the latter case the data may get lost...
4179 80. Added the %p formatting code to string_format() so that it can be used to
4180 print addresses in debug_print(). Adjusted all the address printing in the
4181 debugging in store.c to use %p rather than %d.
4183 81. There was a concern that a line of code in smtp_in.c could overflow a
4184 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4185 initially expressed, the concern was not well-founded, because trailing
4186 spaces are removed early. However, if the trailing spaces were followed by
4187 a NULL, they did not get removed, so the overflow was possible. Two fixes
4190 (a) I re-wrote the offending code in a cleaner fashion.
4191 (b) If an incoming SMTP command contains a NULL character, it is rejected
4194 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4195 runs initgroups(), so that if the Exim user is in any additional groups,
4196 they will be used during message reception.
4202 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob