1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.387 2006/09/05 13:24:10 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).
30 SC/01 Use a glob alias rather than an array ref in eximstats generated
31 parser. This improves both readability and performance.
33 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
34 Don't display local sender or destination tables in eximstats unless
35 there is data to show.
36 Added average volumes into the eximstats top table text output.
38 SC/03 Collect data on the number of addresses (recipients) as well
39 as the number of messages in eximstats.
41 TF/01 Correct an error in the documentation for the redirect router. Exim
42 does (usually) call initgroups() when daemonizing.
44 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
45 with consistent privilege compared to when running as a daemon.
47 TF/03 Note in the spec that $authenticated_id is not set for local
48 submissions from trusted users.
50 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
51 Thanks to Dean Brooks <dean@iglou.com> for the patch.
53 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
54 by adding some example configuration directives to the default
55 configuration file. A little bit of work is required to uncomment the
56 directives and define how usernames and passwords are checked, but
57 there is now a framework to start from.
59 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
60 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
61 without this. I don't know how relevant this is to other LDAP libraries.
63 PH/02 Add the verb name to the "unknown ACL verb" error.
65 PH/03 Magnus Holmgren's patch for filter_prepend_home.
67 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
69 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
70 directory not expanded when it should be if an expanded home directory
71 was set for the address (which is overridden by the transport).
73 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
76 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
77 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
78 because it is too late at that time, and has no effect.
80 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
81 security issue with \' (bugzilla #107). I could not use the
82 PQescapeStringConn() function, because it needs a PGconn value as one of
85 PH/08 When testing addresses using -bt, indicate those final addresses that
86 are duplicates that would not cause an additional delivery. At least one
87 person was confused, thinking that -bt output corresponded to deliveries.
88 (Suppressing duplicates isn't a good idea as you lose the information
89 about possibly different redirections that led to the duplicates.)
91 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
92 systems where poll() doesn't work, in particular OS X.
94 PH/10 Added more information to debugging output for retry time not reached.
96 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
97 operations in malware.c.
99 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
102 PH/13 If write_rejectlog was set false when logging was sent to syslog with
103 syslog_duplication set false, log lines that would normally be written
104 both the the main log and to the reject log were not written to syslog at
107 PH/14 In the default configuration, change the use of "message" in ACL warn
108 statements to "add_header".
110 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
111 not followed by a command (e.g. "seen endif").
113 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
114 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
117 PH/17 Added extra conditions to the default value of delay_warning_condition
121 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
122 { match{$h_precedence:}{(?i)bulk|list|junk} } \
123 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
126 The Auto-Submitted: and various List- headers are standardised, whereas I
127 don't think Precedence: ever was.
129 PH/18 Refactored debugging code in route_finduser() to show more information,
130 in particular, the error code if getpwnam() issues one.
132 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
133 This is apparently needed in addition to the PH/07 change above to avoid
134 any possible encoding problems.
136 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
137 but not after initializing Perl.
139 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
140 output them only if debugging. By default they are written stderr,
141 apparently, which is not desirable.
143 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
146 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
149 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
151 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
152 authenticated or an ident call has been made. Suppress the default
153 values for $authenticated_id and $authenticated_sender (but permit -oMai
154 and -oMas) when testing with -bh.
156 PH/24 Re-jigged the order of the tests in the default configuration so that the
157 tests for valid domains and recipients precede the DNS black list and CSA
158 tests, on the grounds that those ones are more expensive.
160 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
161 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
162 This bug exists in every version of Exim that I still have, right back to
165 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
166 However, an attempt to turn on case-sensitivity in a regex key by
167 including (?-i) didn't work because the subject string was already
168 lowercased, and the effects were non-intuitive. It turns out that a
169 one-line patch can be used to allow (?-i) to work as expected.
175 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
176 other effects) broke the use of negated acl sub-conditions.
178 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
181 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
182 "Deny" causes Exim to reject the incoming connection with a 554 error.
183 Unfortunately, if there is a major crisis, such as a disk failure,
184 tcp-wrappers gives "deny", whereas what one would like would be some
185 kind of temporary error. A kludge has been added to help with this.
186 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
187 554 error is used if errno is still zero or contains ENOENT (which occurs
188 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
191 PH/03 Add -lutil to the default FreeBSD LIBS setting.
193 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
194 errors. Otherwise a message that provokes a temporary error (when other
195 messages do not) can cause a whole host to time out.
197 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
198 addresses were routed directly to files or pipes from a redirect router.
199 File deliveries just didn't batch; pipe deliveries might have suffered
202 PH/06 A failure to get a lock for a hints database would erroneously always say
203 "Failed to get write lock", even when it was really a read lock.
205 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
206 of 0600. This has been changed to use the value of the lockfile_mode
207 option (which defaults to 0600).
209 PH/08 Applied small patch from the Sieve maintainer.
211 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
212 folder from quota calculations, a direct delivery into this folder messed
213 up the contents of the maildirsize file. This was because the regex was
214 used only to exclude .Trash (or whatever) when the size of the mailbox
215 was calculated. There was no check that a delivery was happening into an
216 excluded directory. This bug has been fixed by ignoring all quota
217 processing for deliveries into excluded directories.
219 PH/10 Added the maildirfolder_create_regex option to appendfile.
225 PH/01 The code for finding all the local interface addresses on a FreeBSD
226 system running IPv6 was broken. This may well have applied to all BSD
227 systems, as well as to others that have similar system calls. The broken
228 code found IPv4 interfaces correctly, but gave incorrect values for the
229 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
230 that it would not match correctly against @[] and not recognize the IPv6
233 PH/02 The ipliteral router was not recognizing addresses of the form user@
234 [ipv6:....] because it didn't know about the "ipv6:" prefix.
236 PH/03 Added disable_ipv6.
238 PH/04 Changed $reply_address to use the raw form of the headers instead of the
239 decoded form, because it is most often used to construct To: headers
240 lines in autoreplies, and the decoded form may well be syntactically
241 invalid. However, $reply_address has leading white space removed, and all
242 newlines turned into spaces so that the autoreply transport does not
245 PH/05 If group was specified without a user on a router, and no group or user
246 was specified on a transport, the group from the router was ignored.
248 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
249 for visible compile-time settings that can be used to change these
250 numbers, for those that want even more. Backwards compatibility with old
251 spool files has been maintained. However, going back to a previous Exim
252 release will lost any variables that are in spool files.
254 PH/07 Two small changes when running in the test harness: increase delay when
255 passing a TCP/IP connection to a new process, in case the original
256 process has to generate a bounce, and remove special handling of
257 127.0.0.2 (sic), which is no longer necessary.
259 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
260 be the same on different OS.
262 PH/09 Moved a debug statement in filter processing to avoid a race problem when
265 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
266 whether --show-vars was specified or not
268 JJ/02 exipick: Added support for new ACL variable spool format introduced
271 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
272 syntactically invalid From: or Reply-to: line, and a filter used this to
273 generate an autoreply, and therefore failed to obtain an address for the
274 autoreply, Exim could try to deliver to a non-existent relative file
275 name, causing unrelated and misleading errors. What now happens is that
276 it logs this as a hard delivery error, but does not attempt to create a
279 PH/11 The exinext utility has a -C option for testing purposes, but although
280 the given file was scanned by exinext itself; it wasn't being passed on
281 when Exim was called.
283 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
284 an end-of-file indication when reading a command response.
286 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
287 compiled. In many other places in Exim, IPv6 addresses are always
288 recognized, so I have changed this. It also means that IPv4 domain
289 literals of the form [IPV4:n.n.n.n] are now always recognized.
291 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
292 used if the router is not running as root, for example, when verifying at
293 ACL time, or when using -bh. The debugging output from this situation was
294 non-existent - all you got was a failure to exec. I have made two
297 (a) Failures to set uid/gid, the current directory, or a process leader
298 in a subprocess such as that created by queryprogram now generate
299 suitable debugging ouput when -d is set.
301 (b) The queryprogram router detects when it is not running as root,
302 outputs suitable debugging information if -d is set, and then runs
303 the subprocess without attempting to change uid/gid.
305 PH/15 Minor change to Makefile for building test_host (undocumented testing
308 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
309 additional section of a DNS packet that returns MX or SRV records.
310 Instead, it always explicitly searches for A/AAAA records. This avoids
311 major problems that occur when a DNS server includes only records of one
312 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
313 fixed another bug: if SRV records were looked up and the corresponding
314 address records were *not* found in the additional section, the port
315 values from the SRV records were lost.
317 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
318 using the correct key (the original address) when searching the retry
319 rules in order to find which one to use for generating the retry hint.
321 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
322 adding the default one. Similarly, if it contains a Reply-To: header, the
323 errors_reply_to option, if set, is not used.
325 PH/19 When calculating a retry time, Exim used to measure the "time since
326 failure" by looking at the "first failed" field in the retry record. Now
327 it does not use this if it is later than than the arrival time of the
328 message. Instead it uses the arrival time. This makes for better
329 behaviour in cases where some deliveries succeed, thus re-setting the
330 "first failed" field. An example is a quota failure for a huge message
331 when small messages continue to be delivered. Without this change, the
332 "time since failure" will always be short, possible causing more frequent
333 delivery attempts for the huge message than are intended.
334 [Note: This change was subsequently modified - see PH/04 for 4.62.]
336 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
337 $1, $2, $3) because the numerical variables can be reset during some
338 expansion items (e.g. "match"), thereby losing the authentication data.
340 PH/21 Make -bV show the size of off_t variables so that the test suite can
341 decide whether to run tests for quotas > 2G.
343 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
344 mailbox_size, and mailbox_filecount in the appendfile transport. If a
345 filecount value is greater than 2G or if a quota value is greater than 2G
346 on a system where the size of off_t is not greater than 4, a panic error
349 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
350 never match. The debug and -bh output now contains an explicit error
351 message indicating a malformed IPv4 address or mask.
353 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
354 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
357 PH/25 Do not write to syslog when running in the test harness. The only
358 occasion when this arises is a failure to open the main or panic logs
359 (for which there is an explicit test).
361 PH/26 Added the /no_tell option to "control=freeze".
363 PH/27 If a host name lookup failed very early in a connection, for example, if
364 the IP address matched host_lookup and the reverse lookup yielded a name
365 that did not have a forward lookup, an error message of the form "no IP
366 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
367 could be logged. Now it outputs the IP address instead of "NULL".
369 PH/28 An enabling patch from MH: add new function child_open_exim2() which
370 allows the sender and the authenticated sender to be set when
371 submitting a message from within Exim. Since child_open_exim() is
372 documented for local_scan(), the new function should be too.
374 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
375 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
376 results in an empty string is now treated as unset.
378 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
380 PH/31 Added sender_verify_fail logging option.
382 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
383 needed by RFC 822 but not by RFC 2822 was commented out. I have now
384 tidied the source and removed it altogether.
386 PH/33 When a queue run was abandoned because the load average was too high, a
387 log line was always written; now it is written only if the queue_run log
388 selector is set. In addition, the log line for abandonment now contains
389 information about the queue run such as the pid. This is always present
390 in "start" and "stop" lines but was omitted from the "abandon" line.
392 PH/34 Omit spaces between a header name and the colon in the error message that
393 is given when verify = headers_syntax fails (if there are lots of them,
394 the message gets confusing).
396 PH/35 Change the default for dns_check_names_pattern to allow slashes within
397 names, as there are now some PTR records that contain slashes. This check
398 is only to protect against broken name servers that fall over on strange
399 characters, so the fact that it applies to all lookups doesn't matter.
401 PH/36 Now that the new test suite is complete, we can remove some of the
402 special code in Exim that was needed for the old test suite. For example,
403 sorting DNS records because real resolvers return them in an arbitrary
404 order. The new test suite's fake resolver always returns records in the
407 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
408 bounces) except when queue_only is set, to avoid logging races between
409 the different processes.
411 PH/38 Panic-die if .include specifies a non-absolute path.
413 PH/39 A tweak to the "H" retry rule from its user.
415 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
416 a label. They prevented compilation on older perls.
418 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
419 a warning to be raised on newish perls.
421 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
422 on queue. Changes to match documented behaviour of showing count of
423 messages matching specified criteria.
425 PH/40 Changed the default ident timeout from 30s to 5s.
427 PH/41 Added support for the use of login_cap features, on those BSD systems
428 that have them, for controlling the resources used by pipe deliveries.
430 PH/42 The content-scanning code uses fopen() to create files in which to put
431 message data. Previously it was not paying any attention to the mode of
432 the files. Exim runs with umask(0) because the rest of the code creates
433 files with open(), and sets the required mode explicitly. Thus, these
434 files were ending up world-writeable. This was not a big issue, because,
435 being within the spool directory, they were not world-accessible. I have
436 created a function called modefopen, which takes an additional mode
437 argument. It sets umask(777), creates the file, chmods it to the required
438 mode, then resets the umask. All the relevant calls to fopen() in the
439 content scanning code have been changed to use this function.
441 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
442 to 24 hours. This avoids potential overflow problems when processing G
443 and H retry rules. I suspect nobody ever tinkers with this value.
445 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
447 PH/45 When the plaintext authenticator is running as a client, the server's
448 challenges are checked to ensure they are valid base64 strings. By
449 default, the authentication attempt is cancelled if an invalid string is
450 received. Setting client_ignore_invalid_base64 true ignores these errors.
451 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
452 they are received. Thus, the responses can be made to depend on the
453 challenges. If an invalid string is ignored, an empty string is placed in
456 PH/46 Messages that are created by the autoreply transport now contains a
457 References: header, in accordance with RFCs 2822 and 3834.
459 PH/47 Added authenticated_sender_force to the smtp transport.
461 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
463 PH/49 Installed latest patch from the Sieve maintainer.
465 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
466 was also set, the appendfile transport was unnecessarily scanning a
467 directory of message files (e.g. for maildir delivery) to find the count
468 of files (along with the size), even though it did not need this
469 information. It now does the scan only if it needs to find either the
470 size of the count of files.
472 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
474 PH/52 Two bugs concerned with error handling when the smtp transport is
477 (i) Exim was not creating retry information for temporary errors given
478 for individual recipients after the DATA command when the smtp transport
479 was used in LMTP mode. This meant that they could be retried too
480 frequently, and not timed out correctly.
482 (ii) Exim was setting the flag that allows error details to be returned
483 for LMTP errors on RCPT commands, but not for LMTP errors for individual
484 recipients that were returned after the DATA command.
486 PH/53 This is related to PH/52, but is more general: for any failing address,
487 when detailed error information was permitted to be returned to the
488 sender, but the error was temporary, then after the final timeout, only
489 "retry timeout exceeded" was returned. Now it returns the full error as
490 well as "retry timeout exceeded".
492 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
493 do this, and (what is worse) MTAs that accept it.
495 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
496 will now be deprecated.
498 PH/56 New os.c-cygwin from the Cygwin maintainer.
500 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
501 formats (previously only available in exim formats via -bpr, -bpru,
502 and -bpra. Now also available in native and exiqgrep formats)
504 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
505 with very large, slow to parse queues
507 JJ/08 exipick: added ! as generic prefix to negate any criteria format
509 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
511 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
512 responses to authentication challenges, though it was showing the
513 challenges; (ii) I've removed the CR characters from the debug output for
516 PH/58 Allow for the insertion of a newline as well as a space when a string
517 is turned into more than one encoded-word during RFC 2047 encoding. The
518 Sieve code now uses this.
520 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
521 data_4xx, lost_connection, tls_required.
523 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
524 message was being sent as an SMTP response.
526 PH/61 Add -l and -k options to exicyclog.
528 PH/62 When verifying, if an address was redirected to one new address, so that
529 verification continued, and the new address failed or deferred after
530 having set something in $address_data, the value of $address_data was not
531 passed back to the ACL. This was different to the case when no
532 redirection occurred. The value is now passed back in both cases.
534 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
535 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
536 use login_cap.h, so on its own it isn't the distinguishing feature. The
537 new name refers directly to the setclassresources() function.
539 PH/65 Added configuration files for NetBSD3.
541 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
543 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
544 is preferred over IPv4.
546 PH/68 The bounce_return_message and bounce_return_body options were not being
547 honoured for bounces generated during the reception of non-SMTP messages.
548 In particular, this applied to messages rejected by the ACL. This bug has
549 been fixed. However, if bounce_return_message is true and bounce_return_
550 body is false, the headers that are returned for a non-SMTP message
551 include only those that have been read before the error was detected.
552 (In the case of an ACL rejection, they have all been read.)
554 PH/69 The HTML version of the specification is now built in a directory called
555 spec_html instead of spec.html, because the latter looks like a path with
556 a MIME-type, and this confuses some software.
558 PH/70 Catch two compiler warnings in sieve.c.
560 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
561 function verify_get_ident() calls ip_connect() to connect a socket, but
562 if the "connect()" function timed out, ip_connect() used to close the
563 socket. However, verify_get_ident() also closes the socket later, and in
564 between Exim writes to the log, which may get opened at this point. When
565 the socket was closed in ip_connect(), the log could get the same file
566 descriptor number as the socket. This naturally causes chaos. The fix is
567 not to close the socket in ip_connect(); the socket should be closed by
568 the function that creates it. There was only one place in the code where
569 this was missing, in the iplookup router, which I don't think anybody now
570 uses, but I've fixed it anyway.
572 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
573 well as to direct DNS lookups. Otherwise the handling of names in host
574 lists is inconsistent and therefore confusing.
580 PH/01 Two changes to the default runtime configuration:
582 (1) Move the checks for relay_from_hosts and authenticated clients from
583 after to before the (commented out) DNS black list checks.
585 (2) Add control=submission to the relay_from_hosts and authenticated
586 clients checks, on the grounds that messages accepted by these
587 statements are most likely to be submissions.
589 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
591 (1) Generate an error if the third argument for the ${prvs expansion is
594 (2) Treat a missing third argument of ${prvscheck as if it were an empty
597 (3) Reset the variables that are obtained from the first argument of
598 ${prvscheck and used in the second argument before leaving the code,
599 because their memory is reclaimed, so using them afterwards may do
602 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
603 one (it's much easier than Tom thought :-).
605 (5) Because of (4), we can now allow for the use of $prvscheck_result
606 inside the third argument.
608 PH/03 For some reason, the default setting of PATH when running a command from
609 a pipe transport was just "/usr/bin". I have changed it to
612 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
613 anything to be listed in the output from -bV.
615 PH/05 When a filter generated an autoreply, the entire To: header line was
616 quoted in the delivery log line, like this:
618 => >A.N.Other <ano@some.domain> <original@ddress> ...
620 This has been changed so that it extracts the operative address. There
621 may be more than one such address. If so, they are comma-separated, like
624 => >ano@some.domain,ona@other.domain <original@ddress> ...
626 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
627 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
628 not being looked up in the rDNS to get a host name, Exim was showing the
629 IP address twice in Received: lines, even though the IP addresses were
630 identical. For example:
632 Received: from [1.2.3.4] (helo=[1.2.3.4])
634 However, if the real host name was known, it was omitting the HELO data
635 if it matched the actual IP address. This has been tidied up so that it
636 doesn't show the same IP address twice.
638 PH/07 When both +timestamp and +memory debugging was on, the value given by
639 $tod_xxx expansions could be wrong, because the tod_stamp() function was
640 called by the debug printing, thereby overwriting the timestamp buffer.
641 Debugging no longer uses the tod_stamp() function when +timestamp is set.
643 PH/08 When the original message was included in an autoreply transport, it
644 always said "this is a copy of the message, including all the headers",
645 even if body_only or headers_only was set. It now gives an appropriate
648 PH/09 Applied a patch from the Sieve maintainer which:
650 o fixes some comments
651 o adds the (disabled) notify extension core
652 o adds some debug output for the result of if/elsif tests
653 o points to the current vacation draft in the documentation
654 and documents the missing references header update
658 o fixes a bug in processing the envelope test (when testing
659 multiple envelope elements, the last element determinted the
662 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
663 Electronic Mail") by including:
665 Auto-submitted: auto-generated
667 in the messages that it generates (bounce messages and others, such as
668 warnings). In the case of bounce messages for non-SMTP mesages, there was
669 also a typo: it was using "Auto_submitted" (underscore instead of
670 hyphen). Since every message generated by Exim is necessarily in response
671 to another message, thes have all been changed to:
673 Auto-Submitted: auto-replied
675 in accordance with these statements in the RFC:
677 The auto-replied keyword:
679 - SHOULD be used on messages sent in direct response to another
680 message by an automatic process,
682 - MUST NOT be used on manually-generated messages,
684 - MAY be used on Delivery Status Notifications (DSNs) and Message
685 Disposition Notifications (MDNs),
687 - MUST NOT be used on messages generated by automatic or periodic
688 processes, except for messages which are automatic responses to
691 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
692 to the default Received: header definition.
694 PH/12 Added log selector acl_warn_skipped (default on).
696 PH/13 After a successful wildlsearch lookup, discard the values of numeric
697 variables because (a) they are in the wrong storage pool and (b) even if
698 they were copied, it wouldn't work properly because of the caching.
700 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
701 checking when decoding. Apparently there are clients that generate
702 overlong encoded strings. Why am I not surprised?
704 PH/15 If the first argument of "${if match_address" was not empty, but did not
705 contain an "@" character, Exim crashed. Now it writes a panic log message
706 and treats the condition as false.
708 PH/16 In autoreply, treat an empty string for "once" the same as unset.
710 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
711 extension "envelope-auth". The code is finished and in agreement with
712 other implementations, but there is no documentation so far and in fact,
713 nobody wrote the draft yet. This extension is currently #undef'ed, thus
714 not changing the active code.
716 Print executed "if" and "elsif" statements when debugging is used. This
717 helps a great deal to understand what a filter does.
719 Document more things not specified clearly in RFC3028. I had all this
720 sorted out, when out of a sudden new issues came to my mind. Oops."
722 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
725 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
726 canonical form (as documented). However, after a host name lookup from
727 the IP address, check_host() was doing a simple string comparison with
728 addresses acquired from the DNS when checking that the found name did
729 have the original IP as one of its addresses. Since any found IPv6
730 addresses are likely to be in abbreviated form, the comparison could
731 fail. Luckily, there already exists a function for doing the comparison
732 by converting both addresses to binary, so now that is used instead of
735 PH/20 There was another similar case to PH/19, when a complete host name was
736 given in a host list; looking up its IP address could give an abbreviated
737 form, whereas the current host's name might or might not be abbreviated.
738 The same fix has been applied.
744 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
745 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
748 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
749 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
751 PH/03 Typo: missing ".o" in src/pcre/Makefile.
753 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
754 header line, restrict the check to what is listed in RFCs 2369 and 2929.
755 Also, for "Auto-Submitted", treat anything other than "no" as
756 non-personal, in accordance with RFC 3834. (Previously it treated
757 anything starting "auto-" as non-personal.)
759 TF/01 The control=submission/name=... option had a problem with syntax
760 errors if the name included a slash character. The /name= option
761 now slurps the rest of the string, so it can include any characters
762 but it must come last in the list of options (after /sender_retain
765 PH/05 Some modifications to the interface to the fake nameserver for the new
773 TK/01 Added the "success_on_redirect" address verification option. See
774 NewStuff for rationale and an example.
776 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
778 PH/02 Patch to exigrep to allow it to work on syslog lines.
780 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
781 fread() to skip over the body file's header line, because in Cygwin the
782 header line is locked and is inaccessible.
784 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
785 co-exist for some time) to make it clear that it is the Exim ID that is
786 referenced, not the Message-ID: header line.
788 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
789 string_format() function, because snprintf() does not exist on all
792 PH/06 The use of forbid_filter_existstest now also locks out the use of the
793 ${stat: expansion item.
795 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
796 protocol synchronization error", to keep the pedants happy.
798 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
799 well as for IRIX systems, when gcc is being used. See the host.c source
802 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
804 PH/10 Named domain lists were not working if used in a queue_smtp_domains
807 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
808 transport and to the smtp transport in LMTP mode.
810 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
812 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
813 run a filter in a subprocess. This could lead to confusion in subsequent
814 lookups in the parent process. There should also be a search_tidyup() at
815 the end of the subprocess.
817 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
818 only if the host matched helo_try_verify_hosts, which caused the
819 verification to occur when the EHLO/HELO command was issued. The ACL just
820 tested the remembered result. Now, if a previous verification attempt has
821 not happened, "verify = helo" does it there and then.
823 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
825 TK/03 Fix log output including CR from clamd.
827 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
828 exist provoked a memory error which could cause a segfault.
830 PH/15 Installed PCRE 6.2
832 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
834 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
835 of the problem. Specifically, suggested +O2 rather than +O1 for the
838 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
840 PH/20 If a delivery was routed to a non-standard port by means of an SRV
841 record, the port was not correctly logged when the outgoing_port log
842 selector was set (it logged the transort's default port).
844 PH/21 Added support for host-specific ports to manualroute, queryprogram,
845 fallback_hosts, and "hosts" in the smtp transport.
847 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
848 host errors such as "Connection refused".
850 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
851 authentication with radiusclient 0.4.9:
853 - Error returned from rc_read_config was caught wrongly
854 - Username/password not passed on to radius server due to wrong length.
856 The presumption is that some radiusclient API changes for 4.51/PH/17
857 were not taken care of correctly. The code is still untested by me (my
858 Linux distribution still has 0.3.2 of radiusclient), but it was
859 contributed by a Radius user.
861 PH/24 When doing a callout, the value of $domain wasn't set correctly when
862 expanding the "port" option of the smtp transport.
864 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
865 while reading a MIME header. Thanks to Tom Hughes for a patch.
867 PH/24 Include config.h inside local_scan.h so that configuration settings are
870 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
871 that in an ACL for RCPT (for example), you can examine exactly what was
874 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
875 commands, but it was not correctly comparing the address with the actual
876 client host address. Thus, it would show the EHLO address in Received:
877 header lines when this was not necessary.
879 PH/27 Added the % operator to ${eval:}.
881 PH/28 Exim tries to create and chdir to its spool directory when it starts;
882 it should be ignoring failures (because with -C, for example, it has lost
883 privilege). It wasn't ignoring creation failures other than "already
886 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
889 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
890 because an input line was too long, either on its own, or by virtue of
891 too many continuations, the temporary file was not being removed, and the
892 return code was incorrect.
894 PH/31 Missing "BOOL" in function definition in filtertest.c.
896 PH/32 Applied Sieve patches from the maintainer.
898 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
900 PH/33 Added "verify = not_blind".
902 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
903 Local/Makefile (with some defaults set). These are used in built scripts
904 such as exicyclog, but they have never been used in the exim_install
905 script (though there are many overriding facilities there). I have
906 arranged that the exim_install script now takes note of these two
909 PH/35 Installed configuration files for Dragonfly.
911 PH/36 When a locally submitted message by a trusted user did not contain a
912 From: header, and the sender address was obtained from -f or from an SMTP
913 MAIL command, and the trusted user did not use -F to supply a sender
914 name, $originator_name was incorrectly used when constructing a From:
915 header. Furthermore, $originator_name was used for submission mode
916 messages from external hosts without From: headers in a similar way,
917 which is clearly wrong.
919 PH/37 Added control=suppress_local_fixups.
921 PH/38 When log_selector = +received_sender was set, and the addition of the
922 sender made the log line's construction buffer exactly full, or one byte
923 less than full, an overflow happened when the terminating "\n" was
926 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
927 when the result of a list match is failure because a DNS lookup failed.
929 PH/40 RM_COMMAND is now used in the building process.
931 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
932 the "build-* directories that it finds.
934 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
935 address in a domain literal was a prefix of an interface address.
937 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
938 when verifying a sender address, unless rewrite_headers is false.
940 PH/44 Wrote a long comment about why errors_to addresses are verified as
941 recipients, not senders.
943 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
944 the ratelimit ACL was added.
946 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
948 PH/46 Added extra information about PostgreSQL errors to the error string.
950 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
951 avoiding the need to install special zones in a real server. This is
952 backwards compatible; if it can't find the fake resolver, it drops back.
953 Thus, both old and new test suites can be run.
955 TF/02 Added util/ratelimit.pl
957 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
958 clock is set back in time.
960 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
961 Candler <B.Candler@pobox.com>.
963 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
964 OK for addresses that are the result of redirections.
966 PH/48 A number of further additions for the benefit of the new test suite,
967 including a fake gethostbyname() that interfaces to the fake DNS resolver
970 TF/06 The fix for widen_domains has also been applied to qualify_single and
971 search_parents which are the other dnslookup options that can cause
974 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
977 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
979 TF/07 Exim produced the error message "an SRV record indicated no SMTP
980 service" if it encountered an MX record with an empty target hostname.
981 The message is now "an MX or SRV record indicated no SMTP service".
983 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
984 if the DNS of the sending site is misconfigured. This is quite a
985 common situation. This change restores the behaviour of treating a
986 helo verification defer as a failure.
988 PH/51 If self=fail was set on a router, the bounce message did not include the
989 actual error message.
995 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
997 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
998 command itself ended in error, the underlying message about the transport
999 filter timeout was being overwritten with the pipe command error. Now the
1000 underlying error message should be appended to the second error message.
1002 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1004 PH/02 Reduce the amount of output that "make" produces by default. Full output
1005 can still be requested.
1007 PH/03 The warning log line about a condition test deferring for a "warn" verb
1008 was being output only once per connection, rather than after each
1009 occurrence (because it was using the same function as for successful
1010 "warn" verbs). This seems wrong, so I have changed it.
1012 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1013 it should not have, which might have caused a crash in the right
1014 circumstances, but probably never did.
1016 PH/04 Installed a modified version of Tony Finch's patch to make submission
1017 mode fix the return path as well as the Sender: header line, and to
1018 add a /name= option so that you can make the user's friendly name appear
1021 TF/03 Added the control = fakedefer ACL modifier.
1023 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1024 Mark Lowes for thorough testing.
1026 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1028 TK/03 Merged latest SRS patch from Miles Wilton.
1030 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1031 with the definition in sysexits.h (which is #included earlier).
1032 Fortunately, Exim does not actually use EX_OK. The code used to try to
1033 preserve the sysexits.h value, by assumimg that macro definitions were
1034 scanned for macro replacements. I have been disabused of this notion,
1035 so now the code just undefines EX_OK before #including unistd.h.
1037 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1038 in the smtp transport. When a block could not be written in a single
1039 write() function, the timeout was being re-applied to each part-write.
1040 This seems wrong - if the receiver was accepting one byte at a time it
1041 would take for ever. The timeout is now adjusted when this happens. It
1042 doesn't have to be particularly precise.
1044 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1045 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1047 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1048 without a domain if the check to <postmaster@domain> fails.
1050 SC/01 Eximstats: added -xls and the ability to specify output files
1051 (patch written by Frank Heydlauf).
1053 SC/02 Eximstats: use FileHandles for outputing results.
1055 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1057 SC/04 Eximstats: fixed display of large numbers with -nvr option
1059 SC/05 Eximstats: fixed merging of reports with empty tables.
1061 SC/06 Eximstats: added the -include_original_destination flag
1063 SC/07 Eximstats: removed tabs and trailing whitespace.
1065 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1067 TK/06 MBOX spool code: Add real "From " MBOX separator line
1068 so the .eml file is really in mbox format (even though
1069 most programs do not really care). Patch from Alex Miller.
1071 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1072 The latter is generated from $received_to and is only set if the
1073 message has one envelope recipient. SA can use these headers,
1074 obviously out-of-the-box. Patch from Alex Miller.
1076 PH/08 The ${def test on a variable was returning false if the variable's
1077 value was "0", contrary to what the specification has always said!
1078 The result should be true unless the variable is empty.
1080 PH/09 The syntax error of a character other than { following "${if
1081 def:variable_name" (after optional whitespace) was not being diagnosed.
1082 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1083 accidental colon was present, for example, could give incorrect results.
1085 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1086 result is used (not including appendfile, where other changes are about
1089 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1090 This involved changing a lot of size variables from int to off_t. It
1091 should work with maildirs and everything.
1093 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1094 spamd dying while we are connected to it.
1096 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1099 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1100 function for BDB 4.3.
1102 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1104 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1105 the sources for PCRE have been re-arranged and more files are now
1108 PH/15 The code I had for printing potentially long long variables in PH/11
1109 above was not the best (it lost precision). The length of off_t variables
1110 is now inspected at build time, and an appropriate printing format (%ld
1111 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1112 to be "long long int" or "long int". This is needed for the internal
1113 formatting function string_vformat().
1115 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1116 the configuration file to be ":syslog", then the script "guesses" where
1117 the logs files are, rather than using the compiled in default. In our
1118 case the guess is not the same as the compiled default, so the script
1119 suddenly stopped working when I started to use syslog. The patch checks
1120 to see if log_file_path is "". If so, it attempts to read it from exim
1121 with no configuration file to get the compiled in version, before it
1122 falls back to the previous guessing code."
1124 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1125 implementing BATV in an Exim configuration. See NewStuff for the gory
1128 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1129 Makefile that are specific to HP-UX.
1131 PH/18 If the "use_postmaster" option was set for a recipient callout together
1132 with the "random" option, the postmaster address was used as the MAIL
1133 FROM address for the random test, but not for the subsequent recipient
1134 test. It is now used for both.
1136 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1137 patch removes a few documentation additions to RFC 3028, because the
1138 latest draft now contains them. It adds the new en;ascii-case comparator
1139 and a new error check for 8bit text in MIME parts. Comparator and
1140 require names are now matched exactly. I enabled the subaddress
1141 extension, but it is not well tested yet (read: it works for me)."
1143 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1144 rework some of the code of TK/09 above to avoid the hardwired use of
1145 "%lld" and "long long". Replaced the call to snprintf() with a call to
1148 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1149 records point to non-existent hosts", "retry timeout exceeded", and
1150 "retry time not reached for any host after a long failure period".
1152 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1153 experimental DomainKeys support:
1155 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1156 (2) On an error such as an illegally used "control", the wrong name for
1157 the control was given.
1159 These problems did NOT occur unless DomainKeys support was compiled.
1161 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1163 PH/24 Added ${if match_ip condition.
1165 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1166 databases so that it will be absolutely obvious if a crash occurs in the
1167 DB library. This is a regular occurrence (often caused by mis-matched
1170 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1171 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1172 picked up on a user's system that detects such things. There doesn't seem
1173 to be a gcc warning option for this - only an attribute that has to be
1174 put on the function's prototype. It seems that in Fedora Core 4 they have
1175 set this on a number of new functions. No doubt there will be more in due
1178 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1179 specify a transport. However, if an address that was verified by such a
1180 router was the subject of a callout, Exim crashed because it tried to
1181 read the rcpt_include_affixes from the non-existent transport. Now it
1182 just assumes that the setting of that option is false. This bug was
1183 introduced by 4.51/PH/31.
1185 PH/28 Changed -d+all to exclude +memory, because that information is very
1186 rarely of interest, but it makes the output a lot bigger. People tend to
1187 do -d+all out of habit.
1189 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1190 code in os-type was giving problems when libc.so lives in lib64, like on
1193 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1194 aren't the modern standard, and it seems that some systems' include files
1195 don't always have them. Exim was already checking for some of the newer
1196 ones like T_AAAA, and defining it itself. I've added checks for all the
1197 record types that Exim uses.
1199 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1200 not automatically generating a new one, as it is supposed to. This
1201 prevented TLS from working. If the file did exist, but contained invalid
1202 data, a new version was generated, as expected. It was only the case of a
1203 non-existent file that was broken.
1205 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1206 with a change in libdomainkeys > 0.64.
1208 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1209 from DNS. If the selector record carries the flag, it now has
1210 precedence over the domain-wide flag.
1212 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1214 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1215 the use of a port name that isn't defined in /etc/services) occurred, the
1216 message was deferred as in a normal delivery, and thus remained on the
1217 spool, instead of being failed because of the mua_wrapper setting. This
1218 is now fixed, and I tidied up some of the mua_wrapper messages at the
1221 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1222 the messages in a hash of arrays rather than using individual hashes.
1223 This is a bit cleaner and results in dramatic memory savings, albeit
1224 at a slight CPU cost.
1226 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1227 as requested by Marc Sherman.
1229 SC/10 Eximstats: added histograms for user specified patterns as requested
1232 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1234 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1235 fopen() in the content-scanning modules that did not already have it.
1241 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1242 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1244 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1246 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1248 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1249 to test Sieve filters that use "vacation".
1251 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1252 that changes the way the GnuTLS parameters are stored in the cache file.
1253 The new format can be generated externally. For backward compatibility,
1254 if the data in the cache doesn't make sense, Exim assumes it has read an
1255 old-format file, and it generates new data and writes a new file. This
1256 means that you can't go back to an older release without removing the
1259 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1260 work if there are any delivery delays because "one_time" forces the
1261 parent to be marked "delivered", so its unseen clone is never tried
1262 again. For this reason, Exim now forbids the simultaneous setting of
1265 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1266 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1267 entry if you want to know the details. The fix, however, neglected to
1268 consider the case when local delivery batching is involved. The test for
1269 "previously delivered" was not happening when checking to see if an
1270 address could be batched with a previous (undelivered) one; under
1271 certain circumstances this could lead to multiple deliveries to the same
1274 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1275 in its include files, and this causes problems building Exim.
1277 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1278 header_syntax) but Exim was just ignoring anything given after a slash.
1279 In particular, this caused confusion with an attempt to use "verify =
1280 reverse_host_lookup/defer_ok". An error is now given when options are
1281 supplied for verify items that do not have them. (Maybe reverse_host_
1282 lookup should have a defer_ok option, but that's a different point.)
1284 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1285 defined by RFC 821) to 2048, because there were problems with some AUTH
1286 commands, and RFC 1869 says the size should be increased for extended
1287 SMTP commands that take arguments.
1289 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1292 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1293 "unknown" error; now it says that the functionality isn't in the binary.
1295 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1296 an address' error message when a string expansion fails (syntax or
1297 whatever). Otherwise the password may appear in the log. Following change
1298 PH/42 below, there is no longer a chance of it appearing in a bounce
1301 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1303 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1304 of its addresses was ever tried. (Bugzilla bug #2.)
1306 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1307 the result incorrectly in the debug output. (It correctly added a newline
1308 to what was transported.)
1310 TF/01 Added $received_time.
1312 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1313 commented out examples of how to interface to a virus scanner and to
1314 SpamAssassin. Also added commented examples of av_scanner and
1315 spamd_address settings.
1317 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1318 and controls are allowed in which ACLs. There were a couple of minor
1319 errors. Some of the entries in the conditions table (which is a table of
1320 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1321 negation of where the condition IS allowed.
1323 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1325 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1326 header file does not have a version number, so I've had to invent a new
1327 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1328 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1329 radiusclient), but it was contributed by a Radius user.
1331 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1332 files or directories, for OpenSSL.
1334 PH/19 When an Exim process that is running as root has to create an Exim log
1335 file, it does so in a subprocess that runs as exim:exim so as to get the
1336 ownership right at creation (otherwise, other Exim processes might see
1337 the file with the wrong ownership). There was no test for failure of this
1338 fork() call, which would lead to the process getting stuck as it waited
1339 for a non-existent subprocess. Forks do occasionally fail when resources
1340 run out. I reviewed all the other calls to fork(); they all seem to check
1343 PH/20 When checking for unexpected SMTP input at connect time (before writing
1344 the banner), Exim was not dealing correctly with a non-positive return
1345 from the read() function. If the client had disconnected by this time,
1346 the result was a log entry for a synchronization error with an empty
1347 string after "input=" when read() returned zero. If read() returned -1
1348 (an event I could not check), uninitialized data bytes were printed.
1349 There were reports of junk text (parts of files, etc) appearing after
1352 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1354 PH/22 Added support for macro redefinition, and (re)definition in between
1355 driver and ACL definitions.
1357 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1358 forgetting to use the resulting value; it was using the unexpanded value.
1360 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1361 hadn't been configured. The fix is from Juergen Kreileder, who
1362 understands it better than I do:
1364 "Here's what I see happening with three configured cyrus_sasl
1365 authenticators configured (plain, login, cram-md5):
1367 On startup auth_cyrus_sasl_init() gets called for each of these.
1368 This means three calls to sasl_listmech() without a specified mech_list.
1369 => SASL tests which mechs of all available mechs actually work
1370 => three warnings about OTP not working
1371 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1373 With the patch, sasl_listmech() also gets called three times. But now
1374 SASL's mech_list option is set to the server_mech specified in the the
1375 authenticator. Or in other words, the answer from sasl_listmech()
1376 gets limited to just the mech you're testing for (which is different
1378 => the return list contains just 'plain' or 'login', 'cram-md5' or
1379 nothing depending on the value of ob->server_mech.
1381 I've just tested the patch: Authentication still works fine,
1382 unavailable mechs specified in the exim configuration are still
1383 caught, and the auth.log warnings about OTP are gone."
1385 PH/25 When debugging is enabled, the contents of the command line are added
1386 to the debugging output, even when log_selector=+arguments is not
1389 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1390 answer is "GNU", and only if the return is "GNU/something" is the answer
1393 PH/27 $acl_verify_message is now set immediately after the failure of a
1394 verification in an ACL, and so is available in subsequent modifiers. In
1395 particular, the message can be preserved by coding like this:
1397 warn !verify = sender
1398 set acl_m0 = $acl_verify_message
1400 Previously, $acl_verify_message was set only while expanding "message"
1401 and "log_message" when a very denied access.
1403 PH/28 Modified OS/os.c-Linux with
1405 -#ifndef OS_LOAD_AVERAGE
1406 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1408 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1409 nomenclature these days.)
1411 PH/29 Installed patch from the Sieve maintainer that adds the options
1412 sieve_useraddress and sieve_subaddress to the redirect router.
1414 PH/30 In these circumstances:
1415 . Two addresses routed to the same list of hosts;
1416 . First host does not offer TLS;
1417 . First host accepts first address;
1418 . First host gives temporary error to second address;
1419 . Second host offers TLS and a TLS session is established;
1420 . Second host accepts second address.
1421 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1422 and peerdn, if requested) that were in fact used only for the second
1425 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1426 attention to any local part prefix or suffix that was matched by the
1427 router that accepted the address. It now behaves in the same way as it
1428 does for delivery: the affixes are removed from the local part unless
1429 rcpt_include_affixes is set on the transport.
1431 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1432 timeout during the DATA phase of an incoming message.
1434 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1435 applied a patch sanctioned by the Sieve maintainer.
1437 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1438 the uid or gid is negative. A case of a negative gid caused this to be
1439 noticed. The fix allows for either to be negative.
1441 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1442 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1445 PH/36 The change PH/12 above was broken. Fixed it.
1447 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1448 the grounds that routing the same address twice would always produce the
1449 same answer. This might have been true once, but it is certainly no
1450 longer true now. Routing a child address may depend on the previous
1451 routing that produced that child. Some complicated redirection strategies
1452 went wrong when messages had multiple recipients, and made Exim's
1453 behaviour dependent on the order in which the addresses were given.
1455 I have moved the duplicate checking until after the routing is complete.
1456 Exim scans the addresses that are assigned to local and remote
1457 transports, and removes any duplicates. This means that more work will be
1458 done, as duplicates will always all be routed, but duplicates are
1459 presumably rare, so I don't expect this is of any significance.
1461 For deliveries to pipes, files, and autoreplies, the duplicate checking
1462 still happens during the routing process, since they are not going to be
1465 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1466 It corrects a timeout issue with spamd. This is Ian's comment: "The
1467 background is that sometimes spamd either never reads data from a
1468 connection it has accepted, or it never writes response data. The exiscan
1469 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1470 blindly assumes that writes won't block so it may never time out."
1472 PH/39 Allow G after quota size as well as K and M.
1474 PH/40 The value set for $authenticated_id in an authenticator may not contain
1475 binary zeroes or newlines because the value is written to log lines and
1476 to spool files. There was no check on this. Now the value is run through
1477 the string_printing() function so that such characters are converted to
1478 printable escape sequences.
1480 PH/41 $message_linecount is a new variable that contains the total number of
1481 lines in the message. Compare $body_linecount, which is the count for the
1484 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1485 bounce and delay warning messages, except in certain special cases, which
1488 (a) An SMTP error message from a remote host;
1489 (b) A message specified in a :fail: redirection;
1490 (c) A message specified in a "fail" command in a system filter;
1491 (d) A message specified in a FAIL return from the queryprogram router;
1492 (e) A message specified by the cannot_route_message router option.
1494 In these cases only, Exim does include the error details in bounce and
1495 warning messages. There are also a few cases where bland messages such
1496 as "unrouteable address" or "local delivery error" are given.
1498 PH/43 $value is now also set for the "else" part of a ${run expansion.
1500 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1501 being worked on, but at least Exim now implements the latest version to
1504 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1505 process to complete was treated as a delivery failure, a timeout while
1506 writing the message to the pipe was logged, but erroneously treated as a
1507 successful delivery. Such timeouts include transport filter timeouts. For
1508 consistency with the overall process timeout, these timeouts are now
1509 treated as errors, giving rise to delivery failures by default. However,
1510 there is now a new Boolean option for the pipe transport called
1511 timeout_defer, which, if set TRUE, converts the failures into defers for
1512 both kinds of timeout. A transport filter timeout is now identified in
1515 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1516 systems where "make" and "gmake" are different, calling "gmake" at top
1517 level broke things. I've arranged for the value of $(MAKE) to be passed
1518 from the Makefile to this script so that it can call the same version of
1522 A note about Exim versions 4.44 and 4.50
1523 ----------------------------------------
1525 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1526 changes of various kinds. As a consequence, a big documentation update was
1527 needed. This delayed the release for rather longer than seemed good, especially
1528 in the light of a couple of (minor) security issues. Therefore, the changes
1529 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1530 release. So 4.44 and 4.50 are in effect two different branches that both start
1533 I have left the 4.50 change log unchanged; it contains all the changes since
1534 4.43. The change log for 4.44 is below; many of its items are identical to
1535 those for 4.50. This seems to be the most sensible way to preserve the
1536 historical information.
1542 1. Minor wording change to the doc/README.SIEVE file.
1544 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1545 computation of the current number of files was incorrect.
1547 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1548 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1551 4. Give more explanation in the error message when the command for a transport
1552 filter fails to execute.
1554 5. There are several places where Exim runs a non-Exim command in a
1555 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1556 was being done only for the command run by the queryprogram router. It is
1557 now done for all such subprocesses. The other cases are: ${run, transport
1558 filters, and the commands run by the lmtp and pipe transports.
1560 6. Added CONFIGURE_GROUP build-time option.
1562 7. Some older OS have a limit of 256 on the maximum number of file
1563 descriptors. Exim was using setrlimit() to set 1000 as a large value
1564 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1565 systems. I've change it so that if it can't get 1000, it tries for 256.
1567 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1568 was an oversight, and furthermore, ever since the addition of extra
1569 controls (e.g. 4.43/32), the checks on when to allow different forms of
1570 "control" were broken. There should now be diagnostics for all cases when a
1571 control that does not make sense is encountered.
1573 9. Added the /retain_sender option to "control=submission".
1575 10. $recipients is now available in the predata ACL (oversight).
1577 11. Tidy the search cache before the fork to do a delivery from a message
1578 received from the command line. Otherwise the child will trigger a lookup
1579 failure and thereby defer the delivery if it tries to use (for example) a
1580 cached ldap connection that the parent has called unbind on.
1582 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1583 of $address_data from the recipient verification was clobbered by the
1584 sender verification.
1586 13. The value of address_data from a sender verification is now available in
1587 $sender_address_data in subsequent conditions in the ACL statement.
1589 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1591 15. Added a new option "connect=<time>" to callout options, to set a different
1594 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1595 was its contents. (It was OK if the option was not defined at all.)
1597 17. A "Completed" log line is now written for messages that are removed from
1598 the spool by the -Mrm option.
1600 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1601 information about exactly what failed.
1603 19. Added -dd to debug only the daemon process.
1605 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1606 handles timeouts, both on the server side and network timeouts. Renamed the
1607 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1609 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1610 It is now set to "smtps".
1612 22. $host_address is now set to the target address during the checking of
1613 ignore_target_hosts.
1615 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1616 being passed; this would have caused $sender_host_name to have been used if
1617 matching the list had actually called for a host name (not very likely,
1618 since this list is usually IP addresses). A host name is now passed as
1621 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1622 code that specifies a non-restarting handler (typically sigaction() in
1623 modern systems) in an attempt to fix a rare and obscure crash bug.
1625 25. Narrowed the window for a race in the daemon that could cause it to ignore
1626 SIGCHLD signals. This is not a major problem, because they are used only to
1627 wake it up if nothing else does.
1629 26. A malformed maildirsize file could cause Exim to calculate negative values
1630 for the mailbox size or file count. Odd effects could occur as a result.
1631 The maildirsize information is now recalculated if the size or filecount
1634 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1635 support for a long time. Removed HAVE_SYS_VFS_H.
1637 28. Installed the latest version of exipick from John Jetmore.
1639 29. In an address list, if the pattern was not a regular expression, an empty
1640 subject address (from a bounce message) matched only if the pattern was an
1641 empty string. Non-empty patterns were not even tested. This was the wrong
1642 because it is perfectly reasonable to use an empty address as part of a
1643 database query. An empty address is now tested by patterns that are
1644 lookups. However, all the other forms of pattern expect the subject to
1645 contain a local part and a domain, and therefore, for them, an empty
1646 address still always fails if the pattern is not itself empty.
1648 30. Exim went into a mad DNS loop when attempting to do a callout where the
1649 host was specified on an smtp transport, and looking it up yielded more
1650 than one IP address.
1652 31. Re-factored the code for checking spool and log partition space into a
1653 function that finds that data and another that does the check. The former
1654 is then used to implement four new variables: $spool_space, $log_space,
1655 $spool_inodes, and $log_inodes.
1657 32. The RFC2047 encoding function was originally intended for short strings
1658 such as real names; it was not keeping to the 75-character limit for
1659 encoded words that the RFC imposes. It now respects the limit, and
1660 generates multiple encoded words if necessary. To be on the safe side, I
1661 have increased the buffer size for the ${rfc2047: expansion operator from
1664 33. It is now permitted to omit both strings after an "if" condition; if the
1665 condition is true, the result is "true". As before, when the second string
1666 is omitted, a false condition yields an empty string. This makes it less
1667 cumbersome to write custom ACL and router conditions.
1669 34. Failure to deliver a bounce message always caused it to be frozen, even if
1670 there was an errors_to setting on the router. The errors_to setting is now
1673 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1674 canonical form (fully expanded) before being placed in
1675 $sender_host_address.
1677 36. The table in the code that translates DNS record types into text (T_A to
1678 "A" for instance) was missing entries for NS and CNAME. It is just possible
1679 that this could have caused confusion if both these types were looked up
1680 for the same domain, because the text type is used as part of Exim's
1681 per-process caching. But the chance of anyone hitting this buglet seems
1684 37. The dnsdb lookup has been extended in a number of ways.
1686 (1) There is a new type, "zns", which walks up the domain tree until it
1687 finds some nameserver records. It should be used with care.
1689 (2) There is a new type, "mxh", which is like "mx" except that it returns
1690 just the host names, not the priorities.
1692 (3) It is now possible to give a list of domains (or IP addresses) to be
1693 looked up. The behaviour when one of the lookups defers can be
1694 controlled by a keyword.
1696 (4) It is now possible to specify the separator character for use when
1697 multiple records are returned.
1699 38. The dnslists ACL condition has been extended: it is now possible to supply
1700 a list of IP addresses and/or domains to be looked up in a particular DNS
1703 39. Added log_selector=+queue_time_overall.
1705 40. When running the queue in the test harness, wait just a tad after forking a
1706 delivery process, to get repeatability of debugging output.
1708 41. Include certificate and key file names in error message when GnuTLS fails
1709 to set them up, because the GnuTLS error message doesn't include the name
1710 of the failing file when there is a problem reading it.
1712 42. Allow both -bf and -bF in the same test run.
1714 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1716 44. The "Exiscan patch" is now merged into the mainline Exim source.
1718 45. Sometimes the final signoff response after QUIT could fail to get
1719 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1720 < 0 before doing a fflush(). This bug looks as though it goes back to the
1721 introduction of TLS in release 3.20, but "sometimes" must have been rare
1722 because the tests only now provoked it.
1724 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1725 (this can affect the format of dates).
1727 47. exim_tidydb, when checking for the continued existence of a message for
1728 which it has found a message-specific retry record, was not finding
1729 messages that were in split spool directories. Consequently, it was
1730 deleting retry records that should have stayed in existence.
1732 48. Steve fixed some bugs in eximstats.
1734 49. The SPA authentication driver was not abandoning authentication and moving
1735 on to the next authenticator when an expansion was forced to fail,
1736 contradicting the general specification for all authenticators. Instead it
1737 was generating a temporary error. It now behaves as specified.
1739 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1740 (the order specifies the preference for clients). The order is now AES256,
1741 AES128, 3DES, ARCFOUR128.
1743 51. Small patch to Sieve code - explicitly set From: when generating an
1746 52. Exim crashed if a remote delivery caused a very long error message to be
1747 recorded - for instance if somebody sent an entire SpamAssassin report back
1748 as a large number of 550 error lines. This bug was coincidentally fixed by
1749 increasing the size of one of Exim's internal buffers (big_buffer) that
1750 happened as part of the Exiscan merge. However, to be on the safe side, I
1751 have made the code more robust (and fixed the comments that describe what
1754 53. Now that there can be additional text after "Completed" in log lines (if
1755 the queue_time_overall log selector is set), a one-byte patch to exigrep
1756 was needed to allow it to recognize "Completed" as not the last thing in
1759 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1760 patch that reportedly fixes this has been added. I am not expert enough to
1761 create a test for it. This is what the patch creator wrote:
1763 "I found a little strange behaviour of ldap code when working with
1764 Windows 2003 AD Domain, where users was placed in more than one
1765 Organization Units. When I tried to give exim partial DN, the exit code
1766 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1767 But simultaneously result of request was absolutely normal ldap result,
1768 so I produce this patch..."
1770 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1771 REFERENCE, so I have modified the code to exclude the patch when that macro
1774 55. Some experimental protocols are using DNS PTR records for new purposes. The
1775 keys for these records are domain names, not reversed IP addresses. The
1776 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1777 leaves it alone. Component reversal etc. now happens only for IP addresses.
1780 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1782 57. Double the size of the debug message buffer (to 2048) so that more of very
1783 long debug lines gets shown.
1785 58. The exicyclog utility now does better if the number of log files to keep
1786 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1788 59. Two changes related to the smtp_active_hostname option:
1790 (1) $smtp_active_hostname is now available as a variable.
1791 (2) The default for smtp_banner uses $smtp_active_hostname instead
1792 of $primary_hostname.
1794 60. The host_aton() function is supposed to be passed a string that is known
1795 to be a valid IP address. However, in the case of IPv6 addresses, it was
1796 not checking this. This is a hostage to fortune. Exim now panics and dies
1797 if the condition is not met. A case was found where this could be provoked
1798 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1799 components; fortuitously, this particular loophole had already been fixed
1800 by change 4.50/55 above.
1802 If there are any other similar loopholes, the new check in host_aton()
1803 itself should stop them being exploited. The report I received stated that
1804 data on the command line could provoke the exploit when Exim was running as
1805 exim, but did not say which command line option was involved. All I could
1806 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1807 running as the user.
1810 61. There was a buffer overflow vulnerability in the SPA authentication code
1811 (which came originally from the Samba project). I have added a test to the
1812 spa_base64_to_bits() function which I hope fixes it.
1815 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1816 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1818 63. The daemon start-up calls getloadavg() while still root for those OS that
1819 need the first call to be done as root, but it missed one case: when
1820 deliver_queue_load_max is set with deliver_drop_privilege. This is
1821 necessary for the benefit of the queue runner, because there is no re-exec
1822 when deliver_drop_privilege is set.
1824 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1825 This has been fixed.
1827 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1828 was in use, was not putting the data itself into the right store pool;
1829 consequently, it could be overwritten for a subsequent message in the same
1830 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1833 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1835 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1836 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1837 the function were treating the return as a boolean value, which happened to
1838 work because 0=false and not-0=true, but is not correct code.
1840 68. The host_aton() function was not handling scoped IPv6 addresses (those
1841 with, for example, "%eth0" on the end) correctly.
1843 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1844 negated items (that is, ~something) in unsigned ints. Some compilers
1845 apparently mutter when there is no cast.
1847 70. If an address verification called from an ACL failed, and did not produce a
1848 user-specific message (i.e. there was only a "system" message), nothing was
1849 put in $acl_verify_message. In this situation, it now puts the system
1852 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1853 session; change 4.31/43 added the unwanted input to the log line - except
1854 that it did not do this in the start of session case. It now does.
1856 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1857 This is wrong and can cause the other end to generate a synchronization
1858 error if it is another Exim or anything else that does the synchronization
1859 check. A QUIT command is no longer sent after a timeout.
1861 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1862 during host lookups.
1864 74. The defer_ok option of callout verification was not working if it was used
1865 when verifying addresses in header lines, that is, for this case:
1867 verify = header_sender/callout=defer_ok
1869 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1870 those file descriptors could be used for SMTP connections. If anything
1871 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1872 could be sent to the SMTP client, causing chaos. The daemon now opens
1873 stdin, stdout, and stderr to /dev/null when it puts itself into the
1876 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1877 log by default. The user can override this with suitable Perl magic.
1879 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1880 the log message when discard triggers, was not working for the DATA ACL or
1881 for the non-SMTP ACL.
1883 78. Error message wording change in sieve.c.
1885 79. If smtp_accept_max_per_host was set, the number of connections could be
1886 restricted to fewer than expected, because the daemon was trying to set up
1887 a new connection before checking whether the processes handling previous
1888 connections had finished. The check for completed processes is now done
1889 earlier. On busy systems, this bug wouldn't be noticed because something
1890 else would have woken the daemon, and it would have reaped the completed
1893 80. If a message was submitted locally by a user whose login name contained one
1894 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1895 It caused a spool format error. I have fixed the spool reading code. A
1896 related problem was that the "from" clause in the Received: line became
1897 illegal because of the space(s). It is now covered by ${quote_local_part.
1899 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1902 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1904 83. Patch from Sieve maintainer to fix unterminated string problem in
1905 "vacation" handling.
1907 84. Some minor changes to the Linux configuration files to help with other
1908 OS variants using glibc.
1910 85. One more patch for Sieve to update vacation handling to latest spec.
1913 ----------------------------------------------------
1914 See the note above about the 4.44 and 4.50 releases.
1915 ----------------------------------------------------
1921 1. Change 4.43/35 introduced a bug that caused file counts to be
1922 incorrectly computed when quota_filecount was set in an appendfile
1925 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1926 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1929 3. Give more explanation in the error message when the command for a transport
1930 filter fails to execute.
1932 4. There are several places where Exim runs a non-Exim command in a
1933 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1934 was being done only for the command run by the queryprogram router. It is
1935 now done for all such subprocesses. The other cases are: ${run, transport
1936 filters, and the commands run by the lmtp and pipe transports.
1938 5. Some older OS have a limit of 256 on the maximum number of file
1939 descriptors. Exim was using setrlimit() to set 1000 as a large value
1940 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1941 systems. I've change it so that if it can't get 1000, it tries for 256.
1943 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1944 was an oversight, and furthermore, ever since the addition of extra
1945 controls (e.g. 4.43/32), the checks on when to allow different forms of
1946 "control" were broken. There should now be diagnostics for all cases when a
1947 control that does not make sense is encountered.
1949 7. $recipients is now available in the predata ACL (oversight).
1951 8. Tidy the search cache before the fork to do a delivery from a message
1952 received from the command line. Otherwise the child will trigger a lookup
1953 failure and thereby defer the delivery if it tries to use (for example) a
1954 cached ldap connection that the parent has called unbind on.
1956 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1957 of $address_data from the recipient verification was clobbered by the
1958 sender verification.
1960 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1961 was its contents. (It was OK if the option was not defined at all.)
1963 11. A "Completed" log line is now written for messages that are removed from
1964 the spool by the -Mrm option.
1966 12. $host_address is now set to the target address during the checking of
1967 ignore_target_hosts.
1969 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1970 being passed; this would have caused $sender_host_name to have been used if
1971 matching the list had actually called for a host name (not very likely,
1972 since this list is usually IP addresses). A host name is now passed as
1975 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1976 code that specifies a non-restarting handler (typically sigaction() in
1977 modern systems) in an attempt to fix a rare and obscure crash bug.
1979 15. Narrowed the window for a race in the daemon that could cause it to ignore
1980 SIGCHLD signals. This is not a major problem, because they are used only to
1981 wake it up if nothing else does.
1983 16. A malformed maildirsize file could cause Exim to calculate negative values
1984 for the mailbox size or file count. Odd effects could occur as a result.
1985 The maildirsize information is now recalculated if the size or filecount
1988 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1989 support for a long time. Removed HAVE_SYS_VFS_H.
1991 18. Updated exipick to current release from John Jetmore.
1993 19. Allow an empty sender to be matched against a lookup in an address list.
1994 Previously the only cases considered were a regular expression, or an
1997 20. Exim went into a mad DNS lookup loop when doing a callout where the
1998 host was specified on the transport, if the DNS lookup yielded more than
2001 21. The RFC2047 encoding function was originally intended for short strings
2002 such as real names; it was not keeping to the 75-character limit for
2003 encoded words that the RFC imposes. It now respects the limit, and
2004 generates multiple encoded words if necessary. To be on the safe side, I
2005 have increased the buffer size for the ${rfc2047: expansion operator from
2008 22. Failure to deliver a bounce message always caused it to be frozen, even if
2009 there was an errors_to setting on the router. The errors_to setting is now
2012 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2013 canonical form (fully expanded) before being placed in
2014 $sender_host_address.
2016 24. Updated eximstats to version 1.33
2018 25. Include certificate and key file names in error message when GnuTLS fails
2019 to set them up, because the GnuTLS error message doesn't include the name
2020 of the failing file when there is a problem reading it.
2022 26. Expand error message when OpenSSL has problems setting up cert/key files.
2025 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2026 (this can affect the format of dates).
2028 28. exim_tidydb, when checking for the continued existence of a message for
2029 which it has found a message-specific retry record, was not finding
2030 messages that were in split spool directories. Consequently, it was
2031 deleting retry records that should have stayed in existence.
2033 29. eximstats updated to version 1.35
2034 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2035 1.35 - bugfix such that pie charts by volume are generated correctly
2037 30. The SPA authentication driver was not abandoning authentication and moving
2038 on to the next authenticator when an expansion was forced to fail,
2039 contradicting the general specification for all authenticators. Instead it
2040 was generating a temporary error. It now behaves as specified.
2042 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2043 (the order specifies the preference for clients). The order is now AES256,
2044 AES128, 3DES, ARCFOUR128.
2046 31. Small patch to Sieve code - explicitly set From: when generating an
2049 32. Exim crashed if a remote delivery caused a very long error message to be
2050 recorded - for instance if somebody sent an entire SpamAssassin report back
2051 as a large number of 550 error lines. This bug was coincidentally fixed by
2052 increasing the size of one of Exim's internal buffers (big_buffer) that
2053 happened as part of the Exiscan merge. However, to be on the safe side, I
2054 have made the code more robust (and fixed the comments that describe what
2057 33. Some experimental protocols are using DNS PTR records for new purposes. The
2058 keys for these records are domain names, not reversed IP addresses. The
2059 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2060 leaves it alone. Component reversal etc. now happens only for IP addresses.
2063 34. The host_aton() function is supposed to be passed a string that is known
2064 to be a valid IP address. However, in the case of IPv6 addresses, it was
2065 not checking this. This is a hostage to fortune. Exim now panics and dies
2066 if the condition is not met. A case was found where this could be provoked
2067 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2068 components; fortuitously, this particular loophole had already been fixed
2069 by change 4.50/55 or 4.44/33 above.
2071 If there are any other similar loopholes, the new check in host_aton()
2072 itself should stop them being exploited. The report I received stated that
2073 data on the command line could provoke the exploit when Exim was running as
2074 exim, but did not say which command line option was involved. All I could
2075 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2076 running as the user.
2079 35. There was a buffer overflow vulnerability in the SPA authentication code
2080 (which came originally from the Samba project). I have added a test to the
2081 spa_base64_to_bits() function which I hope fixes it.
2084 36. The daemon start-up calls getloadavg() while still root for those OS that
2085 need the first call to be done as root, but it missed one case: when
2086 deliver_queue_load_max is set with deliver_drop_privilege. This is
2087 necessary for the benefit of the queue runner, because there is no re-exec
2088 when deliver_drop_privilege is set.
2090 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2091 was in use, was not putting the data itself into the right store pool;
2092 consequently, it could be overwritten for a subsequent message in the same
2093 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2096 38. Sometimes the final signoff response after QUIT could fail to get
2097 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2098 < 0 before doing a fflush(). This bug looks as though it goes back to the
2099 introduction of TLS in release 3.20, but "sometimes" must have been rare
2100 because the tests only now provoked it.
2106 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2107 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2108 Now it returns an integer. A number of calls were still expecting a T/F
2109 return. Fortuitously, in all cases, the tests worked in OK situations,
2110 which is the norm. However, things would have gone wrong on any write
2111 failures on the smtp file descriptor. This function is used when sending
2112 messages over SMTP and also when doing verify callouts.
2114 2. When Exim is called to do synchronous delivery of a locally submitted
2115 message (the -odf or -odi options), it no longer closes stderr before doing
2118 3. Implemented the mua_wrapper option.
2120 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2122 5. Implemented the functions header_remove(), header_testname(),
2123 header_add_at_position(), and receive_remove_recipient(), and exported them
2126 6. If an ACL "warn" statement specified the addition of headers, Exim already
2127 inserted X-ACL-Warn: at the start if there was no header name. However, it
2128 was not making this test for the second and subsequent header lines if
2129 there were newlines in the string. This meant that an invalid header could
2130 be inserted if Exim was badly configured.
2132 7. Allow an ACL "warn" statement to add header lines at the start or after all
2133 the Received: headers, as well as at the end.
2135 8. Added the rcpt_4xx retry error code.
2137 9. Added postmaster_mailfrom=xxx to callout verification option.
2139 10. Added mailfrom=xxxx to the callout verification option, for verify=
2142 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2143 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2145 12. Inserted some casts to stop certain compilers warning when using pointer
2146 differences as field lengths or precisions in printf-type calls (mostly
2147 affecting debugging statements).
2149 13. Added optional readline() support for -be (dynamically loaded).
2151 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2152 same clock tick as a message's arrival, so that its received time was the
2153 same as the "first fail" time on the retry record, and that message
2154 remained on the queue past the ultimate address timeout, every queue runner
2155 would try a delivery (because it was past the ultimate address timeout) but
2156 after another failure, the ultimate address timeout, which should have then
2157 bounced the address, did not kick in. This was a "< instead of <=" error;
2158 in most cases the first failure would have been in the next clock tick
2159 after the received time, and all would be well.
2161 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2162 being recognized when the domain list was tested by the match_domain
2163 condition in an expansion string.
2165 16. Added the ${str2b64: operator.
2167 17. Exim was always calling setrlimit() to set a large limit for the number of
2168 processes, without checking whether the existing limit was already
2169 adequate. (It did check for the limit on file descriptors.) Furthermore,
2170 errors from getrlimit() and setrlimit() were being ignored. Now they are
2171 logged to the main and panic logs, but Exim does carry on, to try to do its
2172 job under whatever limits there are.
2174 18. Imported PCRE 5.0.
2176 19. Trivial typo in log message " temporarily refused connection" (the leading
2179 20. If the log selector return_path_on_delivery was set and an address was
2180 redirected to /dev/null, the delivery process crashed because it assumed
2181 that a return path would always be set for a "successful" delivery. In this
2182 case, the whole delivery is bypassed as an optimization, and therefore no
2185 21. Internal re-arrangement: the function for sending a challenge and reading
2186 a response while authentication was assuming a zero-terminated challenge
2187 string. It's now changed to take a pointer and a length, to allow for
2188 binary data in such strings.
2190 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2192 23. Exim was not respecting finduser_retries when seeking the login of the
2193 uid under which it was called; it was always trying 10 times. (The default
2194 setting of finduser_retries is zero.) Also, it was sleeping after the final
2195 failure, which is pointless.
2197 24. Implemented tls_on_connect_ports.
2199 25. Implemented acl_smtp_predata.
2201 26. If the domain in control=submission is set empty, Exim assumes that the
2202 authenticated id is a complete email address when it generates From: or
2203 Sender: header lines.
2205 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2206 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2207 chown and chgrp in /bin and hostname in /usr/bin.
2209 28. Exim was keeping the "process log" file open after each use, just as it
2210 does for the main log. This opens the possibility of it remaining open for
2211 long periods when the USR1 signal hits a daemon. Occasional processlog
2212 errors were reported, that could have been caused by this. Anyway, it seems
2213 much more sensible not to leave this file open at all, so that is what now
2216 29. The long-running daemon process does not normally write to the log once it
2217 has entered its main loop, and it closes the log before doing so. This is
2218 so that log files can straightforwardly be renamed and moved. However,
2219 there are a couple of unusual error situations where the daemon does write
2220 log entries, and I had neglected to close the log afterwards.
2222 30. The text of an SMTP error response that was received during a remote
2223 delivery was being truncated at 512 bytes. This is too short for some of
2224 the long messages that one sometimes sees. I've increased the limit to
2227 31. It is now possible to make retry rules that apply only when a message has a
2228 specific sender, in particular, an empty sender.
2230 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2231 it possible to be selective about when SMTP synchronization is enforced.
2233 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2235 32. Implemented hosts_connection_nolog.
2237 33. Added an ACL for QUIT.
2239 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2242 35. Added mailbox_size and mailbox_filecount to appendfile.
2244 36. Added control = no_multiline_responses to ACLs.
2246 37. There was a bug in the logic of the code that waits for the clock to tick
2247 in the case where the clock went backwards by a substantial amount such
2248 that the microsecond fraction of "now" was more than the microsecond
2249 fraction of "then" (but the whole seconds number was less).
2251 38. Added support for the libradius Radius client library this is found on
2252 FreeBSD (previously only the radiusclient library was supported).
2258 1. When certain lookups returned multiple values in the form name=value, the
2259 quoting of the values was not always being done properly. Specifically:
2260 (a) If the value started with a double quote, but contained no whitespace,
2262 (b) If the value contained whitespace other than a space character (i.e.
2263 tabs or newlines or carriage returns) it was not quoted.
2264 This fix has been applied to the mysql and pgsql lookups by writing a
2265 separate quoting function and calling it from the lookup code. The fix
2266 should probably also be applied to nisplus, ibase and oracle lookups, but
2267 since I cannot test any of those, I have not disturbed their existing code.
2269 2. A hit in the callout cache for a specific address caused a log line with no
2270 reason for rejecting RCPT. Now it says "Previous (cached) callout
2271 verification failure".
2273 3. There was an off-by-one bug in the queryprogram router. An over-long
2274 return line was truncated at 256 instead of 255 characters, thereby
2275 overflowing its buffer with the terminating zero. As well as fixing this, I
2276 have increased the buffer size to 1024 (and made a note to document this).
2278 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2279 when Exim is waiting for an SMTP response from a remote server, Exim
2280 restarts its select() call on the socket, thereby resetting its timeout.
2281 This is not a problem when such interrupts are rare. Somebody set up a cron
2282 job to run exiwhat every 2 minutes, which is less than the normal select()
2283 timeout (5 or 10 minutes). This meant that the select() timeout never
2284 kicked in because it was always reset. I have fixed this by comparing the
2285 time when an interrupt arrives with the time at the start of the first call
2286 to select(). If more time than the timeout has elapsed, the interrupt is
2287 treated as a timeout.
2289 5. Some internal re-factoring in preparation for the addition of Sieve
2290 extensions (by MH). In particular, the "personal" test is moved to a
2291 separate function, and given an option for scanning Cc: and Bcc: (which is
2292 not set for Exim filters).
2294 6. When Exim created an email address using the login of the caller as the
2295 local part (e.g. when creating a From: or Sender: header line), it was not
2296 quoting the local part when it contained special characters such as @.
2298 7. Installed new OpenBSD configuration files.
2300 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2301 try to make them clearer.
2303 9. Callout options, other than the timeout value, were being ignored when
2304 verifying sender addresses in header lines. For example, when using
2306 verify = header_sender/callout=no_cache
2308 the cache was (incorrectly) being used.
2310 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2311 only the Cygwin environment.
2313 11. When return_path_on_delivery was set as a log selector, if different remote
2314 addresses in the same message used different return paths and parallel
2315 remote delivery occurred, the wrong values would sometimes be logged.
2316 (Whenever a remote delivery process finished, the return path value from
2317 the most recently started remote delivery process was logged.)
2319 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2320 lines when AUTH and/or TLS are in use. This is the "received protocol"
2321 field. Exim used to use "asmtp" for authenticated SMTP, without any
2322 indication (in the protocol name) for TLS use. Now it follows the RFC and
2323 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2324 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2325 names appear in log lines as well as in Received: header lines.
2327 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2328 extensions, and comparison tests, and to fix some bugs.
2330 14. Changes to the "personal" filter test:
2332 (1) The test was buggy in that it was just doing the equivalent of
2333 "contains" tests on header lines. For example, if a user's address was
2334 anne@some.where, the "personal" test would incorrectly be true for
2336 To: susanne@some.where
2338 This test is now done by extracting each address from the header in turn,
2339 and checking the entire address. Other tests that are part of "personal"
2340 are now done using regular expressions (for example, to check local parts
2341 of addresses in From: header lines).
2343 (2) The list of non-personal local parts in From: addresses has been
2344 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2345 taken from the Sieve specification recommendations.
2347 (3) If the message contains any header line starting with "List-" it is
2348 treated as non-personal.
2350 (4) The test for "circular" in the Subject: header line has been removed
2351 because it now seems ill-conceived.
2353 15. Minor typos in src/EDITME comments corrected.
2355 16. Installed latest exipick from John Jetmore.
2357 17. If headers_add on a router specified a text string that was too long for
2358 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2359 of string_sprintf() is now avoided.
2361 18. $message_body_size was not set (it was always zero) when running the DATA
2362 ACL and the local_scan() function.
2364 19. For the "mail" command in an Exim filter, no default was being set for
2365 the once_repeat time, causing a random time value to be used if "once" was
2366 specified. (If the value happened to be <= 0, no repeat happened.) The
2367 default is now 0s, meaning "never repeat". The "vacation" command was OK
2368 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2369 (I found it when inspecting the code).
2371 20. There is now an overall timeout for performing a callout verification. It
2372 defaults to 4 times the callout timeout, which applies to individual SMTP
2373 commands during the callout. The overall timeout applies when there is more
2374 than one host that can be tried. The timeout is checked before trying the
2375 next host. This prevents very long delays if there are a large number of
2376 hosts and all are timing out (e.g. when the network connections are timing
2377 out). The value of the overall timeout can be changed by specifying an
2378 additional sub-option for "callout", called "maxwait". For example:
2380 verify = sender/callout=5s,maxwait=20s
2382 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2383 to the end before writing, but this should make it even safer).
2385 22. Exim was forgetting that it had advertised PIPELINING for the second and
2386 subsequent messages on an SMTP connection. It was also not resetting its
2387 memory on STARTTLS and an internal HELO.
2389 23. When Exim logs an SMTP synchronization error within a session, it now
2390 records whether PIPELINING has been advertised or not.
2392 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2393 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2394 rather than long int.
2396 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2398 26. Added the never_mail option to autoreply.
2404 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2405 crash if the getsockname() call failed; this can happen if a connection is
2406 closed very soon after it is established. The problem was simply in the
2407 order in which certain operations were done, causing Exim to try to write
2408 to the SMTP stream before it had set up the file descriptor. The bug has
2409 been fixed by making things happen in the correct order.
2415 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2416 before the connection was closed, thus losing the rejection response.
2418 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2419 some early Solaris releases, but causes trouble in current releases where
2420 socklen_t is defined.
2422 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2425 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2428 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2429 that releasing the top of it at the end releases what was used for sub-
2430 expansions (unless the block got too big). However, discard this block if
2431 the first thing is a variable or header, so that we can use its block when
2432 it is dynamic (useful for very large $message_headers, for example).
2434 6. Lookups now cache *every* query, not just the most recent. A new, separate
2435 store pool is used for this. It can be recovered when all lookup caches are
2436 flushed. Lookups now release memory at the end of their result strings.
2437 This has involved some general refactoring of the lookup sources.
2439 7. Some code has been added to the store_xxx() functions to reduce the amount
2440 of flapping under certain conditions.
2442 8. log_incoming_interface used to affect only the <= reception log lines. Now
2443 it causes the local interface and port to be added to several more SMTP log
2444 lines, for example "SMTP connection from", and rejection lines.
2446 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2448 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2450 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2451 could be overwritten at the end of the current message (or the start of a
2452 new message if it was set in a HELO ACL). The value is now preserved for
2453 the duration of the SMTP connection.
2455 12. If a transport had a headers_rewrite setting, and a matching header line
2456 contained an unqualified address, that address was qualified, even if it
2457 did not match any rewriting rules. The underlying bug was that the values
2458 of the flags that permit the existence of unqualified sender and recipient
2459 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2460 non-local messages, and by -bnq for local messages) were not being
2461 preserved with the message after it was received.
2463 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2464 "next input=" as part of the text comprising the host identity instead of
2465 the correct text. The code was using the same buffer for two different
2466 strings. However, depending on which order the printing function evaluated
2467 its arguments, the bug did not always show up. Under Linux, for example, my
2468 test suite worked just fine.
2470 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2471 doesn't work with some older versions of Perl. It has been changed to "my",
2472 which in any case is probably the better facility to use.
2474 15. A really picky compiler found some instances of statements for creating
2475 error messages that either had too many or two few arguments for the format
2478 16. The size of the buffer for calls to the DNS resolver has been increased
2479 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2480 for addresses that have a lot of PTR records. This alleviates a problem; it
2481 does not fully solve it.
2483 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2484 buffer now truncates the list and logs the incident, which is the same
2485 action as happens when Exim is looking up a host name and its aliases.
2486 Previously in this situation something unpredictable would happen;
2487 sometimes it was "internal error: store_reset failed".
2489 18. If a server dropped the connection unexpectedly when an Exim client was
2490 using GnuTLS and trying to read a response, the client delivery process
2491 crashed while trying to generate an error log message.
2493 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2494 string, for example:
2496 warn message = H1: something\nH2: something
2498 the text was added as a single header line from Exim's point of view
2499 though it ended up OK in the delivered message. However, searching for the
2500 second and subsequent header lines using $h_h2: did not work. This has been
2501 fixed. Similarly, if a system filter added multiple headers in this way,
2502 the routers could not see them.
2504 20. Expanded the error message when iplsearch is called with an invalid key to
2505 suggest using net-iplsearch in a host list.
2507 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2508 ACLs are no longer actually imposed (and a message to that effect is
2511 22. If a "gecos" field in a passwd entry contained escaped characters, in
2512 particular, if it contained a \" sequence, Exim got it wrong when building
2513 a From: or a Sender: header from that name. A second bug also caused
2514 incorrect handling when an unquoted " was present following a character
2515 that needed quoting.
2517 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2518 was not being matched caselessly.
2520 24. Arranged for all hyphens in the exim.8 source to be escaped with
2523 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2524 properly. Recipient callout cache records were still being keyed to include
2525 the sender, even when use_sender was set false. This led to far more
2526 callouts that were necessary. The sender is no longer included in the key
2527 when use_sender is false.
2529 26. Added "control = submission" modifier to ACLs.
2531 27. Added the ${base62d: operator to decode base 62 numbers.
2533 28. dnsdb lookups can now access SRV records.
2535 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2536 the configuration file.
2538 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2539 (-v) mode. This makes the output for a verbose queue run more intelligible.
2541 31. Added a use_postmaster feature to recipient callouts.
2543 32. Added the $body_zerocount variable, containing the number of binary zero
2544 bytes in the message body.
2546 33. The time of last modification of the "new" subdirectory is now used as the
2547 "mailbox time last read" when there is a quota error for a maildir
2550 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2552 35. Added +ignore_unknown as a special item in host lists.
2554 36. Code for decoding IPv6 addresses in host lists is now included, even if
2555 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2556 address was recognized as an IP address, but was then not correctly decoded
2557 into binary, causing unexpected and incorrect effects when compared with
2564 1. Very minor rewording of debugging text in manualroute to say "list of
2565 hosts" instead of "hostlist".
2567 2. If verify=header_syntax was set, and a header line with an unqualified
2568 address (no domain) and a large number of spaces between the end of the
2569 name and the colon was received, the reception process suffered a buffer
2570 overflow, and (when I tested it) crashed. This was caused by some obsolete
2571 code that should have been removed. The fix is to remove it!
2573 3. When running in the test harness, delay a bit after writing a bounce
2574 message to get a bit more predictability in the log output.
2576 4. Added a call to search_tidyup() just before forking a reception process. In
2577 theory, someone could use a lookup in the expansion of smtp_accept_max_
2578 per_host which, without the tidyup, could leave open a database connection.
2580 5. Added the variables $recipient_data and $sender_data which get set from a
2581 lookup success in an ACL "recipients" or "senders" condition, or a router
2582 "senders" option, similar to $domain_data and $local_part_data.
2584 6. Moved the writing of debug_print from before to after the "senders" test
2587 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2588 problems for message scanning, either using a data ACL, or using
2589 local_scan() because the Received: header was not generated till after they
2590 were called (in order to set the time as the time of reception completion).
2591 I have revised the way this works. The header is now generated after the
2592 body is received, but before the ACL or local_scan() are called. After they
2593 are run, the timestamp in the header is updated.
2599 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2600 before starting a queue runner without re-exec. This happened only when
2601 deliver_drop_privilege was set or when the Exim user was set to root. The
2602 effect of the bug was that timeouts during subsequent deliveries caused
2603 crashes instead of being properly handled. The handler is now left at its
2604 default (and expected) setting.
2606 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2607 message, again when deliver_drop_privilege is set or Exim is run as root.
2608 The bug described in (1) was not present in this case, but the tidying up
2609 of the other signals was missing. I have made the two cases consistent.
2611 3. The ignore_target_hosts setting on a manualroute router was being ignored
2612 for hosts that were looked up using the /MX notation.
2614 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2617 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2618 operated on the sender address. After changing the $sender_address to <>
2619 for the sender address verify, Exim was re-instated it as the original
2620 (before rewriting) address, but remembering that it had rewritten it, so it
2621 wasn't rewriting it again. This bug also had the effect of breaking the
2622 sender address verification caching when the sender address was rewritten.
2624 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2625 This has been changed so that if the ip literal address matches
2626 ignore_target_hosts, the router declines.
2628 7. Added expansion conditions match_domain, match_address, and match_local_
2629 part (NOT match_host).
2631 8. The placeholder for the Received: header didn't have a length field set.
2633 9. Added code to Exim itself and to exim_lock to test for a specific race
2634 condition that could lead to file corruption when using MBX delivery. The
2635 issue is with the lockfile that is created in /tmp. If this file is removed
2636 after a process has opened it but before that process has acquired a lock,
2637 there is the potential for a second process to recreate the file and also
2638 acquire a lock. This could lead to two Exim processes writing to the file
2639 at the same time. The added code performs the same test as UW imapd; it
2640 checks after acquiring the lock that its file descriptor still refers to
2641 the same named file.
2643 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2644 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2647 11. Added the smtp_active_hostname option. If used, this will typically be made
2648 to depend on the incoming interface address. Because $interface_address is
2649 not set up until the daemon has forked a reception process, error responses
2650 that can happen earlier (such as "too many connections") no longer contain
2653 12. If an expansion in a condition on a "warn" statement fails because a lookup
2654 defers, the "warn" statement is abandoned, and the next ACL statement is
2655 processed. Previously this caused the whole ACL to be aborted.
2657 13. Added the iplsearch lookup type.
2659 14. Added ident_timeout as a log selector.
2661 15. Added tls_certificate_verified as a log selector.
2663 16. Added a global option tls_require_ciphers (compare the smtp transport
2664 option of the same name). This controls incoming TLS connections.
2666 17. I finally figured out how to make tls_require_ciphers do a similar thing
2667 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2668 before starting the TLS session.
2670 18. Tabs are now shown as \t in -bP output.
2672 19. If the log selector return_path_on_delivery was set, Exim crashed when
2673 bouncing a message because it had too many Received: header lines.
2675 20. If two routers both had headers_remove settings, and the first one included
2676 a superfluous trailing colon, the final name in the first list and the
2677 first name in the second list were incorrectly joined into one item (with a
2678 colon in the middle).
2684 1. Added -C and -D options to the exinext utility, mainly to make it easier
2685 to include in the automated testing, but these could be helpful when
2686 multiple configurations are in use.
2688 2. The exinext utility was not formatting the output nicely when there was
2689 an alternate port involved in the retry record key, nor when there was a
2690 message id as well (for retries that were specific to a specific message
2691 and a specific host). It was also confused by IPv6 addresses, because of
2692 the additional colons they contain. I have fixed the IPv4 problem, and
2693 patched it up to do a reasonable job for IPv6.
2695 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2696 delivery, the log line now contains "pipelined" if PIPELINING was used.
2698 4. An SMTP transport process used to panic and die if the bind() call to set
2699 an explicit outgoing interface failed. This has been changed; it is now
2700 treated in the same way as a connect() failure.
2702 5. A reference to $sender_host_name in the part of a conditional expansion
2703 that was being skipped was still causing a DNS lookup. This no longer
2706 6. The def: expansion condition was not recognizing references to header lines
2707 that used bh_ and bheader_.
2709 7. Added the _cache feature to named lists.
2711 8. The code for checking quota_filecount in the appendfile transport was
2712 allowing one more file than it should have been.
2714 9. For compatibility with Sendmail, the command line option
2722 and sets the incoming protocol and host name (for trusted callers). The
2723 host name and its colon can be omitted when only the protocol is to be set.
2724 Note the Exim already has two private options, -pd and -ps, that refer to
2725 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2726 "s", but I don't think that's a major issue.
2728 10. A number of refactoring changes to the code, none of which should affect
2731 (a) The number of logging options was getting close to filling up the
2732 32-bit word that was used as a bit map. I have split them into two classes:
2733 those that are passed in the argument to log_write(), and those that are
2734 only ever tested independently outside of that function. These are now in
2735 separate 32-bit words, so there is plenty of room for expansion again.
2736 There is no change in the user interface or the logging behaviour.
2738 (b) When building, for example, log lines, the code previously used a
2739 macro that called string_cat() twice, in order to add two strings. This is
2740 not really sufficiently general. Furthermore, there was one instance where
2741 it was actually wrong because one of the argument was used twice, and in
2742 one call a function was used. (As it happened, calling the function twice
2743 did not affect the overall behaviour.) The macro has been replaced by a
2744 function that can join an arbitrary number of extra strings onto a growing
2747 (c) The code for expansion conditions now uses a table and a binary chop
2748 instead of a serial search (which was left over from when there were very
2749 few conditions). Also, it now recognizes conditions like "pam" even when
2750 the relevant support is not compiled in: a suitably worded error message is
2751 given if an attempt is made to use such a condition.
2753 11. Added ${time_interval:xxxxx}.
2755 12. A bug was causing one of the ddress fields not to be passed back correctly
2756 from remote delivery subprocesses. The field in question was not being
2757 subsequently used, so this caused to problems in practice.
2759 13. Added new log selectors queue_time and deliver_time.
2761 14. Might have fixed a bug in maildirsizefile handling that threw up
2762 "unexpected character" debug warnings, and recalculated the data
2763 unnecessarily. In any case, I expanded the warning message to give more
2766 15. Added the message "Restricted characters in address" to the statements in
2767 the default ACL that block characters like @ and % in local parts.
2769 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2770 Three changes have been made:
2772 (a) There was a serious bug; a negative response to MAIL caused the whole
2773 recipient domain to be cached as invalid, thereby blocking all messages
2774 to all local parts at the same domain, from all senders. This bug has
2775 been fixed. The domain is no longer cached after a negative response to
2776 MAIL if the sender used is not empty.
2778 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2781 (c) A new callout option, "use_sender" has been added for people who want
2782 the modified behaviour.
2788 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2791 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2792 indeed breaks things for older releases.
2794 3. Added additional logging to the case where there is a problem reading data
2795 from a filter that is running in a subprocess using a pipe, in order to
2796 try to track down a specific problem.
2798 4. Testing facility fudge: when running in the test harness and attempting
2799 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2800 getting "No route to host". Convert this to a timeout.
2802 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2805 6. Some OS don't have socklen_t but use size_t instead. This affects the
2806 fifth argument of getsockopt() amongst other things. This is now
2807 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2808 can be set for individual OS. I have set it for SunOS5, OSF1, and
2809 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2810 some earlier ones do not.
2812 7. Change 4.30/15 was not doing the test caselessly.
2814 8. The standard form for an IPv6 address literal was being rejected by address
2815 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2816 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2817 this, as well as the form without the "IPv6" on the front (but only when
2818 address literals are enabled, of course).
2820 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2822 10. Exim crashed if a message with an empty sender address specified by -f
2823 encountered a router with an errors_to setting. This could be provoked only
2824 by a command such as
2828 where an empty string was supplied; "<>" did not hit this bug.
2830 11. Installed PCRE release 4.5.
2832 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2833 remained set. It is now erased.
2835 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2836 times from message ids (which are base 36 rather than the normal 62).
2838 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2839 were being counted as actual protocol errors, and logged if the log
2840 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2841 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2842 and DATA following a set of rejected RCPTs do not count as protocol errors.
2843 In other words, Exim assumes they were pipelined, though this may not
2844 actually be the case. Of course, in all cases the client gets an
2845 appropriate error code.
2847 15. If a lookup fails in an ACL condition, a message about the failure may
2848 be available; it is used if testing the ACL cannot continue, because most
2849 such messages specify what the cause of the deferral is. However, some
2850 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2851 that caused an old message to be retained and used if a later statement
2852 caused a defer, replacing the real cause of the deferral.
2854 16. If an IP address had so many PTR records that the DNS lookup buffer
2855 was not large enough to hold them, Exim could crash while trying to process
2856 the truncated data. It now detects and logs this case.
2858 17. Further to 4.21/58, another change has been made: if (and only if) the
2859 first line of a message (the first header line) ends with CRLF, a bare LF
2860 in a subsequent header line has a space inserted after it, so as not to
2861 terminate the header.
2863 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2864 unnecessarily, used atoi() instead of strtol(), and didn't check the
2865 termination when getting file sizes from file names by regex.
2867 19. Completely re-implemented the support for maildirsize files, in the light
2868 of a number of problems with the previous contributed implementation
2869 (4.30/29). In particular:
2871 . If the quota is zero, the maildirsize file is maintained, but no quota is
2874 . If the maildir directory does not exist, it is created before any attempt
2875 to write a maildirsize file.
2877 . The quota value in the file is just a cache; if the quota is changed in
2878 the transport, the new value overrides.
2880 . A regular expression is available for excluding directories from the
2883 20. The autoreply transport checks the characters in options that define the
2884 message's headers; it allows continued headers, but it was checking with
2885 isspace() after an embedded newline instead of explicitly looking for a
2888 21. If all the "regular" hosts to which an address was routed had passed their
2889 expiry times, and had not reached their retry times, the address was
2890 bounced, even if fallback hosts were defined. Now Exim should go on to try
2893 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2894 equivalent code in the SMTP transport. Some hosts send humungous responses
2895 to HELO/EHLO, more than 1024 it seems.
2897 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2898 gives compiler warnings in some environments. I've now done it "properly",
2901 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2902 (because of problems with the built-in one) was declared to return uschar *
2903 instead of char *, causing compiler failure.
2905 25. Fixed a file descriptor leak when processing alias/forward files.
2907 26. Fixed a minor format string issue in dbfn.c.
2909 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2911 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2912 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2915 29. When a sender address is verified, it is cached, to save repeating the test
2916 when there is more than one recipient in a message. However, when the
2917 verification involves a callout, it is possible for different callout
2918 options to be set for different recipients. It is too complicated to keep
2919 track of this in the cache, so now Exim always runs a verification when a
2920 callout is required, relying on the callout cache for the optimization.
2921 The overhead is duplication of the address routing, but this should not be
2924 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2925 to be verified with callout=postmaster, and the main callout worked but the
2926 postmaster check failed, the verification correctly failed. However, if a
2927 subsequent RCPT command asked for sender verification *without* the
2928 postmaster check, incorrect caching caused this verification also to fail,
2931 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2932 it was not caching the DNS options (qualify_single, search_parents) that
2933 were used when the lookup failed. A subsequent lookup with different
2934 options therefore always gave the same answer, though there were cases
2935 where it should not have. (Example: a "domains = !$mx_any" option on a
2936 dnslookup router: the "domains" option is always processed without any
2937 widening, but the router might have qualify_single set.) Now Exim uses the
2938 cached value only when the same options are set.
2940 32. Added John Jetmore's "exipick" utility to the distribution.
2942 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2943 than a timeout (e.g. a certificate is required, and is not provided), an
2944 Exim server now closes the connection immediately. Previously it waited for
2945 the client to close - but if the client is SSL, it seems that they each
2946 wait for each other, leading to a delay before one of them times out.
2948 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2949 maintained 0.8.x compatibility because I don't think many are using it, and
2950 it is clearly obsolete.
2952 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2955 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2956 client certificate was expired. A simple patch fixes this, though I don't
2957 understand the full logic of why the verify callback is called multiple
2960 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2961 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2962 which causes problems with some clients (such as the Certicom SSL Plus
2963 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2964 disables the coutermeasure allowing Eudora to connect."
2966 38. Exim was not checking that a write() to a log file succeeded. This could
2967 lead to Bad Things if a log got too big, in particular if it hit a file
2968 size limit. Exim now panics and dies if it cannot write to a log file, just
2969 as it does if it cannot open a log file.
2971 39. Modified OS/Makefile-Linux so that it now contains
2973 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2975 The two -D definitions ensure that Exim is compiled with large file
2976 support, which makes it possible to handle log files that are bigger than
2979 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2980 instance) a domain was checked against a named list that involved a lookup,
2981 causing $domain_data to be set, then another domain was checked against the
2982 same list, then the first domain was re-checked, the value of $domain_data
2983 after the final check could be wrong. In particular, if the second check
2984 failed, it could be set empty. This bug probably also applied to
2987 41. The strip_trailing_dot option was not being applied to the address given
2988 with the -f command-line option.
2990 42. The code for reading a message's header from the spool was incrementing
2991 $received_count, but never initializing it. This meant that the value was
2992 incorrect (doubled) while delivering a message in the same process in which
2993 it was received. In the most common configuration of Exim, this never
2994 happens - a fresh exec is done - but it can happen when
2995 deliver_drop_privilege is set.
2997 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2998 it now includes up to 150 characters of the unexpected data in the log
3001 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3002 and building data strings. The size of both of these buffers was 10 000
3003 bytes - far larger than anybody would *ever* want, thought I. Needless to
3004 say, somebody hit the limit. I have increased the maximum line length to
3005 20 000 and the maximum data length of concatenated lines to 100 000. I have
3006 also fixed two bugs, because there was no checking on these buffers. Tsk,
3007 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3008 buffer is too small.
3010 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3011 lsearch lookups. Now it does.
3013 46. When parsing a route_list item in a manualroute router, a fixed-length
3014 buffer was used for the list of hosts. I made this 1024 bytes long,
3015 thinking that nobody would ever have a list of hosts that long. Wrong.
3016 Somebody had a whole pile of complicated expansion conditions, and the
3017 string was silently truncated, leading to an expansion error. It turns out
3018 that it is easier to change to an unlimited length (owing to other changes
3019 that have happened since this code was originally written) than to build
3020 structure for giving a limitation error. The length of the item that
3021 expands into the list of hosts is now unlimited.
3023 47. The lsearch lookup could not handle data where the length of text line was
3024 more than 4095 characters. Such lines were truncated, leading to shortened
3025 data being returned. It should now handle lines of any length.
3027 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3028 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3031 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3032 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3033 no longer happens in comment lines. A list of replacements is now placed
3034 at the head of all of the source files, except those whose only change is
3035 to replace PERL_COMMAND in the very first #! line.
3037 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3038 messages on the queue, with a bottom-up merge sort, using code contributed
3039 by Michael Haardt. This should make operations like -bp somewhat faster on
3040 large queues. It won't affect queue runners, except when queue_run_in_order
3043 51. Installed eximstats 1.31 in the distribution.
3045 52. Added support for SRV lookups to the dnslookup router.
3047 53. If an ACL referred to $message_body or $message_body_end, the value was not
3048 reset for any messages that followed in the same SMTP session.
3050 54. The store-handling optimization for building very long strings was not
3051 differentiating between the different store pools. I don't think this
3052 actually made any difference in practice, but I've tidied it.
3054 55. While running the routers to verify a sender address, $sender_address
3055 was still set to the sender address. This is wrong, because when routing to
3056 send a bounce to the sender, it would be empty. Therefore, I have changed
3057 it so that, while verifying a sender address, $sender_address is set to <>.
3058 (There is no change to what happens when verifying a recipient address.)
3060 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3061 target A or AAAA records (if not already returned) without resetting the
3062 qualify_single or search_parents options of the DNS resolver. These are
3063 inappropriate in this case because the targets of MX and SRV records must
3064 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3065 target that, when qualified, matched something in the local domain. These
3066 two options are now turned off when doing these lookups.
3068 57. It seems that at least some releases of Reiserfs (which does not have the
3069 concept of a fixed number of inodes) returns zero and not -1 for the
3070 number of available inodes. This interacted badly with check_spool_inodes,
3071 which assumed that -1 was the "no such thing" setting. What I have done is
3072 to check that the total number of inodes is greater than zero before doing
3073 the test of how many are available.
3075 58. When a "warn" ACL statement has a log_message modifier, the message is
3076 remembered, and not repeated. This is to avoid a lot of repetition when a
3077 message has many recipients that cause the same warning to be written.
3078 Howewer, Exim was preserving the list of already written lines for an
3079 entire SMTP session, which doesn't seem right. The memory is now reset if a
3080 new message is started.
3082 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3083 debugging output unless log_rewrite was also set.
3085 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3086 of (int)(handle) when we know that handle contains (void *)(-1).
3088 61. The Exim daemon panic-logs an error return when it closes the incoming
3089 connection. However "connection reset by peer" seems to be common, and
3090 isn't really an error worthy of noting specially, so that particular error
3093 62. When Exim is trying to find all the local interfaces, it used to panic and
3094 die if the ioctl to get the interface flags failed. However, it seems that
3095 on at least one OS (Solaris 9) it is possible to have an interface that is
3096 included in the list of interfaces, but for which you get a failure error
3097 for this call. This happens when the interface is not "plumbed" into a
3098 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3099 failure of the "get flags" call assumes that the interface is down.
3101 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3102 makes life easier for people who are doing arithmetic on fields extracted
3103 from dates, where you often get leading zeros that should not be
3104 interpreted as octal.
3106 64. Added qualify_domain to the redirect router, to override the global
3109 65. If a pathologically long header line contained very many addresses (the
3110 report of this problem mentioned 10 000) and each of them was rewritten,
3111 Exim could use up a very large amount of memory. (It kept on making new
3112 copies of the header line as it rewrote, and never released the old ones.)
3113 At the expense of a bit more processing, the header rewriting function has
3114 been changed so that it no longer eats memory in this way.
3116 66. The generation of the Received: header has been moved from the time that a
3117 message starts to be received, to the time that it finishes. The timestamp
3118 in the Received: header should now be very close to that of the <= log
3119 line. There are two side-effects of this change:
3121 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3122 logged header lines no longer include the local Received: line, because
3123 it has not yet been created. The same applies to a copy of the message
3124 that is returned to a non-SMTP sender when a message is rejected.
3126 (b) When a filter file is tested using -bf, no additional Received: header
3127 is added to the test message. After some thought, I decided that this
3130 This change does not affect the value of $received_for. It is still set
3131 after address rewriting, but before local_scan() is called.
3133 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3135 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3136 gave an unhelpful panic error message, and a defer error. I have managed to
3137 change this behaviour so that it now rejects any supplied certificate,
3138 which seems right, as the list of acceptable certificates is empty.
3140 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3141 gave an unhelpful defer error. I have not managed to make this reject any
3142 supplied certificates, but the error message it gives is "no certificate
3143 supplied", which is not helpful.
3145 70. exigrep's output now also includes lines that are not associated with any
3146 message, but which match the given pattern. Implemented by a patch from
3147 Martin Sluka, which also tidied up the Perl a bit.
3149 71. Recipient callout verification, like sender verification, was using <> in
3150 the MAIL FROM command. This isn't really the right thing, since the actual
3151 sender may affect whether the remote host accepts the recipient or not. I
3152 have changed it to use the actual sender in the callout; this means that
3153 the cache record is now keyed on a recipient/sender pair, not just the
3154 recipient address. There doesn't seem to be a real danger of callout loops,
3155 since a callout by the remote host to check the sender would use <>.
3156 [SEE ABOVE: changed after hitting problems.]
3158 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3159 temporary errors. However, in the case of such a code being given after
3160 the end of a data transmission (i.e. after ".") Exim was failing to write
3161 a retry record for the message. (Yes, there was some broken host that was
3162 actually sending 8xx at this point.)
3164 73. An unknown lookup type in a host list could cause Exim to panic-die when
3165 the list was checked. (An example that provoked this was putting <; in the
3166 middle of a list instead of at the start.) If this happened during a DATA
3167 ACL check, a -D file could be left lying around. This kind of configuration
3168 error no longer causes Exim to die; instead it causes a defer errror. The
3169 incident is still logged to the main and panic logs.
3171 74. Buglet left over from Exim 3 conversion. The message "too many messages
3172 in one connection" was written to the rejectlog but not the mainlog, except
3173 when address rewriting (yes!) was being logged.
3175 75. Added write_rejectlog option.
3177 76. When a system filter was run not as root (that is, when system_filter_user
3178 was set), the values of the $n variables were not being returned to the
3179 main process; thus, they were not subsequently available in the $sn
3182 77. Added +return_path_on_delivery log selector.
3184 78. A connection timeout was being treated differently from recipients deferred
3185 when testing hosts_max_try with a message that was older than the host's
3186 retry timeout. (The host should not be counted, thus allowing all hosts to
3187 be tried at least once before bouncing.) This may have been the cause of an
3188 occasionally reported bug whereby a message would remain on the queue
3189 longer than the retry timeout, but would be bounced if a delivery was
3190 forced. I say "may" because I never totally pinned down the problem;
3191 setting up timeout/retry tests is difficult. See also the next item.
3193 79. The ultimate address timeout was not being applied to errors that involved
3194 a combination of host plus message (for example, a timeout on a MAIL
3195 command). When an address resolved to a number of possible hosts, and they
3196 were not all tried for each delivery (e.g. because of hosts_max_try), a
3197 message could remain on the queue longer than the retry timeout.
3199 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3202 81. Fixed an obscure SMTP outgoing bug which required at least the following
3203 conditions: (a) there was another message waiting for the same server;
3204 (b) the server returned 5xx to all RCPT commands in the first message so
3205 that the message was not completed; (c) the server dropped the connection
3206 or gave a negative response to the RSET that Exim sends to abort the
3207 transaction. The observed case was a dropped connection after DATA that had
3208 been sent in pipelining mode. That is, the server had advertised PIPELINING
3209 but was not implementing it correctly. The effect of the bug was incorrect
3210 behaviour, such as trying another host, and this could lead to a crash.
3216 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3217 and daemon.c were passed as pointers to ints; they should have been
3218 pointers to socklen_t variables (which are typically unsigned ints).
3220 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3223 3. Fixed a really odd bug that affected only the testing scheme; patching a
3224 certain fixed string in the binary changed the value of another string that
3225 happened to be identical to the end of the original first string.
3227 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3228 name", it returns that address as the IP address. On some operating
3229 systems (e.g. Solaris), it also passes back the IP address string as the
3230 "host name". However, on others (e.g. Linux), it passes back an empty
3231 string. Exim wasn't checking for this, and was changing the host name to an
3232 empty string, assuming it had been canonicized.
3234 5. Although rare, it is permitted to have more than one PTR record for a given
3235 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3236 all the names associated with an address, because they do in Solaris.
3237 However, it seems that they do not in Linux for data that comes from the
3238 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3239 I found this out when I moved to a new Linux workstation and tried to run
3240 the Exim test suite.
3242 To get round this problem I have changed the code so that it now does its
3243 own call to the DNS to look up PTR records when searching for a host name.
3244 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3245 addresses that are only in /etc/hosts are still found.
3247 This behaviour is, however, controlled by an option called host_lookup_
3248 order, which defaults to "bydns:byaddr". If people want to use the other
3249 order, or indeed, just use one or the other means of lookup, they can
3250 specify it in this variable.
3252 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3253 some operating systems, this comes back from gethostbyaddr() as an empty
3254 string, and this is what Exim used to test for. However, it seems that in
3255 other systems, "." is yielded. Exim now tests for this case too.
3257 7. The values of check_spool_space and check_log_space are now held internally
3258 as a number of kilobytes instead of an absolute number of bytes. If a
3259 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3260 kilobyte. This means that much larger values can be stored.
3262 8. Exim monitor: an attempt to get the action menu when not actually pointing
3263 at a message produces an empty menu entitled "No message selected". This
3264 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3265 no entries in it ("Shell widget menu has zero width and/or height"). So I
3266 have added a single, blank menu entry in this case.
3268 9. Added ${quote_local_part.
3270 10. MIME decoding is now applied to the contents of Subject: header lines when
3273 11. Now that a reference to $sender_host_address automatically causes a reverse
3274 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3275 host lookup before query-style lookups in lists that might use this
3276 variable. This has therefore been abolished, and the "net-" prefix is no
3277 longer necessary for query-style lookups.
3279 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3280 have been a typo for LFLAGS, so it has been changed.
3282 13. The install script calls Exim with "-C /dev/null" in order to find the
3283 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3284 to be output. Howeve, since Exim outputs its version number before the
3285 error, it didn't break the script. It just looked ugly. I fixed this by
3286 always allowing "-C /dev/null" if the caller is root.
3288 14. Ignore overlarge ACL variable number when reading spool file - insurance
3289 against a later release with more variables having written the file.
3291 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3292 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3293 this, as well as the form without the "IPv6" on the front.
3295 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3296 OS/Makefile-Darwin file.
3298 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3299 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3301 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3302 variable. However, some OS use ints for inodes. I've added cast to long int
3303 to get rid of the compiler warning.
3305 19. I had forgotten to lock out "/../" in configuration file names when
3306 ALT_CONFIG_PREFIX was set.
3308 20. Routers used for verification do not need to specify transports. However,
3309 if such a router generated a host list, and callout was configured, Exim
3310 crashed, because it could not find a port number from the (non-existent)
3311 transport. It now assumes port 25 in this circumstance.
3313 21. Added the -t option to exigrep.
3315 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3316 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3317 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3318 them is set, LOOKUP_LSEARCH is forced.
3320 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3322 24. Added sender and host information to the "rejected by local_scan()" log
3323 line; previously there was no indication of these.
3325 25. Added .include_if_exists.
3327 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3328 Linux. It turns out that not all file systems support this. Apparently some
3329 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3330 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3331 not supported on the file descriptor, is now ignored when Exim is trying to
3332 sync a directory. This applies only to Linux.
3334 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3336 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3337 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3340 29. Added support for maildirsize files from supplied patch (modified a bit).
3342 30. The use of :fail: followed by an empty string could lead Exim to respond to
3343 sender verification failures with (e.g.):
3345 550 Verification failed for <xxx>
3346 550 Sender verify failed
3348 where the first response line was missing the '-' that indicates it is not
3349 the final line of the response.
3351 31. The loop for finding the name of the user that called Exim had a hardwired
3352 limit of 10; it now uses the value of finduser_retries, which is used for
3353 all other user lookups.
3355 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3358 33. Exim was neglecting to zero errno before one call of strtol() when
3359 expanding a string and expecting an integer value. On some systems this
3360 resulted in spurious "integer overflow" errors. Also, it was casting the
3361 result into an int without checking.
3363 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3364 did not work. The code looks as if it has *never* worked, though it appears
3365 to have been documented since at least releast 1.62. I have made it work.
3367 35. The "timeout_DNS" error in retry rules, also documented since at least
3368 1.62, also never worked. As it isn't clear exactly what this means, and
3369 clearly it isn't a major issue, I have abolished the feature by treating it
3370 as "timeout", and writing a warning to the main and panic logs.
3372 36. The display of retry rules for -brt wasn't always showing the error code
3375 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3376 timeout_connect_A, timeout_connect_MX.
3378 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3379 to the empty sender.
3381 39. The daemon was not analysing the content of -oX till after it had closed
3382 stderr and disconnected from the controlling terminal. This meant that any
3383 syntax errors were only noted on the panic log, and the return code from
3384 the command was 0. By re-arranging the code a little, I've made the
3385 decoding happen first, so such errors now appear on stderr, and the return
3386 code is 1. However, the actual setting up of the sockets still happens in
3387 the disconnected process, so errors there are still only recorded on the
3390 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3391 connections (as happens on some IP stacks) was logged at start up time as
3392 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3393 it from "IPv6 and IPv4", which means that two separate sockets are being
3396 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3397 says whether AF_INET or AF_INET6 was passed as an argument.
3399 42. Exiwhat output was messed up when time zones were included in log
3402 43. Exiwhat now gives more information about the daemon's listening ports,
3403 and whether -tls-on-connect was used.
3405 44. The "port" option of the smtp transport is now expanded.
3407 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3408 silently ignored. Now an error message is written to the main and panic
3411 46. There's a new ACL modifier called "logwrite" which writes to a log file
3412 as soon as it is encountered.
3414 47. Added $local_user_uid and $local_user_gid at routing time.
3416 48. Exim crashed when trying to verify a sender address that was being
3419 49. Exim was recognizing only a space character after ".include". It now also
3420 recognizes a tab character.
3422 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3423 extracting the relevant information from the specification. The man page no
3424 longer contains scrambled data for the -d option, and I've added a section
3425 at the front about calling Exim under different names.
3427 51. Added "extra_headers" argument to the "mail" command in filter files.
3429 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3432 53. Installed eximstats 1.29.
3434 54. Added transport_filter_timeout as a generic transport option.
3436 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3437 Cc: header lines. This was required by RFC 822, but it not required by RFC
3440 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3441 incoming messages that did not have them. Now it does so only if the
3442 message originates locally, that is, if there is no associated remote host
3443 address. When Resent- header lines are present, this applies to the Resent-
3444 lines rather than the non-Resent- lines.
3446 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3447 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3449 58. Messages for configuration errors now include the name of the main
3450 configuration file - useful now that there may be more than one file in a
3451 list (.included file names were always shown).
3453 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3454 for those rare installations that do not start the daemon as root or run it
3455 setuid root. I've cut out the call to initgroups() if the daemon is not
3458 60. The Exim user and group can now be bound into the binary as text strings
3459 that are looked up at the start of Exim's processing.
3461 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3463 62. Added $mailstore_basename variable.
3465 63. Installed patch to sieve.c from Michael Haardt.
3467 64. When Exim failed to open the panic log after failing to open the main log,
3468 the original message it was trying to log was written to stderr and debug
3469 output, but if they were not available (the usual case in production), it
3470 was lost. Now it is written to syslog before the two lines that record the
3471 failures to open the logs.
3473 65. Users' Exim filters run in subprocesses under the user's uid. It is
3474 possible for a "deliver" command or an alias in a "personal" command to
3475 provoke an address rewrite. If logging of address rewriting is configured,
3476 this fails because the process is not running as root or exim. There may be
3477 a better way of dealing with this, but for the moment (because 4.30 needs
3478 to be released), I have disabled address rewrite logging when running a
3479 filter in a non-root, non-exim process.
3485 1. The buildconfig auxiliary program wasn't quoting the value set for
3486 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3487 not defined. This bug was masked in 4.22 by the effect that was fixed in
3490 2. Some messages that were rejected after a message id was allocated were
3491 shown as "incomplete" by exigrep. It no longer does this for messages that
3492 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3494 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3495 have allow_domain_literals set, the ID did not get logged in the <= line.
3496 Domain literals are now always recognized in Message-ID: header lines.
3498 4. The first argument for a ${extract expansion item is the key name or field
3499 number. Leading and trailing spaces in this item were not being ignored,
3500 causing some misleading effects.
3502 5. When deliver_drop_privilege was set, single queue runner processes started
3503 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3504 same command in the process it spins off) were not dropping privilege.
3506 6. When the daemon running as "exim" started a queue runner, it always
3507 re-executed Exim in the spun-off process. This is a waste of effort when
3508 deliver_drop_privilege is set. The new process now just calls the
3509 queue-runner function directly.
3515 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3518 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3519 ignored. Though the use of -oP was forcing the writing of a pid file, it
3520 was always written to the default place.
3522 3. If the message "no IP address found for host xxxx" is generated during
3523 incoming verification, it is now followed by identification of the incoming
3524 connection (so you can more easily find what provoked it).
3526 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3528 5. Added some features to "harden" Exim a bit more against certain attacks:
3530 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3531 be put in Local/Makefile. This is like the never_users runtime option,
3532 but it cannot be overridden. The default setting is "root".
3534 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3535 prefix string with which any file named in a -C command line option
3538 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3539 is retained for -C and -D only if the caller of Exim is root. Without
3540 it, the exim user may also use -C and -D and retain privilege.
3542 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3543 command line option is disabled.
3545 6. Macro names set by the -D option must start with an upper case letter, just
3546 like macro names defined in the configuration file.
3548 7. Added "dereference=" facility to LDAP.
3550 8. Two instances of the typo "uknown" in the source files are fixed.
3552 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3553 the Configure-Makefile script screwed up while processing it.
3555 10. Incorporated PCRE 4.4.
3557 11. The SMTP synchronization check was not operating right at the start of an
3558 SMTP session. For example, it could not catch a HELO sent before the client
3559 waited for the greeting. There is now a check for outstanding input at the
3560 point when the greeting is written. Because of the duplex, asynchronous
3561 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3562 way, but not yet received, when the check is performed.
3564 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3565 on TCP/IP sockets, because this apparently causes some broken clients to
3568 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3569 unchanged) from the Cygwin maintainer.
3571 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3572 was supported instead of testing for mbx. Effectively a typo.
3574 15. The spa authenticator server code was not checking that the input it
3575 received was valid base64.
3577 16. The debug output line for the "set" modifier in ACLs was not showing the
3578 name of the variable that was being set.
3580 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3582 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3583 lookup on its own. Something else was needed to trigger the lookup. For
3584 example, a match in host_lookup or the need for a host name in a host list.
3585 Now, if $sender_host_name is referenced and the host name has not yet been
3586 looked up, a lookup is performed. If the lookup fails, the variable remains
3587 empty, and $host_lookup_failed is set to "1".
3589 19. Added "eqi" as a case-independent comparison operator.
3591 20. The saslauthd authentication condition could segfault if neither service
3592 nor realm was specified.
3594 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3595 error message that was logged was misleading, and incoming SMTP
3596 connections were dropped. The message is now more accurate, and temporary
3597 errors are given to SMTP connections.
3599 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3600 (except RSET) with a 421 error, until QUIT is received. However, it was
3601 failing to send a response to QUIT.
3603 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3604 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3605 ACL is now run after the helo_try_verify_hosts code.
3607 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3608 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3609 case-independent, so other case variants are also recognized). Apparently
3610 some systems use these upper case variants.
3612 25. If more than two messages were waiting for the same host, and a transport
3613 filter was specified for the transport, Exim sent two messages over the
3614 same TCP/IP connection, and then failed with "socket operation on non-
3615 socket" when it tried to send the third.
3617 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3619 27. The extern definition of crypt16() in expand.c was not being excluded when
3620 the OS had its own crypt16() function.
3622 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3623 as a preferred synonym for return_size_limit, both as an option and as an
3626 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3628 30. Changed the default configuration ACL to relax the local part checking rule
3629 for addresses that are not in any local domains. For these addresses,
3630 slashes and pipe symbols are allowed within local parts, but the sequence
3631 /../ is explicitly forbidden.
3633 31. SPA server authentication was not clearing the challenge buffer before
3636 32. log_message in a "warn" ACL statement was writing to the reject log as
3637 well as to the main log, which contradicts the documentation and doesn't
3638 seem right (because no rejection is happening). So I have stopped it.
3640 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3641 However, I am unable to do any testing of this.
3643 34. Fixed an infelicity in the appendfile transport. When checking directories
3644 for a mailbox, to see if any needed to be created, it was accidentally
3645 using path names with one or more superfluous leading slashes; tracing
3646 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3648 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3649 contents are added to the log line that is written for every discarded
3650 recipient. (Previously a log_message setting was ignored.)
3652 36. The ${quote: operator now quotes the string if it is empty.
3654 37. The install script runs exim in order to find its version number. If for
3655 some reason other than non-existence or emptiness, which it checks, it
3656 could not run './exim', it was installing it with an empty version number,
3657 i.e. as "exim-". This error state is now caught, and the installation is
3660 38. An argument was missing from the function that creates an error message
3661 when Exim fails to connect to the socket for saslauthd authentication.
3662 This could cause Exim to crash, or give a corrupted message.
3664 39. Added isip, isip4, and isip6 to ${if conditions.
3666 40. The ACL variables $acl_xx are now saved with the message, and can be
3667 accessed later in routers, transports, and filters.
3669 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3670 strings in the file are not string-expanded.
3672 42. If a MAIL command specified a SIZE value that was too large to fit into an
3673 int variable, the check against message_size_limit failed. Such values are
3674 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3675 day this will have to be increased, but I don't think I want to be around
3676 when emails are that large.
3683 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3684 iconv() is not standard in FreeBSD.
3686 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3687 IPv6 enabled. The observed symptom was a segmentation fault on return from
3688 the function os_common_find_running_interfaces() in src/os.c.
3690 3. In the check_special_case() function in daemon.c I had used "errno" as an
3691 argument name, which causes warnings on some systems. This was basically a
3692 typo, since it was named "eno" in the comments!
3694 4. The code that waits for the clock to tick (at a resolution of some fraction
3695 of a second) so as to ensure message-id uniqueness was always waiting for
3696 at least one whole tick, when it could have waited for less. [This is
3697 almost certainly not relevant at current processor speeds, where it is
3698 unlikely to ever wait at all. But we try to future-proof.]
3700 5. The function that sleeps for a time interval that includes fractions of a
3701 second contained a race. It did not block SIGALRM between setting the
3702 timer, and suspending (a couple of lines later). If the interval was short
3703 and the sigsuspend() was delayed until after it had expired, the suspension
3704 never ended. On busy systems this could lead to processes getting stuck for
3707 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3708 process, before it forks any delivery processes. The open lookup caching
3709 mechanism meant that the open file or database connection was passed into
3710 the delivery process. The problem was that delivery processes always tidy
3711 up cached lookup data. This could cause a problem for the next delivery
3712 process started by the queue runner, because the external queue runner
3713 process does not know about the closure. So the next delivery process
3714 still has data in the lookup cache. In the case of a file lookup, there was
3715 no problem because closing a file descriptor in a subprocess doesn't affect
3716 the parent. However, if the lookup was caching a connection to a database,
3717 the connection was closed, and the second delivery process was likely to
3718 see errors such as "PGSQL: query failed: server closed the connection
3719 unexpectedly". The problem has been fixed by closing all cached lookups
3720 in a queue runner before running a delivery process.
3722 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3723 seem to have the "const" qualifier which it has on other OS. I've
3726 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3727 *of the same type* (client or server) with the same public name that an
3728 error should be diagnosed.
3730 9. When Exim looked up a host name for an IP address, but failed to find the
3731 original IP address when looking up the host name (a safety check), it
3732 output the message "<ip address> does not match any IP for NULL", which was
3733 confusing, to say the least. The bug was that the host name should have
3734 appeared instead of "NULL".
3736 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3737 user that is built into the binary, and the -C or -D options is used, root
3738 privilege is dropped before the configuration file is read. In addition,
3739 logging is switched to stderr instead of the normal log files. If the
3740 configuration then re-defines the Exim user, the unprivileged environment
3741 is probably not what is expected, so Exim logs a panic warning message (but
3744 However, if deliver_drop_privilege is set, the unprivileged state may well
3745 be exactly what is intended, so the warning has been cut out in that case,
3746 and Exim is allowed to try to write to its normal log files.
3752 1. smtp_return_error_details was not giving details for temporary sender
3753 or receiver verification errors.
3755 2. Diagnose a configuration error if two authenticators have the same public
3758 3. Exim used not to create the message log file for a message until the first
3759 delivery attempt. This could be confusing when incoming messages were held
3760 for policy or load reasons. The message log file is now created at the time
3761 the message is received, and an initial "Received" line is written to it.
3763 4. The automatically generated man page for command line options had a minor
3764 bug that caused no ill effects; however, a more serious problem was that
3765 the procedure for building the man page automatically didn't always
3766 operate. Consequently, release 4.20 contains an out-of-date version. This
3767 shouldn't happen again.
3769 5. When building Exim with embedded Perl support, the script that builds the
3770 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3771 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3773 6. The freeze_tell option was not being used for messages that were frozen on
3774 arrival, either by an ACL or by local_scan().
3776 7. Added the smtp_incomplete_transaction log selector.
3778 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3779 was accepting AUTH without a new EHLO.
3781 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3782 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3785 10. Logging of TCP/IP connections (when configured) now happens in the main
3786 daemon process instead of the child process, so that the TCP/IP connection
3787 count is more accurate (but it can never be perfect).
3789 11. The use of "drop" in a nested ACL was not being handled correctly in the
3790 outer ACL. Now, if condition failure induced by the nested "drop" causes
3791 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3792 or "require"), the connection is dropped.
3794 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3795 that yield "discard" can now be used with an "accept" or a "discard" verb,
3796 but an error is generated for any others (because I can't see a useful way
3797 to define what should happen).
3799 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3800 are now processed in the same way as lines in the Exim configuration file.
3801 In particular, continuation lines are supported.
3803 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3805 15. Added -ti meaning -t -i.
3807 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3808 domains, and warn by logging if others are found.
3810 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3811 returned from GIFCONF when Exim is trying to find the list of interfaces on
3812 a host. The code in os.c has been modified to copy each ifreq to an aligned
3813 structure in all cases.
3815 Also, in some cases, the returned ifreq's were being copied to a 'struct
3816 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3817 means the last couple of bytes of an IPv6 address could be chopped if the
3818 ifreq contained only a normal sockaddr (14 bytes storage).
3820 18. Named domain lists were not supported in the hosts_treat_as_local option.
3821 An entry such as +xxxx was not recognized, and was treated as a literal
3824 19. Ensure that header lines added by a DATA ACL are included in the reject log
3825 if the ACL subsequently rejects the message.
3827 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3828 MD5 (which is deprecated).
3830 21. When testing a filter file using -bf, Exim was writing a message when it
3831 took the sender from a "From " line in the message, but it was not doing so
3832 when it took $return_path from a Return-Path: header line. It now does.
3834 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3835 with a valid header line field name (a series of printing characters
3836 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3838 23. Changed "disc" in the source to "disk" to conform to the documentation and
3839 the book and for uniformity.
3841 24. Ignore Sendmail's -Ooption=value command line item.
3843 25. When execve() failed while trying to run a command in a pipe transport,
3844 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3845 could be confused with a return value of 69 from the command itself. This
3846 has been changed to 127, the value the shell returns if it is asked to run
3847 a non-existent command. The wording for the related log line suggests a
3848 non-existent command as the problem.
3850 26. If received_header_text expands to an empty string, do not add a Received:
3851 header line to the message. (Well, it adds a token one on the spool, but
3852 marks it "old" so that it doesn't get used or transmitted.)
3854 27. Installed eximstats 1.28 (addition of -nt option).
3856 28. There was no check for failure on the call to getsockname() in the daemon
3857 code. This can fail if there is a shortage of resources on the system, with
3858 ENOMEM, for example. A temporary error is now given on failure.
3860 29. Contrary to the C standard, it seems that in some environments, the
3861 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3862 program. Exim now does this explicitly; it affects the formatting of
3863 timestamps using strftime().
3865 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3866 complaints. I've now initialized all the variables, to avoid this.
3868 32. Header lines added by a system filter were not being "seen" during
3869 transport-time rewrites.
3871 33. The info_callback() function passed to OpenSSL is set up with type void
3872 (*)(SSL *, int, int), as described somewhere. However, when calling the
3873 function (actually a macro) that sets it up, the type void(*)() is
3874 expected. I've put in a cast to prevent warnings from picky compilers.
3876 34. If a DNS black list lookup found a CNAME record, but there were no A
3877 records associated with the domain it pointed at, Exim crashed.
3879 35. If a DNS black list lookup returned more than one A record, Exim ignored
3880 all but the first. It now scans all returned addresses if a particular IP
3881 value is being sought. In this situation, the contents of the
3882 $dnslist_value variable are a list of all the addresses, separated by a
3885 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3886 to accept a host name and all its aliases if the forward lookup for any of
3887 them yielded the IP address of the incoming connection. Now it accepts only
3888 those names whose forward lookup yields the correct IP address. Any other
3889 names are discarded. This closes a loophole whereby a rogue DNS
3890 administrator could create reverse DNS records to break through a
3891 wildcarded host restriction in an ACL.
3893 37. If a user filter or a system filter that ran in a subprocess used any of
3894 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3895 the wrong values were passed to the pipe command ($thisaddress had the
3896 value of $0, $0 had the value of $1, etc). This bug was introduced by
3897 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3899 38. Improved the line breaking for long SMTP error messages from ACLs.
3900 Previously, if there was no break point between 40 and 75 characters, Exim
3901 left the rest of the message alone. Two changes have been made: (a) I've
3902 reduced the minimum length to 35 characters; (b) if it can't find a break
3903 point between 35 and 75 characters, it looks ahead and uses the first one
3904 that it finds. This may give the occasional overlong line, but at least the
3905 remaining text gets split now.
3907 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3908 file descriptors might be low, and that setting 1000 would always raise it.
3909 It turns out that in some environments, the limit is already over 1000 and
3910 that lowering it causes trouble. So now Exim takes care not to decrease it.
3912 40. When delivering a message, the value of $return_path is set to $sender_
3913 address at the start of routing (routers may change the value). By an
3914 oversight, this default was not being set up when an address was tested by
3915 -bt or -bv, which affected the outcome if any router or filter referred to
3918 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3919 log only when "message" or "log_message" are set. However, if one of the
3920 conditions was an address verification, or a call to a nested ACL, the
3921 messages generated by the underlying test were being passed through. This
3922 no longer happens. The underlying message is available in $acl_verify_
3923 message for both "message" and "log_message" expansions, so it can be
3924 passed through if needed.
3926 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3927 new expansion $bh_ to give the encoded byte string without charset
3928 translation. Translation happens only if iconv() is available; HAVE_ICONV
3929 indicates this at build time. HEADERS_CHARSET gives the charset to
3930 translate to; headers_charset can change it in the configuration, and
3931 "headers charset" can change it in an individual filter file.
3933 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3934 that creates RFC 2047 encoded "words" labels them as that charset instead
3935 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3936 expansion operator; (ii) when Exim creates a From: line for a local
3937 message; (iii) when a header line is rewritten to include a "phrase" part.
3939 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3940 buggy, causing it to skip the first lines of messages whose message ID
3941 ended in 'D'. This would not have bitten before Exim release 4.14, because
3942 message IDs were unlikely to end in 'D' before then. The effect was to have
3943 incorrect size information for certain domains.
3945 45. #include "config.h" was missing at the start of the crypt16.c module. This
3946 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3949 46. If there was a timeout during a "random" callout check, Exim treated it as
3950 a failure of the random address, and carried on sending RSET and the real
3951 address. If the delay was just some slowness somewhere, the response to the
3952 original RCPT would be taken as a response to RSET and so on, causing
3953 mayhem of various kinds.
3955 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3956 when I implemented it. It didn't allow for the fact that some option values
3957 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3958 the right test for positive values.
3960 48. Domain names in DNS records are case-independent. Exim always looks them up
3961 in lower case. Some resolvers return domain names in exactly the case they
3962 appear in the zone file, that is, they may contain uppercase letters. Not
3963 all resolvers do this - some return always lower case. Exim was treating a
3964 change of case by a resolver as a change of domain, similar to a widening
3965 of a domain abbreviation. This triggered its re-routing code and so it was
3966 trying to route what was effectively the same domain again. This normally
3967 caused routing to fail (because the router wouldn't handle the domain
3968 twice). Now Exim checks for this case specially, and just changes the
3969 casing of the domain that it ultimately uses when it transmits the message
3972 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3975 50. If a filter generated a file delivery with a non-absolute name (possible if
3976 no home directory exists for the router), the forbid_file option was not
3979 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3980 the existing equality matching.
3982 52. Exim was using ints instead of ino_t variables in some places where it was
3983 dealing with inode numbers.
3985 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3986 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3987 called TMPDIR, and if it finds it is different, it changes its value.
3989 54. The smtp_printf() function is now made available to local_scan() so
3990 additional output lines can be written before returning. There is also an
3991 smtp_fflush() function to enable the detection of a dropped connection.
3992 The variables smtp_input and smtp_batched_input are exported to
3995 55. Changed the default runtime configuration: the message "Unknown user"
3996 has been removed from the ACL, and instead placed on the localuser router,
3997 using the cannot_route_message feature. This means that any verification
3998 failures that generate their own messages won't get overridden. Similarly,
3999 the "Unrouteable address" message that was in the ACL for unverifiable
4000 relay addresses has also been removed.
4002 56. Added hosts_avoid_esmtp to the smtp transport.
4004 57. The exicyclog script was not checking for the esoteric option
4005 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4006 will work only if exicyclog is run under the appropriate euid.
4008 58. Following a discussion on the list, the rules by which Exim recognises line
4009 endings on incoming messages have been changed. The -dropcr and drop_cr
4010 options are now no-ops, retained only for backwards compatibility. The
4011 following line terminators are recognized: LF CRLF CR. However, special
4012 processing applies to CR:
4014 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4015 nor a local message in the state where . is a terminator.
4017 (ii) If a bare CR is encountered in a header line, an extra space is added
4018 after the line terminator so as not to end the header. The reasoning
4019 behind this is that bare CRs in header lines are most likely either
4020 to be mistakes, or people trying to play silly games.
4022 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4023 was being incorrectly given as 18 bytes larger than it should have been.
4024 This is a VOB (very old bug).
4026 60. This may never have affected anything current, but just in case it has:
4027 When the local host is found other than at the start of a list of hosts,
4028 the local host, those with the same MX, and any that follow, are discarded.
4029 When the list in question was part of a longer list of hosts, the following
4030 hosts (not currently being processed) were also being discarded. This no
4031 longer happens. I'm not sure if this situation could ever has previously
4034 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4037 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4038 header. This is something that is recommended in a new Internet Draft, and
4039 is something that is documented as being done by Sendmail. There are two
4040 possible values. For messages generated by the autoreply transport, Exim
4043 Auto-Submitted: auto-replied
4045 whereas for all other generated messages (e.g. bounces) it adds
4047 Auto-Submitted: auto-generated
4049 63. The "personal" condition in filters now includes a test for the
4050 Auto-Submitted: header. If it contains the string "auto-" the message it
4051 not considered personal.
4053 64. Added rcpt_include_affixes as a generic transport option.
4055 65. Added queue_only_override (default true).
4057 66. Added the syslog_duplication option.
4059 67. If what should have been the first header line of a message consisted of
4060 a space followed by a colon, Exim was mis-interpreting it as a header line.
4061 It isn't of course - it is syntactically invalid and should therefore be
4062 treated as the start of the message body. The misbehaviour could have
4063 caused a number of strange effects, including loss of data in subsequent
4064 header lines, and spool format errors.
4066 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4067 client host had authenticated. This control can now be exercised by an ACL
4068 for more flexibility.
4070 69. By default, callouts do not happen when testing with -bh. There is now a
4071 variant, -bhc, which does actually run the callout code, including
4072 consulting and updating the callout cache.
4074 70. Added support for saslauthd authentication, courtesy of Alexander
4077 71. If statvfs() failed on the spool or log directories while checking their
4078 size for availability, Exim confusingly gave the error "space shortage".
4079 Furthermore, in debugging mode it crashed with a floating point exception.
4080 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4081 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4082 serious problem, Exim now writes to the main and panic logs when this
4083 happens, with details of the failure. It then refuses to accept the
4084 incoming message, giving the message "spool directory problem" or "log
4085 directory problem" with a 421 code for SMTP messages.
4087 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4088 0, 1, and 2 exist, because some OS complain for execs without them (see
4089 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4090 descriptors. However, the code omitted to check that the open succeeded,
4091 causing mysterious errors if for some reason the permissions on /dev/null
4092 got screwed. Now Exim writes a message to the main and panic logs, and
4093 bombs out if it can't open /dev/null.
4095 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4096 interact so that it is all more flexible. It is supposed to remain
4097 backwards compatible. Also added extra_local_interfaces.
4099 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4100 to bomb out with an assertion failure - to the client this appears as a
4101 connection drop. This problem occurs in the part of the code that was taken
4102 from the Samba project. Fortunately, the assertion is in a very simple
4103 function, so I have fixed this by reproducing the function inline in the
4104 one place where it is called, and arranging for authentication to fail
4105 instead of killing the process with assert().
4107 75. The SPA client code was not working when the server requested OEM rather
4108 than Unicode encoding.
4110 76. Added code to make require_files with a specific uid setting more usable in
4111 the case where statting the file as root fails - usually a non-root-mounted
4112 NFS file system. When this happens and the failure is EACCES, Exim now
4113 forks a subprocess and does the per-uid checking as the relevant uid.
4115 77. Added process_log_path.
4117 78. If log_file_path was not explicitly set, a setting of check_log_space or
4118 check_log_inodes was ignored.
4120 79. If a space check for the spool or log partitions fails, the incident is now
4121 logged. Of course, in the latter case the data may get lost...
4123 80. Added the %p formatting code to string_format() so that it can be used to
4124 print addresses in debug_print(). Adjusted all the address printing in the
4125 debugging in store.c to use %p rather than %d.
4127 81. There was a concern that a line of code in smtp_in.c could overflow a
4128 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4129 initially expressed, the concern was not well-founded, because trailing
4130 spaces are removed early. However, if the trailing spaces were followed by
4131 a NULL, they did not get removed, so the overflow was possible. Two fixes
4134 (a) I re-wrote the offending code in a cleaner fashion.
4135 (b) If an incoming SMTP command contains a NULL character, it is rejected
4138 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4139 runs initgroups(), so that if the Exim user is in any additional groups,
4140 they will be used during message reception.
4146 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob