1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.382 2006/07/27 10:36:34 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
24 with consistent privilege compared to when running as a daemon.
26 TF/03 Note in the spec that $authenticated_id is not set for local
27 submissions from trusted users.
29 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
30 Thanks to Dean Brooks <dean@iglou.com> for the patch.
32 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
33 by adding some example configuration directives to the default
34 configuration file. A little bit of work is required to uncomment the
35 directives and define how usernames and passwords are checked, but
36 there is now a framework to start from.
38 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
39 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
40 without this. I don't know how relevant this is to other LDAP libraries.
42 PH/02 Add the verb name to the "unknown ACL verb" error.
44 PH/03 Magnus Holmgren's patch for filter_prepend_home.
46 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
48 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
49 directory not expanded when it should be if an expanded home directory
50 was set for the address (which is overridden by the transport).
52 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
55 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
56 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
57 because it is too late at that time, and has no effect.
59 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
60 security issue with \' (bugzilla #107). I could not use the
61 PQescapeStringConn() function, because it needs a PGconn value as one of
64 PH/08 When testing addresses using -bt, indicate those final addresses that
65 are duplicates that would not cause an additional delivery. At least one
66 person was confused, thinking that -bt output corresponded to deliveries.
67 (Suppressing duplicates isn't a good idea as you lose the information
68 about possibly different redirections that led to the duplicates.)
70 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
71 systems where poll() doesn't work, in particular OS X.
73 PH/10 Added more information to debugging output for retry time not reached.
75 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
76 operations in malware.c.
78 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
81 PH/13 If write_rejectlog was set false when logging was sent to syslog with
82 syslog_duplication set false, log lines that would normally be written
83 both the the main log and to the reject log were not written to syslog at
86 PH/14 In the default configuration, change the use of "message" in ACL warn
87 statements to "add_header".
89 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
90 not followed by a command (e.g. "seen endif").
92 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
93 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
96 PH/17 Added an extra condition to the default value of delay_warning_condition
100 { match{$h_precedence:}{(?i)bulk|list|junk} }
101 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} }
104 The Auto-Submitted: header is standardised, whereas I don't think
105 Precedence: ever was.
107 PH/18 Refactored debugging code in route_finduser() to show more information,
108 in particular, the error code if getpwnam() issues one.
110 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
111 This is apparently needed in addition to the PH/07 change above to avoid
112 any possible encoding problems.
114 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
115 but not after initializing Perl.
117 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
118 output them only if debugging. By default they are written stderr,
119 apparently, which is not desirable.
121 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
124 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
127 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
129 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
130 authenticated or an ident call has been made. Suppress the default
131 values for $authenticated_id and $authenticated_sender (but permit -oMai
132 and -oMas) when testing with -bh.
134 PH/24 Re-jigged the order of the tests in the default configuration so that the
135 tests for valid domains and recipients precede the DNS black list and CSA
136 tests, on the grounds that those ones are more expensive.
142 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
143 other effects) broke the use of negated acl sub-conditions.
145 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
148 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
149 "Deny" causes Exim to reject the incoming connection with a 554 error.
150 Unfortunately, if there is a major crisis, such as a disk failure,
151 tcp-wrappers gives "deny", whereas what one would like would be some
152 kind of temporary error. A kludge has been added to help with this.
153 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
154 554 error is used if errno is still zero or contains ENOENT (which occurs
155 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
158 PH/03 Add -lutil to the default FreeBSD LIBS setting.
160 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
161 errors. Otherwise a message that provokes a temporary error (when other
162 messages do not) can cause a whole host to time out.
164 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
165 addresses were routed directly to files or pipes from a redirect router.
166 File deliveries just didn't batch; pipe deliveries might have suffered
169 PH/06 A failure to get a lock for a hints database would erroneously always say
170 "Failed to get write lock", even when it was really a read lock.
172 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
173 of 0600. This has been changed to use the value of the lockfile_mode
174 option (which defaults to 0600).
176 PH/08 Applied small patch from the Sieve maintainer.
178 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
179 folder from quota calculations, a direct delivery into this folder messed
180 up the contents of the maildirsize file. This was because the regex was
181 used only to exclude .Trash (or whatever) when the size of the mailbox
182 was calculated. There was no check that a delivery was happening into an
183 excluded directory. This bug has been fixed by ignoring all quota
184 processing for deliveries into excluded directories.
186 PH/10 Added the maildirfolder_create_regex option to appendfile.
192 PH/01 The code for finding all the local interface addresses on a FreeBSD
193 system running IPv6 was broken. This may well have applied to all BSD
194 systems, as well as to others that have similar system calls. The broken
195 code found IPv4 interfaces correctly, but gave incorrect values for the
196 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
197 that it would not match correctly against @[] and not recognize the IPv6
200 PH/02 The ipliteral router was not recognizing addresses of the form user@
201 [ipv6:....] because it didn't know about the "ipv6:" prefix.
203 PH/03 Added disable_ipv6.
205 PH/04 Changed $reply_address to use the raw form of the headers instead of the
206 decoded form, because it is most often used to construct To: headers
207 lines in autoreplies, and the decoded form may well be syntactically
208 invalid. However, $reply_address has leading white space removed, and all
209 newlines turned into spaces so that the autoreply transport does not
212 PH/05 If group was specified without a user on a router, and no group or user
213 was specified on a transport, the group from the router was ignored.
215 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
216 for visible compile-time settings that can be used to change these
217 numbers, for those that want even more. Backwards compatibility with old
218 spool files has been maintained. However, going back to a previous Exim
219 release will lost any variables that are in spool files.
221 PH/07 Two small changes when running in the test harness: increase delay when
222 passing a TCP/IP connection to a new process, in case the original
223 process has to generate a bounce, and remove special handling of
224 127.0.0.2 (sic), which is no longer necessary.
226 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
227 be the same on different OS.
229 PH/09 Moved a debug statement in filter processing to avoid a race problem when
232 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
233 whether --show-vars was specified or not
235 JJ/02 exipick: Added support for new ACL variable spool format introduced
238 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
239 syntactically invalid From: or Reply-to: line, and a filter used this to
240 generate an autoreply, and therefore failed to obtain an address for the
241 autoreply, Exim could try to deliver to a non-existent relative file
242 name, causing unrelated and misleading errors. What now happens is that
243 it logs this as a hard delivery error, but does not attempt to create a
246 PH/11 The exinext utility has a -C option for testing purposes, but although
247 the given file was scanned by exinext itself; it wasn't being passed on
248 when Exim was called.
250 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
251 an end-of-file indication when reading a command response.
253 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
254 compiled. In many other places in Exim, IPv6 addresses are always
255 recognized, so I have changed this. It also means that IPv4 domain
256 literals of the form [IPV4:n.n.n.n] are now always recognized.
258 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
259 used if the router is not running as root, for example, when verifying at
260 ACL time, or when using -bh. The debugging output from this situation was
261 non-existent - all you got was a failure to exec. I have made two
264 (a) Failures to set uid/gid, the current directory, or a process leader
265 in a subprocess such as that created by queryprogram now generate
266 suitable debugging ouput when -d is set.
268 (b) The queryprogram router detects when it is not running as root,
269 outputs suitable debugging information if -d is set, and then runs
270 the subprocess without attempting to change uid/gid.
272 PH/15 Minor change to Makefile for building test_host (undocumented testing
275 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
276 additional section of a DNS packet that returns MX or SRV records.
277 Instead, it always explicitly searches for A/AAAA records. This avoids
278 major problems that occur when a DNS server includes only records of one
279 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
280 fixed another bug: if SRV records were looked up and the corresponding
281 address records were *not* found in the additional section, the port
282 values from the SRV records were lost.
284 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
285 using the correct key (the original address) when searching the retry
286 rules in order to find which one to use for generating the retry hint.
288 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
289 adding the default one. Similarly, if it contains a Reply-To: header, the
290 errors_reply_to option, if set, is not used.
292 PH/19 When calculating a retry time, Exim used to measure the "time since
293 failure" by looking at the "first failed" field in the retry record. Now
294 it does not use this if it is later than than the arrival time of the
295 message. Instead it uses the arrival time. This makes for better
296 behaviour in cases where some deliveries succeed, thus re-setting the
297 "first failed" field. An example is a quota failure for a huge message
298 when small messages continue to be delivered. Without this change, the
299 "time since failure" will always be short, possible causing more frequent
300 delivery attempts for the huge message than are intended.
301 [Note: This change was subsequently modified - see PH/04 for 4.62.]
303 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
304 $1, $2, $3) because the numerical variables can be reset during some
305 expansion items (e.g. "match"), thereby losing the authentication data.
307 PH/21 Make -bV show the size of off_t variables so that the test suite can
308 decide whether to run tests for quotas > 2G.
310 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
311 mailbox_size, and mailbox_filecount in the appendfile transport. If a
312 filecount value is greater than 2G or if a quota value is greater than 2G
313 on a system where the size of off_t is not greater than 4, a panic error
316 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
317 never match. The debug and -bh output now contains an explicit error
318 message indicating a malformed IPv4 address or mask.
320 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
321 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
324 PH/25 Do not write to syslog when running in the test harness. The only
325 occasion when this arises is a failure to open the main or panic logs
326 (for which there is an explicit test).
328 PH/26 Added the /no_tell option to "control=freeze".
330 PH/27 If a host name lookup failed very early in a connection, for example, if
331 the IP address matched host_lookup and the reverse lookup yielded a name
332 that did not have a forward lookup, an error message of the form "no IP
333 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
334 could be logged. Now it outputs the IP address instead of "NULL".
336 PH/28 An enabling patch from MH: add new function child_open_exim2() which
337 allows the sender and the authenticated sender to be set when
338 submitting a message from within Exim. Since child_open_exim() is
339 documented for local_scan(), the new function should be too.
341 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
342 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
343 results in an empty string is now treated as unset.
345 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
347 PH/31 Added sender_verify_fail logging option.
349 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
350 needed by RFC 822 but not by RFC 2822 was commented out. I have now
351 tidied the source and removed it altogether.
353 PH/33 When a queue run was abandoned because the load average was too high, a
354 log line was always written; now it is written only if the queue_run log
355 selector is set. In addition, the log line for abandonment now contains
356 information about the queue run such as the pid. This is always present
357 in "start" and "stop" lines but was omitted from the "abandon" line.
359 PH/34 Omit spaces between a header name and the colon in the error message that
360 is given when verify = headers_syntax fails (if there are lots of them,
361 the message gets confusing).
363 PH/35 Change the default for dns_check_names_pattern to allow slashes within
364 names, as there are now some PTR records that contain slashes. This check
365 is only to protect against broken name servers that fall over on strange
366 characters, so the fact that it applies to all lookups doesn't matter.
368 PH/36 Now that the new test suite is complete, we can remove some of the
369 special code in Exim that was needed for the old test suite. For example,
370 sorting DNS records because real resolvers return them in an arbitrary
371 order. The new test suite's fake resolver always returns records in the
374 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
375 bounces) except when queue_only is set, to avoid logging races between
376 the different processes.
378 PH/38 Panic-die if .include specifies a non-absolute path.
380 PH/39 A tweak to the "H" retry rule from its user.
382 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
383 a label. They prevented compilation on older perls.
385 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
386 a warning to be raised on newish perls.
388 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
389 on queue. Changes to match documented behaviour of showing count of
390 messages matching specified criteria.
392 PH/40 Changed the default ident timeout from 30s to 5s.
394 PH/41 Added support for the use of login_cap features, on those BSD systems
395 that have them, for controlling the resources used by pipe deliveries.
397 PH/42 The content-scanning code uses fopen() to create files in which to put
398 message data. Previously it was not paying any attention to the mode of
399 the files. Exim runs with umask(0) because the rest of the code creates
400 files with open(), and sets the required mode explicitly. Thus, these
401 files were ending up world-writeable. This was not a big issue, because,
402 being within the spool directory, they were not world-accessible. I have
403 created a function called modefopen, which takes an additional mode
404 argument. It sets umask(777), creates the file, chmods it to the required
405 mode, then resets the umask. All the relevant calls to fopen() in the
406 content scanning code have been changed to use this function.
408 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
409 to 24 hours. This avoids potential overflow problems when processing G
410 and H retry rules. I suspect nobody ever tinkers with this value.
412 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
414 PH/45 When the plaintext authenticator is running as a client, the server's
415 challenges are checked to ensure they are valid base64 strings. By
416 default, the authentication attempt is cancelled if an invalid string is
417 received. Setting client_ignore_invalid_base64 true ignores these errors.
418 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
419 they are received. Thus, the responses can be made to depend on the
420 challenges. If an invalid string is ignored, an empty string is placed in
423 PH/46 Messages that are created by the autoreply transport now contains a
424 References: header, in accordance with RFCs 2822 and 3834.
426 PH/47 Added authenticated_sender_force to the smtp transport.
428 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
430 PH/49 Installed latest patch from the Sieve maintainer.
432 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
433 was also set, the appendfile transport was unnecessarily scanning a
434 directory of message files (e.g. for maildir delivery) to find the count
435 of files (along with the size), even though it did not need this
436 information. It now does the scan only if it needs to find either the
437 size of the count of files.
439 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
441 PH/52 Two bugs concerned with error handling when the smtp transport is
444 (i) Exim was not creating retry information for temporary errors given
445 for individual recipients after the DATA command when the smtp transport
446 was used in LMTP mode. This meant that they could be retried too
447 frequently, and not timed out correctly.
449 (ii) Exim was setting the flag that allows error details to be returned
450 for LMTP errors on RCPT commands, but not for LMTP errors for individual
451 recipients that were returned after the DATA command.
453 PH/53 This is related to PH/52, but is more general: for any failing address,
454 when detailed error information was permitted to be returned to the
455 sender, but the error was temporary, then after the final timeout, only
456 "retry timeout exceeded" was returned. Now it returns the full error as
457 well as "retry timeout exceeded".
459 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
460 do this, and (what is worse) MTAs that accept it.
462 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
463 will now be deprecated.
465 PH/56 New os.c-cygwin from the Cygwin maintainer.
467 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
468 formats (previously only available in exim formats via -bpr, -bpru,
469 and -bpra. Now also available in native and exiqgrep formats)
471 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
472 with very large, slow to parse queues
474 JJ/08 exipick: added ! as generic prefix to negate any criteria format
476 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
478 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
479 responses to authentication challenges, though it was showing the
480 challenges; (ii) I've removed the CR characters from the debug output for
483 PH/58 Allow for the insertion of a newline as well as a space when a string
484 is turned into more than one encoded-word during RFC 2047 encoding. The
485 Sieve code now uses this.
487 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
488 data_4xx, lost_connection, tls_required.
490 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
491 message was being sent as an SMTP response.
493 PH/61 Add -l and -k options to exicyclog.
495 PH/62 When verifying, if an address was redirected to one new address, so that
496 verification continued, and the new address failed or deferred after
497 having set something in $address_data, the value of $address_data was not
498 passed back to the ACL. This was different to the case when no
499 redirection occurred. The value is now passed back in both cases.
501 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
502 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
503 use login_cap.h, so on its own it isn't the distinguishing feature. The
504 new name refers directly to the setclassresources() function.
506 PH/65 Added configuration files for NetBSD3.
508 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
510 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
511 is preferred over IPv4.
513 PH/68 The bounce_return_message and bounce_return_body options were not being
514 honoured for bounces generated during the reception of non-SMTP messages.
515 In particular, this applied to messages rejected by the ACL. This bug has
516 been fixed. However, if bounce_return_message is true and bounce_return_
517 body is false, the headers that are returned for a non-SMTP message
518 include only those that have been read before the error was detected.
519 (In the case of an ACL rejection, they have all been read.)
521 PH/69 The HTML version of the specification is now built in a directory called
522 spec_html instead of spec.html, because the latter looks like a path with
523 a MIME-type, and this confuses some software.
525 PH/70 Catch two compiler warnings in sieve.c.
527 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
528 function verify_get_ident() calls ip_connect() to connect a socket, but
529 if the "connect()" function timed out, ip_connect() used to close the
530 socket. However, verify_get_ident() also closes the socket later, and in
531 between Exim writes to the log, which may get opened at this point. When
532 the socket was closed in ip_connect(), the log could get the same file
533 descriptor number as the socket. This naturally causes chaos. The fix is
534 not to close the socket in ip_connect(); the socket should be closed by
535 the function that creates it. There was only one place in the code where
536 this was missing, in the iplookup router, which I don't think anybody now
537 uses, but I've fixed it anyway.
539 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
540 well as to direct DNS lookups. Otherwise the handling of names in host
541 lists is inconsistent and therefore confusing.
547 PH/01 Two changes to the default runtime configuration:
549 (1) Move the checks for relay_from_hosts and authenticated clients from
550 after to before the (commented out) DNS black list checks.
552 (2) Add control=submission to the relay_from_hosts and authenticated
553 clients checks, on the grounds that messages accepted by these
554 statements are most likely to be submissions.
556 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
558 (1) Generate an error if the third argument for the ${prvs expansion is
561 (2) Treat a missing third argument of ${prvscheck as if it were an empty
564 (3) Reset the variables that are obtained from the first argument of
565 ${prvscheck and used in the second argument before leaving the code,
566 because their memory is reclaimed, so using them afterwards may do
569 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
570 one (it's much easier than Tom thought :-).
572 (5) Because of (4), we can now allow for the use of $prvscheck_result
573 inside the third argument.
575 PH/03 For some reason, the default setting of PATH when running a command from
576 a pipe transport was just "/usr/bin". I have changed it to
579 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
580 anything to be listed in the output from -bV.
582 PH/05 When a filter generated an autoreply, the entire To: header line was
583 quoted in the delivery log line, like this:
585 => >A.N.Other <ano@some.domain> <original@ddress> ...
587 This has been changed so that it extracts the operative address. There
588 may be more than one such address. If so, they are comma-separated, like
591 => >ano@some.domain,ona@other.domain <original@ddress> ...
593 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
594 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
595 not being looked up in the rDNS to get a host name, Exim was showing the
596 IP address twice in Received: lines, even though the IP addresses were
597 identical. For example:
599 Received: from [1.2.3.4] (helo=[1.2.3.4])
601 However, if the real host name was known, it was omitting the HELO data
602 if it matched the actual IP address. This has been tidied up so that it
603 doesn't show the same IP address twice.
605 PH/07 When both +timestamp and +memory debugging was on, the value given by
606 $tod_xxx expansions could be wrong, because the tod_stamp() function was
607 called by the debug printing, thereby overwriting the timestamp buffer.
608 Debugging no longer uses the tod_stamp() function when +timestamp is set.
610 PH/08 When the original message was included in an autoreply transport, it
611 always said "this is a copy of the message, including all the headers",
612 even if body_only or headers_only was set. It now gives an appropriate
615 PH/09 Applied a patch from the Sieve maintainer which:
617 o fixes some comments
618 o adds the (disabled) notify extension core
619 o adds some debug output for the result of if/elsif tests
620 o points to the current vacation draft in the documentation
621 and documents the missing references header update
625 o fixes a bug in processing the envelope test (when testing
626 multiple envelope elements, the last element determinted the
629 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
630 Electronic Mail") by including:
632 Auto-submitted: auto-generated
634 in the messages that it generates (bounce messages and others, such as
635 warnings). In the case of bounce messages for non-SMTP mesages, there was
636 also a typo: it was using "Auto_submitted" (underscore instead of
637 hyphen). Since every message generated by Exim is necessarily in response
638 to another message, thes have all been changed to:
640 Auto-Submitted: auto-replied
642 in accordance with these statements in the RFC:
644 The auto-replied keyword:
646 - SHOULD be used on messages sent in direct response to another
647 message by an automatic process,
649 - MUST NOT be used on manually-generated messages,
651 - MAY be used on Delivery Status Notifications (DSNs) and Message
652 Disposition Notifications (MDNs),
654 - MUST NOT be used on messages generated by automatic or periodic
655 processes, except for messages which are automatic responses to
658 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
659 to the default Received: header definition.
661 PH/12 Added log selector acl_warn_skipped (default on).
663 PH/13 After a successful wildlsearch lookup, discard the values of numeric
664 variables because (a) they are in the wrong storage pool and (b) even if
665 they were copied, it wouldn't work properly because of the caching.
667 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
668 checking when decoding. Apparently there are clients that generate
669 overlong encoded strings. Why am I not surprised?
671 PH/15 If the first argument of "${if match_address" was not empty, but did not
672 contain an "@" character, Exim crashed. Now it writes a panic log message
673 and treats the condition as false.
675 PH/16 In autoreply, treat an empty string for "once" the same as unset.
677 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
678 extension "envelope-auth". The code is finished and in agreement with
679 other implementations, but there is no documentation so far and in fact,
680 nobody wrote the draft yet. This extension is currently #undef'ed, thus
681 not changing the active code.
683 Print executed "if" and "elsif" statements when debugging is used. This
684 helps a great deal to understand what a filter does.
686 Document more things not specified clearly in RFC3028. I had all this
687 sorted out, when out of a sudden new issues came to my mind. Oops."
689 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
692 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
693 canonical form (as documented). However, after a host name lookup from
694 the IP address, check_host() was doing a simple string comparison with
695 addresses acquired from the DNS when checking that the found name did
696 have the original IP as one of its addresses. Since any found IPv6
697 addresses are likely to be in abbreviated form, the comparison could
698 fail. Luckily, there already exists a function for doing the comparison
699 by converting both addresses to binary, so now that is used instead of
702 PH/20 There was another similar case to PH/19, when a complete host name was
703 given in a host list; looking up its IP address could give an abbreviated
704 form, whereas the current host's name might or might not be abbreviated.
705 The same fix has been applied.
711 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
712 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
715 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
716 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
718 PH/03 Typo: missing ".o" in src/pcre/Makefile.
720 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
721 header line, restrict the check to what is listed in RFCs 2369 and 2929.
722 Also, for "Auto-Submitted", treat anything other than "no" as
723 non-personal, in accordance with RFC 3834. (Previously it treated
724 anything starting "auto-" as non-personal.)
726 TF/01 The control=submission/name=... option had a problem with syntax
727 errors if the name included a slash character. The /name= option
728 now slurps the rest of the string, so it can include any characters
729 but it must come last in the list of options (after /sender_retain
732 PH/05 Some modifications to the interface to the fake nameserver for the new
740 TK/01 Added the "success_on_redirect" address verification option. See
741 NewStuff for rationale and an example.
743 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
745 PH/02 Patch to exigrep to allow it to work on syslog lines.
747 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
748 fread() to skip over the body file's header line, because in Cygwin the
749 header line is locked and is inaccessible.
751 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
752 co-exist for some time) to make it clear that it is the Exim ID that is
753 referenced, not the Message-ID: header line.
755 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
756 string_format() function, because snprintf() does not exist on all
759 PH/06 The use of forbid_filter_existstest now also locks out the use of the
760 ${stat: expansion item.
762 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
763 protocol synchronization error", to keep the pedants happy.
765 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
766 well as for IRIX systems, when gcc is being used. See the host.c source
769 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
771 PH/10 Named domain lists were not working if used in a queue_smtp_domains
774 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
775 transport and to the smtp transport in LMTP mode.
777 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
779 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
780 run a filter in a subprocess. This could lead to confusion in subsequent
781 lookups in the parent process. There should also be a search_tidyup() at
782 the end of the subprocess.
784 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
785 only if the host matched helo_try_verify_hosts, which caused the
786 verification to occur when the EHLO/HELO command was issued. The ACL just
787 tested the remembered result. Now, if a previous verification attempt has
788 not happened, "verify = helo" does it there and then.
790 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
792 TK/03 Fix log output including CR from clamd.
794 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
795 exist provoked a memory error which could cause a segfault.
797 PH/15 Installed PCRE 6.2
799 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
801 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
802 of the problem. Specifically, suggested +O2 rather than +O1 for the
805 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
807 PH/20 If a delivery was routed to a non-standard port by means of an SRV
808 record, the port was not correctly logged when the outgoing_port log
809 selector was set (it logged the transort's default port).
811 PH/21 Added support for host-specific ports to manualroute, queryprogram,
812 fallback_hosts, and "hosts" in the smtp transport.
814 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
815 host errors such as "Connection refused".
817 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
818 authentication with radiusclient 0.4.9:
820 - Error returned from rc_read_config was caught wrongly
821 - Username/password not passed on to radius server due to wrong length.
823 The presumption is that some radiusclient API changes for 4.51/PH/17
824 were not taken care of correctly. The code is still untested by me (my
825 Linux distribution still has 0.3.2 of radiusclient), but it was
826 contributed by a Radius user.
828 PH/24 When doing a callout, the value of $domain wasn't set correctly when
829 expanding the "port" option of the smtp transport.
831 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
832 while reading a MIME header. Thanks to Tom Hughes for a patch.
834 PH/24 Include config.h inside local_scan.h so that configuration settings are
837 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
838 that in an ACL for RCPT (for example), you can examine exactly what was
841 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
842 commands, but it was not correctly comparing the address with the actual
843 client host address. Thus, it would show the EHLO address in Received:
844 header lines when this was not necessary.
846 PH/27 Added the % operator to ${eval:}.
848 PH/28 Exim tries to create and chdir to its spool directory when it starts;
849 it should be ignoring failures (because with -C, for example, it has lost
850 privilege). It wasn't ignoring creation failures other than "already
853 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
856 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
857 because an input line was too long, either on its own, or by virtue of
858 too many continuations, the temporary file was not being removed, and the
859 return code was incorrect.
861 PH/31 Missing "BOOL" in function definition in filtertest.c.
863 PH/32 Applied Sieve patches from the maintainer.
865 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
867 PH/33 Added "verify = not_blind".
869 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
870 Local/Makefile (with some defaults set). These are used in built scripts
871 such as exicyclog, but they have never been used in the exim_install
872 script (though there are many overriding facilities there). I have
873 arranged that the exim_install script now takes note of these two
876 PH/35 Installed configuration files for Dragonfly.
878 PH/36 When a locally submitted message by a trusted user did not contain a
879 From: header, and the sender address was obtained from -f or from an SMTP
880 MAIL command, and the trusted user did not use -F to supply a sender
881 name, $originator_name was incorrectly used when constructing a From:
882 header. Furthermore, $originator_name was used for submission mode
883 messages from external hosts without From: headers in a similar way,
884 which is clearly wrong.
886 PH/37 Added control=suppress_local_fixups.
888 PH/38 When log_selector = +received_sender was set, and the addition of the
889 sender made the log line's construction buffer exactly full, or one byte
890 less than full, an overflow happened when the terminating "\n" was
893 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
894 when the result of a list match is failure because a DNS lookup failed.
896 PH/40 RM_COMMAND is now used in the building process.
898 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
899 the "build-* directories that it finds.
901 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
902 address in a domain literal was a prefix of an interface address.
904 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
905 when verifying a sender address, unless rewrite_headers is false.
907 PH/44 Wrote a long comment about why errors_to addresses are verified as
908 recipients, not senders.
910 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
911 the ratelimit ACL was added.
913 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
915 PH/46 Added extra information about PostgreSQL errors to the error string.
917 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
918 avoiding the need to install special zones in a real server. This is
919 backwards compatible; if it can't find the fake resolver, it drops back.
920 Thus, both old and new test suites can be run.
922 TF/02 Added util/ratelimit.pl
924 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
925 clock is set back in time.
927 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
928 Candler <B.Candler@pobox.com>.
930 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
931 OK for addresses that are the result of redirections.
933 PH/48 A number of further additions for the benefit of the new test suite,
934 including a fake gethostbyname() that interfaces to the fake DNS resolver
937 TF/06 The fix for widen_domains has also been applied to qualify_single and
938 search_parents which are the other dnslookup options that can cause
941 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
944 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
946 TF/07 Exim produced the error message "an SRV record indicated no SMTP
947 service" if it encountered an MX record with an empty target hostname.
948 The message is now "an MX or SRV record indicated no SMTP service".
950 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
951 if the DNS of the sending site is misconfigured. This is quite a
952 common situation. This change restores the behaviour of treating a
953 helo verification defer as a failure.
955 PH/51 If self=fail was set on a router, the bounce message did not include the
956 actual error message.
962 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
964 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
965 command itself ended in error, the underlying message about the transport
966 filter timeout was being overwritten with the pipe command error. Now the
967 underlying error message should be appended to the second error message.
969 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
971 PH/02 Reduce the amount of output that "make" produces by default. Full output
972 can still be requested.
974 PH/03 The warning log line about a condition test deferring for a "warn" verb
975 was being output only once per connection, rather than after each
976 occurrence (because it was using the same function as for successful
977 "warn" verbs). This seems wrong, so I have changed it.
979 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
980 it should not have, which might have caused a crash in the right
981 circumstances, but probably never did.
983 PH/04 Installed a modified version of Tony Finch's patch to make submission
984 mode fix the return path as well as the Sender: header line, and to
985 add a /name= option so that you can make the user's friendly name appear
988 TF/03 Added the control = fakedefer ACL modifier.
990 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
991 Mark Lowes for thorough testing.
993 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
995 TK/03 Merged latest SRS patch from Miles Wilton.
997 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
998 with the definition in sysexits.h (which is #included earlier).
999 Fortunately, Exim does not actually use EX_OK. The code used to try to
1000 preserve the sysexits.h value, by assumimg that macro definitions were
1001 scanned for macro replacements. I have been disabused of this notion,
1002 so now the code just undefines EX_OK before #including unistd.h.
1004 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1005 in the smtp transport. When a block could not be written in a single
1006 write() function, the timeout was being re-applied to each part-write.
1007 This seems wrong - if the receiver was accepting one byte at a time it
1008 would take for ever. The timeout is now adjusted when this happens. It
1009 doesn't have to be particularly precise.
1011 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1012 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1014 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1015 without a domain if the check to <postmaster@domain> fails.
1017 SC/01 Eximstats: added -xls and the ability to specify output files
1018 (patch written by Frank Heydlauf).
1020 SC/02 Eximstats: use FileHandles for outputing results.
1022 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1024 SC/04 Eximstats: fixed display of large numbers with -nvr option
1026 SC/05 Eximstats: fixed merging of reports with empty tables.
1028 SC/06 Eximstats: added the -include_original_destination flag
1030 SC/07 Eximstats: removed tabs and trailing whitespace.
1032 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1034 TK/06 MBOX spool code: Add real "From " MBOX separator line
1035 so the .eml file is really in mbox format (even though
1036 most programs do not really care). Patch from Alex Miller.
1038 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1039 The latter is generated from $received_to and is only set if the
1040 message has one envelope recipient. SA can use these headers,
1041 obviously out-of-the-box. Patch from Alex Miller.
1043 PH/08 The ${def test on a variable was returning false if the variable's
1044 value was "0", contrary to what the specification has always said!
1045 The result should be true unless the variable is empty.
1047 PH/09 The syntax error of a character other than { following "${if
1048 def:variable_name" (after optional whitespace) was not being diagnosed.
1049 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1050 accidental colon was present, for example, could give incorrect results.
1052 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1053 result is used (not including appendfile, where other changes are about
1056 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1057 This involved changing a lot of size variables from int to off_t. It
1058 should work with maildirs and everything.
1060 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1061 spamd dying while we are connected to it.
1063 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1066 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1067 function for BDB 4.3.
1069 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1071 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1072 the sources for PCRE have been re-arranged and more files are now
1075 PH/15 The code I had for printing potentially long long variables in PH/11
1076 above was not the best (it lost precision). The length of off_t variables
1077 is now inspected at build time, and an appropriate printing format (%ld
1078 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1079 to be "long long int" or "long int". This is needed for the internal
1080 formatting function string_vformat().
1082 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1083 the configuration file to be ":syslog", then the script "guesses" where
1084 the logs files are, rather than using the compiled in default. In our
1085 case the guess is not the same as the compiled default, so the script
1086 suddenly stopped working when I started to use syslog. The patch checks
1087 to see if log_file_path is "". If so, it attempts to read it from exim
1088 with no configuration file to get the compiled in version, before it
1089 falls back to the previous guessing code."
1091 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1092 implementing BATV in an Exim configuration. See NewStuff for the gory
1095 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1096 Makefile that are specific to HP-UX.
1098 PH/18 If the "use_postmaster" option was set for a recipient callout together
1099 with the "random" option, the postmaster address was used as the MAIL
1100 FROM address for the random test, but not for the subsequent recipient
1101 test. It is now used for both.
1103 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1104 patch removes a few documentation additions to RFC 3028, because the
1105 latest draft now contains them. It adds the new en;ascii-case comparator
1106 and a new error check for 8bit text in MIME parts. Comparator and
1107 require names are now matched exactly. I enabled the subaddress
1108 extension, but it is not well tested yet (read: it works for me)."
1110 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1111 rework some of the code of TK/09 above to avoid the hardwired use of
1112 "%lld" and "long long". Replaced the call to snprintf() with a call to
1115 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1116 records point to non-existent hosts", "retry timeout exceeded", and
1117 "retry time not reached for any host after a long failure period".
1119 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1120 experimental DomainKeys support:
1122 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1123 (2) On an error such as an illegally used "control", the wrong name for
1124 the control was given.
1126 These problems did NOT occur unless DomainKeys support was compiled.
1128 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1130 PH/24 Added ${if match_ip condition.
1132 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1133 databases so that it will be absolutely obvious if a crash occurs in the
1134 DB library. This is a regular occurrence (often caused by mis-matched
1137 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1138 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1139 picked up on a user's system that detects such things. There doesn't seem
1140 to be a gcc warning option for this - only an attribute that has to be
1141 put on the function's prototype. It seems that in Fedora Core 4 they have
1142 set this on a number of new functions. No doubt there will be more in due
1145 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1146 specify a transport. However, if an address that was verified by such a
1147 router was the subject of a callout, Exim crashed because it tried to
1148 read the rcpt_include_affixes from the non-existent transport. Now it
1149 just assumes that the setting of that option is false. This bug was
1150 introduced by 4.51/PH/31.
1152 PH/28 Changed -d+all to exclude +memory, because that information is very
1153 rarely of interest, but it makes the output a lot bigger. People tend to
1154 do -d+all out of habit.
1156 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1157 code in os-type was giving problems when libc.so lives in lib64, like on
1160 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1161 aren't the modern standard, and it seems that some systems' include files
1162 don't always have them. Exim was already checking for some of the newer
1163 ones like T_AAAA, and defining it itself. I've added checks for all the
1164 record types that Exim uses.
1166 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1167 not automatically generating a new one, as it is supposed to. This
1168 prevented TLS from working. If the file did exist, but contained invalid
1169 data, a new version was generated, as expected. It was only the case of a
1170 non-existent file that was broken.
1172 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1173 with a change in libdomainkeys > 0.64.
1175 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1176 from DNS. If the selector record carries the flag, it now has
1177 precedence over the domain-wide flag.
1179 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1181 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1182 the use of a port name that isn't defined in /etc/services) occurred, the
1183 message was deferred as in a normal delivery, and thus remained on the
1184 spool, instead of being failed because of the mua_wrapper setting. This
1185 is now fixed, and I tidied up some of the mua_wrapper messages at the
1188 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1189 the messages in a hash of arrays rather than using individual hashes.
1190 This is a bit cleaner and results in dramatic memory savings, albeit
1191 at a slight CPU cost.
1193 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1194 as requested by Marc Sherman.
1196 SC/10 Eximstats: added histograms for user specified patterns as requested
1199 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1201 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1202 fopen() in the content-scanning modules that did not already have it.
1208 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1209 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1211 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1213 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1215 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1216 to test Sieve filters that use "vacation".
1218 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1219 that changes the way the GnuTLS parameters are stored in the cache file.
1220 The new format can be generated externally. For backward compatibility,
1221 if the data in the cache doesn't make sense, Exim assumes it has read an
1222 old-format file, and it generates new data and writes a new file. This
1223 means that you can't go back to an older release without removing the
1226 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1227 work if there are any delivery delays because "one_time" forces the
1228 parent to be marked "delivered", so its unseen clone is never tried
1229 again. For this reason, Exim now forbids the simultaneous setting of
1232 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1233 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1234 entry if you want to know the details. The fix, however, neglected to
1235 consider the case when local delivery batching is involved. The test for
1236 "previously delivered" was not happening when checking to see if an
1237 address could be batched with a previous (undelivered) one; under
1238 certain circumstances this could lead to multiple deliveries to the same
1241 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1242 in its include files, and this causes problems building Exim.
1244 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1245 header_syntax) but Exim was just ignoring anything given after a slash.
1246 In particular, this caused confusion with an attempt to use "verify =
1247 reverse_host_lookup/defer_ok". An error is now given when options are
1248 supplied for verify items that do not have them. (Maybe reverse_host_
1249 lookup should have a defer_ok option, but that's a different point.)
1251 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1252 defined by RFC 821) to 2048, because there were problems with some AUTH
1253 commands, and RFC 1869 says the size should be increased for extended
1254 SMTP commands that take arguments.
1256 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1259 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1260 "unknown" error; now it says that the functionality isn't in the binary.
1262 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1263 an address' error message when a string expansion fails (syntax or
1264 whatever). Otherwise the password may appear in the log. Following change
1265 PH/42 below, there is no longer a chance of it appearing in a bounce
1268 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1270 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1271 of its addresses was ever tried. (Bugzilla bug #2.)
1273 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1274 the result incorrectly in the debug output. (It correctly added a newline
1275 to what was transported.)
1277 TF/01 Added $received_time.
1279 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1280 commented out examples of how to interface to a virus scanner and to
1281 SpamAssassin. Also added commented examples of av_scanner and
1282 spamd_address settings.
1284 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1285 and controls are allowed in which ACLs. There were a couple of minor
1286 errors. Some of the entries in the conditions table (which is a table of
1287 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1288 negation of where the condition IS allowed.
1290 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1292 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1293 header file does not have a version number, so I've had to invent a new
1294 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1295 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1296 radiusclient), but it was contributed by a Radius user.
1298 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1299 files or directories, for OpenSSL.
1301 PH/19 When an Exim process that is running as root has to create an Exim log
1302 file, it does so in a subprocess that runs as exim:exim so as to get the
1303 ownership right at creation (otherwise, other Exim processes might see
1304 the file with the wrong ownership). There was no test for failure of this
1305 fork() call, which would lead to the process getting stuck as it waited
1306 for a non-existent subprocess. Forks do occasionally fail when resources
1307 run out. I reviewed all the other calls to fork(); they all seem to check
1310 PH/20 When checking for unexpected SMTP input at connect time (before writing
1311 the banner), Exim was not dealing correctly with a non-positive return
1312 from the read() function. If the client had disconnected by this time,
1313 the result was a log entry for a synchronization error with an empty
1314 string after "input=" when read() returned zero. If read() returned -1
1315 (an event I could not check), uninitialized data bytes were printed.
1316 There were reports of junk text (parts of files, etc) appearing after
1319 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1321 PH/22 Added support for macro redefinition, and (re)definition in between
1322 driver and ACL definitions.
1324 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1325 forgetting to use the resulting value; it was using the unexpanded value.
1327 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1328 hadn't been configured. The fix is from Juergen Kreileder, who
1329 understands it better than I do:
1331 "Here's what I see happening with three configured cyrus_sasl
1332 authenticators configured (plain, login, cram-md5):
1334 On startup auth_cyrus_sasl_init() gets called for each of these.
1335 This means three calls to sasl_listmech() without a specified mech_list.
1336 => SASL tests which mechs of all available mechs actually work
1337 => three warnings about OTP not working
1338 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1340 With the patch, sasl_listmech() also gets called three times. But now
1341 SASL's mech_list option is set to the server_mech specified in the the
1342 authenticator. Or in other words, the answer from sasl_listmech()
1343 gets limited to just the mech you're testing for (which is different
1345 => the return list contains just 'plain' or 'login', 'cram-md5' or
1346 nothing depending on the value of ob->server_mech.
1348 I've just tested the patch: Authentication still works fine,
1349 unavailable mechs specified in the exim configuration are still
1350 caught, and the auth.log warnings about OTP are gone."
1352 PH/25 When debugging is enabled, the contents of the command line are added
1353 to the debugging output, even when log_selector=+arguments is not
1356 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1357 answer is "GNU", and only if the return is "GNU/something" is the answer
1360 PH/27 $acl_verify_message is now set immediately after the failure of a
1361 verification in an ACL, and so is available in subsequent modifiers. In
1362 particular, the message can be preserved by coding like this:
1364 warn !verify = sender
1365 set acl_m0 = $acl_verify_message
1367 Previously, $acl_verify_message was set only while expanding "message"
1368 and "log_message" when a very denied access.
1370 PH/28 Modified OS/os.c-Linux with
1372 -#ifndef OS_LOAD_AVERAGE
1373 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1375 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1376 nomenclature these days.)
1378 PH/29 Installed patch from the Sieve maintainer that adds the options
1379 sieve_useraddress and sieve_subaddress to the redirect router.
1381 PH/30 In these circumstances:
1382 . Two addresses routed to the same list of hosts;
1383 . First host does not offer TLS;
1384 . First host accepts first address;
1385 . First host gives temporary error to second address;
1386 . Second host offers TLS and a TLS session is established;
1387 . Second host accepts second address.
1388 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1389 and peerdn, if requested) that were in fact used only for the second
1392 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1393 attention to any local part prefix or suffix that was matched by the
1394 router that accepted the address. It now behaves in the same way as it
1395 does for delivery: the affixes are removed from the local part unless
1396 rcpt_include_affixes is set on the transport.
1398 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1399 timeout during the DATA phase of an incoming message.
1401 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1402 applied a patch sanctioned by the Sieve maintainer.
1404 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1405 the uid or gid is negative. A case of a negative gid caused this to be
1406 noticed. The fix allows for either to be negative.
1408 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1409 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1412 PH/36 The change PH/12 above was broken. Fixed it.
1414 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1415 the grounds that routing the same address twice would always produce the
1416 same answer. This might have been true once, but it is certainly no
1417 longer true now. Routing a child address may depend on the previous
1418 routing that produced that child. Some complicated redirection strategies
1419 went wrong when messages had multiple recipients, and made Exim's
1420 behaviour dependent on the order in which the addresses were given.
1422 I have moved the duplicate checking until after the routing is complete.
1423 Exim scans the addresses that are assigned to local and remote
1424 transports, and removes any duplicates. This means that more work will be
1425 done, as duplicates will always all be routed, but duplicates are
1426 presumably rare, so I don't expect this is of any significance.
1428 For deliveries to pipes, files, and autoreplies, the duplicate checking
1429 still happens during the routing process, since they are not going to be
1432 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1433 It corrects a timeout issue with spamd. This is Ian's comment: "The
1434 background is that sometimes spamd either never reads data from a
1435 connection it has accepted, or it never writes response data. The exiscan
1436 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1437 blindly assumes that writes won't block so it may never time out."
1439 PH/39 Allow G after quota size as well as K and M.
1441 PH/40 The value set for $authenticated_id in an authenticator may not contain
1442 binary zeroes or newlines because the value is written to log lines and
1443 to spool files. There was no check on this. Now the value is run through
1444 the string_printing() function so that such characters are converted to
1445 printable escape sequences.
1447 PH/41 $message_linecount is a new variable that contains the total number of
1448 lines in the message. Compare $body_linecount, which is the count for the
1451 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1452 bounce and delay warning messages, except in certain special cases, which
1455 (a) An SMTP error message from a remote host;
1456 (b) A message specified in a :fail: redirection;
1457 (c) A message specified in a "fail" command in a system filter;
1458 (d) A message specified in a FAIL return from the queryprogram router;
1459 (e) A message specified by the cannot_route_message router option.
1461 In these cases only, Exim does include the error details in bounce and
1462 warning messages. There are also a few cases where bland messages such
1463 as "unrouteable address" or "local delivery error" are given.
1465 PH/43 $value is now also set for the "else" part of a ${run expansion.
1467 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1468 being worked on, but at least Exim now implements the latest version to
1471 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1472 process to complete was treated as a delivery failure, a timeout while
1473 writing the message to the pipe was logged, but erroneously treated as a
1474 successful delivery. Such timeouts include transport filter timeouts. For
1475 consistency with the overall process timeout, these timeouts are now
1476 treated as errors, giving rise to delivery failures by default. However,
1477 there is now a new Boolean option for the pipe transport called
1478 timeout_defer, which, if set TRUE, converts the failures into defers for
1479 both kinds of timeout. A transport filter timeout is now identified in
1482 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1483 systems where "make" and "gmake" are different, calling "gmake" at top
1484 level broke things. I've arranged for the value of $(MAKE) to be passed
1485 from the Makefile to this script so that it can call the same version of
1489 A note about Exim versions 4.44 and 4.50
1490 ----------------------------------------
1492 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1493 changes of various kinds. As a consequence, a big documentation update was
1494 needed. This delayed the release for rather longer than seemed good, especially
1495 in the light of a couple of (minor) security issues. Therefore, the changes
1496 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1497 release. So 4.44 and 4.50 are in effect two different branches that both start
1500 I have left the 4.50 change log unchanged; it contains all the changes since
1501 4.43. The change log for 4.44 is below; many of its items are identical to
1502 those for 4.50. This seems to be the most sensible way to preserve the
1503 historical information.
1509 1. Minor wording change to the doc/README.SIEVE file.
1511 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1512 computation of the current number of files was incorrect.
1514 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1515 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1518 4. Give more explanation in the error message when the command for a transport
1519 filter fails to execute.
1521 5. There are several places where Exim runs a non-Exim command in a
1522 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1523 was being done only for the command run by the queryprogram router. It is
1524 now done for all such subprocesses. The other cases are: ${run, transport
1525 filters, and the commands run by the lmtp and pipe transports.
1527 6. Added CONFIGURE_GROUP build-time option.
1529 7. Some older OS have a limit of 256 on the maximum number of file
1530 descriptors. Exim was using setrlimit() to set 1000 as a large value
1531 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1532 systems. I've change it so that if it can't get 1000, it tries for 256.
1534 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1535 was an oversight, and furthermore, ever since the addition of extra
1536 controls (e.g. 4.43/32), the checks on when to allow different forms of
1537 "control" were broken. There should now be diagnostics for all cases when a
1538 control that does not make sense is encountered.
1540 9. Added the /retain_sender option to "control=submission".
1542 10. $recipients is now available in the predata ACL (oversight).
1544 11. Tidy the search cache before the fork to do a delivery from a message
1545 received from the command line. Otherwise the child will trigger a lookup
1546 failure and thereby defer the delivery if it tries to use (for example) a
1547 cached ldap connection that the parent has called unbind on.
1549 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1550 of $address_data from the recipient verification was clobbered by the
1551 sender verification.
1553 13. The value of address_data from a sender verification is now available in
1554 $sender_address_data in subsequent conditions in the ACL statement.
1556 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1558 15. Added a new option "connect=<time>" to callout options, to set a different
1561 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1562 was its contents. (It was OK if the option was not defined at all.)
1564 17. A "Completed" log line is now written for messages that are removed from
1565 the spool by the -Mrm option.
1567 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1568 information about exactly what failed.
1570 19. Added -dd to debug only the daemon process.
1572 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1573 handles timeouts, both on the server side and network timeouts. Renamed the
1574 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1576 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1577 It is now set to "smtps".
1579 22. $host_address is now set to the target address during the checking of
1580 ignore_target_hosts.
1582 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1583 being passed; this would have caused $sender_host_name to have been used if
1584 matching the list had actually called for a host name (not very likely,
1585 since this list is usually IP addresses). A host name is now passed as
1588 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1589 code that specifies a non-restarting handler (typically sigaction() in
1590 modern systems) in an attempt to fix a rare and obscure crash bug.
1592 25. Narrowed the window for a race in the daemon that could cause it to ignore
1593 SIGCHLD signals. This is not a major problem, because they are used only to
1594 wake it up if nothing else does.
1596 26. A malformed maildirsize file could cause Exim to calculate negative values
1597 for the mailbox size or file count. Odd effects could occur as a result.
1598 The maildirsize information is now recalculated if the size or filecount
1601 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1602 support for a long time. Removed HAVE_SYS_VFS_H.
1604 28. Installed the latest version of exipick from John Jetmore.
1606 29. In an address list, if the pattern was not a regular expression, an empty
1607 subject address (from a bounce message) matched only if the pattern was an
1608 empty string. Non-empty patterns were not even tested. This was the wrong
1609 because it is perfectly reasonable to use an empty address as part of a
1610 database query. An empty address is now tested by patterns that are
1611 lookups. However, all the other forms of pattern expect the subject to
1612 contain a local part and a domain, and therefore, for them, an empty
1613 address still always fails if the pattern is not itself empty.
1615 30. Exim went into a mad DNS loop when attempting to do a callout where the
1616 host was specified on an smtp transport, and looking it up yielded more
1617 than one IP address.
1619 31. Re-factored the code for checking spool and log partition space into a
1620 function that finds that data and another that does the check. The former
1621 is then used to implement four new variables: $spool_space, $log_space,
1622 $spool_inodes, and $log_inodes.
1624 32. The RFC2047 encoding function was originally intended for short strings
1625 such as real names; it was not keeping to the 75-character limit for
1626 encoded words that the RFC imposes. It now respects the limit, and
1627 generates multiple encoded words if necessary. To be on the safe side, I
1628 have increased the buffer size for the ${rfc2047: expansion operator from
1631 33. It is now permitted to omit both strings after an "if" condition; if the
1632 condition is true, the result is "true". As before, when the second string
1633 is omitted, a false condition yields an empty string. This makes it less
1634 cumbersome to write custom ACL and router conditions.
1636 34. Failure to deliver a bounce message always caused it to be frozen, even if
1637 there was an errors_to setting on the router. The errors_to setting is now
1640 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1641 canonical form (fully expanded) before being placed in
1642 $sender_host_address.
1644 36. The table in the code that translates DNS record types into text (T_A to
1645 "A" for instance) was missing entries for NS and CNAME. It is just possible
1646 that this could have caused confusion if both these types were looked up
1647 for the same domain, because the text type is used as part of Exim's
1648 per-process caching. But the chance of anyone hitting this buglet seems
1651 37. The dnsdb lookup has been extended in a number of ways.
1653 (1) There is a new type, "zns", which walks up the domain tree until it
1654 finds some nameserver records. It should be used with care.
1656 (2) There is a new type, "mxh", which is like "mx" except that it returns
1657 just the host names, not the priorities.
1659 (3) It is now possible to give a list of domains (or IP addresses) to be
1660 looked up. The behaviour when one of the lookups defers can be
1661 controlled by a keyword.
1663 (4) It is now possible to specify the separator character for use when
1664 multiple records are returned.
1666 38. The dnslists ACL condition has been extended: it is now possible to supply
1667 a list of IP addresses and/or domains to be looked up in a particular DNS
1670 39. Added log_selector=+queue_time_overall.
1672 40. When running the queue in the test harness, wait just a tad after forking a
1673 delivery process, to get repeatability of debugging output.
1675 41. Include certificate and key file names in error message when GnuTLS fails
1676 to set them up, because the GnuTLS error message doesn't include the name
1677 of the failing file when there is a problem reading it.
1679 42. Allow both -bf and -bF in the same test run.
1681 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1683 44. The "Exiscan patch" is now merged into the mainline Exim source.
1685 45. Sometimes the final signoff response after QUIT could fail to get
1686 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1687 < 0 before doing a fflush(). This bug looks as though it goes back to the
1688 introduction of TLS in release 3.20, but "sometimes" must have been rare
1689 because the tests only now provoked it.
1691 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1692 (this can affect the format of dates).
1694 47. exim_tidydb, when checking for the continued existence of a message for
1695 which it has found a message-specific retry record, was not finding
1696 messages that were in split spool directories. Consequently, it was
1697 deleting retry records that should have stayed in existence.
1699 48. Steve fixed some bugs in eximstats.
1701 49. The SPA authentication driver was not abandoning authentication and moving
1702 on to the next authenticator when an expansion was forced to fail,
1703 contradicting the general specification for all authenticators. Instead it
1704 was generating a temporary error. It now behaves as specified.
1706 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1707 (the order specifies the preference for clients). The order is now AES256,
1708 AES128, 3DES, ARCFOUR128.
1710 51. Small patch to Sieve code - explicitly set From: when generating an
1713 52. Exim crashed if a remote delivery caused a very long error message to be
1714 recorded - for instance if somebody sent an entire SpamAssassin report back
1715 as a large number of 550 error lines. This bug was coincidentally fixed by
1716 increasing the size of one of Exim's internal buffers (big_buffer) that
1717 happened as part of the Exiscan merge. However, to be on the safe side, I
1718 have made the code more robust (and fixed the comments that describe what
1721 53. Now that there can be additional text after "Completed" in log lines (if
1722 the queue_time_overall log selector is set), a one-byte patch to exigrep
1723 was needed to allow it to recognize "Completed" as not the last thing in
1726 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1727 patch that reportedly fixes this has been added. I am not expert enough to
1728 create a test for it. This is what the patch creator wrote:
1730 "I found a little strange behaviour of ldap code when working with
1731 Windows 2003 AD Domain, where users was placed in more than one
1732 Organization Units. When I tried to give exim partial DN, the exit code
1733 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1734 But simultaneously result of request was absolutely normal ldap result,
1735 so I produce this patch..."
1737 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1738 REFERENCE, so I have modified the code to exclude the patch when that macro
1741 55. Some experimental protocols are using DNS PTR records for new purposes. The
1742 keys for these records are domain names, not reversed IP addresses. The
1743 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1744 leaves it alone. Component reversal etc. now happens only for IP addresses.
1747 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1749 57. Double the size of the debug message buffer (to 2048) so that more of very
1750 long debug lines gets shown.
1752 58. The exicyclog utility now does better if the number of log files to keep
1753 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1755 59. Two changes related to the smtp_active_hostname option:
1757 (1) $smtp_active_hostname is now available as a variable.
1758 (2) The default for smtp_banner uses $smtp_active_hostname instead
1759 of $primary_hostname.
1761 60. The host_aton() function is supposed to be passed a string that is known
1762 to be a valid IP address. However, in the case of IPv6 addresses, it was
1763 not checking this. This is a hostage to fortune. Exim now panics and dies
1764 if the condition is not met. A case was found where this could be provoked
1765 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1766 components; fortuitously, this particular loophole had already been fixed
1767 by change 4.50/55 above.
1769 If there are any other similar loopholes, the new check in host_aton()
1770 itself should stop them being exploited. The report I received stated that
1771 data on the command line could provoke the exploit when Exim was running as
1772 exim, but did not say which command line option was involved. All I could
1773 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1774 running as the user.
1777 61. There was a buffer overflow vulnerability in the SPA authentication code
1778 (which came originally from the Samba project). I have added a test to the
1779 spa_base64_to_bits() function which I hope fixes it.
1782 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1783 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1785 63. The daemon start-up calls getloadavg() while still root for those OS that
1786 need the first call to be done as root, but it missed one case: when
1787 deliver_queue_load_max is set with deliver_drop_privilege. This is
1788 necessary for the benefit of the queue runner, because there is no re-exec
1789 when deliver_drop_privilege is set.
1791 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1792 This has been fixed.
1794 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1795 was in use, was not putting the data itself into the right store pool;
1796 consequently, it could be overwritten for a subsequent message in the same
1797 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1800 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1802 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1803 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1804 the function were treating the return as a boolean value, which happened to
1805 work because 0=false and not-0=true, but is not correct code.
1807 68. The host_aton() function was not handling scoped IPv6 addresses (those
1808 with, for example, "%eth0" on the end) correctly.
1810 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1811 negated items (that is, ~something) in unsigned ints. Some compilers
1812 apparently mutter when there is no cast.
1814 70. If an address verification called from an ACL failed, and did not produce a
1815 user-specific message (i.e. there was only a "system" message), nothing was
1816 put in $acl_verify_message. In this situation, it now puts the system
1819 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1820 session; change 4.31/43 added the unwanted input to the log line - except
1821 that it did not do this in the start of session case. It now does.
1823 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1824 This is wrong and can cause the other end to generate a synchronization
1825 error if it is another Exim or anything else that does the synchronization
1826 check. A QUIT command is no longer sent after a timeout.
1828 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1829 during host lookups.
1831 74. The defer_ok option of callout verification was not working if it was used
1832 when verifying addresses in header lines, that is, for this case:
1834 verify = header_sender/callout=defer_ok
1836 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1837 those file descriptors could be used for SMTP connections. If anything
1838 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1839 could be sent to the SMTP client, causing chaos. The daemon now opens
1840 stdin, stdout, and stderr to /dev/null when it puts itself into the
1843 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1844 log by default. The user can override this with suitable Perl magic.
1846 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1847 the log message when discard triggers, was not working for the DATA ACL or
1848 for the non-SMTP ACL.
1850 78. Error message wording change in sieve.c.
1852 79. If smtp_accept_max_per_host was set, the number of connections could be
1853 restricted to fewer than expected, because the daemon was trying to set up
1854 a new connection before checking whether the processes handling previous
1855 connections had finished. The check for completed processes is now done
1856 earlier. On busy systems, this bug wouldn't be noticed because something
1857 else would have woken the daemon, and it would have reaped the completed
1860 80. If a message was submitted locally by a user whose login name contained one
1861 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1862 It caused a spool format error. I have fixed the spool reading code. A
1863 related problem was that the "from" clause in the Received: line became
1864 illegal because of the space(s). It is now covered by ${quote_local_part.
1866 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1869 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1871 83. Patch from Sieve maintainer to fix unterminated string problem in
1872 "vacation" handling.
1874 84. Some minor changes to the Linux configuration files to help with other
1875 OS variants using glibc.
1877 85. One more patch for Sieve to update vacation handling to latest spec.
1880 ----------------------------------------------------
1881 See the note above about the 4.44 and 4.50 releases.
1882 ----------------------------------------------------
1888 1. Change 4.43/35 introduced a bug that caused file counts to be
1889 incorrectly computed when quota_filecount was set in an appendfile
1892 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1893 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1896 3. Give more explanation in the error message when the command for a transport
1897 filter fails to execute.
1899 4. There are several places where Exim runs a non-Exim command in a
1900 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1901 was being done only for the command run by the queryprogram router. It is
1902 now done for all such subprocesses. The other cases are: ${run, transport
1903 filters, and the commands run by the lmtp and pipe transports.
1905 5. Some older OS have a limit of 256 on the maximum number of file
1906 descriptors. Exim was using setrlimit() to set 1000 as a large value
1907 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1908 systems. I've change it so that if it can't get 1000, it tries for 256.
1910 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1911 was an oversight, and furthermore, ever since the addition of extra
1912 controls (e.g. 4.43/32), the checks on when to allow different forms of
1913 "control" were broken. There should now be diagnostics for all cases when a
1914 control that does not make sense is encountered.
1916 7. $recipients is now available in the predata ACL (oversight).
1918 8. Tidy the search cache before the fork to do a delivery from a message
1919 received from the command line. Otherwise the child will trigger a lookup
1920 failure and thereby defer the delivery if it tries to use (for example) a
1921 cached ldap connection that the parent has called unbind on.
1923 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1924 of $address_data from the recipient verification was clobbered by the
1925 sender verification.
1927 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1928 was its contents. (It was OK if the option was not defined at all.)
1930 11. A "Completed" log line is now written for messages that are removed from
1931 the spool by the -Mrm option.
1933 12. $host_address is now set to the target address during the checking of
1934 ignore_target_hosts.
1936 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1937 being passed; this would have caused $sender_host_name to have been used if
1938 matching the list had actually called for a host name (not very likely,
1939 since this list is usually IP addresses). A host name is now passed as
1942 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1943 code that specifies a non-restarting handler (typically sigaction() in
1944 modern systems) in an attempt to fix a rare and obscure crash bug.
1946 15. Narrowed the window for a race in the daemon that could cause it to ignore
1947 SIGCHLD signals. This is not a major problem, because they are used only to
1948 wake it up if nothing else does.
1950 16. A malformed maildirsize file could cause Exim to calculate negative values
1951 for the mailbox size or file count. Odd effects could occur as a result.
1952 The maildirsize information is now recalculated if the size or filecount
1955 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1956 support for a long time. Removed HAVE_SYS_VFS_H.
1958 18. Updated exipick to current release from John Jetmore.
1960 19. Allow an empty sender to be matched against a lookup in an address list.
1961 Previously the only cases considered were a regular expression, or an
1964 20. Exim went into a mad DNS lookup loop when doing a callout where the
1965 host was specified on the transport, if the DNS lookup yielded more than
1968 21. The RFC2047 encoding function was originally intended for short strings
1969 such as real names; it was not keeping to the 75-character limit for
1970 encoded words that the RFC imposes. It now respects the limit, and
1971 generates multiple encoded words if necessary. To be on the safe side, I
1972 have increased the buffer size for the ${rfc2047: expansion operator from
1975 22. Failure to deliver a bounce message always caused it to be frozen, even if
1976 there was an errors_to setting on the router. The errors_to setting is now
1979 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1980 canonical form (fully expanded) before being placed in
1981 $sender_host_address.
1983 24. Updated eximstats to version 1.33
1985 25. Include certificate and key file names in error message when GnuTLS fails
1986 to set them up, because the GnuTLS error message doesn't include the name
1987 of the failing file when there is a problem reading it.
1989 26. Expand error message when OpenSSL has problems setting up cert/key files.
1992 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1993 (this can affect the format of dates).
1995 28. exim_tidydb, when checking for the continued existence of a message for
1996 which it has found a message-specific retry record, was not finding
1997 messages that were in split spool directories. Consequently, it was
1998 deleting retry records that should have stayed in existence.
2000 29. eximstats updated to version 1.35
2001 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2002 1.35 - bugfix such that pie charts by volume are generated correctly
2004 30. The SPA authentication driver was not abandoning authentication and moving
2005 on to the next authenticator when an expansion was forced to fail,
2006 contradicting the general specification for all authenticators. Instead it
2007 was generating a temporary error. It now behaves as specified.
2009 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2010 (the order specifies the preference for clients). The order is now AES256,
2011 AES128, 3DES, ARCFOUR128.
2013 31. Small patch to Sieve code - explicitly set From: when generating an
2016 32. Exim crashed if a remote delivery caused a very long error message to be
2017 recorded - for instance if somebody sent an entire SpamAssassin report back
2018 as a large number of 550 error lines. This bug was coincidentally fixed by
2019 increasing the size of one of Exim's internal buffers (big_buffer) that
2020 happened as part of the Exiscan merge. However, to be on the safe side, I
2021 have made the code more robust (and fixed the comments that describe what
2024 33. Some experimental protocols are using DNS PTR records for new purposes. The
2025 keys for these records are domain names, not reversed IP addresses. The
2026 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2027 leaves it alone. Component reversal etc. now happens only for IP addresses.
2030 34. The host_aton() function is supposed to be passed a string that is known
2031 to be a valid IP address. However, in the case of IPv6 addresses, it was
2032 not checking this. This is a hostage to fortune. Exim now panics and dies
2033 if the condition is not met. A case was found where this could be provoked
2034 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2035 components; fortuitously, this particular loophole had already been fixed
2036 by change 4.50/55 or 4.44/33 above.
2038 If there are any other similar loopholes, the new check in host_aton()
2039 itself should stop them being exploited. The report I received stated that
2040 data on the command line could provoke the exploit when Exim was running as
2041 exim, but did not say which command line option was involved. All I could
2042 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2043 running as the user.
2046 35. There was a buffer overflow vulnerability in the SPA authentication code
2047 (which came originally from the Samba project). I have added a test to the
2048 spa_base64_to_bits() function which I hope fixes it.
2051 36. The daemon start-up calls getloadavg() while still root for those OS that
2052 need the first call to be done as root, but it missed one case: when
2053 deliver_queue_load_max is set with deliver_drop_privilege. This is
2054 necessary for the benefit of the queue runner, because there is no re-exec
2055 when deliver_drop_privilege is set.
2057 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2058 was in use, was not putting the data itself into the right store pool;
2059 consequently, it could be overwritten for a subsequent message in the same
2060 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2063 38. Sometimes the final signoff response after QUIT could fail to get
2064 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2065 < 0 before doing a fflush(). This bug looks as though it goes back to the
2066 introduction of TLS in release 3.20, but "sometimes" must have been rare
2067 because the tests only now provoked it.
2073 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2074 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2075 Now it returns an integer. A number of calls were still expecting a T/F
2076 return. Fortuitously, in all cases, the tests worked in OK situations,
2077 which is the norm. However, things would have gone wrong on any write
2078 failures on the smtp file descriptor. This function is used when sending
2079 messages over SMTP and also when doing verify callouts.
2081 2. When Exim is called to do synchronous delivery of a locally submitted
2082 message (the -odf or -odi options), it no longer closes stderr before doing
2085 3. Implemented the mua_wrapper option.
2087 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2089 5. Implemented the functions header_remove(), header_testname(),
2090 header_add_at_position(), and receive_remove_recipient(), and exported them
2093 6. If an ACL "warn" statement specified the addition of headers, Exim already
2094 inserted X-ACL-Warn: at the start if there was no header name. However, it
2095 was not making this test for the second and subsequent header lines if
2096 there were newlines in the string. This meant that an invalid header could
2097 be inserted if Exim was badly configured.
2099 7. Allow an ACL "warn" statement to add header lines at the start or after all
2100 the Received: headers, as well as at the end.
2102 8. Added the rcpt_4xx retry error code.
2104 9. Added postmaster_mailfrom=xxx to callout verification option.
2106 10. Added mailfrom=xxxx to the callout verification option, for verify=
2109 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2110 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2112 12. Inserted some casts to stop certain compilers warning when using pointer
2113 differences as field lengths or precisions in printf-type calls (mostly
2114 affecting debugging statements).
2116 13. Added optional readline() support for -be (dynamically loaded).
2118 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2119 same clock tick as a message's arrival, so that its received time was the
2120 same as the "first fail" time on the retry record, and that message
2121 remained on the queue past the ultimate address timeout, every queue runner
2122 would try a delivery (because it was past the ultimate address timeout) but
2123 after another failure, the ultimate address timeout, which should have then
2124 bounced the address, did not kick in. This was a "< instead of <=" error;
2125 in most cases the first failure would have been in the next clock tick
2126 after the received time, and all would be well.
2128 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2129 being recognized when the domain list was tested by the match_domain
2130 condition in an expansion string.
2132 16. Added the ${str2b64: operator.
2134 17. Exim was always calling setrlimit() to set a large limit for the number of
2135 processes, without checking whether the existing limit was already
2136 adequate. (It did check for the limit on file descriptors.) Furthermore,
2137 errors from getrlimit() and setrlimit() were being ignored. Now they are
2138 logged to the main and panic logs, but Exim does carry on, to try to do its
2139 job under whatever limits there are.
2141 18. Imported PCRE 5.0.
2143 19. Trivial typo in log message " temporarily refused connection" (the leading
2146 20. If the log selector return_path_on_delivery was set and an address was
2147 redirected to /dev/null, the delivery process crashed because it assumed
2148 that a return path would always be set for a "successful" delivery. In this
2149 case, the whole delivery is bypassed as an optimization, and therefore no
2152 21. Internal re-arrangement: the function for sending a challenge and reading
2153 a response while authentication was assuming a zero-terminated challenge
2154 string. It's now changed to take a pointer and a length, to allow for
2155 binary data in such strings.
2157 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2159 23. Exim was not respecting finduser_retries when seeking the login of the
2160 uid under which it was called; it was always trying 10 times. (The default
2161 setting of finduser_retries is zero.) Also, it was sleeping after the final
2162 failure, which is pointless.
2164 24. Implemented tls_on_connect_ports.
2166 25. Implemented acl_smtp_predata.
2168 26. If the domain in control=submission is set empty, Exim assumes that the
2169 authenticated id is a complete email address when it generates From: or
2170 Sender: header lines.
2172 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2173 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2174 chown and chgrp in /bin and hostname in /usr/bin.
2176 28. Exim was keeping the "process log" file open after each use, just as it
2177 does for the main log. This opens the possibility of it remaining open for
2178 long periods when the USR1 signal hits a daemon. Occasional processlog
2179 errors were reported, that could have been caused by this. Anyway, it seems
2180 much more sensible not to leave this file open at all, so that is what now
2183 29. The long-running daemon process does not normally write to the log once it
2184 has entered its main loop, and it closes the log before doing so. This is
2185 so that log files can straightforwardly be renamed and moved. However,
2186 there are a couple of unusual error situations where the daemon does write
2187 log entries, and I had neglected to close the log afterwards.
2189 30. The text of an SMTP error response that was received during a remote
2190 delivery was being truncated at 512 bytes. This is too short for some of
2191 the long messages that one sometimes sees. I've increased the limit to
2194 31. It is now possible to make retry rules that apply only when a message has a
2195 specific sender, in particular, an empty sender.
2197 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2198 it possible to be selective about when SMTP synchronization is enforced.
2200 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2202 32. Implemented hosts_connection_nolog.
2204 33. Added an ACL for QUIT.
2206 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2209 35. Added mailbox_size and mailbox_filecount to appendfile.
2211 36. Added control = no_multiline_responses to ACLs.
2213 37. There was a bug in the logic of the code that waits for the clock to tick
2214 in the case where the clock went backwards by a substantial amount such
2215 that the microsecond fraction of "now" was more than the microsecond
2216 fraction of "then" (but the whole seconds number was less).
2218 38. Added support for the libradius Radius client library this is found on
2219 FreeBSD (previously only the radiusclient library was supported).
2225 1. When certain lookups returned multiple values in the form name=value, the
2226 quoting of the values was not always being done properly. Specifically:
2227 (a) If the value started with a double quote, but contained no whitespace,
2229 (b) If the value contained whitespace other than a space character (i.e.
2230 tabs or newlines or carriage returns) it was not quoted.
2231 This fix has been applied to the mysql and pgsql lookups by writing a
2232 separate quoting function and calling it from the lookup code. The fix
2233 should probably also be applied to nisplus, ibase and oracle lookups, but
2234 since I cannot test any of those, I have not disturbed their existing code.
2236 2. A hit in the callout cache for a specific address caused a log line with no
2237 reason for rejecting RCPT. Now it says "Previous (cached) callout
2238 verification failure".
2240 3. There was an off-by-one bug in the queryprogram router. An over-long
2241 return line was truncated at 256 instead of 255 characters, thereby
2242 overflowing its buffer with the terminating zero. As well as fixing this, I
2243 have increased the buffer size to 1024 (and made a note to document this).
2245 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2246 when Exim is waiting for an SMTP response from a remote server, Exim
2247 restarts its select() call on the socket, thereby resetting its timeout.
2248 This is not a problem when such interrupts are rare. Somebody set up a cron
2249 job to run exiwhat every 2 minutes, which is less than the normal select()
2250 timeout (5 or 10 minutes). This meant that the select() timeout never
2251 kicked in because it was always reset. I have fixed this by comparing the
2252 time when an interrupt arrives with the time at the start of the first call
2253 to select(). If more time than the timeout has elapsed, the interrupt is
2254 treated as a timeout.
2256 5. Some internal re-factoring in preparation for the addition of Sieve
2257 extensions (by MH). In particular, the "personal" test is moved to a
2258 separate function, and given an option for scanning Cc: and Bcc: (which is
2259 not set for Exim filters).
2261 6. When Exim created an email address using the login of the caller as the
2262 local part (e.g. when creating a From: or Sender: header line), it was not
2263 quoting the local part when it contained special characters such as @.
2265 7. Installed new OpenBSD configuration files.
2267 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2268 try to make them clearer.
2270 9. Callout options, other than the timeout value, were being ignored when
2271 verifying sender addresses in header lines. For example, when using
2273 verify = header_sender/callout=no_cache
2275 the cache was (incorrectly) being used.
2277 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2278 only the Cygwin environment.
2280 11. When return_path_on_delivery was set as a log selector, if different remote
2281 addresses in the same message used different return paths and parallel
2282 remote delivery occurred, the wrong values would sometimes be logged.
2283 (Whenever a remote delivery process finished, the return path value from
2284 the most recently started remote delivery process was logged.)
2286 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2287 lines when AUTH and/or TLS are in use. This is the "received protocol"
2288 field. Exim used to use "asmtp" for authenticated SMTP, without any
2289 indication (in the protocol name) for TLS use. Now it follows the RFC and
2290 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2291 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2292 names appear in log lines as well as in Received: header lines.
2294 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2295 extensions, and comparison tests, and to fix some bugs.
2297 14. Changes to the "personal" filter test:
2299 (1) The test was buggy in that it was just doing the equivalent of
2300 "contains" tests on header lines. For example, if a user's address was
2301 anne@some.where, the "personal" test would incorrectly be true for
2303 To: susanne@some.where
2305 This test is now done by extracting each address from the header in turn,
2306 and checking the entire address. Other tests that are part of "personal"
2307 are now done using regular expressions (for example, to check local parts
2308 of addresses in From: header lines).
2310 (2) The list of non-personal local parts in From: addresses has been
2311 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2312 taken from the Sieve specification recommendations.
2314 (3) If the message contains any header line starting with "List-" it is
2315 treated as non-personal.
2317 (4) The test for "circular" in the Subject: header line has been removed
2318 because it now seems ill-conceived.
2320 15. Minor typos in src/EDITME comments corrected.
2322 16. Installed latest exipick from John Jetmore.
2324 17. If headers_add on a router specified a text string that was too long for
2325 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2326 of string_sprintf() is now avoided.
2328 18. $message_body_size was not set (it was always zero) when running the DATA
2329 ACL and the local_scan() function.
2331 19. For the "mail" command in an Exim filter, no default was being set for
2332 the once_repeat time, causing a random time value to be used if "once" was
2333 specified. (If the value happened to be <= 0, no repeat happened.) The
2334 default is now 0s, meaning "never repeat". The "vacation" command was OK
2335 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2336 (I found it when inspecting the code).
2338 20. There is now an overall timeout for performing a callout verification. It
2339 defaults to 4 times the callout timeout, which applies to individual SMTP
2340 commands during the callout. The overall timeout applies when there is more
2341 than one host that can be tried. The timeout is checked before trying the
2342 next host. This prevents very long delays if there are a large number of
2343 hosts and all are timing out (e.g. when the network connections are timing
2344 out). The value of the overall timeout can be changed by specifying an
2345 additional sub-option for "callout", called "maxwait". For example:
2347 verify = sender/callout=5s,maxwait=20s
2349 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2350 to the end before writing, but this should make it even safer).
2352 22. Exim was forgetting that it had advertised PIPELINING for the second and
2353 subsequent messages on an SMTP connection. It was also not resetting its
2354 memory on STARTTLS and an internal HELO.
2356 23. When Exim logs an SMTP synchronization error within a session, it now
2357 records whether PIPELINING has been advertised or not.
2359 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2360 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2361 rather than long int.
2363 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2365 26. Added the never_mail option to autoreply.
2371 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2372 crash if the getsockname() call failed; this can happen if a connection is
2373 closed very soon after it is established. The problem was simply in the
2374 order in which certain operations were done, causing Exim to try to write
2375 to the SMTP stream before it had set up the file descriptor. The bug has
2376 been fixed by making things happen in the correct order.
2382 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2383 before the connection was closed, thus losing the rejection response.
2385 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2386 some early Solaris releases, but causes trouble in current releases where
2387 socklen_t is defined.
2389 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2392 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2395 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2396 that releasing the top of it at the end releases what was used for sub-
2397 expansions (unless the block got too big). However, discard this block if
2398 the first thing is a variable or header, so that we can use its block when
2399 it is dynamic (useful for very large $message_headers, for example).
2401 6. Lookups now cache *every* query, not just the most recent. A new, separate
2402 store pool is used for this. It can be recovered when all lookup caches are
2403 flushed. Lookups now release memory at the end of their result strings.
2404 This has involved some general refactoring of the lookup sources.
2406 7. Some code has been added to the store_xxx() functions to reduce the amount
2407 of flapping under certain conditions.
2409 8. log_incoming_interface used to affect only the <= reception log lines. Now
2410 it causes the local interface and port to be added to several more SMTP log
2411 lines, for example "SMTP connection from", and rejection lines.
2413 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2415 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2417 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2418 could be overwritten at the end of the current message (or the start of a
2419 new message if it was set in a HELO ACL). The value is now preserved for
2420 the duration of the SMTP connection.
2422 12. If a transport had a headers_rewrite setting, and a matching header line
2423 contained an unqualified address, that address was qualified, even if it
2424 did not match any rewriting rules. The underlying bug was that the values
2425 of the flags that permit the existence of unqualified sender and recipient
2426 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2427 non-local messages, and by -bnq for local messages) were not being
2428 preserved with the message after it was received.
2430 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2431 "next input=" as part of the text comprising the host identity instead of
2432 the correct text. The code was using the same buffer for two different
2433 strings. However, depending on which order the printing function evaluated
2434 its arguments, the bug did not always show up. Under Linux, for example, my
2435 test suite worked just fine.
2437 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2438 doesn't work with some older versions of Perl. It has been changed to "my",
2439 which in any case is probably the better facility to use.
2441 15. A really picky compiler found some instances of statements for creating
2442 error messages that either had too many or two few arguments for the format
2445 16. The size of the buffer for calls to the DNS resolver has been increased
2446 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2447 for addresses that have a lot of PTR records. This alleviates a problem; it
2448 does not fully solve it.
2450 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2451 buffer now truncates the list and logs the incident, which is the same
2452 action as happens when Exim is looking up a host name and its aliases.
2453 Previously in this situation something unpredictable would happen;
2454 sometimes it was "internal error: store_reset failed".
2456 18. If a server dropped the connection unexpectedly when an Exim client was
2457 using GnuTLS and trying to read a response, the client delivery process
2458 crashed while trying to generate an error log message.
2460 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2461 string, for example:
2463 warn message = H1: something\nH2: something
2465 the text was added as a single header line from Exim's point of view
2466 though it ended up OK in the delivered message. However, searching for the
2467 second and subsequent header lines using $h_h2: did not work. This has been
2468 fixed. Similarly, if a system filter added multiple headers in this way,
2469 the routers could not see them.
2471 20. Expanded the error message when iplsearch is called with an invalid key to
2472 suggest using net-iplsearch in a host list.
2474 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2475 ACLs are no longer actually imposed (and a message to that effect is
2478 22. If a "gecos" field in a passwd entry contained escaped characters, in
2479 particular, if it contained a \" sequence, Exim got it wrong when building
2480 a From: or a Sender: header from that name. A second bug also caused
2481 incorrect handling when an unquoted " was present following a character
2482 that needed quoting.
2484 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2485 was not being matched caselessly.
2487 24. Arranged for all hyphens in the exim.8 source to be escaped with
2490 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2491 properly. Recipient callout cache records were still being keyed to include
2492 the sender, even when use_sender was set false. This led to far more
2493 callouts that were necessary. The sender is no longer included in the key
2494 when use_sender is false.
2496 26. Added "control = submission" modifier to ACLs.
2498 27. Added the ${base62d: operator to decode base 62 numbers.
2500 28. dnsdb lookups can now access SRV records.
2502 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2503 the configuration file.
2505 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2506 (-v) mode. This makes the output for a verbose queue run more intelligible.
2508 31. Added a use_postmaster feature to recipient callouts.
2510 32. Added the $body_zerocount variable, containing the number of binary zero
2511 bytes in the message body.
2513 33. The time of last modification of the "new" subdirectory is now used as the
2514 "mailbox time last read" when there is a quota error for a maildir
2517 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2519 35. Added +ignore_unknown as a special item in host lists.
2521 36. Code for decoding IPv6 addresses in host lists is now included, even if
2522 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2523 address was recognized as an IP address, but was then not correctly decoded
2524 into binary, causing unexpected and incorrect effects when compared with
2531 1. Very minor rewording of debugging text in manualroute to say "list of
2532 hosts" instead of "hostlist".
2534 2. If verify=header_syntax was set, and a header line with an unqualified
2535 address (no domain) and a large number of spaces between the end of the
2536 name and the colon was received, the reception process suffered a buffer
2537 overflow, and (when I tested it) crashed. This was caused by some obsolete
2538 code that should have been removed. The fix is to remove it!
2540 3. When running in the test harness, delay a bit after writing a bounce
2541 message to get a bit more predictability in the log output.
2543 4. Added a call to search_tidyup() just before forking a reception process. In
2544 theory, someone could use a lookup in the expansion of smtp_accept_max_
2545 per_host which, without the tidyup, could leave open a database connection.
2547 5. Added the variables $recipient_data and $sender_data which get set from a
2548 lookup success in an ACL "recipients" or "senders" condition, or a router
2549 "senders" option, similar to $domain_data and $local_part_data.
2551 6. Moved the writing of debug_print from before to after the "senders" test
2554 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2555 problems for message scanning, either using a data ACL, or using
2556 local_scan() because the Received: header was not generated till after they
2557 were called (in order to set the time as the time of reception completion).
2558 I have revised the way this works. The header is now generated after the
2559 body is received, but before the ACL or local_scan() are called. After they
2560 are run, the timestamp in the header is updated.
2566 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2567 before starting a queue runner without re-exec. This happened only when
2568 deliver_drop_privilege was set or when the Exim user was set to root. The
2569 effect of the bug was that timeouts during subsequent deliveries caused
2570 crashes instead of being properly handled. The handler is now left at its
2571 default (and expected) setting.
2573 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2574 message, again when deliver_drop_privilege is set or Exim is run as root.
2575 The bug described in (1) was not present in this case, but the tidying up
2576 of the other signals was missing. I have made the two cases consistent.
2578 3. The ignore_target_hosts setting on a manualroute router was being ignored
2579 for hosts that were looked up using the /MX notation.
2581 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2584 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2585 operated on the sender address. After changing the $sender_address to <>
2586 for the sender address verify, Exim was re-instated it as the original
2587 (before rewriting) address, but remembering that it had rewritten it, so it
2588 wasn't rewriting it again. This bug also had the effect of breaking the
2589 sender address verification caching when the sender address was rewritten.
2591 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2592 This has been changed so that if the ip literal address matches
2593 ignore_target_hosts, the router declines.
2595 7. Added expansion conditions match_domain, match_address, and match_local_
2596 part (NOT match_host).
2598 8. The placeholder for the Received: header didn't have a length field set.
2600 9. Added code to Exim itself and to exim_lock to test for a specific race
2601 condition that could lead to file corruption when using MBX delivery. The
2602 issue is with the lockfile that is created in /tmp. If this file is removed
2603 after a process has opened it but before that process has acquired a lock,
2604 there is the potential for a second process to recreate the file and also
2605 acquire a lock. This could lead to two Exim processes writing to the file
2606 at the same time. The added code performs the same test as UW imapd; it
2607 checks after acquiring the lock that its file descriptor still refers to
2608 the same named file.
2610 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2611 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2614 11. Added the smtp_active_hostname option. If used, this will typically be made
2615 to depend on the incoming interface address. Because $interface_address is
2616 not set up until the daemon has forked a reception process, error responses
2617 that can happen earlier (such as "too many connections") no longer contain
2620 12. If an expansion in a condition on a "warn" statement fails because a lookup
2621 defers, the "warn" statement is abandoned, and the next ACL statement is
2622 processed. Previously this caused the whole ACL to be aborted.
2624 13. Added the iplsearch lookup type.
2626 14. Added ident_timeout as a log selector.
2628 15. Added tls_certificate_verified as a log selector.
2630 16. Added a global option tls_require_ciphers (compare the smtp transport
2631 option of the same name). This controls incoming TLS connections.
2633 17. I finally figured out how to make tls_require_ciphers do a similar thing
2634 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2635 before starting the TLS session.
2637 18. Tabs are now shown as \t in -bP output.
2639 19. If the log selector return_path_on_delivery was set, Exim crashed when
2640 bouncing a message because it had too many Received: header lines.
2642 20. If two routers both had headers_remove settings, and the first one included
2643 a superfluous trailing colon, the final name in the first list and the
2644 first name in the second list were incorrectly joined into one item (with a
2645 colon in the middle).
2651 1. Added -C and -D options to the exinext utility, mainly to make it easier
2652 to include in the automated testing, but these could be helpful when
2653 multiple configurations are in use.
2655 2. The exinext utility was not formatting the output nicely when there was
2656 an alternate port involved in the retry record key, nor when there was a
2657 message id as well (for retries that were specific to a specific message
2658 and a specific host). It was also confused by IPv6 addresses, because of
2659 the additional colons they contain. I have fixed the IPv4 problem, and
2660 patched it up to do a reasonable job for IPv6.
2662 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2663 delivery, the log line now contains "pipelined" if PIPELINING was used.
2665 4. An SMTP transport process used to panic and die if the bind() call to set
2666 an explicit outgoing interface failed. This has been changed; it is now
2667 treated in the same way as a connect() failure.
2669 5. A reference to $sender_host_name in the part of a conditional expansion
2670 that was being skipped was still causing a DNS lookup. This no longer
2673 6. The def: expansion condition was not recognizing references to header lines
2674 that used bh_ and bheader_.
2676 7. Added the _cache feature to named lists.
2678 8. The code for checking quota_filecount in the appendfile transport was
2679 allowing one more file than it should have been.
2681 9. For compatibility with Sendmail, the command line option
2689 and sets the incoming protocol and host name (for trusted callers). The
2690 host name and its colon can be omitted when only the protocol is to be set.
2691 Note the Exim already has two private options, -pd and -ps, that refer to
2692 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2693 "s", but I don't think that's a major issue.
2695 10. A number of refactoring changes to the code, none of which should affect
2698 (a) The number of logging options was getting close to filling up the
2699 32-bit word that was used as a bit map. I have split them into two classes:
2700 those that are passed in the argument to log_write(), and those that are
2701 only ever tested independently outside of that function. These are now in
2702 separate 32-bit words, so there is plenty of room for expansion again.
2703 There is no change in the user interface or the logging behaviour.
2705 (b) When building, for example, log lines, the code previously used a
2706 macro that called string_cat() twice, in order to add two strings. This is
2707 not really sufficiently general. Furthermore, there was one instance where
2708 it was actually wrong because one of the argument was used twice, and in
2709 one call a function was used. (As it happened, calling the function twice
2710 did not affect the overall behaviour.) The macro has been replaced by a
2711 function that can join an arbitrary number of extra strings onto a growing
2714 (c) The code for expansion conditions now uses a table and a binary chop
2715 instead of a serial search (which was left over from when there were very
2716 few conditions). Also, it now recognizes conditions like "pam" even when
2717 the relevant support is not compiled in: a suitably worded error message is
2718 given if an attempt is made to use such a condition.
2720 11. Added ${time_interval:xxxxx}.
2722 12. A bug was causing one of the ddress fields not to be passed back correctly
2723 from remote delivery subprocesses. The field in question was not being
2724 subsequently used, so this caused to problems in practice.
2726 13. Added new log selectors queue_time and deliver_time.
2728 14. Might have fixed a bug in maildirsizefile handling that threw up
2729 "unexpected character" debug warnings, and recalculated the data
2730 unnecessarily. In any case, I expanded the warning message to give more
2733 15. Added the message "Restricted characters in address" to the statements in
2734 the default ACL that block characters like @ and % in local parts.
2736 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2737 Three changes have been made:
2739 (a) There was a serious bug; a negative response to MAIL caused the whole
2740 recipient domain to be cached as invalid, thereby blocking all messages
2741 to all local parts at the same domain, from all senders. This bug has
2742 been fixed. The domain is no longer cached after a negative response to
2743 MAIL if the sender used is not empty.
2745 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2748 (c) A new callout option, "use_sender" has been added for people who want
2749 the modified behaviour.
2755 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2758 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2759 indeed breaks things for older releases.
2761 3. Added additional logging to the case where there is a problem reading data
2762 from a filter that is running in a subprocess using a pipe, in order to
2763 try to track down a specific problem.
2765 4. Testing facility fudge: when running in the test harness and attempting
2766 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2767 getting "No route to host". Convert this to a timeout.
2769 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2772 6. Some OS don't have socklen_t but use size_t instead. This affects the
2773 fifth argument of getsockopt() amongst other things. This is now
2774 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2775 can be set for individual OS. I have set it for SunOS5, OSF1, and
2776 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2777 some earlier ones do not.
2779 7. Change 4.30/15 was not doing the test caselessly.
2781 8. The standard form for an IPv6 address literal was being rejected by address
2782 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2783 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2784 this, as well as the form without the "IPv6" on the front (but only when
2785 address literals are enabled, of course).
2787 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2789 10. Exim crashed if a message with an empty sender address specified by -f
2790 encountered a router with an errors_to setting. This could be provoked only
2791 by a command such as
2795 where an empty string was supplied; "<>" did not hit this bug.
2797 11. Installed PCRE release 4.5.
2799 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2800 remained set. It is now erased.
2802 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2803 times from message ids (which are base 36 rather than the normal 62).
2805 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2806 were being counted as actual protocol errors, and logged if the log
2807 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2808 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2809 and DATA following a set of rejected RCPTs do not count as protocol errors.
2810 In other words, Exim assumes they were pipelined, though this may not
2811 actually be the case. Of course, in all cases the client gets an
2812 appropriate error code.
2814 15. If a lookup fails in an ACL condition, a message about the failure may
2815 be available; it is used if testing the ACL cannot continue, because most
2816 such messages specify what the cause of the deferral is. However, some
2817 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2818 that caused an old message to be retained and used if a later statement
2819 caused a defer, replacing the real cause of the deferral.
2821 16. If an IP address had so many PTR records that the DNS lookup buffer
2822 was not large enough to hold them, Exim could crash while trying to process
2823 the truncated data. It now detects and logs this case.
2825 17. Further to 4.21/58, another change has been made: if (and only if) the
2826 first line of a message (the first header line) ends with CRLF, a bare LF
2827 in a subsequent header line has a space inserted after it, so as not to
2828 terminate the header.
2830 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2831 unnecessarily, used atoi() instead of strtol(), and didn't check the
2832 termination when getting file sizes from file names by regex.
2834 19. Completely re-implemented the support for maildirsize files, in the light
2835 of a number of problems with the previous contributed implementation
2836 (4.30/29). In particular:
2838 . If the quota is zero, the maildirsize file is maintained, but no quota is
2841 . If the maildir directory does not exist, it is created before any attempt
2842 to write a maildirsize file.
2844 . The quota value in the file is just a cache; if the quota is changed in
2845 the transport, the new value overrides.
2847 . A regular expression is available for excluding directories from the
2850 20. The autoreply transport checks the characters in options that define the
2851 message's headers; it allows continued headers, but it was checking with
2852 isspace() after an embedded newline instead of explicitly looking for a
2855 21. If all the "regular" hosts to which an address was routed had passed their
2856 expiry times, and had not reached their retry times, the address was
2857 bounced, even if fallback hosts were defined. Now Exim should go on to try
2860 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2861 equivalent code in the SMTP transport. Some hosts send humungous responses
2862 to HELO/EHLO, more than 1024 it seems.
2864 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2865 gives compiler warnings in some environments. I've now done it "properly",
2868 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2869 (because of problems with the built-in one) was declared to return uschar *
2870 instead of char *, causing compiler failure.
2872 25. Fixed a file descriptor leak when processing alias/forward files.
2874 26. Fixed a minor format string issue in dbfn.c.
2876 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2878 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2879 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2882 29. When a sender address is verified, it is cached, to save repeating the test
2883 when there is more than one recipient in a message. However, when the
2884 verification involves a callout, it is possible for different callout
2885 options to be set for different recipients. It is too complicated to keep
2886 track of this in the cache, so now Exim always runs a verification when a
2887 callout is required, relying on the callout cache for the optimization.
2888 The overhead is duplication of the address routing, but this should not be
2891 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2892 to be verified with callout=postmaster, and the main callout worked but the
2893 postmaster check failed, the verification correctly failed. However, if a
2894 subsequent RCPT command asked for sender verification *without* the
2895 postmaster check, incorrect caching caused this verification also to fail,
2898 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2899 it was not caching the DNS options (qualify_single, search_parents) that
2900 were used when the lookup failed. A subsequent lookup with different
2901 options therefore always gave the same answer, though there were cases
2902 where it should not have. (Example: a "domains = !$mx_any" option on a
2903 dnslookup router: the "domains" option is always processed without any
2904 widening, but the router might have qualify_single set.) Now Exim uses the
2905 cached value only when the same options are set.
2907 32. Added John Jetmore's "exipick" utility to the distribution.
2909 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2910 than a timeout (e.g. a certificate is required, and is not provided), an
2911 Exim server now closes the connection immediately. Previously it waited for
2912 the client to close - but if the client is SSL, it seems that they each
2913 wait for each other, leading to a delay before one of them times out.
2915 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2916 maintained 0.8.x compatibility because I don't think many are using it, and
2917 it is clearly obsolete.
2919 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2922 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2923 client certificate was expired. A simple patch fixes this, though I don't
2924 understand the full logic of why the verify callback is called multiple
2927 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2928 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2929 which causes problems with some clients (such as the Certicom SSL Plus
2930 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2931 disables the coutermeasure allowing Eudora to connect."
2933 38. Exim was not checking that a write() to a log file succeeded. This could
2934 lead to Bad Things if a log got too big, in particular if it hit a file
2935 size limit. Exim now panics and dies if it cannot write to a log file, just
2936 as it does if it cannot open a log file.
2938 39. Modified OS/Makefile-Linux so that it now contains
2940 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2942 The two -D definitions ensure that Exim is compiled with large file
2943 support, which makes it possible to handle log files that are bigger than
2946 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2947 instance) a domain was checked against a named list that involved a lookup,
2948 causing $domain_data to be set, then another domain was checked against the
2949 same list, then the first domain was re-checked, the value of $domain_data
2950 after the final check could be wrong. In particular, if the second check
2951 failed, it could be set empty. This bug probably also applied to
2954 41. The strip_trailing_dot option was not being applied to the address given
2955 with the -f command-line option.
2957 42. The code for reading a message's header from the spool was incrementing
2958 $received_count, but never initializing it. This meant that the value was
2959 incorrect (doubled) while delivering a message in the same process in which
2960 it was received. In the most common configuration of Exim, this never
2961 happens - a fresh exec is done - but it can happen when
2962 deliver_drop_privilege is set.
2964 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2965 it now includes up to 150 characters of the unexpected data in the log
2968 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2969 and building data strings. The size of both of these buffers was 10 000
2970 bytes - far larger than anybody would *ever* want, thought I. Needless to
2971 say, somebody hit the limit. I have increased the maximum line length to
2972 20 000 and the maximum data length of concatenated lines to 100 000. I have
2973 also fixed two bugs, because there was no checking on these buffers. Tsk,
2974 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2975 buffer is too small.
2977 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2978 lsearch lookups. Now it does.
2980 46. When parsing a route_list item in a manualroute router, a fixed-length
2981 buffer was used for the list of hosts. I made this 1024 bytes long,
2982 thinking that nobody would ever have a list of hosts that long. Wrong.
2983 Somebody had a whole pile of complicated expansion conditions, and the
2984 string was silently truncated, leading to an expansion error. It turns out
2985 that it is easier to change to an unlimited length (owing to other changes
2986 that have happened since this code was originally written) than to build
2987 structure for giving a limitation error. The length of the item that
2988 expands into the list of hosts is now unlimited.
2990 47. The lsearch lookup could not handle data where the length of text line was
2991 more than 4095 characters. Such lines were truncated, leading to shortened
2992 data being returned. It should now handle lines of any length.
2994 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2995 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2998 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2999 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3000 no longer happens in comment lines. A list of replacements is now placed
3001 at the head of all of the source files, except those whose only change is
3002 to replace PERL_COMMAND in the very first #! line.
3004 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3005 messages on the queue, with a bottom-up merge sort, using code contributed
3006 by Michael Haardt. This should make operations like -bp somewhat faster on
3007 large queues. It won't affect queue runners, except when queue_run_in_order
3010 51. Installed eximstats 1.31 in the distribution.
3012 52. Added support for SRV lookups to the dnslookup router.
3014 53. If an ACL referred to $message_body or $message_body_end, the value was not
3015 reset for any messages that followed in the same SMTP session.
3017 54. The store-handling optimization for building very long strings was not
3018 differentiating between the different store pools. I don't think this
3019 actually made any difference in practice, but I've tidied it.
3021 55. While running the routers to verify a sender address, $sender_address
3022 was still set to the sender address. This is wrong, because when routing to
3023 send a bounce to the sender, it would be empty. Therefore, I have changed
3024 it so that, while verifying a sender address, $sender_address is set to <>.
3025 (There is no change to what happens when verifying a recipient address.)
3027 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3028 target A or AAAA records (if not already returned) without resetting the
3029 qualify_single or search_parents options of the DNS resolver. These are
3030 inappropriate in this case because the targets of MX and SRV records must
3031 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3032 target that, when qualified, matched something in the local domain. These
3033 two options are now turned off when doing these lookups.
3035 57. It seems that at least some releases of Reiserfs (which does not have the
3036 concept of a fixed number of inodes) returns zero and not -1 for the
3037 number of available inodes. This interacted badly with check_spool_inodes,
3038 which assumed that -1 was the "no such thing" setting. What I have done is
3039 to check that the total number of inodes is greater than zero before doing
3040 the test of how many are available.
3042 58. When a "warn" ACL statement has a log_message modifier, the message is
3043 remembered, and not repeated. This is to avoid a lot of repetition when a
3044 message has many recipients that cause the same warning to be written.
3045 Howewer, Exim was preserving the list of already written lines for an
3046 entire SMTP session, which doesn't seem right. The memory is now reset if a
3047 new message is started.
3049 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3050 debugging output unless log_rewrite was also set.
3052 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3053 of (int)(handle) when we know that handle contains (void *)(-1).
3055 61. The Exim daemon panic-logs an error return when it closes the incoming
3056 connection. However "connection reset by peer" seems to be common, and
3057 isn't really an error worthy of noting specially, so that particular error
3060 62. When Exim is trying to find all the local interfaces, it used to panic and
3061 die if the ioctl to get the interface flags failed. However, it seems that
3062 on at least one OS (Solaris 9) it is possible to have an interface that is
3063 included in the list of interfaces, but for which you get a failure error
3064 for this call. This happens when the interface is not "plumbed" into a
3065 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3066 failure of the "get flags" call assumes that the interface is down.
3068 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3069 makes life easier for people who are doing arithmetic on fields extracted
3070 from dates, where you often get leading zeros that should not be
3071 interpreted as octal.
3073 64. Added qualify_domain to the redirect router, to override the global
3076 65. If a pathologically long header line contained very many addresses (the
3077 report of this problem mentioned 10 000) and each of them was rewritten,
3078 Exim could use up a very large amount of memory. (It kept on making new
3079 copies of the header line as it rewrote, and never released the old ones.)
3080 At the expense of a bit more processing, the header rewriting function has
3081 been changed so that it no longer eats memory in this way.
3083 66. The generation of the Received: header has been moved from the time that a
3084 message starts to be received, to the time that it finishes. The timestamp
3085 in the Received: header should now be very close to that of the <= log
3086 line. There are two side-effects of this change:
3088 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3089 logged header lines no longer include the local Received: line, because
3090 it has not yet been created. The same applies to a copy of the message
3091 that is returned to a non-SMTP sender when a message is rejected.
3093 (b) When a filter file is tested using -bf, no additional Received: header
3094 is added to the test message. After some thought, I decided that this
3097 This change does not affect the value of $received_for. It is still set
3098 after address rewriting, but before local_scan() is called.
3100 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3102 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3103 gave an unhelpful panic error message, and a defer error. I have managed to
3104 change this behaviour so that it now rejects any supplied certificate,
3105 which seems right, as the list of acceptable certificates is empty.
3107 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3108 gave an unhelpful defer error. I have not managed to make this reject any
3109 supplied certificates, but the error message it gives is "no certificate
3110 supplied", which is not helpful.
3112 70. exigrep's output now also includes lines that are not associated with any
3113 message, but which match the given pattern. Implemented by a patch from
3114 Martin Sluka, which also tidied up the Perl a bit.
3116 71. Recipient callout verification, like sender verification, was using <> in
3117 the MAIL FROM command. This isn't really the right thing, since the actual
3118 sender may affect whether the remote host accepts the recipient or not. I
3119 have changed it to use the actual sender in the callout; this means that
3120 the cache record is now keyed on a recipient/sender pair, not just the
3121 recipient address. There doesn't seem to be a real danger of callout loops,
3122 since a callout by the remote host to check the sender would use <>.
3123 [SEE ABOVE: changed after hitting problems.]
3125 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3126 temporary errors. However, in the case of such a code being given after
3127 the end of a data transmission (i.e. after ".") Exim was failing to write
3128 a retry record for the message. (Yes, there was some broken host that was
3129 actually sending 8xx at this point.)
3131 73. An unknown lookup type in a host list could cause Exim to panic-die when
3132 the list was checked. (An example that provoked this was putting <; in the
3133 middle of a list instead of at the start.) If this happened during a DATA
3134 ACL check, a -D file could be left lying around. This kind of configuration
3135 error no longer causes Exim to die; instead it causes a defer errror. The
3136 incident is still logged to the main and panic logs.
3138 74. Buglet left over from Exim 3 conversion. The message "too many messages
3139 in one connection" was written to the rejectlog but not the mainlog, except
3140 when address rewriting (yes!) was being logged.
3142 75. Added write_rejectlog option.
3144 76. When a system filter was run not as root (that is, when system_filter_user
3145 was set), the values of the $n variables were not being returned to the
3146 main process; thus, they were not subsequently available in the $sn
3149 77. Added +return_path_on_delivery log selector.
3151 78. A connection timeout was being treated differently from recipients deferred
3152 when testing hosts_max_try with a message that was older than the host's
3153 retry timeout. (The host should not be counted, thus allowing all hosts to
3154 be tried at least once before bouncing.) This may have been the cause of an
3155 occasionally reported bug whereby a message would remain on the queue
3156 longer than the retry timeout, but would be bounced if a delivery was
3157 forced. I say "may" because I never totally pinned down the problem;
3158 setting up timeout/retry tests is difficult. See also the next item.
3160 79. The ultimate address timeout was not being applied to errors that involved
3161 a combination of host plus message (for example, a timeout on a MAIL
3162 command). When an address resolved to a number of possible hosts, and they
3163 were not all tried for each delivery (e.g. because of hosts_max_try), a
3164 message could remain on the queue longer than the retry timeout.
3166 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3169 81. Fixed an obscure SMTP outgoing bug which required at least the following
3170 conditions: (a) there was another message waiting for the same server;
3171 (b) the server returned 5xx to all RCPT commands in the first message so
3172 that the message was not completed; (c) the server dropped the connection
3173 or gave a negative response to the RSET that Exim sends to abort the
3174 transaction. The observed case was a dropped connection after DATA that had
3175 been sent in pipelining mode. That is, the server had advertised PIPELINING
3176 but was not implementing it correctly. The effect of the bug was incorrect
3177 behaviour, such as trying another host, and this could lead to a crash.
3183 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3184 and daemon.c were passed as pointers to ints; they should have been
3185 pointers to socklen_t variables (which are typically unsigned ints).
3187 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3190 3. Fixed a really odd bug that affected only the testing scheme; patching a
3191 certain fixed string in the binary changed the value of another string that
3192 happened to be identical to the end of the original first string.
3194 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3195 name", it returns that address as the IP address. On some operating
3196 systems (e.g. Solaris), it also passes back the IP address string as the
3197 "host name". However, on others (e.g. Linux), it passes back an empty
3198 string. Exim wasn't checking for this, and was changing the host name to an
3199 empty string, assuming it had been canonicized.
3201 5. Although rare, it is permitted to have more than one PTR record for a given
3202 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3203 all the names associated with an address, because they do in Solaris.
3204 However, it seems that they do not in Linux for data that comes from the
3205 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3206 I found this out when I moved to a new Linux workstation and tried to run
3207 the Exim test suite.
3209 To get round this problem I have changed the code so that it now does its
3210 own call to the DNS to look up PTR records when searching for a host name.
3211 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3212 addresses that are only in /etc/hosts are still found.
3214 This behaviour is, however, controlled by an option called host_lookup_
3215 order, which defaults to "bydns:byaddr". If people want to use the other
3216 order, or indeed, just use one or the other means of lookup, they can
3217 specify it in this variable.
3219 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3220 some operating systems, this comes back from gethostbyaddr() as an empty
3221 string, and this is what Exim used to test for. However, it seems that in
3222 other systems, "." is yielded. Exim now tests for this case too.
3224 7. The values of check_spool_space and check_log_space are now held internally
3225 as a number of kilobytes instead of an absolute number of bytes. If a
3226 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3227 kilobyte. This means that much larger values can be stored.
3229 8. Exim monitor: an attempt to get the action menu when not actually pointing
3230 at a message produces an empty menu entitled "No message selected". This
3231 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3232 no entries in it ("Shell widget menu has zero width and/or height"). So I
3233 have added a single, blank menu entry in this case.
3235 9. Added ${quote_local_part.
3237 10. MIME decoding is now applied to the contents of Subject: header lines when
3240 11. Now that a reference to $sender_host_address automatically causes a reverse
3241 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3242 host lookup before query-style lookups in lists that might use this
3243 variable. This has therefore been abolished, and the "net-" prefix is no
3244 longer necessary for query-style lookups.
3246 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3247 have been a typo for LFLAGS, so it has been changed.
3249 13. The install script calls Exim with "-C /dev/null" in order to find the
3250 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3251 to be output. Howeve, since Exim outputs its version number before the
3252 error, it didn't break the script. It just looked ugly. I fixed this by
3253 always allowing "-C /dev/null" if the caller is root.
3255 14. Ignore overlarge ACL variable number when reading spool file - insurance
3256 against a later release with more variables having written the file.
3258 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3259 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3260 this, as well as the form without the "IPv6" on the front.
3262 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3263 OS/Makefile-Darwin file.
3265 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3266 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3268 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3269 variable. However, some OS use ints for inodes. I've added cast to long int
3270 to get rid of the compiler warning.
3272 19. I had forgotten to lock out "/../" in configuration file names when
3273 ALT_CONFIG_PREFIX was set.
3275 20. Routers used for verification do not need to specify transports. However,
3276 if such a router generated a host list, and callout was configured, Exim
3277 crashed, because it could not find a port number from the (non-existent)
3278 transport. It now assumes port 25 in this circumstance.
3280 21. Added the -t option to exigrep.
3282 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3283 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3284 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3285 them is set, LOOKUP_LSEARCH is forced.
3287 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3289 24. Added sender and host information to the "rejected by local_scan()" log
3290 line; previously there was no indication of these.
3292 25. Added .include_if_exists.
3294 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3295 Linux. It turns out that not all file systems support this. Apparently some
3296 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3297 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3298 not supported on the file descriptor, is now ignored when Exim is trying to
3299 sync a directory. This applies only to Linux.
3301 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3303 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3304 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3307 29. Added support for maildirsize files from supplied patch (modified a bit).
3309 30. The use of :fail: followed by an empty string could lead Exim to respond to
3310 sender verification failures with (e.g.):
3312 550 Verification failed for <xxx>
3313 550 Sender verify failed
3315 where the first response line was missing the '-' that indicates it is not
3316 the final line of the response.
3318 31. The loop for finding the name of the user that called Exim had a hardwired
3319 limit of 10; it now uses the value of finduser_retries, which is used for
3320 all other user lookups.
3322 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3325 33. Exim was neglecting to zero errno before one call of strtol() when
3326 expanding a string and expecting an integer value. On some systems this
3327 resulted in spurious "integer overflow" errors. Also, it was casting the
3328 result into an int without checking.
3330 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3331 did not work. The code looks as if it has *never* worked, though it appears
3332 to have been documented since at least releast 1.62. I have made it work.
3334 35. The "timeout_DNS" error in retry rules, also documented since at least
3335 1.62, also never worked. As it isn't clear exactly what this means, and
3336 clearly it isn't a major issue, I have abolished the feature by treating it
3337 as "timeout", and writing a warning to the main and panic logs.
3339 36. The display of retry rules for -brt wasn't always showing the error code
3342 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3343 timeout_connect_A, timeout_connect_MX.
3345 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3346 to the empty sender.
3348 39. The daemon was not analysing the content of -oX till after it had closed
3349 stderr and disconnected from the controlling terminal. This meant that any
3350 syntax errors were only noted on the panic log, and the return code from
3351 the command was 0. By re-arranging the code a little, I've made the
3352 decoding happen first, so such errors now appear on stderr, and the return
3353 code is 1. However, the actual setting up of the sockets still happens in
3354 the disconnected process, so errors there are still only recorded on the
3357 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3358 connections (as happens on some IP stacks) was logged at start up time as
3359 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3360 it from "IPv6 and IPv4", which means that two separate sockets are being
3363 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3364 says whether AF_INET or AF_INET6 was passed as an argument.
3366 42. Exiwhat output was messed up when time zones were included in log
3369 43. Exiwhat now gives more information about the daemon's listening ports,
3370 and whether -tls-on-connect was used.
3372 44. The "port" option of the smtp transport is now expanded.
3374 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3375 silently ignored. Now an error message is written to the main and panic
3378 46. There's a new ACL modifier called "logwrite" which writes to a log file
3379 as soon as it is encountered.
3381 47. Added $local_user_uid and $local_user_gid at routing time.
3383 48. Exim crashed when trying to verify a sender address that was being
3386 49. Exim was recognizing only a space character after ".include". It now also
3387 recognizes a tab character.
3389 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3390 extracting the relevant information from the specification. The man page no
3391 longer contains scrambled data for the -d option, and I've added a section
3392 at the front about calling Exim under different names.
3394 51. Added "extra_headers" argument to the "mail" command in filter files.
3396 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3399 53. Installed eximstats 1.29.
3401 54. Added transport_filter_timeout as a generic transport option.
3403 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3404 Cc: header lines. This was required by RFC 822, but it not required by RFC
3407 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3408 incoming messages that did not have them. Now it does so only if the
3409 message originates locally, that is, if there is no associated remote host
3410 address. When Resent- header lines are present, this applies to the Resent-
3411 lines rather than the non-Resent- lines.
3413 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3414 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3416 58. Messages for configuration errors now include the name of the main
3417 configuration file - useful now that there may be more than one file in a
3418 list (.included file names were always shown).
3420 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3421 for those rare installations that do not start the daemon as root or run it
3422 setuid root. I've cut out the call to initgroups() if the daemon is not
3425 60. The Exim user and group can now be bound into the binary as text strings
3426 that are looked up at the start of Exim's processing.
3428 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3430 62. Added $mailstore_basename variable.
3432 63. Installed patch to sieve.c from Michael Haardt.
3434 64. When Exim failed to open the panic log after failing to open the main log,
3435 the original message it was trying to log was written to stderr and debug
3436 output, but if they were not available (the usual case in production), it
3437 was lost. Now it is written to syslog before the two lines that record the
3438 failures to open the logs.
3440 65. Users' Exim filters run in subprocesses under the user's uid. It is
3441 possible for a "deliver" command or an alias in a "personal" command to
3442 provoke an address rewrite. If logging of address rewriting is configured,
3443 this fails because the process is not running as root or exim. There may be
3444 a better way of dealing with this, but for the moment (because 4.30 needs
3445 to be released), I have disabled address rewrite logging when running a
3446 filter in a non-root, non-exim process.
3452 1. The buildconfig auxiliary program wasn't quoting the value set for
3453 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3454 not defined. This bug was masked in 4.22 by the effect that was fixed in
3457 2. Some messages that were rejected after a message id was allocated were
3458 shown as "incomplete" by exigrep. It no longer does this for messages that
3459 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3461 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3462 have allow_domain_literals set, the ID did not get logged in the <= line.
3463 Domain literals are now always recognized in Message-ID: header lines.
3465 4. The first argument for a ${extract expansion item is the key name or field
3466 number. Leading and trailing spaces in this item were not being ignored,
3467 causing some misleading effects.
3469 5. When deliver_drop_privilege was set, single queue runner processes started
3470 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3471 same command in the process it spins off) were not dropping privilege.
3473 6. When the daemon running as "exim" started a queue runner, it always
3474 re-executed Exim in the spun-off process. This is a waste of effort when
3475 deliver_drop_privilege is set. The new process now just calls the
3476 queue-runner function directly.
3482 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3485 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3486 ignored. Though the use of -oP was forcing the writing of a pid file, it
3487 was always written to the default place.
3489 3. If the message "no IP address found for host xxxx" is generated during
3490 incoming verification, it is now followed by identification of the incoming
3491 connection (so you can more easily find what provoked it).
3493 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3495 5. Added some features to "harden" Exim a bit more against certain attacks:
3497 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3498 be put in Local/Makefile. This is like the never_users runtime option,
3499 but it cannot be overridden. The default setting is "root".
3501 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3502 prefix string with which any file named in a -C command line option
3505 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3506 is retained for -C and -D only if the caller of Exim is root. Without
3507 it, the exim user may also use -C and -D and retain privilege.
3509 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3510 command line option is disabled.
3512 6. Macro names set by the -D option must start with an upper case letter, just
3513 like macro names defined in the configuration file.
3515 7. Added "dereference=" facility to LDAP.
3517 8. Two instances of the typo "uknown" in the source files are fixed.
3519 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3520 the Configure-Makefile script screwed up while processing it.
3522 10. Incorporated PCRE 4.4.
3524 11. The SMTP synchronization check was not operating right at the start of an
3525 SMTP session. For example, it could not catch a HELO sent before the client
3526 waited for the greeting. There is now a check for outstanding input at the
3527 point when the greeting is written. Because of the duplex, asynchronous
3528 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3529 way, but not yet received, when the check is performed.
3531 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3532 on TCP/IP sockets, because this apparently causes some broken clients to
3535 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3536 unchanged) from the Cygwin maintainer.
3538 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3539 was supported instead of testing for mbx. Effectively a typo.
3541 15. The spa authenticator server code was not checking that the input it
3542 received was valid base64.
3544 16. The debug output line for the "set" modifier in ACLs was not showing the
3545 name of the variable that was being set.
3547 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3549 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3550 lookup on its own. Something else was needed to trigger the lookup. For
3551 example, a match in host_lookup or the need for a host name in a host list.
3552 Now, if $sender_host_name is referenced and the host name has not yet been
3553 looked up, a lookup is performed. If the lookup fails, the variable remains
3554 empty, and $host_lookup_failed is set to "1".
3556 19. Added "eqi" as a case-independent comparison operator.
3558 20. The saslauthd authentication condition could segfault if neither service
3559 nor realm was specified.
3561 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3562 error message that was logged was misleading, and incoming SMTP
3563 connections were dropped. The message is now more accurate, and temporary
3564 errors are given to SMTP connections.
3566 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3567 (except RSET) with a 421 error, until QUIT is received. However, it was
3568 failing to send a response to QUIT.
3570 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3571 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3572 ACL is now run after the helo_try_verify_hosts code.
3574 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3575 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3576 case-independent, so other case variants are also recognized). Apparently
3577 some systems use these upper case variants.
3579 25. If more than two messages were waiting for the same host, and a transport
3580 filter was specified for the transport, Exim sent two messages over the
3581 same TCP/IP connection, and then failed with "socket operation on non-
3582 socket" when it tried to send the third.
3584 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3586 27. The extern definition of crypt16() in expand.c was not being excluded when
3587 the OS had its own crypt16() function.
3589 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3590 as a preferred synonym for return_size_limit, both as an option and as an
3593 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3595 30. Changed the default configuration ACL to relax the local part checking rule
3596 for addresses that are not in any local domains. For these addresses,
3597 slashes and pipe symbols are allowed within local parts, but the sequence
3598 /../ is explicitly forbidden.
3600 31. SPA server authentication was not clearing the challenge buffer before
3603 32. log_message in a "warn" ACL statement was writing to the reject log as
3604 well as to the main log, which contradicts the documentation and doesn't
3605 seem right (because no rejection is happening). So I have stopped it.
3607 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3608 However, I am unable to do any testing of this.
3610 34. Fixed an infelicity in the appendfile transport. When checking directories
3611 for a mailbox, to see if any needed to be created, it was accidentally
3612 using path names with one or more superfluous leading slashes; tracing
3613 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3615 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3616 contents are added to the log line that is written for every discarded
3617 recipient. (Previously a log_message setting was ignored.)
3619 36. The ${quote: operator now quotes the string if it is empty.
3621 37. The install script runs exim in order to find its version number. If for
3622 some reason other than non-existence or emptiness, which it checks, it
3623 could not run './exim', it was installing it with an empty version number,
3624 i.e. as "exim-". This error state is now caught, and the installation is
3627 38. An argument was missing from the function that creates an error message
3628 when Exim fails to connect to the socket for saslauthd authentication.
3629 This could cause Exim to crash, or give a corrupted message.
3631 39. Added isip, isip4, and isip6 to ${if conditions.
3633 40. The ACL variables $acl_xx are now saved with the message, and can be
3634 accessed later in routers, transports, and filters.
3636 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3637 strings in the file are not string-expanded.
3639 42. If a MAIL command specified a SIZE value that was too large to fit into an
3640 int variable, the check against message_size_limit failed. Such values are
3641 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3642 day this will have to be increased, but I don't think I want to be around
3643 when emails are that large.
3650 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3651 iconv() is not standard in FreeBSD.
3653 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3654 IPv6 enabled. The observed symptom was a segmentation fault on return from
3655 the function os_common_find_running_interfaces() in src/os.c.
3657 3. In the check_special_case() function in daemon.c I had used "errno" as an
3658 argument name, which causes warnings on some systems. This was basically a
3659 typo, since it was named "eno" in the comments!
3661 4. The code that waits for the clock to tick (at a resolution of some fraction
3662 of a second) so as to ensure message-id uniqueness was always waiting for
3663 at least one whole tick, when it could have waited for less. [This is
3664 almost certainly not relevant at current processor speeds, where it is
3665 unlikely to ever wait at all. But we try to future-proof.]
3667 5. The function that sleeps for a time interval that includes fractions of a
3668 second contained a race. It did not block SIGALRM between setting the
3669 timer, and suspending (a couple of lines later). If the interval was short
3670 and the sigsuspend() was delayed until after it had expired, the suspension
3671 never ended. On busy systems this could lead to processes getting stuck for
3674 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3675 process, before it forks any delivery processes. The open lookup caching
3676 mechanism meant that the open file or database connection was passed into
3677 the delivery process. The problem was that delivery processes always tidy
3678 up cached lookup data. This could cause a problem for the next delivery
3679 process started by the queue runner, because the external queue runner
3680 process does not know about the closure. So the next delivery process
3681 still has data in the lookup cache. In the case of a file lookup, there was
3682 no problem because closing a file descriptor in a subprocess doesn't affect
3683 the parent. However, if the lookup was caching a connection to a database,
3684 the connection was closed, and the second delivery process was likely to
3685 see errors such as "PGSQL: query failed: server closed the connection
3686 unexpectedly". The problem has been fixed by closing all cached lookups
3687 in a queue runner before running a delivery process.
3689 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3690 seem to have the "const" qualifier which it has on other OS. I've
3693 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3694 *of the same type* (client or server) with the same public name that an
3695 error should be diagnosed.
3697 9. When Exim looked up a host name for an IP address, but failed to find the
3698 original IP address when looking up the host name (a safety check), it
3699 output the message "<ip address> does not match any IP for NULL", which was
3700 confusing, to say the least. The bug was that the host name should have
3701 appeared instead of "NULL".
3703 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3704 user that is built into the binary, and the -C or -D options is used, root
3705 privilege is dropped before the configuration file is read. In addition,
3706 logging is switched to stderr instead of the normal log files. If the
3707 configuration then re-defines the Exim user, the unprivileged environment
3708 is probably not what is expected, so Exim logs a panic warning message (but
3711 However, if deliver_drop_privilege is set, the unprivileged state may well
3712 be exactly what is intended, so the warning has been cut out in that case,
3713 and Exim is allowed to try to write to its normal log files.
3719 1. smtp_return_error_details was not giving details for temporary sender
3720 or receiver verification errors.
3722 2. Diagnose a configuration error if two authenticators have the same public
3725 3. Exim used not to create the message log file for a message until the first
3726 delivery attempt. This could be confusing when incoming messages were held
3727 for policy or load reasons. The message log file is now created at the time
3728 the message is received, and an initial "Received" line is written to it.
3730 4. The automatically generated man page for command line options had a minor
3731 bug that caused no ill effects; however, a more serious problem was that
3732 the procedure for building the man page automatically didn't always
3733 operate. Consequently, release 4.20 contains an out-of-date version. This
3734 shouldn't happen again.
3736 5. When building Exim with embedded Perl support, the script that builds the
3737 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3738 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3740 6. The freeze_tell option was not being used for messages that were frozen on
3741 arrival, either by an ACL or by local_scan().
3743 7. Added the smtp_incomplete_transaction log selector.
3745 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3746 was accepting AUTH without a new EHLO.
3748 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3749 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3752 10. Logging of TCP/IP connections (when configured) now happens in the main
3753 daemon process instead of the child process, so that the TCP/IP connection
3754 count is more accurate (but it can never be perfect).
3756 11. The use of "drop" in a nested ACL was not being handled correctly in the
3757 outer ACL. Now, if condition failure induced by the nested "drop" causes
3758 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3759 or "require"), the connection is dropped.
3761 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3762 that yield "discard" can now be used with an "accept" or a "discard" verb,
3763 but an error is generated for any others (because I can't see a useful way
3764 to define what should happen).
3766 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3767 are now processed in the same way as lines in the Exim configuration file.
3768 In particular, continuation lines are supported.
3770 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3772 15. Added -ti meaning -t -i.
3774 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3775 domains, and warn by logging if others are found.
3777 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3778 returned from GIFCONF when Exim is trying to find the list of interfaces on
3779 a host. The code in os.c has been modified to copy each ifreq to an aligned
3780 structure in all cases.
3782 Also, in some cases, the returned ifreq's were being copied to a 'struct
3783 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3784 means the last couple of bytes of an IPv6 address could be chopped if the
3785 ifreq contained only a normal sockaddr (14 bytes storage).
3787 18. Named domain lists were not supported in the hosts_treat_as_local option.
3788 An entry such as +xxxx was not recognized, and was treated as a literal
3791 19. Ensure that header lines added by a DATA ACL are included in the reject log
3792 if the ACL subsequently rejects the message.
3794 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3795 MD5 (which is deprecated).
3797 21. When testing a filter file using -bf, Exim was writing a message when it
3798 took the sender from a "From " line in the message, but it was not doing so
3799 when it took $return_path from a Return-Path: header line. It now does.
3801 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3802 with a valid header line field name (a series of printing characters
3803 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3805 23. Changed "disc" in the source to "disk" to conform to the documentation and
3806 the book and for uniformity.
3808 24. Ignore Sendmail's -Ooption=value command line item.
3810 25. When execve() failed while trying to run a command in a pipe transport,
3811 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3812 could be confused with a return value of 69 from the command itself. This
3813 has been changed to 127, the value the shell returns if it is asked to run
3814 a non-existent command. The wording for the related log line suggests a
3815 non-existent command as the problem.
3817 26. If received_header_text expands to an empty string, do not add a Received:
3818 header line to the message. (Well, it adds a token one on the spool, but
3819 marks it "old" so that it doesn't get used or transmitted.)
3821 27. Installed eximstats 1.28 (addition of -nt option).
3823 28. There was no check for failure on the call to getsockname() in the daemon
3824 code. This can fail if there is a shortage of resources on the system, with
3825 ENOMEM, for example. A temporary error is now given on failure.
3827 29. Contrary to the C standard, it seems that in some environments, the
3828 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3829 program. Exim now does this explicitly; it affects the formatting of
3830 timestamps using strftime().
3832 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3833 complaints. I've now initialized all the variables, to avoid this.
3835 32. Header lines added by a system filter were not being "seen" during
3836 transport-time rewrites.
3838 33. The info_callback() function passed to OpenSSL is set up with type void
3839 (*)(SSL *, int, int), as described somewhere. However, when calling the
3840 function (actually a macro) that sets it up, the type void(*)() is
3841 expected. I've put in a cast to prevent warnings from picky compilers.
3843 34. If a DNS black list lookup found a CNAME record, but there were no A
3844 records associated with the domain it pointed at, Exim crashed.
3846 35. If a DNS black list lookup returned more than one A record, Exim ignored
3847 all but the first. It now scans all returned addresses if a particular IP
3848 value is being sought. In this situation, the contents of the
3849 $dnslist_value variable are a list of all the addresses, separated by a
3852 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3853 to accept a host name and all its aliases if the forward lookup for any of
3854 them yielded the IP address of the incoming connection. Now it accepts only
3855 those names whose forward lookup yields the correct IP address. Any other
3856 names are discarded. This closes a loophole whereby a rogue DNS
3857 administrator could create reverse DNS records to break through a
3858 wildcarded host restriction in an ACL.
3860 37. If a user filter or a system filter that ran in a subprocess used any of
3861 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3862 the wrong values were passed to the pipe command ($thisaddress had the
3863 value of $0, $0 had the value of $1, etc). This bug was introduced by
3864 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3866 38. Improved the line breaking for long SMTP error messages from ACLs.
3867 Previously, if there was no break point between 40 and 75 characters, Exim
3868 left the rest of the message alone. Two changes have been made: (a) I've
3869 reduced the minimum length to 35 characters; (b) if it can't find a break
3870 point between 35 and 75 characters, it looks ahead and uses the first one
3871 that it finds. This may give the occasional overlong line, but at least the
3872 remaining text gets split now.
3874 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3875 file descriptors might be low, and that setting 1000 would always raise it.
3876 It turns out that in some environments, the limit is already over 1000 and
3877 that lowering it causes trouble. So now Exim takes care not to decrease it.
3879 40. When delivering a message, the value of $return_path is set to $sender_
3880 address at the start of routing (routers may change the value). By an
3881 oversight, this default was not being set up when an address was tested by
3882 -bt or -bv, which affected the outcome if any router or filter referred to
3885 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3886 log only when "message" or "log_message" are set. However, if one of the
3887 conditions was an address verification, or a call to a nested ACL, the
3888 messages generated by the underlying test were being passed through. This
3889 no longer happens. The underlying message is available in $acl_verify_
3890 message for both "message" and "log_message" expansions, so it can be
3891 passed through if needed.
3893 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3894 new expansion $bh_ to give the encoded byte string without charset
3895 translation. Translation happens only if iconv() is available; HAVE_ICONV
3896 indicates this at build time. HEADERS_CHARSET gives the charset to
3897 translate to; headers_charset can change it in the configuration, and
3898 "headers charset" can change it in an individual filter file.
3900 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3901 that creates RFC 2047 encoded "words" labels them as that charset instead
3902 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3903 expansion operator; (ii) when Exim creates a From: line for a local
3904 message; (iii) when a header line is rewritten to include a "phrase" part.
3906 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3907 buggy, causing it to skip the first lines of messages whose message ID
3908 ended in 'D'. This would not have bitten before Exim release 4.14, because
3909 message IDs were unlikely to end in 'D' before then. The effect was to have
3910 incorrect size information for certain domains.
3912 45. #include "config.h" was missing at the start of the crypt16.c module. This
3913 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3916 46. If there was a timeout during a "random" callout check, Exim treated it as
3917 a failure of the random address, and carried on sending RSET and the real
3918 address. If the delay was just some slowness somewhere, the response to the
3919 original RCPT would be taken as a response to RSET and so on, causing
3920 mayhem of various kinds.
3922 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3923 when I implemented it. It didn't allow for the fact that some option values
3924 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3925 the right test for positive values.
3927 48. Domain names in DNS records are case-independent. Exim always looks them up
3928 in lower case. Some resolvers return domain names in exactly the case they
3929 appear in the zone file, that is, they may contain uppercase letters. Not
3930 all resolvers do this - some return always lower case. Exim was treating a
3931 change of case by a resolver as a change of domain, similar to a widening
3932 of a domain abbreviation. This triggered its re-routing code and so it was
3933 trying to route what was effectively the same domain again. This normally
3934 caused routing to fail (because the router wouldn't handle the domain
3935 twice). Now Exim checks for this case specially, and just changes the
3936 casing of the domain that it ultimately uses when it transmits the message
3939 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3942 50. If a filter generated a file delivery with a non-absolute name (possible if
3943 no home directory exists for the router), the forbid_file option was not
3946 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3947 the existing equality matching.
3949 52. Exim was using ints instead of ino_t variables in some places where it was
3950 dealing with inode numbers.
3952 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3953 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3954 called TMPDIR, and if it finds it is different, it changes its value.
3956 54. The smtp_printf() function is now made available to local_scan() so
3957 additional output lines can be written before returning. There is also an
3958 smtp_fflush() function to enable the detection of a dropped connection.
3959 The variables smtp_input and smtp_batched_input are exported to
3962 55. Changed the default runtime configuration: the message "Unknown user"
3963 has been removed from the ACL, and instead placed on the localuser router,
3964 using the cannot_route_message feature. This means that any verification
3965 failures that generate their own messages won't get overridden. Similarly,
3966 the "Unrouteable address" message that was in the ACL for unverifiable
3967 relay addresses has also been removed.
3969 56. Added hosts_avoid_esmtp to the smtp transport.
3971 57. The exicyclog script was not checking for the esoteric option
3972 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3973 will work only if exicyclog is run under the appropriate euid.
3975 58. Following a discussion on the list, the rules by which Exim recognises line
3976 endings on incoming messages have been changed. The -dropcr and drop_cr
3977 options are now no-ops, retained only for backwards compatibility. The
3978 following line terminators are recognized: LF CRLF CR. However, special
3979 processing applies to CR:
3981 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3982 nor a local message in the state where . is a terminator.
3984 (ii) If a bare CR is encountered in a header line, an extra space is added
3985 after the line terminator so as not to end the header. The reasoning
3986 behind this is that bare CRs in header lines are most likely either
3987 to be mistakes, or people trying to play silly games.
3989 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3990 was being incorrectly given as 18 bytes larger than it should have been.
3991 This is a VOB (very old bug).
3993 60. This may never have affected anything current, but just in case it has:
3994 When the local host is found other than at the start of a list of hosts,
3995 the local host, those with the same MX, and any that follow, are discarded.
3996 When the list in question was part of a longer list of hosts, the following
3997 hosts (not currently being processed) were also being discarded. This no
3998 longer happens. I'm not sure if this situation could ever has previously
4001 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4004 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4005 header. This is something that is recommended in a new Internet Draft, and
4006 is something that is documented as being done by Sendmail. There are two
4007 possible values. For messages generated by the autoreply transport, Exim
4010 Auto-Submitted: auto-replied
4012 whereas for all other generated messages (e.g. bounces) it adds
4014 Auto-Submitted: auto-generated
4016 63. The "personal" condition in filters now includes a test for the
4017 Auto-Submitted: header. If it contains the string "auto-" the message it
4018 not considered personal.
4020 64. Added rcpt_include_affixes as a generic transport option.
4022 65. Added queue_only_override (default true).
4024 66. Added the syslog_duplication option.
4026 67. If what should have been the first header line of a message consisted of
4027 a space followed by a colon, Exim was mis-interpreting it as a header line.
4028 It isn't of course - it is syntactically invalid and should therefore be
4029 treated as the start of the message body. The misbehaviour could have
4030 caused a number of strange effects, including loss of data in subsequent
4031 header lines, and spool format errors.
4033 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4034 client host had authenticated. This control can now be exercised by an ACL
4035 for more flexibility.
4037 69. By default, callouts do not happen when testing with -bh. There is now a
4038 variant, -bhc, which does actually run the callout code, including
4039 consulting and updating the callout cache.
4041 70. Added support for saslauthd authentication, courtesy of Alexander
4044 71. If statvfs() failed on the spool or log directories while checking their
4045 size for availability, Exim confusingly gave the error "space shortage".
4046 Furthermore, in debugging mode it crashed with a floating point exception.
4047 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4048 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4049 serious problem, Exim now writes to the main and panic logs when this
4050 happens, with details of the failure. It then refuses to accept the
4051 incoming message, giving the message "spool directory problem" or "log
4052 directory problem" with a 421 code for SMTP messages.
4054 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4055 0, 1, and 2 exist, because some OS complain for execs without them (see
4056 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4057 descriptors. However, the code omitted to check that the open succeeded,
4058 causing mysterious errors if for some reason the permissions on /dev/null
4059 got screwed. Now Exim writes a message to the main and panic logs, and
4060 bombs out if it can't open /dev/null.
4062 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4063 interact so that it is all more flexible. It is supposed to remain
4064 backwards compatible. Also added extra_local_interfaces.
4066 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4067 to bomb out with an assertion failure - to the client this appears as a
4068 connection drop. This problem occurs in the part of the code that was taken
4069 from the Samba project. Fortunately, the assertion is in a very simple
4070 function, so I have fixed this by reproducing the function inline in the
4071 one place where it is called, and arranging for authentication to fail
4072 instead of killing the process with assert().
4074 75. The SPA client code was not working when the server requested OEM rather
4075 than Unicode encoding.
4077 76. Added code to make require_files with a specific uid setting more usable in
4078 the case where statting the file as root fails - usually a non-root-mounted
4079 NFS file system. When this happens and the failure is EACCES, Exim now
4080 forks a subprocess and does the per-uid checking as the relevant uid.
4082 77. Added process_log_path.
4084 78. If log_file_path was not explicitly set, a setting of check_log_space or
4085 check_log_inodes was ignored.
4087 79. If a space check for the spool or log partitions fails, the incident is now
4088 logged. Of course, in the latter case the data may get lost...
4090 80. Added the %p formatting code to string_format() so that it can be used to
4091 print addresses in debug_print(). Adjusted all the address printing in the
4092 debugging in store.c to use %p rather than %d.
4094 81. There was a concern that a line of code in smtp_in.c could overflow a
4095 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4096 initially expressed, the concern was not well-founded, because trailing
4097 spaces are removed early. However, if the trailing spaces were followed by
4098 a NULL, they did not get removed, so the overflow was possible. Two fixes
4101 (a) I re-wrote the offending code in a cleaner fashion.
4102 (b) If an incoming SMTP command contains a NULL character, it is rejected
4105 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4106 runs initgroups(), so that if the Exim user is in any additional groups,
4107 they will be used during message reception.
4113 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob