Tidy queue run abandon log message.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
3eef829e 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.301 2006/02/14 15:24:10 ph10 Exp $
495ae4b0
PH
2
3Change log file for Exim from version 4.21
4-------------------------------------------
5
214e2000
PH
6Exim version 4.61
7-----------------
8
9PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
15 addresses as local.
16
f9daeae0
PH
17PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
19
7e66e54d
PH
20PH/03 Added disable_ipv6.
21
c8ea1597
PH
22PH/04 Changed $reply_address to use the raw form of the headers instead of the
23 decoded form, because it is most often used to construct To: headers
24 lines in autoreplies, and the decoded form may well be syntactically
25 invalid. However, $reply_address has leading white space removed, and all
26 newlines turned into spaces so that the autoreply transport does not
27 grumble.
28
911f6fde
PH
29PH/05 If group was specified without a user on a router, and no group or user
30 was specified on a transport, the group from the router was ignored.
31
47ca6d6c
PH
32PH/06 Increased the number of ACL variables to 20 of each type, and arranged
33 for visible compile-time settings that can be used to change these
34 numbers, for those that want even more. Backwards compatibility with old
35 spool files has been maintained. However, going back to a previous Exim
36 release will lost any variables that are in spool files.
37
ed0e9820
PH
38PH/07 Two small changes when running in the test harness: increase delay when
39 passing a TCP/IP connection to a new process, in case the original
40 process has to generate a bounce, and remove special handling of
41 127.0.0.2 (sic), which is no longer necessary.
42
eff37e47
PH
43PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
44 be the same on different OS.
45
1921d2ea
PH
46PH/09 Moved a debug statement in filter processing to avoid a race problem when
47 testing.
48
b3f69ca8
JJ
49JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
50 whether --show-vars was specified or not
51
52JJ/02 exipick: Added support for new ACL variable spool format introduced
53 in 4.61-PH/06
54
424a1c63
PH
55PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
56 syntactically invalid From: or Reply-to: line, and a filter used this to
57 generate an autoreply, and therefore failed to obtain an address for the
58 autoreply, Exim could try to deliver to a non-existent relative file
59 name, causing unrelated and misleading errors. What now happens is that
60 it logs this as a hard delivery error, but does not attempt to create a
61 bounce message.
62
7a100415
PH
63PH/11 The exinext utility has a -C option for testing purposes, but although
64 the given file was scanned by exinext itself; it wasn't being passed on
65 when Exim was called.
66
19b9dc85
PH
67PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
68 an end-of-file indication when reading a command response.
69
309bd837
PH
70PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
71 compiled. In many other places in Exim, IPv6 addresses are always
72 recognized, so I have changed this. It also means that IPv4 domain
73 literals of the form [IPV4:n.n.n.n] are now always recognized.
74
59e82a2a
PH
75PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
76 used if the router is not running as root, for example, when verifying at
77 ACL time, or when using -bh. The debugging output from this situation was
78 non-existent - all you got was a failure to exec. I have made two
79 changes:
80
81 (a) Failures to set uid/gid, the current directory, or a process leader
82 in a subprocess such as that created by queryprogram now generate
83 suitable debugging ouput when -d is set.
84
85 (b) The queryprogram router detects when it is not running as root,
86 outputs suitable debugging information if -d is set, and then runs
87 the subprocess without attempting to change uid/gid.
88
9edc04ce
PH
89PH/15 Minor change to Makefile for building test_host (undocumented testing
90 feature).
91
1349e1e5
PH
92PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
93 additional section of a DNS packet that returns MX or SRV records.
94 Instead, it always explicitly searches for A/AAAA records. This avoids
95 major problems that occur when a DNS server includes only records of one
96 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
97 fixed another bug: if SRV records were looked up and the corresponding
98 address records were *not* found in the additional section, the port
99 values from the SRV records were lost.
100
ea49d0e1
PH
101PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
102 using the correct key (the original address) when searching the retry
103 rules in order to find which one to use for generating the retry hint.
104
064a94c9
PH
105PH/18 If quota_warn_message contains a From: header, Exim now refrains from
106 adding the default one. Similarly, if it contains a Reply-To: header, the
107 errors_reply_to option, if set, is not used.
108
727071f8
PH
109PH/19 When calculating a retry time, Exim used to measure the "time since
110 failure" by looking at the "first failed" field in the retry record. Now
111 it does not use this if it is later than than the arrival time of the
112 message. Instead it uses the arrival time. This makes for better
113 behaviour in cases where some deliveries succeed, thus re-setting the
114 "first failed" field. An example is a quota failure for a huge message
115 when small messages continue to be delivered. Without this change, the
116 "time since failure" will always be short, possible causing more frequent
117 delivery attempts for the huge message than are intended.
118
f78eb7c6
PH
119PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
120 $1, $2, $3) because the numerical variables can be reset during some
121 expansion items (e.g. "match"), thereby losing the authentication data.
122
21c28500
PH
123PH/21 Make -bV show the size of off_t variables so that the test suite can
124 decide whether to run tests for quotas > 2G.
125
126PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
127 mailbox_size, and mailbox_filecount in the appendfile transport. If a
128 filecount value is greater than 2G or if a quota value is greater than 2G
129 on a system where the size of off_t is not greater than 4, a panic error
130 is given.
131
1688f43b
PH
132PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
133 never match. The debug and -bh output now contains an explicit error
134 message indicating a malformed IPv4 address or mask.
135
136PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
137 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
138 PH/23 above applies.
139
9675b384
PH
140PH/25 Do not write to syslog when running in the test harness. The only
141 occasion when this arises is a failure to open the main or panic logs
142 (for which there is an explicit test).
143
6a3f1455
PH
144PH/26 Added the /no_tell option to "control=freeze".
145
dac79d3e
PH
146PH/27 If a host name lookup failed very early in a connection, for example, if
147 the IP address matched host_lookup and the reverse lookup yielded a name
148 that did not have a forward lookup, an error message of the form "no IP
149 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
150 could be logged. Now it outputs the IP address instead of "NULL".
1349e1e5 151
5977a0b3
PH
152PH/28 An enabling patch from MH: add new function child_open_exim2() which
153 allows the sender and the authenticated sender to be set when
154 submitting a message from within Exim. Since child_open_exim() is
155 documented for local_scan(), the new function should be too.
156
c91535f3
PH
157PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
158 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
159 results in an empty string is now treated as unset.
160
0d46a8c8
PH
161PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
162
278c6e6c
PH
163PH/31 Added sender_verify_fail logging option.
164
2cbb4081
PH
165PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
166 needed by RFC 822 but not by RFC 2822 was commented out. I have now
167 tidied the source and removed it altogether.
168
3eef829e
PH
169PH/33 When a queue run was abandoned because the load average was too high, a
170 log line was always written; now it is written only if the queue_run log
171 selector is set. In addition, the log line for abandonment now contains
172 information about the queue run such as the pid. This is always present
173 in "start" and "stop" lines but was omitted from the "abandon" line.
174
214e2000 175
5de37277
PH
176Exim version 4.60
177-----------------
178
cc38ddbf
PH
179PH/01 Two changes to the default runtime configuration:
180
181 (1) Move the checks for relay_from_hosts and authenticated clients from
182 after to before the (commented out) DNS black list checks.
183
184 (2) Add control=submission to the relay_from_hosts and authenticated
185 clients checks, on the grounds that messages accepted by these
186 statements are most likely to be submissions.
5de37277 187
72fdd6ae
PH
188PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
189
190 (1) Generate an error if the third argument for the ${prvs expansion is
191 not a single digit.
192
193 (2) Treat a missing third argument of ${prvscheck as if it were an empty
194 string.
195
196 (3) Reset the variables that are obtained from the first argument of
197 ${prvscheck and used in the second argument before leaving the code,
198 because their memory is reclaimed, so using them afterwards may do
199 silly things.
200
201 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
202 one (it's much easier than Tom thought :-).
203
204 (5) Because of (4), we can now allow for the use of $prvscheck_result
205 inside the third argument.
cb9328de 206
cb741023
PH
207PH/03 For some reason, the default setting of PATH when running a command from
208 a pipe transport was just "/usr/bin". I have changed it to
209 "/bin:/usr/bin".
210
f174f16e
PH
211PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
212 anything to be listed in the output from -bV.
b2f5a032 213
c25242d7
PH
214PH/05 When a filter generated an autoreply, the entire To: header line was
215 quoted in the delivery log line, like this:
216
217 => >A.N.Other <ano@some.domain> <original@ddress> ...
218
219 This has been changed so that it extracts the operative address. There
220 may be more than one such address. If so, they are comma-separated, like
221 this:
222
223 => >ano@some.domain,ona@other.domain <original@ddress> ...
224
82c19f95
PH
225PH/06 When a client host used a correct literal IP address in a HELO or EHLO
226 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
227 not being looked up in the rDNS to get a host name, Exim was showing the
228 IP address twice in Received: lines, even though the IP addresses were
229 identical. For example:
230
231 Received: from [1.2.3.4] (helo=[1.2.3.4])
232
233 However, if the real host name was known, it was omitting the HELO data
234 if it matched the actual IP address. This has been tidied up so that it
235 doesn't show the same IP address twice.
236
d7ffbc12
PH
237PH/07 When both +timestamp and +memory debugging was on, the value given by
238 $tod_xxx expansions could be wrong, because the tod_stamp() function was
239 called by the debug printing, thereby overwriting the timestamp buffer.
240 Debugging no longer uses the tod_stamp() function when +timestamp is set.
241
9f526266
PH
242PH/08 When the original message was included in an autoreply transport, it
243 always said "this is a copy of the message, including all the headers",
244 even if body_only or headers_only was set. It now gives an appropriate
245 message.
246
87fcc8b9
PH
247PH/09 Applied a patch from the Sieve maintainer which:
248
249 o fixes some comments
250 o adds the (disabled) notify extension core
251 o adds some debug output for the result of if/elsif tests
252 o points to the current vacation draft in the documentation
253 and documents the missing references header update
254
255 and most important:
256
257 o fixes a bug in processing the envelope test (when testing
258 multiple envelope elements, the last element determinted the
259 result)
260
456682f5
PH
261PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
262 Electronic Mail") by including:
263
264 Auto-submitted: auto-generated
265
266 in the messages that it generates (bounce messages and others, such as
267 warnings). In the case of bounce messages for non-SMTP mesages, there was
268 also a typo: it was using "Auto_submitted" (underscore instead of
269 hyphen). Since every message generated by Exim is necessarily in response
270 to another message, thes have all been changed to:
271
272 Auto-Submitted: auto-replied
273
274 in accordance with these statements in the RFC:
275
276 The auto-replied keyword:
277
278 - SHOULD be used on messages sent in direct response to another
279 message by an automatic process,
280
281 - MUST NOT be used on manually-generated messages,
282
283 - MAY be used on Delivery Status Notifications (DSNs) and Message
284 Disposition Notifications (MDNs),
285
286 - MUST NOT be used on messages generated by automatic or periodic
287 processes, except for messages which are automatic responses to
288 other messages.
289
3e46c1aa
PH
290PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
291 to the default Received: header definition.
456682f5 292
49826d12
PH
293PH/12 Added log selector acl_warn_skipped (default on).
294
eba0c039
PH
295PH/13 After a successful wildlsearch lookup, discard the values of numeric
296 variables because (a) they are in the wrong storage pool and (b) even if
297 they were copied, it wouldn't work properly because of the caching.
298
a0d6ba8a
PH
299PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
300 checking when decoding. Apparently there are clients that generate
301 overlong encoded strings. Why am I not surprised?
302
f0917727
PH
303PH/15 If the first argument of "${if match_address" was not empty, but did not
304 contain an "@" character, Exim crashed. Now it writes a panic log message
305 and treats the condition as false.
306
096fee00
PH
307PH/16 In autoreply, treat an empty string for "once" the same as unset.
308
024bd3c2
PH
309PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
310 extension "envelope-auth". The code is finished and in agreement with
311 other implementations, but there is no documentation so far and in fact,
312 nobody wrote the draft yet. This extension is currently #undef'ed, thus
313 not changing the active code.
314
315 Print executed "if" and "elsif" statements when debugging is used. This
316 helps a great deal to understand what a filter does.
317
318 Document more things not specified clearly in RFC3028. I had all this
319 sorted out, when out of a sudden new issues came to my mind. Oops."
320
df199fec
PH
321PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
322 (Bugzilla #53).
323
d27f1df3
PH
324PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
325 canonical form (as documented). However, after a host name lookup from
326 the IP address, check_host() was doing a simple string comparison with
327 addresses acquired from the DNS when checking that the found name did
328 have the original IP as one of its addresses. Since any found IPv6
329 addresses are likely to be in abbreviated form, the comparison could
330 fail. Luckily, there already exists a function for doing the comparison
331 by converting both addresses to binary, so now that is used instead of
332 the text comparison.
333
96776534
PH
334PH/20 There was another similar case to PH/19, when a complete host name was
335 given in a host list; looking up its IP address could give an abbreviated
336 form, whereas the current host's name might or might not be abbreviated.
337 The same fix has been applied.
338
5de37277 339
9a799bc0
PH
340Exim version 4.54
341-----------------
342
343PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
344 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
345 It now does.
346
99a4b039
PH
347PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
348 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
349
350PH/03 Typo: missing ".o" in src/pcre/Makefile.
351
4b233853
PH
352PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
353 header line, restrict the check to what is listed in RFCs 2369 and 2929.
354 Also, for "Auto-Submitted", treat anything other than "no" as
355 non-personal, in accordance with RFC 3834. (Previously it treated
356 anything starting "auto-" as non-personal.)
357
8857ccfd
PH
358TF/01 The control=submission/name=... option had a problem with syntax
359 errors if the name included a slash character. The /name= option
360 now slurps the rest of the string, so it can include any characters
361 but it must come last in the list of options (after /sender_retain
362 or /domain=).
363
433a2980
PH
364PH/05 Some modifications to the interface to the fake nameserver for the new
365 testing suite.
366
3e46c1aa 367
9a799bc0 368
e3a311ba
TK
369Exim version 4.53
370-----------------
371
372TK/01 Added the "success_on_redirect" address verification option. See
373 NewStuff for rationale and an example.
374
13b685f9
PH
375PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
376
395ff96d
PH
377PH/02 Patch to exigrep to allow it to work on syslog lines.
378
5b68f6e4
PH
379PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
380 fread() to skip over the body file's header line, because in Cygwin the
381 header line is locked and is inaccessible.
382
1ab52c69
PH
383PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
384 co-exist for some time) to make it clear that it is the Exim ID that is
385 referenced, not the Message-ID: header line.
386
b07e6aa3
PH
387PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
388 string_format() function, because snprintf() does not exist on all
389 operating systems.
390
254e032f
PH
391PH/06 The use of forbid_filter_existstest now also locks out the use of the
392 ${stat: expansion item.
393
3af76a81
PH
394PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
395 protocol synchronization error", to keep the pedants happy.
396
2548ba04
PH
397PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
398 well as for IRIX systems, when gcc is being used. See the host.c source
399 file for comments.
400
b6c6011d
PH
401PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
402
cf39cf57
PH
403PH/10 Named domain lists were not working if used in a queue_smtp_domains
404 setting.
405
f1513293
PH
406PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
407 transport and to the smtp transport in LMTP mode.
408
727549a4
PH
409TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
410
af46795e
PH
411PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
412 run a filter in a subprocess. This could lead to confusion in subsequent
413 lookups in the parent process. There should also be a search_tidyup() at
414 the end of the subprocess.
415
d7b47fd0
PH
416PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
417 only if the host matched helo_try_verify_hosts, which caused the
418 verification to occur when the EHLO/HELO command was issued. The ACL just
419 tested the remembered result. Now, if a previous verification attempt has
420 not happened, "verify = helo" does it there and then.
421
ee744174
JJ
422JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
423
b582ab87
PH
424TK/03 Fix log output including CR from clamd.
425
41a13e0a
PH
426PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
427 exist provoked a memory error which could cause a segfault.
428
f625cc5a
PH
429PH/15 Installed PCRE 6.2
430
431PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
432
21f7af35
PH
433PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
434 of the problem. Specifically, suggested +O2 rather than +O1 for the
435 HP-UX compiler.
436
31480e42
PH
437PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
438
2d280592
PH
439PH/20 If a delivery was routed to a non-standard port by means of an SRV
440 record, the port was not correctly logged when the outgoing_port log
441 selector was set (it logged the transort's default port).
442
7cd1141b
PH
443PH/21 Added support for host-specific ports to manualroute, queryprogram,
444 fallback_hosts, and "hosts" in the smtp transport.
445
446PH/22 If the log selector "outgoing_port" is set, the port is now also given on
447 host errors such as "Connection refused".
448
750af86e
PH
449PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
450 authentication with radiusclient 0.4.9:
451
452 - Error returned from rc_read_config was caught wrongly
453 - Username/password not passed on to radius server due to wrong length.
454
455 The presumption is that some radiusclient API changes for 4.51/PH/17
456 were not taken care of correctly. The code is still untested by me (my
457 Linux distribution still has 0.3.2 of radiusclient), but it was
458 contributed by a Radius user.
459
460PH/24 When doing a callout, the value of $domain wasn't set correctly when
461 expanding the "port" option of the smtp transport.
462
4304270b
TK
463TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
464 while reading a MIME header. Thanks to Tom Hughes for a patch.
465
750af86e
PH
466PH/24 Include config.h inside local_scan.h so that configuration settings are
467 available.
468
64ffc24f
PH
469PH/25 Make $smtp_command_argument available after all SMTP commands. This means
470 that in an ACL for RCPT (for example), you can examine exactly what was
471 received.
472
5dd9625b
PH
473PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
474 commands, but it was not correctly comparing the address with the actual
475 client host address. Thus, it would show the EHLO address in Received:
476 header lines when this was not necessary.
477
5591031b
PH
478PH/27 Added the % operator to ${eval:}.
479
ba18e66a
PH
480PH/28 Exim tries to create and chdir to its spool directory when it starts;
481 it should be ignoring failures (because with -C, for example, it has lost
482 privilege). It wasn't ignoring creation failures other than "already
483 exists".
484
9cec981f
PH
485PH/29 Added "crypteq" to the list of supported features that Exim outputs when
486 -bV or -d is used.
487
aa2b5c79
PH
488PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
489 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
490 too many continuations, the temporary file was not being removed, and the
491 return code was incorrect.
aa2b5c79 492
48a53b7f
PH
493PH/31 Missing "BOOL" in function definition in filtertest.c.
494
1c59d63b
PH
495PH/32 Applied Sieve patches from the maintainer.
496
671012da
TK
497TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
498
1509d3a8
PH
499PH/33 Added "verify = not_blind".
500
501PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
502 Local/Makefile (with some defaults set). These are used in built scripts
503 such as exicyclog, but they have never been used in the exim_install
504 script (though there are many overriding facilities there). I have
505 arranged that the exim_install script now takes note of these two
506 settings.
507
508PH/35 Installed configuration files for Dragonfly.
509
2fe1a124
PH
510PH/36 When a locally submitted message by a trusted user did not contain a
511 From: header, and the sender address was obtained from -f or from an SMTP
512 MAIL command, and the trusted user did not use -F to supply a sender
513 name, $originator_name was incorrectly used when constructing a From:
514 header. Furthermore, $originator_name was used for submission mode
515 messages from external hosts without From: headers in a similar way,
516 which is clearly wrong.
517
8800895a
PH
518PH/37 Added control=suppress_local_fixups.
519
ccfdb010
PH
520PH/38 When log_selector = +received_sender was set, and the addition of the
521 sender made the log line's construction buffer exactly full, or one byte
522 less than full, an overflow happened when the terminating "\n" was
523 subsequently added.
524
1130bfb0
PH
525PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
526 when the result of a list match is failure because a DNS lookup failed.
527
ebcb507f
PH
528PH/40 RM_COMMAND is now used in the building process.
529
c35e155c
PH
530PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
531 the "build-* directories that it finds.
532
95d1f782
PH
533PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
534 address in a domain literal was a prefix of an interface address.
535
fd6de02e
PH
536PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
537 when verifying a sender address, unless rewrite_headers is false.
538
58de37c5
PH
539PH/44 Wrote a long comment about why errors_to addresses are verified as
540 recipients, not senders.
541
261cf466
TF
542TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
543 the ratelimit ACL was added.
544
3ee512ff
PH
545PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
546
e08c430f
PH
547PH/46 Added extra information about PostgreSQL errors to the error string.
548
bef5a11f
PH
549PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
550 avoiding the need to install special zones in a real server. This is
551 backwards compatible; if it can't find the fake resolver, it drops back.
552 Thus, both old and new test suites can be run.
553
7546de58
TF
554TF/02 Added util/ratelimit.pl
555
e5d5a95f
TF
556TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
557 clock is set back in time.
558
2e88a017
TF
559TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
560 Candler <B.Candler@pobox.com>.
561
a5f65aa4
TF
562TF/05 The fix for PH/43 was not completely correct; widen_domains is always
563 OK for addresses that are the result of redirections.
564
e7726cbf
PH
565PH/48 A number of further additions for the benefit of the new test suite,
566 including a fake gethostbyname() that interfaces to the fake DNS resolver
567 (see PH/47 above).
568
a7fdad5b
TF
569TF/06 The fix for widen_domains has also been applied to qualify_single and
570 search_parents which are the other dnslookup options that can cause
571 header rewrites.
572
6af56900
PH
573PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
574 type ("H").
575
0925ede6
PH
576PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
577
66afa403
TF
578TF/07 Exim produced the error message "an SRV record indicated no SMTP
579 service" if it encountered an MX record with an empty target hostname.
580 The message is now "an MX or SRV record indicated no SMTP service".
581
0154e85a
TF
582TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
583 if the DNS of the sending site is misconfigured. This is quite a
584 common situation. This change restores the behaviour of treating a
585 helo verification defer as a failure.
586
16f12c76
PH
587PH/51 If self=fail was set on a router, the bounce message did not include the
588 actual error message.
589
bbe902f0 590
e5a9dba6
PH
591Exim version 4.52
592-----------------
593
594TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
595
22c3b60b
PH
596PH/01 When a transport filter timed out in a pipe delivery, and the pipe
597 command itself ended in error, the underlying message about the transport
598 filter timeout was being overwritten with the pipe command error. Now the
599 underlying error message should be appended to the second error message.
600
06a9b4b5
PH
601TK/01 Fix poll() being unavailable on Mac OSX 10.2.
602
c1ac6996
PH
603PH/02 Reduce the amount of output that "make" produces by default. Full output
604 can still be requested.
605
9c7a242c
PH
606PH/03 The warning log line about a condition test deferring for a "warn" verb
607 was being output only once per connection, rather than after each
608 occurrence (because it was using the same function as for successful
609 "warn" verbs). This seems wrong, so I have changed it.
610
87ba3f5f
PH
611TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
612 it should not have, which might have caused a crash in the right
613 circumstances, but probably never did.
614
615PH/04 Installed a modified version of Tony Finch's patch to make submission
616 mode fix the return path as well as the Sender: header line, and to
617 add a /name= option so that you can make the user's friendly name appear
618 in the header line.
619
29aba418
TF
620TF/03 Added the control = fakedefer ACL modifier.
621
fe0dab11
TF
622TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
623 Mark Lowes for thorough testing.
870f6ba8 624
11d337a4
TK
625TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
626
627TK/03 Merged latest SRS patch from Miles Wilton.
628
415c8f3b
PH
629PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
630 with the definition in sysexits.h (which is #included earlier).
631 Fortunately, Exim does not actually use EX_OK. The code used to try to
632 preserve the sysexits.h value, by assumimg that macro definitions were
633 scanned for macro replacements. I have been disabused of this notion,
634 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 635
958541e9
PH
636PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
637 in the smtp transport. When a block could not be written in a single
638 write() function, the timeout was being re-applied to each part-write.
639 This seems wrong - if the receiver was accepting one byte at a time it
640 would take for ever. The timeout is now adjusted when this happens. It
641 doesn't have to be particularly precise.
642
c206415f
TK
643TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
644 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
645
2a4be8f9
PH
646PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
647 without a domain if the check to <postmaster@domain> fails.
648
1cba11c5
SC
649SC/01 Eximstats: added -xls and the ability to specify output files
650 (patch written by Frank Heydlauf).
651
652SC/02 Eximstats: use FileHandles for outputing results.
653
654SC/03 Eximstats: allow any combination of xls, txt, and html output.
655
656SC/04 Eximstats: fixed display of large numbers with -nvr option
657
658SC/05 Eximstats: fixed merging of reports with empty tables.
659
660SC/06 Eximstats: added the -include_original_destination flag
661
662SC/07 Eximstats: removed tabs and trailing whitespace.
663
1005d00e
TK
664TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
665
666TK/06 MBOX spool code: Add real "From " MBOX separator line
667 so the .eml file is really in mbox format (even though
668 most programs do not really care). Patch from Alex Miller.
669
670TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
671 The latter is generated from $received_to and is only set if the
672 message has one envelope recipient. SA can use these headers,
673 obviously out-of-the-box. Patch from Alex Miller.
674
9b4768fa
PH
675PH/08 The ${def test on a variable was returning false if the variable's
676 value was "0", contrary to what the specification has always said!
677 The result should be true unless the variable is empty.
678
679PH/09 The syntax error of a character other than { following "${if
680 def:variable_name" (after optional whitespace) was not being diagnosed.
681 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
682 accidental colon was present, for example, could give incorrect results.
683
0d7eb84a
PH
684PH/10 Tidied the code in a number of places where the st_size field of a stat()
685 result is used (not including appendfile, where other changes are about
686 to be made).
687
688PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
689 This involved changing a lot of size variables from int to off_t. It
690 should work with maildirs and everything.
691
40727bee
TK
692TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
693 spamd dying while we are connected to it.
694
554d2369
TF
695TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
696 <jgh@wizmail.org>
697
1f922db1
PH
698PH/12 Applied Alex Kiernan's patch for the API change for the error callback
699 function for BDB 4.3.
700
ef213c3b
PH
701PH/13 Changed auto_thaw such that it does not apply to bounce messages.
702
8ac170f3
PH
703PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
704 the sources for PCRE have been re-arranged and more files are now
705 involved.
706
b1c749bb
PH
707PH/15 The code I had for printing potentially long long variables in PH/11
708 above was not the best (it lost precision). The length of off_t variables
709 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
710 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
711 to be "long long int" or "long int". This is needed for the internal
712 formatting function string_vformat().
b1c749bb 713
4aac9b49
PH
714PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
715 the configuration file to be ":syslog", then the script "guesses" where
716 the logs files are, rather than using the compiled in default. In our
717 case the guess is not the same as the compiled default, so the script
718 suddenly stopped working when I started to use syslog. The patch checks
719 to see if log_file_path is "". If so, it attempts to read it from exim
720 with no configuration file to get the compiled in version, before it
721 falls back to the previous guessing code."
722
294520c8
TK
723TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
724 implementing BATV in an Exim configuration. See NewStuff for the gory
725 details.
726
5bd022fe
PH
727PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
728 Makefile that are specific to HP-UX.
729
90e9ce59
PH
730PH/18 If the "use_postmaster" option was set for a recipient callout together
731 with the "random" option, the postmaster address was used as the MAIL
732 FROM address for the random test, but not for the subsequent recipient
733 test. It is now used for both.
734
5ea81592
PH
735PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
736 patch removes a few documentation additions to RFC 3028, because the
737 latest draft now contains them. It adds the new en;ascii-case comparator
738 and a new error check for 8bit text in MIME parts. Comparator and
739 require names are now matched exactly. I enabled the subaddress
740 extension, but it is not well tested yet (read: it works for me)."
741
c6c2dc1d
PH
742PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
743 rework some of the code of TK/09 above to avoid the hardwired use of
744 "%lld" and "long long". Replaced the call to snprintf() with a call to
745 string_vformat().
746
fffffe4c
PH
747PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
748 records point to non-existent hosts", "retry timeout exceeded", and
749 "retry time not reached for any host after a long failure period".
ca02eafb 750
9a26b6b2
PH
751PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
752 experimental DomainKeys support:
753
754 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
755 (2) On an error such as an illegally used "control", the wrong name for
756 the control was given.
757
758 These problems did NOT occur unless DomainKeys support was compiled.
759
4aee0225
PH
760PH/23 Added daemon_startup_retries and daemon_startup_sleep.
761
32d668a5
PH
762PH/24 Added ${if match_ip condition.
763
8187c3f3
PH
764PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
765 databases so that it will be absolutely obvious if a crash occurs in the
766 DB library. This is a regular occurrence (often caused by mis-matched
767 db.h files).
768
ff790e47 769PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
770 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
771 picked up on a user's system that detects such things. There doesn't seem
772 to be a gcc warning option for this - only an attribute that has to be
773 put on the function's prototype. It seems that in Fedora Core 4 they have
774 set this on a number of new functions. No doubt there will be more in due
775 course.
ff790e47 776
5417f6d1
PH
777PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
778 specify a transport. However, if an address that was verified by such a
779 router was the subject of a callout, Exim crashed because it tried to
780 read the rcpt_include_affixes from the non-existent transport. Now it
781 just assumes that the setting of that option is false. This bug was
782 introduced by 4.51/PH/31.
783
59cf8544
PH
784PH/28 Changed -d+all to exclude +memory, because that information is very
785 rarely of interest, but it makes the output a lot bigger. People tend to
786 do -d+all out of habit.
787
e7ad8a65
PH
788PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
789 code in os-type was giving problems when libc.so lives in lib64, like on
790 x86_64 Fedora Core.
791
ade42478
PH
792PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
793 aren't the modern standard, and it seems that some systems' include files
794 don't always have them. Exim was already checking for some of the newer
795 ones like T_AAAA, and defining it itself. I've added checks for all the
796 record types that Exim uses.
797
182ad5cf
PH
798PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
799 not automatically generating a new one, as it is supposed to. This
800 prevented TLS from working. If the file did exist, but contained invalid
801 data, a new version was generated, as expected. It was only the case of a
802 non-existent file that was broken.
803
b0d9fc80
TK
804TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
805 with a change in libdomainkeys > 0.64.
806
807TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
808 from DNS. If the selector record carries the flag, it now has
809 precedence over the domain-wide flag.
810
811TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
812
47c7a64a
PH
813PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
814 the use of a port name that isn't defined in /etc/services) occurred, the
815 message was deferred as in a normal delivery, and thus remained on the
816 spool, instead of being failed because of the mua_wrapper setting. This
817 is now fixed, and I tidied up some of the mua_wrapper messages at the
818 same time.
819
a388bce4
SC
820SC/08 Eximstats: whilst parsing the mainlog(s), store information about
821 the messages in a hash of arrays rather than using individual hashes.
822 This is a bit cleaner and results in dramatic memory savings, albeit
823 at a slight CPU cost.
824
825SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
826 as requested by Marc Sherman.
827
828SC/10 Eximstats: added histograms for user specified patterns as requested
829 by Marc Sherman.
830
0793e4ed
SC
831SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
832
c58b88df
PH
833PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
834 fopen() in the content-scanning modules that did not already have it.
835
e7ad8a65 836
7982096b
PH
837Exim version 4.51
838-----------------
839
1a46a8c5
PH
840TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
841 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
842
2f079f46 843TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
844
845TK/03 Fix ACL "regex" condition not being available in MIME ACL.
846
847PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
848 to test Sieve filters that use "vacation".
849
850PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
851 that changes the way the GnuTLS parameters are stored in the cache file.
852 The new format can be generated externally. For backward compatibility,
853 if the data in the cache doesn't make sense, Exim assumes it has read an
854 old-format file, and it generates new data and writes a new file. This
855 means that you can't go back to an older release without removing the
856 file.
857
858PH/03 A redirect router that has both "unseen" and "one_time" set does not
859 work if there are any delivery delays because "one_time" forces the
860 parent to be marked "delivered", so its unseen clone is never tried
861 again. For this reason, Exim now forbids the simultaneous setting of
862 these two options.
863
864PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
865 redirected to themselves ("homonym" addresses). Read the long ChangeLog
866 entry if you want to know the details. The fix, however, neglected to
867 consider the case when local delivery batching is involved. The test for
868 "previously delivered" was not happening when checking to see if an
869 address could be batched with a previous (undelivered) one; under
870 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 871 address.
1a46a8c5
PH
872
873PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
874 in its include files, and this causes problems building Exim.
875
876PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
877 header_syntax) but Exim was just ignoring anything given after a slash.
878 In particular, this caused confusion with an attempt to use "verify =
879 reverse_host_lookup/defer_ok". An error is now given when options are
880 supplied for verify items that do not have them. (Maybe reverse_host_
881 lookup should have a defer_ok option, but that's a different point.)
882
883PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
884 defined by RFC 821) to 2048, because there were problems with some AUTH
885 commands, and RFC 1869 says the size should be increased for extended
886 SMTP commands that take arguments.
887
888PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
889 Finch).
890
891PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
892 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 893
49c2d5ea
PH
894PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
895 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
896 whatever). Otherwise the password may appear in the log. Following change
897 PH/42 below, there is no longer a chance of it appearing in a bounce
898 message.
49c2d5ea 899
bf759a8b
PH
900PH/11 Installed exipick version 20050225.0 from John Jetmore.
901
83364d30
PH
902PH/12 If the last host in a fallback_hosts list was multihomed, only the first
903 of its addresses was ever tried. (Bugzilla bug #2.)
904
7999bbd7
PH
905PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
906 the result incorrectly in the debug output. (It correctly added a newline
907 to what was transported.)
908
7dbf77c9
PH
909TF/01 Added $received_time.
910
74e0617f
PH
911PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
912 commented out examples of how to interface to a virus scanner and to
913 SpamAssassin. Also added commented examples of av_scanner and
914 spamd_address settings.
915
2f079f46
PH
916PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
917 and controls are allowed in which ACLs. There were a couple of minor
918 errors. Some of the entries in the conditions table (which is a table of
919 where they are NOT allowed) were getting very unwieldy; rewrote them as a
920 negation of where the condition IS allowed.
921
8c841523
PH
922PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
923
7766a4f0
PH
924PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
925 header file does not have a version number, so I've had to invent a new
926 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
927 API. The code is untested by me (my Linux distribution still has 0.3.2 of
928 radiusclient), but it was contributed by a Radius user.
929
8b417f2c
PH
930PH/18 Installed Lars Mainka's patch for the support of CRL collections in
931 files or directories, for OpenSSL.
932
901f42cb
PH
933PH/19 When an Exim process that is running as root has to create an Exim log
934 file, it does so in a subprocess that runs as exim:exim so as to get the
935 ownership right at creation (otherwise, other Exim processes might see
936 the file with the wrong ownership). There was no test for failure of this
937 fork() call, which would lead to the process getting stuck as it waited
938 for a non-existent subprocess. Forks do occasionally fail when resources
939 run out. I reviewed all the other calls to fork(); they all seem to check
940 for failure.
941
f9b9210e
PH
942PH/20 When checking for unexpected SMTP input at connect time (before writing
943 the banner), Exim was not dealing correctly with a non-positive return
944 from the read() function. If the client had disconnected by this time,
945 the result was a log entry for a synchronization error with an empty
946 string after "input=" when read() returned zero. If read() returned -1
947 (an event I could not check), uninitialized data bytes were printed.
948 There were reports of junk text (parts of files, etc) appearing after
949 "input=".
950
54cdb463
PH
951PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
952
cf00dad6
PH
953PH/22 Added support for macro redefinition, and (re)definition in between
954 driver and ACL definitions.
955
acb1b346
PH
956PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
957 forgetting to use the resulting value; it was using the unexpanded value.
958
c5ddb310
PH
959PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
960 hadn't been configured. The fix is from Juergen Kreileder, who
961 understands it better than I do:
962
963 "Here's what I see happening with three configured cyrus_sasl
964 authenticators configured (plain, login, cram-md5):
965
966 On startup auth_cyrus_sasl_init() gets called for each of these.
967 This means three calls to sasl_listmech() without a specified mech_list.
968 => SASL tests which mechs of all available mechs actually work
969 => three warnings about OTP not working
970 => the returned list contains: plain, login, cram-md5, digest-md5, ...
971
972 With the patch, sasl_listmech() also gets called three times. But now
973 SASL's mech_list option is set to the server_mech specified in the the
974 authenticator. Or in other words, the answer from sasl_listmech()
975 gets limited to just the mech you're testing for (which is different
976 for each call.)
977 => the return list contains just 'plain' or 'login', 'cram-md5' or
978 nothing depending on the value of ob->server_mech.
979
980 I've just tested the patch: Authentication still works fine,
981 unavailable mechs specified in the exim configuration are still
982 caught, and the auth.log warnings about OTP are gone."
983
31619da6
PH
984PH/25 When debugging is enabled, the contents of the command line are added
985 to the debugging output, even when log_selector=+arguments is not
986 specified.
987
bebaf0fc
PH
988PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
989 answer is "GNU", and only if the return is "GNU/something" is the answer
990 "Linux".
991
475fe28a
PH
992PH/27 $acl_verify_message is now set immediately after the failure of a
993 verification in an ACL, and so is available in subsequent modifiers. In
994 particular, the message can be preserved by coding like this:
995
996 warn !verify = sender
997 set acl_m0 = $acl_verify_message
998
999 Previously, $acl_verify_message was set only while expanding "message"
1000 and "log_message" when a very denied access.
1001
7e8bec7a
PH
1002PH/28 Modified OS/os.c-Linux with
1003
1004 -#ifndef OS_LOAD_AVERAGE
1005 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1006
1007 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1008 nomenclature these days.)
1009
e4a89c47
PH
1010PH/29 Installed patch from the Sieve maintainer that adds the options
1011 sieve_useraddress and sieve_subaddress to the redirect router.
1012
5ca2a9a1
PH
1013PH/30 In these circumstances:
1014 . Two addresses routed to the same list of hosts;
1015 . First host does not offer TLS;
1016 . First host accepts first address;
1017 . First host gives temporary error to second address;
1018 . Second host offers TLS and a TLS session is established;
1019 . Second host accepts second address.
1020 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1021 and peerdn, if requested) that were in fact used only for the second
1022 address.
7e8bec7a 1023
c688b954
PH
1024PH/31 When doing a callout as part of verifying an address, Exim was not paying
1025 attention to any local part prefix or suffix that was matched by the
1026 router that accepted the address. It now behaves in the same way as it
1027 does for delivery: the affixes are removed from the local part unless
1028 rcpt_include_affixes is set on the transport.
1029
fed77020
PH
1030PH/32 Add the sender address, as F=<...>, to the log line when logging a
1031 timeout during the DATA phase of an incoming message.
1032
7fe1560f
PH
1033PH/33 Sieve envelope tests were broken for match types other than :is. I have
1034 applied a patch sanctioned by the Sieve maintainer.
c688b954 1035
ebb6e6d5
PH
1036PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1037 the uid or gid is negative. A case of a negative gid caused this to be
1038 noticed. The fix allows for either to be negative.
1039
9c4e8f60
PH
1040PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1041 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1042 overlooked.
1043
1044PH/36 The change PH/12 above was broken. Fixed it.
1045
d7174846
PH
1046PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1047 the grounds that routing the same address twice would always produce the
1048 same answer. This might have been true once, but it is certainly no
1049 longer true now. Routing a child address may depend on the previous
1050 routing that produced that child. Some complicated redirection strategies
1051 went wrong when messages had multiple recipients, and made Exim's
1052 behaviour dependent on the order in which the addresses were given.
1053
1054 I have moved the duplicate checking until after the routing is complete.
1055 Exim scans the addresses that are assigned to local and remote
1056 transports, and removes any duplicates. This means that more work will be
1057 done, as duplicates will always all be routed, but duplicates are
1058 presumably rare, so I don't expect this is of any significance.
1059
1060 For deliveries to pipes, files, and autoreplies, the duplicate checking
1061 still happens during the routing process, since they are not going to be
1062 routed further.
1063
cfe75fc3
PH
1064PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1065 It corrects a timeout issue with spamd. This is Ian's comment: "The
1066 background is that sometimes spamd either never reads data from a
1067 connection it has accepted, or it never writes response data. The exiscan
1068 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1069 blindly assumes that writes won't block so it may never time out."
1070
be22d70e
PH
1071PH/39 Allow G after quota size as well as K and M.
1072
0612b098
PH
1073PH/40 The value set for $authenticated_id in an authenticator may not contain
1074 binary zeroes or newlines because the value is written to log lines and
1075 to spool files. There was no check on this. Now the value is run through
1076 the string_printing() function so that such characters are converted to
1077 printable escape sequences.
1078
2e0c1448
PH
1079PH/41 $message_linecount is a new variable that contains the total number of
1080 lines in the message. Compare $body_linecount, which is the count for the
1081 body only.
1082
447d236c
PH
1083PH/42 Exim no longer gives details of delivery errors for specific addresses in
1084 bounce and delay warning messages, except in certain special cases, which
1085 are as follows:
1086
1087 (a) An SMTP error message from a remote host;
1088 (b) A message specified in a :fail: redirection;
1089 (c) A message specified in a "fail" command in a system filter;
1090 (d) A message specified in a FAIL return from the queryprogram router;
1091 (e) A message specified by the cannot_route_message router option.
1092
1093 In these cases only, Exim does include the error details in bounce and
1094 warning messages. There are also a few cases where bland messages such
1095 as "unrouteable address" or "local delivery error" are given.
1096
d20976dc
PH
1097PH/43 $value is now also set for the "else" part of a ${run expansion.
1098
f656d135
PH
1099PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1100 being worked on, but at least Exim now implements the latest version to
1101 play with."
1102
2e2a30b4
PH
1103PH/45 In a pipe transport, although a timeout while waiting for the pipe
1104 process to complete was treated as a delivery failure, a timeout while
1105 writing the message to the pipe was logged, but erroneously treated as a
1106 successful delivery. Such timeouts include transport filter timeouts. For
1107 consistency with the overall process timeout, these timeouts are now
1108 treated as errors, giving rise to delivery failures by default. However,
1109 there is now a new Boolean option for the pipe transport called
1110 timeout_defer, which, if set TRUE, converts the failures into defers for
1111 both kinds of timeout. A transport filter timeout is now identified in
1112 the log output.
1113
9176e9f0
PH
1114PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1115 systems where "make" and "gmake" are different, calling "gmake" at top
1116 level broke things. I've arranged for the value of $(MAKE) to be passed
1117 from the Makefile to this script so that it can call the same version of
1118 "make".
1119
7982096b 1120
bbe902f0
PH
1121A note about Exim versions 4.44 and 4.50
1122----------------------------------------
1123
1124Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1125changes of various kinds. As a consequence, a big documentation update was
1126needed. This delayed the release for rather longer than seemed good, especially
1127in the light of a couple of (minor) security issues. Therefore, the changes
1128that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1129release. So 4.44 and 4.50 are in effect two different branches that both start
1130from 4.43.
1131
1132I have left the 4.50 change log unchanged; it contains all the changes since
11334.43. The change log for 4.44 is below; many of its items are identical to
1134those for 4.50. This seems to be the most sensible way to preserve the
1135historical information.
1136
1137
f7b63901 1138Exim version 4.50
495ae4b0
PH
1139-----------------
1140
5fe762f6
PH
1141 1. Minor wording change to the doc/README.SIEVE file.
1142
139059f6 1143 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1144 computation of the current number of files was incorrect.
495ae4b0 1145
7086e875
PH
1146 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1147 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1148 place.
1149
35af9f61
PH
1150 4. Give more explanation in the error message when the command for a transport
1151 filter fails to execute.
1152
b668c215
PH
1153 5. There are several places where Exim runs a non-Exim command in a
1154 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1155 was being done only for the command run by the queryprogram router. It is
1156 now done for all such subprocesses. The other cases are: ${run, transport
1157 filters, and the commands run by the lmtp and pipe transports.
1158
a494b1e1
PH
1159 6. Added CONFIGURE_GROUP build-time option.
1160
1161 7. Some older OS have a limit of 256 on the maximum number of file
1162 descriptors. Exim was using setrlimit() to set 1000 as a large value
1163 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1164 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1165
c5fcb476
PH
1166 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1167 was an oversight, and furthermore, ever since the addition of extra
1168 controls (e.g. 4.43/32), the checks on when to allow different forms of
1169 "control" were broken. There should now be diagnostics for all cases when a
1170 control that does not make sense is encountered.
1171
69358f02
PH
1172 9. Added the /retain_sender option to "control=submission".
1173
5be20824
PH
117410. $recipients is now available in the predata ACL (oversight).
1175
eb2c0248
PH
117611. Tidy the search cache before the fork to do a delivery from a message
1177 received from the command line. Otherwise the child will trigger a lookup
1178 failure and thereby defer the delivery if it tries to use (for example) a
1179 cached ldap connection that the parent has called unbind on.
1180
2a3eea10
PH
118112. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1182 of $address_data from the recipient verification was clobbered by the
1183 sender verification.
1184
118513. The value of address_data from a sender verification is now available in
1186 $sender_address_data in subsequent conditions in the ACL statement.
1187
23c7ff99
PH
118814. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1189
4deaf07d
PH
119015. Added a new option "connect=<time>" to callout options, to set a different
1191 connection timeout.
1192
926e1192
PH
119316. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1194 was its contents. (It was OK if the option was not defined at all.)
1195
650edc6f
PH
119617. A "Completed" log line is now written for messages that are removed from
1197 the spool by the -Mrm option.
1198
2c7db3f5
PH
119918. New variables $sender_verify_failure and $recipient_verify_failure contain
1200 information about exactly what failed.
1201
3d235903
PH
120219. Added -dd to debug only the daemon process.
1203
7c7ad977
PH
120420. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1205 handles timeouts, both on the server side and network timeouts. Renamed the
1206 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1207
981756db
PH
120821. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1209 It is now set to "smtps".
1210
d4eb88df
PH
121122. $host_address is now set to the target address during the checking of
1212 ignore_target_hosts.
1213
121423. When checking ignore_target_hosts for an ipliteral router, no host name was
1215 being passed; this would have caused $sender_host_name to have been used if
1216 matching the list had actually called for a host name (not very likely,
1217 since this list is usually IP addresses). A host name is now passed as
1218 "[x.x.x.x]".
1219
7d468ab8
PH
122024. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1221 code that specifies a non-restarting handler (typically sigaction() in
1222 modern systems) in an attempt to fix a rare and obscure crash bug.
1223
122425. Narrowed the window for a race in the daemon that could cause it to ignore
1225 SIGCHLD signals. This is not a major problem, because they are used only to
1226 wake it up if nothing else does.
1227
62c0818f
PH
122826. A malformed maildirsize file could cause Exim to calculate negative values
1229 for the mailbox size or file count. Odd effects could occur as a result.
1230 The maildirsize information is now recalculated if the size or filecount
1231 end up negative.
1232
26034054
PH
123327. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1234 support for a long time. Removed HAVE_SYS_VFS_H.
1235
af66f652
PH
123628. Installed the latest version of exipick from John Jetmore.
1237
90af77f4
PH
123829. In an address list, if the pattern was not a regular expression, an empty
1239 subject address (from a bounce message) matched only if the pattern was an
1240 empty string. Non-empty patterns were not even tested. This was the wrong
1241 because it is perfectly reasonable to use an empty address as part of a
1242 database query. An empty address is now tested by patterns that are
1243 lookups. However, all the other forms of pattern expect the subject to
1244 contain a local part and a domain, and therefore, for them, an empty
1245 address still always fails if the pattern is not itself empty.
1246
d8ef3577
PH
124730. Exim went into a mad DNS loop when attempting to do a callout where the
1248 host was specified on an smtp transport, and looking it up yielded more
1249 than one IP address.
1250
5cb8cbc6
PH
125131. Re-factored the code for checking spool and log partition space into a
1252 function that finds that data and another that does the check. The former
1253 is then used to implement four new variables: $spool_space, $log_space,
1254 $spool_inodes, and $log_inodes.
1255
14702f5b
PH
125632. The RFC2047 encoding function was originally intended for short strings
1257 such as real names; it was not keeping to the 75-character limit for
1258 encoded words that the RFC imposes. It now respects the limit, and
1259 generates multiple encoded words if necessary. To be on the safe side, I
1260 have increased the buffer size for the ${rfc2047: expansion operator from
1261 1024 to 2048 bytes.
1262
063b1e99
PH
126333. It is now permitted to omit both strings after an "if" condition; if the
1264 condition is true, the result is "true". As before, when the second string
1265 is omitted, a false condition yields an empty string. This makes it less
1266 cumbersome to write custom ACL and router conditions.
1267
652e1b65
PH
126834. Failure to deliver a bounce message always caused it to be frozen, even if
1269 there was an errors_to setting on the router. The errors_to setting is now
1270 respected.
1271
6f0c9a4f
PH
127235. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1273 canonical form (fully expanded) before being placed in
1274 $sender_host_address.
1275
33397d19
PH
127636. The table in the code that translates DNS record types into text (T_A to
1277 "A" for instance) was missing entries for NS and CNAME. It is just possible
1278 that this could have caused confusion if both these types were looked up
1279 for the same domain, because the text type is used as part of Exim's
1280 per-process caching. But the chance of anyone hitting this buglet seems
1281 very small.
1282
7bb56e1f
PH
128337. The dnsdb lookup has been extended in a number of ways.
1284
1285 (1) There is a new type, "zns", which walks up the domain tree until it
1286 finds some nameserver records. It should be used with care.
1287
ea3bc19b
PH
1288 (2) There is a new type, "mxh", which is like "mx" except that it returns
1289 just the host names, not the priorities.
1290
1291 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1292 looked up. The behaviour when one of the lookups defers can be
1293 controlled by a keyword.
7bb56e1f 1294
ea3bc19b 1295 (4) It is now possible to specify the separator character for use when
7bb56e1f 1296 multiple records are returned.
33397d19 1297
0bcb2a0e
PH
129838. The dnslists ACL condition has been extended: it is now possible to supply
1299 a list of IP addresses and/or domains to be looked up in a particular DNS
1300 domain.
1301
2ac0e484
PH
130239. Added log_selector=+queue_time_overall.
1303
4e1fde53
PH
130440. When running the queue in the test harness, wait just a tad after forking a
1305 delivery process, to get repeatability of debugging output.
1306
de365ded
PH
130741. Include certificate and key file names in error message when GnuTLS fails
1308 to set them up, because the GnuTLS error message doesn't include the name
1309 of the failing file when there is a problem reading it.
1310
f05da2e8
PH
131142. Allow both -bf and -bF in the same test run.
1312
d6453af2
PH
131343. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1314
f7b63901
PH
131544. The "Exiscan patch" is now merged into the mainline Exim source.
1316
131745. Sometimes the final signoff response after QUIT could fail to get
1318 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1319 < 0 before doing a fflush(). This bug looks as though it goes back to the
1320 introduction of TLS in release 3.20, but "sometimes" must have been rare
1321 because the tests only now provoked it.
1322
a444213a
PH
132346. Reset the locale to "C" after calling embedded Perl, in case it was changed
1324 (this can affect the format of dates).
1325
0ec020ea
PH
132647. exim_tidydb, when checking for the continued existence of a message for
1327 which it has found a message-specific retry record, was not finding
1328 messages that were in split spool directories. Consequently, it was
1329 deleting retry records that should have stayed in existence.
1330
b1206957
PH
133148. Steve fixed some bugs in eximstats.
1332
133349. The SPA authentication driver was not abandoning authentication and moving
1334 on to the next authenticator when an expansion was forced to fail,
1335 contradicting the general specification for all authenticators. Instead it
1336 was generating a temporary error. It now behaves as specified.
1337
26dd5a95
PH
133850. The default ordering of permitted cipher suites for GnuTLS was pessimal
1339 (the order specifies the preference for clients). The order is now AES256,
1340 AES128, 3DES, ARCFOUR128.
1341
343b2385
PH
134251. Small patch to Sieve code - explicitly set From: when generating an
1343 autoreply.
1344
1c5466b9
PH
134552. Exim crashed if a remote delivery caused a very long error message to be
1346 recorded - for instance if somebody sent an entire SpamAssassin report back
1347 as a large number of 550 error lines. This bug was coincidentally fixed by
1348 increasing the size of one of Exim's internal buffers (big_buffer) that
1349 happened as part of the Exiscan merge. However, to be on the safe side, I
1350 have made the code more robust (and fixed the comments that describe what
1351 is going on).
1352
55ee9ee3
PH
135353. Now that there can be additional text after "Completed" in log lines (if
1354 the queue_time_overall log selector is set), a one-byte patch to exigrep
1355 was needed to allow it to recognize "Completed" as not the last thing in
1356 the line.
1357
d38f8232
PH
135854. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1359 patch that reportedly fixes this has been added. I am not expert enough to
1360 create a test for it. This is what the patch creator wrote:
1361
1362 "I found a little strange behaviour of ldap code when working with
1363 Windows 2003 AD Domain, where users was placed in more than one
1364 Organization Units. When I tried to give exim partial DN, the exit code
1365 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1366 But simultaneously result of request was absolutely normal ldap result,
1367 so I produce this patch..."
1368
3295e65b
PH
1369 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1370 REFERENCE, so I have modified the code to exclude the patch when that macro
1371 is not defined.
1372
7102e136
PH
137355. Some experimental protocols are using DNS PTR records for new purposes. The
1374 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1375 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1376 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1377 CAN-2005-0021
7102e136 1378
3ca0ba97
PH
137956. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1380
c2bcbe20
PH
138157. Double the size of the debug message buffer (to 2048) so that more of very
1382 long debug lines gets shown.
1383
18ce445d
PH
138458. The exicyclog utility now does better if the number of log files to keep
1385 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1386
1f5b4c3d
PH
138759. Two changes related to the smtp_active_hostname option:
1388
1389 (1) $smtp_active_hostname is now available as a variable.
1390 (2) The default for smtp_banner uses $smtp_active_hostname instead
1391 of $primary_hostname.
1392
b975ba52
PH
139360. The host_aton() function is supposed to be passed a string that is known
1394 to be a valid IP address. However, in the case of IPv6 addresses, it was
1395 not checking this. This is a hostage to fortune. Exim now panics and dies
1396 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1397 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1398 components; fortuitously, this particular loophole had already been fixed
1399 by change 4.50/55 above.
1400
1401 If there are any other similar loopholes, the new check in host_aton()
1402 itself should stop them being exploited. The report I received stated that
1403 data on the command line could provoke the exploit when Exim was running as
1404 exim, but did not say which command line option was involved. All I could
1405 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1406 running as the user.
ea3a6f44 1407 CAN-2005-0021
85b87bc2
PH
1408
140961. There was a buffer overflow vulnerability in the SPA authentication code
1410 (which came originally from the Samba project). I have added a test to the
1411 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1412 CAN-2005-0022
b975ba52 1413
17ffcae7
PH
141462. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1415 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1416
d95f9fdb
PH
141763. The daemon start-up calls getloadavg() while still root for those OS that
1418 need the first call to be done as root, but it missed one case: when
1419 deliver_queue_load_max is set with deliver_drop_privilege. This is
1420 necessary for the benefit of the queue runner, because there is no re-exec
1421 when deliver_drop_privilege is set.
1422
86b8287f
PH
142364. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1424 This has been fixed.
1425
60dc5e56
PH
142665. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1427 was in use, was not putting the data itself into the right store pool;
1428 consequently, it could be overwritten for a subsequent message in the same
1429 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1430 the caching.)
1431
533244af
PH
143266. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1433
a5a28604
PH
143467. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1435 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1436 the function were treating the return as a boolean value, which happened to
1437 work because 0=false and not-0=true, but is not correct code.
1438
7e634d24
PH
143968. The host_aton() function was not handling scoped IPv6 addresses (those
1440 with, for example, "%eth0" on the end) correctly.
1441
3e11c26b
PH
144269. Fixed some compiler warnings in acl.c for the bitmaps specified with
1443 negated items (that is, ~something) in unsigned ints. Some compilers
1444 apparently mutter when there is no cast.
1445
6729cf78
PH
144670. If an address verification called from an ACL failed, and did not produce a
1447 user-specific message (i.e. there was only a "system" message), nothing was
1448 put in $acl_verify_message. In this situation, it now puts the system
1449 message there.
1450
00f00ca5
PH
145171. Change 4.23/11 added synchronization checking at the start of an SMTP
1452 session; change 4.31/43 added the unwanted input to the log line - except
1453 that it did not do this in the start of session case. It now does.
1454
c9bdd01c
PH
145572. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1456 This is wrong and can cause the other end to generate a synchronization
1457 error if it is another Exim or anything else that does the synchronization
1458 check. A QUIT command is no longer sent after a timeout.
1459
d43194df
PH
146073. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1461 during host lookups.
1462
fe5b5d0b
PH
146374. The defer_ok option of callout verification was not working if it was used
1464 when verifying addresses in header lines, that is, for this case:
1465
1466 verify = header_sender/callout=defer_ok
1467
76a2d7ba
PH
146875. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1469 those file descriptors could be used for SMTP connections. If anything
1470 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1471 could be sent to the SMTP client, causing chaos. The daemon now opens
1472 stdin, stdout, and stderr to /dev/null when it puts itself into the
1473 background.
1474
147576. Arrange for output from Perl's "warn" command to be written to Exim's main
1476 log by default. The user can override this with suitable Perl magic.
1477
04f7d5b9
PH
147877. The use of log_message on a "discard" ACL verb, which is supposed to add to
1479 the log message when discard triggers, was not working for the DATA ACL or
1480 for the non-SMTP ACL.
1481
bc60667e
PH
148278. Error message wording change in sieve.c.
1483
bb6e88ff
PH
148479. If smtp_accept_max_per_host was set, the number of connections could be
1485 restricted to fewer than expected, because the daemon was trying to set up
1486 a new connection before checking whether the processes handling previous
1487 connections had finished. The check for completed processes is now done
1488 earlier. On busy systems, this bug wouldn't be noticed because something
1489 else would have woken the daemon, and it would have reaped the completed
1490 process earlier.
1491
1e70f85b
PH
149280. If a message was submitted locally by a user whose login name contained one
1493 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1494 It caused a spool format error. I have fixed the spool reading code. A
1495 related problem was that the "from" clause in the Received: line became
1496 illegal because of the space(s). It is now covered by ${quote_local_part.
1497
149881. Included the latest eximstats from Steve (adds average sizes to HTML Top
1499 tables).
1500
4e01f9d6
PH
150182. Updated OS/Makefile-AIX as per message from Mike Meredith.
1502
1ee1cef2
PH
150383. Patch from Sieve maintainer to fix unterminated string problem in
1504 "vacation" handling.
1505
6e2b4ccc
PH
150684. Some minor changes to the Linux configuration files to help with other
1507 OS variants using glibc.
1508
8e669ac1
PH
150985. One more patch for Sieve to update vacation handling to latest spec.
1510
495ae4b0 1511
bbe902f0
PH
1512----------------------------------------------------
1513See the note above about the 4.44 and 4.50 releases.
1514----------------------------------------------------
1515
1516
1517Exim version 4.44
1518-----------------
1519
1520 1. Change 4.43/35 introduced a bug that caused file counts to be
1521 incorrectly computed when quota_filecount was set in an appendfile
1522 transport
1523
1524 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1525 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1526 place.
1527
1528 3. Give more explanation in the error message when the command for a transport
1529 filter fails to execute.
1530
1531 4. There are several places where Exim runs a non-Exim command in a
1532 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1533 was being done only for the command run by the queryprogram router. It is
1534 now done for all such subprocesses. The other cases are: ${run, transport
1535 filters, and the commands run by the lmtp and pipe transports.
1536
1537 5. Some older OS have a limit of 256 on the maximum number of file
1538 descriptors. Exim was using setrlimit() to set 1000 as a large value
1539 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1540 systems. I've change it so that if it can't get 1000, it tries for 256.
1541
1542 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1543 was an oversight, and furthermore, ever since the addition of extra
1544 controls (e.g. 4.43/32), the checks on when to allow different forms of
1545 "control" were broken. There should now be diagnostics for all cases when a
1546 control that does not make sense is encountered.
1547
1548 7. $recipients is now available in the predata ACL (oversight).
1549
1550 8. Tidy the search cache before the fork to do a delivery from a message
1551 received from the command line. Otherwise the child will trigger a lookup
1552 failure and thereby defer the delivery if it tries to use (for example) a
1553 cached ldap connection that the parent has called unbind on.
1554
1555 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1556 of $address_data from the recipient verification was clobbered by the
1557 sender verification.
1558
155910. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1560 was its contents. (It was OK if the option was not defined at all.)
1561
156211. A "Completed" log line is now written for messages that are removed from
1563 the spool by the -Mrm option.
1564
156512. $host_address is now set to the target address during the checking of
1566 ignore_target_hosts.
1567
156813. When checking ignore_target_hosts for an ipliteral router, no host name was
1569 being passed; this would have caused $sender_host_name to have been used if
1570 matching the list had actually called for a host name (not very likely,
1571 since this list is usually IP addresses). A host name is now passed as
1572 "[x.x.x.x]".
1573
157414. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1575 code that specifies a non-restarting handler (typically sigaction() in
1576 modern systems) in an attempt to fix a rare and obscure crash bug.
1577
157815. Narrowed the window for a race in the daemon that could cause it to ignore
1579 SIGCHLD signals. This is not a major problem, because they are used only to
1580 wake it up if nothing else does.
1581
158216. A malformed maildirsize file could cause Exim to calculate negative values
1583 for the mailbox size or file count. Odd effects could occur as a result.
1584 The maildirsize information is now recalculated if the size or filecount
1585 end up negative.
1586
158717. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1588 support for a long time. Removed HAVE_SYS_VFS_H.
1589
ea3a6f44 159018. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1591
159219. Allow an empty sender to be matched against a lookup in an address list.
1593 Previously the only cases considered were a regular expression, or an
1594 empty pattern.
1595
159620. Exim went into a mad DNS lookup loop when doing a callout where the
1597 host was specified on the transport, if the DNS lookup yielded more than
1598 one IP address.
1599
ea3a6f44
NM
160021. The RFC2047 encoding function was originally intended for short strings
1601 such as real names; it was not keeping to the 75-character limit for
1602 encoded words that the RFC imposes. It now respects the limit, and
1603 generates multiple encoded words if necessary. To be on the safe side, I
1604 have increased the buffer size for the ${rfc2047: expansion operator from
1605 1024 to 2048 bytes.
bbe902f0 1606
ea3a6f44
NM
160722. Failure to deliver a bounce message always caused it to be frozen, even if
1608 there was an errors_to setting on the router. The errors_to setting is now
1609 respected.
bbe902f0
PH
1610
161123. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1612 canonical form (fully expanded) before being placed in
1613 $sender_host_address.
1614
161524. Updated eximstats to version 1.33
1616
ea3a6f44
NM
161725. Include certificate and key file names in error message when GnuTLS fails
1618 to set them up, because the GnuTLS error message doesn't include the name
1619 of the failing file when there is a problem reading it.
bbe902f0
PH
1620
162126. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1622 As per change 25.
bbe902f0 1623
ea3a6f44
NM
162427. Reset the locale to "C" after calling embedded Perl, in case it was changed
1625 (this can affect the format of dates).
bbe902f0 1626
ea3a6f44
NM
162728. exim_tidydb, when checking for the continued existence of a message for
1628 which it has found a message-specific retry record, was not finding
1629 messages that were in split spool directories. Consequently, it was
1630 deleting retry records that should have stayed in existence.
bbe902f0
PH
1631
163229. eximstats updated to version 1.35
1633 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1634 1.35 - bugfix such that pie charts by volume are generated correctly
1635
ea3a6f44
NM
163630. The SPA authentication driver was not abandoning authentication and moving
1637 on to the next authenticator when an expansion was forced to fail,
1638 contradicting the general specification for all authenticators. Instead it
1639 was generating a temporary error. It now behaves as specified.
bbe902f0 1640
ea3a6f44
NM
164131. The default ordering of permitted cipher suites for GnuTLS was pessimal
1642 (the order specifies the preference for clients). The order is now AES256,
1643 AES128, 3DES, ARCFOUR128.
bbe902f0 1644
ea3a6f44
NM
164531. Small patch to Sieve code - explicitly set From: when generating an
1646 autoreply.
bbe902f0 1647
ea3a6f44
NM
164832. Exim crashed if a remote delivery caused a very long error message to be
1649 recorded - for instance if somebody sent an entire SpamAssassin report back
1650 as a large number of 550 error lines. This bug was coincidentally fixed by
1651 increasing the size of one of Exim's internal buffers (big_buffer) that
1652 happened as part of the Exiscan merge. However, to be on the safe side, I
1653 have made the code more robust (and fixed the comments that describe what
1654 is going on).
bbe902f0 1655
ea3a6f44
NM
165633. Some experimental protocols are using DNS PTR records for new purposes. The
1657 keys for these records are domain names, not reversed IP addresses. The
1658 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1659 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1660 CAN-2005-0021
1661
ea3a6f44
NM
166234. The host_aton() function is supposed to be passed a string that is known
1663 to be a valid IP address. However, in the case of IPv6 addresses, it was
1664 not checking this. This is a hostage to fortune. Exim now panics and dies
1665 if the condition is not met. A case was found where this could be provoked
1666 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1667 components; fortuitously, this particular loophole had already been fixed
1668 by change 4.50/55 or 4.44/33 above.
1669
1670 If there are any other similar loopholes, the new check in host_aton()
1671 itself should stop them being exploited. The report I received stated that
1672 data on the command line could provoke the exploit when Exim was running as
1673 exim, but did not say which command line option was involved. All I could
1674 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1675 running as the user.
bbe902f0
PH
1676 CAN-2005-0021
1677
ea3a6f44
NM
167835. There was a buffer overflow vulnerability in the SPA authentication code
1679 (which came originally from the Samba project). I have added a test to the
1680 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1681 CAN-2005-0022
1682
ea3a6f44
NM
168336. The daemon start-up calls getloadavg() while still root for those OS that
1684 need the first call to be done as root, but it missed one case: when
1685 deliver_queue_load_max is set with deliver_drop_privilege. This is
1686 necessary for the benefit of the queue runner, because there is no re-exec
1687 when deliver_drop_privilege is set.
bbe902f0 1688
ea3a6f44
NM
168937. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1690 was in use, was not putting the data itself into the right store pool;
1691 consequently, it could be overwritten for a subsequent message in the same
1692 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1693 the caching.)
bbe902f0 1694
ea3a6f44
NM
169538. Sometimes the final signoff response after QUIT could fail to get
1696 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1697 < 0 before doing a fflush(). This bug looks as though it goes back to the
1698 introduction of TLS in release 3.20, but "sometimes" must have been rare
1699 because the tests only now provoked it.
bbe902f0
PH
1700
1701
495ae4b0
PH
1702Exim version 4.43
1703-----------------
1704
1705 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1706 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1707 Now it returns an integer. A number of calls were still expecting a T/F
1708 return. Fortuitously, in all cases, the tests worked in OK situations,
1709 which is the norm. However, things would have gone wrong on any write
1710 failures on the smtp file descriptor. This function is used when sending
1711 messages over SMTP and also when doing verify callouts.
1712
1713 2. When Exim is called to do synchronous delivery of a locally submitted
1714 message (the -odf or -odi options), it no longer closes stderr before doing
1715 the delivery.
1716
1717 3. Implemented the mua_wrapper option.
1718
1719 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1720
1721 5. Implemented the functions header_remove(), header_testname(),
1722 header_add_at_position(), and receive_remove_recipient(), and exported them
1723 to local_scan().
1724
1725 6. If an ACL "warn" statement specified the addition of headers, Exim already
1726 inserted X-ACL-Warn: at the start if there was no header name. However, it
1727 was not making this test for the second and subsequent header lines if
1728 there were newlines in the string. This meant that an invalid header could
1729 be inserted if Exim was badly configured.
1730
1731 7. Allow an ACL "warn" statement to add header lines at the start or after all
1732 the Received: headers, as well as at the end.
1733
1734 8. Added the rcpt_4xx retry error code.
1735
1736 9. Added postmaster_mailfrom=xxx to callout verification option.
1737
173810. Added mailfrom=xxxx to the callout verification option, for verify=
1739 header_sender only.
1740
174111. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1742 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1743
174412. Inserted some casts to stop certain compilers warning when using pointer
1745 differences as field lengths or precisions in printf-type calls (mostly
1746 affecting debugging statements).
1747
174813. Added optional readline() support for -be (dynamically loaded).
1749
175014. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1751 same clock tick as a message's arrival, so that its received time was the
1752 same as the "first fail" time on the retry record, and that message
1753 remained on the queue past the ultimate address timeout, every queue runner
1754 would try a delivery (because it was past the ultimate address timeout) but
1755 after another failure, the ultimate address timeout, which should have then
1756 bounced the address, did not kick in. This was a "< instead of <=" error;
1757 in most cases the first failure would have been in the next clock tick
1758 after the received time, and all would be well.
1759
176015. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1761 being recognized when the domain list was tested by the match_domain
1762 condition in an expansion string.
1763
176416. Added the ${str2b64: operator.
1765
176617. Exim was always calling setrlimit() to set a large limit for the number of
1767 processes, without checking whether the existing limit was already
1768 adequate. (It did check for the limit on file descriptors.) Furthermore,
1769 errors from getrlimit() and setrlimit() were being ignored. Now they are
1770 logged to the main and panic logs, but Exim does carry on, to try to do its
1771 job under whatever limits there are.
1772
177318. Imported PCRE 5.0.
1774
177519. Trivial typo in log message " temporarily refused connection" (the leading
1776 space).
1777
177820. If the log selector return_path_on_delivery was set and an address was
1779 redirected to /dev/null, the delivery process crashed because it assumed
1780 that a return path would always be set for a "successful" delivery. In this
1781 case, the whole delivery is bypassed as an optimization, and therefore no
1782 return path is set.
1783
178421. Internal re-arrangement: the function for sending a challenge and reading
1785 a response while authentication was assuming a zero-terminated challenge
1786 string. It's now changed to take a pointer and a length, to allow for
1787 binary data in such strings.
1788
178922. Added the cyrus_sasl authenticator (code supplied by MBM).
1790
179123. Exim was not respecting finduser_retries when seeking the login of the
1792 uid under which it was called; it was always trying 10 times. (The default
1793 setting of finduser_retries is zero.) Also, it was sleeping after the final
1794 failure, which is pointless.
1795
179624. Implemented tls_on_connect_ports.
1797
179825. Implemented acl_smtp_predata.
1799
180026. If the domain in control=submission is set empty, Exim assumes that the
1801 authenticated id is a complete email address when it generates From: or
1802 Sender: header lines.
1803
180427. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1805 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1806 chown and chgrp in /bin and hostname in /usr/bin.
1807
180828. Exim was keeping the "process log" file open after each use, just as it
1809 does for the main log. This opens the possibility of it remaining open for
1810 long periods when the USR1 signal hits a daemon. Occasional processlog
1811 errors were reported, that could have been caused by this. Anyway, it seems
1812 much more sensible not to leave this file open at all, so that is what now
1813 happens.
1814
181529. The long-running daemon process does not normally write to the log once it
1816 has entered its main loop, and it closes the log before doing so. This is
1817 so that log files can straightforwardly be renamed and moved. However,
1818 there are a couple of unusual error situations where the daemon does write
1819 log entries, and I had neglected to close the log afterwards.
1820
182130. The text of an SMTP error response that was received during a remote
1822 delivery was being truncated at 512 bytes. This is too short for some of
1823 the long messages that one sometimes sees. I've increased the limit to
1824 1024.
1825
182631. It is now possible to make retry rules that apply only when a message has a
1827 specific sender, in particular, an empty sender.
1828
182932. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1830 it possible to be selective about when SMTP synchronization is enforced.
1831
183233. Added "control = caseful_local_part" and "control = "caselower_local_part".
1833
183432. Implemented hosts_connection_nolog.
1835
183633. Added an ACL for QUIT.
1837
183834. Setting "delay_warning=" to disable warnings was not working; it gave a
1839 syntax error.
1840
184135. Added mailbox_size and mailbox_filecount to appendfile.
1842
184336. Added control = no_multiline_responses to ACLs.
1844
184537. There was a bug in the logic of the code that waits for the clock to tick
1846 in the case where the clock went backwards by a substantial amount such
1847 that the microsecond fraction of "now" was more than the microsecond
1848 fraction of "then" (but the whole seconds number was less).
1849
185038. Added support for the libradius Radius client library this is found on
1851 FreeBSD (previously only the radiusclient library was supported).
1852
1853
1854Exim version 4.42
1855-----------------
1856
1857 1. When certain lookups returned multiple values in the form name=value, the
1858 quoting of the values was not always being done properly. Specifically:
1859 (a) If the value started with a double quote, but contained no whitespace,
1860 it was not quoted.
1861 (b) If the value contained whitespace other than a space character (i.e.
1862 tabs or newlines or carriage returns) it was not quoted.
1863 This fix has been applied to the mysql and pgsql lookups by writing a
1864 separate quoting function and calling it from the lookup code. The fix
1865 should probably also be applied to nisplus, ibase and oracle lookups, but
1866 since I cannot test any of those, I have not disturbed their existing code.
1867
1868 2. A hit in the callout cache for a specific address caused a log line with no
1869 reason for rejecting RCPT. Now it says "Previous (cached) callout
1870 verification failure".
1871
1872 3. There was an off-by-one bug in the queryprogram router. An over-long
1873 return line was truncated at 256 instead of 255 characters, thereby
1874 overflowing its buffer with the terminating zero. As well as fixing this, I
1875 have increased the buffer size to 1024 (and made a note to document this).
1876
1877 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1878 when Exim is waiting for an SMTP response from a remote server, Exim
1879 restarts its select() call on the socket, thereby resetting its timeout.
1880 This is not a problem when such interrupts are rare. Somebody set up a cron
1881 job to run exiwhat every 2 minutes, which is less than the normal select()
1882 timeout (5 or 10 minutes). This meant that the select() timeout never
1883 kicked in because it was always reset. I have fixed this by comparing the
1884 time when an interrupt arrives with the time at the start of the first call
1885 to select(). If more time than the timeout has elapsed, the interrupt is
1886 treated as a timeout.
1887
1888 5. Some internal re-factoring in preparation for the addition of Sieve
1889 extensions (by MH). In particular, the "personal" test is moved to a
1890 separate function, and given an option for scanning Cc: and Bcc: (which is
1891 not set for Exim filters).
1892
1893 6. When Exim created an email address using the login of the caller as the
1894 local part (e.g. when creating a From: or Sender: header line), it was not
1895 quoting the local part when it contained special characters such as @.
1896
1897 7. Installed new OpenBSD configuration files.
1898
1899 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1900 try to make them clearer.
1901
1902 9. Callout options, other than the timeout value, were being ignored when
1903 verifying sender addresses in header lines. For example, when using
1904
1905 verify = header_sender/callout=no_cache
1906
1907 the cache was (incorrectly) being used.
1908
190910. Added a missing instance of ${EXE} to the exim_install script; this affects
1910 only the Cygwin environment.
1911
191211. When return_path_on_delivery was set as a log selector, if different remote
1913 addresses in the same message used different return paths and parallel
1914 remote delivery occurred, the wrong values would sometimes be logged.
1915 (Whenever a remote delivery process finished, the return path value from
1916 the most recently started remote delivery process was logged.)
1917
191812. RFC 3848 specifies standard names for the "with" phrase in Received: header
1919 lines when AUTH and/or TLS are in use. This is the "received protocol"
1920 field. Exim used to use "asmtp" for authenticated SMTP, without any
1921 indication (in the protocol name) for TLS use. Now it follows the RFC and
1922 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1923 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1924 names appear in log lines as well as in Received: header lines.
1925
192613. Installed MH's patches for Sieve to add the "copy" and "vacation"
1927 extensions, and comparison tests, and to fix some bugs.
1928
192914. Changes to the "personal" filter test:
1930
1931 (1) The test was buggy in that it was just doing the equivalent of
1932 "contains" tests on header lines. For example, if a user's address was
1933 anne@some.where, the "personal" test would incorrectly be true for
1934
1935 To: susanne@some.where
1936
1937 This test is now done by extracting each address from the header in turn,
1938 and checking the entire address. Other tests that are part of "personal"
1939 are now done using regular expressions (for example, to check local parts
1940 of addresses in From: header lines).
1941
1942 (2) The list of non-personal local parts in From: addresses has been
1943 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1944 taken from the Sieve specification recommendations.
1945
1946 (3) If the message contains any header line starting with "List-" it is
1947 treated as non-personal.
1948
1949 (4) The test for "circular" in the Subject: header line has been removed
1950 because it now seems ill-conceived.
1951
195215. Minor typos in src/EDITME comments corrected.
1953
195416. Installed latest exipick from John Jetmore.
1955
195617. If headers_add on a router specified a text string that was too long for
1957 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1958 of string_sprintf() is now avoided.
1959
196018. $message_body_size was not set (it was always zero) when running the DATA
1961 ACL and the local_scan() function.
1962
196319. For the "mail" command in an Exim filter, no default was being set for
1964 the once_repeat time, causing a random time value to be used if "once" was
1965 specified. (If the value happened to be <= 0, no repeat happened.) The
1966 default is now 0s, meaning "never repeat". The "vacation" command was OK
1967 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1968 (I found it when inspecting the code).
1969
197020. There is now an overall timeout for performing a callout verification. It
1971 defaults to 4 times the callout timeout, which applies to individual SMTP
1972 commands during the callout. The overall timeout applies when there is more
1973 than one host that can be tried. The timeout is checked before trying the
1974 next host. This prevents very long delays if there are a large number of
1975 hosts and all are timing out (e.g. when the network connections are timing
1976 out). The value of the overall timeout can be changed by specifying an
1977 additional sub-option for "callout", called "maxwait". For example:
1978
1979 verify = sender/callout=5s,maxwait=20s
1980
198121. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1982 to the end before writing, but this should make it even safer).
1983
198422. Exim was forgetting that it had advertised PIPELINING for the second and
1985 subsequent messages on an SMTP connection. It was also not resetting its
1986 memory on STARTTLS and an internal HELO.
1987
198823. When Exim logs an SMTP synchronization error within a session, it now
1989 records whether PIPELINING has been advertised or not.
1990
199124. Added 3 instances of "(long int)" casts to time_t variables that were being
1992 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1993 rather than long int.
1994
199525. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1996
199726. Added the never_mail option to autoreply.
1998
1999
2000Exim version 4.41
2001-----------------
2002
2003 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2004 crash if the getsockname() call failed; this can happen if a connection is
2005 closed very soon after it is established. The problem was simply in the
2006 order in which certain operations were done, causing Exim to try to write
2007 to the SMTP stream before it had set up the file descriptor. The bug has
2008 been fixed by making things happen in the correct order.
2009
2010
2011Exim version 4.40
2012-----------------
2013
2014 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2015 before the connection was closed, thus losing the rejection response.
2016
2017 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2018 some early Solaris releases, but causes trouble in current releases where
2019 socklen_t is defined.
2020
2021 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2022 always exist.
2023
2024 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2025 configured.
2026
2027 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2028 that releasing the top of it at the end releases what was used for sub-
2029 expansions (unless the block got too big). However, discard this block if
2030 the first thing is a variable or header, so that we can use its block when
2031 it is dynamic (useful for very large $message_headers, for example).
2032
2033 6. Lookups now cache *every* query, not just the most recent. A new, separate
2034 store pool is used for this. It can be recovered when all lookup caches are
2035 flushed. Lookups now release memory at the end of their result strings.
2036 This has involved some general refactoring of the lookup sources.
2037
2038 7. Some code has been added to the store_xxx() functions to reduce the amount
2039 of flapping under certain conditions.
2040
2041 8. log_incoming_interface used to affect only the <= reception log lines. Now
2042 it causes the local interface and port to be added to several more SMTP log
2043 lines, for example "SMTP connection from", and rejection lines.
2044
2045 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2046
204710. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2048
204911. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2050 could be overwritten at the end of the current message (or the start of a
2051 new message if it was set in a HELO ACL). The value is now preserved for
2052 the duration of the SMTP connection.
2053
205412. If a transport had a headers_rewrite setting, and a matching header line
2055 contained an unqualified address, that address was qualified, even if it
2056 did not match any rewriting rules. The underlying bug was that the values
2057 of the flags that permit the existence of unqualified sender and recipient
2058 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2059 non-local messages, and by -bnq for local messages) were not being
2060 preserved with the message after it was received.
2061
206213. When Exim was logging an SMTP synchronization error, it could sometimes log
2063 "next input=" as part of the text comprising the host identity instead of
2064 the correct text. The code was using the same buffer for two different
2065 strings. However, depending on which order the printing function evaluated
2066 its arguments, the bug did not always show up. Under Linux, for example, my
2067 test suite worked just fine.
2068
206914. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2070 doesn't work with some older versions of Perl. It has been changed to "my",
2071 which in any case is probably the better facility to use.
2072
207315. A really picky compiler found some instances of statements for creating
2074 error messages that either had too many or two few arguments for the format
2075 string.
2076
207716. The size of the buffer for calls to the DNS resolver has been increased
2078 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2079 for addresses that have a lot of PTR records. This alleviates a problem; it
2080 does not fully solve it.
2081
208217. A dnsdb lookup for PTR records that receives more data than will fit in the
2083 buffer now truncates the list and logs the incident, which is the same
2084 action as happens when Exim is looking up a host name and its aliases.
2085 Previously in this situation something unpredictable would happen;
2086 sometimes it was "internal error: store_reset failed".
2087
208818. If a server dropped the connection unexpectedly when an Exim client was
2089 using GnuTLS and trying to read a response, the client delivery process
2090 crashed while trying to generate an error log message.
2091
209219. If a "warn" verb in an ACL added multiple headers to a message in a single
2093 string, for example:
2094
2095 warn message = H1: something\nH2: something
2096
2097 the text was added as a single header line from Exim's point of view
2098 though it ended up OK in the delivered message. However, searching for the
2099 second and subsequent header lines using $h_h2: did not work. This has been
2100 fixed. Similarly, if a system filter added multiple headers in this way,
2101 the routers could not see them.
2102
210320. Expanded the error message when iplsearch is called with an invalid key to
2104 suggest using net-iplsearch in a host list.
2105
210621. When running tests using -bh, any delays imposed by "delay" modifiers in
2107 ACLs are no longer actually imposed (and a message to that effect is
2108 output).
2109
211022. If a "gecos" field in a passwd entry contained escaped characters, in
2111 particular, if it contained a \" sequence, Exim got it wrong when building
2112 a From: or a Sender: header from that name. A second bug also caused
2113 incorrect handling when an unquoted " was present following a character
2114 that needed quoting.
2115
211623. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2117 was not being matched caselessly.
2118
211924. Arranged for all hyphens in the exim.8 source to be escaped with
2120 backslashes.
2121
212225. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2123 properly. Recipient callout cache records were still being keyed to include
2124 the sender, even when use_sender was set false. This led to far more
2125 callouts that were necessary. The sender is no longer included in the key
2126 when use_sender is false.
2127
212826. Added "control = submission" modifier to ACLs.
2129
213027. Added the ${base62d: operator to decode base 62 numbers.
2131
213228. dnsdb lookups can now access SRV records.
2133
213429. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2135 the configuration file.
2136
213730. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2138 (-v) mode. This makes the output for a verbose queue run more intelligible.
2139
214031. Added a use_postmaster feature to recipient callouts.
2141
214232. Added the $body_zerocount variable, containing the number of binary zero
2143 bytes in the message body.
2144
214533. The time of last modification of the "new" subdirectory is now used as the
2146 "mailbox time last read" when there is a quota error for a maildir
2147 delivery.
2148
214934. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2150
215135. Added +ignore_unknown as a special item in host lists.
2152
215336. Code for decoding IPv6 addresses in host lists is now included, even if
2154 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2155 address was recognized as an IP address, but was then not correctly decoded
2156 into binary, causing unexpected and incorrect effects when compared with
2157 another IP address.
2158
2159
2160Exim version 4.34
2161-----------------
2162
2163 1. Very minor rewording of debugging text in manualroute to say "list of
2164 hosts" instead of "hostlist".
2165
2166 2. If verify=header_syntax was set, and a header line with an unqualified
2167 address (no domain) and a large number of spaces between the end of the
2168 name and the colon was received, the reception process suffered a buffer
2169 overflow, and (when I tested it) crashed. This was caused by some obsolete
2170 code that should have been removed. The fix is to remove it!
2171
2172 3. When running in the test harness, delay a bit after writing a bounce
2173 message to get a bit more predictability in the log output.
2174
2175 4. Added a call to search_tidyup() just before forking a reception process. In
2176 theory, someone could use a lookup in the expansion of smtp_accept_max_
2177 per_host which, without the tidyup, could leave open a database connection.
2178
2179 5. Added the variables $recipient_data and $sender_data which get set from a
2180 lookup success in an ACL "recipients" or "senders" condition, or a router
2181 "senders" option, similar to $domain_data and $local_part_data.
2182
2183 6. Moved the writing of debug_print from before to after the "senders" test
2184 for routers.
2185
2186 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2187 problems for message scanning, either using a data ACL, or using
2188 local_scan() because the Received: header was not generated till after they
2189 were called (in order to set the time as the time of reception completion).
2190 I have revised the way this works. The header is now generated after the
2191 body is received, but before the ACL or local_scan() are called. After they
2192 are run, the timestamp in the header is updated.
2193
2194
2195Exim version 4.33
2196-----------------
2197
2198 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2199 before starting a queue runner without re-exec. This happened only when
2200 deliver_drop_privilege was set or when the Exim user was set to root. The
2201 effect of the bug was that timeouts during subsequent deliveries caused
2202 crashes instead of being properly handled. The handler is now left at its
2203 default (and expected) setting.
2204
2205 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2206 message, again when deliver_drop_privilege is set or Exim is run as root.
2207 The bug described in (1) was not present in this case, but the tidying up
2208 of the other signals was missing. I have made the two cases consistent.
2209
2210 3. The ignore_target_hosts setting on a manualroute router was being ignored
2211 for hosts that were looked up using the /MX notation.
2212
2213 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2214 in domain lists.
2215
2216 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2217 operated on the sender address. After changing the $sender_address to <>
2218 for the sender address verify, Exim was re-instated it as the original
2219 (before rewriting) address, but remembering that it had rewritten it, so it
2220 wasn't rewriting it again. This bug also had the effect of breaking the
2221 sender address verification caching when the sender address was rewritten.
2222
2223 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2224 This has been changed so that if the ip literal address matches
2225 ignore_target_hosts, the router declines.
2226
2227 7. Added expansion conditions match_domain, match_address, and match_local_
2228 part (NOT match_host).
2229
2230 8. The placeholder for the Received: header didn't have a length field set.
2231
2232 9. Added code to Exim itself and to exim_lock to test for a specific race
2233 condition that could lead to file corruption when using MBX delivery. The
2234 issue is with the lockfile that is created in /tmp. If this file is removed
2235 after a process has opened it but before that process has acquired a lock,
2236 there is the potential for a second process to recreate the file and also
2237 acquire a lock. This could lead to two Exim processes writing to the file
2238 at the same time. The added code performs the same test as UW imapd; it
2239 checks after acquiring the lock that its file descriptor still refers to
2240 the same named file.
2241
224210. The buffer for building added header lines was of fixed size, 8192 bytes.
2243 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2244 when Exim is built.
2245
224611. Added the smtp_active_hostname option. If used, this will typically be made
2247 to depend on the incoming interface address. Because $interface_address is
2248 not set up until the daemon has forked a reception process, error responses
2249 that can happen earlier (such as "too many connections") no longer contain
2250 a host name.
2251
225212. If an expansion in a condition on a "warn" statement fails because a lookup
2253 defers, the "warn" statement is abandoned, and the next ACL statement is
2254 processed. Previously this caused the whole ACL to be aborted.
2255
225613. Added the iplsearch lookup type.
2257
225814. Added ident_timeout as a log selector.
2259
226015. Added tls_certificate_verified as a log selector.
2261
226216. Added a global option tls_require_ciphers (compare the smtp transport
2263 option of the same name). This controls incoming TLS connections.
2264
226517. I finally figured out how to make tls_require_ciphers do a similar thing
2266 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2267 before starting the TLS session.
2268
226918. Tabs are now shown as \t in -bP output.
2270
227119. If the log selector return_path_on_delivery was set, Exim crashed when
2272 bouncing a message because it had too many Received: header lines.
2273
227420. If two routers both had headers_remove settings, and the first one included
2275 a superfluous trailing colon, the final name in the first list and the
2276 first name in the second list were incorrectly joined into one item (with a
2277 colon in the middle).
2278
2279
2280Exim version 4.32
2281-----------------
2282
2283 1. Added -C and -D options to the exinext utility, mainly to make it easier
2284 to include in the automated testing, but these could be helpful when
2285 multiple configurations are in use.
2286
2287 2. The exinext utility was not formatting the output nicely when there was
2288 an alternate port involved in the retry record key, nor when there was a
2289 message id as well (for retries that were specific to a specific message
2290 and a specific host). It was also confused by IPv6 addresses, because of
2291 the additional colons they contain. I have fixed the IPv4 problem, and
2292 patched it up to do a reasonable job for IPv6.
2293
2294 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2295 delivery, the log line now contains "pipelined" if PIPELINING was used.
2296
2297 4. An SMTP transport process used to panic and die if the bind() call to set
2298 an explicit outgoing interface failed. This has been changed; it is now
2299 treated in the same way as a connect() failure.
2300
2301 5. A reference to $sender_host_name in the part of a conditional expansion
2302 that was being skipped was still causing a DNS lookup. This no longer
2303 occurs.
2304
2305 6. The def: expansion condition was not recognizing references to header lines
2306 that used bh_ and bheader_.
2307
2308 7. Added the _cache feature to named lists.
2309
2310 8. The code for checking quota_filecount in the appendfile transport was
2311 allowing one more file than it should have been.
2312
2313 9. For compatibility with Sendmail, the command line option
2314
2315 -prval:sval
2316
2317 is equivalent to
2318
2319 -oMr rval -oMs sval
2320
2321 and sets the incoming protocol and host name (for trusted callers). The
2322 host name and its colon can be omitted when only the protocol is to be set.
2323 Note the Exim already has two private options, -pd and -ps, that refer to
2324 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2325 "s", but I don't think that's a major issue.
2326
232710. A number of refactoring changes to the code, none of which should affect
2328 Exim's behaviour:
2329
2330 (a) The number of logging options was getting close to filling up the
2331 32-bit word that was used as a bit map. I have split them into two classes:
2332 those that are passed in the argument to log_write(), and those that are
2333 only ever tested independently outside of that function. These are now in
2334 separate 32-bit words, so there is plenty of room for expansion again.
2335 There is no change in the user interface or the logging behaviour.
2336
2337 (b) When building, for example, log lines, the code previously used a
2338 macro that called string_cat() twice, in order to add two strings. This is
2339 not really sufficiently general. Furthermore, there was one instance where
2340 it was actually wrong because one of the argument was used twice, and in
2341 one call a function was used. (As it happened, calling the function twice
2342 did not affect the overall behaviour.) The macro has been replaced by a
2343 function that can join an arbitrary number of extra strings onto a growing
2344 string.
2345
2346 (c) The code for expansion conditions now uses a table and a binary chop
2347 instead of a serial search (which was left over from when there were very
2348 few conditions). Also, it now recognizes conditions like "pam" even when
2349 the relevant support is not compiled in: a suitably worded error message is
2350 given if an attempt is made to use such a condition.
2351
235211. Added ${time_interval:xxxxx}.
2353
235412. A bug was causing one of the ddress fields not to be passed back correctly
2355 from remote delivery subprocesses. The field in question was not being
2356 subsequently used, so this caused to problems in practice.
2357
235813. Added new log selectors queue_time and deliver_time.
2359
236014. Might have fixed a bug in maildirsizefile handling that threw up
2361 "unexpected character" debug warnings, and recalculated the data
2362 unnecessarily. In any case, I expanded the warning message to give more
2363 information.
2364
236515. Added the message "Restricted characters in address" to the statements in
2366 the default ACL that block characters like @ and % in local parts.
2367
236816. Change 71 for release 4.31 proved to be much less benign that I imagined.
2369 Three changes have been made:
2370
2371 (a) There was a serious bug; a negative response to MAIL caused the whole
2372 recipient domain to be cached as invalid, thereby blocking all messages
2373 to all local parts at the same domain, from all senders. This bug has
2374 been fixed. The domain is no longer cached after a negative response to
2375 MAIL if the sender used is not empty.
2376
2377 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2378 been restored.
2379
2380 (c) A new callout option, "use_sender" has been added for people who want
2381 the modified behaviour.
2382
2383
2384Exim version 4.31
2385-----------------
2386
2387 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2388 Larry Rosenman.
2389
2390 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2391 indeed breaks things for older releases.
2392
2393 3. Added additional logging to the case where there is a problem reading data
2394 from a filter that is running in a subprocess using a pipe, in order to
2395 try to track down a specific problem.
2396
2397 4. Testing facility fudge: when running in the test harness and attempting
2398 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2399 getting "No route to host". Convert this to a timeout.
2400
2401 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2402 warning.
2403
2404 6. Some OS don't have socklen_t but use size_t instead. This affects the
2405 fifth argument of getsockopt() amongst other things. This is now
2406 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2407 can be set for individual OS. I have set it for SunOS5, OSF1, and
2408 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2409 some earlier ones do not.
2410
2411 7. Change 4.30/15 was not doing the test caselessly.
2412
2413 8. The standard form for an IPv6 address literal was being rejected by address
2414 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2415 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2416 this, as well as the form without the "IPv6" on the front (but only when
2417 address literals are enabled, of course).
2418
2419 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2420
242110. Exim crashed if a message with an empty sender address specified by -f
2422 encountered a router with an errors_to setting. This could be provoked only
2423 by a command such as
2424
2425 exim -f "" ...
2426
2427 where an empty string was supplied; "<>" did not hit this bug.
2428
242911. Installed PCRE release 4.5.
2430
243112. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2432 remained set. It is now erased.
2433
243413. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2435 times from message ids (which are base 36 rather than the normal 62).
2436
243714. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2438 were being counted as actual protocol errors, and logged if the log
2439 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2440 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2441 and DATA following a set of rejected RCPTs do not count as protocol errors.
2442 In other words, Exim assumes they were pipelined, though this may not
2443 actually be the case. Of course, in all cases the client gets an
2444 appropriate error code.
2445
244615. If a lookup fails in an ACL condition, a message about the failure may
2447 be available; it is used if testing the ACL cannot continue, because most
2448 such messages specify what the cause of the deferral is. However, some
2449 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2450 that caused an old message to be retained and used if a later statement
2451 caused a defer, replacing the real cause of the deferral.
2452
245316. If an IP address had so many PTR records that the DNS lookup buffer
2454 was not large enough to hold them, Exim could crash while trying to process
2455 the truncated data. It now detects and logs this case.
2456
245717. Further to 4.21/58, another change has been made: if (and only if) the
2458 first line of a message (the first header line) ends with CRLF, a bare LF
2459 in a subsequent header line has a space inserted after it, so as not to
2460 terminate the header.
2461
246218. Refactoring: tidied an ugly bit of code in appendfile that copied data
2463 unnecessarily, used atoi() instead of strtol(), and didn't check the
2464 termination when getting file sizes from file names by regex.
2465
246619. Completely re-implemented the support for maildirsize files, in the light
2467 of a number of problems with the previous contributed implementation
2468 (4.30/29). In particular:
2469
2470 . If the quota is zero, the maildirsize file is maintained, but no quota is
2471 imposed.
2472
2473 . If the maildir directory does not exist, it is created before any attempt
2474 to write a maildirsize file.
2475
2476 . The quota value in the file is just a cache; if the quota is changed in
2477 the transport, the new value overrides.
2478
2479 . A regular expression is available for excluding directories from the
2480 count.
2481
248220. The autoreply transport checks the characters in options that define the
2483 message's headers; it allows continued headers, but it was checking with
2484 isspace() after an embedded newline instead of explicitly looking for a
2485 space or a tab.
2486
248721. If all the "regular" hosts to which an address was routed had passed their
2488 expiry times, and had not reached their retry times, the address was
2489 bounced, even if fallback hosts were defined. Now Exim should go on to try
2490 the fallback hosts.
2491
249222. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2493 equivalent code in the SMTP transport. Some hosts send humungous responses
2494 to HELO/EHLO, more than 1024 it seems.
2495
249623. Refactoring: code in filter.c used (void *) for "any old type" but this
2497 gives compiler warnings in some environments. I've now done it "properly",
2498 using a union.
2499
250024. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2501 (because of problems with the built-in one) was declared to return uschar *
2502 instead of char *, causing compiler failure.
2503
250425. Fixed a file descriptor leak when processing alias/forward files.
2505
250626. Fixed a minor format string issue in dbfn.c.
2507
250827. Typo in exim.c: ("dmbnz" for "dbmnz").
2509
251028. If a filter file refered to $h_xxx or $message_headers, and the headers
2511 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2512 become corrupted.
2513
251429. When a sender address is verified, it is cached, to save repeating the test
2515 when there is more than one recipient in a message. However, when the
2516 verification involves a callout, it is possible for different callout
2517 options to be set for different recipients. It is too complicated to keep
2518 track of this in the cache, so now Exim always runs a verification when a
2519 callout is required, relying on the callout cache for the optimization.
2520 The overhead is duplication of the address routing, but this should not be
2521 too great.
2522
252330. Fixed a bug in callout caching. If a RCPT command caused the sender address
2524 to be verified with callout=postmaster, and the main callout worked but the
2525 postmaster check failed, the verification correctly failed. However, if a
2526 subsequent RCPT command asked for sender verification *without* the
2527 postmaster check, incorrect caching caused this verification also to fail,
2528 incorrectly.
2529
253031. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2531 it was not caching the DNS options (qualify_single, search_parents) that
2532 were used when the lookup failed. A subsequent lookup with different
2533 options therefore always gave the same answer, though there were cases
2534 where it should not have. (Example: a "domains = !$mx_any" option on a
2535 dnslookup router: the "domains" option is always processed without any
2536 widening, but the router might have qualify_single set.) Now Exim uses the
2537 cached value only when the same options are set.
2538
253932. Added John Jetmore's "exipick" utility to the distribution.
2540
254133. GnuTLS: When an attempt to start a TLS session fails for any reason other
2542 than a timeout (e.g. a certificate is required, and is not provided), an
2543 Exim server now closes the connection immediately. Previously it waited for
2544 the client to close - but if the client is SSL, it seems that they each
2545 wait for each other, leading to a delay before one of them times out.
2546
254734: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2548 maintained 0.8.x compatibility because I don't think many are using it, and
2549 it is clearly obsolete.
2550
255135. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2552 transport.
2553
255436. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2555 client certificate was expired. A simple patch fixes this, though I don't
2556 understand the full logic of why the verify callback is called multiple
2557 times.
2558
255937. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2560 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2561 which causes problems with some clients (such as the Certicom SSL Plus
2562 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2563 disables the coutermeasure allowing Eudora to connect."
2564
256538. Exim was not checking that a write() to a log file succeeded. This could
2566 lead to Bad Things if a log got too big, in particular if it hit a file
2567 size limit. Exim now panics and dies if it cannot write to a log file, just
2568 as it does if it cannot open a log file.
2569
257039. Modified OS/Makefile-Linux so that it now contains
2571
2572 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2573
2574 The two -D definitions ensure that Exim is compiled with large file
2575 support, which makes it possible to handle log files that are bigger than
2576 2^31.
2577
257840. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2579 instance) a domain was checked against a named list that involved a lookup,
2580 causing $domain_data to be set, then another domain was checked against the
2581 same list, then the first domain was re-checked, the value of $domain_data
2582 after the final check could be wrong. In particular, if the second check
2583 failed, it could be set empty. This bug probably also applied to
2584 $localpart_data.
2585
258641. The strip_trailing_dot option was not being applied to the address given
2587 with the -f command-line option.
2588
258942. The code for reading a message's header from the spool was incrementing
2590 $received_count, but never initializing it. This meant that the value was
2591 incorrect (doubled) while delivering a message in the same process in which
2592 it was received. In the most common configuration of Exim, this never
2593 happens - a fresh exec is done - but it can happen when
2594 deliver_drop_privilege is set.
2595
259643. When Exim logs an SMTP synchronization error - client data sent too soon -
2597 it now includes up to 150 characters of the unexpected data in the log
2598 line.
2599
260044. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2601 and building data strings. The size of both of these buffers was 10 000
2602 bytes - far larger than anybody would *ever* want, thought I. Needless to
2603 say, somebody hit the limit. I have increased the maximum line length to
2604 20 000 and the maximum data length of concatenated lines to 100 000. I have
2605 also fixed two bugs, because there was no checking on these buffers. Tsk,
2606 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2607 buffer is too small.
2608
260945. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2610 lsearch lookups. Now it does.
2611
261246. When parsing a route_list item in a manualroute router, a fixed-length
2613 buffer was used for the list of hosts. I made this 1024 bytes long,
2614 thinking that nobody would ever have a list of hosts that long. Wrong.
2615 Somebody had a whole pile of complicated expansion conditions, and the
2616 string was silently truncated, leading to an expansion error. It turns out
2617 that it is easier to change to an unlimited length (owing to other changes
2618 that have happened since this code was originally written) than to build
2619 structure for giving a limitation error. The length of the item that
2620 expands into the list of hosts is now unlimited.
2621
262247. The lsearch lookup could not handle data where the length of text line was
2623 more than 4095 characters. Such lines were truncated, leading to shortened
2624 data being returned. It should now handle lines of any length.
2625
262648. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2627 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2628 ACL").
2629
263049. Cosmetic tidy to scripts like exicyclog that are generated by globally
2631 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2632 no longer happens in comment lines. A list of replacements is now placed
2633 at the head of all of the source files, except those whose only change is
2634 to replace PERL_COMMAND in the very first #! line.
2635
263650. Replaced the slow insertion sort in queue.c, for sorting the list of
2637 messages on the queue, with a bottom-up merge sort, using code contributed
2638 by Michael Haardt. This should make operations like -bp somewhat faster on
2639 large queues. It won't affect queue runners, except when queue_run_in_order
2640 is set.
2641
264251. Installed eximstats 1.31 in the distribution.
2643
264452. Added support for SRV lookups to the dnslookup router.
2645
264653. If an ACL referred to $message_body or $message_body_end, the value was not
2647 reset for any messages that followed in the same SMTP session.
2648
264954. The store-handling optimization for building very long strings was not
2650 differentiating between the different store pools. I don't think this
2651 actually made any difference in practice, but I've tidied it.
2652
265355. While running the routers to verify a sender address, $sender_address
2654 was still set to the sender address. This is wrong, because when routing to
2655 send a bounce to the sender, it would be empty. Therefore, I have changed
2656 it so that, while verifying a sender address, $sender_address is set to <>.
2657 (There is no change to what happens when verifying a recipient address.)
2658
265956. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2660 target A or AAAA records (if not already returned) without resetting the
2661 qualify_single or search_parents options of the DNS resolver. These are
2662 inappropriate in this case because the targets of MX and SRV records must
2663 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2664 target that, when qualified, matched something in the local domain. These
2665 two options are now turned off when doing these lookups.
2666
266757. It seems that at least some releases of Reiserfs (which does not have the
2668 concept of a fixed number of inodes) returns zero and not -1 for the
2669 number of available inodes. This interacted badly with check_spool_inodes,
2670 which assumed that -1 was the "no such thing" setting. What I have done is
2671 to check that the total number of inodes is greater than zero before doing
2672 the test of how many are available.
2673
267458. When a "warn" ACL statement has a log_message modifier, the message is
2675 remembered, and not repeated. This is to avoid a lot of repetition when a
2676 message has many recipients that cause the same warning to be written.
2677 Howewer, Exim was preserving the list of already written lines for an
2678 entire SMTP session, which doesn't seem right. The memory is now reset if a
2679 new message is started.
2680
268159. The "rewrite" debugging flag was not showing the result of rewriting in the
2682 debugging output unless log_rewrite was also set.
2683
268460. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2685 of (int)(handle) when we know that handle contains (void *)(-1).
2686
268761. The Exim daemon panic-logs an error return when it closes the incoming
2688 connection. However "connection reset by peer" seems to be common, and
2689 isn't really an error worthy of noting specially, so that particular error
2690 is no long logged.
2691
269262. When Exim is trying to find all the local interfaces, it used to panic and
2693 die if the ioctl to get the interface flags failed. However, it seems that
2694 on at least one OS (Solaris 9) it is possible to have an interface that is
2695 included in the list of interfaces, but for which you get a failure error
2696 for this call. This happens when the interface is not "plumbed" into a
2697 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2698 failure of the "get flags" call assumes that the interface is down.
2699
270063. Added a ${eval10: operator, which assumes all numbers are decimal. This
2701 makes life easier for people who are doing arithmetic on fields extracted
2702 from dates, where you often get leading zeros that should not be
2703 interpreted as octal.
2704
270564. Added qualify_domain to the redirect router, to override the global
2706 setting.
2707
270865. If a pathologically long header line contained very many addresses (the
2709 report of this problem mentioned 10 000) and each of them was rewritten,
2710 Exim could use up a very large amount of memory. (It kept on making new
2711 copies of the header line as it rewrote, and never released the old ones.)
2712 At the expense of a bit more processing, the header rewriting function has
2713 been changed so that it no longer eats memory in this way.
2714
271566. The generation of the Received: header has been moved from the time that a
2716 message starts to be received, to the time that it finishes. The timestamp
2717 in the Received: header should now be very close to that of the <= log
2718 line. There are two side-effects of this change:
2719
2720 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2721 logged header lines no longer include the local Received: line, because
2722 it has not yet been created. The same applies to a copy of the message
2723 that is returned to a non-SMTP sender when a message is rejected.
2724
2725 (b) When a filter file is tested using -bf, no additional Received: header
2726 is added to the test message. After some thought, I decided that this
2727 is a bug fix.
2728
2729 This change does not affect the value of $received_for. It is still set
2730 after address rewriting, but before local_scan() is called.
2731
273267. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2733
273468. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2735 gave an unhelpful panic error message, and a defer error. I have managed to
2736 change this behaviour so that it now rejects any supplied certificate,
2737 which seems right, as the list of acceptable certificates is empty.
2738
273969. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2740 gave an unhelpful defer error. I have not managed to make this reject any
2741 supplied certificates, but the error message it gives is "no certificate
2742 supplied", which is not helpful.
2743
274470. exigrep's output now also includes lines that are not associated with any
2745 message, but which match the given pattern. Implemented by a patch from
2746 Martin Sluka, which also tidied up the Perl a bit.
2747
274871. Recipient callout verification, like sender verification, was using <> in
2749 the MAIL FROM command. This isn't really the right thing, since the actual
2750 sender may affect whether the remote host accepts the recipient or not. I
2751 have changed it to use the actual sender in the callout; this means that
2752 the cache record is now keyed on a recipient/sender pair, not just the
2753 recipient address. There doesn't seem to be a real danger of callout loops,
2754 since a callout by the remote host to check the sender would use <>.
2755 [SEE ABOVE: changed after hitting problems.]
2756
275772. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2758 temporary errors. However, in the case of such a code being given after
2759 the end of a data transmission (i.e. after ".") Exim was failing to write
2760 a retry record for the message. (Yes, there was some broken host that was
2761 actually sending 8xx at this point.)
2762
276373. An unknown lookup type in a host list could cause Exim to panic-die when
2764 the list was checked. (An example that provoked this was putting <; in the
2765 middle of a list instead of at the start.) If this happened during a DATA
2766 ACL check, a -D file could be left lying around. This kind of configuration
2767 error no longer causes Exim to die; instead it causes a defer errror. The
2768 incident is still logged to the main and panic logs.
2769
277074. Buglet left over from Exim 3 conversion. The message "too many messages
2771 in one connection" was written to the rejectlog but not the mainlog, except
2772 when address rewriting (yes!) was being logged.
2773
277475. Added write_rejectlog option.
2775
277676. When a system filter was run not as root (that is, when system_filter_user
2777 was set), the values of the $n variables were not being returned to the
2778 main process; thus, they were not subsequently available in the $sn
2779 variables.
2780
278177. Added +return_path_on_delivery log selector.
2782
278378. A connection timeout was being treated differently from recipients deferred
2784 when testing hosts_max_try with a message that was older than the host's
2785 retry timeout. (The host should not be counted, thus allowing all hosts to
2786 be tried at least once before bouncing.) This may have been the cause of an
2787 occasionally reported bug whereby a message would remain on the queue
2788 longer than the retry timeout, but would be bounced if a delivery was
2789 forced. I say "may" because I never totally pinned down the problem;
2790 setting up timeout/retry tests is difficult. See also the next item.
2791
279279. The ultimate address timeout was not being applied to errors that involved
2793 a combination of host plus message (for example, a timeout on a MAIL
2794 command). When an address resolved to a number of possible hosts, and they
2795 were not all tried for each delivery (e.g. because of hosts_max_try), a
2796 message could remain on the queue longer than the retry timeout.
2797
279880. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2799 Haardt.
2800
280181. Fixed an obscure SMTP outgoing bug which required at least the following
2802 conditions: (a) there was another message waiting for the same server;
2803 (b) the server returned 5xx to all RCPT commands in the first message so
2804 that the message was not completed; (c) the server dropped the connection
2805 or gave a negative response to the RSET that Exim sends to abort the
2806 transaction. The observed case was a dropped connection after DATA that had
2807 been sent in pipelining mode. That is, the server had advertised PIPELINING
2808 but was not implementing it correctly. The effect of the bug was incorrect
2809 behaviour, such as trying another host, and this could lead to a crash.
2810
2811
2812Exim version 4.30
2813-----------------
2814
2815 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2816 and daemon.c were passed as pointers to ints; they should have been
2817 pointers to socklen_t variables (which are typically unsigned ints).
2818
2819 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2820 fixed.
2821
2822 3. Fixed a really odd bug that affected only the testing scheme; patching a
2823 certain fixed string in the binary changed the value of another string that
2824 happened to be identical to the end of the original first string.
2825
2826 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2827 name", it returns that address as the IP address. On some operating
2828 systems (e.g. Solaris), it also passes back the IP address string as the
2829 "host name". However, on others (e.g. Linux), it passes back an empty
2830 string. Exim wasn't checking for this, and was changing the host name to an
2831 empty string, assuming it had been canonicized.
2832
2833 5. Although rare, it is permitted to have more than one PTR record for a given
2834 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2835 all the names associated with an address, because they do in Solaris.
2836 However, it seems that they do not in Linux for data that comes from the
2837 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2838 I found this out when I moved to a new Linux workstation and tried to run
2839 the Exim test suite.
2840
2841 To get round this problem I have changed the code so that it now does its
2842 own call to the DNS to look up PTR records when searching for a host name.
2843 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2844 addresses that are only in /etc/hosts are still found.
2845
2846 This behaviour is, however, controlled by an option called host_lookup_
2847 order, which defaults to "bydns:byaddr". If people want to use the other
2848 order, or indeed, just use one or the other means of lookup, they can
2849 specify it in this variable.
2850
2851 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2852 some operating systems, this comes back from gethostbyaddr() as an empty
2853 string, and this is what Exim used to test for. However, it seems that in
2854 other systems, "." is yielded. Exim now tests for this case too.
2855
2856 7. The values of check_spool_space and check_log_space are now held internally
2857 as a number of kilobytes instead of an absolute number of bytes. If a
2858 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2859 kilobyte. This means that much larger values can be stored.
2860
2861 8. Exim monitor: an attempt to get the action menu when not actually pointing
2862 at a message produces an empty menu entitled "No message selected". This
2863 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2864 no entries in it ("Shell widget menu has zero width and/or height"). So I
2865 have added a single, blank menu entry in this case.
2866
2867 9. Added ${quote_local_part.
2868
286910. MIME decoding is now applied to the contents of Subject: header lines when
2870 they are logged.
2871
287211. Now that a reference to $sender_host_address automatically causes a reverse
2873 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2874 host lookup before query-style lookups in lists that might use this
2875 variable. This has therefore been abolished, and the "net-" prefix is no
2876 longer necessary for query-style lookups.
2877
287812. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2879 have been a typo for LFLAGS, so it has been changed.
2880
288113. The install script calls Exim with "-C /dev/null" in order to find the
2882 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2883 to be output. Howeve, since Exim outputs its version number before the
2884 error, it didn't break the script. It just looked ugly. I fixed this by
2885 always allowing "-C /dev/null" if the caller is root.
2886
288714. Ignore overlarge ACL variable number when reading spool file - insurance
2888 against a later release with more variables having written the file.
2889
289015. The standard form for an IPv6 address literal was being rejected by EHLO.
2891 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2892 this, as well as the form without the "IPv6" on the front.
2893
289416. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2895 OS/Makefile-Darwin file.
2896
289717. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2898 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2899
290018. After change 4.21/52, "%ld" was used to format the contents of the $inode
2901 variable. However, some OS use ints for inodes. I've added cast to long int
2902 to get rid of the compiler warning.
2903
290419. I had forgotten to lock out "/../" in configuration file names when
2905 ALT_CONFIG_PREFIX was set.
2906
290720. Routers used for verification do not need to specify transports. However,
2908 if such a router generated a host list, and callout was configured, Exim
2909 crashed, because it could not find a port number from the (non-existent)
2910 transport. It now assumes port 25 in this circumstance.
2911
291221. Added the -t option to exigrep.
2913
291422. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2915 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2916 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2917 them is set, LOOKUP_LSEARCH is forced.
2918
291923. "exim -bV" now outputs a list of lookups that are included in the binary.
2920
292124. Added sender and host information to the "rejected by local_scan()" log
2922 line; previously there was no indication of these.
2923
292425. Added .include_if_exists.
2925
292626. Change 3.952/11 added an explicit directory sync on top of a file sync for
2927 Linux. It turns out that not all file systems support this. Apparently some
2928 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2929 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2930 not supported on the file descriptor, is now ignored when Exim is trying to
2931 sync a directory. This applies only to Linux.
2932
293327. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2934
293528. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2936 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2937 this.
2938
293929. Added support for maildirsize files from supplied patch (modified a bit).
2940
294130. The use of :fail: followed by an empty string could lead Exim to respond to
2942 sender verification failures with (e.g.):
2943
2944 550 Verification failed for <xxx>
2945 550 Sender verify failed
2946
2947 where the first response line was missing the '-' that indicates it is not
2948 the final line of the response.
2949
295031. The loop for finding the name of the user that called Exim had a hardwired
2951 limit of 10; it now uses the value of finduser_retries, which is used for
2952 all other user lookups.
2953
295432. Added $received_count variable, available in data and not_smtp ACLs, and at
2955 delivery time.
2956
295733. Exim was neglecting to zero errno before one call of strtol() when
2958 expanding a string and expecting an integer value. On some systems this
2959 resulted in spurious "integer overflow" errors. Also, it was casting the
2960 result into an int without checking.
2961
296234. Testing for a connection timeout using "timeout_connect" in the retry rules
2963 did not work. The code looks as if it has *never* worked, though it appears
2964 to have been documented since at least releast 1.62. I have made it work.
2965
296635. The "timeout_DNS" error in retry rules, also documented since at least
2967 1.62, also never worked. As it isn't clear exactly what this means, and
2968 clearly it isn't a major issue, I have abolished the feature by treating it
2969 as "timeout", and writing a warning to the main and panic logs.
2970
297136. The display of retry rules for -brt wasn't always showing the error code
2972 correctly.
2973
297437. Added new error conditions to retry rules: timeout_A, timeout_MX,
2975 timeout_connect_A, timeout_connect_MX.
2976
297738. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2978 to the empty sender.
2979
298039. The daemon was not analysing the content of -oX till after it had closed
2981 stderr and disconnected from the controlling terminal. This meant that any
2982 syntax errors were only noted on the panic log, and the return code from
2983 the command was 0. By re-arranging the code a little, I've made the
2984 decoding happen first, so such errors now appear on stderr, and the return
2985 code is 1. However, the actual setting up of the sockets still happens in
2986 the disconnected process, so errors there are still only recorded on the
2987 panic log.
2988
298940. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2990 connections (as happens on some IP stacks) was logged at start up time as
2991 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2992 it from "IPv6 and IPv4", which means that two separate sockets are being
2993 used.
2994
299541. The debug output for gethostbyname2() or getipnodebyname() failures now
2996 says whether AF_INET or AF_INET6 was passed as an argument.
2997
299842. Exiwhat output was messed up when time zones were included in log
2999 timestamps.
3000
300143. Exiwhat now gives more information about the daemon's listening ports,
3002 and whether -tls-on-connect was used.
3003
300444. The "port" option of the smtp transport is now expanded.
3005
300645. A "message" modifier in a "warn" statement in a non-message ACL was being
3007 silently ignored. Now an error message is written to the main and panic
3008 logs.
3009
301046. There's a new ACL modifier called "logwrite" which writes to a log file
3011 as soon as it is encountered.
3012
301347. Added $local_user_uid and $local_user_gid at routing time.
3014
301548. Exim crashed when trying to verify a sender address that was being
3016 rewritten to "<>".
3017
301849. Exim was recognizing only a space character after ".include". It now also
3019 recognizes a tab character.
3020
302150. Fixed several bugs in the Perl script that creates the exim.8 man page by
3022 extracting the relevant information from the specification. The man page no
3023 longer contains scrambled data for the -d option, and I've added a section
3024 at the front about calling Exim under different names.
3025
302651. Added "extra_headers" argument to the "mail" command in filter files.
3027
302852. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3029 crash.
3030
303153. Installed eximstats 1.29.
3032
303354. Added transport_filter_timeout as a generic transport option.
3034
303555. Exim no longer adds an empty Bcc: header to messages that have no To: or
3036 Cc: header lines. This was required by RFC 822, but it not required by RFC
3037 2822.
3038
303956. Exim used to add From:, Date:, and Message-Id: header lines to any
3040 incoming messages that did not have them. Now it does so only if the
3041 message originates locally, that is, if there is no associated remote host
3042 address. When Resent- header lines are present, this applies to the Resent-
3043 lines rather than the non-Resent- lines.
3044
304557. Drop incoming SMTP connection after too many syntax or protocol errors. The
3046 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3047
304858. Messages for configuration errors now include the name of the main
3049 configuration file - useful now that there may be more than one file in a
3050 list (.included file names were always shown).
3051
305259. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3053 for those rare installations that do not start the daemon as root or run it
3054 setuid root. I've cut out the call to initgroups() if the daemon is not
3055 root at that time.
3056
305760. The Exim user and group can now be bound into the binary as text strings
3058 that are looked up at the start of Exim's processing.
3059
306061. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3061
306262. Added $mailstore_basename variable.
3063
306463. Installed patch to sieve.c from Michael Haardt.
3065
306664. When Exim failed to open the panic log after failing to open the main log,
3067 the original message it was trying to log was written to stderr and debug
3068 output, but if they were not available (the usual case in production), it
3069 was lost. Now it is written to syslog before the two lines that record the
3070 failures to open the logs.
3071
307265. Users' Exim filters run in subprocesses under the user's uid. It is
3073 possible for a "deliver" command or an alias in a "personal" command to
3074 provoke an address rewrite. If logging of address rewriting is configured,
3075 this fails because the process is not running as root or exim. There may be
3076 a better way of dealing with this, but for the moment (because 4.30 needs
3077 to be released), I have disabled address rewrite logging when running a
3078 filter in a non-root, non-exim process.
3079
3080
3081Exim version 4.24
3082-----------------
3083
3084 1. The buildconfig auxiliary program wasn't quoting the value set for
3085 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3086 not defined. This bug was masked in 4.22 by the effect that was fixed in
3087 change 4.23/1.
3088
3089 2. Some messages that were rejected after a message id was allocated were
3090 shown as "incomplete" by exigrep. It no longer does this for messages that
3091 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3092
3093 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3094 have allow_domain_literals set, the ID did not get logged in the <= line.
3095 Domain literals are now always recognized in Message-ID: header lines.
3096
3097 4. The first argument for a ${extract expansion item is the key name or field
3098 number. Leading and trailing spaces in this item were not being ignored,
3099 causing some misleading effects.
3100
3101 5. When deliver_drop_privilege was set, single queue runner processes started
3102 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3103 same command in the process it spins off) were not dropping privilege.
3104
3105 6. When the daemon running as "exim" started a queue runner, it always
3106 re-executed Exim in the spun-off process. This is a waste of effort when
3107 deliver_drop_privilege is set. The new process now just calls the
3108 queue-runner function directly.
3109
3110
3111Exim version 4.23
3112-----------------
3113
3114 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3115 HEADERS_CHARSET.
3116
3117 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3118 ignored. Though the use of -oP was forcing the writing of a pid file, it
3119 was always written to the default place.
3120
3121 3. If the message "no IP address found for host xxxx" is generated during
3122 incoming verification, it is now followed by identification of the incoming
3123 connection (so you can more easily find what provoked it).
3124
3125 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3126
3127 5. Added some features to "harden" Exim a bit more against certain attacks:
3128
3129 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3130 be put in Local/Makefile. This is like the never_users runtime option,
3131 but it cannot be overridden. The default setting is "root".
3132
3133 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3134 prefix string with which any file named in a -C command line option
3135 must start.
3136
3137 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3138 is retained for -C and -D only if the caller of Exim is root. Without
3139 it, the exim user may also use -C and -D and retain privilege.
3140
3141 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3142 command line option is disabled.
3143
3144 6. Macro names set by the -D option must start with an upper case letter, just
3145 like macro names defined in the configuration file.
3146
3147 7. Added "dereference=" facility to LDAP.
3148
3149 8. Two instances of the typo "uknown" in the source files are fixed.
3150
3151 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3152 the Configure-Makefile script screwed up while processing it.
3153
315410. Incorporated PCRE 4.4.
3155
315611. The SMTP synchronization check was not operating right at the start of an
3157 SMTP session. For example, it could not catch a HELO sent before the client
3158 waited for the greeting. There is now a check for outstanding input at the
3159 point when the greeting is written. Because of the duplex, asynchronous
3160 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3161 way, but not yet received, when the check is performed.
3162
316312. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3164 on TCP/IP sockets, because this apparently causes some broken clients to
3165 timeout.
3166
316713. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3168 unchanged) from the Cygwin maintainer.
3169
317014. The code for -bV that shows what is in the binary showed "mbx" when maildir
3171 was supported instead of testing for mbx. Effectively a typo.
3172
317315. The spa authenticator server code was not checking that the input it
3174 received was valid base64.
3175
317616. The debug output line for the "set" modifier in ACLs was not showing the
3177 name of the variable that was being set.
3178
317917. Code tidy: the variable type "vtype_string" was never used. Removed it.
3180
318118. Previously, a reference to $sender_host_name did not cause a DNS reverse
3182 lookup on its own. Something else was needed to trigger the lookup. For
3183 example, a match in host_lookup or the need for a host name in a host list.
3184 Now, if $sender_host_name is referenced and the host name has not yet been
3185 looked up, a lookup is performed. If the lookup fails, the variable remains
3186 empty, and $host_lookup_failed is set to "1".
3187
318819. Added "eqi" as a case-independent comparison operator.
3189
319020. The saslauthd authentication condition could segfault if neither service
3191 nor realm was specified.
3192
319321. If an overflowing value such as "2048M" was set for message_size_limit, the
3194 error message that was logged was misleading, and incoming SMTP
3195 connections were dropped. The message is now more accurate, and temporary
3196 errors are given to SMTP connections.
3197
319822. In some error situations (such as 21 above) Exim rejects all SMTP commands
3199 (except RSET) with a 421 error, until QUIT is received. However, it was
3200 failing to send a response to QUIT.
3201
320223. The HELO ACL was being run before the code for helo_try_verify_hosts,
3203 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3204 ACL is now run after the helo_try_verify_hosts code.
3205
320624. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3207 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3208 case-independent, so other case variants are also recognized). Apparently
3209 some systems use these upper case variants.
3210
321125. If more than two messages were waiting for the same host, and a transport
3212 filter was specified for the transport, Exim sent two messages over the
3213 same TCP/IP connection, and then failed with "socket operation on non-
3214 socket" when it tried to send the third.
3215
321626. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3217
321827. The extern definition of crypt16() in expand.c was not being excluded when
3219 the OS had its own crypt16() function.
3220
322128. Added bounce_return_body as a new option, and bounce_return_size_limit
3222 as a preferred synonym for return_size_limit, both as an option and as an
3223 expansion variable.
3224
322529. Added LIBS=-liconv to OS/Makefile-OSF1.
3226
322730. Changed the default configuration ACL to relax the local part checking rule
3228 for addresses that are not in any local domains. For these addresses,
3229 slashes and pipe symbols are allowed within local parts, but the sequence
3230 /../ is explicitly forbidden.
3231
323231. SPA server authentication was not clearing the challenge buffer before
3233 using it.
3234
323532. log_message in a "warn" ACL statement was writing to the reject log as
3236 well as to the main log, which contradicts the documentation and doesn't
3237 seem right (because no rejection is happening). So I have stopped it.
3238
323933. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3240 However, I am unable to do any testing of this.
3241
324234. Fixed an infelicity in the appendfile transport. When checking directories
3243 for a mailbox, to see if any needed to be created, it was accidentally
3244 using path names with one or more superfluous leading slashes; tracing
3245 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3246
324735. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3248 contents are added to the log line that is written for every discarded
3249 recipient. (Previously a log_message setting was ignored.)
3250
325136. The ${quote: operator now quotes the string if it is empty.
3252
325337. The install script runs exim in order to find its version number. If for
3254 some reason other than non-existence or emptiness, which it checks, it
3255 could not run './exim', it was installing it with an empty version number,
3256 i.e. as "exim-". This error state is now caught, and the installation is
3257 aborted.
3258
325938. An argument was missing from the function that creates an error message
3260 when Exim fails to connect to the socket for saslauthd authentication.
3261 This could cause Exim to crash, or give a corrupted message.
3262
326339. Added isip, isip4, and isip6 to ${if conditions.
3264
326540. The ACL variables $acl_xx are now saved with the message, and can be
3266 accessed later in routers, transports, and filters.
3267
326841. The new lookup type nwildlsearch is like wildlsearch, except that the key
3269 strings in the file are not string-expanded.
3270
327142. If a MAIL command specified a SIZE value that was too large to fit into an
3272 int variable, the check against message_size_limit failed. Such values are
3273 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3274 day this will have to be increased, but I don't think I want to be around
3275 when emails are that large.
3276
3277
3278
3279Exim version 4.22
3280-----------------
3281
3282 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3283 iconv() is not standard in FreeBSD.
3284
3285 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3286 IPv6 enabled. The observed symptom was a segmentation fault on return from
3287 the function os_common_find_running_interfaces() in src/os.c.
3288
3289 3. In the check_special_case() function in daemon.c I had used "errno" as an
3290 argument name, which causes warnings on some systems. This was basically a
3291 typo, since it was named "eno" in the comments!
3292
3293 4. The code that waits for the clock to tick (at a resolution of some fraction
3294 of a second) so as to ensure message-id uniqueness was always waiting for
3295 at least one whole tick, when it could have waited for less. [This is
3296 almost certainly not relevant at current processor speeds, where it is
3297 unlikely to ever wait at all. But we try to future-proof.]
3298
3299 5. The function that sleeps for a time interval that includes fractions of a
3300 second contained a race. It did not block SIGALRM between setting the
3301 timer, and suspending (a couple of lines later). If the interval was short
3302 and the sigsuspend() was delayed until after it had expired, the suspension
3303 never ended. On busy systems this could lead to processes getting stuck for
3304 ever.
3305
3306 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3307 process, before it forks any delivery processes. The open lookup caching
3308 mechanism meant that the open file or database connection was passed into
3309 the delivery process. The problem was that delivery processes always tidy
3310 up cached lookup data. This could cause a problem for the next delivery
3311 process started by the queue runner, because the external queue runner
3312 process does not know about the closure. So the next delivery process
3313 still has data in the lookup cache. In the case of a file lookup, there was
3314 no problem because closing a file descriptor in a subprocess doesn't affect
3315 the parent. However, if the lookup was caching a connection to a database,
3316 the connection was closed, and the second delivery process was likely to
3317 see errors such as "PGSQL: query failed: server closed the connection
3318 unexpectedly". The problem has been fixed by closing all cached lookups
3319 in a queue runner before running a delivery process.
3320
3321 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3322 seem to have the "const" qualifier which it has on other OS. I've
3323 parameterised it.
3324
3325 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3326 *of the same type* (client or server) with the same public name that an
3327 error should be diagnosed.
3328
3329 9. When Exim looked up a host name for an IP address, but failed to find the
3330 original IP address when looking up the host name (a safety check), it
3331 output the message "<ip address> does not match any IP for NULL", which was
3332 confusing, to say the least. The bug was that the host name should have
3333 appeared instead of "NULL".
3334
333510. Since release 3.03, if Exim is called by a uid other than root or the Exim
3336 user that is built into the binary, and the -C or -D options is used, root
3337 privilege is dropped before the configuration file is read. In addition,
3338 logging is switched to stderr instead of the normal log files. If the
3339 configuration then re-defines the Exim user, the unprivileged environment
3340 is probably not what is expected, so Exim logs a panic warning message (but
3341 proceeds).
3342
3343 However, if deliver_drop_privilege is set, the unprivileged state may well
3344 be exactly what is intended, so the warning has been cut out in that case,
3345 and Exim is allowed to try to write to its normal log files.
3346
3347
3348Exim version 4.21
3349-----------------
3350
3351 1. smtp_return_error_details was not giving details for temporary sender
3352 or receiver verification errors.
3353
3354 2. Diagnose a configuration error if two authenticators have the same public
3355 name.
3356
3357 3. Exim used not to create the message log file for a message until the first
3358 delivery attempt. This could be confusing when incoming messages were held
3359 for policy or load reasons. The message log file is now created at the time
3360 the message is received, and an initial "Received" line is written to it.
3361
3362 4. The automatically generated man page for command line options had a minor
3363 bug that caused no ill effects; however, a more serious problem was that
3364 the procedure for building the man page automatically didn't always
3365 operate. Consequently, release 4.20 contains an out-of-date version. This
3366 shouldn't happen again.
3367
3368 5. When building Exim with embedded Perl support, the script that builds the
3369 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3370 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3371
3372 6. The freeze_tell option was not being used for messages that were frozen on
3373 arrival, either by an ACL or by local_scan().
3374
3375 7. Added the smtp_incomplete_transaction log selector.
3376
3377 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3378 was accepting AUTH without a new EHLO.
3379
3380 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3381 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3382 RFC.
3383
338410. Logging of TCP/IP connections (when configured) now happens in the main
3385 daemon process instead of the child process, so that the TCP/IP connection
3386 count is more accurate (but it can never be perfect).
3387
338811. The use of "drop" in a nested ACL was not being handled correctly in the
3389 outer ACL. Now, if condition failure induced by the nested "drop" causes
3390 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3391 or "require"), the connection is dropped.
3392
339312. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3394 that yield "discard" can now be used with an "accept" or a "discard" verb,
3395 but an error is generated for any others (because I can't see a useful way
3396 to define what should happen).
3397
339813. When an ACL is read dynamically from a file (or anywhere else), the lines
3399 are now processed in the same way as lines in the Exim configuration file.
3400 In particular, continuation lines are supported.
3401
340214. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3403
340415. Added -ti meaning -t -i.
3405
340616. Check for letters, digits, hyphens, and dots in the names of dnslist
3407 domains, and warn by logging if others are found.
3408
340917. At least on BSD, alignment is not guarenteed for the array of ifreq's
3410 returned from GIFCONF when Exim is trying to find the list of interfaces on
3411 a host. The code in os.c has been modified to copy each ifreq to an aligned
3412 structure in all cases.
3413
3414 Also, in some cases, the returned ifreq's were being copied to a 'struct
3415 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3416 means the last couple of bytes of an IPv6 address could be chopped if the
3417 ifreq contained only a normal sockaddr (14 bytes storage).
3418
341918. Named domain lists were not supported in the hosts_treat_as_local option.
3420 An entry such as +xxxx was not recognized, and was treated as a literal
3421 domain name.
3422
342319. Ensure that header lines added by a DATA ACL are included in the reject log
3424 if the ACL subsequently rejects the message.
3425
342620. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3427 MD5 (which is deprecated).
3428
342921. When testing a filter file using -bf, Exim was writing a message when it
3430 took the sender from a "From " line in the message, but it was not doing so
3431 when it took $return_path from a Return-Path: header line. It now does.
3432
343322. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3434 with a valid header line field name (a series of printing characters
3435 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3436
343723. Changed "disc" in the source to "disk" to conform to the documentation and
3438 the book and for uniformity.
3439
344024. Ignore Sendmail's -Ooption=value command line item.
3441
344225. When execve() failed while trying to run a command in a pipe transport,
3443 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3444 could be confused with a return value of 69 from the command itself. This
3445 has been changed to 127, the value the shell returns if it is asked to run
3446 a non-existent command. The wording for the related log line suggests a
3447 non-existent command as the problem.
3448
344926. If received_header_text expands to an empty string, do not add a Received:
3450 header line to the message. (Well, it adds a token one on the spool, but
3451 marks it "old" so that it doesn't get used or transmitted.)
3452
345327. Installed eximstats 1.28 (addition of -nt option).
3454
345528. There was no check for failure on the call to getsockname() in the daemon
3456 code. This can fail if there is a shortage of resources on the system, with
3457 ENOMEM, for example. A temporary error is now given on failure.
3458
345929. Contrary to the C standard, it seems that in some environments, the
3460 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3461 program. Exim now does this explicitly; it affects the formatting of
3462 timestamps using strftime().
3463
346430. If exiqsumm was given junk data, it threw up some uninitialized variable
3465 complaints. I've now initialized all the variables, to avoid this.
3466
346732. Header lines added by a system filter were not being "seen" during
3468 transport-time rewrites.
3469
347033. The info_callback() function passed to OpenSSL is set up with type void
3471 (*)(SSL *, int, int), as described somewhere. However, when calling the
3472 function (actually a macro) that sets it up, the type void(*)() is
3473 expected. I've put in a cast to prevent warnings from picky compilers.
3474
347534. If a DNS black list lookup found a CNAME record, but there were no A
3476 records associated with the domain it pointed at, Exim crashed.
3477
347835. If a DNS black list lookup returned more than one A record, Exim ignored
3479 all but the first. It now scans all returned addresses if a particular IP
3480 value is being sought. In this situation, the contents of the
3481 $dnslist_value variable are a list of all the addresses, separated by a
3482 comma and a space.
3483
348436. Tightened up the rules for host name lookups using reverse DNS. Exim used
3485 to accept a host name and all its aliases if the forward lookup for any of
3486 them yielded the IP address of the incoming connection. Now it accepts only
3487 those names whose forward lookup yields the correct IP address. Any other
3488 names are discarded. This closes a loophole whereby a rogue DNS
3489 administrator could create reverse DNS records to break through a
3490 wildcarded host restriction in an ACL.
3491
349237. If a user filter or a system filter that ran in a subprocess used any of
3493 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3494 the wrong values were passed to the pipe command ($thisaddress had the
3495 value of $0, $0 had the value of $1, etc). This bug was introduced by
3496 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3497
349838. Improved the line breaking for long SMTP error messages from ACLs.
3499 Previously, if there was no break point between 40 and 75 characters, Exim
3500 left the rest of the message alone. Two changes have been made: (a) I've
3501 reduced the minimum length to 35 characters; (b) if it can't find a break
3502 point between 35 and 75 characters, it looks ahead and uses the first one
3503 that it finds. This may give the occasional overlong line, but at least the
3504 remaining text gets split now.
3505
350639. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3507 file descriptors might be low, and that setting 1000 would always raise it.
3508 It turns out that in some environments, the limit is already over 1000 and
3509 that lowering it causes trouble. So now Exim takes care not to decrease it.
3510
351140. When delivering a message, the value of $return_path is set to $sender_
3512 address at the start of routing (routers may change the value). By an
3513 oversight, this default was not being set up when an address was tested by
3514 -bt or -bv, which affected the outcome if any router or filter referred to
3515 $return_path.
3516
351741. The idea of the "warn" ACL verb is that it adds a header or writes to the
3518 log only when "message" or "log_message" are set. However, if one of the
3519 conditions was an address verification, or a call to a nested ACL, the
3520 messages generated by the underlying test were being passed through. This
3521 no longer happens. The underlying message is available in $acl_verify_
3522 message for both "message" and "log_message" expansions, so it can be
3523 passed through if needed.
3524
352542. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3526 new expansion $bh_ to give the encoded byte string without charset
3527 translation. Translation happens only if iconv() is available; HAVE_ICONV
3528 indicates this at build time. HEADERS_CHARSET gives the charset to
3529 translate to; headers_charset can change it in the configuration, and
3530 "headers charset" can change it in an individual filter file.
3531
353243. Now that we have a default RFC 2047 charset (see above), the code in Exim
3533 that creates RFC 2047 encoded "words" labels them as that charset instead
3534 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3535 expansion operator; (ii) when Exim creates a From: line for a local
3536 message; (iii) when a header line is rewritten to include a "phrase" part.
3537
353844. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3539 buggy, causing it to skip the first lines of messages whose message ID
3540 ended in 'D'. This would not have bitten before Exim release 4.14, because
3541 message IDs were unlikely to end in 'D' before then. The effect was to have
3542 incorrect size information for certain domains.
3543
354445. #include "config.h" was missing at the start of the crypt16.c module. This
3545 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3546 noticed.
3547
354846. If there was a timeout during a "random" callout check, Exim treated it as
3549 a failure of the random address, and carried on sending RSET and the real
3550 address. If the delay was just some slowness somewhere, the response to the
3551 original RCPT would be taken as a response to RSET and so on, causing
3552 mayhem of various kinds.
3553
355447. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3555 when I implemented it. It didn't allow for the fact that some option values
3556 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3557 the right test for positive values.
3558
355948. Domain names in DNS records are case-independent. Exim always looks them up
3560 in lower case. Some resolvers return domain names in exactly the case they
3561 appear in the zone file, that is, they may contain uppercase letters. Not
3562 all resolvers do this - some return always lower case. Exim was treating a
3563 change of case by a resolver as a change of domain, similar to a widening
3564 of a domain abbreviation. This triggered its re-routing code and so it was
3565 trying to route what was effectively the same domain again. This normally
3566 caused routing to fail (because the router wouldn't handle the domain
3567 twice). Now Exim checks for this case specially, and just changes the
3568 casing of the domain that it ultimately uses when it transmits the message
3569 envelope.
3570
357149. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3572 module.
3573
357450. If a filter generated a file delivery with a non-absolute name (possible if
3575 no home directory exists for the router), the forbid_file option was not
3576 forbidding it.
3577
357851. Added '&' feature to dnslists, to provide bit mask matching in addition to
3579 the existing equality matching.
3580
358152. Exim was using ints instead of ino_t variables in some places where it was
3582 dealing with inode numbers.
3583
358453. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3585 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3586 called TMPDIR, and if it finds it is different, it changes its value.
3587
358854. The smtp_printf() function is now made available to local_scan() so
3589 additional output lines can be written before returning. There is also an
3590 smtp_fflush() function to enable the detection of a dropped connection.
3591 The variables smtp_input and smtp_batched_input are exported to
3592 local_scan().
3593
359455. Changed the default runtime configuration: the message "Unknown user"
3595 has been removed from the ACL, and instead placed on the localuser router,
3596 using the cannot_route_message feature. This means that any verification
3597 failures that generate their own messages won't get overridden. Similarly,
3598 the "Unrouteable address" message that was in the ACL for unverifiable
3599 relay addresses has also been removed.
3600
360156. Added hosts_avoid_esmtp to the smtp transport.
3602
360357. The exicyclog script was not checking for the esoteric option
3604 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3605 will work only if exicyclog is run under the appropriate euid.
3606
360758. Following a discussion on the list, the rules by which Exim recognises line
3608 endings on incoming messages have been changed. The -dropcr and drop_cr
3609 options are now no-ops, retained only for backwards compatibility. The
3610 following line terminators are recognized: LF CRLF CR. However, special
3611 processing applies to CR:
3612
3613 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3614 nor a local message in the state where . is a terminator.
3615
3616 (ii) If a bare CR is encountered in a header line, an extra space is added
3617 after the line terminator so as not to end the header. The reasoning
3618 behind this is that bare CRs in header lines are most likely either
3619 to be mistakes, or people trying to play silly games.
3620
362159. The size of a message, as listed by "-bp" or in the Exim monitor window,
3622 was being incorrectly given as 18 bytes larger than it should have been.
3623 This is a VOB (very old bug).
3624
362560. This may never have affected anything current, but just in case it has:
3626 When the local host is found other than at the start of a list of hosts,
3627 the local host, those with the same MX, and any that follow, are discarded.
3628 When the list in question was part of a longer list of hosts, the following
3629 hosts (not currently being processed) were also being discarded. This no
3630 longer happens. I'm not sure if this situation could ever has previously
3631 arisen.
3632
363361. Added the "/MX" feature to lists of hosts in the manualroute and query
3634 program routers.
3635
363662. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3637 header. This is something that is recommended in a new Internet Draft, and
3638 is something that is documented as being done by Sendmail. There are two
3639 possible values. For messages generated by the autoreply transport, Exim
3640 adds:
3641
3642 Auto-Submitted: auto-replied
3643
3644 whereas for all other generated messages (e.g. bounces) it adds
3645
3646 Auto-Submitted: auto-generated
3647
364863. The "personal" condition in filters now includes a test for the
3649 Auto-Submitted: header. If it contains the string "auto-" the message it
3650 not considered personal.
3651
365264. Added rcpt_include_affixes as a generic transport option.
3653
365465. Added queue_only_override (default true).
3655
365666. Added the syslog_duplication option.
3657
365867. If what should have been the first header line of a message consisted of
3659 a space followed by a colon, Exim was mis-interpreting it as a header line.
3660 It isn't of course - it is syntactically invalid and should therefore be
3661 treated as the start of the message body. The misbehaviour could have
3662 caused a number of strange effects, including loss of data in subsequent
3663 header lines, and spool format errors.
3664
366568. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3666 client host had authenticated. This control can now be exercised by an ACL
3667 for more flexibility.
3668
366969. By default, callouts do not happen when testing with -bh. There is now a
3670 variant, -bhc, which does actually run the callout code, including
3671 consulting and updating the callout cache.
3672
367370. Added support for saslauthd authentication, courtesy of Alexander
3674 Sabourenkov.
3675
367671. If statvfs() failed on the spool or log directories while checking their
3677 size for availability, Exim confusingly gave the error "space shortage".
3678 Furthermore, in debugging mode it crashed with a floating point exception.
3679 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3680 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3681 serious problem, Exim now writes to the main and panic logs when this