Fix typo in usage: seond -> second
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
c887c79e 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.339 2006/04/04 17:10:07 fanf2 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
214e2000
PH
12Exim version 4.61
13-----------------
14
15PH/01 The code for finding all the local interface addresses on a FreeBSD
16 system running IPv6 was broken. This may well have applied to all BSD
17 systems, as well as to others that have similar system calls. The broken
18 code found IPv4 interfaces correctly, but gave incorrect values for the
19 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
20 that it would not match correctly against @[] and not recognize the IPv6
21 addresses as local.
22
f9daeae0
PH
23PH/02 The ipliteral router was not recognizing addresses of the form user@
24 [ipv6:....] because it didn't know about the "ipv6:" prefix.
25
7e66e54d
PH
26PH/03 Added disable_ipv6.
27
c8ea1597
PH
28PH/04 Changed $reply_address to use the raw form of the headers instead of the
29 decoded form, because it is most often used to construct To: headers
30 lines in autoreplies, and the decoded form may well be syntactically
31 invalid. However, $reply_address has leading white space removed, and all
32 newlines turned into spaces so that the autoreply transport does not
33 grumble.
34
911f6fde
PH
35PH/05 If group was specified without a user on a router, and no group or user
36 was specified on a transport, the group from the router was ignored.
37
47ca6d6c
PH
38PH/06 Increased the number of ACL variables to 20 of each type, and arranged
39 for visible compile-time settings that can be used to change these
40 numbers, for those that want even more. Backwards compatibility with old
41 spool files has been maintained. However, going back to a previous Exim
42 release will lost any variables that are in spool files.
43
ed0e9820
PH
44PH/07 Two small changes when running in the test harness: increase delay when
45 passing a TCP/IP connection to a new process, in case the original
46 process has to generate a bounce, and remove special handling of
47 127.0.0.2 (sic), which is no longer necessary.
48
eff37e47
PH
49PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
50 be the same on different OS.
51
1921d2ea
PH
52PH/09 Moved a debug statement in filter processing to avoid a race problem when
53 testing.
54
b3f69ca8
JJ
55JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
56 whether --show-vars was specified or not
57
58JJ/02 exipick: Added support for new ACL variable spool format introduced
59 in 4.61-PH/06
60
424a1c63
PH
61PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
62 syntactically invalid From: or Reply-to: line, and a filter used this to
63 generate an autoreply, and therefore failed to obtain an address for the
64 autoreply, Exim could try to deliver to a non-existent relative file
65 name, causing unrelated and misleading errors. What now happens is that
66 it logs this as a hard delivery error, but does not attempt to create a
67 bounce message.
68
7a100415
PH
69PH/11 The exinext utility has a -C option for testing purposes, but although
70 the given file was scanned by exinext itself; it wasn't being passed on
71 when Exim was called.
72
19b9dc85
PH
73PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
74 an end-of-file indication when reading a command response.
75
309bd837
PH
76PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
77 compiled. In many other places in Exim, IPv6 addresses are always
78 recognized, so I have changed this. It also means that IPv4 domain
79 literals of the form [IPV4:n.n.n.n] are now always recognized.
80
59e82a2a
PH
81PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
82 used if the router is not running as root, for example, when verifying at
83 ACL time, or when using -bh. The debugging output from this situation was
84 non-existent - all you got was a failure to exec. I have made two
85 changes:
86
87 (a) Failures to set uid/gid, the current directory, or a process leader
88 in a subprocess such as that created by queryprogram now generate
89 suitable debugging ouput when -d is set.
90
91 (b) The queryprogram router detects when it is not running as root,
92 outputs suitable debugging information if -d is set, and then runs
93 the subprocess without attempting to change uid/gid.
94
9edc04ce
PH
95PH/15 Minor change to Makefile for building test_host (undocumented testing
96 feature).
97
1349e1e5
PH
98PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
99 additional section of a DNS packet that returns MX or SRV records.
100 Instead, it always explicitly searches for A/AAAA records. This avoids
101 major problems that occur when a DNS server includes only records of one
102 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
103 fixed another bug: if SRV records were looked up and the corresponding
104 address records were *not* found in the additional section, the port
105 values from the SRV records were lost.
106
ea49d0e1
PH
107PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
108 using the correct key (the original address) when searching the retry
109 rules in order to find which one to use for generating the retry hint.
110
064a94c9
PH
111PH/18 If quota_warn_message contains a From: header, Exim now refrains from
112 adding the default one. Similarly, if it contains a Reply-To: header, the
113 errors_reply_to option, if set, is not used.
114
727071f8
PH
115PH/19 When calculating a retry time, Exim used to measure the "time since
116 failure" by looking at the "first failed" field in the retry record. Now
117 it does not use this if it is later than than the arrival time of the
118 message. Instead it uses the arrival time. This makes for better
119 behaviour in cases where some deliveries succeed, thus re-setting the
120 "first failed" field. An example is a quota failure for a huge message
121 when small messages continue to be delivered. Without this change, the
122 "time since failure" will always be short, possible causing more frequent
123 delivery attempts for the huge message than are intended.
124
f78eb7c6
PH
125PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
126 $1, $2, $3) because the numerical variables can be reset during some
127 expansion items (e.g. "match"), thereby losing the authentication data.
128
21c28500
PH
129PH/21 Make -bV show the size of off_t variables so that the test suite can
130 decide whether to run tests for quotas > 2G.
131
132PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
133 mailbox_size, and mailbox_filecount in the appendfile transport. If a
134 filecount value is greater than 2G or if a quota value is greater than 2G
135 on a system where the size of off_t is not greater than 4, a panic error
136 is given.
137
1688f43b
PH
138PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
139 never match. The debug and -bh output now contains an explicit error
140 message indicating a malformed IPv4 address or mask.
141
142PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
143 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
144 PH/23 above applies.
145
9675b384
PH
146PH/25 Do not write to syslog when running in the test harness. The only
147 occasion when this arises is a failure to open the main or panic logs
148 (for which there is an explicit test).
149
6a3f1455
PH
150PH/26 Added the /no_tell option to "control=freeze".
151
dac79d3e
PH
152PH/27 If a host name lookup failed very early in a connection, for example, if
153 the IP address matched host_lookup and the reverse lookup yielded a name
154 that did not have a forward lookup, an error message of the form "no IP
155 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
156 could be logged. Now it outputs the IP address instead of "NULL".
1349e1e5 157
5977a0b3
PH
158PH/28 An enabling patch from MH: add new function child_open_exim2() which
159 allows the sender and the authenticated sender to be set when
160 submitting a message from within Exim. Since child_open_exim() is
161 documented for local_scan(), the new function should be too.
162
c91535f3
PH
163PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
164 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
165 results in an empty string is now treated as unset.
166
0d46a8c8
PH
167PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
168
278c6e6c
PH
169PH/31 Added sender_verify_fail logging option.
170
2cbb4081
PH
171PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
172 needed by RFC 822 but not by RFC 2822 was commented out. I have now
173 tidied the source and removed it altogether.
174
3eef829e
PH
175PH/33 When a queue run was abandoned because the load average was too high, a
176 log line was always written; now it is written only if the queue_run log
177 selector is set. In addition, the log line for abandonment now contains
178 information about the queue run such as the pid. This is always present
179 in "start" and "stop" lines but was omitted from the "abandon" line.
180
1ab95fa6
PH
181PH/34 Omit spaces between a header name and the colon in the error message that
182 is given when verify = headers_syntax fails (if there are lots of them,
183 the message gets confusing).
184
230205fc
PH
185PH/35 Change the default for dns_check_names_pattern to allow slashes within
186 names, as there are now some PTR records that contain slashes. This check
187 is only to protect against broken name servers that fall over on strange
188 characters, so the fact that it applies to all lookups doesn't matter.
189
75e0e026
PH
190PH/36 Now that the new test suite is complete, we can remove some of the
191 special code in Exim that was needed for the old test suite. For example,
192 sorting DNS records because real resolvers return them in an arbitrary
193 order. The new test suite's fake resolver always returns records in the
194 same order.
195
196PH/37 When running in the test harness, use -odi for submitted messages (e.g.
197 bounces) except when queue_only is set, to avoid logging races between
198 the different processes.
199
145396a6
PH
200PH/38 Panic-die if .include specifies a non-absolute path.
201
3cd34f13
PH
202PH/39 A tweak to the "H" retry rule from its user.
203
11121d3d
JJ
204JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
205 a label. They prevented compilation on older perls.
206
207JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
208 a warning to be raised on newish perls.
209
210JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
211 on queue. Changes to match documented behaviour of showing count of
212 messages matching specified criteria.
213
8def5aaf
PH
214PH/40 Changed the default ident timeout from 30s to 5s.
215
929ba01c
PH
216PH/41 Added support for the use of login_cap features, on those BSD systems
217 that have them, for controlling the resources used by pipe deliveries.
218
2632889e
PH
219PH/42 The content-scanning code uses fopen() to create files in which to put
220 message data. Previously it was not paying any attention to the mode of
221 the files. Exim runs with umask(0) because the rest of the code creates
222 files with open(), and sets the required mode explicitly. Thus, these
223 files were ending up world-writeable. This was not a big issue, because,
224 being within the spool directory, they were not world-accessible. I have
225 created a function called modefopen, which takes an additional mode
226 argument. It sets umask(777), creates the file, chmods it to the required
227 mode, then resets the umask. All the relevant calls to fopen() in the
228 content scanning code have been changed to use this function.
229
944a9c55
PH
230PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
231 to 24 hours. This avoids potential overflow problems when processing G
232 and H retry rules. I suspect nobody ever tinkers with this value.
233
4a23603b
PH
234PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
235
4730f942
PH
236PH/45 When the plaintext authenticator is running as a client, the server's
237 challenges are checked to ensure they are valid base64 strings. By
238 default, the authentication attempt is cancelled if an invalid string is
239 received. Setting client_ignore_invalid_base64 true ignores these errors.
240 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
241 they are received. Thus, the responses can be made to depend on the
242 challenges. If an invalid string is ignored, an empty string is placed in
243 the variable.
244
30dba1e6
PH
245PH/46 Messages that are created by the autoreply transport now contains a
246 References: header, in accordance with RFCs 2822 and 3834.
247
382afc6b
PH
248PH/47 Added authenticated_sender_force to the smtp transport.
249
a86229cf
PH
250PH/48 The ${prvs expansion was broken on systems where time_t was long long.
251
50c99ba6
PH
252PH/49 Installed latest patch from the Sieve maintainer.
253
d35e429d
PH
254PH/50 When an Exim quota was set without a file count quota, and mailbox_size
255 was also set, the appendfile transport was unnecessarily scanning a
256 directory of message files (e.g. for maildir delivery) to find the count
257 of files (along with the size), even though it did not need this
258 information. It now does the scan only if it needs to find either the
259 size of the count of files.
260
f90d018c
PH
261PH/51 Added ${time_eval: to convert Exim time strings into seconds.
262
75def545
PH
263PH/52 Two bugs concerned with error handling when the smtp transport is
264 used in LMTP mode:
265
266 (i) Exim was not creating retry information for temporary errors given
267 for individual recipients after the DATA command when the smtp transport
268 was used in LMTP mode. This meant that they could be retried too
269 frequently, and not timed out correctly.
270
271 (ii) Exim was setting the flag that allows error details to be returned
272 for LMTP errors on RCPT commands, but not for LMTP errors for individual
273 recipients that were returned after the DATA command.
274
275PH/53 This is related to PH/52, but is more general: for any failing address,
276 when detailed error information was permitted to be returned to the
277 sender, but the error was temporary, then after the final timeout, only
278 "retry timeout exceeded" was returned. Now it returns the full error as
279 well as "retry timeout exceeded".
280
c46782ef
PH
281PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
282 do this, and (what is worse) MTAs that accept it.
283
71fafd95
PH
284PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
285 will now be deprecated.
286
2c5db4fd
PH
287PH/56 New os.c-cygwin from the Cygwin maintainer.
288
9cf6b11a
JJ
289JJ/06 exipick: added --unsorted option to allow unsorted output in all output
290 formats (previously only available in exim formats via -bpr, -bpru,
291 and -bpra. Now also available in native and exiqgrep formats)
292
293JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
294 with very large, slow to parse queues
295
296JJ/08 exipick: added ! as generic prefix to negate any criteria format
297
298JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
299
898d150f
PH
300PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
301 responses to authentication challenges, though it was showing the
302 challenges; (ii) I've removed the CR characters from the debug output for
303 SMTP output lines.
304
46218253
PH
305PH/58 Allow for the insertion of a newline as well as a space when a string
306 is turned into more than one encoded-word during RFC 2047 encoding. The
307 Sieve code now uses this.
308
e97957bc
PH
309PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
310 data_4xx, lost_connection, tls_required.
311
81e509d7
PH
312PH/60 When a VRFY deferred or FAILED, the log message rather than the user
313 message was being sent as an SMTP response.
314
3d240ff7
PH
315PH/61 Add -l and -k options to exicyclog.
316
b37c4101
PH
317PH/62 When verifying, if an address was redirected to one new address, so that
318 verification continued, and the new address failed or deferred after
319 having set something in $address_data, the value of $address_data was not
320 passed back to the ACL. This was different to the case when no
321 redirection occurred. The value is now passed back in both cases.
322
79378e0f
PH
323PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
324 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
325 use login_cap.h, so on its own it isn't the distinguishing feature. The
326 new name refers directly to the setclassresources() function.
327
e49c7bb4
PH
328PH/65 Added configuration files for NetBSD3.
329
d114ec46
PH
330PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
331
f3d7df6c
PH
332PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
333 is preferred over IPv4.
334
715ab376
PH
335PH/68 The bounce_return_message and bounce_return_body options were not being
336 honoured for bounces generated during the reception of non-SMTP messages.
337 In particular, this applied to messages rejected by the ACL. This bug has
338 been fixed. However, if bounce_return_message is true and bounce_return_
339 body is false, the headers that are returned for a non-SMTP message
340 include only those that have been read before the error was detected.
341 (In the case of an ACL rejection, they have all been read.)
342
6b31b150
PH
343PH/69 The HTML version of the specification is now built in a directory called
344 spec_html instead of spec.html, because the latter looks like a path with
345 a MIME-type, and this confuses some software.
346
347PH/70 Catch two compiler warnings in sieve.c.
348
d515a917
PH
349PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
350 function verify_get_ident() calls ip_connect() to connect a socket, but
351 if the "connect()" function timed out, ip_connect() used to close the
352 socket. However, verify_get_ident() also closes the socket later, and in
353 between Exim writes to the log, which may get opened at this point. When
354 the socket was closed in ip_connect(), the log could get the same file
355 descriptor number as the socket. This naturally causes chaos. The fix is
356 not to close the socket in ip_connect(); the socket should be closed by
357 the function that creates it. There was only one place in the code where
358 this was missing, in the iplookup router, which I don't think anybody now
359 uses, but I've fixed it anyway.
360
9b8fadde
PH
361PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
362 well as to direct DNS lookups. Otherwise the handling of names in host
363 lists is inconsistent and therefore confusing.
364
214e2000 365
5de37277
PH
366Exim version 4.60
367-----------------
368
cc38ddbf
PH
369PH/01 Two changes to the default runtime configuration:
370
371 (1) Move the checks for relay_from_hosts and authenticated clients from
372 after to before the (commented out) DNS black list checks.
373
374 (2) Add control=submission to the relay_from_hosts and authenticated
375 clients checks, on the grounds that messages accepted by these
376 statements are most likely to be submissions.
5de37277 377
72fdd6ae
PH
378PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
379
380 (1) Generate an error if the third argument for the ${prvs expansion is
381 not a single digit.
382
383 (2) Treat a missing third argument of ${prvscheck as if it were an empty
384 string.
385
386 (3) Reset the variables that are obtained from the first argument of
387 ${prvscheck and used in the second argument before leaving the code,
388 because their memory is reclaimed, so using them afterwards may do
389 silly things.
390
391 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
392 one (it's much easier than Tom thought :-).
393
394 (5) Because of (4), we can now allow for the use of $prvscheck_result
395 inside the third argument.
cb9328de 396
cb741023
PH
397PH/03 For some reason, the default setting of PATH when running a command from
398 a pipe transport was just "/usr/bin". I have changed it to
399 "/bin:/usr/bin".
400
f174f16e
PH
401PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
402 anything to be listed in the output from -bV.
b2f5a032 403
c25242d7
PH
404PH/05 When a filter generated an autoreply, the entire To: header line was
405 quoted in the delivery log line, like this:
406
407 => >A.N.Other <ano@some.domain> <original@ddress> ...
408
409 This has been changed so that it extracts the operative address. There
410 may be more than one such address. If so, they are comma-separated, like
411 this:
412
413 => >ano@some.domain,ona@other.domain <original@ddress> ...
414
82c19f95
PH
415PH/06 When a client host used a correct literal IP address in a HELO or EHLO
416 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
417 not being looked up in the rDNS to get a host name, Exim was showing the
418 IP address twice in Received: lines, even though the IP addresses were
419 identical. For example:
420
421 Received: from [1.2.3.4] (helo=[1.2.3.4])
422
423 However, if the real host name was known, it was omitting the HELO data
424 if it matched the actual IP address. This has been tidied up so that it
425 doesn't show the same IP address twice.
426
d7ffbc12
PH
427PH/07 When both +timestamp and +memory debugging was on, the value given by
428 $tod_xxx expansions could be wrong, because the tod_stamp() function was
429 called by the debug printing, thereby overwriting the timestamp buffer.
430 Debugging no longer uses the tod_stamp() function when +timestamp is set.
431
9f526266
PH
432PH/08 When the original message was included in an autoreply transport, it
433 always said "this is a copy of the message, including all the headers",
434 even if body_only or headers_only was set. It now gives an appropriate
435 message.
436
87fcc8b9
PH
437PH/09 Applied a patch from the Sieve maintainer which:
438
439 o fixes some comments
440 o adds the (disabled) notify extension core
441 o adds some debug output for the result of if/elsif tests
442 o points to the current vacation draft in the documentation
443 and documents the missing references header update
444
445 and most important:
446
447 o fixes a bug in processing the envelope test (when testing
448 multiple envelope elements, the last element determinted the
449 result)
450
456682f5
PH
451PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
452 Electronic Mail") by including:
453
454 Auto-submitted: auto-generated
455
456 in the messages that it generates (bounce messages and others, such as
457 warnings). In the case of bounce messages for non-SMTP mesages, there was
458 also a typo: it was using "Auto_submitted" (underscore instead of
459 hyphen). Since every message generated by Exim is necessarily in response
460 to another message, thes have all been changed to:
461
462 Auto-Submitted: auto-replied
463
464 in accordance with these statements in the RFC:
465
466 The auto-replied keyword:
467
468 - SHOULD be used on messages sent in direct response to another
469 message by an automatic process,
470
471 - MUST NOT be used on manually-generated messages,
472
473 - MAY be used on Delivery Status Notifications (DSNs) and Message
474 Disposition Notifications (MDNs),
475
476 - MUST NOT be used on messages generated by automatic or periodic
477 processes, except for messages which are automatic responses to
478 other messages.
479
3e46c1aa
PH
480PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
481 to the default Received: header definition.
456682f5 482
49826d12
PH
483PH/12 Added log selector acl_warn_skipped (default on).
484
eba0c039
PH
485PH/13 After a successful wildlsearch lookup, discard the values of numeric
486 variables because (a) they are in the wrong storage pool and (b) even if
487 they were copied, it wouldn't work properly because of the caching.
488
a0d6ba8a
PH
489PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
490 checking when decoding. Apparently there are clients that generate
491 overlong encoded strings. Why am I not surprised?
492
f0917727
PH
493PH/15 If the first argument of "${if match_address" was not empty, but did not
494 contain an "@" character, Exim crashed. Now it writes a panic log message
495 and treats the condition as false.
496
096fee00
PH
497PH/16 In autoreply, treat an empty string for "once" the same as unset.
498
024bd3c2
PH
499PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
500 extension "envelope-auth". The code is finished and in agreement with
501 other implementations, but there is no documentation so far and in fact,
502 nobody wrote the draft yet. This extension is currently #undef'ed, thus
503 not changing the active code.
504
505 Print executed "if" and "elsif" statements when debugging is used. This
506 helps a great deal to understand what a filter does.
507
508 Document more things not specified clearly in RFC3028. I had all this
509 sorted out, when out of a sudden new issues came to my mind. Oops."
510
df199fec
PH
511PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
512 (Bugzilla #53).
513
d27f1df3
PH
514PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
515 canonical form (as documented). However, after a host name lookup from
516 the IP address, check_host() was doing a simple string comparison with
517 addresses acquired from the DNS when checking that the found name did
518 have the original IP as one of its addresses. Since any found IPv6
519 addresses are likely to be in abbreviated form, the comparison could
520 fail. Luckily, there already exists a function for doing the comparison
521 by converting both addresses to binary, so now that is used instead of
522 the text comparison.
523
96776534
PH
524PH/20 There was another similar case to PH/19, when a complete host name was
525 given in a host list; looking up its IP address could give an abbreviated
526 form, whereas the current host's name might or might not be abbreviated.
527 The same fix has been applied.
528
5de37277 529
9a799bc0
PH
530Exim version 4.54
531-----------------
532
533PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
534 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
535 It now does.
536
99a4b039
PH
537PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
538 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
539
540PH/03 Typo: missing ".o" in src/pcre/Makefile.
541
4b233853
PH
542PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
543 header line, restrict the check to what is listed in RFCs 2369 and 2929.
544 Also, for "Auto-Submitted", treat anything other than "no" as
545 non-personal, in accordance with RFC 3834. (Previously it treated
546 anything starting "auto-" as non-personal.)
547
8857ccfd
PH
548TF/01 The control=submission/name=... option had a problem with syntax
549 errors if the name included a slash character. The /name= option
550 now slurps the rest of the string, so it can include any characters
551 but it must come last in the list of options (after /sender_retain
552 or /domain=).
553
433a2980
PH
554PH/05 Some modifications to the interface to the fake nameserver for the new
555 testing suite.
556
3e46c1aa 557
9a799bc0 558
e3a311ba
TK
559Exim version 4.53
560-----------------
561
562TK/01 Added the "success_on_redirect" address verification option. See
563 NewStuff for rationale and an example.
564
13b685f9
PH
565PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
566
395ff96d
PH
567PH/02 Patch to exigrep to allow it to work on syslog lines.
568
5b68f6e4
PH
569PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
570 fread() to skip over the body file's header line, because in Cygwin the
571 header line is locked and is inaccessible.
572
1ab52c69
PH
573PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
574 co-exist for some time) to make it clear that it is the Exim ID that is
575 referenced, not the Message-ID: header line.
576
b07e6aa3
PH
577PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
578 string_format() function, because snprintf() does not exist on all
579 operating systems.
580
254e032f
PH
581PH/06 The use of forbid_filter_existstest now also locks out the use of the
582 ${stat: expansion item.
583
3af76a81
PH
584PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
585 protocol synchronization error", to keep the pedants happy.
586
2548ba04
PH
587PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
588 well as for IRIX systems, when gcc is being used. See the host.c source
589 file for comments.
590
b6c6011d
PH
591PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
592
cf39cf57
PH
593PH/10 Named domain lists were not working if used in a queue_smtp_domains
594 setting.
595
f1513293
PH
596PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
597 transport and to the smtp transport in LMTP mode.
598
727549a4
PH
599TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
600
af46795e
PH
601PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
602 run a filter in a subprocess. This could lead to confusion in subsequent
603 lookups in the parent process. There should also be a search_tidyup() at
604 the end of the subprocess.
605
d7b47fd0
PH
606PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
607 only if the host matched helo_try_verify_hosts, which caused the
608 verification to occur when the EHLO/HELO command was issued. The ACL just
609 tested the remembered result. Now, if a previous verification attempt has
610 not happened, "verify = helo" does it there and then.
611
ee744174
JJ
612JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
613
b582ab87
PH
614TK/03 Fix log output including CR from clamd.
615
41a13e0a
PH
616PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
617 exist provoked a memory error which could cause a segfault.
618
f625cc5a
PH
619PH/15 Installed PCRE 6.2
620
621PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
622
21f7af35
PH
623PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
624 of the problem. Specifically, suggested +O2 rather than +O1 for the
625 HP-UX compiler.
626
31480e42
PH
627PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
628
2d280592
PH
629PH/20 If a delivery was routed to a non-standard port by means of an SRV
630 record, the port was not correctly logged when the outgoing_port log
631 selector was set (it logged the transort's default port).
632
7cd1141b
PH
633PH/21 Added support for host-specific ports to manualroute, queryprogram,
634 fallback_hosts, and "hosts" in the smtp transport.
635
636PH/22 If the log selector "outgoing_port" is set, the port is now also given on
637 host errors such as "Connection refused".
638
750af86e
PH
639PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
640 authentication with radiusclient 0.4.9:
641
642 - Error returned from rc_read_config was caught wrongly
643 - Username/password not passed on to radius server due to wrong length.
644
645 The presumption is that some radiusclient API changes for 4.51/PH/17
646 were not taken care of correctly. The code is still untested by me (my
647 Linux distribution still has 0.3.2 of radiusclient), but it was
648 contributed by a Radius user.
649
650PH/24 When doing a callout, the value of $domain wasn't set correctly when
651 expanding the "port" option of the smtp transport.
652
4304270b
TK
653TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
654 while reading a MIME header. Thanks to Tom Hughes for a patch.
655
750af86e
PH
656PH/24 Include config.h inside local_scan.h so that configuration settings are
657 available.
658
64ffc24f
PH
659PH/25 Make $smtp_command_argument available after all SMTP commands. This means
660 that in an ACL for RCPT (for example), you can examine exactly what was
661 received.
662
5dd9625b
PH
663PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
664 commands, but it was not correctly comparing the address with the actual
665 client host address. Thus, it would show the EHLO address in Received:
666 header lines when this was not necessary.
667
5591031b
PH
668PH/27 Added the % operator to ${eval:}.
669
ba18e66a
PH
670PH/28 Exim tries to create and chdir to its spool directory when it starts;
671 it should be ignoring failures (because with -C, for example, it has lost
672 privilege). It wasn't ignoring creation failures other than "already
673 exists".
674
9cec981f
PH
675PH/29 Added "crypteq" to the list of supported features that Exim outputs when
676 -bV or -d is used.
677
aa2b5c79
PH
678PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
679 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
680 too many continuations, the temporary file was not being removed, and the
681 return code was incorrect.
aa2b5c79 682
48a53b7f
PH
683PH/31 Missing "BOOL" in function definition in filtertest.c.
684
1c59d63b
PH
685PH/32 Applied Sieve patches from the maintainer.
686
671012da
TK
687TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
688
1509d3a8
PH
689PH/33 Added "verify = not_blind".
690
691PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
692 Local/Makefile (with some defaults set). These are used in built scripts
693 such as exicyclog, but they have never been used in the exim_install
694 script (though there are many overriding facilities there). I have
695 arranged that the exim_install script now takes note of these two
696 settings.
697
698PH/35 Installed configuration files for Dragonfly.
699
2fe1a124
PH
700PH/36 When a locally submitted message by a trusted user did not contain a
701 From: header, and the sender address was obtained from -f or from an SMTP
702 MAIL command, and the trusted user did not use -F to supply a sender
703 name, $originator_name was incorrectly used when constructing a From:
704 header. Furthermore, $originator_name was used for submission mode
705 messages from external hosts without From: headers in a similar way,
706 which is clearly wrong.
707
8800895a
PH
708PH/37 Added control=suppress_local_fixups.
709
ccfdb010
PH
710PH/38 When log_selector = +received_sender was set, and the addition of the
711 sender made the log line's construction buffer exactly full, or one byte
712 less than full, an overflow happened when the terminating "\n" was
713 subsequently added.
714
1130bfb0
PH
715PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
716 when the result of a list match is failure because a DNS lookup failed.
717
ebcb507f
PH
718PH/40 RM_COMMAND is now used in the building process.
719
c35e155c
PH
720PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
721 the "build-* directories that it finds.
722
95d1f782
PH
723PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
724 address in a domain literal was a prefix of an interface address.
725
fd6de02e
PH
726PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
727 when verifying a sender address, unless rewrite_headers is false.
728
58de37c5
PH
729PH/44 Wrote a long comment about why errors_to addresses are verified as
730 recipients, not senders.
731
261cf466
TF
732TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
733 the ratelimit ACL was added.
734
3ee512ff
PH
735PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
736
e08c430f
PH
737PH/46 Added extra information about PostgreSQL errors to the error string.
738
bef5a11f
PH
739PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
740 avoiding the need to install special zones in a real server. This is
741 backwards compatible; if it can't find the fake resolver, it drops back.
742 Thus, both old and new test suites can be run.
743
7546de58
TF
744TF/02 Added util/ratelimit.pl
745
e5d5a95f
TF
746TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
747 clock is set back in time.
748
2e88a017
TF
749TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
750 Candler <B.Candler@pobox.com>.
751
a5f65aa4
TF
752TF/05 The fix for PH/43 was not completely correct; widen_domains is always
753 OK for addresses that are the result of redirections.
754
e7726cbf
PH
755PH/48 A number of further additions for the benefit of the new test suite,
756 including a fake gethostbyname() that interfaces to the fake DNS resolver
757 (see PH/47 above).
758
a7fdad5b
TF
759TF/06 The fix for widen_domains has also been applied to qualify_single and
760 search_parents which are the other dnslookup options that can cause
761 header rewrites.
762
6af56900
PH
763PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
764 type ("H").
765
0925ede6
PH
766PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
767
66afa403
TF
768TF/07 Exim produced the error message "an SRV record indicated no SMTP
769 service" if it encountered an MX record with an empty target hostname.
770 The message is now "an MX or SRV record indicated no SMTP service".
771
0154e85a
TF
772TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
773 if the DNS of the sending site is misconfigured. This is quite a
774 common situation. This change restores the behaviour of treating a
775 helo verification defer as a failure.
776
16f12c76
PH
777PH/51 If self=fail was set on a router, the bounce message did not include the
778 actual error message.
779
bbe902f0 780
e5a9dba6
PH
781Exim version 4.52
782-----------------
783
784TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
785
22c3b60b
PH
786PH/01 When a transport filter timed out in a pipe delivery, and the pipe
787 command itself ended in error, the underlying message about the transport
788 filter timeout was being overwritten with the pipe command error. Now the
789 underlying error message should be appended to the second error message.
790
06a9b4b5
PH
791TK/01 Fix poll() being unavailable on Mac OSX 10.2.
792
c1ac6996
PH
793PH/02 Reduce the amount of output that "make" produces by default. Full output
794 can still be requested.
795
9c7a242c
PH
796PH/03 The warning log line about a condition test deferring for a "warn" verb
797 was being output only once per connection, rather than after each
798 occurrence (because it was using the same function as for successful
799 "warn" verbs). This seems wrong, so I have changed it.
800
87ba3f5f
PH
801TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
802 it should not have, which might have caused a crash in the right
803 circumstances, but probably never did.
804
805PH/04 Installed a modified version of Tony Finch's patch to make submission
806 mode fix the return path as well as the Sender: header line, and to
807 add a /name= option so that you can make the user's friendly name appear
808 in the header line.
809
29aba418
TF
810TF/03 Added the control = fakedefer ACL modifier.
811
fe0dab11
TF
812TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
813 Mark Lowes for thorough testing.
870f6ba8 814
11d337a4
TK
815TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
816
817TK/03 Merged latest SRS patch from Miles Wilton.
818
415c8f3b
PH
819PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
820 with the definition in sysexits.h (which is #included earlier).
821 Fortunately, Exim does not actually use EX_OK. The code used to try to
822 preserve the sysexits.h value, by assumimg that macro definitions were
823 scanned for macro replacements. I have been disabused of this notion,
824 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 825
958541e9
PH
826PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
827 in the smtp transport. When a block could not be written in a single
828 write() function, the timeout was being re-applied to each part-write.
829 This seems wrong - if the receiver was accepting one byte at a time it
830 would take for ever. The timeout is now adjusted when this happens. It
831 doesn't have to be particularly precise.
832
c206415f
TK
833TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
834 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
835
2a4be8f9
PH
836PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
837 without a domain if the check to <postmaster@domain> fails.
838
1cba11c5
SC
839SC/01 Eximstats: added -xls and the ability to specify output files
840 (patch written by Frank Heydlauf).
841
842SC/02 Eximstats: use FileHandles for outputing results.
843
844SC/03 Eximstats: allow any combination of xls, txt, and html output.
845
846SC/04 Eximstats: fixed display of large numbers with -nvr option
847
848SC/05 Eximstats: fixed merging of reports with empty tables.
849
850SC/06 Eximstats: added the -include_original_destination flag
851
852SC/07 Eximstats: removed tabs and trailing whitespace.
853
1005d00e
TK
854TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
855
856TK/06 MBOX spool code: Add real "From " MBOX separator line
857 so the .eml file is really in mbox format (even though
858 most programs do not really care). Patch from Alex Miller.
859
860TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
861 The latter is generated from $received_to and is only set if the
862 message has one envelope recipient. SA can use these headers,
863 obviously out-of-the-box. Patch from Alex Miller.
864
9b4768fa
PH
865PH/08 The ${def test on a variable was returning false if the variable's
866 value was "0", contrary to what the specification has always said!
867 The result should be true unless the variable is empty.
868
869PH/09 The syntax error of a character other than { following "${if
870 def:variable_name" (after optional whitespace) was not being diagnosed.
871 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
872 accidental colon was present, for example, could give incorrect results.
873
0d7eb84a
PH
874PH/10 Tidied the code in a number of places where the st_size field of a stat()
875 result is used (not including appendfile, where other changes are about
876 to be made).
877
878PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
879 This involved changing a lot of size variables from int to off_t. It
880 should work with maildirs and everything.
881
40727bee
TK
882TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
883 spamd dying while we are connected to it.
884
554d2369
TF
885TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
886 <jgh@wizmail.org>
887
1f922db1
PH
888PH/12 Applied Alex Kiernan's patch for the API change for the error callback
889 function for BDB 4.3.
890
ef213c3b
PH
891PH/13 Changed auto_thaw such that it does not apply to bounce messages.
892
8ac170f3
PH
893PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
894 the sources for PCRE have been re-arranged and more files are now
895 involved.
896
b1c749bb
PH
897PH/15 The code I had for printing potentially long long variables in PH/11
898 above was not the best (it lost precision). The length of off_t variables
899 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
900 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
901 to be "long long int" or "long int". This is needed for the internal
902 formatting function string_vformat().
b1c749bb 903
4aac9b49
PH
904PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
905 the configuration file to be ":syslog", then the script "guesses" where
906 the logs files are, rather than using the compiled in default. In our
907 case the guess is not the same as the compiled default, so the script
908 suddenly stopped working when I started to use syslog. The patch checks
909 to see if log_file_path is "". If so, it attempts to read it from exim
910 with no configuration file to get the compiled in version, before it
911 falls back to the previous guessing code."
912
294520c8
TK
913TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
914 implementing BATV in an Exim configuration. See NewStuff for the gory
915 details.
916
5bd022fe
PH
917PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
918 Makefile that are specific to HP-UX.
919
90e9ce59
PH
920PH/18 If the "use_postmaster" option was set for a recipient callout together
921 with the "random" option, the postmaster address was used as the MAIL
922 FROM address for the random test, but not for the subsequent recipient
923 test. It is now used for both.
924
5ea81592
PH
925PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
926 patch removes a few documentation additions to RFC 3028, because the
927 latest draft now contains them. It adds the new en;ascii-case comparator
928 and a new error check for 8bit text in MIME parts. Comparator and
929 require names are now matched exactly. I enabled the subaddress
930 extension, but it is not well tested yet (read: it works for me)."
931
c6c2dc1d
PH
932PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
933 rework some of the code of TK/09 above to avoid the hardwired use of
934 "%lld" and "long long". Replaced the call to snprintf() with a call to
935 string_vformat().
936
fffffe4c
PH
937PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
938 records point to non-existent hosts", "retry timeout exceeded", and
939 "retry time not reached for any host after a long failure period".
ca02eafb 940
9a26b6b2
PH
941PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
942 experimental DomainKeys support:
943
944 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
945 (2) On an error such as an illegally used "control", the wrong name for
946 the control was given.
947
948 These problems did NOT occur unless DomainKeys support was compiled.
949
4aee0225
PH
950PH/23 Added daemon_startup_retries and daemon_startup_sleep.
951
32d668a5
PH
952PH/24 Added ${if match_ip condition.
953
8187c3f3
PH
954PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
955 databases so that it will be absolutely obvious if a crash occurs in the
956 DB library. This is a regular occurrence (often caused by mis-matched
957 db.h files).
958
ff790e47 959PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
960 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
961 picked up on a user's system that detects such things. There doesn't seem
962 to be a gcc warning option for this - only an attribute that has to be
963 put on the function's prototype. It seems that in Fedora Core 4 they have
964 set this on a number of new functions. No doubt there will be more in due
965 course.
ff790e47 966
5417f6d1
PH
967PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
968 specify a transport. However, if an address that was verified by such a
969 router was the subject of a callout, Exim crashed because it tried to
970 read the rcpt_include_affixes from the non-existent transport. Now it
971 just assumes that the setting of that option is false. This bug was
972 introduced by 4.51/PH/31.
973
59cf8544
PH
974PH/28 Changed -d+all to exclude +memory, because that information is very
975 rarely of interest, but it makes the output a lot bigger. People tend to
976 do -d+all out of habit.
977
e7ad8a65
PH
978PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
979 code in os-type was giving problems when libc.so lives in lib64, like on
980 x86_64 Fedora Core.
981
ade42478
PH
982PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
983 aren't the modern standard, and it seems that some systems' include files
984 don't always have them. Exim was already checking for some of the newer
985 ones like T_AAAA, and defining it itself. I've added checks for all the
986 record types that Exim uses.
987
182ad5cf
PH
988PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
989 not automatically generating a new one, as it is supposed to. This
990 prevented TLS from working. If the file did exist, but contained invalid
991 data, a new version was generated, as expected. It was only the case of a
992 non-existent file that was broken.
993
b0d9fc80
TK
994TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
995 with a change in libdomainkeys > 0.64.
996
997TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
998 from DNS. If the selector record carries the flag, it now has
999 precedence over the domain-wide flag.
1000
1001TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1002
47c7a64a
PH
1003PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1004 the use of a port name that isn't defined in /etc/services) occurred, the
1005 message was deferred as in a normal delivery, and thus remained on the
1006 spool, instead of being failed because of the mua_wrapper setting. This
1007 is now fixed, and I tidied up some of the mua_wrapper messages at the
1008 same time.
1009
a388bce4
SC
1010SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1011 the messages in a hash of arrays rather than using individual hashes.
1012 This is a bit cleaner and results in dramatic memory savings, albeit
1013 at a slight CPU cost.
1014
1015SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1016 as requested by Marc Sherman.
1017
1018SC/10 Eximstats: added histograms for user specified patterns as requested
1019 by Marc Sherman.
1020
0793e4ed
SC
1021SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1022
c58b88df
PH
1023PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1024 fopen() in the content-scanning modules that did not already have it.
1025
e7ad8a65 1026
7982096b
PH
1027Exim version 4.51
1028-----------------
1029
1a46a8c5
PH
1030TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1031 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1032
2f079f46 1033TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
1034
1035TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1036
1037PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1038 to test Sieve filters that use "vacation".
1039
1040PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1041 that changes the way the GnuTLS parameters are stored in the cache file.
1042 The new format can be generated externally. For backward compatibility,
1043 if the data in the cache doesn't make sense, Exim assumes it has read an
1044 old-format file, and it generates new data and writes a new file. This
1045 means that you can't go back to an older release without removing the
1046 file.
1047
1048PH/03 A redirect router that has both "unseen" and "one_time" set does not
1049 work if there are any delivery delays because "one_time" forces the
1050 parent to be marked "delivered", so its unseen clone is never tried
1051 again. For this reason, Exim now forbids the simultaneous setting of
1052 these two options.
1053
1054PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1055 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1056 entry if you want to know the details. The fix, however, neglected to
1057 consider the case when local delivery batching is involved. The test for
1058 "previously delivered" was not happening when checking to see if an
1059 address could be batched with a previous (undelivered) one; under
1060 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 1061 address.
1a46a8c5
PH
1062
1063PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1064 in its include files, and this causes problems building Exim.
1065
1066PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1067 header_syntax) but Exim was just ignoring anything given after a slash.
1068 In particular, this caused confusion with an attempt to use "verify =
1069 reverse_host_lookup/defer_ok". An error is now given when options are
1070 supplied for verify items that do not have them. (Maybe reverse_host_
1071 lookup should have a defer_ok option, but that's a different point.)
1072
1073PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1074 defined by RFC 821) to 2048, because there were problems with some AUTH
1075 commands, and RFC 1869 says the size should be increased for extended
1076 SMTP commands that take arguments.
1077
1078PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1079 Finch).
1080
1081PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1082 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 1083
49c2d5ea
PH
1084PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1085 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
1086 whatever). Otherwise the password may appear in the log. Following change
1087 PH/42 below, there is no longer a chance of it appearing in a bounce
1088 message.
49c2d5ea 1089
bf759a8b
PH
1090PH/11 Installed exipick version 20050225.0 from John Jetmore.
1091
83364d30
PH
1092PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1093 of its addresses was ever tried. (Bugzilla bug #2.)
1094
7999bbd7
PH
1095PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1096 the result incorrectly in the debug output. (It correctly added a newline
1097 to what was transported.)
1098
7dbf77c9
PH
1099TF/01 Added $received_time.
1100
74e0617f
PH
1101PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1102 commented out examples of how to interface to a virus scanner and to
1103 SpamAssassin. Also added commented examples of av_scanner and
1104 spamd_address settings.
1105
2f079f46
PH
1106PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1107 and controls are allowed in which ACLs. There were a couple of minor
1108 errors. Some of the entries in the conditions table (which is a table of
1109 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1110 negation of where the condition IS allowed.
1111
8c841523
PH
1112PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1113
7766a4f0
PH
1114PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1115 header file does not have a version number, so I've had to invent a new
1116 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1117 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1118 radiusclient), but it was contributed by a Radius user.
1119
8b417f2c
PH
1120PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1121 files or directories, for OpenSSL.
1122
901f42cb
PH
1123PH/19 When an Exim process that is running as root has to create an Exim log
1124 file, it does so in a subprocess that runs as exim:exim so as to get the
1125 ownership right at creation (otherwise, other Exim processes might see
1126 the file with the wrong ownership). There was no test for failure of this
1127 fork() call, which would lead to the process getting stuck as it waited
1128 for a non-existent subprocess. Forks do occasionally fail when resources
1129 run out. I reviewed all the other calls to fork(); they all seem to check
1130 for failure.
1131
f9b9210e
PH
1132PH/20 When checking for unexpected SMTP input at connect time (before writing
1133 the banner), Exim was not dealing correctly with a non-positive return
1134 from the read() function. If the client had disconnected by this time,
1135 the result was a log entry for a synchronization error with an empty
1136 string after "input=" when read() returned zero. If read() returned -1
1137 (an event I could not check), uninitialized data bytes were printed.
1138 There were reports of junk text (parts of files, etc) appearing after
1139 "input=".
1140
54cdb463
PH
1141PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1142
cf00dad6
PH
1143PH/22 Added support for macro redefinition, and (re)definition in between
1144 driver and ACL definitions.
1145
acb1b346
PH
1146PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1147 forgetting to use the resulting value; it was using the unexpanded value.
1148
c5ddb310
PH
1149PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1150 hadn't been configured. The fix is from Juergen Kreileder, who
1151 understands it better than I do:
1152
1153 "Here's what I see happening with three configured cyrus_sasl
1154 authenticators configured (plain, login, cram-md5):
1155
1156 On startup auth_cyrus_sasl_init() gets called for each of these.
1157 This means three calls to sasl_listmech() without a specified mech_list.
1158 => SASL tests which mechs of all available mechs actually work
1159 => three warnings about OTP not working
1160 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1161
1162 With the patch, sasl_listmech() also gets called three times. But now
1163 SASL's mech_list option is set to the server_mech specified in the the
1164 authenticator. Or in other words, the answer from sasl_listmech()
1165 gets limited to just the mech you're testing for (which is different
1166 for each call.)
1167 => the return list contains just 'plain' or 'login', 'cram-md5' or
1168 nothing depending on the value of ob->server_mech.
1169
1170 I've just tested the patch: Authentication still works fine,
1171 unavailable mechs specified in the exim configuration are still
1172 caught, and the auth.log warnings about OTP are gone."
1173
31619da6
PH
1174PH/25 When debugging is enabled, the contents of the command line are added
1175 to the debugging output, even when log_selector=+arguments is not
1176 specified.
1177
bebaf0fc
PH
1178PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1179 answer is "GNU", and only if the return is "GNU/something" is the answer
1180 "Linux".
1181
475fe28a
PH
1182PH/27 $acl_verify_message is now set immediately after the failure of a
1183 verification in an ACL, and so is available in subsequent modifiers. In
1184 particular, the message can be preserved by coding like this:
1185
1186 warn !verify = sender
1187 set acl_m0 = $acl_verify_message
1188
1189 Previously, $acl_verify_message was set only while expanding "message"
1190 and "log_message" when a very denied access.
1191
7e8bec7a
PH
1192PH/28 Modified OS/os.c-Linux with
1193
1194 -#ifndef OS_LOAD_AVERAGE
1195 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1196
1197 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1198 nomenclature these days.)
1199
e4a89c47
PH
1200PH/29 Installed patch from the Sieve maintainer that adds the options
1201 sieve_useraddress and sieve_subaddress to the redirect router.
1202
5ca2a9a1
PH
1203PH/30 In these circumstances:
1204 . Two addresses routed to the same list of hosts;
1205 . First host does not offer TLS;
1206 . First host accepts first address;
1207 . First host gives temporary error to second address;
1208 . Second host offers TLS and a TLS session is established;
1209 . Second host accepts second address.
1210 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1211 and peerdn, if requested) that were in fact used only for the second
1212 address.
7e8bec7a 1213
c688b954
PH
1214PH/31 When doing a callout as part of verifying an address, Exim was not paying
1215 attention to any local part prefix or suffix that was matched by the
1216 router that accepted the address. It now behaves in the same way as it
1217 does for delivery: the affixes are removed from the local part unless
1218 rcpt_include_affixes is set on the transport.
1219
fed77020
PH
1220PH/32 Add the sender address, as F=<...>, to the log line when logging a
1221 timeout during the DATA phase of an incoming message.
1222
7fe1560f
PH
1223PH/33 Sieve envelope tests were broken for match types other than :is. I have
1224 applied a patch sanctioned by the Sieve maintainer.
c688b954 1225
ebb6e6d5
PH
1226PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1227 the uid or gid is negative. A case of a negative gid caused this to be
1228 noticed. The fix allows for either to be negative.
1229
9c4e8f60
PH
1230PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1231 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1232 overlooked.
1233
1234PH/36 The change PH/12 above was broken. Fixed it.
1235
d7174846
PH
1236PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1237 the grounds that routing the same address twice would always produce the
1238 same answer. This might have been true once, but it is certainly no
1239 longer true now. Routing a child address may depend on the previous
1240 routing that produced that child. Some complicated redirection strategies
1241 went wrong when messages had multiple recipients, and made Exim's
1242 behaviour dependent on the order in which the addresses were given.
1243
1244 I have moved the duplicate checking until after the routing is complete.
1245 Exim scans the addresses that are assigned to local and remote
1246 transports, and removes any duplicates. This means that more work will be
1247 done, as duplicates will always all be routed, but duplicates are
1248 presumably rare, so I don't expect this is of any significance.
1249
1250 For deliveries to pipes, files, and autoreplies, the duplicate checking
1251 still happens during the routing process, since they are not going to be
1252 routed further.
1253
cfe75fc3
PH
1254PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1255 It corrects a timeout issue with spamd. This is Ian's comment: "The
1256 background is that sometimes spamd either never reads data from a
1257 connection it has accepted, or it never writes response data. The exiscan
1258 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1259 blindly assumes that writes won't block so it may never time out."
1260
be22d70e
PH
1261PH/39 Allow G after quota size as well as K and M.
1262
0612b098
PH
1263PH/40 The value set for $authenticated_id in an authenticator may not contain
1264 binary zeroes or newlines because the value is written to log lines and
1265 to spool files. There was no check on this. Now the value is run through
1266 the string_printing() function so that such characters are converted to
1267 printable escape sequences.
1268
2e0c1448
PH
1269PH/41 $message_linecount is a new variable that contains the total number of
1270 lines in the message. Compare $body_linecount, which is the count for the
1271 body only.
1272
447d236c
PH
1273PH/42 Exim no longer gives details of delivery errors for specific addresses in
1274 bounce and delay warning messages, except in certain special cases, which
1275 are as follows:
1276
1277 (a) An SMTP error message from a remote host;
1278 (b) A message specified in a :fail: redirection;
1279 (c) A message specified in a "fail" command in a system filter;
1280 (d) A message specified in a FAIL return from the queryprogram router;
1281 (e) A message specified by the cannot_route_message router option.
1282
1283 In these cases only, Exim does include the error details in bounce and
1284 warning messages. There are also a few cases where bland messages such
1285 as "unrouteable address" or "local delivery error" are given.
1286
d20976dc
PH
1287PH/43 $value is now also set for the "else" part of a ${run expansion.
1288
f656d135
PH
1289PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1290 being worked on, but at least Exim now implements the latest version to
1291 play with."
1292
2e2a30b4
PH
1293PH/45 In a pipe transport, although a timeout while waiting for the pipe
1294 process to complete was treated as a delivery failure, a timeout while
1295 writing the message to the pipe was logged, but erroneously treated as a
1296 successful delivery. Such timeouts include transport filter timeouts. For
1297 consistency with the overall process timeout, these timeouts are now
1298 treated as errors, giving rise to delivery failures by default. However,
1299 there is now a new Boolean option for the pipe transport called
1300 timeout_defer, which, if set TRUE, converts the failures into defers for
1301 both kinds of timeout. A transport filter timeout is now identified in
1302 the log output.
1303
9176e9f0
PH
1304PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1305 systems where "make" and "gmake" are different, calling "gmake" at top
1306 level broke things. I've arranged for the value of $(MAKE) to be passed
1307 from the Makefile to this script so that it can call the same version of
1308 "make".
1309
7982096b 1310
bbe902f0
PH
1311A note about Exim versions 4.44 and 4.50
1312----------------------------------------
1313
1314Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1315changes of various kinds. As a consequence, a big documentation update was
1316needed. This delayed the release for rather longer than seemed good, especially
1317in the light of a couple of (minor) security issues. Therefore, the changes
1318that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1319release. So 4.44 and 4.50 are in effect two different branches that both start
1320from 4.43.
1321
1322I have left the 4.50 change log unchanged; it contains all the changes since
13234.43. The change log for 4.44 is below; many of its items are identical to
1324those for 4.50. This seems to be the most sensible way to preserve the
1325historical information.
1326
1327
f7b63901 1328Exim version 4.50
495ae4b0
PH
1329-----------------
1330
5fe762f6
PH
1331 1. Minor wording change to the doc/README.SIEVE file.
1332
139059f6 1333 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1334 computation of the current number of files was incorrect.
495ae4b0 1335
7086e875
PH
1336 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1337 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1338 place.
1339
35af9f61
PH
1340 4. Give more explanation in the error message when the command for a transport
1341 filter fails to execute.
1342
b668c215
PH
1343 5. There are several places where Exim runs a non-Exim command in a
1344 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1345 was being done only for the command run by the queryprogram router. It is
1346 now done for all such subprocesses. The other cases are: ${run, transport
1347 filters, and the commands run by the lmtp and pipe transports.
1348
a494b1e1
PH
1349 6. Added CONFIGURE_GROUP build-time option.
1350
1351 7. Some older OS have a limit of 256 on the maximum number of file
1352 descriptors. Exim was using setrlimit() to set 1000 as a large value
1353 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1354 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1355
c5fcb476
PH
1356 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1357 was an oversight, and furthermore, ever since the addition of extra
1358 controls (e.g. 4.43/32), the checks on when to allow different forms of
1359 "control" were broken. There should now be diagnostics for all cases when a
1360 control that does not make sense is encountered.
1361
69358f02
PH
1362 9. Added the /retain_sender option to "control=submission".
1363
5be20824
PH
136410. $recipients is now available in the predata ACL (oversight).
1365
eb2c0248
PH
136611. Tidy the search cache before the fork to do a delivery from a message
1367 received from the command line. Otherwise the child will trigger a lookup
1368 failure and thereby defer the delivery if it tries to use (for example) a
1369 cached ldap connection that the parent has called unbind on.
1370
2a3eea10
PH
137112. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1372 of $address_data from the recipient verification was clobbered by the
1373 sender verification.
1374
137513. The value of address_data from a sender verification is now available in
1376 $sender_address_data in subsequent conditions in the ACL statement.
1377
23c7ff99
PH
137814. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1379
4deaf07d
PH
138015. Added a new option "connect=<time>" to callout options, to set a different
1381 connection timeout.
1382
926e1192
PH
138316. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1384 was its contents. (It was OK if the option was not defined at all.)
1385
650edc6f
PH
138617. A "Completed" log line is now written for messages that are removed from
1387 the spool by the -Mrm option.
1388
2c7db3f5
PH
138918. New variables $sender_verify_failure and $recipient_verify_failure contain
1390 information about exactly what failed.
1391
3d235903
PH
139219. Added -dd to debug only the daemon process.
1393
7c7ad977
PH
139420. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1395 handles timeouts, both on the server side and network timeouts. Renamed the
1396 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1397
981756db
PH
139821. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1399 It is now set to "smtps".
1400
d4eb88df
PH
140122. $host_address is now set to the target address during the checking of
1402 ignore_target_hosts.
1403
140423. When checking ignore_target_hosts for an ipliteral router, no host name was
1405 being passed; this would have caused $sender_host_name to have been used if
1406 matching the list had actually called for a host name (not very likely,
1407 since this list is usually IP addresses). A host name is now passed as
1408 "[x.x.x.x]".
1409
7d468ab8
PH
141024. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1411 code that specifies a non-restarting handler (typically sigaction() in
1412 modern systems) in an attempt to fix a rare and obscure crash bug.
1413
141425. Narrowed the window for a race in the daemon that could cause it to ignore
1415 SIGCHLD signals. This is not a major problem, because they are used only to
1416 wake it up if nothing else does.
1417
62c0818f
PH
141826. A malformed maildirsize file could cause Exim to calculate negative values
1419 for the mailbox size or file count. Odd effects could occur as a result.
1420 The maildirsize information is now recalculated if the size or filecount
1421 end up negative.
1422
26034054
PH
142327. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1424 support for a long time. Removed HAVE_SYS_VFS_H.
1425
af66f652
PH
142628. Installed the latest version of exipick from John Jetmore.
1427
90af77f4
PH
142829. In an address list, if the pattern was not a regular expression, an empty
1429 subject address (from a bounce message) matched only if the pattern was an
1430 empty string. Non-empty patterns were not even tested. This was the wrong
1431 because it is perfectly reasonable to use an empty address as part of a
1432 database query. An empty address is now tested by patterns that are
1433 lookups. However, all the other forms of pattern expect the subject to
1434 contain a local part and a domain, and therefore, for them, an empty
1435 address still always fails if the pattern is not itself empty.
1436
d8ef3577
PH
143730. Exim went into a mad DNS loop when attempting to do a callout where the
1438 host was specified on an smtp transport, and looking it up yielded more
1439 than one IP address.
1440
5cb8cbc6
PH
144131. Re-factored the code for checking spool and log partition space into a
1442 function that finds that data and another that does the check. The former
1443 is then used to implement four new variables: $spool_space, $log_space,
1444 $spool_inodes, and $log_inodes.
1445
14702f5b
PH
144632. The RFC2047 encoding function was originally intended for short strings
1447 such as real names; it was not keeping to the 75-character limit for
1448 encoded words that the RFC imposes. It now respects the limit, and
1449 generates multiple encoded words if necessary. To be on the safe side, I
1450 have increased the buffer size for the ${rfc2047: expansion operator from
1451 1024 to 2048 bytes.
1452
063b1e99
PH
145333. It is now permitted to omit both strings after an "if" condition; if the
1454 condition is true, the result is "true". As before, when the second string
1455 is omitted, a false condition yields an empty string. This makes it less
1456 cumbersome to write custom ACL and router conditions.
1457
652e1b65
PH
145834. Failure to deliver a bounce message always caused it to be frozen, even if
1459 there was an errors_to setting on the router. The errors_to setting is now
1460 respected.
1461
6f0c9a4f
PH
146235. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1463 canonical form (fully expanded) before being placed in
1464 $sender_host_address.
1465
33397d19
PH
146636. The table in the code that translates DNS record types into text (T_A to
1467 "A" for instance) was missing entries for NS and CNAME. It is just possible
1468 that this could have caused confusion if both these types were looked up
1469 for the same domain, because the text type is used as part of Exim's
1470 per-process caching. But the chance of anyone hitting this buglet seems
1471 very small.
1472
7bb56e1f
PH
147337. The dnsdb lookup has been extended in a number of ways.
1474
1475 (1) There is a new type, "zns", which walks up the domain tree until it
1476 finds some nameserver records. It should be used with care.
1477
ea3bc19b
PH
1478 (2) There is a new type, "mxh", which is like "mx" except that it returns
1479 just the host names, not the priorities.
1480
1481 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1482 looked up. The behaviour when one of the lookups defers can be
1483 controlled by a keyword.
7bb56e1f 1484
ea3bc19b 1485 (4) It is now possible to specify the separator character for use when
7bb56e1f 1486 multiple records are returned.
33397d19 1487
0bcb2a0e
PH
148838. The dnslists ACL condition has been extended: it is now possible to supply
1489 a list of IP addresses and/or domains to be looked up in a particular DNS
1490 domain.
1491
2ac0e484
PH
149239. Added log_selector=+queue_time_overall.
1493
4e1fde53
PH
149440. When running the queue in the test harness, wait just a tad after forking a
1495 delivery process, to get repeatability of debugging output.
1496
de365ded
PH
149741. Include certificate and key file names in error message when GnuTLS fails
1498 to set them up, because the GnuTLS error message doesn't include the name
1499 of the failing file when there is a problem reading it.
1500
f05da2e8
PH
150142. Allow both -bf and -bF in the same test run.
1502
d6453af2
PH
150343. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1504
f7b63901
PH
150544. The "Exiscan patch" is now merged into the mainline Exim source.
1506
150745. Sometimes the final signoff response after QUIT could fail to get
1508 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1509 < 0 before doing a fflush(). This bug looks as though it goes back to the
1510 introduction of TLS in release 3.20, but "sometimes" must have been rare
1511 because the tests only now provoked it.
1512
a444213a
PH
151346. Reset the locale to "C" after calling embedded Perl, in case it was changed
1514 (this can affect the format of dates).
1515
0ec020ea
PH
151647. exim_tidydb, when checking for the continued existence of a message for
1517 which it has found a message-specific retry record, was not finding
1518 messages that were in split spool directories. Consequently, it was
1519 deleting retry records that should have stayed in existence.
1520
b1206957
PH
152148. Steve fixed some bugs in eximstats.
1522
152349. The SPA authentication driver was not abandoning authentication and moving
1524 on to the next authenticator when an expansion was forced to fail,
1525 contradicting the general specification for all authenticators. Instead it
1526 was generating a temporary error. It now behaves as specified.
1527
26dd5a95
PH
152850. The default ordering of permitted cipher suites for GnuTLS was pessimal
1529 (the order specifies the preference for clients). The order is now AES256,
1530 AES128, 3DES, ARCFOUR128.
1531
343b2385
PH
153251. Small patch to Sieve code - explicitly set From: when generating an
1533 autoreply.
1534
1c5466b9
PH
153552. Exim crashed if a remote delivery caused a very long error message to be
1536 recorded - for instance if somebody sent an entire SpamAssassin report back
1537 as a large number of 550 error lines. This bug was coincidentally fixed by
1538 increasing the size of one of Exim's internal buffers (big_buffer) that
1539 happened as part of the Exiscan merge. However, to be on the safe side, I
1540 have made the code more robust (and fixed the comments that describe what
1541 is going on).
1542
55ee9ee3
PH
154353. Now that there can be additional text after "Completed" in log lines (if
1544 the queue_time_overall log selector is set), a one-byte patch to exigrep
1545 was needed to allow it to recognize "Completed" as not the last thing in
1546 the line.
1547
d38f8232
PH
154854. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1549 patch that reportedly fixes this has been added. I am not expert enough to
1550 create a test for it. This is what the patch creator wrote:
1551
1552 "I found a little strange behaviour of ldap code when working with
1553 Windows 2003 AD Domain, where users was placed in more than one
1554 Organization Units. When I tried to give exim partial DN, the exit code
1555 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1556 But simultaneously result of request was absolutely normal ldap result,
1557 so I produce this patch..."
1558
3295e65b
PH
1559 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1560 REFERENCE, so I have modified the code to exclude the patch when that macro
1561 is not defined.
1562
7102e136
PH
156355. Some experimental protocols are using DNS PTR records for new purposes. The
1564 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1565 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1566 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1567 CAN-2005-0021
7102e136 1568
3ca0ba97
PH
156956. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1570
c2bcbe20
PH
157157. Double the size of the debug message buffer (to 2048) so that more of very
1572 long debug lines gets shown.
1573
18ce445d
PH
157458. The exicyclog utility now does better if the number of log files to keep
1575 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1576
1f5b4c3d
PH
157759. Two changes related to the smtp_active_hostname option:
1578
1579 (1) $smtp_active_hostname is now available as a variable.
1580 (2) The default for smtp_banner uses $smtp_active_hostname instead
1581 of $primary_hostname.
1582
b975ba52
PH
158360. The host_aton() function is supposed to be passed a string that is known
1584 to be a valid IP address. However, in the case of IPv6 addresses, it was
1585 not checking this. This is a hostage to fortune. Exim now panics and dies
1586 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1587 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1588 components; fortuitously, this particular loophole had already been fixed
1589 by change 4.50/55 above.
1590
1591 If there are any other similar loopholes, the new check in host_aton()
1592 itself should stop them being exploited. The report I received stated that
1593 data on the command line could provoke the exploit when Exim was running as
1594 exim, but did not say which command line option was involved. All I could
1595 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1596 running as the user.
ea3a6f44 1597 CAN-2005-0021
85b87bc2
PH
1598
159961. There was a buffer overflow vulnerability in the SPA authentication code
1600 (which came originally from the Samba project). I have added a test to the
1601 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1602 CAN-2005-0022
b975ba52 1603
17ffcae7
PH
160462. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1605 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1606
d95f9fdb
PH
160763. The daemon start-up calls getloadavg() while still root for those OS that
1608 need the first call to be done as root, but it missed one case: when
1609 deliver_queue_load_max is set with deliver_drop_privilege. This is
1610 necessary for the benefit of the queue runner, because there is no re-exec
1611 when deliver_drop_privilege is set.
1612
86b8287f
PH
161364. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1614 This has been fixed.
1615
60dc5e56
PH
161665. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1617 was in use, was not putting the data itself into the right store pool;
1618 consequently, it could be overwritten for a subsequent message in the same
1619 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1620 the caching.)
1621
533244af
PH
162266. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1623
a5a28604
PH
162467. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1625 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1626 the function were treating the return as a boolean value, which happened to
1627 work because 0=false and not-0=true, but is not correct code.
1628
7e634d24
PH
162968. The host_aton() function was not handling scoped IPv6 addresses (those
1630 with, for example, "%eth0" on the end) correctly.
1631
3e11c26b
PH
163269. Fixed some compiler warnings in acl.c for the bitmaps specified with
1633 negated items (that is, ~something) in unsigned ints. Some compilers
1634 apparently mutter when there is no cast.
1635
6729cf78
PH
163670. If an address verification called from an ACL failed, and did not produce a
1637 user-specific message (i.e. there was only a "system" message), nothing was
1638 put in $acl_verify_message. In this situation, it now puts the system
1639 message there.
1640
00f00ca5
PH
164171. Change 4.23/11 added synchronization checking at the start of an SMTP
1642 session; change 4.31/43 added the unwanted input to the log line - except
1643 that it did not do this in the start of session case. It now does.
1644
c9bdd01c
PH
164572. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1646 This is wrong and can cause the other end to generate a synchronization
1647 error if it is another Exim or anything else that does the synchronization
1648 check. A QUIT command is no longer sent after a timeout.
1649
d43194df
PH
165073. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1651 during host lookups.
1652
fe5b5d0b
PH
165374. The defer_ok option of callout verification was not working if it was used
1654 when verifying addresses in header lines, that is, for this case:
1655
1656 verify = header_sender/callout=defer_ok
1657
76a2d7ba
PH
165875. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1659 those file descriptors could be used for SMTP connections. If anything
1660 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1661 could be sent to the SMTP client, causing chaos. The daemon now opens
1662 stdin, stdout, and stderr to /dev/null when it puts itself into the
1663 background.
1664
166576. Arrange for output from Perl's "warn" command to be written to Exim's main
1666 log by default. The user can override this with suitable Perl magic.
1667
04f7d5b9
PH
166877. The use of log_message on a "discard" ACL verb, which is supposed to add to
1669 the log message when discard triggers, was not working for the DATA ACL or
1670 for the non-SMTP ACL.
1671
bc60667e
PH
167278. Error message wording change in sieve.c.
1673
bb6e88ff
PH
167479. If smtp_accept_max_per_host was set, the number of connections could be
1675 restricted to fewer than expected, because the daemon was trying to set up
1676 a new connection before checking whether the processes handling previous
1677 connections had finished. The check for completed processes is now done
1678 earlier. On busy systems, this bug wouldn't be noticed because something
1679 else would have woken the daemon, and it would have reaped the completed
1680 process earlier.
1681
1e70f85b
PH
168280. If a message was submitted locally by a user whose login name contained one
1683 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1684 It caused a spool format error. I have fixed the spool reading code. A
1685 related problem was that the "from" clause in the Received: line became
1686 illegal because of the space(s). It is now covered by ${quote_local_part.
1687
168881. Included the latest eximstats from Steve (adds average sizes to HTML Top
1689 tables).
1690
4e01f9d6
PH
169182. Updated OS/Makefile-AIX as per message from Mike Meredith.
1692
1ee1cef2
PH
169383. Patch from Sieve maintainer to fix unterminated string problem in
1694 "vacation" handling.
1695
6e2b4ccc
PH
169684. Some minor changes to the Linux configuration files to help with other
1697 OS variants using glibc.
1698
8e669ac1
PH
169985. One more patch for Sieve to update vacation handling to latest spec.
1700
495ae4b0 1701
bbe902f0
PH
1702----------------------------------------------------
1703See the note above about the 4.44 and 4.50 releases.
1704----------------------------------------------------
1705
1706
1707Exim version 4.44
1708-----------------
1709
1710 1. Change 4.43/35 introduced a bug that caused file counts to be
1711 incorrectly computed when quota_filecount was set in an appendfile
1712 transport
1713
1714 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1715 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1716 place.
1717
1718 3. Give more explanation in the error message when the command for a transport
1719 filter fails to execute.
1720
1721 4. There are several places where Exim runs a non-Exim command in a
1722 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1723 was being done only for the command run by the queryprogram router. It is
1724 now done for all such subprocesses. The other cases are: ${run, transport
1725 filters, and the commands run by the lmtp and pipe transports.
1726
1727 5. Some older OS have a limit of 256 on the maximum number of file
1728 descriptors. Exim was using setrlimit() to set 1000 as a large value
1729 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1730 systems. I've change it so that if it can't get 1000, it tries for 256.
1731
1732 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1733 was an oversight, and furthermore, ever since the addition of extra
1734 controls (e.g. 4.43/32), the checks on when to allow different forms of
1735 "control" were broken. There should now be diagnostics for all cases when a
1736 control that does not make sense is encountered.
1737
1738 7. $recipients is now available in the predata ACL (oversight).
1739
1740 8. Tidy the search cache before the fork to do a delivery from a message
1741 received from the command line. Otherwise the child will trigger a lookup
1742 failure and thereby defer the delivery if it tries to use (for example) a
1743 cached ldap connection that the parent has called unbind on.
1744
1745 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1746 of $address_data from the recipient verification was clobbered by the
1747 sender verification.
1748
174910. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1750 was its contents. (It was OK if the option was not defined at all.)
1751
175211. A "Completed" log line is now written for messages that are removed from
1753 the spool by the -Mrm option.
1754
175512. $host_address is now set to the target address during the checking of
1756 ignore_target_hosts.
1757
175813. When checking ignore_target_hosts for an ipliteral router, no host name was
1759 being passed; this would have caused $sender_host_name to have been used if
1760 matching the list had actually called for a host name (not very likely,
1761 since this list is usually IP addresses). A host name is now passed as
1762 "[x.x.x.x]".
1763
176414. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1765 code that specifies a non-restarting handler (typically sigaction() in
1766 modern systems) in an attempt to fix a rare and obscure crash bug.
1767
176815. Narrowed the window for a race in the daemon that could cause it to ignore
1769 SIGCHLD signals. This is not a major problem, because they are used only to
1770 wake it up if nothing else does.
1771
177216. A malformed maildirsize file could cause Exim to calculate negative values
1773 for the mailbox size or file count. Odd effects could occur as a result.
1774 The maildirsize information is now recalculated if the size or filecount
1775 end up negative.
1776
177717. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1778 support for a long time. Removed HAVE_SYS_VFS_H.
1779
ea3a6f44 178018. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1781
178219. Allow an empty sender to be matched against a lookup in an address list.
1783 Previously the only cases considered were a regular expression, or an
1784 empty pattern.
1785
178620. Exim went into a mad DNS lookup loop when doing a callout where the
1787 host was specified on the transport, if the DNS lookup yielded more than
1788 one IP address.
1789
ea3a6f44
NM
179021. The RFC2047 encoding function was originally intended for short strings
1791 such as real names; it was not keeping to the 75-character limit for
1792 encoded words that the RFC imposes. It now respects the limit, and
1793 generates multiple encoded words if necessary. To be on the safe side, I
1794 have increased the buffer size for the ${rfc2047: expansion operator from
1795 1024 to 2048 bytes.
bbe902f0 1796
ea3a6f44
NM
179722. Failure to deliver a bounce message always caused it to be frozen, even if
1798 there was an errors_to setting on the router. The errors_to setting is now
1799 respected.
bbe902f0
PH
1800
180123. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1802 canonical form (fully expanded) before being placed in
1803 $sender_host_address.
1804
180524. Updated eximstats to version 1.33
1806
ea3a6f44
NM
180725. Include certificate and key file names in error message when GnuTLS fails
1808 to set them up, because the GnuTLS error message doesn't include the name
1809 of the failing file when there is a problem reading it.
bbe902f0
PH
1810
181126. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1812 As per change 25.
bbe902f0 1813
ea3a6f44
NM
181427. Reset the locale to "C" after calling embedded Perl, in case it was changed
1815 (this can affect the format of dates).
bbe902f0 1816
ea3a6f44
NM
181728. exim_tidydb, when checking for the continued existence of a message for
1818 which it has found a message-specific retry record, was not finding
1819 messages that were in split spool directories. Consequently, it was
1820 deleting retry records that should have stayed in existence.
bbe902f0
PH
1821
182229. eximstats updated to version 1.35
1823 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1824 1.35 - bugfix such that pie charts by volume are generated correctly
1825
ea3a6f44
NM
182630. The SPA authentication driver was not abandoning authentication and moving
1827 on to the next authenticator when an expansion was forced to fail,
1828 contradicting the general specification for all authenticators. Instead it
1829 was generating a temporary error. It now behaves as specified.
bbe902f0 1830
ea3a6f44
NM
183131. The default ordering of permitted cipher suites for GnuTLS was pessimal
1832 (the order specifies the preference for clients). The order is now AES256,
1833 AES128, 3DES, ARCFOUR128.
bbe902f0 1834
ea3a6f44
NM
183531. Small patch to Sieve code - explicitly set From: when generating an
1836 autoreply.
bbe902f0 1837
ea3a6f44
NM
183832. Exim crashed if a remote delivery caused a very long error message to be
1839 recorded - for instance if somebody sent an entire SpamAssassin report back
1840 as a large number of 550 error lines. This bug was coincidentally fixed by
1841 increasing the size of one of Exim's internal buffers (big_buffer) that
1842 happened as part of the Exiscan merge. However, to be on the safe side, I
1843 have made the code more robust (and fixed the comments that describe what
1844 is going on).
bbe902f0 1845
ea3a6f44
NM
184633. Some experimental protocols are using DNS PTR records for new purposes. The
1847 keys for these records are domain names, not reversed IP addresses. The
1848 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1849 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1850 CAN-2005-0021
1851
ea3a6f44
NM
185234. The host_aton() function is supposed to be passed a string that is known
1853 to be a valid IP address. However, in the case of IPv6 addresses, it was
1854 not checking this. This is a hostage to fortune. Exim now panics and dies
1855 if the condition is not met. A case was found where this could be provoked
1856 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1857 components; fortuitously, this particular loophole had already been fixed
1858 by change 4.50/55 or 4.44/33 above.
1859
1860 If there are any other similar loopholes, the new check in host_aton()
1861 itself should stop them being exploited. The report I received stated that
1862 data on the command line could provoke the exploit when Exim was running as
1863 exim, but did not say which command line option was involved. All I could
1864 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1865 running as the user.
bbe902f0
PH
1866 CAN-2005-0021
1867
ea3a6f44
NM
186835. There was a buffer overflow vulnerability in the SPA authentication code
1869 (which came originally from the Samba project). I have added a test to the
1870 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1871 CAN-2005-0022
1872
ea3a6f44
NM
187336. The daemon start-up calls getloadavg() while still root for those OS that
1874 need the first call to be done as root, but it missed one case: when
1875 deliver_queue_load_max is set with deliver_drop_privilege. This is
1876 necessary for the benefit of the queue runner, because there is no re-exec
1877 when deliver_drop_privilege is set.
bbe902f0 1878
ea3a6f44
NM
187937. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1880 was in use, was not putting the data itself into the right store pool;
1881 consequently, it could be overwritten for a subsequent message in the same
1882 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1883 the caching.)
bbe902f0 1884
ea3a6f44
NM
188538. Sometimes the final signoff response after QUIT could fail to get
1886 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1887 < 0 before doing a fflush(). This bug looks as though it goes back to the
1888 introduction of TLS in release 3.20, but "sometimes" must have been rare
1889 because the tests only now provoked it.
bbe902f0
PH
1890
1891
495ae4b0
PH
1892Exim version 4.43
1893-----------------
1894
1895 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1896 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1897 Now it returns an integer. A number of calls were still expecting a T/F
1898 return. Fortuitously, in all cases, the tests worked in OK situations,
1899 which is the norm. However, things would have gone wrong on any write
1900 failures on the smtp file descriptor. This function is used when sending
1901 messages over SMTP and also when doing verify callouts.
1902
1903 2. When Exim is called to do synchronous delivery of a locally submitted
1904 message (the -odf or -odi options), it no longer closes stderr before doing
1905 the delivery.
1906
1907 3. Implemented the mua_wrapper option.
1908
1909 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1910
1911 5. Implemented the functions header_remove(), header_testname(),
1912 header_add_at_position(), and receive_remove_recipient(), and exported them
1913 to local_scan().
1914
1915 6. If an ACL "warn" statement specified the addition of headers, Exim already
1916 inserted X-ACL-Warn: at the start if there was no header name. However, it
1917 was not making this test for the second and subsequent header lines if
1918 there were newlines in the string. This meant that an invalid header could
1919 be inserted if Exim was badly configured.
1920
1921 7. Allow an ACL "warn" statement to add header lines at the start or after all
1922 the Received: headers, as well as at the end.
1923
1924 8. Added the rcpt_4xx retry error code.
1925
1926 9. Added postmaster_mailfrom=xxx to callout verification option.
1927
192810. Added mailfrom=xxxx to the callout verification option, for verify=
1929 header_sender only.
1930
193111. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1932 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1933
193412. Inserted some casts to stop certain compilers warning when using pointer
1935 differences as field lengths or precisions in printf-type calls (mostly
1936 affecting debugging statements).
1937
193813. Added optional readline() support for -be (dynamically loaded).
1939
194014. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1941 same clock tick as a message's arrival, so that its received time was the
1942 same as the "first fail" time on the retry record, and that message
1943 remained on the queue past the ultimate address timeout, every queue runner
1944 would try a delivery (because it was past the ultimate address timeout) but
1945 after another failure, the ultimate address timeout, which should have then
1946 bounced the address, did not kick in. This was a "< instead of <=" error;
1947 in most cases the first failure would have been in the next clock tick
1948 after the received time, and all would be well.
1949
195015. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1951 being recognized when the domain list was tested by the match_domain
1952 condition in an expansion string.
1953
195416. Added the ${str2b64: operator.
1955
195617. Exim was always calling setrlimit() to set a large limit for the number of
1957 processes, without checking whether the existing limit was already
1958 adequate. (It did check for the limit on file descriptors.) Furthermore,
1959 errors from getrlimit() and setrlimit() were being ignored. Now they are
1960 logged to the main and panic logs, but Exim does carry on, to try to do its
1961 job under whatever limits there are.
1962
196318. Imported PCRE 5.0.
1964
196519. Trivial typo in log message " temporarily refused connection" (the leading
1966 space).
1967
196820. If the log selector return_path_on_delivery was set and an address was
1969 redirected to /dev/null, the delivery process crashed because it assumed
1970 that a return path would always be set for a "successful" delivery. In this
1971 case, the whole delivery is bypassed as an optimization, and therefore no
1972 return path is set.
1973
197421. Internal re-arrangement: the function for sending a challenge and reading
1975 a response while authentication was assuming a zero-terminated challenge
1976 string. It's now changed to take a pointer and a length, to allow for
1977 binary data in such strings.
1978
197922. Added the cyrus_sasl authenticator (code supplied by MBM).
1980
198123. Exim was not respecting finduser_retries when seeking the login of the
1982 uid under which it was called; it was always trying 10 times. (The default
1983 setting of finduser_retries is zero.) Also, it was sleeping after the final
1984 failure, which is pointless.
1985
198624. Implemented tls_on_connect_ports.
1987
198825. Implemented acl_smtp_predata.
1989
199026. If the domain in control=submission is set empty, Exim assumes that the
1991 authenticated id is a complete email address when it generates From: or
1992 Sender: header lines.
1993
199427. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1995 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1996 chown and chgrp in /bin and hostname in /usr/bin.
1997
199828. Exim was keeping the "process log" file open after each use, just as it
1999 does for the main log. This opens the possibility of it remaining open for
2000 long periods when the USR1 signal hits a daemon. Occasional processlog
2001 errors were reported, that could have been caused by this. Anyway, it seems
2002 much more sensible not to leave this file open at all, so that is what now
2003 happens.
2004
200529. The long-running daemon process does not normally write to the log once it
2006 has entered its main loop, and it closes the log before doing so. This is
2007 so that log files can straightforwardly be renamed and moved. However,
2008 there are a couple of unusual error situations where the daemon does write
2009 log entries, and I had neglected to close the log afterwards.
2010
201130. The text of an SMTP error response that was received during a remote
2012 delivery was being truncated at 512 bytes. This is too short for some of
2013 the long messages that one sometimes sees. I've increased the limit to
2014 1024.
2015
201631. It is now possible to make retry rules that apply only when a message has a
2017 specific sender, in particular, an empty sender.
2018
201932. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2020 it possible to be selective about when SMTP synchronization is enforced.
2021
202233. Added "control = caseful_local_part" and "control = "caselower_local_part".
2023
202432. Implemented hosts_connection_nolog.
2025
202633. Added an ACL for QUIT.
2027
202834. Setting "delay_warning=" to disable warnings was not working; it gave a
2029 syntax error.
2030
203135. Added mailbox_size and mailbox_filecount to appendfile.
2032
203336. Added control = no_multiline_responses to ACLs.
2034
203537. There was a bug in the logic of the code that waits for the clock to tick
2036 in the case where the clock went backwards by a substantial amount such
2037 that the microsecond fraction of "now" was more than the microsecond
2038 fraction of "then" (but the whole seconds number was less).
2039
204038. Added support for the libradius Radius client library this is found on
2041 FreeBSD (previously only the radiusclient library was supported).
2042
2043
2044Exim version 4.42
2045-----------------
2046
2047 1. When certain lookups returned multiple values in the form name=value, the
2048 quoting of the values was not always being done properly. Specifically:
2049 (a) If the value started with a double quote, but contained no whitespace,
2050 it was not quoted.
2051 (b) If the value contained whitespace other than a space character (i.e.
2052 tabs or newlines or carriage returns) it was not quoted.
2053 This fix has been applied to the mysql and pgsql lookups by writing a
2054 separate quoting function and calling it from the lookup code. The fix
2055 should probably also be applied to nisplus, ibase and oracle lookups, but
2056 since I cannot test any of those, I have not disturbed their existing code.
2057
2058 2. A hit in the callout cache for a specific address caused a log line with no
2059 reason for rejecting RCPT. Now it says "Previous (cached) callout
2060 verification failure".
2061
2062 3. There was an off-by-one bug in the queryprogram router. An over-long
2063 return line was truncated at 256 instead of 255 characters, thereby
2064 overflowing its buffer with the terminating zero. As well as fixing this, I
2065 have increased the buffer size to 1024 (and made a note to document this).
2066
2067 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2068 when Exim is waiting for an SMTP response from a remote server, Exim
2069 restarts its select() call on the socket, thereby resetting its timeout.
2070 This is not a problem when such interrupts are rare. Somebody set up a cron
2071 job to run exiwhat every 2 minutes, which is less than the normal select()
2072 timeout (5 or 10 minutes). This meant that the select() timeout never
2073 kicked in because it was always reset. I have fixed this by comparing the
2074 time when an interrupt arrives with the time at the start of the first call
2075 to select(). If more time than the timeout has elapsed, the interrupt is
2076 treated as a timeout.
2077
2078 5. Some internal re-factoring in preparation for the addition of Sieve
2079 extensions (by MH). In particular, the "personal" test is moved to a
2080 separate function, and given an option for scanning Cc: and Bcc: (which is
2081 not set for Exim filters).
2082
2083 6. When Exim created an email address using the login of the caller as the
2084 local part (e.g. when creating a From: or Sender: header line), it was not
2085 quoting the local part when it contained special characters such as @.
2086
2087 7. Installed new OpenBSD configuration files.
2088
2089 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2090 try to make them clearer.
2091
2092 9. Callout options, other than the timeout value, were being ignored when
2093 verifying sender addresses in header lines. For example, when using
2094
2095 verify = header_sender/callout=no_cache
2096
2097 the cache was (incorrectly) being used.
2098
209910. Added a missing instance of ${EXE} to the exim_install script; this affects
2100 only the Cygwin environment.
2101
210211. When return_path_on_delivery was set as a log selector, if different remote
2103 addresses in the same message used different return paths and parallel
2104 remote delivery occurred, the wrong values would sometimes be logged.
2105 (Whenever a remote delivery process finished, the return path value from
2106 the most recently started remote delivery process was logged.)
2107
210812. RFC 3848 specifies standard names for the "with" phrase in Received: header
2109 lines when AUTH and/or TLS are in use. This is the "received protocol"
2110 field. Exim used to use "asmtp" for authenticated SMTP, without any
2111 indication (in the protocol name) for TLS use. Now it follows the RFC and
2112 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2113 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2114 names appear in log lines as well as in Received: header lines.
2115
211613. Installed MH's patches for Sieve to add the "copy" and "vacation"
2117 extensions, and comparison tests, and to fix some bugs.
2118
211914. Changes to the "personal" filter test:
2120
2121 (1) The test was buggy in that it was just doing the equivalent of
2122 "contains" tests on header lines. For example, if a user's address was
2123 anne@some.where, the "personal" test would incorrectly be true for
2124
2125 To: susanne@some.where
2126
2127 This test is now done by extracting each address from the header in turn,
2128 and checking the entire address. Other tests that are part of "personal"
2129 are now done using regular expressions (for example, to check local parts
2130 of addresses in From: header lines).
2131
2132 (2) The list of non-personal local parts in From: addresses has been
2133 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2134 taken from the Sieve specification recommendations.
2135
2136 (3) If the message contains any header line starting with "List-" it is
2137 treated as non-personal.
2138
2139 (4) The test for "circular" in the Subject: header line has been removed
2140 because it now seems ill-conceived.
2141
214215. Minor typos in src/EDITME comments corrected.
2143
214416. Installed latest exipick from John Jetmore.
2145
214617. If headers_add on a router specified a text string that was too long for
2147 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2148 of string_sprintf() is now avoided.
2149
215018. $message_body_size was not set (it was always zero) when running the DATA
2151 ACL and the local_scan() function.
2152
215319. For the "mail" command in an Exim filter, no default was being set for
2154 the once_repeat time, causing a random time value to be used if "once" was
2155 specified. (If the value happened to be <= 0, no repeat happened.) The
2156 default is now 0s, meaning "never repeat". The "vacation" command was OK
2157 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2158 (I found it when inspecting the code).
2159
216020. There is now an overall timeout for performing a callout verification. It
2161 defaults to 4 times the callout timeout, which applies to individual SMTP
2162 commands during the callout. The overall timeout applies when there is more
2163 than one host that can be tried. The timeout is checked before trying the
2164 next host. This prevents very long delays if there are a large number of
2165 hosts and all are timing out (e.g. when the network connections are timing
2166 out). The value of the overall timeout can be changed by specifying an
2167 additional sub-option for "callout", called "maxwait". For example:
2168
2169 verify = sender/callout=5s,maxwait=20s
2170
217121. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2172 to the end before writing, but this should make it even safer).
2173
217422. Exim was forgetting that it had advertised PIPELINING for the second and
2175 subsequent messages on an SMTP connection. It was also not resetting its
2176 memory on STARTTLS and an internal HELO.
2177
217823. When Exim logs an SMTP synchronization error within a session, it now
2179 records whether PIPELINING has been advertised or not.
2180
218124. Added 3 instances of "(long int)" casts to time_t variables that were being
2182 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2183 rather than long int.
2184
218525. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2186
218726. Added the never_mail option to autoreply.
2188
2189
2190Exim version 4.41
2191-----------------
2192
2193 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2194 crash if the getsockname() call failed; this can happen if a connection is
2195 closed very soon after it is established. The problem was simply in the
2196 order in which certain operations were done, causing Exim to try to write
2197 to the SMTP stream before it had set up the file descriptor. The bug has
2198 been fixed by making things happen in the correct order.
2199
2200
2201Exim version 4.40
2202-----------------
2203
2204 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2205 before the connection was closed, thus losing the rejection response.
2206
2207 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2208 some early Solaris releases, but causes trouble in current releases where
2209 socklen_t is defined.
2210
2211 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2212 always exist.
2213
2214 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2215 configured.
2216
2217 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2218 that releasing the top of it at the end releases what was used for sub-
2219 expansions (unless the block got too big). However, discard this block if
2220 the first thing is a variable or header, so that we can use its block when
2221 it is dynamic (useful for very large $message_headers, for example).
2222
2223 6. Lookups now cache *every* query, not just the most recent. A new, separate
2224 store pool is used for this. It can be recovered when all lookup caches are
2225 flushed. Lookups now release memory at the end of their result strings.
2226 This has involved some general refactoring of the lookup sources.
2227
2228 7. Some code has been added to the store_xxx() functions to reduce the amount
2229 of flapping under certain conditions.
2230
2231 8. log_incoming_interface used to affect only the <= reception log lines. Now
2232 it causes the local interface and port to be added to several more SMTP log
2233 lines, for example "SMTP connection from", and rejection lines.
2234
2235 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2236
223710. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2238
223911. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2240 could be overwritten at the end of the current message (or the start of a
2241 new message if it was set in a HELO ACL). The value is now preserved for
2242 the duration of the SMTP connection.
2243
224412. If a transport had a headers_rewrite setting, and a matching header line
2245 contained an unqualified address, that address was qualified, even if it
2246 did not match any rewriting rules. The underlying bug was that the values
2247 of the flags that permit the existence of unqualified sender and recipient
2248 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2249 non-local messages, and by -bnq for local messages) were not being
2250 preserved with the message after it was received.
2251
225213. When Exim was logging an SMTP synchronization error, it could sometimes log
2253 "next input=" as part of the text comprising the host identity instead of
2254 the correct text. The code was using the same buffer for two different
2255 strings. However, depending on which order the printing function evaluated
2256 its arguments, the bug did not always show up. Under Linux, for example, my
2257 test suite worked just fine.
2258
225914. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2260 doesn't work with some older versions of Perl. It has been changed to "my",
2261 which in any case is probably the better facility to use.
2262
226315. A really picky compiler found some instances of statements for creating
2264 error messages that either had too many or two few arguments for the format
2265 string.
2266
226716. The size of the buffer for calls to the DNS resolver has been increased
2268 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2269 for addresses that have a lot of PTR records. This alleviates a problem; it
2270 does not fully solve it.
2271
227217. A dnsdb lookup for PTR records that receives more data than will fit in the
2273 buffer now truncates the list and logs the incident, which is the same
2274 action as happens when Exim is looking up a host name and its aliases.
2275 Previously in this situation something unpredictable would happen;
2276 sometimes it was "internal error: store_reset failed".
2277
227818. If a server dropped the connection unexpectedly when an Exim client was
2279 using GnuTLS and trying to read a response, the client delivery process
2280 crashed while trying to generate an error log message.
2281
228219. If a "warn" verb in an ACL added multiple headers to a message in a single
2283 string, for example:
2284
2285 warn message = H1: something\nH2: something
2286
2287 the text was added as a single header line from Exim's point of view
2288 though it ended up OK in the delivered message. However, searching for the
2289 second and subsequent header lines using $h_h2: did not work. This has been
2290 fixed. Similarly, if a system filter added multiple headers in this way,
2291 the routers could not see them.
2292
229320. Expanded the error message when iplsearch is called with an invalid key to
2294 suggest using net-iplsearch in a host list.
2295
229621. When running tests using -bh, any delays imposed by "delay" modifiers in
2297 ACLs are no longer actually imposed (and a message to that effect is
2298 output).
2299
230022. If a "gecos" field in a passwd entry contained escaped characters, in
2301 particular, if it contained a \" sequence, Exim got it wrong when building
2302 a From: or a Sender: header from that name. A second bug also caused
2303 incorrect handling when an unquoted " was present following a character
2304 that needed quoting.
2305
230623. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2307 was not being matched caselessly.
2308
230924. Arranged for all hyphens in the exim.8 source to be escaped with
2310 backslashes.
2311
231225. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2313 properly. Recipient callout cache records were still being keyed to include
2314 the sender, even when use_sender was set false. This led to far more
2315 callouts that were necessary. The sender is no longer included in the key
2316 when use_sender is false.
2317
231826. Added "control = submission" modifier to ACLs.
2319
232027. Added the ${base62d: operator to decode base 62 numbers.
2321
232228. dnsdb lookups can now access SRV records.
2323
232429. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2325 the configuration file.
2326
232730. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2328 (-v) mode. This makes the output for a verbose queue run more intelligible.
2329
233031. Added a use_postmaster feature to recipient callouts.
2331
233232. Added the $body_zerocount variable, containing the number of binary zero
2333 bytes in the message body.
2334
233533. The time of last modification of the "new" subdirectory is now used as the
2336 "mailbox time last read" when there is a quota error for a maildir
2337 delivery.
2338
233934. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2340
234135. Added +ignore_unknown as a special item in host lists.
2342
234336. Code for decoding IPv6 addresses in host lists is now included, even if
2344 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2345 address was recognized as an IP address, but was then not correctly decoded
2346 into binary, causing unexpected and incorrect effects when compared with
2347 another IP address.
2348
2349
2350Exim version 4.34
2351-----------------
2352
2353 1. Very minor rewording of debugging text in manualroute to say "list of
2354 hosts" instead of "hostlist".
2355
2356 2. If verify=header_syntax was set, and a header line with an unqualified
2357 address (no domain) and a large number of spaces between the end of the
2358 name and the colon was received, the reception process suffered a buffer
2359 overflow, and (when I tested it) crashed. This was caused by some obsolete
2360 code that should have been removed. The fix is to remove it!
2361
2362 3. When running in the test harness, delay a bit after writing a bounce
2363 message to get a bit more predictability in the log output.
2364
2365 4. Added a call to search_tidyup() just before forking a reception process. In
2366 theory, someone could use a lookup in the expansion of smtp_accept_max_
2367 per_host which, without the tidyup, could leave open a database connection.
2368
2369 5. Added the variables $recipient_data and $sender_data which get set from a
2370 lookup success in an ACL "recipients" or "senders" condition, or a router
2371 "senders" option, similar to $domain_data and $local_part_data.
2372
2373 6. Moved the writing of debug_print from before to after the "senders" test
2374 for routers.
2375
2376 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2377 problems for message scanning, either using a data ACL, or using
2378 local_scan() because the Received: header was not generated till after they
2379 were called (in order to set the time as the time of reception completion).
2380 I have revised the way this works. The header is now generated after the
2381 body is received, but before the ACL or local_scan() are called. After they
2382 are run, the timestamp in the header is updated.
2383
2384
2385Exim version 4.33
2386-----------------
2387
2388 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2389 before starting a queue runner without re-exec. This happened only when
2390 deliver_drop_privilege was set or when the Exim user was set to root. The
2391 effect of the bug was that timeouts during subsequent deliveries caused
2392 crashes instead of being properly handled. The handler is now left at its
2393 default (and expected) setting.
2394
2395 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2396 message, again when deliver_drop_privilege is set or Exim is run as root.
2397 The bug described in (1) was not present in this case, but the tidying up
2398 of the other signals was missing. I have made the two cases consistent.
2399
2400 3. The ignore_target_hosts setting on a manualroute router was being ignored
2401 for hosts that were looked up using the /MX notation.
2402
2403 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2404 in domain lists.
2405
2406 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2407 operated on the sender address. After changing the $sender_address to <>
2408 for the sender address verify, Exim was re-instated it as the original
2409 (before rewriting) address, but remembering that it had rewritten it, so it
2410 wasn't rewriting it again. This bug also had the effect of breaking the
2411 sender address verification caching when the sender address was rewritten.
2412
2413 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2414 This has been changed so that if the ip literal address matches
2415 ignore_target_hosts, the router declines.
2416
2417 7. Added expansion conditions match_domain, match_address, and match_local_
2418 part (NOT match_host).
2419
2420 8. The placeholder for the Received: header didn't have a length field set.
2421
2422 9. Added code to Exim itself and to exim_lock to test for a specific race
2423 condition that could lead to file corruption when using MBX delivery. The
2424 issue is with the lockfile that is created in /tmp. If this file is removed
2425 after a process has opened it but before that process has acquired a lock,
2426 there is the potential for a second process to recreate the file and also
2427 acquire a lock. This could lead to two Exim processes writing to the file
2428 at the same time. The added code performs the same test as UW imapd; it
2429 checks after acquiring the lock that its file descriptor still refers to
2430 the same named file.
2431
243210. The buffer for building added header lines was of fixed size, 8192 bytes.
2433 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2434 when Exim is built.
2435
243611. Added the smtp_active_hostname option. If used, this will typically be made
2437 to depend on the incoming interface address. Because $interface_address is
2438 not set up until the daemon has forked a reception process, error responses
2439 that can happen earlier (such as "too many connections") no longer contain
2440 a host name.
2441
244212. If an expansion in a condition on a "warn" statement fails because a lookup
2443 defers, the "warn" statement is abandoned, and the next ACL statement is
2444 processed. Previously this caused the whole ACL to be aborted.
2445
244613. Added the iplsearch lookup type.
2447
244814. Added ident_timeout as a log selector.
2449
245015. Added tls_certificate_verified as a log selector.
2451
245216. Added a global option tls_require_ciphers (compare the smtp transport
2453 option of the same name). This controls incoming TLS connections.
2454
245517. I finally figured out how to make tls_require_ciphers do a similar thing
2456 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2457 before starting the TLS session.
2458
245918. Tabs are now shown as \t in -bP output.
2460
246119. If the log selector return_path_on_delivery was set, Exim crashed when
2462 bouncing a message because it had too many Received: header lines.
2463
246420. If two routers both had headers_remove settings, and the first one included
2465 a superfluous trailing colon, the final name in the first list and the
2466 first name in the second list were incorrectly joined into one item (with a
2467 colon in the middle).
2468
2469
2470Exim version 4.32
2471-----------------
2472
2473 1. Added -C and -D options to the exinext utility, mainly to make it easier
2474 to include in the automated testing, but these could be helpful when
2475 multiple configurations are in use.
2476
2477 2. The exinext utility was not formatting the output nicely when there was
2478 an alternate port involved in the retry record key, nor when there was a
2479 message id as well (for retries that were specific to a specific message
2480 and a specific host). It was also confused by IPv6 addresses, because of
2481 the additional colons they contain. I have fixed the IPv4 problem, and
2482 patched it up to do a reasonable job for IPv6.
2483
2484 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2485 delivery, the log line now contains "pipelined" if PIPELINING was used.
2486
2487 4. An SMTP transport process used to panic and die if the bind() call to set
2488 an explicit outgoing interface failed. This has been changed; it is now
2489 treated in the same way as a connect() failure.
2490
2491 5. A reference to $sender_host_name in the part of a conditional expansion
2492 that was being skipped was still causing a DNS lookup. This no longer
2493 occurs.
2494
2495 6. The def: expansion condition was not recognizing references to header lines
2496 that used bh_ and bheader_.
2497
2498 7. Added the _cache feature to named lists.
2499
2500 8. The code for checking quota_filecount in the appendfile transport was
2501 allowing one more file than it should have been.
2502
2503 9. For compatibility with Sendmail, the command line option
2504
2505 -prval:sval
2506
2507 is equivalent to
2508
2509 -oMr rval -oMs sval
2510
2511 and sets the incoming protocol and host name (for trusted callers). The
2512 host name and its colon can be omitted when only the protocol is to be set.
2513 Note the Exim already has two private options, -pd and -ps, that refer to
2514 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2515 "s", but I don't think that's a major issue.
2516
251710. A number of refactoring changes to the code, none of which should affect
2518 Exim's behaviour:
2519
2520 (a) The number of logging options was getting close to filling up the
2521 32-bit word that was used as a bit map. I have split them into two classes:
2522 those that are passed in the argument to log_write(), and those that are
2523 only ever tested independently outside of that function. These are now in
2524 separate 32-bit words, so there is plenty of room for expansion again.
2525 There is no change in the user interface or the logging behaviour.
2526
2527 (b) When building, for example, log lines, the code previously used a
2528 macro that called string_cat() twice, in order to add two strings. This is
2529 not really sufficiently general. Furthermore, there was one instance where
2530 it was actually wrong because one of the argument was used twice, and in
2531 one call a function was used. (As it happened, calling the function twice
2532 did not affect the overall behaviour.) The macro has been replaced by a
2533 function that can join an arbitrary number of extra strings onto a growing
2534 string.
2535
2536 (c) The code for expansion conditions now uses a table and a binary chop
2537 instead of a serial search (which was left over from when there were very
2538 few conditions). Also, it now recognizes conditions like "pam" even when
2539 the relevant support is not compiled in: a suitably worded error message is
2540 given if an attempt is made to use such a condition.
2541
254211. Added ${time_interval:xxxxx}.
2543
254412. A bug was causing one of the ddress fields not to be passed back correctly
2545 from remote delivery subprocesses. The field in question was not being
2546 subsequently used, so this caused to problems in practice.
2547
254813. Added new log selectors queue_time and deliver_time.
2549
255014. Might have fixed a bug in maildirsizefile handling that threw up
2551 "unexpected character" debug warnings, and recalculated the data
2552 unnecessarily. In any case, I expanded the warning message to give more
2553 information.
2554
255515. Added the message "Restricted characters in address" to the statements in
2556 the default ACL that block characters like @ and % in local parts.
2557
255816. Change 71 for release 4.31 proved to be much less benign that I imagined.
2559 Three changes have been made:
2560
2561 (a) There was a serious bug; a negative response to MAIL caused the whole
2562 recipient domain to be cached as invalid, thereby blocking all messages
2563 to all local parts at the same domain, from all senders. This bug has
2564 been fixed. The domain is no longer cached after a negative response to
2565 MAIL if the sender used is not empty.
2566
2567 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2568 been restored.
2569
2570 (c) A new callout option, "use_sender" has been added for people who want
2571 the modified behaviour.
2572
2573
2574Exim version 4.31
2575-----------------
2576
2577 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2578 Larry Rosenman.
2579
2580 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2581 indeed breaks things for older releases.
2582
2583 3. Added additional logging to the case where there is a problem reading data
2584 from a filter that is running in a subprocess using a pipe, in order to
2585 try to track down a specific problem.
2586
2587 4. Testing facility fudge: when running in the test harness and attempting
2588 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2589 getting "No route to host". Convert this to a timeout.
2590
2591 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2592 warning.
2593
2594 6. Some OS don't have socklen_t but use size_t instead. This affects the
2595 fifth argument of getsockopt() amongst other things. This is now
2596 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2597 can be set for individual OS. I have set it for SunOS5, OSF1, and
2598 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2599 some earlier ones do not.
2600
2601 7. Change 4.30/15 was not doing the test caselessly.
2602
2603 8. The standard form for an IPv6 address literal was being rejected by address
2604 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2605 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2606 this, as well as the form without the "IPv6" on the front (but only when
2607 address literals are enabled, of course).
2608
2609 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2610
261110. Exim crashed if a message with an empty sender address specified by -f
2612 encountered a router with an errors_to setting. This could be provoked only
2613 by a command such as
2614
2615 exim -f "" ...
2616
2617 where an empty string was supplied; "<>" did not hit this bug.
2618
261911. Installed PCRE release 4.5.
2620
262112. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2622 remained set. It is now erased.
2623
262413. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2625 times from message ids (which are base 36 rather than the normal 62).
2626
262714. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2628 were being counted as actual protocol errors, and logged if the log
2629 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2630 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2631 and DATA following a set of rejected RCPTs do not count as protocol errors.
2632 In other words, Exim assumes they were pipelined, though this may not
2633 actually be the case. Of course, in all cases the client gets an
2634 appropriate error code.
2635
263615. If a lookup fails in an ACL condition, a message about the failure may
2637 be available; it is used if testing the ACL cannot continue, because most
2638 such messages specify what the cause of the deferral is. However, some
2639 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2640 that caused an old message to be retained and used if a later statement
2641 caused a defer, replacing the real cause of the deferral.
2642
264316. If an IP address had so many PTR records that the DNS lookup buffer
2644 was not large enough to hold them, Exim could crash while trying to process
2645 the truncated data. It now detects and logs this case.
2646
264717. Further to 4.21/58, another change has been made: if (and only if) the
2648 first line of a message (the first header line) ends with CRLF, a bare LF
2649 in a subsequent header line has a space inserted after it, so as not to
2650 terminate the header.
2651
265218. Refactoring: tidied an ugly bit of code in appendfile that copied data
2653 unnecessarily, used atoi() instead of strtol(), and didn't check the
2654 termination when getting file sizes from file names by regex.
2655
265619. Completely re-implemented the support for maildirsize files, in the light
2657 of a number of problems with the previous contributed implementation
2658 (4.30/29). In particular:
2659
2660 . If the quota is zero, the maildirsize file is maintained, but no quota is
2661 imposed.
2662
2663 . If the maildir directory does not exist, it is created before any attempt
2664 to write a maildirsize file.
2665
2666 . The quota value in the file is just a cache; if the quota is changed in
2667 the transport, the new value overrides.
2668
2669 . A regular expression is available for excluding directories from the
2670 count.
2671
267220. The autoreply transport checks the characters in options that define the
2673 message's headers; it allows continued headers, but it was checking with
2674 isspace() after an embedded newline instead of explicitly looking for a
2675 space or a tab.
2676
267721. If all the "regular" hosts to which an address was routed had passed their
2678 expiry times, and had not reached their retry times, the address was
2679 bounced, even if fallback hosts were defined. Now Exim should go on to try
2680 the fallback hosts.
2681
268222. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2683 equivalent code in the SMTP transport. Some hosts send humungous responses
2684 to HELO/EHLO, more than 1024 it seems.
2685
268623. Refactoring: code in filter.c used (void *) for "any old type" but this
2687 gives compiler warnings in some environments. I've now done it "properly",
2688 using a union.
2689
269024. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2691 (because of problems with the built-in one) was declared to return uschar *
2692 instead of char *, causing compiler failure.
2693
269425. Fixed a file descriptor leak when processing alias/forward files.
2695
269626. Fixed a minor format string issue in dbfn.c.
2697
269827. Typo in exim.c: ("dmbnz" for "dbmnz").
2699
270028. If a filter file refered to $h_xxx or $message_headers, and the headers
2701 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2702 become corrupted.
2703
270429. When a sender address is verified, it is cached, to save repeating the test
2705 when there is more than one recipient in a message. However, when the
2706 verification involves a callout, it is possible for different callout
2707 options to be set for different recipients. It is too complicated to keep
2708 track of this in the cache, so now Exim always runs a verification when a
2709 callout is required, relying on the callout cache for the optimization.
2710 The overhead is duplication of the address routing, but this should not be
2711 too great.
2712
271330. Fixed a bug in callout caching. If a RCPT command caused the sender address
2714 to be verified with callout=postmaster, and the main callout worked but the
2715 postmaster check failed, the verification correctly failed. However, if a
2716 subsequent RCPT command asked for sender verification *without* the
2717 postmaster check, incorrect caching caused this verification also to fail,
2718 incorrectly.
2719
272031. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2721 it was not caching the DNS options (qualify_single, search_parents) that
2722 were used when the lookup failed. A subsequent lookup with different
2723 options therefore always gave the same answer, though there were cases
2724 where it should not have. (Example: a "domains = !$mx_any" option on a
2725 dnslookup router: the "domains" option is always processed without any
2726 widening, but the router might have qualify_single set.) Now Exim uses the
2727 cached value only when the same options are set.
2728
272932. Added John Jetmore's "exipick" utility to the distribution.
2730
273133. GnuTLS: When an attempt to start a TLS session fails for any reason other
2732 than a timeout (e.g. a certificate is required, and is not provided), an
2733 Exim server now closes the connection immediately. Previously it waited for
2734 the client to close - but if the client is SSL, it seems that they each
2735 wait for each other, leading to a delay before one of them times out.
2736
273734: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2738 maintained 0.8.x compatibility because I don't think many are using it, and
2739 it is clearly obsolete.
2740
274135. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2742 transport.
2743
274436. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2745 client certificate was expired. A simple patch fixes this, though I don't
2746 understand the full logic of why the verify callback is called multiple
2747 times.
2748
274937. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2750 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2751 which causes problems with some clients (such as the Certicom SSL Plus
2752 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2753 disables the coutermeasure allowing Eudora to connect."
2754
275538. Exim was not checking that a write() to a log file succeeded. This could
2756 lead to Bad Things if a log got too big, in particular if it hit a file
2757 size limit. Exim now panics and dies if it cannot write to a log file, just
2758 as it does if it cannot open a log file.
2759
276039. Modified OS/Makefile-Linux so that it now contains
2761
2762 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2763
2764 The two -D definitions ensure that Exim is compiled with large file
2765 support, which makes it possible to handle log files that are bigger than
2766 2^31.
2767
276840. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2769 instance) a domain was checked against a named list that involved a lookup,
2770 causing $domain_data to be set, then another domain was checked against the
2771 same list, then the first domain was re-checked, the value of $domain_data
2772 after the final check could be wrong. In particular, if the second check
2773 failed, it could be set empty. This bug probably also applied to
2774 $localpart_data.
2775
277641. The strip_trailing_dot option was not being applied to the address given
2777 with the -f command-line option.
2778
277942. The code for reading a message's header from the spool was incrementing
2780 $received_count, but never initializing it. This meant that the value was
2781 incorrect (doubled) while delivering a message in the same process in which
2782 it was received. In the most common configuration of Exim, this never
2783 happens - a fresh exec is done - but it can happen when
2784 deliver_drop_privilege is set.
2785
278643. When Exim logs an SMTP synchronization error - client data sent too soon -
2787 it now includes up to 150 characters of the unexpected data in the log
2788 line.
2789
279044. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2791 and building data strings. The size of both of these buffers was 10 000
2792 bytes - far larger than anybody would *ever* want, thought I. Needless to
2793 say, somebody hit the limit. I have increased the maximum line length to
2794 20 000 and the maximum data length of concatenated lines to 100 000. I have
2795 also fixed two bugs, because there was no checking on these buffers. Tsk,
2796 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2797 buffer is too small.
2798
279945. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2800 lsearch lookups. Now it does.
2801
280246. When parsing a route_list item in a manualroute router, a fixed-length
2803 buffer was used for the list of hosts. I made this 1024 bytes long,
2804 thinking that nobody would ever have a list of hosts that long. Wrong.
2805 Somebody had a whole pile of complicated expansion conditions, and the
2806 string was silently truncated, leading to an expansion error. It turns out
2807 that it is easier to change to an unlimited length (owing to other changes
2808 that have happened since this code was originally written) than to build
2809 structure for giving a limitation error. The length of the item that
2810 expands into the list of hosts is now unlimited.
2811
281247. The lsearch lookup could not handle data where the length of text line was
2813 more than 4095 characters. Such lines were truncated, leading to shortened
2814 data being returned. It should now handle lines of any length.
2815
281648. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2817 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2818 ACL").
2819
282049. Cosmetic tidy to scripts like exicyclog that are generated by globally
2821 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2822 no longer happens in comment lines. A list of replacements is now placed
2823 at the head of all of the source files, except those whose only change is
2824 to replace PERL_COMMAND in the very first #! line.
2825
282650. Replaced the slow insertion sort in queue.c, for sorting the list of
2827 messages on the queue, with a bottom-up merge sort, using code contributed
2828 by Michael Haardt. This should make operations like -bp somewhat faster on
2829 large queues. It won't affect queue runners, except when queue_run_in_order
2830 is set.
2831
283251. Installed eximstats 1.31 in the distribution.
2833
283452. Added support for SRV lookups to the dnslookup router.
2835
283653. If an ACL referred to $message_body or $message_body_end, the value was not
2837 reset for any messages that followed in the same SMTP session.
2838
283954. The store-handling optimization for building very long strings was not
2840 differentiating between the different store pools. I don't think this
2841 actually made any difference in practice, but I've tidied it.
2842
284355. While running the routers to verify a sender address, $sender_address
2844 was still set to the sender address. This is wrong, because when routing to
2845 send a bounce to the sender, it would be empty. Therefore, I have changed
2846 it so that, while verifying a sender address, $sender_address is set to <>.
2847 (There is no change to what happens when verifying a recipient address.)
2848
284956. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2850 target A or AAAA records (if not already returned) without resetting the
2851 qualify_single or search_parents options of the DNS resolver. These are
2852 inappropriate in this case because the targets of MX and SRV records must
2853 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2854 target that, when qualified, matched something in the local domain. These
2855 two options are now turned off when doing these lookups.
2856
285757. It seems that at least some releases of Reiserfs (which does not have the
2858 concept of a fixed number of inodes) returns zero and not -1 for the
2859 number of available inodes. This interacted badly with check_spool_inodes,
2860 which assumed that -1 was the "no such thing" setting. What I have done is
2861 to check that the total number of inodes is greater than zero before doing
2862 the test of how many are available.
2863
286458. When a "warn" ACL statement has a log_message modifier, the message is
2865 remembered, and not repeated. This is to avoid a lot of repetition when a
2866 message has many recipients that cause the same warning to be written.
2867 Howewer, Exim was preserving the list of already written lines for an
2868 entire SMTP session, which doesn't seem right. The memory is now reset if a
2869 new message is started.
2870
287159. The "rewrite" debugging flag was not showing the result of rewriting in the
2872 debugging output unless log_rewrite was also set.
2873
287460. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2875 of (int)(handle) when we know that handle contains (void *)(-1).
2876
287761. The Exim daemon panic-logs an error return when it closes the incoming
2878 connection. However "connection reset by peer" seems to be common, and
2879 isn't really an error worthy of noting specially, so that particular error
2880 is no long logged.
2881
288262. When Exim is trying to find all the local interfaces, it used to panic and
2883 die if the ioctl to get the interface flags failed. However, it seems that
2884 on at least one OS (Solaris 9) it is possible to have an interface that is
2885 included in the list of interfaces, but for which you get a failure error
2886 for this call. This happens when the interface is not "plumbed" into a
2887 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2888 failure of the "get flags" call assumes that the interface is down.
2889
289063. Added a ${eval10: operator, which assumes all numbers are decimal. This
2891 makes life easier for people who are doing arithmetic on fields extracted
2892 from dates, where you often get leading zeros that should not be
2893 interpreted as octal.
2894
289564. Added qualify_domain to the redirect router, to override the global
2896 setting.
2897
289865. If a pathologically long header line contained very many addresses (the
2899 report of this problem mentioned 10 000) and each of them was rewritten,
2900 Exim could use up a very large amount of memory. (It kept on making new
2901 copies of the header line as it rewrote, and never released the old ones.)
2902 At the expense of a bit more processing, the header rewriting function has
2903 been changed so that it no longer eats memory in this way.
2904
290566. The generation of the Received: header has been moved from the time that a
2906 message starts to be received, to the time that it finishes. The timestamp
2907 in the Received: header should now be very close to that of the <= log
2908 line. There are two side-effects of this change:
2909
2910 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2911 logged header lines no longer include the local Received: line, because
2912 it has not yet been created. The same applies to a copy of the message
2913 that is returned to a non-SMTP sender when a message is rejected.
2914
2915 (b) When a filter file is tested using -bf, no additional Received: header
2916 is added to the test message. After some thought, I decided that this
2917 is a bug fix.
2918
2919 This change does not affect the value of $received_for. It is still set
2920 after address rewriting, but before local_scan() is called.
2921
292267. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2923
292468. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2925 gave an unhelpful panic error message, and a defer error. I have managed to
2926 change this behaviour so that it now rejects any supplied certificate,
2927 which seems right, as the list of acceptable certificates is empty.
2928
292969. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2930 gave an unhelpful defer error. I have not managed to make this reject any
2931 supplied certificates, but the error message it gives is "no certificate
2932 supplied", which is not helpful.
2933
293470. exigrep's output now also includes lines that are not associated with any
2935 message, but which match the given pattern. Implemented by a patch from
2936 Martin Sluka, which also tidied up the Perl a bit.
2937
293871. Recipient callout verification, like sender verification, was using <> in
2939 the MAIL FROM command. This isn't really the right thing, since the actual
2940 sender may affect whether the remote host accepts the recipient or not. I
2941 have changed it to use the actual sender in the callout; this means that
2942 the cache record is now keyed on a recipient/sender pair, not just the
2943 recipient address. There doesn't seem to be a real danger of callout loops,
2944 since a callout by the remote host to check the sender would use <>.
2945 [SEE ABOVE: changed after hitting problems.]
2946
294772. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2948 temporary errors. However, in the case of such a code being given after
2949 the end of a data transmission (i.e. after ".") Exim was failing to write
2950 a retry record for the message. (Yes, there was some broken host that was
2951 actually sending 8xx at this point.)
2952
295373. An unknown lookup type in a host list could cause Exim to panic-die when
2954 the list was checked. (An example that provoked this was putting <; in the
2955 middle of a list instead of at the start.) If this happened during a DATA
2956 ACL check, a -D file could be left lying around. This kind of configuration
2957 error no longer causes Exim to die; instead it causes a defer errror. The
2958 incident is still logged to the main and panic logs.
2959
296074. Buglet left over from Exim 3 conversion. The message "too many messages
2961 in one connection" was written to the rejectlog but not the mainlog, except
2962 when address rewriting (yes!) was being logged.
2963
296475. Added write_rejectlog option.
2965
296676. When a system filter was run not as root (that is, when system_filter_user
2967 was set), the values of the $n variables were not being returned to the
2968 main process; thus, they were not subsequently available in the $sn
2969 variables.
2970
297177. Added +return_path_on_delivery log selector.
2972
297378. A connection timeout was being treated differently from recipients deferred
2974 when testing hosts_max_try with a message that was older than the host's
2975 retry timeout. (The host should not be counted, thus allowing all hosts to
2976 be tried at least once before bouncing.) This may have been the cause of an
2977 occasionally reported bug whereby a message would remain on the queue
2978 longer than the retry timeout, but would be bounced if a delivery was
2979 forced. I say "may" because I never totally pinned down the problem;
2980 setting up timeout/retry tests is difficult. See also the next item.
2981
298279. The ultimate address timeout was not being applied to errors that involved
2983 a combination of host plus message (for example, a timeout on a MAIL
2984 command). When an address resolved to a number of possible hosts, and they
2985 were not all tried for each delivery (e.g. because of hosts_max_try), a
2986 message could remain on the queue longer than the retry timeout.
2987
298880. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2989 Haardt.
2990
299181. Fixed an obscure SMTP outgoing bug which required at least the following
2992 conditions: (a) there was another message waiting for the same server;
2993 (b) the server returned 5xx to all RCPT commands in the first message so
2994 that the message was not completed; (c) the server dropped the connection
2995 or gave a negative response to the RSET that Exim sends to abort the
2996 transaction. The observed case was a dropped connection after DATA that had
2997 been sent in pipelining mode. That is, the server had advertised PIPELINING
2998 but was not implementing it correctly. The effect of the bug was incorrect
2999 behaviour, such as trying another host, and this could lead to a crash.
3000
3001
3002Exim version 4.30
3003-----------------
3004
3005 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3006 and daemon.c were passed as pointers to ints; they should have been
3007 pointers to socklen_t variables (which are typically unsigned ints).
3008
3009 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3010 fixed.
3011
3012 3. Fixed a really odd bug that affected only the testing scheme; patching a
3013 certain fixed string in the binary changed the value of another string that
3014 happened to be identical to the end of the original first string.
3015
3016 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3017 name", it returns that address as the IP address. On some operating
3018 systems (e.g. Solaris), it also passes back the IP address string as the
3019 "host name". However, on others (e.g. Linux), it passes back an empty
3020 string. Exim wasn't checking for this, and was changing the host name to an
3021 empty string, assuming it had been canonicized.
3022
3023 5. Although rare, it is permitted to have more than one PTR record for a given
3024 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3025 all the names associated with an address, because they do in Solaris.
3026 However, it seems that they do not in Linux for data that comes from the
3027 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3028 I found this out when I moved to a new Linux workstation and tried to run
3029 the Exim test suite.
3030
3031 To get round this problem I have changed the code so that it now does its
3032 own call to the DNS to look up PTR records when searching for a host name.
3033 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3034 addresses that are only in /etc/hosts are still found.
3035
3036 This behaviour is, however, controlled by an option called host_lookup_
3037 order, which defaults to "bydns:byaddr". If people want to use the other
3038 order, or indeed, just use one or the other means of lookup, they can
3039 specify it in this variable.
3040
3041 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3042 some operating systems, this comes back from gethostbyaddr() as an empty
3043 string, and this is what Exim used to test for. However, it seems that in
3044 other systems, "." is yielded. Exim now tests for this case too.
3045
3046 7. The values of check_spool_space and check_log_space are now held internally
3047 as a number of kilobytes instead of an absolute number of bytes. If a
3048 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3049 kilobyte. This means that much larger values can be stored.
3050
3051 8. Exim monitor: an attempt to get the action menu when not actually pointing
3052 at a message produces an empty menu entitled "No message selected". This
3053 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3054 no entries in it ("Shell widget menu has zero width and/or height"). So I
3055 have added a single, blank menu entry in this case.
3056
3057 9. Added ${quote_local_part.
3058
305910. MIME decoding is now applied to the contents of Subject: header lines when
3060 they are logged.
3061
306211. Now that a reference to $sender_host_address automatically causes a reverse
3063 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3064 host lookup before query-style lookups in lists that might use this
3065 variable. This has therefore been abolished, and the "net-" prefix is no
3066 longer necessary for query-style lookups.
3067
306812. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3069 have been a typo for LFLAGS, so it has been changed.
3070
307113. The install script calls Exim with "-C /dev/null" in order to find the
3072 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3073 to be output. Howeve, since Exim outputs its version number before the
3074 error, it didn't break the script. It just looked ugly. I fixed this by
3075 always allowing "-C /dev/null" if the caller is root.
3076
307714. Ignore overlarge ACL variable number when reading spool file - insurance
3078 against a later release with more variables having written the file.
3079
308015. The standard form for an IPv6 address literal was being rejected by EHLO.
3081 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3082 this, as well as the form without the "IPv6" on the front.
3083
308416. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3085 OS/Makefile-Darwin file.
3086
308717. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3088 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3089
309018. After change 4.21/52, "%ld" was used to format the contents of the $inode
3091 variable. However, some OS use ints for inodes. I've added cast to long int
3092 to get rid of the compiler warning.
3093
309419. I had forgotten to lock out "/../" in configuration file names when
3095 ALT_CONFIG_PREFIX was set.
3096
309720. Routers used for verification do not need to specify transports. However,
3098 if such a router generated a host list, and callout was configured, Exim
3099 crashed, because it could not find a port number from the (non-existent)
3100 transport. It now assumes port 25 in this circumstance.
3101
310221. Added the -t option to exigrep.
3103
310422. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3105 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3106 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3107 them is set, LOOKUP_LSEARCH is forced.
3108
310923. "exim -bV" now outputs a list of lookups that are included in the binary.
3110
311124. Added sender and host information to the "rejected by local_scan()" log
3112 line; previously there was no indication of these.
3113
311425. Added .include_if_exists.
3115
311626. Change 3.952/11 added an explicit directory sync on top of a file sync for
3117 Linux. It turns out that not all file systems support this. Apparently some
3118 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3119 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3120 not supported on the file descriptor, is now ignored when Exim is trying to
3121 sync a directory. This applies only to Linux.
3122
312327. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3124
312528. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3126 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3127 this.
3128
312929. Added support for maildirsize files from supplied patch (modified a bit).
3130
313130. The use of :fail: followed by an empty string could lead Exim to respond to
3132 sender verification failures with (e.g.):
3133
3134 550 Verification failed for <xxx>
3135 550 Sender verify failed
3136
3137 where the first response line was missing the '-' that indicates it is not
3138 the final line of the response.
3139
314031. The loop for finding the name of the user that called Exim had a hardwired
3141 limit of 10; it now uses the value of finduser_retries, which is used for
3142 all other user lookups.
3143
314432. Added $received_count variable, available in data and not_smtp ACLs, and at
3145 delivery time.
3146
314733. Exim was neglecting to zero errno before one call of strtol() when
3148 expanding a string and expecting an integer value. On some systems this
3149 resulted in spurious "integer overflow" errors. Also, it was casting the
3150 result into an int without checking.
3151
315234. Testing for a connection timeout using "timeout_connect" in the retry rules
3153 did not work. The code looks as if it has *never* worked, though it appears
3154 to have been documented since at least releast 1.62. I have made it work.
3155
315635. The "timeout_DNS" error in retry rules, also documented since at least
3157 1.62, also never worked. As it isn't clear exactly what this means, and
3158 clearly it isn't a major issue, I have abolished the feature by treating it
3159 as "timeout", and writing a warning to the main and panic logs.
3160
316136. The display of retry rules for -brt wasn't always showing the error code
3162 correctly.
3163
316437. Added new error conditions to retry rules: timeout_A, timeout_MX,
3165 timeout_connect_A, timeout_connect_MX.
3166
316738. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3168 to the empty sender.
3169
317039. The daemon was not analysing the content of -oX till after it had closed
3171 stderr and disconnected from the controlling terminal. This meant that any
3172 syntax errors were only noted on the panic log, and the return code from
3173 the command was 0. By re-arranging the code a little, I've made the
3174 decoding happen first, so such errors now appear on stderr, and the return
3175 code is 1. However, the actual setting up of the sockets still happens in
3176 the disconnected process, so errors there are still only recorded on the
3177 panic log.
3178
317940. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3180 connections (as happens on some IP stacks) was logged at start up time as
3181 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3182 it from "IPv6 and IPv4", which means that two separate sockets are being
3183 used.
3184
318541. The debug output for gethostbyname2() or getipnodebyname() failures now
3186 says whether AF_INET or AF_INET6 was passed as an argument.
3187
318842. Exiwhat output was messed up when time zones were included in log
3189 timestamps.
3190
319143. Exiwhat now gives more information about the daemon's listening ports,
3192 and whether -tls-on-connect was used.
3193
319444. The "port" option of the smtp transport is now expanded.
3195
319645. A "message" modifier in a "warn" statement in a non-message ACL was being
3197 silently ignored. Now an error message is written to the main and panic
3198 logs.
3199
320046. There's a new ACL modifier called "logwrite" which writes to a log file
3201 as soon as it is encountered.
3202
320347. Added $local_user_uid and $local_user_gid at routing time.
3204
320548. Exim crashed when trying to verify a sender address that was being
3206 rewritten to "<>".
3207
320849. Exim was recognizing only a space character after ".include". It now also
3209 recognizes a tab character.
3210
321150. Fixed several bugs in the Perl script that creates the exim.8 man page by
3212 extracting the relevant information from the specification. The man page no
3213 longer contains scrambled data for the -d option, and I've added a section
3214 at the front about calling Exim under different names.
3215
321651. Added "extra_headers" argument to the "mail" command in filter files.
3217
321852. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3219 crash.
3220
322153. Installed eximstats 1.29.
3222
322354. Added transport_filter_timeout as a generic transport option.
3224
322555. Exim no longer adds an empty Bcc: header to messages that have no To: or
3226 Cc: header lines. This was required by RFC 822, but it not required by RFC
3227 2822.
3228
322956. Exim used to add From:, Date:, and Message-Id: header lines to any
3230 incoming messages that did not have them. Now it does so only if the
3231 message originates locally, that is, if there is no associated remote host
3232 address. When Resent- header lines are present, this applies to the Resent-
3233 lines rather than the non-Resent- lines.
3234
323557. Drop incoming SMTP connection after too many syntax or protocol errors. The
3236 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3237
323858. Messages for configuration errors now include the name of the main
3239 configuration file - useful now that there may be more than one file in a
3240 list (.included file names were always shown).
3241
324259. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3243 for those rare installations that do not start the daemon as root or run it
3244 setuid root. I've cut out the call to initgroups() if the daemon is not
3245 root at that time.
3246
324760. The Exim user and group can now be bound into the binary as text strings
3248 that are looked up at the start of Exim's processing.
3249
325061. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3251
325262. Added $mailstore_basename variable.
3253
325463. Installed patch to sieve.c from Michael Haardt.
3255
325664. When Exim failed to open the panic log after failing to open the main log,
3257 the original message it was trying to log was written to stderr and debug
3258 output, but if they were not available (the usual case in production), it
3259 was lost. Now it is written to syslog before the two lines that record the
3260 failures to open the logs.
3261
326265. Users' Exim filters run in subprocesses under the user's uid. It is
3263 possible for a "deliver" command or an alias in a "personal" command to
3264 provoke an address rewrite. If logging of address rewriting is configured,
3265 this fails because the process is not running as root or exim. There may be
3266 a better way of dealing with this, but for the moment (because 4.30 needs
3267 to be released), I have disabled address rewrite logging when running a
3268 filter in a non-root, non-exim process.
3269
3270
3271Exim version 4.24
3272-----------------
3273
3274 1. The buildconfig auxiliary program wasn't quoting the value set for
3275 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3276 not defined. This bug was masked in 4.22 by the effect that was fixed in
3277 change 4.23/1.
3278
3279 2. Some messages that were rejected after a message id was allocated were
3280 shown as "incomplete" by exigrep. It no longer does this for messages that
3281 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3282
3283 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3284 have allow_domain_literals set, the ID did not get logged in the <= line.
3285 Domain literals are now always recognized in Message-ID: header lines.
3286
3287 4. The first argument for a ${extract expansion item is the key name or field
3288 number. Leading and trailing spaces in this item were not being ignored,
3289 causing some misleading effects.
3290
3291 5. When deliver_drop_privilege was set, single queue runner processes started
3292 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3293 same command in the process it spins off) were not dropping privilege.
3294
3295 6. When the daemon running as "exim" started a queue runner, it always
3296 re-executed Exim in the spun-off process. This is a waste of effort when
3297 deliver_drop_privilege is set. The new process now just calls the
3298 queue-runner function directly.
3299
3300
3301Exim version 4.23
3302-----------------
3303
3304 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3305 HEADERS_CHARSET.
3306
3307 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3308 ignored. Though the use of -oP was forcing the writing of a pid file, it
3309 was always written to the default place.
3310
3311 3. If the message "no IP address found for host xxxx" is generated during
3312 incoming verification, it is now followed by identification of the incoming
3313 connection (so you can more easily find what provoked it).
3314
3315 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3316
3317 5. Added some features to "harden" Exim a bit more against certain attacks:
3318
3319 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3320 be put in Local/Makefile. This is like the never_users runtime option,
3321 but it cannot be overridden. The default setting is "root".
3322
3323 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3324 prefix string with which any file named in a -C command line option
3325 must start.
3326
3327 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3328 is retained for -C and -D only if the caller of Exim is root. Without
3329 it, the exim user may also use -C and -D and retain privilege.
3330
3331 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3332 command line option is disabled.
3333
3334 6. Macro names set by the -D option must start with an upper case letter, just
3335 like macro names defined in the configuration file.
3336
3337 7. Added "dereference=" facility to LDAP.
3338
3339 8. Two instances of the typo "uknown" in the source files are fixed.
3340
3341 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3342 the Configure-Makefile script screwed up while processing it.
3343
334410. Incorporated PCRE 4.4.
3345
334611. The SMTP synchronization check was not operating right at the start of an
3347 SMTP session. For example, it could not catch a HELO sent before the client
3348 waited for the greeting. There is now a check for outstanding input at the
3349 point when the greeting is written. Because of the duplex, asynchronous
3350 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3351 way, but not yet received, when the check is performed.
3352
335312. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3354 on TCP/IP sockets, because this apparently causes some broken clients to
3355 timeout.
3356
335713. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3358 unchanged) from the Cygwin maintainer.
3359
336014. The code for -bV that shows what is in the binary showed "mbx" when maildir
3361 was supported instead of testing for mbx. Effectively a typo.
3362
336315. The spa authenticator server code was not checking that the input it
3364 received was valid base64.
3365
336616. The debug output line for the "set" modifier in ACLs was not showing the
3367 name of the variable that was being set.
3368
336917. Code tidy: the variable type "vtype_string" was never used. Removed it.
3370
337118. Previously, a reference to $sender_host_name did not cause a DNS reverse
3372 lookup on its own. Something else was needed to trigger the lookup. For
3373 example, a match in host_lookup or the need for a host name in a host list.
3374 Now, if $sender_host_name is referenced and the host name has not yet been
3375 looked up, a lookup is performed. If the lookup fails, the variable remains
3376 empty, and $host_lookup_failed is set to "1".
3377
337819. Added "eqi" as a case-independent comparison operator.
3379
338020. The saslauthd authentication condition could segfault if neither service
3381 nor realm was specified.
3382
338321. If an overflowing value such as "2048M" was set for message_size_limit, the
3384 error message that was logged was misleading, and incoming SMTP
3385 connections were dropped. The message is now more accurate, and temporary
3386 errors are given to SMTP connections.
3387
338822. In some error situations (such as 21 above) Exim rejects all SMTP commands
3389 (except RSET) with a 421 error, until QUIT is received. However, it was
3390 failing to send a response to QUIT.
3391
339223. The HELO ACL was being run before the code for helo_try_verify_hosts,
3393 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3394 ACL is now run after the helo_try_verify_hosts code.
3395
339624. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3397 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3398 case-independent, so other case variants are also recognized). Apparently
3399 some systems use these upper case variants.
3400
340125. If more than two messages were waiting for the same host, and a transport
3402 filter was specified for the transport, Exim sent two messages over the
3403 same TCP/IP connection, and then failed with "socket operation on non-
3404 socket" when it tried to send the third.
3405
340626. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3407
340827. The extern definition of crypt16() in expand.c was not being excluded when
3409 the OS had its own crypt16() function.
3410
341128. Added bounce_return_body as a new option, and bounce_return_size_limit
3412 as a preferred synonym for return_size_limit, both as an option and as an
3413 expansion variable.
3414
341529. Added LIBS=-liconv to OS/Makefile-OSF1.
3416
341730. Changed the default configuration ACL to relax the local part checking rule
3418 for addresses that are not in any local domains. For these addresses,
3419 slashes and pipe symbols are allowed within local parts, but the sequence
3420 /../ is explicitly forbidden.
3421
342231. SPA server authentication was not clearing the challenge buffer before
3423 using it.
3424
342532. log_message in a "warn" ACL statement was writing to the reject log as
3426 well as to the main log, which contradicts the documentation and doesn't
3427 seem right (because no rejection is happening). So I have stopped it.
3428
342933. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3430 However, I am unable to do any testing of this.
3431
343234. Fixed an infelicity in the appendfile transport. When checking directories
3433 for a mailbox, to see if any needed to be created, it was accidentally
3434 using path names with one or more superfluous leading slashes; tracing
3435 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3436
343735. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3438 contents are added to the log line that is written for every discarded
3439 recipient. (Previously a log_message setting was ignored.)
3440
344136. The ${quote: operator now quotes the string if it is empty.
3442
344337. The install script runs exim in order to find its version number. If for
3444 some reason other than non-existence or emptiness, which it checks, it
3445 could not run './exim', it was installing it with an empty version number,
3446 i.e. as "exim-". This error state is now caught, and the installation is
3447 aborted.
3448
344938. An argument was missing from the function that creates an error message
3450 when Exim fails to connect to the socket for saslauthd authentication.
3451 This could cause Exim to crash, or give a corrupted message.
3452
345339. Added isip, isip4, and isip6 to ${if conditions.
3454
345540. The ACL variables $acl_xx are now saved with the message, and can be
3456 accessed later in routers, transports, and filters.
3457
345841. The new lookup type nwildlsearch is like wildlsearch, except that the key
3459 strings in the file are not string-expanded.
3460
346142. If a MAIL command specified a SIZE value that was too large to fit into an
3462 int variable, the check against message_size_limit failed. Such values are
3463 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3464 day this will have to be increased, but I don't think I want to be around
3465 when emails are that large.
3466
3467
3468
3469Exim version 4.22
3470-----------------
3471
3472 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3473 iconv() is not standard in FreeBSD.
3474
3475 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3476 IPv6 enabled. The observed symptom was a segmentation fault on return from
3477 the function os_common_find_running_interfaces() in src/os.c.
3478
3479 3. In the check_special_case() function in daemon.c I had used "errno" as an
3480 argument name, which causes warnings on some systems. This was basically a
3481 typo, since it was named "eno" in the comments!
3482
3483 4. The code that waits for the clock to tick (at a resolution of some fraction
3484 of a second) so as to ensure message-id uniqueness was always waiting for
3485 at least one whole tick, when it could have waited for less. [This is
3486 almost certainly not relevant at current processor speeds, where it is
3487 unlikely to ever wait at all. But we try to future-proof.]
3488
3489 5. The function that sleeps for a time interval that includes fractions of a
3490 second contained a race. It did not block SIGALRM between setting the
3491 timer, and suspending (a couple of lines later). If the interval was short
3492 and the sigsuspend() was delayed until after it had expired, the suspension
3493 never ended. On busy systems this could lead to processes getting stuck for
3494 ever.
3495
3496 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3497 process, before it forks any delivery processes. The open lookup caching
3498 mechanism meant that the open file or database connection was passed into
3499 the delivery process. The problem was that delivery processes always tidy
3500 up cached lookup data. This could cause a problem for the next delivery
3501 process started by the queue runner, because the external queue runner
3502 process does not know about the closure. So the next delivery process
3503 still has data in the lookup cache. In the case of a file lookup, there was
3504 no problem because closing a file descriptor in a subprocess doesn't affect
3505 the parent. However, if the lookup was caching a connection to a database,
3506 the connection was closed, and the second delivery process was likely to
3507 see errors such as "PGSQL: query failed: server closed the connection
3508 unexpectedly". The problem has been fixed by closing all cached lookups
3509 in a queue runner before running a delivery process.
3510
3511 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3512 seem to have the "const" qualifier which it has on other OS. I've
3513 parameterised it.
3514
3515 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3516 *of the same type* (client or server) with the same public name that an
3517 error should be diagnosed.
3518
3519 9. When Exim looked up a host name for an IP address, but failed to find the
3520 original IP address when looking up the host name (a safety check), it
3521 output the message "<ip address> does not match any IP for NULL", which was
3522 confusing, to say the least. The bug was that the host name should have
3523 appeared instead of "NULL".
3524
352510. Since release 3.03, if Exim is called by a uid other than root or the Exim
3526 user that is built into the binary, and the -C or -D options is used, root
3527 privilege is dropped before the configuration file is read. In addition,
3528 logging is switched to stderr instead of the normal log files. If the
3529 configuration then re-defines the Exim user, the unprivileged environment
3530 is probably not what is expected, so Exim logs a panic warning message (but
3531 proceeds).
3532
3533 However, if deliver_drop_privilege is set, the unprivileged state may well
3534 be exactly what is intended, so the warning has been cut out in that case,
3535 and Exim is allowed to try to write to its normal log files.
3536
3537
3538Exim version 4.21
3539-----------------
3540
3541 1. smtp_return_error_details was not giving details for temporary sender
3542 or receiver verification errors.
3543
3544 2. Diagnose a configuration error if two authenticators have the same public
3545 name.
3546
3547 3. Exim used not to create the message log file for a message until the first
3548 delivery attempt. This could be confusing when incoming messages were held
3549 for policy or load reasons. The message log file is now created at the time
3550 the message is received, and an initial "Received" line is written to it.
3551
3552 4. The automatically generated man page for command line options had a minor
3553 bug that caused no ill effects; however, a more serious problem was that
3554 the procedure for building the man page automatically didn't always
3555 operate. Consequently, release 4.20 contains an out-of-date version. This
3556 shouldn't happen again.
3557
3558 5. When building Exim with embedded Perl support, the script that builds the
3559 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3560 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3561
3562 6. The freeze_tell option was not being used for messages that were frozen on
3563 arrival, either by an ACL or by local_scan().
3564
3565 7. Added the smtp_incomplete_transaction log selector.
3566
3567 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3568 was accepting AUTH without a new EHLO.
3569
3570 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3571 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3572 RFC.
3573
357410. Logging of TCP/IP connections (when configured) now happens in the main
3575 daemon process instead of the child process, so that the TCP/IP connection
3576 count is more accurate (but it can never be perfect).
3577
357811. The use of "drop" in a nested ACL was not being handled correctly in the
3579 outer ACL. Now, if condition failure induced by the nested "drop" causes
3580 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3581 or "require"), the connection is dropped.
3582
358312. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3584 that yield "discard" can now be used with an "accept" or a "discard" verb,
3585 but an error is generated for any others (because I can't see a useful way
3586 to define what should happen).
3587
358813. When an ACL is read dynamically from a file (or anywhere else), the lines
3589 are now processed in the same way as lines in the Exim configuration file.
3590 In particular, continuation lines are supported.
3591
359214. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3593
359415. Added -ti meaning -t -i.
3595
359616. Check for letters, digits, hyphens, and dots in the names of dnslist
3597 domains, and warn by logging if others are found.
3598
359917. At least on BSD, alignment is not guarenteed for the array of ifreq's
3600 returned from GIFCONF when Exim is trying to find the list of interfaces on
3601 a host. The code in os.c has been modified to copy each ifreq to an aligned
3602 structure in all cases.
3603
3604 Also, in some cases, the returned ifreq's were being copied to a 'struct
3605 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3606 means the last couple of bytes of an IPv6 address could be chopped if the
3607 ifreq contained only a normal sockaddr (14 bytes storage).
3608
360918. Named domain lists were not supported in the hosts_treat_as_local option.
3610 An entry such as +xxxx was not recognized, and was treated as a literal
3611 domain name.
3612
361319. Ensure that header lines added by a DATA ACL are included in the reject log
3614 if the ACL subsequently rejects the message.
3615
361620. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3617 MD5 (which is deprecated).
3618
361921. When testing a filter file using -bf, Exim was writing a message when it
3620 took the sender from a "From " line in the message, but it was not doing so
3621 when it took $return_path from a Return-Path: header line. It now does.
3622
362322. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3624 with a valid header line field name (a series of printing characters
3625 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3626
362723. Changed "disc" in the source to "disk" to conform to the documentation and
3628 the book and for uniformity.
3629
363024. Ignore Sendmail's -Ooption=value command line item.
3631
363225. When execve() failed while trying to run a command in a pipe transport,
3633 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3634 could be confused with a return value of 69 from the command itself. This
3635 has been changed to 127, the value the shell returns if it is asked to run
3636 a non-existent command. The wording for the related log line suggests a
3637 non-existent command as the problem.
3638
363926. If received_header_text expands to an empty string, do not add a Received:
3640 header line to the message. (Well, it adds a token one on the spool, but
3641 marks it "old" so that it doesn't get used or transmitted.)
3642
364327. Installed eximstats 1.28 (addition of -nt option).
3644
364528. There was no check for failure on the call to getsockname() in the daemon
3646 code. This can fail if there is a shortage of resources on the system, with
3647 ENOMEM, for example. A temporary error is now given on failure.
3648
364929. Contrary to the C standard, it seems that in some environments, the
3650 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3651 program. Exim now does this explicitly; it affects the formatting of
3652 timestamps using strftime().
3653
365430. If exiqsumm was given junk data, it threw up some uninitialized variable
3655 complaints. I've now initialized all the variables, to avoid this.
3656
365732. Header lines added by a system filter were not being "seen" during
3658 transport-time rewrites.
3659
366033. The info_callback() function passed to OpenSSL is set up with type void
3661 (*)(SSL *, int, int), as described somewhere. However, when calling the
3662 function (actually a macro) that sets it up, the type void(*)() is