Modify (i.e. fix :-) change 4.61/PH/19: Using message arrival time for
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
dd16e114 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.343 2006/04/20 10:57:46 ph10 Exp $
495ae4b0
PH
2
3Change log file for Exim from version 4.21
4-------------------------------------------
5
c887c79e
TF
6Exim version 4.62
7-----------------
8
9TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
10 other effects) broke the use of negated acl sub-conditions.
11
1cce3af8
PH
12PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
13 patch).
14
afb3eaaf
PH
15PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
16 "Deny" causes Exim to reject the incoming connection with a 554 error.
17 Unfortunately, if there is a major crisis, such as a disk failure,
18 tcp-wrappers gives "deny", whereas what one would like would be some
19 kind of temporary error. A kludge has been added to help with this.
20 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
21 554 error is used if errno is still zero or contains ENOENT (which occurs
22 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
23 451 error is used.
24
e173618b
PH
25PH/03 Add -lutil to the default FreeBSD LIBS setting.
26
dd16e114
PH
27PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
28 errors. Otherwise a message that provokes a temporary error (when other
29 messages do not) can cause a whole host to time out.
30
1cce3af8 31
214e2000
PH
32Exim version 4.61
33-----------------
34
35PH/01 The code for finding all the local interface addresses on a FreeBSD
36 system running IPv6 was broken. This may well have applied to all BSD
37 systems, as well as to others that have similar system calls. The broken
38 code found IPv4 interfaces correctly, but gave incorrect values for the
39 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
40 that it would not match correctly against @[] and not recognize the IPv6
41 addresses as local.
42
f9daeae0
PH
43PH/02 The ipliteral router was not recognizing addresses of the form user@
44 [ipv6:....] because it didn't know about the "ipv6:" prefix.
45
7e66e54d
PH
46PH/03 Added disable_ipv6.
47
c8ea1597
PH
48PH/04 Changed $reply_address to use the raw form of the headers instead of the
49 decoded form, because it is most often used to construct To: headers
50 lines in autoreplies, and the decoded form may well be syntactically
51 invalid. However, $reply_address has leading white space removed, and all
52 newlines turned into spaces so that the autoreply transport does not
53 grumble.
54
911f6fde
PH
55PH/05 If group was specified without a user on a router, and no group or user
56 was specified on a transport, the group from the router was ignored.
57
47ca6d6c
PH
58PH/06 Increased the number of ACL variables to 20 of each type, and arranged
59 for visible compile-time settings that can be used to change these
60 numbers, for those that want even more. Backwards compatibility with old
61 spool files has been maintained. However, going back to a previous Exim
62 release will lost any variables that are in spool files.
63
ed0e9820
PH
64PH/07 Two small changes when running in the test harness: increase delay when
65 passing a TCP/IP connection to a new process, in case the original
66 process has to generate a bounce, and remove special handling of
67 127.0.0.2 (sic), which is no longer necessary.
68
eff37e47
PH
69PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
70 be the same on different OS.
71
1921d2ea
PH
72PH/09 Moved a debug statement in filter processing to avoid a race problem when
73 testing.
74
b3f69ca8
JJ
75JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
76 whether --show-vars was specified or not
77
78JJ/02 exipick: Added support for new ACL variable spool format introduced
79 in 4.61-PH/06
80
424a1c63
PH
81PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
82 syntactically invalid From: or Reply-to: line, and a filter used this to
83 generate an autoreply, and therefore failed to obtain an address for the
84 autoreply, Exim could try to deliver to a non-existent relative file
85 name, causing unrelated and misleading errors. What now happens is that
86 it logs this as a hard delivery error, but does not attempt to create a
87 bounce message.
88
7a100415
PH
89PH/11 The exinext utility has a -C option for testing purposes, but although
90 the given file was scanned by exinext itself; it wasn't being passed on
91 when Exim was called.
92
19b9dc85
PH
93PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
94 an end-of-file indication when reading a command response.
95
309bd837
PH
96PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
97 compiled. In many other places in Exim, IPv6 addresses are always
98 recognized, so I have changed this. It also means that IPv4 domain
99 literals of the form [IPV4:n.n.n.n] are now always recognized.
100
59e82a2a
PH
101PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
102 used if the router is not running as root, for example, when verifying at
103 ACL time, or when using -bh. The debugging output from this situation was
104 non-existent - all you got was a failure to exec. I have made two
105 changes:
106
107 (a) Failures to set uid/gid, the current directory, or a process leader
108 in a subprocess such as that created by queryprogram now generate
109 suitable debugging ouput when -d is set.
110
111 (b) The queryprogram router detects when it is not running as root,
112 outputs suitable debugging information if -d is set, and then runs
113 the subprocess without attempting to change uid/gid.
114
9edc04ce
PH
115PH/15 Minor change to Makefile for building test_host (undocumented testing
116 feature).
117
1349e1e5
PH
118PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
119 additional section of a DNS packet that returns MX or SRV records.
120 Instead, it always explicitly searches for A/AAAA records. This avoids
121 major problems that occur when a DNS server includes only records of one
122 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
123 fixed another bug: if SRV records were looked up and the corresponding
124 address records were *not* found in the additional section, the port
125 values from the SRV records were lost.
126
ea49d0e1
PH
127PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
128 using the correct key (the original address) when searching the retry
129 rules in order to find which one to use for generating the retry hint.
130
064a94c9
PH
131PH/18 If quota_warn_message contains a From: header, Exim now refrains from
132 adding the default one. Similarly, if it contains a Reply-To: header, the
133 errors_reply_to option, if set, is not used.
134
727071f8
PH
135PH/19 When calculating a retry time, Exim used to measure the "time since
136 failure" by looking at the "first failed" field in the retry record. Now
137 it does not use this if it is later than than the arrival time of the
138 message. Instead it uses the arrival time. This makes for better
139 behaviour in cases where some deliveries succeed, thus re-setting the
140 "first failed" field. An example is a quota failure for a huge message
141 when small messages continue to be delivered. Without this change, the
142 "time since failure" will always be short, possible causing more frequent
143 delivery attempts for the huge message than are intended.
dd16e114 144 [Note: This change was subsequently modified - see PH/04 for 4.62.]
727071f8 145
f78eb7c6
PH
146PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
147 $1, $2, $3) because the numerical variables can be reset during some
148 expansion items (e.g. "match"), thereby losing the authentication data.
149
21c28500
PH
150PH/21 Make -bV show the size of off_t variables so that the test suite can
151 decide whether to run tests for quotas > 2G.
152
153PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
154 mailbox_size, and mailbox_filecount in the appendfile transport. If a
155 filecount value is greater than 2G or if a quota value is greater than 2G
156 on a system where the size of off_t is not greater than 4, a panic error
157 is given.
158
1688f43b
PH
159PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
160 never match. The debug and -bh output now contains an explicit error
161 message indicating a malformed IPv4 address or mask.
162
163PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
164 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
165 PH/23 above applies.
166
9675b384
PH
167PH/25 Do not write to syslog when running in the test harness. The only
168 occasion when this arises is a failure to open the main or panic logs
169 (for which there is an explicit test).
170
6a3f1455
PH
171PH/26 Added the /no_tell option to "control=freeze".
172
dac79d3e
PH
173PH/27 If a host name lookup failed very early in a connection, for example, if
174 the IP address matched host_lookup and the reverse lookup yielded a name
175 that did not have a forward lookup, an error message of the form "no IP
176 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
177 could be logged. Now it outputs the IP address instead of "NULL".
1349e1e5 178
5977a0b3
PH
179PH/28 An enabling patch from MH: add new function child_open_exim2() which
180 allows the sender and the authenticated sender to be set when
181 submitting a message from within Exim. Since child_open_exim() is
182 documented for local_scan(), the new function should be too.
183
c91535f3
PH
184PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
185 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
186 results in an empty string is now treated as unset.
187
0d46a8c8
PH
188PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
189
278c6e6c
PH
190PH/31 Added sender_verify_fail logging option.
191
2cbb4081
PH
192PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
193 needed by RFC 822 but not by RFC 2822 was commented out. I have now
194 tidied the source and removed it altogether.
195
3eef829e
PH
196PH/33 When a queue run was abandoned because the load average was too high, a
197 log line was always written; now it is written only if the queue_run log
198 selector is set. In addition, the log line for abandonment now contains
199 information about the queue run such as the pid. This is always present
200 in "start" and "stop" lines but was omitted from the "abandon" line.
201
1ab95fa6
PH
202PH/34 Omit spaces between a header name and the colon in the error message that
203 is given when verify = headers_syntax fails (if there are lots of them,
204 the message gets confusing).
205
230205fc
PH
206PH/35 Change the default for dns_check_names_pattern to allow slashes within
207 names, as there are now some PTR records that contain slashes. This check
208 is only to protect against broken name servers that fall over on strange
209 characters, so the fact that it applies to all lookups doesn't matter.
210
75e0e026
PH
211PH/36 Now that the new test suite is complete, we can remove some of the
212 special code in Exim that was needed for the old test suite. For example,
213 sorting DNS records because real resolvers return them in an arbitrary
214 order. The new test suite's fake resolver always returns records in the
215 same order.
216
217PH/37 When running in the test harness, use -odi for submitted messages (e.g.
218 bounces) except when queue_only is set, to avoid logging races between
219 the different processes.
220
145396a6
PH
221PH/38 Panic-die if .include specifies a non-absolute path.
222
3cd34f13
PH
223PH/39 A tweak to the "H" retry rule from its user.
224
11121d3d
JJ
225JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
226 a label. They prevented compilation on older perls.
227
228JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
229 a warning to be raised on newish perls.
230
231JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
232 on queue. Changes to match documented behaviour of showing count of
233 messages matching specified criteria.
234
8def5aaf
PH
235PH/40 Changed the default ident timeout from 30s to 5s.
236
929ba01c
PH
237PH/41 Added support for the use of login_cap features, on those BSD systems
238 that have them, for controlling the resources used by pipe deliveries.
239
2632889e
PH
240PH/42 The content-scanning code uses fopen() to create files in which to put
241 message data. Previously it was not paying any attention to the mode of
242 the files. Exim runs with umask(0) because the rest of the code creates
243 files with open(), and sets the required mode explicitly. Thus, these
244 files were ending up world-writeable. This was not a big issue, because,
245 being within the spool directory, they were not world-accessible. I have
246 created a function called modefopen, which takes an additional mode
247 argument. It sets umask(777), creates the file, chmods it to the required
248 mode, then resets the umask. All the relevant calls to fopen() in the
249 content scanning code have been changed to use this function.
250
944a9c55
PH
251PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
252 to 24 hours. This avoids potential overflow problems when processing G
253 and H retry rules. I suspect nobody ever tinkers with this value.
254
4a23603b
PH
255PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
256
4730f942
PH
257PH/45 When the plaintext authenticator is running as a client, the server's
258 challenges are checked to ensure they are valid base64 strings. By
259 default, the authentication attempt is cancelled if an invalid string is
260 received. Setting client_ignore_invalid_base64 true ignores these errors.
261 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
262 they are received. Thus, the responses can be made to depend on the
263 challenges. If an invalid string is ignored, an empty string is placed in
264 the variable.
265
30dba1e6
PH
266PH/46 Messages that are created by the autoreply transport now contains a
267 References: header, in accordance with RFCs 2822 and 3834.
268
382afc6b
PH
269PH/47 Added authenticated_sender_force to the smtp transport.
270
a86229cf
PH
271PH/48 The ${prvs expansion was broken on systems where time_t was long long.
272
50c99ba6
PH
273PH/49 Installed latest patch from the Sieve maintainer.
274
d35e429d
PH
275PH/50 When an Exim quota was set without a file count quota, and mailbox_size
276 was also set, the appendfile transport was unnecessarily scanning a
277 directory of message files (e.g. for maildir delivery) to find the count
278 of files (along with the size), even though it did not need this
279 information. It now does the scan only if it needs to find either the
280 size of the count of files.
281
f90d018c
PH
282PH/51 Added ${time_eval: to convert Exim time strings into seconds.
283
75def545
PH
284PH/52 Two bugs concerned with error handling when the smtp transport is
285 used in LMTP mode:
286
287 (i) Exim was not creating retry information for temporary errors given
288 for individual recipients after the DATA command when the smtp transport
289 was used in LMTP mode. This meant that they could be retried too
290 frequently, and not timed out correctly.
291
292 (ii) Exim was setting the flag that allows error details to be returned
293 for LMTP errors on RCPT commands, but not for LMTP errors for individual
294 recipients that were returned after the DATA command.
295
296PH/53 This is related to PH/52, but is more general: for any failing address,
297 when detailed error information was permitted to be returned to the
298 sender, but the error was temporary, then after the final timeout, only
299 "retry timeout exceeded" was returned. Now it returns the full error as
300 well as "retry timeout exceeded".
301
c46782ef
PH
302PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
303 do this, and (what is worse) MTAs that accept it.
304
71fafd95
PH
305PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
306 will now be deprecated.
307
2c5db4fd
PH
308PH/56 New os.c-cygwin from the Cygwin maintainer.
309
9cf6b11a
JJ
310JJ/06 exipick: added --unsorted option to allow unsorted output in all output
311 formats (previously only available in exim formats via -bpr, -bpru,
312 and -bpra. Now also available in native and exiqgrep formats)
313
314JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
315 with very large, slow to parse queues
316
317JJ/08 exipick: added ! as generic prefix to negate any criteria format
318
319JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
320
898d150f
PH
321PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
322 responses to authentication challenges, though it was showing the
323 challenges; (ii) I've removed the CR characters from the debug output for
324 SMTP output lines.
325
46218253
PH
326PH/58 Allow for the insertion of a newline as well as a space when a string
327 is turned into more than one encoded-word during RFC 2047 encoding. The
328 Sieve code now uses this.
329
e97957bc
PH
330PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
331 data_4xx, lost_connection, tls_required.
332
81e509d7
PH
333PH/60 When a VRFY deferred or FAILED, the log message rather than the user
334 message was being sent as an SMTP response.
335
3d240ff7
PH
336PH/61 Add -l and -k options to exicyclog.
337
b37c4101
PH
338PH/62 When verifying, if an address was redirected to one new address, so that
339 verification continued, and the new address failed or deferred after
340 having set something in $address_data, the value of $address_data was not
341 passed back to the ACL. This was different to the case when no
342 redirection occurred. The value is now passed back in both cases.
343
79378e0f
PH
344PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
345 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
346 use login_cap.h, so on its own it isn't the distinguishing feature. The
347 new name refers directly to the setclassresources() function.
348
e49c7bb4
PH
349PH/65 Added configuration files for NetBSD3.
350
d114ec46
PH
351PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
352
f3d7df6c
PH
353PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
354 is preferred over IPv4.
355
715ab376
PH
356PH/68 The bounce_return_message and bounce_return_body options were not being
357 honoured for bounces generated during the reception of non-SMTP messages.
358 In particular, this applied to messages rejected by the ACL. This bug has
359 been fixed. However, if bounce_return_message is true and bounce_return_
360 body is false, the headers that are returned for a non-SMTP message
361 include only those that have been read before the error was detected.
362 (In the case of an ACL rejection, they have all been read.)
363
6b31b150
PH
364PH/69 The HTML version of the specification is now built in a directory called
365 spec_html instead of spec.html, because the latter looks like a path with
366 a MIME-type, and this confuses some software.
367
368PH/70 Catch two compiler warnings in sieve.c.
369
d515a917
PH
370PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
371 function verify_get_ident() calls ip_connect() to connect a socket, but
372 if the "connect()" function timed out, ip_connect() used to close the
373 socket. However, verify_get_ident() also closes the socket later, and in
374 between Exim writes to the log, which may get opened at this point. When
375 the socket was closed in ip_connect(), the log could get the same file
376 descriptor number as the socket. This naturally causes chaos. The fix is
377 not to close the socket in ip_connect(); the socket should be closed by
378 the function that creates it. There was only one place in the code where
379 this was missing, in the iplookup router, which I don't think anybody now
380 uses, but I've fixed it anyway.
381
9b8fadde
PH
382PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
383 well as to direct DNS lookups. Otherwise the handling of names in host
384 lists is inconsistent and therefore confusing.
385
214e2000 386
5de37277
PH
387Exim version 4.60
388-----------------
389
cc38ddbf
PH
390PH/01 Two changes to the default runtime configuration:
391
392 (1) Move the checks for relay_from_hosts and authenticated clients from
393 after to before the (commented out) DNS black list checks.
394
395 (2) Add control=submission to the relay_from_hosts and authenticated
396 clients checks, on the grounds that messages accepted by these
397 statements are most likely to be submissions.
5de37277 398
72fdd6ae
PH
399PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
400
401 (1) Generate an error if the third argument for the ${prvs expansion is
402 not a single digit.
403
404 (2) Treat a missing third argument of ${prvscheck as if it were an empty
405 string.
406
407 (3) Reset the variables that are obtained from the first argument of
408 ${prvscheck and used in the second argument before leaving the code,
409 because their memory is reclaimed, so using them afterwards may do
410 silly things.
411
412 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
413 one (it's much easier than Tom thought :-).
414
415 (5) Because of (4), we can now allow for the use of $prvscheck_result
416 inside the third argument.
cb9328de 417
cb741023
PH
418PH/03 For some reason, the default setting of PATH when running a command from
419 a pipe transport was just "/usr/bin". I have changed it to
420 "/bin:/usr/bin".
421
f174f16e
PH
422PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
423 anything to be listed in the output from -bV.
b2f5a032 424
c25242d7
PH
425PH/05 When a filter generated an autoreply, the entire To: header line was
426 quoted in the delivery log line, like this:
427
428 => >A.N.Other <ano@some.domain> <original@ddress> ...
429
430 This has been changed so that it extracts the operative address. There
431 may be more than one such address. If so, they are comma-separated, like
432 this:
433
434 => >ano@some.domain,ona@other.domain <original@ddress> ...
435
82c19f95
PH
436PH/06 When a client host used a correct literal IP address in a HELO or EHLO
437 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
438 not being looked up in the rDNS to get a host name, Exim was showing the
439 IP address twice in Received: lines, even though the IP addresses were
440 identical. For example:
441
442 Received: from [1.2.3.4] (helo=[1.2.3.4])
443
444 However, if the real host name was known, it was omitting the HELO data
445 if it matched the actual IP address. This has been tidied up so that it
446 doesn't show the same IP address twice.
447
d7ffbc12
PH
448PH/07 When both +timestamp and +memory debugging was on, the value given by
449 $tod_xxx expansions could be wrong, because the tod_stamp() function was
450 called by the debug printing, thereby overwriting the timestamp buffer.
451 Debugging no longer uses the tod_stamp() function when +timestamp is set.
452
9f526266
PH
453PH/08 When the original message was included in an autoreply transport, it
454 always said "this is a copy of the message, including all the headers",
455 even if body_only or headers_only was set. It now gives an appropriate
456 message.
457
87fcc8b9
PH
458PH/09 Applied a patch from the Sieve maintainer which:
459
460 o fixes some comments
461 o adds the (disabled) notify extension core
462 o adds some debug output for the result of if/elsif tests
463 o points to the current vacation draft in the documentation
464 and documents the missing references header update
465
466 and most important:
467
468 o fixes a bug in processing the envelope test (when testing
469 multiple envelope elements, the last element determinted the
470 result)
471
456682f5
PH
472PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
473 Electronic Mail") by including:
474
475 Auto-submitted: auto-generated
476
477 in the messages that it generates (bounce messages and others, such as
478 warnings). In the case of bounce messages for non-SMTP mesages, there was
479 also a typo: it was using "Auto_submitted" (underscore instead of
480 hyphen). Since every message generated by Exim is necessarily in response
481 to another message, thes have all been changed to:
482
483 Auto-Submitted: auto-replied
484
485 in accordance with these statements in the RFC:
486
487 The auto-replied keyword:
488
489 - SHOULD be used on messages sent in direct response to another
490 message by an automatic process,
491
492 - MUST NOT be used on manually-generated messages,
493
494 - MAY be used on Delivery Status Notifications (DSNs) and Message
495 Disposition Notifications (MDNs),
496
497 - MUST NOT be used on messages generated by automatic or periodic
498 processes, except for messages which are automatic responses to
499 other messages.
500
3e46c1aa
PH
501PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
502 to the default Received: header definition.
456682f5 503
49826d12
PH
504PH/12 Added log selector acl_warn_skipped (default on).
505
eba0c039
PH
506PH/13 After a successful wildlsearch lookup, discard the values of numeric
507 variables because (a) they are in the wrong storage pool and (b) even if
508 they were copied, it wouldn't work properly because of the caching.
509
a0d6ba8a
PH
510PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
511 checking when decoding. Apparently there are clients that generate
512 overlong encoded strings. Why am I not surprised?
513
f0917727
PH
514PH/15 If the first argument of "${if match_address" was not empty, but did not
515 contain an "@" character, Exim crashed. Now it writes a panic log message
516 and treats the condition as false.
517
096fee00
PH
518PH/16 In autoreply, treat an empty string for "once" the same as unset.
519
024bd3c2
PH
520PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
521 extension "envelope-auth". The code is finished and in agreement with
522 other implementations, but there is no documentation so far and in fact,
523 nobody wrote the draft yet. This extension is currently #undef'ed, thus
524 not changing the active code.
525
526 Print executed "if" and "elsif" statements when debugging is used. This
527 helps a great deal to understand what a filter does.
528
529 Document more things not specified clearly in RFC3028. I had all this
530 sorted out, when out of a sudden new issues came to my mind. Oops."
531
df199fec
PH
532PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
533 (Bugzilla #53).
534
d27f1df3
PH
535PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
536 canonical form (as documented). However, after a host name lookup from
537 the IP address, check_host() was doing a simple string comparison with
538 addresses acquired from the DNS when checking that the found name did
539 have the original IP as one of its addresses. Since any found IPv6
540 addresses are likely to be in abbreviated form, the comparison could
541 fail. Luckily, there already exists a function for doing the comparison
542 by converting both addresses to binary, so now that is used instead of
543 the text comparison.
544
96776534
PH
545PH/20 There was another similar case to PH/19, when a complete host name was
546 given in a host list; looking up its IP address could give an abbreviated
547 form, whereas the current host's name might or might not be abbreviated.
548 The same fix has been applied.
549
5de37277 550
9a799bc0
PH
551Exim version 4.54
552-----------------
553
554PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
555 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
556 It now does.
557
99a4b039
PH
558PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
559 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
560
561PH/03 Typo: missing ".o" in src/pcre/Makefile.
562
4b233853
PH
563PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
564 header line, restrict the check to what is listed in RFCs 2369 and 2929.
565 Also, for "Auto-Submitted", treat anything other than "no" as
566 non-personal, in accordance with RFC 3834. (Previously it treated
567 anything starting "auto-" as non-personal.)
568
8857ccfd
PH
569TF/01 The control=submission/name=... option had a problem with syntax
570 errors if the name included a slash character. The /name= option
571 now slurps the rest of the string, so it can include any characters
572 but it must come last in the list of options (after /sender_retain
573 or /domain=).
574
433a2980
PH
575PH/05 Some modifications to the interface to the fake nameserver for the new
576 testing suite.
577
3e46c1aa 578
9a799bc0 579
e3a311ba
TK
580Exim version 4.53
581-----------------
582
583TK/01 Added the "success_on_redirect" address verification option. See
584 NewStuff for rationale and an example.
585
13b685f9
PH
586PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
587
395ff96d
PH
588PH/02 Patch to exigrep to allow it to work on syslog lines.
589
5b68f6e4
PH
590PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
591 fread() to skip over the body file's header line, because in Cygwin the
592 header line is locked and is inaccessible.
593
1ab52c69
PH
594PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
595 co-exist for some time) to make it clear that it is the Exim ID that is
596 referenced, not the Message-ID: header line.
597
b07e6aa3
PH
598PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
599 string_format() function, because snprintf() does not exist on all
600 operating systems.
601
254e032f
PH
602PH/06 The use of forbid_filter_existstest now also locks out the use of the
603 ${stat: expansion item.
604
3af76a81
PH
605PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
606 protocol synchronization error", to keep the pedants happy.
607
2548ba04
PH
608PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
609 well as for IRIX systems, when gcc is being used. See the host.c source
610 file for comments.
611
b6c6011d
PH
612PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
613
cf39cf57
PH
614PH/10 Named domain lists were not working if used in a queue_smtp_domains
615 setting.
616
f1513293
PH
617PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
618 transport and to the smtp transport in LMTP mode.
619
727549a4
PH
620TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
621
af46795e
PH
622PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
623 run a filter in a subprocess. This could lead to confusion in subsequent
624 lookups in the parent process. There should also be a search_tidyup() at
625 the end of the subprocess.
626
d7b47fd0
PH
627PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
628 only if the host matched helo_try_verify_hosts, which caused the
629 verification to occur when the EHLO/HELO command was issued. The ACL just
630 tested the remembered result. Now, if a previous verification attempt has
631 not happened, "verify = helo" does it there and then.
632
ee744174
JJ
633JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
634
b582ab87
PH
635TK/03 Fix log output including CR from clamd.
636
41a13e0a
PH
637PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
638 exist provoked a memory error which could cause a segfault.
639
f625cc5a
PH
640PH/15 Installed PCRE 6.2
641
642PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
643
21f7af35
PH
644PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
645 of the problem. Specifically, suggested +O2 rather than +O1 for the
646 HP-UX compiler.
647
31480e42
PH
648PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
649
2d280592
PH
650PH/20 If a delivery was routed to a non-standard port by means of an SRV
651 record, the port was not correctly logged when the outgoing_port log
652 selector was set (it logged the transort's default port).
653
7cd1141b
PH
654PH/21 Added support for host-specific ports to manualroute, queryprogram,
655 fallback_hosts, and "hosts" in the smtp transport.
656
657PH/22 If the log selector "outgoing_port" is set, the port is now also given on
658 host errors such as "Connection refused".
659
750af86e
PH
660PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
661 authentication with radiusclient 0.4.9:
662
663 - Error returned from rc_read_config was caught wrongly
664 - Username/password not passed on to radius server due to wrong length.
665
666 The presumption is that some radiusclient API changes for 4.51/PH/17
667 were not taken care of correctly. The code is still untested by me (my
668 Linux distribution still has 0.3.2 of radiusclient), but it was
669 contributed by a Radius user.
670
671PH/24 When doing a callout, the value of $domain wasn't set correctly when
672 expanding the "port" option of the smtp transport.
673
4304270b
TK
674TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
675 while reading a MIME header. Thanks to Tom Hughes for a patch.
676
750af86e
PH
677PH/24 Include config.h inside local_scan.h so that configuration settings are
678 available.
679
64ffc24f
PH
680PH/25 Make $smtp_command_argument available after all SMTP commands. This means
681 that in an ACL for RCPT (for example), you can examine exactly what was
682 received.
683
5dd9625b
PH
684PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
685 commands, but it was not correctly comparing the address with the actual
686 client host address. Thus, it would show the EHLO address in Received:
687 header lines when this was not necessary.
688
5591031b
PH
689PH/27 Added the % operator to ${eval:}.
690
ba18e66a
PH
691PH/28 Exim tries to create and chdir to its spool directory when it starts;
692 it should be ignoring failures (because with -C, for example, it has lost
693 privilege). It wasn't ignoring creation failures other than "already
694 exists".
695
9cec981f
PH
696PH/29 Added "crypteq" to the list of supported features that Exim outputs when
697 -bV or -d is used.
698
aa2b5c79
PH
699PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
700 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
701 too many continuations, the temporary file was not being removed, and the
702 return code was incorrect.
aa2b5c79 703
48a53b7f
PH
704PH/31 Missing "BOOL" in function definition in filtertest.c.
705
1c59d63b
PH
706PH/32 Applied Sieve patches from the maintainer.
707
671012da
TK
708TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
709
1509d3a8
PH
710PH/33 Added "verify = not_blind".
711
712PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
713 Local/Makefile (with some defaults set). These are used in built scripts
714 such as exicyclog, but they have never been used in the exim_install
715 script (though there are many overriding facilities there). I have
716 arranged that the exim_install script now takes note of these two
717 settings.
718
719PH/35 Installed configuration files for Dragonfly.
720
2fe1a124
PH
721PH/36 When a locally submitted message by a trusted user did not contain a
722 From: header, and the sender address was obtained from -f or from an SMTP
723 MAIL command, and the trusted user did not use -F to supply a sender
724 name, $originator_name was incorrectly used when constructing a From:
725 header. Furthermore, $originator_name was used for submission mode
726 messages from external hosts without From: headers in a similar way,
727 which is clearly wrong.
728
8800895a
PH
729PH/37 Added control=suppress_local_fixups.
730
ccfdb010
PH
731PH/38 When log_selector = +received_sender was set, and the addition of the
732 sender made the log line's construction buffer exactly full, or one byte
733 less than full, an overflow happened when the terminating "\n" was
734 subsequently added.
735
1130bfb0
PH
736PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
737 when the result of a list match is failure because a DNS lookup failed.
738
ebcb507f
PH
739PH/40 RM_COMMAND is now used in the building process.
740
c35e155c
PH
741PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
742 the "build-* directories that it finds.
743
95d1f782
PH
744PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
745 address in a domain literal was a prefix of an interface address.
746
fd6de02e
PH
747PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
748 when verifying a sender address, unless rewrite_headers is false.
749
58de37c5
PH
750PH/44 Wrote a long comment about why errors_to addresses are verified as
751 recipients, not senders.
752
261cf466
TF
753TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
754 the ratelimit ACL was added.
755
3ee512ff
PH
756PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
757
e08c430f
PH
758PH/46 Added extra information about PostgreSQL errors to the error string.
759
bef5a11f
PH
760PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
761 avoiding the need to install special zones in a real server. This is
762 backwards compatible; if it can't find the fake resolver, it drops back.
763 Thus, both old and new test suites can be run.
764
7546de58
TF
765TF/02 Added util/ratelimit.pl
766
e5d5a95f
TF
767TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
768 clock is set back in time.
769
2e88a017
TF
770TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
771 Candler <B.Candler@pobox.com>.
772
a5f65aa4
TF
773TF/05 The fix for PH/43 was not completely correct; widen_domains is always
774 OK for addresses that are the result of redirections.
775
e7726cbf
PH
776PH/48 A number of further additions for the benefit of the new test suite,
777 including a fake gethostbyname() that interfaces to the fake DNS resolver
778 (see PH/47 above).
779
a7fdad5b
TF
780TF/06 The fix for widen_domains has also been applied to qualify_single and
781 search_parents which are the other dnslookup options that can cause
782 header rewrites.
783
6af56900
PH
784PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
785 type ("H").
786
0925ede6
PH
787PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
788
66afa403
TF
789TF/07 Exim produced the error message "an SRV record indicated no SMTP
790 service" if it encountered an MX record with an empty target hostname.
791 The message is now "an MX or SRV record indicated no SMTP service".
792
0154e85a
TF
793TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
794 if the DNS of the sending site is misconfigured. This is quite a
795 common situation. This change restores the behaviour of treating a
796 helo verification defer as a failure.
797
16f12c76
PH
798PH/51 If self=fail was set on a router, the bounce message did not include the
799 actual error message.
800
bbe902f0 801
e5a9dba6
PH
802Exim version 4.52
803-----------------
804
805TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
806
22c3b60b
PH
807PH/01 When a transport filter timed out in a pipe delivery, and the pipe
808 command itself ended in error, the underlying message about the transport
809 filter timeout was being overwritten with the pipe command error. Now the
810 underlying error message should be appended to the second error message.
811
06a9b4b5
PH
812TK/01 Fix poll() being unavailable on Mac OSX 10.2.
813
c1ac6996
PH
814PH/02 Reduce the amount of output that "make" produces by default. Full output
815 can still be requested.
816
9c7a242c
PH
817PH/03 The warning log line about a condition test deferring for a "warn" verb
818 was being output only once per connection, rather than after each
819 occurrence (because it was using the same function as for successful
820 "warn" verbs). This seems wrong, so I have changed it.
821
87ba3f5f
PH
822TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
823 it should not have, which might have caused a crash in the right
824 circumstances, but probably never did.
825
826PH/04 Installed a modified version of Tony Finch's patch to make submission
827 mode fix the return path as well as the Sender: header line, and to
828 add a /name= option so that you can make the user's friendly name appear
829 in the header line.
830
29aba418
TF
831TF/03 Added the control = fakedefer ACL modifier.
832
fe0dab11
TF
833TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
834 Mark Lowes for thorough testing.
870f6ba8 835
11d337a4
TK
836TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
837
838TK/03 Merged latest SRS patch from Miles Wilton.
839
415c8f3b
PH
840PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
841 with the definition in sysexits.h (which is #included earlier).
842 Fortunately, Exim does not actually use EX_OK. The code used to try to
843 preserve the sysexits.h value, by assumimg that macro definitions were
844 scanned for macro replacements. I have been disabused of this notion,
845 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 846
958541e9
PH
847PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
848 in the smtp transport. When a block could not be written in a single
849 write() function, the timeout was being re-applied to each part-write.
850 This seems wrong - if the receiver was accepting one byte at a time it
851 would take for ever. The timeout is now adjusted when this happens. It
852 doesn't have to be particularly precise.
853
c206415f
TK
854TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
855 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
856
2a4be8f9
PH
857PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
858 without a domain if the check to <postmaster@domain> fails.
859
1cba11c5
SC
860SC/01 Eximstats: added -xls and the ability to specify output files
861 (patch written by Frank Heydlauf).
862
863SC/02 Eximstats: use FileHandles for outputing results.
864
865SC/03 Eximstats: allow any combination of xls, txt, and html output.
866
867SC/04 Eximstats: fixed display of large numbers with -nvr option
868
869SC/05 Eximstats: fixed merging of reports with empty tables.
870
871SC/06 Eximstats: added the -include_original_destination flag
872
873SC/07 Eximstats: removed tabs and trailing whitespace.
874
1005d00e
TK
875TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
876
877TK/06 MBOX spool code: Add real "From " MBOX separator line
878 so the .eml file is really in mbox format (even though
879 most programs do not really care). Patch from Alex Miller.
880
881TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
882 The latter is generated from $received_to and is only set if the
883 message has one envelope recipient. SA can use these headers,
884 obviously out-of-the-box. Patch from Alex Miller.
885
9b4768fa
PH
886PH/08 The ${def test on a variable was returning false if the variable's
887 value was "0", contrary to what the specification has always said!
888 The result should be true unless the variable is empty.
889
890PH/09 The syntax error of a character other than { following "${if
891 def:variable_name" (after optional whitespace) was not being diagnosed.
892 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
893 accidental colon was present, for example, could give incorrect results.
894
0d7eb84a
PH
895PH/10 Tidied the code in a number of places where the st_size field of a stat()
896 result is used (not including appendfile, where other changes are about
897 to be made).
898
899PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
900 This involved changing a lot of size variables from int to off_t. It
901 should work with maildirs and everything.
902
40727bee
TK
903TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
904 spamd dying while we are connected to it.
905
554d2369
TF
906TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
907 <jgh@wizmail.org>
908
1f922db1
PH
909PH/12 Applied Alex Kiernan's patch for the API change for the error callback
910 function for BDB 4.3.
911
ef213c3b
PH
912PH/13 Changed auto_thaw such that it does not apply to bounce messages.
913
8ac170f3
PH
914PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
915 the sources for PCRE have been re-arranged and more files are now
916 involved.
917
b1c749bb
PH
918PH/15 The code I had for printing potentially long long variables in PH/11
919 above was not the best (it lost precision). The length of off_t variables
920 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
921 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
922 to be "long long int" or "long int". This is needed for the internal
923 formatting function string_vformat().
b1c749bb 924
4aac9b49
PH
925PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
926 the configuration file to be ":syslog", then the script "guesses" where
927 the logs files are, rather than using the compiled in default. In our
928 case the guess is not the same as the compiled default, so the script
929 suddenly stopped working when I started to use syslog. The patch checks
930 to see if log_file_path is "". If so, it attempts to read it from exim
931 with no configuration file to get the compiled in version, before it
932 falls back to the previous guessing code."
933
294520c8
TK
934TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
935 implementing BATV in an Exim configuration. See NewStuff for the gory
936 details.
937
5bd022fe
PH
938PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
939 Makefile that are specific to HP-UX.
940
90e9ce59
PH
941PH/18 If the "use_postmaster" option was set for a recipient callout together
942 with the "random" option, the postmaster address was used as the MAIL
943 FROM address for the random test, but not for the subsequent recipient
944 test. It is now used for both.
945
5ea81592
PH
946PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
947 patch removes a few documentation additions to RFC 3028, because the
948 latest draft now contains them. It adds the new en;ascii-case comparator
949 and a new error check for 8bit text in MIME parts. Comparator and
950 require names are now matched exactly. I enabled the subaddress
951 extension, but it is not well tested yet (read: it works for me)."
952
c6c2dc1d
PH
953PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
954 rework some of the code of TK/09 above to avoid the hardwired use of
955 "%lld" and "long long". Replaced the call to snprintf() with a call to
956 string_vformat().
957
fffffe4c
PH
958PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
959 records point to non-existent hosts", "retry timeout exceeded", and
960 "retry time not reached for any host after a long failure period".
ca02eafb 961
9a26b6b2
PH
962PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
963 experimental DomainKeys support:
964
965 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
966 (2) On an error such as an illegally used "control", the wrong name for
967 the control was given.
968
969 These problems did NOT occur unless DomainKeys support was compiled.
970
4aee0225
PH
971PH/23 Added daemon_startup_retries and daemon_startup_sleep.
972
32d668a5
PH
973PH/24 Added ${if match_ip condition.
974
8187c3f3
PH
975PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
976 databases so that it will be absolutely obvious if a crash occurs in the
977 DB library. This is a regular occurrence (often caused by mis-matched
978 db.h files).
979
ff790e47 980PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
981 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
982 picked up on a user's system that detects such things. There doesn't seem
983 to be a gcc warning option for this - only an attribute that has to be
984 put on the function's prototype. It seems that in Fedora Core 4 they have
985 set this on a number of new functions. No doubt there will be more in due
986 course.
ff790e47 987
5417f6d1
PH
988PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
989 specify a transport. However, if an address that was verified by such a
990 router was the subject of a callout, Exim crashed because it tried to
991 read the rcpt_include_affixes from the non-existent transport. Now it
992 just assumes that the setting of that option is false. This bug was
993 introduced by 4.51/PH/31.
994
59cf8544
PH
995PH/28 Changed -d+all to exclude +memory, because that information is very
996 rarely of interest, but it makes the output a lot bigger. People tend to
997 do -d+all out of habit.
998
e7ad8a65
PH
999PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1000 code in os-type was giving problems when libc.so lives in lib64, like on
1001 x86_64 Fedora Core.
1002
ade42478
PH
1003PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1004 aren't the modern standard, and it seems that some systems' include files
1005 don't always have them. Exim was already checking for some of the newer
1006 ones like T_AAAA, and defining it itself. I've added checks for all the
1007 record types that Exim uses.
1008
182ad5cf
PH
1009PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1010 not automatically generating a new one, as it is supposed to. This
1011 prevented TLS from working. If the file did exist, but contained invalid
1012 data, a new version was generated, as expected. It was only the case of a
1013 non-existent file that was broken.
1014
b0d9fc80
TK
1015TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1016 with a change in libdomainkeys > 0.64.
1017
1018TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1019 from DNS. If the selector record carries the flag, it now has
1020 precedence over the domain-wide flag.
1021
1022TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1023
47c7a64a
PH
1024PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1025 the use of a port name that isn't defined in /etc/services) occurred, the
1026 message was deferred as in a normal delivery, and thus remained on the
1027 spool, instead of being failed because of the mua_wrapper setting. This
1028 is now fixed, and I tidied up some of the mua_wrapper messages at the
1029 same time.
1030
a388bce4
SC
1031SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1032 the messages in a hash of arrays rather than using individual hashes.
1033 This is a bit cleaner and results in dramatic memory savings, albeit
1034 at a slight CPU cost.
1035
1036SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1037 as requested by Marc Sherman.
1038
1039SC/10 Eximstats: added histograms for user specified patterns as requested
1040 by Marc Sherman.
1041
0793e4ed
SC
1042SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1043
c58b88df
PH
1044PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1045 fopen() in the content-scanning modules that did not already have it.
1046
e7ad8a65 1047
7982096b
PH
1048Exim version 4.51
1049-----------------
1050
1a46a8c5
PH
1051TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1052 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1053
2f079f46 1054TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
1055
1056TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1057
1058PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1059 to test Sieve filters that use "vacation".
1060
1061PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1062 that changes the way the GnuTLS parameters are stored in the cache file.
1063 The new format can be generated externally. For backward compatibility,
1064 if the data in the cache doesn't make sense, Exim assumes it has read an
1065 old-format file, and it generates new data and writes a new file. This
1066 means that you can't go back to an older release without removing the
1067 file.
1068
1069PH/03 A redirect router that has both "unseen" and "one_time" set does not
1070 work if there are any delivery delays because "one_time" forces the
1071 parent to be marked "delivered", so its unseen clone is never tried
1072 again. For this reason, Exim now forbids the simultaneous setting of
1073 these two options.
1074
1075PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1076 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1077 entry if you want to know the details. The fix, however, neglected to
1078 consider the case when local delivery batching is involved. The test for
1079 "previously delivered" was not happening when checking to see if an
1080 address could be batched with a previous (undelivered) one; under
1081 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 1082 address.
1a46a8c5
PH
1083
1084PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1085 in its include files, and this causes problems building Exim.
1086
1087PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1088 header_syntax) but Exim was just ignoring anything given after a slash.
1089 In particular, this caused confusion with an attempt to use "verify =
1090 reverse_host_lookup/defer_ok". An error is now given when options are
1091 supplied for verify items that do not have them. (Maybe reverse_host_
1092 lookup should have a defer_ok option, but that's a different point.)
1093
1094PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1095 defined by RFC 821) to 2048, because there were problems with some AUTH
1096 commands, and RFC 1869 says the size should be increased for extended
1097 SMTP commands that take arguments.
1098
1099PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1100 Finch).
1101
1102PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1103 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 1104
49c2d5ea
PH
1105PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1106 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
1107 whatever). Otherwise the password may appear in the log. Following change
1108 PH/42 below, there is no longer a chance of it appearing in a bounce
1109 message.
49c2d5ea 1110
bf759a8b
PH
1111PH/11 Installed exipick version 20050225.0 from John Jetmore.
1112
83364d30
PH
1113PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1114 of its addresses was ever tried. (Bugzilla bug #2.)
1115
7999bbd7
PH
1116PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1117 the result incorrectly in the debug output. (It correctly added a newline
1118 to what was transported.)
1119
7dbf77c9
PH
1120TF/01 Added $received_time.
1121
74e0617f
PH
1122PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1123 commented out examples of how to interface to a virus scanner and to
1124 SpamAssassin. Also added commented examples of av_scanner and
1125 spamd_address settings.
1126
2f079f46
PH
1127PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1128 and controls are allowed in which ACLs. There were a couple of minor
1129 errors. Some of the entries in the conditions table (which is a table of
1130 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1131 negation of where the condition IS allowed.
1132
8c841523
PH
1133PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1134
7766a4f0
PH
1135PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1136 header file does not have a version number, so I've had to invent a new
1137 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1138 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1139 radiusclient), but it was contributed by a Radius user.
1140
8b417f2c
PH
1141PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1142 files or directories, for OpenSSL.
1143
901f42cb
PH
1144PH/19 When an Exim process that is running as root has to create an Exim log
1145 file, it does so in a subprocess that runs as exim:exim so as to get the
1146 ownership right at creation (otherwise, other Exim processes might see
1147 the file with the wrong ownership). There was no test for failure of this
1148 fork() call, which would lead to the process getting stuck as it waited
1149 for a non-existent subprocess. Forks do occasionally fail when resources
1150 run out. I reviewed all the other calls to fork(); they all seem to check
1151 for failure.
1152
f9b9210e
PH
1153PH/20 When checking for unexpected SMTP input at connect time (before writing
1154 the banner), Exim was not dealing correctly with a non-positive return
1155 from the read() function. If the client had disconnected by this time,
1156 the result was a log entry for a synchronization error with an empty
1157 string after "input=" when read() returned zero. If read() returned -1
1158 (an event I could not check), uninitialized data bytes were printed.
1159 There were reports of junk text (parts of files, etc) appearing after
1160 "input=".
1161
54cdb463
PH
1162PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1163
cf00dad6
PH
1164PH/22 Added support for macro redefinition, and (re)definition in between
1165 driver and ACL definitions.
1166
acb1b346
PH
1167PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1168 forgetting to use the resulting value; it was using the unexpanded value.
1169
c5ddb310
PH
1170PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1171 hadn't been configured. The fix is from Juergen Kreileder, who
1172 understands it better than I do:
1173
1174 "Here's what I see happening with three configured cyrus_sasl
1175 authenticators configured (plain, login, cram-md5):
1176
1177 On startup auth_cyrus_sasl_init() gets called for each of these.
1178 This means three calls to sasl_listmech() without a specified mech_list.
1179 => SASL tests which mechs of all available mechs actually work
1180 => three warnings about OTP not working
1181 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1182
1183 With the patch, sasl_listmech() also gets called three times. But now
1184 SASL's mech_list option is set to the server_mech specified in the the
1185 authenticator. Or in other words, the answer from sasl_listmech()
1186 gets limited to just the mech you're testing for (which is different
1187 for each call.)
1188 => the return list contains just 'plain' or 'login', 'cram-md5' or
1189 nothing depending on the value of ob->server_mech.
1190
1191 I've just tested the patch: Authentication still works fine,
1192 unavailable mechs specified in the exim configuration are still
1193 caught, and the auth.log warnings about OTP are gone."
1194
31619da6
PH
1195PH/25 When debugging is enabled, the contents of the command line are added
1196 to the debugging output, even when log_selector=+arguments is not
1197 specified.
1198
bebaf0fc
PH
1199PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1200 answer is "GNU", and only if the return is "GNU/something" is the answer
1201 "Linux".
1202
475fe28a
PH
1203PH/27 $acl_verify_message is now set immediately after the failure of a
1204 verification in an ACL, and so is available in subsequent modifiers. In
1205 particular, the message can be preserved by coding like this:
1206
1207 warn !verify = sender
1208 set acl_m0 = $acl_verify_message
1209
1210 Previously, $acl_verify_message was set only while expanding "message"
1211 and "log_message" when a very denied access.
1212
7e8bec7a
PH
1213PH/28 Modified OS/os.c-Linux with
1214
1215 -#ifndef OS_LOAD_AVERAGE
1216 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1217
1218 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1219 nomenclature these days.)
1220
e4a89c47
PH
1221PH/29 Installed patch from the Sieve maintainer that adds the options
1222 sieve_useraddress and sieve_subaddress to the redirect router.
1223
5ca2a9a1
PH
1224PH/30 In these circumstances:
1225 . Two addresses routed to the same list of hosts;
1226 . First host does not offer TLS;
1227 . First host accepts first address;
1228 . First host gives temporary error to second address;
1229 . Second host offers TLS and a TLS session is established;
1230 . Second host accepts second address.
1231 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1232 and peerdn, if requested) that were in fact used only for the second
1233 address.
7e8bec7a 1234
c688b954
PH
1235PH/31 When doing a callout as part of verifying an address, Exim was not paying
1236 attention to any local part prefix or suffix that was matched by the
1237 router that accepted the address. It now behaves in the same way as it
1238 does for delivery: the affixes are removed from the local part unless
1239 rcpt_include_affixes is set on the transport.
1240
fed77020
PH
1241PH/32 Add the sender address, as F=<...>, to the log line when logging a
1242 timeout during the DATA phase of an incoming message.
1243
7fe1560f
PH
1244PH/33 Sieve envelope tests were broken for match types other than :is. I have
1245 applied a patch sanctioned by the Sieve maintainer.
c688b954 1246
ebb6e6d5
PH
1247PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1248 the uid or gid is negative. A case of a negative gid caused this to be
1249 noticed. The fix allows for either to be negative.
1250
9c4e8f60
PH
1251PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1252 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1253 overlooked.
1254
1255PH/36 The change PH/12 above was broken. Fixed it.
1256
d7174846
PH
1257PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1258 the grounds that routing the same address twice would always produce the
1259 same answer. This might have been true once, but it is certainly no
1260 longer true now. Routing a child address may depend on the previous
1261 routing that produced that child. Some complicated redirection strategies
1262 went wrong when messages had multiple recipients, and made Exim's
1263 behaviour dependent on the order in which the addresses were given.
1264
1265 I have moved the duplicate checking until after the routing is complete.
1266 Exim scans the addresses that are assigned to local and remote
1267 transports, and removes any duplicates. This means that more work will be
1268 done, as duplicates will always all be routed, but duplicates are
1269 presumably rare, so I don't expect this is of any significance.
1270
1271 For deliveries to pipes, files, and autoreplies, the duplicate checking
1272 still happens during the routing process, since they are not going to be
1273 routed further.
1274
cfe75fc3
PH
1275PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1276 It corrects a timeout issue with spamd. This is Ian's comment: "The
1277 background is that sometimes spamd either never reads data from a
1278 connection it has accepted, or it never writes response data. The exiscan
1279 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1280 blindly assumes that writes won't block so it may never time out."
1281
be22d70e
PH
1282PH/39 Allow G after quota size as well as K and M.
1283
0612b098
PH
1284PH/40 The value set for $authenticated_id in an authenticator may not contain
1285 binary zeroes or newlines because the value is written to log lines and
1286 to spool files. There was no check on this. Now the value is run through
1287 the string_printing() function so that such characters are converted to
1288 printable escape sequences.
1289
2e0c1448
PH
1290PH/41 $message_linecount is a new variable that contains the total number of
1291 lines in the message. Compare $body_linecount, which is the count for the
1292 body only.
1293
447d236c
PH
1294PH/42 Exim no longer gives details of delivery errors for specific addresses in
1295 bounce and delay warning messages, except in certain special cases, which
1296 are as follows:
1297
1298 (a) An SMTP error message from a remote host;
1299 (b) A message specified in a :fail: redirection;
1300 (c) A message specified in a "fail" command in a system filter;
1301 (d) A message specified in a FAIL return from the queryprogram router;
1302 (e) A message specified by the cannot_route_message router option.
1303
1304 In these cases only, Exim does include the error details in bounce and
1305 warning messages. There are also a few cases where bland messages such
1306 as "unrouteable address" or "local delivery error" are given.
1307
d20976dc
PH
1308PH/43 $value is now also set for the "else" part of a ${run expansion.
1309
f656d135
PH
1310PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1311 being worked on, but at least Exim now implements the latest version to
1312 play with."
1313
2e2a30b4
PH
1314PH/45 In a pipe transport, although a timeout while waiting for the pipe
1315 process to complete was treated as a delivery failure, a timeout while
1316 writing the message to the pipe was logged, but erroneously treated as a
1317 successful delivery. Such timeouts include transport filter timeouts. For
1318 consistency with the overall process timeout, these timeouts are now
1319 treated as errors, giving rise to delivery failures by default. However,
1320 there is now a new Boolean option for the pipe transport called
1321 timeout_defer, which, if set TRUE, converts the failures into defers for
1322 both kinds of timeout. A transport filter timeout is now identified in
1323 the log output.
1324
9176e9f0
PH
1325PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1326 systems where "make" and "gmake" are different, calling "gmake" at top
1327 level broke things. I've arranged for the value of $(MAKE) to be passed
1328 from the Makefile to this script so that it can call the same version of
1329 "make".
1330
7982096b 1331
bbe902f0
PH
1332A note about Exim versions 4.44 and 4.50
1333----------------------------------------
1334
1335Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1336changes of various kinds. As a consequence, a big documentation update was
1337needed. This delayed the release for rather longer than seemed good, especially
1338in the light of a couple of (minor) security issues. Therefore, the changes
1339that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1340release. So 4.44 and 4.50 are in effect two different branches that both start
1341from 4.43.
1342
1343I have left the 4.50 change log unchanged; it contains all the changes since
13444.43. The change log for 4.44 is below; many of its items are identical to
1345those for 4.50. This seems to be the most sensible way to preserve the
1346historical information.
1347
1348
f7b63901 1349Exim version 4.50
495ae4b0
PH
1350-----------------
1351
5fe762f6
PH
1352 1. Minor wording change to the doc/README.SIEVE file.
1353
139059f6 1354 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1355 computation of the current number of files was incorrect.
495ae4b0 1356
7086e875
PH
1357 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1358 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1359 place.
1360
35af9f61
PH
1361 4. Give more explanation in the error message when the command for a transport
1362 filter fails to execute.
1363
b668c215
PH
1364 5. There are several places where Exim runs a non-Exim command in a
1365 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1366 was being done only for the command run by the queryprogram router. It is
1367 now done for all such subprocesses. The other cases are: ${run, transport
1368 filters, and the commands run by the lmtp and pipe transports.
1369
a494b1e1
PH
1370 6. Added CONFIGURE_GROUP build-time option.
1371
1372 7. Some older OS have a limit of 256 on the maximum number of file
1373 descriptors. Exim was using setrlimit() to set 1000 as a large value
1374 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1375 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1376
c5fcb476
PH
1377 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1378 was an oversight, and furthermore, ever since the addition of extra
1379 controls (e.g. 4.43/32), the checks on when to allow different forms of
1380 "control" were broken. There should now be diagnostics for all cases when a
1381 control that does not make sense is encountered.
1382
69358f02
PH
1383 9. Added the /retain_sender option to "control=submission".
1384
5be20824
PH
138510. $recipients is now available in the predata ACL (oversight).
1386
eb2c0248
PH
138711. Tidy the search cache before the fork to do a delivery from a message
1388 received from the command line. Otherwise the child will trigger a lookup
1389 failure and thereby defer the delivery if it tries to use (for example) a
1390 cached ldap connection that the parent has called unbind on.
1391
2a3eea10
PH
139212. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1393 of $address_data from the recipient verification was clobbered by the
1394 sender verification.
1395
139613. The value of address_data from a sender verification is now available in
1397 $sender_address_data in subsequent conditions in the ACL statement.
1398
23c7ff99
PH
139914. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1400
4deaf07d
PH
140115. Added a new option "connect=<time>" to callout options, to set a different
1402 connection timeout.
1403
926e1192
PH
140416. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1405 was its contents. (It was OK if the option was not defined at all.)
1406
650edc6f
PH
140717. A "Completed" log line is now written for messages that are removed from
1408 the spool by the -Mrm option.
1409
2c7db3f5
PH
141018. New variables $sender_verify_failure and $recipient_verify_failure contain
1411 information about exactly what failed.
1412
3d235903
PH
141319. Added -dd to debug only the daemon process.
1414
7c7ad977
PH
141520. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1416 handles timeouts, both on the server side and network timeouts. Renamed the
1417 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1418
981756db
PH
141921. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1420 It is now set to "smtps".
1421
d4eb88df
PH
142222. $host_address is now set to the target address during the checking of
1423 ignore_target_hosts.
1424
142523. When checking ignore_target_hosts for an ipliteral router, no host name was
1426 being passed; this would have caused $sender_host_name to have been used if
1427 matching the list had actually called for a host name (not very likely,
1428 since this list is usually IP addresses). A host name is now passed as
1429 "[x.x.x.x]".
1430
7d468ab8
PH
143124. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1432 code that specifies a non-restarting handler (typically sigaction() in
1433 modern systems) in an attempt to fix a rare and obscure crash bug.
1434
143525. Narrowed the window for a race in the daemon that could cause it to ignore
1436 SIGCHLD signals. This is not a major problem, because they are used only to
1437 wake it up if nothing else does.
1438
62c0818f
PH
143926. A malformed maildirsize file could cause Exim to calculate negative values
1440 for the mailbox size or file count. Odd effects could occur as a result.
1441 The maildirsize information is now recalculated if the size or filecount
1442 end up negative.
1443
26034054
PH
144427. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1445 support for a long time. Removed HAVE_SYS_VFS_H.
1446
af66f652
PH
144728. Installed the latest version of exipick from John Jetmore.
1448
90af77f4
PH
144929. In an address list, if the pattern was not a regular expression, an empty
1450 subject address (from a bounce message) matched only if the pattern was an
1451 empty string. Non-empty patterns were not even tested. This was the wrong
1452 because it is perfectly reasonable to use an empty address as part of a
1453 database query. An empty address is now tested by patterns that are
1454 lookups. However, all the other forms of pattern expect the subject to
1455 contain a local part and a domain, and therefore, for them, an empty
1456 address still always fails if the pattern is not itself empty.
1457
d8ef3577
PH
145830. Exim went into a mad DNS loop when attempting to do a callout where the
1459 host was specified on an smtp transport, and looking it up yielded more
1460 than one IP address.
1461
5cb8cbc6
PH
146231. Re-factored the code for checking spool and log partition space into a
1463 function that finds that data and another that does the check. The former
1464 is then used to implement four new variables: $spool_space, $log_space,
1465 $spool_inodes, and $log_inodes.
1466
14702f5b
PH
146732. The RFC2047 encoding function was originally intended for short strings
1468 such as real names; it was not keeping to the 75-character limit for
1469 encoded words that the RFC imposes. It now respects the limit, and
1470 generates multiple encoded words if necessary. To be on the safe side, I
1471 have increased the buffer size for the ${rfc2047: expansion operator from
1472 1024 to 2048 bytes.
1473
063b1e99
PH
147433. It is now permitted to omit both strings after an "if" condition; if the
1475 condition is true, the result is "true". As before, when the second string
1476 is omitted, a false condition yields an empty string. This makes it less
1477 cumbersome to write custom ACL and router conditions.
1478
652e1b65
PH
147934. Failure to deliver a bounce message always caused it to be frozen, even if
1480 there was an errors_to setting on the router. The errors_to setting is now
1481 respected.
1482
6f0c9a4f
PH
148335. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1484 canonical form (fully expanded) before being placed in
1485 $sender_host_address.
1486
33397d19
PH
148736. The table in the code that translates DNS record types into text (T_A to
1488 "A" for instance) was missing entries for NS and CNAME. It is just possible
1489 that this could have caused confusion if both these types were looked up
1490 for the same domain, because the text type is used as part of Exim's
1491 per-process caching. But the chance of anyone hitting this buglet seems
1492 very small.
1493
7bb56e1f
PH
149437. The dnsdb lookup has been extended in a number of ways.
1495
1496 (1) There is a new type, "zns", which walks up the domain tree until it
1497 finds some nameserver records. It should be used with care.
1498
ea3bc19b
PH
1499 (2) There is a new type, "mxh", which is like "mx" except that it returns
1500 just the host names, not the priorities.
1501
1502 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1503 looked up. The behaviour when one of the lookups defers can be
1504 controlled by a keyword.
7bb56e1f 1505
ea3bc19b 1506 (4) It is now possible to specify the separator character for use when
7bb56e1f 1507 multiple records are returned.
33397d19 1508
0bcb2a0e
PH
150938. The dnslists ACL condition has been extended: it is now possible to supply
1510 a list of IP addresses and/or domains to be looked up in a particular DNS
1511 domain.
1512
2ac0e484
PH
151339. Added log_selector=+queue_time_overall.
1514
4e1fde53
PH
151540. When running the queue in the test harness, wait just a tad after forking a
1516 delivery process, to get repeatability of debugging output.
1517
de365ded
PH
151841. Include certificate and key file names in error message when GnuTLS fails
1519 to set them up, because the GnuTLS error message doesn't include the name
1520 of the failing file when there is a problem reading it.
1521
f05da2e8
PH
152242. Allow both -bf and -bF in the same test run.
1523
d6453af2
PH
152443. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1525
f7b63901
PH
152644. The "Exiscan patch" is now merged into the mainline Exim source.
1527
152845. Sometimes the final signoff response after QUIT could fail to get
1529 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1530 < 0 before doing a fflush(). This bug looks as though it goes back to the
1531 introduction of TLS in release 3.20, but "sometimes" must have been rare
1532 because the tests only now provoked it.
1533
a444213a
PH
153446. Reset the locale to "C" after calling embedded Perl, in case it was changed
1535 (this can affect the format of dates).
1536
0ec020ea
PH
153747. exim_tidydb, when checking for the continued existence of a message for
1538 which it has found a message-specific retry record, was not finding
1539 messages that were in split spool directories. Consequently, it was
1540 deleting retry records that should have stayed in existence.
1541
b1206957
PH
154248. Steve fixed some bugs in eximstats.
1543
154449. The SPA authentication driver was not abandoning authentication and moving
1545 on to the next authenticator when an expansion was forced to fail,
1546 contradicting the general specification for all authenticators. Instead it
1547 was generating a temporary error. It now behaves as specified.
1548
26dd5a95
PH
154950. The default ordering of permitted cipher suites for GnuTLS was pessimal
1550 (the order specifies the preference for clients). The order is now AES256,
1551 AES128, 3DES, ARCFOUR128.
1552
343b2385
PH
155351. Small patch to Sieve code - explicitly set From: when generating an
1554 autoreply.
1555
1c5466b9
PH
155652. Exim crashed if a remote delivery caused a very long error message to be
1557 recorded - for instance if somebody sent an entire SpamAssassin report back
1558 as a large number of 550 error lines. This bug was coincidentally fixed by
1559 increasing the size of one of Exim's internal buffers (big_buffer) that
1560 happened as part of the Exiscan merge. However, to be on the safe side, I
1561 have made the code more robust (and fixed the comments that describe what
1562 is going on).
1563
55ee9ee3
PH
156453. Now that there can be additional text after "Completed" in log lines (if
1565 the queue_time_overall log selector is set), a one-byte patch to exigrep
1566 was needed to allow it to recognize "Completed" as not the last thing in
1567 the line.
1568
d38f8232
PH
156954. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1570 patch that reportedly fixes this has been added. I am not expert enough to
1571 create a test for it. This is what the patch creator wrote:
1572
1573 "I found a little strange behaviour of ldap code when working with
1574 Windows 2003 AD Domain, where users was placed in more than one
1575 Organization Units. When I tried to give exim partial DN, the exit code
1576 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1577 But simultaneously result of request was absolutely normal ldap result,
1578 so I produce this patch..."
1579
3295e65b
PH
1580 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1581 REFERENCE, so I have modified the code to exclude the patch when that macro
1582 is not defined.
1583
7102e136
PH
158455. Some experimental protocols are using DNS PTR records for new purposes. The
1585 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1586 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1587 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1588 CAN-2005-0021
7102e136 1589
3ca0ba97
PH
159056. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1591
c2bcbe20
PH
159257. Double the size of the debug message buffer (to 2048) so that more of very
1593 long debug lines gets shown.
1594
18ce445d
PH
159558. The exicyclog utility now does better if the number of log files to keep
1596 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1597
1f5b4c3d
PH
159859. Two changes related to the smtp_active_hostname option:
1599
1600 (1) $smtp_active_hostname is now available as a variable.
1601 (2) The default for smtp_banner uses $smtp_active_hostname instead
1602 of $primary_hostname.
1603
b975ba52
PH
160460. The host_aton() function is supposed to be passed a string that is known
1605 to be a valid IP address. However, in the case of IPv6 addresses, it was
1606 not checking this. This is a hostage to fortune. Exim now panics and dies
1607 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1608 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1609 components; fortuitously, this particular loophole had already been fixed
1610 by change 4.50/55 above.
1611
1612 If there are any other similar loopholes, the new check in host_aton()
1613 itself should stop them being exploited. The report I received stated that
1614 data on the command line could provoke the exploit when Exim was running as
1615 exim, but did not say which command line option was involved. All I could
1616 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1617 running as the user.
ea3a6f44 1618 CAN-2005-0021
85b87bc2
PH
1619
162061. There was a buffer overflow vulnerability in the SPA authentication code
1621 (which came originally from the Samba project). I have added a test to the
1622 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1623 CAN-2005-0022
b975ba52 1624
17ffcae7
PH
162562. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1626 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1627
d95f9fdb
PH
162863. The daemon start-up calls getloadavg() while still root for those OS that
1629 need the first call to be done as root, but it missed one case: when
1630 deliver_queue_load_max is set with deliver_drop_privilege. This is
1631 necessary for the benefit of the queue runner, because there is no re-exec
1632 when deliver_drop_privilege is set.
1633
86b8287f
PH
163464. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1635 This has been fixed.
1636
60dc5e56
PH
163765. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1638 was in use, was not putting the data itself into the right store pool;
1639 consequently, it could be overwritten for a subsequent message in the same
1640 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1641 the caching.)
1642
533244af
PH
164366. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1644
a5a28604
PH
164567. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1646 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1647 the function were treating the return as a boolean value, which happened to
1648 work because 0=false and not-0=true, but is not correct code.
1649
7e634d24
PH
165068. The host_aton() function was not handling scoped IPv6 addresses (those
1651 with, for example, "%eth0" on the end) correctly.
1652
3e11c26b
PH
165369. Fixed some compiler warnings in acl.c for the bitmaps specified with
1654 negated items (that is, ~something) in unsigned ints. Some compilers
1655 apparently mutter when there is no cast.
1656
6729cf78
PH
165770. If an address verification called from an ACL failed, and did not produce a
1658 user-specific message (i.e. there was only a "system" message), nothing was
1659 put in $acl_verify_message. In this situation, it now puts the system
1660 message there.
1661
00f00ca5
PH
166271. Change 4.23/11 added synchronization checking at the start of an SMTP
1663 session; change 4.31/43 added the unwanted input to the log line - except
1664 that it did not do this in the start of session case. It now does.
1665
c9bdd01c
PH
166672. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1667 This is wrong and can cause the other end to generate a synchronization
1668 error if it is another Exim or anything else that does the synchronization
1669 check. A QUIT command is no longer sent after a timeout.
1670
d43194df
PH
167173. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1672 during host lookups.
1673
fe5b5d0b
PH
167474. The defer_ok option of callout verification was not working if it was used
1675 when verifying addresses in header lines, that is, for this case:
1676
1677 verify = header_sender/callout=defer_ok
1678
76a2d7ba
PH
167975. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1680 those file descriptors could be used for SMTP connections. If anything
1681 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1682 could be sent to the SMTP client, causing chaos. The daemon now opens
1683 stdin, stdout, and stderr to /dev/null when it puts itself into the
1684 background.
1685
168676. Arrange for output from Perl's "warn" command to be written to Exim's main
1687 log by default. The user can override this with suitable Perl magic.
1688
04f7d5b9
PH
168977. The use of log_message on a "discard" ACL verb, which is supposed to add to
1690 the log message when discard triggers, was not working for the DATA ACL or
1691 for the non-SMTP ACL.
1692
bc60667e
PH
169378. Error message wording change in sieve.c.
1694
bb6e88ff
PH
169579. If smtp_accept_max_per_host was set, the number of connections could be
1696 restricted to fewer than expected, because the daemon was trying to set up
1697 a new connection before checking whether the processes handling previous
1698 connections had finished. The check for completed processes is now done
1699 earlier. On busy systems, this bug wouldn't be noticed because something
1700 else would have woken the daemon, and it would have reaped the completed
1701 process earlier.
1702
1e70f85b
PH
170380. If a message was submitted locally by a user whose login name contained one
1704 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1705 It caused a spool format error. I have fixed the spool reading code. A
1706 related problem was that the "from" clause in the Received: line became
1707 illegal because of the space(s). It is now covered by ${quote_local_part.
1708
170981. Included the latest eximstats from Steve (adds average sizes to HTML Top
1710 tables).
1711
4e01f9d6
PH
171282. Updated OS/Makefile-AIX as per message from Mike Meredith.
1713
1ee1cef2
PH
171483. Patch from Sieve maintainer to fix unterminated string problem in
1715 "vacation" handling.
1716
6e2b4ccc
PH
171784. Some minor changes to the Linux configuration files to help with other
1718 OS variants using glibc.
1719
8e669ac1
PH
172085. One more patch for Sieve to update vacation handling to latest spec.
1721
495ae4b0 1722
bbe902f0
PH
1723----------------------------------------------------
1724See the note above about the 4.44 and 4.50 releases.
1725----------------------------------------------------
1726
1727
1728Exim version 4.44
1729-----------------
1730
1731 1. Change 4.43/35 introduced a bug that caused file counts to be
1732 incorrectly computed when quota_filecount was set in an appendfile
1733 transport
1734
1735 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1736 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1737 place.
1738
1739 3. Give more explanation in the error message when the command for a transport
1740 filter fails to execute.
1741
1742 4. There are several places where Exim runs a non-Exim command in a
1743 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1744 was being done only for the command run by the queryprogram router. It is
1745 now done for all such subprocesses. The other cases are: ${run, transport
1746 filters, and the commands run by the lmtp and pipe transports.
1747
1748 5. Some older OS have a limit of 256 on the maximum number of file
1749 descriptors. Exim was using setrlimit() to set 1000 as a large value
1750 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1751 systems. I've change it so that if it can't get 1000, it tries for 256.
1752
1753 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1754 was an oversight, and furthermore, ever since the addition of extra
1755 controls (e.g. 4.43/32), the checks on when to allow different forms of
1756 "control" were broken. There should now be diagnostics for all cases when a
1757 control that does not make sense is encountered.
1758
1759 7. $recipients is now available in the predata ACL (oversight).
1760
1761 8. Tidy the search cache before the fork to do a delivery from a message
1762 received from the command line. Otherwise the child will trigger a lookup
1763 failure and thereby defer the delivery if it tries to use (for example) a
1764 cached ldap connection that the parent has called unbind on.
1765
1766 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1767 of $address_data from the recipient verification was clobbered by the
1768 sender verification.
1769
177010. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1771 was its contents. (It was OK if the option was not defined at all.)
1772
177311. A "Completed" log line is now written for messages that are removed from
1774 the spool by the -Mrm option.
1775
177612. $host_address is now set to the target address during the checking of
1777 ignore_target_hosts.
1778
177913. When checking ignore_target_hosts for an ipliteral router, no host name was
1780 being passed; this would have caused $sender_host_name to have been used if
1781 matching the list had actually called for a host name (not very likely,
1782 since this list is usually IP addresses). A host name is now passed as
1783 "[x.x.x.x]".
1784
178514. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1786 code that specifies a non-restarting handler (typically sigaction() in
1787 modern systems) in an attempt to fix a rare and obscure crash bug.
1788
178915. Narrowed the window for a race in the daemon that could cause it to ignore
1790 SIGCHLD signals. This is not a major problem, because they are used only to
1791 wake it up if nothing else does.
1792
179316. A malformed maildirsize file could cause Exim to calculate negative values
1794 for the mailbox size or file count. Odd effects could occur as a result.
1795 The maildirsize information is now recalculated if the size or filecount
1796 end up negative.
1797
179817. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1799 support for a long time. Removed HAVE_SYS_VFS_H.
1800
ea3a6f44 180118. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1802
180319. Allow an empty sender to be matched against a lookup in an address list.
1804 Previously the only cases considered were a regular expression, or an
1805 empty pattern.
1806
180720. Exim went into a mad DNS lookup loop when doing a callout where the
1808 host was specified on the transport, if the DNS lookup yielded more than
1809 one IP address.
1810
ea3a6f44
NM
181121. The RFC2047 encoding function was originally intended for short strings
1812 such as real names; it was not keeping to the 75-character limit for
1813 encoded words that the RFC imposes. It now respects the limit, and
1814 generates multiple encoded words if necessary. To be on the safe side, I
1815 have increased the buffer size for the ${rfc2047: expansion operator from
1816 1024 to 2048 bytes.
bbe902f0 1817
ea3a6f44
NM
181822. Failure to deliver a bounce message always caused it to be frozen, even if
1819 there was an errors_to setting on the router. The errors_to setting is now
1820 respected.
bbe902f0
PH
1821
182223. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1823 canonical form (fully expanded) before being placed in
1824 $sender_host_address.
1825
182624. Updated eximstats to version 1.33
1827
ea3a6f44
NM
182825. Include certificate and key file names in error message when GnuTLS fails
1829 to set them up, because the GnuTLS error message doesn't include the name
1830 of the failing file when there is a problem reading it.
bbe902f0
PH
1831
183226. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1833 As per change 25.
bbe902f0 1834
ea3a6f44
NM
183527. Reset the locale to "C" after calling embedded Perl, in case it was changed
1836 (this can affect the format of dates).
bbe902f0 1837
ea3a6f44
NM
183828. exim_tidydb, when checking for the continued existence of a message for
1839 which it has found a message-specific retry record, was not finding
1840 messages that were in split spool directories. Consequently, it was
1841 deleting retry records that should have stayed in existence.
bbe902f0
PH
1842
184329. eximstats updated to version 1.35
1844 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1845 1.35 - bugfix such that pie charts by volume are generated correctly
1846
ea3a6f44
NM
184730. The SPA authentication driver was not abandoning authentication and moving
1848 on to the next authenticator when an expansion was forced to fail,
1849 contradicting the general specification for all authenticators. Instead it
1850 was generating a temporary error. It now behaves as specified.
bbe902f0 1851
ea3a6f44
NM
185231. The default ordering of permitted cipher suites for GnuTLS was pessimal
1853 (the order specifies the preference for clients). The order is now AES256,
1854 AES128, 3DES, ARCFOUR128.
bbe902f0 1855
ea3a6f44
NM
185631. Small patch to Sieve code - explicitly set From: when generating an
1857 autoreply.
bbe902f0 1858
ea3a6f44
NM
185932. Exim crashed if a remote delivery caused a very long error message to be
1860 recorded - for instance if somebody sent an entire SpamAssassin report back
1861 as a large number of 550 error lines. This bug was coincidentally fixed by
1862 increasing the size of one of Exim's internal buffers (big_buffer) that
1863 happened as part of the Exiscan merge. However, to be on the safe side, I
1864 have made the code more robust (and fixed the comments that describe what
1865 is going on).
bbe902f0 1866
ea3a6f44
NM
186733. Some experimental protocols are using DNS PTR records for new purposes. The
1868 keys for these records are domain names, not reversed IP addresses. The
1869 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1870 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1871 CAN-2005-0021
1872
ea3a6f44
NM
187334. The host_aton() function is supposed to be passed a string that is known
1874 to be a valid IP address. However, in the case of IPv6 addresses, it was
1875 not checking this. This is a hostage to fortune. Exim now panics and dies
1876 if the condition is not met. A case was found where this could be provoked
1877 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1878 components; fortuitously, this particular loophole had already been fixed
1879 by change 4.50/55 or 4.44/33 above.
1880
1881 If there are any other similar loopholes, the new check in host_aton()
1882 itself should stop them being exploited. The report I received stated that
1883 data on the command line could provoke the exploit when Exim was running as
1884 exim, but did not say which command line option was involved. All I could
1885 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1886 running as the user.
bbe902f0
PH
1887 CAN-2005-0021
1888
ea3a6f44
NM
188935. There was a buffer overflow vulnerability in the SPA authentication code
1890 (which came originally from the Samba project). I have added a test to the
1891 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1892 CAN-2005-0022
1893
ea3a6f44
NM
189436. The daemon start-up calls getloadavg() while still root for those OS that
1895 need the first call to be done as root, but it missed one case: when
1896 deliver_queue_load_max is set with deliver_drop_privilege. This is
1897 necessary for the benefit of the queue runner, because there is no re-exec
1898 when deliver_drop_privilege is set.
bbe902f0 1899
ea3a6f44
NM
190037. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1901 was in use, was not putting the data itself into the right store pool;
1902 consequently, it could be overwritten for a subsequent message in the same
1903 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1904 the caching.)
bbe902f0 1905
ea3a6f44
NM
190638. Sometimes the final signoff response after QUIT could fail to get
1907 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1908 < 0 before doing a fflush(). This bug looks as though it goes back to the
1909 introduction of TLS in release 3.20, but "sometimes" must have been rare
1910 because the tests only now provoked it.
bbe902f0
PH
1911
1912
495ae4b0
PH
1913Exim version 4.43
1914-----------------
1915
1916 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1917 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1918 Now it returns an integer. A number of calls were still expecting a T/F
1919 return. Fortuitously, in all cases, the tests worked in OK situations,
1920 which is the norm. However, things would have gone wrong on any write
1921 failures on the smtp file descriptor. This function is used when sending
1922 messages over SMTP and also when doing verify callouts.
1923
1924 2. When Exim is called to do synchronous delivery of a locally submitted
1925 message (the -odf or -odi options), it no longer closes stderr before doing
1926 the delivery.
1927
1928 3. Implemented the mua_wrapper option.
1929
1930 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1931
1932 5. Implemented the functions header_remove(), header_testname(),
1933 header_add_at_position(), and receive_remove_recipient(), and exported them
1934 to local_scan().
1935
1936 6. If an ACL "warn" statement specified the addition of headers, Exim already
1937 inserted X-ACL-Warn: at the start if there was no header name. However, it
1938 was not making this test for the second and subsequent header lines if
1939 there were newlines in the string. This meant that an invalid header could
1940 be inserted if Exim was badly configured.
1941
1942 7. Allow an ACL "warn" statement to add header lines at the start or after all
1943 the Received: headers, as well as at the end.
1944
1945 8. Added the rcpt_4xx retry error code.
1946
1947 9. Added postmaster_mailfrom=xxx to callout verification option.
1948
194910. Added mailfrom=xxxx to the callout verification option, for verify=
1950 header_sender only.
1951
195211. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1953 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1954
195512. Inserted some casts to stop certain compilers warning when using pointer
1956 differences as field lengths or precisions in printf-type calls (mostly
1957 affecting debugging statements).
1958
195913. Added optional readline() support for -be (dynamically loaded).
1960
196114. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1962 same clock tick as a message's arrival, so that its received time was the
1963 same as the "first fail" time on the retry record, and that message
1964 remained on the queue past the ultimate address timeout, every queue runner
1965 would try a delivery (because it was past the ultimate address timeout) but
1966 after another failure, the ultimate address timeout, which should have then
1967 bounced the address, did not kick in. This was a "< instead of <=" error;
1968 in most cases the first failure would have been in the next clock tick
1969 after the received time, and all would be well.
1970
197115. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1972 being recognized when the domain list was tested by the match_domain
1973 condition in an expansion string.
1974
197516. Added the ${str2b64: operator.
1976
197717. Exim was always calling setrlimit() to set a large limit for the number of
1978 processes, without checking whether the existing limit was already
1979 adequate. (It did check for the limit on file descriptors.) Furthermore,
1980 errors from getrlimit() and setrlimit() were being ignored. Now they are
1981 logged to the main and panic logs, but Exim does carry on, to try to do its
1982 job under whatever limits there are.
1983
198418. Imported PCRE 5.0.
1985
198619. Trivial typo in log message " temporarily refused connection" (the leading
1987 space).
1988
198920. If the log selector return_path_on_delivery was set and an address was
1990 redirected to /dev/null, the delivery process crashed because it assumed
1991 that a return path would always be set for a "successful" delivery. In this
1992 case, the whole delivery is bypassed as an optimization, and therefore no
1993 return path is set.
1994
199521. Internal re-arrangement: the function for sending a challenge and reading
1996 a response while authentication was assuming a zero-terminated challenge
1997 string. It's now changed to take a pointer and a length, to allow for
1998 binary data in such strings.
1999
200022. Added the cyrus_sasl authenticator (code supplied by MBM).
2001
200223. Exim was not respecting finduser_retries when seeking the login of the
2003 uid under which it was called; it was always trying 10 times. (The default
2004 setting of finduser_retries is zero.) Also, it was sleeping after the final
2005 failure, which is pointless.
2006
200724. Implemented tls_on_connect_ports.
2008
200925. Implemented acl_smtp_predata.
2010
201126. If the domain in control=submission is set empty, Exim assumes that the
2012 authenticated id is a complete email address when it generates From: or
2013 Sender: header lines.
2014
201527. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2016 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2017 chown and chgrp in /bin and hostname in /usr/bin.
2018
201928. Exim was keeping the "process log" file open after each use, just as it
2020 does for the main log. This opens the possibility of it remaining open for
2021 long periods when the USR1 signal hits a daemon. Occasional processlog
2022 errors were reported, that could have been caused by this. Anyway, it seems
2023 much more sensible not to leave this file open at all, so that is what now
2024 happens.
2025
202629. The long-running daemon process does not normally write to the log once it
2027 has entered its main loop, and it closes the log before doing so. This is
2028 so that log files can straightforwardly be renamed and moved. However,
2029 there are a couple of unusual error situations where the daemon does write
2030 log entries, and I had neglected to close the log afterwards.
2031
203230. The text of an SMTP error response that was received during a remote
2033 delivery was being truncated at 512 bytes. This is too short for some of
2034 the long messages that one sometimes sees. I've increased the limit to
2035 1024.
2036
203731. It is now possible to make retry rules that apply only when a message has a
2038 specific sender, in particular, an empty sender.
2039
204032. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2041 it possible to be selective about when SMTP synchronization is enforced.
2042
204333. Added "control = caseful_local_part" and "control = "caselower_local_part".
2044
204532. Implemented hosts_connection_nolog.
2046
204733. Added an ACL for QUIT.
2048
204934. Setting "delay_warning=" to disable warnings was not working; it gave a
2050 syntax error.
2051
205235. Added mailbox_size and mailbox_filecount to appendfile.
2053
205436. Added control = no_multiline_responses to ACLs.
2055
205637. There was a bug in the logic of the code that waits for the clock to tick
2057 in the case where the clock went backwards by a substantial amount such
2058 that the microsecond fraction of "now" was more than the microsecond
2059 fraction of "then" (but the whole seconds number was less).
2060
206138. Added support for the libradius Radius client library this is found on
2062 FreeBSD (previously only the radiusclient library was supported).
2063
2064
2065Exim version 4.42
2066-----------------
2067
2068 1. When certain lookups returned multiple values in the form name=value, the
2069 quoting of the values was not always being done properly. Specifically:
2070 (a) If the value started with a double quote, but contained no whitespace,
2071 it was not quoted.
2072 (b) If the value contained whitespace other than a space character (i.e.
2073 tabs or newlines or carriage returns) it was not quoted.
2074 This fix has been applied to the mysql and pgsql lookups by writing a
2075 separate quoting function and calling it from the lookup code. The fix
2076 should probably also be applied to nisplus, ibase and oracle lookups, but
2077 since I cannot test any of those, I have not disturbed their existing code.
2078
2079 2. A hit in the callout cache for a specific address caused a log line with no
2080 reason for rejecting RCPT. Now it says "Previous (cached) callout
2081 verification failure".
2082
2083 3. There was an off-by-one bug in the queryprogram router. An over-long
2084 return line was truncated at 256 instead of 255 characters, thereby
2085 overflowing its buffer with the terminating zero. As well as fixing this, I
2086 have increased the buffer size to 1024 (and made a note to document this).
2087
2088 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2089 when Exim is waiting for an SMTP response from a remote server, Exim
2090 restarts its select() call on the socket, thereby resetting its timeout.
2091 This is not a problem when such interrupts are rare. Somebody set up a cron
2092 job to run exiwhat every 2 minutes, which is less than the normal select()
2093 timeout (5 or 10 minutes). This meant that the select() timeout never
2094 kicked in because it was always reset. I have fixed this by comparing the
2095 time when an interrupt arrives with the time at the start of the first call
2096 to select(). If more time than the timeout has elapsed, the interrupt is
2097 treated as a timeout.
2098
2099 5. Some internal re-factoring in preparation for the addition of Sieve
2100 extensions (by MH). In particular, the "personal" test is moved to a
2101 separate function, and given an option for scanning Cc: and Bcc: (which is
2102 not set for Exim filters).
2103
2104 6. When Exim created an email address using the login of the caller as the
2105 local part (e.g. when creating a From: or Sender: header line), it was not
2106 quoting the local part when it contained special characters such as @.
2107
2108 7. Installed new OpenBSD configuration files.
2109
2110 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2111 try to make them clearer.
2112
2113 9. Callout options, other than the timeout value, were being ignored when
2114 verifying sender addresses in header lines. For example, when using
2115
2116 verify = header_sender/callout=no_cache
2117
2118 the cache was (incorrectly) being used.
2119
212010. Added a missing instance of ${EXE} to the exim_install script; this affects
2121 only the Cygwin environment.
2122
212311. When return_path_on_delivery was set as a log selector, if different remote
2124 addresses in the same message used different return paths and parallel
2125 remote delivery occurred, the wrong values would sometimes be logged.
2126 (Whenever a remote delivery process finished, the return path value from
2127 the most recently started remote delivery process was logged.)
2128
212912. RFC 3848 specifies standard names for the "with" phrase in Received: header
2130 lines when AUTH and/or TLS are in use. This is the "received protocol"
2131 field. Exim used to use "asmtp" for authenticated SMTP, without any
2132 indication (in the protocol name) for TLS use. Now it follows the RFC and
2133 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2134 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2135 names appear in log lines as well as in Received: header lines.
2136
213713. Installed MH's patches for Sieve to add the "copy" and "vacation"
2138 extensions, and comparison tests, and to fix some bugs.
2139
214014. Changes to the "personal" filter test:
2141
2142 (1) The test was buggy in that it was just doing the equivalent of
2143 "contains" tests on header lines. For example, if a user's address was
2144 anne@some.where, the "personal" test would incorrectly be true for
2145
2146 To: susanne@some.where
2147
2148 This test is now done by extracting each address from the header in turn,
2149 and checking the entire address. Other tests that are part of "personal"
2150 are now done using regular expressions (for example, to check local parts
2151 of addresses in From: header lines).
2152
2153 (2) The list of non-personal local parts in From: addresses has been
2154 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2155 taken from the Sieve specification recommendations.
2156
2157 (3) If the message contains any header line starting with "List-" it is
2158 treated as non-personal.
2159
2160 (4) The test for "circular" in the Subject: header line has been removed
2161 because it now seems ill-conceived.
2162
216315. Minor typos in src/EDITME comments corrected.
2164
216516. Installed latest exipick from John Jetmore.
2166
216717. If headers_add on a router specified a text string that was too long for
2168 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2169 of string_sprintf() is now avoided.
2170
217118. $message_body_size was not set (it was always zero) when running the DATA
2172 ACL and the local_scan() function.
2173
217419. For the "mail" command in an Exim filter, no default was being set for
2175 the once_repeat time, causing a random time value to be used if "once" was
2176 specified. (If the value happened to be <= 0, no repeat happened.) The
2177 default is now 0s, meaning "never repeat". The "vacation" command was OK
2178 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2179 (I found it when inspecting the code).
2180
218120. There is now an overall timeout for performing a callout verification. It
2182 defaults to 4 times the callout timeout, which applies to individual SMTP
2183 commands during the callout. The overall timeout applies when there is more
2184 than one host that can be tried. The timeout is checked before trying the
2185 next host. This prevents very long delays if there are a large number of
2186 hosts and all are timing out (e.g. when the network connections are timing
2187 out). The value of the overall timeout can be changed by specifying an
2188 additional sub-option for "callout", called "maxwait". For example:
2189
2190 verify = sender/callout=5s,maxwait=20s
2191
219221. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2193 to the end before writing, but this should make it even safer).
2194
219522. Exim was forgetting that it had advertised PIPELINING for the second and
2196 subsequent messages on an SMTP connection. It was also not resetting its
2197 memory on STARTTLS and an internal HELO.
2198
219923. When Exim logs an SMTP synchronization error within a session, it now
2200 records whether PIPELINING has been advertised or not.
2201
220224. Added 3 instances of "(long int)" casts to time_t variables that were being
2203 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2204 rather than long int.
2205
220625. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2207
220826. Added the never_mail option to autoreply.
2209
2210
2211Exim version 4.41
2212-----------------
2213
2214 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2215 crash if the getsockname() call failed; this can happen if a connection is
2216 closed very soon after it is established. The problem was simply in the
2217 order in which certain operations were done, causing Exim to try to write
2218 to the SMTP stream before it had set up the file descriptor. The bug has
2219 been fixed by making things happen in the correct order.
2220
2221
2222Exim version 4.40
2223-----------------
2224
2225 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2226 before the connection was closed, thus losing the rejection response.
2227
2228 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2229 some early Solaris releases, but causes trouble in current releases where
2230 socklen_t is defined.
2231
2232 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2233 always exist.
2234
2235 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2236 configured.
2237
2238 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2239 that releasing the top of it at the end releases what was used for sub-
2240 expansions (unless the block got too big). However, discard this block if
2241 the first thing is a variable or header, so that we can use its block when
2242 it is dynamic (useful for very large $message_headers, for example).
2243
2244 6. Lookups now cache *every* query, not just the most recent. A new, separate
2245 store pool is used for this. It can be recovered when all lookup caches are
2246 flushed. Lookups now release memory at the end of their result strings.
2247 This has involved some general refactoring of the lookup sources.
2248
2249 7. Some code has been added to the store_xxx() functions to reduce the amount
2250 of flapping under certain conditions.
2251
2252 8. log_incoming_interface used to affect only the <= reception log lines. Now
2253 it causes the local interface and port to be added to several more SMTP log
2254 lines, for example "SMTP connection from", and rejection lines.
2255
2256 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2257
225810. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2259
226011. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2261 could be overwritten at the end of the current message (or the start of a
2262 new message if it was set in a HELO ACL). The value is now preserved for
2263 the duration of the SMTP connection.
2264
226512. If a transport had a headers_rewrite setting, and a matching header line
2266 contained an unqualified address, that address was qualified, even if it
2267 did not match any rewriting rules. The underlying bug was that the values
2268 of the flags that permit the existence of unqualified sender and recipient
2269 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2270 non-local messages, and by -bnq for local messages) were not being
2271 preserved with the message after it was received.
2272
227313. When Exim was logging an SMTP synchronization error, it could sometimes log
2274 "next input=" as part of the text comprising the host identity instead of
2275 the correct text. The code was using the same buffer for two different
2276 strings. However, depending on which order the printing function evaluated
2277 its arguments, the bug did not always show up. Under Linux, for example, my
2278 test suite worked just fine.
2279
228014. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2281 doesn't work with some older versions of Perl. It has been changed to "my",
2282 which in any case is probably the better facility to use.
2283
228415. A really picky compiler found some instances of statements for creating
2285 error messages that either had too many or two few arguments for the format
2286 string.
2287
228816. The size of the buffer for calls to the DNS resolver has been increased
2289 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2290 for addresses that have a lot of PTR records. This alleviates a problem; it
2291 does not fully solve it.
2292
229317. A dnsdb lookup for PTR records that receives more data than will fit in the
2294 buffer now truncates the list and logs the incident, which is the same
2295 action as happens when Exim is looking up a host name and its aliases.
2296 Previously in this situation something unpredictable would happen;
2297 sometimes it was "internal error: store_reset failed".
2298
229918. If a server dropped the connection unexpectedly when an Exim client was
2300 using GnuTLS and trying to read a response, the client delivery process
2301 crashed while trying to generate an error log message.
2302
230319. If a "warn" verb in an ACL added multiple headers to a message in a single
2304 string, for example:
2305
2306 warn message = H1: something\nH2: something
2307
2308 the text was added as a single header line from Exim's point of view
2309 though it ended up OK in the delivered message. However, searching for the
2310 second and subsequent header lines using $h_h2: did not work. This has been
2311 fixed. Similarly, if a system filter added multiple headers in this way,
2312 the routers could not see them.
2313
231420. Expanded the error message when iplsearch is called with an invalid key to
2315 suggest using net-iplsearch in a host list.
2316
231721. When running tests using -bh, any delays imposed by "delay" modifiers in
2318 ACLs are no longer actually imposed (and a message to that effect is
2319 output).
2320
232122. If a "gecos" field in a passwd entry contained escaped characters, in
2322 particular, if it contained a \" sequence, Exim got it wrong when building
2323 a From: or a Sender: header from that name. A second bug also caused
2324 incorrect handling when an unquoted " was present following a character
2325 that needed quoting.
2326
232723. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2328 was not being matched caselessly.
2329
233024. Arranged for all hyphens in the exim.8 source to be escaped with
2331 backslashes.
2332
233325. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2334 properly. Recipient callout cache records were still being keyed to include
2335 the sender, even when use_sender was set false. This led to far more
2336 callouts that were necessary. The sender is no longer included in the key
2337 when use_sender is false.
2338
233926. Added "control = submission" modifier to ACLs.
2340
234127. Added the ${base62d: operator to decode base 62 numbers.
2342
234328. dnsdb lookups can now access SRV records.
2344
234529. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2346 the configuration file.
2347
234830. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2349 (-v) mode. This makes the output for a verbose queue run more intelligible.
2350
235131. Added a use_postmaster feature to recipient callouts.
2352
235332. Added the $body_zerocount variable, containing the number of binary zero
2354 bytes in the message body.
2355
235633. The time of last modification of the "new" subdirectory is now used as the
2357 "mailbox time last read" when there is a quota error for a maildir
2358 delivery.
2359
236034. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2361
236235. Added +ignore_unknown as a special item in host lists.
2363
236436. Code for decoding IPv6 addresses in host lists is now included, even if
2365 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2366 address was recognized as an IP address, but was then not correctly decoded
2367 into binary, causing unexpected and incorrect effects when compared with
2368 another IP address.
2369
2370
2371Exim version 4.34
2372-----------------
2373
2374 1. Very minor rewording of debugging text in manualroute to say "list of
2375 hosts" instead of "hostlist".
2376
2377 2. If verify=header_syntax was set, and a header line with an unqualified
2378 address (no domain) and a large number of spaces between the end of the
2379 name and the colon was received, the reception process suffered a buffer
2380 overflow, and (when I tested it) crashed. This was caused by some obsolete
2381 code that should have been removed. The fix is to remove it!
2382
2383 3. When running in the test harness, delay a bit after writing a bounce
2384 message to get a bit more predictability in the log output.
2385
2386 4. Added a call to search_tidyup() just before forking a reception process. In
2387 theory, someone could use a lookup in the expansion of smtp_accept_max_
2388 per_host which, without the tidyup, could leave open a database connection.
2389
2390 5. Added the variables $recipient_data and $sender_data which get set from a
2391 lookup success in an ACL "recipients" or "senders" condition, or a router
2392 "senders" option, similar to $domain_data and $local_part_data.
2393
2394 6. Moved the writing of debug_print from before to after the "senders" test
2395 for routers.
2396
2397 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2398 problems for message scanning, either using a data ACL, or using
2399 local_scan() because the Received: header was not generated till after they
2400 were called (in order to set the time as the time of reception completion).
2401 I have revised the way this works. The header is now generated after the
2402 body is received, but before the ACL or local_scan() are called. After they
2403 are run, the timestamp in the header is updated.
2404
2405
2406Exim version 4.33
2407-----------------
2408
2409 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2410 before starting a queue runner without re-exec. This happened only when
2411 deliver_drop_privilege was set or when the Exim user was set to root. The
2412 effect of the bug was that timeouts during subsequent deliveries caused
2413 crashes instead of being properly handled. The handler is now left at its
2414 default (and expected) setting.
2415
2416 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2417 message, again when deliver_drop_privilege is set or Exim is run as root.
2418 The bug described in (1) was not present in this case, but the tidying up
2419 of the other signals was missing. I have made the two cases consistent.
2420
2421 3. The ignore_target_hosts setting on a manualroute router was being ignored
2422 for hosts that were looked up using the /MX notation.
2423
2424 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2425 in domain lists.
2426
2427 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2428 operated on the sender address. After changing the $sender_address to <>
2429 for the sender address verify, Exim was re-instated it as the original
2430 (before rewriting) address, but remembering that it had rewritten it, so it
2431 wasn't rewriting it again. This bug also had the effect of breaking the
2432 sender address verification caching when the sender address was rewritten.
2433
2434 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2435 This has been changed so that if the ip literal address matches
2436 ignore_target_hosts, the router declines.
2437
2438 7. Added expansion conditions match_domain, match_address, and match_local_
2439 part (NOT match_host).
2440
2441 8. The placeholder for the Received: header didn't have a length field set.
2442
2443 9. Added code to Exim itself and to exim_lock to test for a specific race
2444 condition that could lead to file corruption when using MBX delivery. The
2445 issue is with the lockfile that is created in /tmp. If this file is removed
2446 after a process has opened it but before that process has acquired a lock,
2447 there is the potential for a second process to recreate the file and also
2448 acquire a lock. This could lead to two Exim processes writing to the file
2449 at the same time. The added code performs the same test as UW imapd; it
2450 checks after acquiring the lock that its file descriptor still refers to
2451 the same named file.
2452
245310. The buffer for building added header lines was of fixed size, 8192 bytes.
2454 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2455 when Exim is built.
2456
245711. Added the smtp_active_hostname option. If used, this will typically be made
2458 to depend on the incoming interface address. Because $interface_address is
2459 not set up until the daemon has forked a reception process, error responses
2460 that can happen earlier (such as "too many connections") no longer contain
2461 a host name.
2462
246312. If an expansion in a condition on a "warn" statement fails because a lookup
2464 defers, the "warn" statement is abandoned, and the next ACL statement is
2465 processed. Previously this caused the whole ACL to be aborted.
2466
246713. Added the iplsearch lookup type.
2468
246914. Added ident_timeout as a log selector.
2470
247115. Added tls_certificate_verified as a log selector.
2472
247316. Added a global option tls_require_ciphers (compare the smtp transport
2474 option of the same name). This controls incoming TLS connections.
2475
247617. I finally figured out how to make tls_require_ciphers do a similar thing
2477 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2478 before starting the TLS session.
2479
248018. Tabs are now shown as \t in -bP output.
2481
248219. If the log selector return_path_on_delivery was set, Exim crashed when
2483 bouncing a message because it had too many Received: header lines.
2484
248520. If two routers both had headers_remove settings, and the first one included
2486 a superfluous trailing colon, the final name in the first list and the
2487 first name in the second list were incorrectly joined into one item (with a
2488 colon in the middle).
2489
2490
2491Exim version 4.32
2492-----------------
2493
2494 1. Added -C and -D options to the exinext utility, mainly to make it easier
2495 to include in the automated testing, but these could be helpful when
2496 multiple configurations are in use.
2497
2498 2. The exinext utility was not formatting the output nicely when there was
2499 an alternate port involved in the retry record key, nor when there was a
2500 message id as well (for retries that were specific to a specific message
2501 and a specific host). It was also confused by IPv6 addresses, because of
2502 the additional colons they contain. I have fixed the IPv4 problem, and
2503 patched it up to do a reasonable job for IPv6.
2504
2505 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2506 delivery, the log line now contains "pipelined" if PIPELINING was used.
2507
2508 4. An SMTP transport process used to panic and die if the bind() call to set
2509 an explicit outgoing interface failed. This has been changed; it is now
2510 treated in the same way as a connect() failure.
2511
2512 5. A reference to $sender_host_name in the part of a conditional expansion
2513 that was being skipped was still causing a DNS lookup. This no longer
2514 occurs.
2515
2516 6. The def: expansion condition was not recognizing references to header lines
2517 that used bh_ and bheader_.
2518
2519 7. Added the _cache feature to named lists.
2520
2521 8. The code for checking quota_filecount in the appendfile transport was
2522 allowing one more file than it should have been.
2523
2524 9. For compatibility with Sendmail, the command line option
2525
2526 -prval:sval
2527
2528 is equivalent to
2529
2530 -oMr rval -oMs sval
2531
2532 and sets the incoming protocol and host name (for trusted callers). The
2533 host name and its colon can be omitted when only the protocol is to be set.
2534 Note the Exim already has two private options, -pd and -ps, that refer to
2535 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2536 "s", but I don't think that's a major issue.
2537
253810. A number of refactoring changes to the code, none of which should affect
2539 Exim's behaviour:
2540
2541 (a) The number of logging options was getting close to filling up the
2542 32-bit word that was used as a bit map. I have split them into two classes:
2543 those that are passed in the argument to log_write(), and those that are
2544 only ever tested independently outside of that function. These are now in
2545 separate 32-bit words, so there is plenty of room for expansion again.
2546 There is no change in the user interface or the logging behaviour.
2547
2548 (b) When building, for example, log lines, the code previously used a
2549 macro that called string_cat() twice, in order to add two strings. This is
2550 not really sufficiently general. Furthermore, there was one instance where
2551 it was actually wrong because one of the argument was used twice, and in
2552 one call a function was used. (As it happened, calling the function twice
2553 did not affect the overall behaviour.) The macro has been replaced by a
2554 function that can join an arbitrary number of extra strings onto a growing
2555 string.
2556
2557 (c) The code for expansion conditions now uses a table and a binary chop
2558 instead of a serial search (which was left over from when there were very
2559 few conditions). Also, it now recognizes conditions like "pam" even when
2560 the relevant support is not compiled in: a suitably worded error message is
2561 given if an attempt is made to use such a condition.
2562
256311. Added ${time_interval:xxxxx}.
2564
256512. A bug was causing one of the ddress fields not to be passed back correctly
2566 from remote delivery subprocesses. The field in question was not being
2567 subsequently used, so this caused to problems in practice.
2568
256913. Added new log selectors queue_time and deliver_time.
2570
257114. Might have fixed a bug in maildirsizefile handling that threw up
2572 "unexpected character" debug warnings, and recalculated the data
2573 unnecessarily. In any case, I expanded the warning message to give more
2574 information.
2575
257615. Added the message "Restricted characters in address" to the statements in
2577 the default ACL that block characters like @ and % in local parts.
2578
257916. Change 71 for release 4.31 proved to be much less benign that I imagined.
2580 Three changes have been made:
2581
2582 (a) There was a serious bug; a negative response to MAIL caused the whole
2583 recipient domain to be cached as invalid, thereby blocking all messages
2584 to all local parts at the same domain, from all senders. This bug has
2585 been fixed. The domain is no longer cached after a negative response to
2586 MAIL if the sender used is not empty.
2587
2588 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2589 been restored.
2590
2591 (c) A new callout option, "use_sender" has been added for people who want
2592 the modified behaviour.
2593
2594
2595Exim version 4.31
2596-----------------
2597
2598 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2599 Larry Rosenman.
2600
2601 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2602 indeed breaks things for older releases.
2603
2604 3. Added additional logging to the case where there is a problem reading data
2605 from a filter that is running in a subprocess using a pipe, in order to
2606 try to track down a specific problem.
2607
2608 4. Testing facility fudge: when running in the test harness and attempting
2609 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2610 getting "No route to host". Convert this to a timeout.
2611
2612 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2613 warning.
2614
2615 6. Some OS don't have socklen_t but use size_t instead. This affects the
2616 fifth argument of getsockopt() amongst other things. This is now
2617 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2618 can be set for individual OS. I have set it for SunOS5, OSF1, and
2619 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2620 some earlier ones do not.
2621
2622 7. Change 4.30/15 was not doing the test caselessly.
2623
2624 8. The standard form for an IPv6 address literal was being rejected by address
2625 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2626 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2627 this, as well as the form without the "IPv6" on the front (but only when
2628 address literals are enabled, of course).
2629
2630 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2631
263210. Exim crashed if a message with an empty sender address specified by -f
2633 encountered a router with an errors_to setting. This could be provoked only
2634 by a command such as
2635
2636 exim -f "" ...
2637
2638 where an empty string was supplied; "<>" did not hit this bug.
2639
264011. Installed PCRE release 4.5.
2641
264212. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2643 remained set. It is now erased.
2644
264513. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2646 times from message ids (which are base 36 rather than the normal 62).
2647
264814. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2649 were being counted as actual protocol errors, and logged if the log
2650 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2651 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2652 and DATA following a set of rejected RCPTs do not count as protocol errors.
2653 In other words, Exim assumes they were pipelined, though this may not
2654 actually be the case. Of course, in all cases the client gets an
2655 appropriate error code.
2656
265715. If a lookup fails in an ACL condition, a message about the failure may
2658 be available; it is used if testing the ACL cannot continue, because most
2659 such messages specify what the cause of the deferral is. However, some
2660 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2661 that caused an old message to be retained and used if a later statement
2662 caused a defer, replacing the real cause of the deferral.
2663
266416. If an IP address had so many PTR records that the DNS lookup buffer
2665 was not large enough to hold them, Exim could crash while trying to process
2666 the truncated data. It now detects and logs this case.
2667
266817. Further to 4.21/58, another change has been made: if (and only if) the
2669 first line of a message (the first header line) ends with CRLF, a bare LF
2670 in a subsequent header line has a space inserted after it, so as not to
2671 terminate the header.
2672
267318. Refactoring: tidied an ugly bit of code in appendfile that copied data
2674 unnecessarily, used atoi() instead of strtol(), and didn't check the
2675 termination when getting file sizes from file names by regex.
2676
267719. Completely re-implemented the support for maildirsize files, in the light
2678 of a number of problems with the previous contributed implementation
2679 (4.30/29). In particular:
2680
2681 . If the quota is zero, the maildirsize file is maintained, but no quota is
2682 imposed.
2683
2684 . If the maildir directory does not exist, it is created before any attempt
2685 to write a maildirsize file.
2686
2687 . The quota value in the file is just a cache; if the quota is changed in
2688 the transport, the new value overrides.
2689
2690 . A regular expression is available for excluding directories from the
2691 count.
2692
269320. The autoreply transport checks the characters in options that define the
2694 message's headers; it allows continued headers, but it was checking with
2695 isspace() after an embedded newline instead of explicitly looking for a
2696 space or a tab.
2697
269821. If all the "regular" hosts to which an address was routed had passed their
2699 expiry times, and had not reached their retry times, the address was
2700 bounced, even if fallback hosts were defined. Now Exim should go on to try
2701 the fallback hosts.
2702
270322. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2704 equivalent code in the SMTP transport. Some hosts send humungous responses
2705 to HELO/EHLO, more than 1024 it seems.
2706
270723. Refactoring: code in filter.c used (void *) for "any old type" but this
2708 gives compiler warnings in some environments. I've now done it "properly",
2709 using a union.
2710
271124. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2712 (because of problems with the built-in one) was declared to return uschar *
2713 instead of char *, causing compiler failure.
2714
271525. Fixed a file descriptor leak when processing alias/forward files.
2716
271726. Fixed a minor format string issue in dbfn.c.
2718
271927. Typo in exim.c: ("dmbnz" for "dbmnz").
2720
272128. If a filter file refered to $h_xxx or $message_headers, and the headers
2722 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2723 become corrupted.
2724
272529. When a sender address is verified, it is cached, to save repeating the test
2726 when there is more than one recipient in a message. However, when the
2727 verification involves a callout, it is possible for different callout
2728 options to be set for different recipients. It is too complicated to keep
2729 track of this in the cache, so now Exim always runs a verification when a
2730 callout is required, relying on the callout cache for the optimization.
2731 The overhead is duplication of the address routing, but this should not be
2732 too great.
2733
273430. Fixed a bug in callout caching. If a RCPT command caused the sender address
2735 to be verified with callout=postmaster, and the main callout worked but the
2736 postmaster check failed, the verification correctly failed. However, if a
2737 subsequent RCPT command asked for sender verification *without* the
2738 postmaster check, incorrect caching caused this verification also to fail,
2739 incorrectly.
2740
274131. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2742 it was not caching the DNS options (qualify_single, search_parents) that
2743 were used when the lookup failed. A subsequent lookup with different
2744 options therefore always gave the same answer, though there were cases
2745 where it should not have. (Example: a "domains = !$mx_any" option on a
2746 dnslookup router: the "domains" option is always processed without any
2747 widening, but the router might have qualify_single set.) Now Exim uses the
2748 cached value only when the same options are set.
2749
275032. Added John Jetmore's "exipick" utility to the distribution.
2751
275233. GnuTLS: When an attempt to start a TLS session fails for any reason other
2753 than a timeout (e.g. a certificate is required, and is not provided), an
2754 Exim server now closes the connection immediately. Previously it waited for
2755 the client to close - but if the client is SSL, it seems that they each
2756 wait for each other, leading to a delay before one of them times out.
2757
275834: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2759 maintained 0.8.x compatibility because I don't think many are using it, and
2760 it is clearly obsolete.
2761
276235. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2763 transport.
2764
276536. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2766 client certificate was expired. A simple patch fixes this, though I don't
2767 understand the full logic of why the verify callback is called multiple
2768 times.
2769
277037. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2771 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2772 which causes problems with some clients (such as the Certicom SSL Plus
2773 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2774 disables the coutermeasure allowing Eudora to connect."
2775
277638. Exim was not checking that a write() to a log file succeeded. This could
2777 lead to Bad Things if a log got too big, in particular if it hit a file
2778 size limit. Exim now panics and dies if it cannot write to a log file, just
2779 as it does if it cannot open a log file.
2780
278139. Modified OS/Makefile-Linux so that it now contains
2782
2783 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2784
2785 The two -D definitions ensure that Exim is compiled with large file
2786 support, which makes it possible to handle log files that are bigger than
2787 2^31.
2788
278940. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2790 instance) a domain was checked against a named list that involved a lookup,
2791 causing $domain_data to be set, then another domain was checked against the
2792 same list, then the first domain was re-checked, the value of $domain_data
2793 after the final check could be wrong. In particular, if the second check
2794 failed, it could be set empty. This bug probably also applied to
2795 $localpart_data.
2796
279741. The strip_trailing_dot option was not being applied to the address given
2798 with the -f command-line option.
2799
280042. The code for reading a message's header from the spool was incrementing
2801 $received_count, but never initializing it. This meant that the value was
2802 incorrect (doubled) while delivering a message in the same process in which
2803 it was received. In the most common configuration of Exim, this never
2804 happens - a fresh exec is done - but it can happen when
2805 deliver_drop_privilege is set.
2806
280743. When Exim logs an SMTP synchronization error - client data sent too soon -
2808 it now includes up to 150 characters of the unexpected data in the log
2809 line.
2810
281144. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2812 and building data strings. The size of both of these buffers was 10 000
2813 bytes - far larger than anybody would *ever* want, thought I. Needless to
2814 say, somebody hit the limit. I have increased the maximum line length to
2815 20 000 and the maximum data length of concatenated lines to 100 000. I have
2816 also fixed two bugs, because there was no checking on these buffers. Tsk,
2817 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2818 buffer is too small.
2819
282045. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2821 lsearch lookups. Now it does.
2822
282346. When parsing a route_list item in a manualroute router, a fixed-length
2824 buffer was used for the list of hosts. I made this 1024 bytes long,
2825 thinking that nobody would ever have a list of hosts that long. Wrong.
2826 Somebody had a whole pile of complicated expansion conditions, and the
2827 string was silently truncated, leading to an expansion error. It turns out
2828 that it is easier to change to an unlimited length (owing to other changes
2829 that have happened since this code was originally written) than to build
2830 structure for giving a limitation error. The length of the item that
2831 expands into the list of hosts is now unlimited.
2832
283347. The lsearch lookup could not handle data where the length of text line was
2834 more than 4095 characters. Such lines were truncated, leading to shortened
2835 data being returned. It should now handle lines of any length.
2836
283748. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2838 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2839 ACL").
2840
284149. Cosmetic tidy to scripts like exicyclog that are generated by globally
2842 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2843 no longer happens in comment lines. A list of replacements is now placed
2844 at the head of all of the source files, except those whose only change is
2845 to replace PERL_COMMAND in the very first #! line.
2846
284750. Replaced the slow insertion sort in queue.c, for sorting the list of
2848 messages on the queue, with a bottom-up merge sort, using code contributed
2849 by Michael Haardt. This should make operations like -bp somewhat faster on
2850 large queues. It won't affect queue runners, except when queue_run_in_order
2851 is set.
2852
285351. Installed eximstats 1.31 in the distribution.
2854
285552. Added support for SRV lookups to the dnslookup router.
2856
285753. If an ACL referred to $message_body or $message_body_end, the value was not
2858 reset for any messages that followed in the same SMTP session.
2859
286054. The store-handling optimization for building very long strings was not
2861 differentiating between the different store pools. I don't think this
2862 actually made any difference in practice, but I've tidied it.
2863
286455. While running the routers to verify a sender address, $sender_address
2865 was still set to the sender address. This is wrong, because when routing to
2866 send a bounce to the sender, it would be empty. Therefore, I have changed
2867 it so that, while verifying a sender address, $sender_address is set to <>.
2868 (There is no change to what happens when verifying a recipient address.)
2869
287056. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2871 target A or AAAA records (if not already returned) without resetting the
2872 qualify_single or search_parents options of the DNS resolver. These are
2873 inappropriate in this case because the targets of MX and SRV records must
2874 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2875 target that, when qualified, matched something in the local domain. These
2876 two options are now turned off when doing these lookups.
2877
287857. It seems that at least some releases of Reiserfs (which does not have the
2879 concept of a fixed number of inodes) returns zero and not -1 for the
2880 number of available inodes. This interacted badly with check_spool_inodes,
2881 which assumed that -1 was the "no such thing" setting. What I have done is
2882 to check that the total number of inodes is greater than zero before doing
2883 the test of how many are available.
2884
288558. When a "warn" ACL statement has a log_message modifier, the message is
2886 remembered, and not repeated. This is to avoid a lot of repetition when a
2887 message has many recipients that cause the same warning to be written.
2888 Howewer, Exim was preserving the list of already written lines for an
2889 entire SMTP session, which doesn't seem right. The memory is now reset if a
2890 new message is started.
2891
289259. The "rewrite" debugging flag was not showing the result of rewriting in the
2893 debugging output unless log_rewrite was also set.
2894
289560. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2896 of (int)(handle) when we know that handle contains (void *)(-1).
2897
289861. The Exim daemon panic-logs an error return when it closes the incoming
2899 connection. However "connection reset by peer" seems to be common, and
2900 isn't really an error worthy of noting specially, so that particular error
2901 is no long logged.
2902
290362. When Exim is trying to find all the local interfaces, it used to panic and
2904 die if the ioctl to get the interface flags failed. However, it seems that
2905 on at least one OS (Solaris 9) it is possible to have an interface that is
2906 included in the list of interfaces, but for which you get a failure error
2907 for this call. This happens when the interface is not "plumbed" into a
2908 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2909 failure of the "get flags" call assumes that the interface is down.
2910
291163. Added a ${eval10: operator, which assumes all numbers are decimal. This
2912 makes life easier for people who are doing arithmetic on fields extracted
2913 from dates, where you often get leading zeros that should not be
2914 interpreted as octal.
2915
291664. Added qualify_domain to the redirect router, to override the global
2917 setting.
2918
291965. If a pathologically long header line contained very many addresses (the
2920 report of this problem mentioned 10 000) and each of them was rewritten,
2921 Exim could use up a very large amount of memory. (It kept on making new
2922 copies of the header line as it rewrote, and never released the old ones.)
2923 At the expense of a bit more processing, the header rewriting function has
2924 been changed so that it no longer eats memory in this way.
2925
292666. The generation of the Received: header has been moved from the time that a
2927 message starts to be received, to the time that it finishes. The timestamp
2928 in the Received: header should now be very close to that of the <= log
2929 line. There are two side-effects of this change:
2930
2931 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2932 logged header lines no longer include the local Received: line, because
2933 it has not yet been created. The same applies to a copy of the message
2934 that is returned to a non-SMTP sender when a message is rejected.
2935
2936 (b) When a filter file is tested using -bf, no additional Received: header
2937 is added to the test message. After some thought, I decided that this
2938 is a bug fix.
2939
2940 This change does not affect the value of $received_for. It is still set
2941 after address rewriting, but before local_scan() is called.
2942
294367. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2944
294568. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2946 gave an unhelpful panic error message, and a defer error. I have managed to
2947 change this behaviour so that it now rejects any supplied certificate,
2948 which seems right, as the list of acceptable certificates is empty.
2949
295069. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2951 gave an unhelpful defer error. I have not managed to make this reject any
2952 supplied certificates, but the error message it gives is "no certificate
2953 supplied", which is not helpful.
2954
295570. exigrep's output now also includes lines that are not associated with any
2956 message, but which match the given pattern. Implemented by a patch from
2957 Martin Sluka, which also tidied up the Perl a bit.
2958
295971. Recipient callout verification, like sender verification, was using <> in
2960 the MAIL FROM command. This isn't really the right thing, since the actual
2961 sender may affect whether the remote host accepts the recipient or not. I
2962 have changed it to use the actual sender in the callout; this means that
2963 the cache record is now keyed on a recipient/sender pair, not just the
2964 recipient address. There doesn't seem to be a real danger of callout loops,
2965 since a callout by the remote host to check the sender would use <>.
2966 [SEE ABOVE: changed after hitting problems.]
2967
296872. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2969 temporary errors. However, in the case of such a code being given after
2970 the end of a data transmission (i.e. after ".") Exim was failing to write
2971 a retry record for the message. (Yes, there was some broken host that was
2972 actually sending 8xx at this point.)
2973
297473. An unknown lookup type in a host list could cause Exim to panic-die when
2975 the list was checked. (An example that provoked this was putting <; in the
2976 middle of a list instead of at the start.) If this happened during a DATA
2977 ACL check, a -D file could be left lying around. This kind of configuration
2978 error no longer causes Exim to die; instead it causes a defer errror. The
2979 incident is still logged to the main and panic logs.
2980
298174. Buglet left over from Exim 3 conversion. The message "too many messages
2982 in one connection" was written to the rejectlog but not the mainlog, except
2983 when address rewriting (yes!) was being logged.
2984
298575. Added write_rejectlog option.
2986
298776. When a system filter was run not as root (that is, when system_filter_user
2988 was set), the values of the $n variables were not being returned to the
2989 main process; thus, they were not subsequently available in the $sn
2990 variables.
2991
299277. Added +return_path_on_delivery log selector.
2993
299478. A connection timeout was being treated differently from recipients deferred
2995 when testing hosts_max_try with a message that was older than the host's
2996 retry timeout. (The host should not be counted, thus allowing all hosts to
2997 be tried at least once before bouncing.) This may have been the cause of an
2998 occasionally reported bug whereby a message would remain on the queue
2999 longer than the retry timeout, but would be bounced if a delivery was
3000 forced. I say "may" because I never totally pinned down the problem;
3001 setting up timeout/retry tests is difficult. See also the next item.
3002
300379. The ultimate address timeout was not being applied to errors that involved
3004 a combination of host plus message (for example, a timeout on a MAIL
3005 command). When an address resolved to a number of possible hosts, and they
3006 were not all tried for each delivery (e.g. because of hosts_max_try), a
3007 message could remain on the queue longer than the retry timeout.
3008
300980. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3010 Haardt.
3011
301281. Fixed an obscure SMTP outgoing bug which required at least the following
3013 conditions: (a) there was another message waiting for the same server;
3014 (b) the server returned 5xx to all RCPT commands in the first message so
3015 that the message was not completed; (c) the server dropped the connection
3016 or gave a negative response to the RSET that Exim sends to abort the
3017 transaction. The observed case was a dropped connection after DATA that had
3018 been sent in pipelining mode. That is, the server had advertised PIPELINING
3019 but was not implementing it correctly. The effect of the bug was incorrect
3020 behaviour, such as trying another host, and this could lead to a crash.
3021
3022
3023Exim version 4.30
3024-----------------
3025
3026 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3027 and daemon.c were passed as pointers to ints; they should have been
3028 pointers to socklen_t variables (which are typically unsigned ints).
3029
3030 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3031 fixed.
3032
3033 3. Fixed a really odd bug that affected only the testing scheme; patching a
3034 certain fixed string in the binary changed the value of another string that
3035 happened to be identical to the end of the original first string.
3036
3037 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3038 name", it returns that address as the IP address. On some operating
3039 systems (e.g. Solaris), it also passes back the IP address string as the
3040 "host name". However, on others (e.g. Linux), it passes back an empty
3041 string. Exim wasn't checking for this, and was changing the host name to an
3042 empty string, assuming it had been canonicized.
3043
3044 5. Although rare, it is permitted to have more than one PTR record for a given
3045 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3046 all the names associated with an address, because they do in Solaris.
3047 However, it seems that they do not in Linux for data that comes from the
3048 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3049 I found this out when I moved to a new Linux workstation and tried to run
3050 the Exim test suite.
3051
3052 To get round this problem I have changed the code so that it now does its
3053 own call to the DNS to look up PTR records when searching for a host name.
3054 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3055 addresses that are only in /etc/hosts are still found.
3056
3057 This behaviour is, however, controlled by an option called host_lookup_
3058 order, which defaults to "bydns:byaddr". If people want to use the other
3059 order, or indeed, just use one or the other means of lookup, they can
3060 specify it in this variable.
3061
3062 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3063 some operating systems, this comes back from gethostbyaddr() as an empty
3064 string, and this is what Exim used to test for. However, it seems that in
3065 other systems, "." is yielded. Exim now tests for this case too.
3066
3067 7. The values of check_spool_space and check_log_space are now held internally
3068 as a number of kilobytes instead of an absolute number of bytes. If a
3069 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3070 kilobyte. This means that much larger values can be stored.
3071
3072 8. Exim monitor: an attempt to get the action menu when not actually pointing
3073 at a message produces an empty menu entitled "No message selected". This
3074 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3075 no entries in it ("Shell widget menu has zero width and/or height"). So I
3076 have added a single, blank menu entry in this case.
3077
3078 9. Added ${quote_local_part.
3079
308010. MIME decoding is now applied to the contents of Subject: header lines when
3081 they are logged.
3082
308311. Now that a reference to $sender_host_address automatically causes a reverse
3084 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3085 host lookup before query-style lookups in lists that might use this
3086 variable. This has therefore been abolished, and the "net-" prefix is no
3087 longer necessary for query-style lookups.
3088
308912. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3090 have been a typo for LFLAGS, so it has been changed.
3091
309213. The install script calls Exim with "-C /dev/null" in order to find the
3093 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3094 to be output. Howeve, since Exim outputs its version number before the
3095 error, it didn't break the script. It just looked ugly. I fixed this by
3096 always allowing "-C /dev/null" if the caller is root.
3097
309814. Ignore overlarge ACL variable number when reading spool file - insurance
3099 against a later release with more variables having written the file.
3100
310115. The standard form for an IPv6 address literal was being rejected by EHLO.
3102 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3103 this, as well as the form without the "IPv6" on the front.
3104
310516. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3106 OS/Makefile-Darwin file.
3107
310817. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3109 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3110
311118. After change 4.21/52, "%ld" was used to format the contents of the $inode
3112 variable. However, some OS use ints for inodes. I've added cast to long int
3113 to get rid of the compiler warning.
3114
311519. I had forgotten to lock out "/../" in configuration file names when
3116 ALT_CONFIG_PREFIX was set.
3117
311820. Routers used for verification do not need to specify transports. However,
3119 if such a router generated a host list, and callout was configured, Exim
3120 crashed, because it could not find a port number from the (non-existent)
3121 transport. It now assumes port 25 in this circumstance.
3122
312321. Added the -t option to exigrep.
3124
312522. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3126 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3127 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3128 them is set, LOOKUP_LSEARCH is forced.
3129
313023. "exim -bV" now outputs a list of lookups that are included in the binary.
3131
313224. Added sender and host information to the "rejected by local_scan()" log
3133 line; previously there was no indication of these.
3134
313525. Added .include_if_exists.
3136
313726. Change 3.952/11 added an explicit directory sync on top of a file sync for
3138 Linux. It turns out that not all file systems support this. Apparently some
3139 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3140 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3141 not supported on the file descriptor, is now ignored when Exim is trying to
3142 sync a directory. This applies only to Linux.
3143
314427. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3145
314628. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3147 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3148 this.
3149
315029. Added support for maildirsize files from supplied patch (modified a bit).
3151
315230. The use of :fail: followed by an empty string could lead Exim to respond to
3153 sender verification failures with (e.g.):
3154
3155 550 Verification failed for <xxx>
3156 550 Sender verify failed
3157
3158 where the first response line was missing the '-' that indicates it is not
3159 the final line of the response.
3160
316131. The loop for finding the name of the user that called Exim had a hardwired
3162 limit of 10; it now uses the value of finduser_retries, which is used for
3163 all other user lookups.
3164
316532. Added $received_count variable, available in data and not_smtp ACLs, and at
3166 delivery time.
3167
316833. Exim was neglecting to zero errno before one call of strtol() when
3169 expanding a string and expecting an integer value. On some systems this
3170 resulted in spurious "integer overflow" errors. Also, it was casting the
3171 result into an int without checking.
3172
317334. Testing for a connection timeout using "timeout_connect" in the retry rules
3174 did not work. The code looks as if it has *never* worked, though it appears
3175 to have been documented since at least releast 1.62. I have made it work.
3176
317735. The "timeout_DNS" error in retry rules, also documented since at least
3178 1.62, also never worked. As it isn't clear exactly what this means, and
3179 clearly it isn't a major issue, I have abolished the feature by treating it
3180 as "timeout", and writing a warning to the main and panic logs.
3181
318236. The display of retry rules for -brt wasn't always showing the error code
3183 correctly.
3184
318537. Added new error conditions to retry rules: timeout_A, timeout_MX,
3186 timeout_connect_A, timeout_connect_MX.
3187
318838. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3189 to the empty sender.
3190
319139. The daemon was not analysing the content of -oX till after it had closed
3192 stderr and disconnected from the controlling terminal. This meant that any
3193 syntax errors were only noted on the panic log, and the return code from
3194 the command was 0. By re-arranging the code a little, I've made the
3195 decoding happen first, so such errors now appear on stderr, and the return
3196 code is 1. However, the actual setting up of the sockets still happens in
3197 the disconnected process, so errors there are still only recorded on the
3198 panic log.
3199
320040. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3201 connections (as happens on some IP stacks) was logged at start up time as
3202 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3203 it from "IPv6 and IPv4", which means that two separate sockets are being
3204 used.
3205
320641. The debug output for gethostbyname2() or getipnodebyname() failures now
3207 says whether AF_INET or AF_INET6 was passed as an argument.
3208
320942. Exiwhat output was messed up when time zones were included in log
3210 timestamps.
3211
321243. Exiwhat now gives more information about the daemon's listening ports,
3213 and whether -tls-on-connect was used.
3214
321544. The "port" option of the smtp transport is now expanded.
3216
321745. A "message" modifier in a "warn" statement in a non-message ACL was being
3218 silently ignored. Now an error message is written to the main and panic
3219 logs.
3220
322146. There's a new ACL modifier called "logwrite" which writes to a log file
3222 as soon as it is encountered.
3223
322447. Added $local_user_uid and $local_user_gid at routing time.
3225
322648. Exim crashed when trying to verify a sender address that was being
3227 rewritten to "<>".
3228
322949. Exim was recognizing only a space character after ".include". It now also
3230 recognizes a tab character.
3231
323250. Fixed several bugs in the Perl script that creates the exim.8 man page by
3233 extracting the relevant information from the specification. The man page no
3234 longer contains scrambled data for the -d option, and I've added a section
3235 at the front about calling Exim under different names.
3236
323751. Added "extra_headers" argument to the "mail" command in filter files.
3238
323952. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3240 crash.
3241
324253. Installed eximstats 1.29.
3243
324454. Added transport_filter_timeout as a generic transport option.
3245
324655. Exim no longer adds an empty Bcc: header to messages that have no To: or
3247 Cc: header lines. This was required by RFC 822, but it not required by RFC
3248 2822.
3249
325056. Exim used to add From:, Date:, and Message-Id: header lines to any
3251 incoming messages that did not have them. Now it does so only if the
3252 message originates locally, that is, if there is no associated remote host
3253 address. When Resent- header lines are present, this applies to the Resent-
3254 lines rather than the non-Resent- lines.
3255
325657. Drop incoming SMTP connection after too many syntax or protocol errors. The
3257 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3258
325958. Messages for configuration errors now include the name of the main
3260 configuration file - useful now that there may be more than one file in a
3261 list (.included file names were always shown).
3262
326359. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3264 for those rare installations that do not start the daemon as root or run it
3265 setuid root. I've cut out the call to initgroups() if the daemon is not
3266 root at that time.
3267
326860. The Exim user and group can now be bound into the binary as text strings
3269 that are looked up at the start of Exim's processing.
3270
327161. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3272
327362. Added $mailstore_basename variable.
3274
327563. Installed patch to sieve.c from Michael Haardt.
3276
327764. When Exim failed to open the panic log after failing to open the main log,
3278 the original message it was trying to log was written to stderr and debug
3279 output, but if they were not available (the usual case in production), it
3280 was lost. Now it is written to syslog before the two lines that record the
3281 failures to open the logs.
3282
328365. Users' Exim filters run in subprocesses under the user's uid. It is
3284 possible for a "deliver" command or an alias in a "personal" command to
3285 provoke an address rewrite. If logging of address rewriting is configured,
3286 this fails because the process is not running as root or exim. There may be
3287 a better way of dealing with this, but for the moment (because 4.30 needs
3288 to be released), I have disabled address rewrite logging when running a
3289 filter in a non-root, non-exim process.
3290
3291
3292Exim version 4.24
3293-----------------
3294
3295 1. The buildconfig auxiliary program wasn't quoting the value set for
3296 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3297 not defined. This bug was masked in 4.22 by the effect that was fixed in
3298 change 4.23/1.
3299
3300 2. Some messages that were rejected after a message id was allocated were
3301 shown as "incomplete" by exigrep. It no longer does this for messages that
3302 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3303
3304 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3305 have allow_domain_literals set, the ID did not get logged in the <= line.
3306 Domain literals are now always recognized in Message-ID: header lines.
3307
3308 4. The first argument for a ${extract expansion item is the key name or field
3309 number. Leading and trailing spaces in this item were not being ignored,
3310 causing some misleading effects.
3311
3312 5. When deliver_drop_privilege was set, single queue runner processes started
3313 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3314 same command in the process it spins off) were not dropping privilege.
3315
3316 6. When the daemon running as "exim" started a queue runner, it always
3317 re-executed Exim in the spun-off process. This is a waste of effort when
3318 deliver_drop_privilege is set. The new process now just calls the
3319 queue-runner function directly.
3320
3321
3322Exim version 4.23
3323-----------------
3324
3325 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3326 HEADERS_CHARSET.
3327
3328 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3329 ignored. Though the use of -oP was forcing the writing of a pid file, it
3330 was always written to the default place.
3331
3332 3. If the message "no IP address found for host xxxx" is generated during
3333 incoming verification, it is now followed by identification of the incoming
3334 connection (so you can more easily find what provoked it).
3335
3336 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3337
3338 5. Added some features to "harden" Exim a bit more against certain attacks:
3339
3340 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3341 be put in Local/Makefile. This is like the never_users runtime option,
3342 but it cannot be overridden. The default setting is "root".
3343
3344 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3345 prefix string with which any file named in a -C command line option
3346 must start.
3347
3348 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3349 is retained for -C and -D only if the caller of Exim is root. Without
3350 it, the exim user may also use -C and -D and retain privilege.
3351
3352 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3353 command line option is disabled.
3354
3355 6. Macro names set by the -D option must start with an upper case letter, just
3356 like macro names defined in the configuration file.
3357
3358 7. Added "dereference=" facility to LDAP.
3359
3360 8. Two instances of the typo "uknown" in the source files are fixed.
3361
3362 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3363 the Configure-Makefile script screwed up while processing it.
3364
336510. Incorporated PCRE 4.4.
3366
336711. The SMTP synchronization check was not operating right at the start of an
3368 SMTP session. For example, it could not catch a HELO sent before the client
3369 waited for the greeting. There is now a check for outstanding input at the
3370 point when the greeting is written. Because of the duplex, asynchronous
3371 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3372 way, but not yet received, when the check is performed.
3373
337412. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3375 on TCP/IP sockets, because this apparently causes some broken clients to
3376 timeout.
3377
337813. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3379 unchanged) from the Cygwin maintainer.
3380
338114. The code for -bV that shows what is in the binary showed "mbx" when maildir
3382 was supported instead of testing for mbx. Effectively a typo.
3383
338415. The spa authenticator server code was not checking that the input it
3385 received was valid base64.
3386
338716. The debug output line for the "set" modifier in ACLs was not showing the
3388 name of the variable that was being set.
3389
339017. Code tidy: the variable type "vtype_string" was never used. Removed it.
3391
339218. Previously, a reference to $sender_host_name did not cause a DNS reverse
3393 lookup on its own. Something else was needed to trigger the lookup. For
3394 example, a match in host_lookup or the need for a host name in a host list.
3395 Now, if $sender_host_name is referenced and the host name has not yet been
3396 looked up, a lookup is performed. If the lookup fails, the variable remains
3397 empty, and $host_lookup_failed is set to "1".
3398
339919. Added "eqi" as a case-independent comparison operator.
3400
340120. The saslauthd authentication condition could segfault if neither service
3402 nor realm was specified.
3403
340421. If an overflowing value such as "2048M" was set for message_size_limit, the
3405 error message that was logged was misleading, and incoming SMTP
3406 connections were dropped. The message is now more accurate, and temporary
3407 errors are given to SMTP connections.
3408
340922. In some error situations (such as 21 above) Exim rejects all SMTP commands
3410 (except RSET) with a 421 error, until QUIT is received. However, it was
3411 failing to send a response to QUIT.
3412
341323. The HELO ACL was being run before the code for helo_try_verify_hosts,
3414 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3415 ACL is now run after the helo_try_verify_hosts code.
3416
341724. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3418 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3419 case-independent, so other case variants are also recognized). Apparently
3420 some systems use these upper case variants.
3421
342225. If more than two messages were waiting for the same host, and a transport
3423 filter was specified for the transport, Exim sent two messages over the
3424 same TCP/IP connection, and then failed with "socket operation on non-
3425 socket" when it tried to send the third.
3426
342726. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3428
342927. The extern definition of crypt16() in expand.c was not being excluded when
3430 the OS had its own crypt16() function.
3431
343228. Added bounce_return_body as a new option, and bounce_return_size_limit
3433 as a preferred synonym for return_size_limit, both as an option and as an
3434 expansion variable.
3435
343629. Added LIBS=-liconv to OS/Makefile-OSF1.
3437
343830. Changed the default configuration ACL to relax the local part checking rule
3439 for addresses that are not in any local domains. For these addresses,
3440 slashes and pipe symbols are allowed within local parts, but the sequence
3441 /../ is explicitly forbidden.
3442
344331. SPA server authentication was not clearing the challenge buffer before
3444 using it.
3445
344632. log_message in a "warn" ACL statement was writing to the reject log as
3447 well as to the main log, which contradicts the documentation and doesn't
3448 seem right (because no rejection is happening). So I have stopped it.
3449
345033. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3451 However, I am unable to do any testing of this.
3452
345334. Fixed an infelicity in the appendfile transport. When checking directories
3454 for a mailbox, to see if any needed to be created, it was accidentally
3455 using path names with one or more superfluous leading slashes; tracing
3456 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3457
345835. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3459 contents are added to the log line that is written for every discarded
3460 recipient. (Previously a log_message setting was ignored.)
3461
346236. The ${quote: operator now quotes the string if it is empty.
3463
346437. The install script runs exim in order to find its version number. If for
3465 some reason other than non-existence or emptiness, which it checks, it
3466 could not run './exim', it was installing it with an empty version number,
3467 i.e. as "exim-". This error state is now caught, and the installation is
3468 aborted.
3469
347038. An argument was missing from the function that creates an error message
3471 when Exim fails to connect to the socket for saslauthd authentication.
3472 This could cause Exim to crash, or give a corrupted message.
3473
347439. Added isip, isip4, and isip6 to ${if conditions.
3475
347640. The ACL variables $acl_xx are now saved with the message, and can be
3477 accessed later in routers, transports, and filters.
3478
347941. The new lookup type nwildlsearch is like wildlsearch, except that the key
3480 strings in the file are not string-expanded.
3481
348242. If a MAIL command specified a SIZE value that was too large to fit into an
3483 int variable, the check against message_size_limit failed. Such values are
3484 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3485 day this will have to be increased, but I don't think I want to be around
3486 when emails are that large.
3487
3488
3489
3490Exim version 4.22
3491-----------------
3492
3493 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3494 iconv() is not standard in FreeBSD.
3495
3496 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3497 IPv6 enabled. The observed symptom was a segmentation fault on return from
3498 the function os_common_find_running_interfaces() in src/os.c.
3499
3500 3. In the check_special_case() function in daemon.c I had used "errno" as an
3501 argument name, which causes warnings on some systems. This was basically a
3502 typo, since it was named "eno" in the comments!
3503
3504 4. The code that waits for the clock to tick (at a resolution of some fraction
3505 of a second) so as to ensure message-id uniqueness was always waiting for
3506 at least one whole tick, when it could have waited for less. [This is
3507 almost certainly not relevant at current processor speeds, where it is
3508 unlikely to ever wait at all. But we try to future-proof.]
3509
3510 5. The function that sleeps for a time interval that includes fractions of a
3511 second contained a race. It did not block SIGALRM between setting the
3512 timer, and suspending (a couple of lines later). If the interval was short
3513 and the sigsuspend() was delayed until after it had expired, the suspension
3514 never ended. On busy systems this could lead to processes getting stuck for
3515 ever.
3516
3517 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3518 process, before it forks any delivery processes. The open lookup caching
3519 mechanism meant that the open file or database connection was passed into
3520 the delivery process. The problem was that delivery processes always tidy
3521 up cached lookup data. This could cause a problem for the next delivery
3522 process started by the queue runner, because the external queue runner
3523 process does not know about the closure. So the next delivery process
3524 still has data in the lookup cache. In the case of a file lookup, there was
3525 no problem because closing a file descriptor in a subprocess doesn't affect
3526 the parent. However, if the lookup was caching a connection to a database,
3527 the connection was closed, and the second delivery process was likely to
3528 see errors such as "PGSQL: query failed: server closed the connection
3529 unexpectedly". The problem has been fixed by closing all cached lookups
3530 in a queue runner before running a delivery process.
3531
3532 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3533 seem to have the "const" qualifier which it has on other OS. I've
3534 parameterised it.
3535
3536 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3537 *of the same type* (client or server) with the same public name that an
3538 error should be diagnosed.
3539
3540 9. When Exim looked up a host name for an IP address, but failed to find the
3541 original IP address when looking up the host name (a safety check), it
3542 output the message "<ip address> does not match any IP for NULL", which was
3543 confusing, to say the least. The bug was that the host name should have
3544 appeared instead of "NULL".
3545
354610. Since release 3.03, if Exim is called by a uid other than root or the Exim
3547 user that is built into the binary, and the -C or -D options is used, root
3548 privilege is dropped before the configuration file is read. In addition,
3549 logging is switched to stderr instead of the normal log files. If the
3550 configuration then re-defines the Exim user, the unprivileged environment
3551 is probably not what is expected, so Exim logs a panic warning message (but
3552 proceeds).
3553
3554 However, if deliver_drop_privilege is set, the unprivileged state may well
3555 be exactly what is intended, so the warning has been cut out in that case,
3556 and Exim is allowed to try to write to its normal log files.
3557
3558
3559Exim version 4.21
3560-----------------
3561
3562 1. smtp_return_error_details was not giving details for temporary sender
3563 or receiver verification errors.
3564
3565 2. Diagnose a configuration error if two authenticators have the same public
3566 name.
3567
3568 3. Exim used not to create the message log file for a message until the first
3569 delivery attempt. This could be confusing when incoming messages were held
3570 for policy or load reasons. The message log file is now created at the time
3571 the message is received, and an initial "Received" line is written to it.
3572
3573 4. The automatically generated man page for command line options had a minor
3574 bug that caused no ill effects; however, a more serious problem was that
3575 the procedure for building the man page automatically didn't always
3576 operate. Consequently, release 4.20 contains an out-of-date version. This
3577 shouldn't happen again.
3578
3579 5. When building Exim with embedded Perl support, the script that builds the
3580 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3581 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3582
3583 6. The freeze_tell option was not being used for messages that were frozen on
3584 arrival, either by an ACL or by local_scan().
3585
3586 7. Added the smtp_incomplete_transaction log selector.
3587
3588 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3589 was accepting AUTH without a new EHLO.
3590
3591 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3592 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3593 RFC.
3594
359510. Logging of TCP/IP connections (when configured) now happens in the main
3596 daemon process instead of the child process, so that the TCP/IP connection
3597 count is more accurate (but it can never be perfect).
3598
359911. The use of "drop" in a nested ACL was not being handled correctly in the
3600 outer ACL. Now, if condition failure induced by the nested "drop" causes
3601 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3602 or "require"), the connection is dropped.
3603
360412. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3605 that yield "discard" can now be used with an "accept" or a "discard" verb,
3606 but an error is generated for any others (because I can't see a useful way
3607 to define what should happen).
3608
360913. When an ACL is read dynamically from a file (or anywhere else), the lines
3610 are now processed in the same way as lines in the Exim configuration file.
3611 In particular, continuation lines are supported.
3612
361314. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3614
361515. Added -ti meaning -t -i.
3616
361716. Check for letters, digits, hyphens, and dots in the names of dnslist
3618 domains, and warn by logging if others are found.
3619
362017. At least on BSD, alignment is not guarenteed for the array of ifreq's
3621 returned from GIFCONF when Exim is trying to find the list of interfaces on
3622 a host. The code in os.c has been modified to copy each ifreq to an aligned
3623 structure in all cases.
3624
3625 Also, in some cases, the returned ifreq's were being copied to a 'struct
3626 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3627 means the last couple of bytes of an IPv6 address could be chopped if the
3628 ifreq contained only a normal sockaddr (14 bytes storage).
3629
363018. Named domain lists were not supported in the hosts_treat_as_local option.
3631 An entry such as +xxxx was not recognized, and was treated as a literal
3632 domain name.
3633
363419. Ensure that header lines added by a DATA ACL are included in the reject log
3635 if the ACL subsequently rejects the message.
3636
363720. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3638 MD5 (which is deprecated).
3639
364021. When testing a filter file using -bf, Exim was writing a message when it
3641 took the sender from a "From " line in the message, but it was not doing so
3642 when it took $return_path from a Return-Path: header line. It now does.
3643
364422. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3645 with a valid header line field name (a series of printing characters
3646 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3647
364823. Changed "disc" in the source to "disk" to conform to the documentation and
3649 the book and for uniformity.
3650
365124. Ignore Sendmail's -Ooption=value command line item.
3652
365325. When execve() failed while trying to run a command in a pipe transport,
3654 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3655 could be confused with a return value of 69 from the command itself. This
3656 has been changed to 127, the value the shell returns if it is asked to run
3657 a non-existent command. The wording for the related log line suggests a
3658 non-existent command as the problem.
3659
366026. If received_header_text expands to an empty string, do not add a Received:
3661 header line to the message. (Well, it adds a token one on the spool, but
3662 marks it "old" so that it doesn't get used or transmitted.)
3663
366427. Installed eximstats 1.28 (addition of -nt option).
3665
366628. There was no check for failure on the call to getsockname() in the daemon
3667 code. This can fail if there is a shortage of resources on the system, with