Fix eximon buffer overflow bug.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
0d46a8c8 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.298 2006/02/14 14:26:14 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
214e2000 163
5de37277
PH
164Exim version 4.60
165-----------------
166
cc38ddbf
PH
167PH/01 Two changes to the default runtime configuration:
168
169 (1) Move the checks for relay_from_hosts and authenticated clients from
170 after to before the (commented out) DNS black list checks.
171
172 (2) Add control=submission to the relay_from_hosts and authenticated
173 clients checks, on the grounds that messages accepted by these
174 statements are most likely to be submissions.
5de37277 175
72fdd6ae
PH
176PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
177
178 (1) Generate an error if the third argument for the ${prvs expansion is
179 not a single digit.
180
181 (2) Treat a missing third argument of ${prvscheck as if it were an empty
182 string.
183
184 (3) Reset the variables that are obtained from the first argument of
185 ${prvscheck and used in the second argument before leaving the code,
186 because their memory is reclaimed, so using them afterwards may do
187 silly things.
188
189 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
190 one (it's much easier than Tom thought :-).
191
192 (5) Because of (4), we can now allow for the use of $prvscheck_result
193 inside the third argument.
cb9328de 194
cb741023
PH
195PH/03 For some reason, the default setting of PATH when running a command from
196 a pipe transport was just "/usr/bin". I have changed it to
197 "/bin:/usr/bin".
198
f174f16e
PH
199PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
200 anything to be listed in the output from -bV.
b2f5a032 201
c25242d7
PH
202PH/05 When a filter generated an autoreply, the entire To: header line was
203 quoted in the delivery log line, like this:
204
205 => >A.N.Other <ano@some.domain> <original@ddress> ...
206
207 This has been changed so that it extracts the operative address. There
208 may be more than one such address. If so, they are comma-separated, like
209 this:
210
211 => >ano@some.domain,ona@other.domain <original@ddress> ...
212
82c19f95
PH
213PH/06 When a client host used a correct literal IP address in a HELO or EHLO
214 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
215 not being looked up in the rDNS to get a host name, Exim was showing the
216 IP address twice in Received: lines, even though the IP addresses were
217 identical. For example:
218
219 Received: from [1.2.3.4] (helo=[1.2.3.4])
220
221 However, if the real host name was known, it was omitting the HELO data
222 if it matched the actual IP address. This has been tidied up so that it
223 doesn't show the same IP address twice.
224
d7ffbc12
PH
225PH/07 When both +timestamp and +memory debugging was on, the value given by
226 $tod_xxx expansions could be wrong, because the tod_stamp() function was
227 called by the debug printing, thereby overwriting the timestamp buffer.
228 Debugging no longer uses the tod_stamp() function when +timestamp is set.
229
9f526266
PH
230PH/08 When the original message was included in an autoreply transport, it
231 always said "this is a copy of the message, including all the headers",
232 even if body_only or headers_only was set. It now gives an appropriate
233 message.
234
87fcc8b9
PH
235PH/09 Applied a patch from the Sieve maintainer which:
236
237 o fixes some comments
238 o adds the (disabled) notify extension core
239 o adds some debug output for the result of if/elsif tests
240 o points to the current vacation draft in the documentation
241 and documents the missing references header update
242
243 and most important:
244
245 o fixes a bug in processing the envelope test (when testing
246 multiple envelope elements, the last element determinted the
247 result)
248
456682f5
PH
249PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
250 Electronic Mail") by including:
251
252 Auto-submitted: auto-generated
253
254 in the messages that it generates (bounce messages and others, such as
255 warnings). In the case of bounce messages for non-SMTP mesages, there was
256 also a typo: it was using "Auto_submitted" (underscore instead of
257 hyphen). Since every message generated by Exim is necessarily in response
258 to another message, thes have all been changed to:
259
260 Auto-Submitted: auto-replied
261
262 in accordance with these statements in the RFC:
263
264 The auto-replied keyword:
265
266 - SHOULD be used on messages sent in direct response to another
267 message by an automatic process,
268
269 - MUST NOT be used on manually-generated messages,
270
271 - MAY be used on Delivery Status Notifications (DSNs) and Message
272 Disposition Notifications (MDNs),
273
274 - MUST NOT be used on messages generated by automatic or periodic
275 processes, except for messages which are automatic responses to
276 other messages.
277
3e46c1aa
PH
278PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
279 to the default Received: header definition.
456682f5 280
49826d12
PH
281PH/12 Added log selector acl_warn_skipped (default on).
282
eba0c039
PH
283PH/13 After a successful wildlsearch lookup, discard the values of numeric
284 variables because (a) they are in the wrong storage pool and (b) even if
285 they were copied, it wouldn't work properly because of the caching.
286
a0d6ba8a
PH
287PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
288 checking when decoding. Apparently there are clients that generate
289 overlong encoded strings. Why am I not surprised?
290
f0917727
PH
291PH/15 If the first argument of "${if match_address" was not empty, but did not
292 contain an "@" character, Exim crashed. Now it writes a panic log message
293 and treats the condition as false.
294
096fee00
PH
295PH/16 In autoreply, treat an empty string for "once" the same as unset.
296
024bd3c2
PH
297PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
298 extension "envelope-auth". The code is finished and in agreement with
299 other implementations, but there is no documentation so far and in fact,
300 nobody wrote the draft yet. This extension is currently #undef'ed, thus
301 not changing the active code.
302
303 Print executed "if" and "elsif" statements when debugging is used. This
304 helps a great deal to understand what a filter does.
305
306 Document more things not specified clearly in RFC3028. I had all this
307 sorted out, when out of a sudden new issues came to my mind. Oops."
308
df199fec
PH
309PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
310 (Bugzilla #53).
311
d27f1df3
PH
312PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
313 canonical form (as documented). However, after a host name lookup from
314 the IP address, check_host() was doing a simple string comparison with
315 addresses acquired from the DNS when checking that the found name did
316 have the original IP as one of its addresses. Since any found IPv6
317 addresses are likely to be in abbreviated form, the comparison could
318 fail. Luckily, there already exists a function for doing the comparison
319 by converting both addresses to binary, so now that is used instead of
320 the text comparison.
321
96776534
PH
322PH/20 There was another similar case to PH/19, when a complete host name was
323 given in a host list; looking up its IP address could give an abbreviated
324 form, whereas the current host's name might or might not be abbreviated.
325 The same fix has been applied.
326
5de37277 327
9a799bc0
PH
328Exim version 4.54
329-----------------
330
331PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
332 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
333 It now does.
334
99a4b039
PH
335PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
336 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
337
338PH/03 Typo: missing ".o" in src/pcre/Makefile.
339
4b233853
PH
340PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
341 header line, restrict the check to what is listed in RFCs 2369 and 2929.
342 Also, for "Auto-Submitted", treat anything other than "no" as
343 non-personal, in accordance with RFC 3834. (Previously it treated
344 anything starting "auto-" as non-personal.)
345
8857ccfd
PH
346TF/01 The control=submission/name=... option had a problem with syntax
347 errors if the name included a slash character. The /name= option
348 now slurps the rest of the string, so it can include any characters
349 but it must come last in the list of options (after /sender_retain
350 or /domain=).
351
433a2980
PH
352PH/05 Some modifications to the interface to the fake nameserver for the new
353 testing suite.
354
3e46c1aa 355
9a799bc0 356
e3a311ba
TK
357Exim version 4.53
358-----------------
359
360TK/01 Added the "success_on_redirect" address verification option. See
361 NewStuff for rationale and an example.
362
13b685f9
PH
363PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
364
395ff96d
PH
365PH/02 Patch to exigrep to allow it to work on syslog lines.
366
5b68f6e4
PH
367PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
368 fread() to skip over the body file's header line, because in Cygwin the
369 header line is locked and is inaccessible.
370
1ab52c69
PH
371PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
372 co-exist for some time) to make it clear that it is the Exim ID that is
373 referenced, not the Message-ID: header line.
374
b07e6aa3
PH
375PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
376 string_format() function, because snprintf() does not exist on all
377 operating systems.
378
254e032f
PH
379PH/06 The use of forbid_filter_existstest now also locks out the use of the
380 ${stat: expansion item.
381
3af76a81
PH
382PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
383 protocol synchronization error", to keep the pedants happy.
384
2548ba04
PH
385PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
386 well as for IRIX systems, when gcc is being used. See the host.c source
387 file for comments.
388
b6c6011d
PH
389PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
390
cf39cf57
PH
391PH/10 Named domain lists were not working if used in a queue_smtp_domains
392 setting.
393
f1513293
PH
394PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
395 transport and to the smtp transport in LMTP mode.
396
727549a4
PH
397TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
398
af46795e
PH
399PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
400 run a filter in a subprocess. This could lead to confusion in subsequent
401 lookups in the parent process. There should also be a search_tidyup() at
402 the end of the subprocess.
403
d7b47fd0
PH
404PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
405 only if the host matched helo_try_verify_hosts, which caused the
406 verification to occur when the EHLO/HELO command was issued. The ACL just
407 tested the remembered result. Now, if a previous verification attempt has
408 not happened, "verify = helo" does it there and then.
409
ee744174
JJ
410JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
411
b582ab87
PH
412TK/03 Fix log output including CR from clamd.
413
41a13e0a
PH
414PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
415 exist provoked a memory error which could cause a segfault.
416
f625cc5a
PH
417PH/15 Installed PCRE 6.2
418
419PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
420
21f7af35
PH
421PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
422 of the problem. Specifically, suggested +O2 rather than +O1 for the
423 HP-UX compiler.
424
31480e42
PH
425PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
426
2d280592
PH
427PH/20 If a delivery was routed to a non-standard port by means of an SRV
428 record, the port was not correctly logged when the outgoing_port log
429 selector was set (it logged the transort's default port).
430
7cd1141b
PH
431PH/21 Added support for host-specific ports to manualroute, queryprogram,
432 fallback_hosts, and "hosts" in the smtp transport.
433
434PH/22 If the log selector "outgoing_port" is set, the port is now also given on
435 host errors such as "Connection refused".
436
750af86e
PH
437PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
438 authentication with radiusclient 0.4.9:
439
440 - Error returned from rc_read_config was caught wrongly
441 - Username/password not passed on to radius server due to wrong length.
442
443 The presumption is that some radiusclient API changes for 4.51/PH/17
444 were not taken care of correctly. The code is still untested by me (my
445 Linux distribution still has 0.3.2 of radiusclient), but it was
446 contributed by a Radius user.
447
448PH/24 When doing a callout, the value of $domain wasn't set correctly when
449 expanding the "port" option of the smtp transport.
450
4304270b
TK
451TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
452 while reading a MIME header. Thanks to Tom Hughes for a patch.
453
750af86e
PH
454PH/24 Include config.h inside local_scan.h so that configuration settings are
455 available.
456
64ffc24f
PH
457PH/25 Make $smtp_command_argument available after all SMTP commands. This means
458 that in an ACL for RCPT (for example), you can examine exactly what was
459 received.
460
5dd9625b
PH
461PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
462 commands, but it was not correctly comparing the address with the actual
463 client host address. Thus, it would show the EHLO address in Received:
464 header lines when this was not necessary.
465
5591031b
PH
466PH/27 Added the % operator to ${eval:}.
467
ba18e66a
PH
468PH/28 Exim tries to create and chdir to its spool directory when it starts;
469 it should be ignoring failures (because with -C, for example, it has lost
470 privilege). It wasn't ignoring creation failures other than "already
471 exists".
472
9cec981f
PH
473PH/29 Added "crypteq" to the list of supported features that Exim outputs when
474 -bV or -d is used.
475
aa2b5c79
PH
476PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
477 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
478 too many continuations, the temporary file was not being removed, and the
479 return code was incorrect.
aa2b5c79 480
48a53b7f
PH
481PH/31 Missing "BOOL" in function definition in filtertest.c.
482
1c59d63b
PH
483PH/32 Applied Sieve patches from the maintainer.
484
671012da
TK
485TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
486
1509d3a8
PH
487PH/33 Added "verify = not_blind".
488
489PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
490 Local/Makefile (with some defaults set). These are used in built scripts
491 such as exicyclog, but they have never been used in the exim_install
492 script (though there are many overriding facilities there). I have
493 arranged that the exim_install script now takes note of these two
494 settings.
495
496PH/35 Installed configuration files for Dragonfly.
497
2fe1a124
PH
498PH/36 When a locally submitted message by a trusted user did not contain a
499 From: header, and the sender address was obtained from -f or from an SMTP
500 MAIL command, and the trusted user did not use -F to supply a sender
501 name, $originator_name was incorrectly used when constructing a From:
502 header. Furthermore, $originator_name was used for submission mode
503 messages from external hosts without From: headers in a similar way,
504 which is clearly wrong.
505
8800895a
PH
506PH/37 Added control=suppress_local_fixups.
507
ccfdb010
PH
508PH/38 When log_selector = +received_sender was set, and the addition of the
509 sender made the log line's construction buffer exactly full, or one byte
510 less than full, an overflow happened when the terminating "\n" was
511 subsequently added.
512
1130bfb0
PH
513PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
514 when the result of a list match is failure because a DNS lookup failed.
515
ebcb507f
PH
516PH/40 RM_COMMAND is now used in the building process.
517
c35e155c
PH
518PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
519 the "build-* directories that it finds.
520
95d1f782
PH
521PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
522 address in a domain literal was a prefix of an interface address.
523
fd6de02e
PH
524PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
525 when verifying a sender address, unless rewrite_headers is false.
526
58de37c5
PH
527PH/44 Wrote a long comment about why errors_to addresses are verified as
528 recipients, not senders.
529
261cf466
TF
530TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
531 the ratelimit ACL was added.
532
3ee512ff
PH
533PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
534
e08c430f
PH
535PH/46 Added extra information about PostgreSQL errors to the error string.
536
bef5a11f
PH
537PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
538 avoiding the need to install special zones in a real server. This is
539 backwards compatible; if it can't find the fake resolver, it drops back.
540 Thus, both old and new test suites can be run.
541
7546de58
TF
542TF/02 Added util/ratelimit.pl
543
e5d5a95f
TF
544TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
545 clock is set back in time.
546
2e88a017
TF
547TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
548 Candler <B.Candler@pobox.com>.
549
a5f65aa4
TF
550TF/05 The fix for PH/43 was not completely correct; widen_domains is always
551 OK for addresses that are the result of redirections.
552
e7726cbf
PH
553PH/48 A number of further additions for the benefit of the new test suite,
554 including a fake gethostbyname() that interfaces to the fake DNS resolver
555 (see PH/47 above).
556
a7fdad5b
TF
557TF/06 The fix for widen_domains has also been applied to qualify_single and
558 search_parents which are the other dnslookup options that can cause
559 header rewrites.
560
6af56900
PH
561PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
562 type ("H").
563
0925ede6
PH
564PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
565
66afa403
TF
566TF/07 Exim produced the error message "an SRV record indicated no SMTP
567 service" if it encountered an MX record with an empty target hostname.
568 The message is now "an MX or SRV record indicated no SMTP service".
569
0154e85a
TF
570TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
571 if the DNS of the sending site is misconfigured. This is quite a
572 common situation. This change restores the behaviour of treating a
573 helo verification defer as a failure.
574
16f12c76
PH
575PH/51 If self=fail was set on a router, the bounce message did not include the
576 actual error message.
577
bbe902f0 578
e5a9dba6
PH
579Exim version 4.52
580-----------------
581
582TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
583
22c3b60b
PH
584PH/01 When a transport filter timed out in a pipe delivery, and the pipe
585 command itself ended in error, the underlying message about the transport
586 filter timeout was being overwritten with the pipe command error. Now the
587 underlying error message should be appended to the second error message.
588
06a9b4b5
PH
589TK/01 Fix poll() being unavailable on Mac OSX 10.2.
590
c1ac6996
PH
591PH/02 Reduce the amount of output that "make" produces by default. Full output
592 can still be requested.
593
9c7a242c
PH
594PH/03 The warning log line about a condition test deferring for a "warn" verb
595 was being output only once per connection, rather than after each
596 occurrence (because it was using the same function as for successful
597 "warn" verbs). This seems wrong, so I have changed it.
598
87ba3f5f
PH
599TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
600 it should not have, which might have caused a crash in the right
601 circumstances, but probably never did.
602
603PH/04 Installed a modified version of Tony Finch's patch to make submission
604 mode fix the return path as well as the Sender: header line, and to
605 add a /name= option so that you can make the user's friendly name appear
606 in the header line.
607
29aba418
TF
608TF/03 Added the control = fakedefer ACL modifier.
609
fe0dab11
TF
610TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
611 Mark Lowes for thorough testing.
870f6ba8 612
11d337a4
TK
613TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
614
615TK/03 Merged latest SRS patch from Miles Wilton.
616
415c8f3b
PH
617PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
618 with the definition in sysexits.h (which is #included earlier).
619 Fortunately, Exim does not actually use EX_OK. The code used to try to
620 preserve the sysexits.h value, by assumimg that macro definitions were
621 scanned for macro replacements. I have been disabused of this notion,
622 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 623
958541e9
PH
624PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
625 in the smtp transport. When a block could not be written in a single
626 write() function, the timeout was being re-applied to each part-write.
627 This seems wrong - if the receiver was accepting one byte at a time it
628 would take for ever. The timeout is now adjusted when this happens. It
629 doesn't have to be particularly precise.
630
c206415f
TK
631TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
632 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
633
2a4be8f9
PH
634PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
635 without a domain if the check to <postmaster@domain> fails.
636
1cba11c5
SC
637SC/01 Eximstats: added -xls and the ability to specify output files
638 (patch written by Frank Heydlauf).
639
640SC/02 Eximstats: use FileHandles for outputing results.
641
642SC/03 Eximstats: allow any combination of xls, txt, and html output.
643
644SC/04 Eximstats: fixed display of large numbers with -nvr option
645
646SC/05 Eximstats: fixed merging of reports with empty tables.
647
648SC/06 Eximstats: added the -include_original_destination flag
649
650SC/07 Eximstats: removed tabs and trailing whitespace.
651
1005d00e
TK
652TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
653
654TK/06 MBOX spool code: Add real "From " MBOX separator line
655 so the .eml file is really in mbox format (even though
656 most programs do not really care). Patch from Alex Miller.
657
658TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
659 The latter is generated from $received_to and is only set if the
660 message has one envelope recipient. SA can use these headers,
661 obviously out-of-the-box. Patch from Alex Miller.
662
9b4768fa
PH
663PH/08 The ${def test on a variable was returning false if the variable's
664 value was "0", contrary to what the specification has always said!
665 The result should be true unless the variable is empty.
666
667PH/09 The syntax error of a character other than { following "${if
668 def:variable_name" (after optional whitespace) was not being diagnosed.
669 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
670 accidental colon was present, for example, could give incorrect results.
671
0d7eb84a
PH
672PH/10 Tidied the code in a number of places where the st_size field of a stat()
673 result is used (not including appendfile, where other changes are about
674 to be made).
675
676PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
677 This involved changing a lot of size variables from int to off_t. It
678 should work with maildirs and everything.
679
40727bee
TK
680TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
681 spamd dying while we are connected to it.
682
554d2369
TF
683TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
684 <jgh@wizmail.org>
685
1f922db1
PH
686PH/12 Applied Alex Kiernan's patch for the API change for the error callback
687 function for BDB 4.3.
688
ef213c3b
PH
689PH/13 Changed auto_thaw such that it does not apply to bounce messages.
690
8ac170f3
PH
691PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
692 the sources for PCRE have been re-arranged and more files are now
693 involved.
694
b1c749bb
PH
695PH/15 The code I had for printing potentially long long variables in PH/11
696 above was not the best (it lost precision). The length of off_t variables
697 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
698 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
699 to be "long long int" or "long int". This is needed for the internal
700 formatting function string_vformat().
b1c749bb 701
4aac9b49
PH
702PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
703 the configuration file to be ":syslog", then the script "guesses" where
704 the logs files are, rather than using the compiled in default. In our
705 case the guess is not the same as the compiled default, so the script
706 suddenly stopped working when I started to use syslog. The patch checks
707 to see if log_file_path is "". If so, it attempts to read it from exim
708 with no configuration file to get the compiled in version, before it
709 falls back to the previous guessing code."
710
294520c8
TK
711TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
712 implementing BATV in an Exim configuration. See NewStuff for the gory
713 details.
714
5bd022fe
PH
715PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
716 Makefile that are specific to HP-UX.
717
90e9ce59
PH
718PH/18 If the "use_postmaster" option was set for a recipient callout together
719 with the "random" option, the postmaster address was used as the MAIL
720 FROM address for the random test, but not for the subsequent recipient
721 test. It is now used for both.
722
5ea81592
PH
723PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
724 patch removes a few documentation additions to RFC 3028, because the
725 latest draft now contains them. It adds the new en;ascii-case comparator
726 and a new error check for 8bit text in MIME parts. Comparator and
727 require names are now matched exactly. I enabled the subaddress
728 extension, but it is not well tested yet (read: it works for me)."
729
c6c2dc1d
PH
730PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
731 rework some of the code of TK/09 above to avoid the hardwired use of
732 "%lld" and "long long". Replaced the call to snprintf() with a call to
733 string_vformat().
734
fffffe4c
PH
735PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
736 records point to non-existent hosts", "retry timeout exceeded", and
737 "retry time not reached for any host after a long failure period".
ca02eafb 738
9a26b6b2
PH
739PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
740 experimental DomainKeys support:
741
742 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
743 (2) On an error such as an illegally used "control", the wrong name for
744 the control was given.
745
746 These problems did NOT occur unless DomainKeys support was compiled.
747
4aee0225
PH
748PH/23 Added daemon_startup_retries and daemon_startup_sleep.
749
32d668a5
PH
750PH/24 Added ${if match_ip condition.
751
8187c3f3
PH
752PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
753 databases so that it will be absolutely obvious if a crash occurs in the
754 DB library. This is a regular occurrence (often caused by mis-matched
755 db.h files).
756
ff790e47 757PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
758 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
759 picked up on a user's system that detects such things. There doesn't seem
760 to be a gcc warning option for this - only an attribute that has to be
761 put on the function's prototype. It seems that in Fedora Core 4 they have
762 set this on a number of new functions. No doubt there will be more in due
763 course.
ff790e47 764
5417f6d1
PH
765PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
766 specify a transport. However, if an address that was verified by such a
767 router was the subject of a callout, Exim crashed because it tried to
768 read the rcpt_include_affixes from the non-existent transport. Now it
769 just assumes that the setting of that option is false. This bug was
770 introduced by 4.51/PH/31.
771
59cf8544
PH
772PH/28 Changed -d+all to exclude +memory, because that information is very
773 rarely of interest, but it makes the output a lot bigger. People tend to
774 do -d+all out of habit.
775
e7ad8a65
PH
776PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
777 code in os-type was giving problems when libc.so lives in lib64, like on
778 x86_64 Fedora Core.
779
ade42478
PH
780PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
781 aren't the modern standard, and it seems that some systems' include files
782 don't always have them. Exim was already checking for some of the newer
783 ones like T_AAAA, and defining it itself. I've added checks for all the
784 record types that Exim uses.
785
182ad5cf
PH
786PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
787 not automatically generating a new one, as it is supposed to. This
788 prevented TLS from working. If the file did exist, but contained invalid
789 data, a new version was generated, as expected. It was only the case of a
790 non-existent file that was broken.
791
b0d9fc80
TK
792TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
793 with a change in libdomainkeys > 0.64.
794
795TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
796 from DNS. If the selector record carries the flag, it now has
797 precedence over the domain-wide flag.
798
799TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
800
47c7a64a
PH
801PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
802 the use of a port name that isn't defined in /etc/services) occurred, the
803 message was deferred as in a normal delivery, and thus remained on the
804 spool, instead of being failed because of the mua_wrapper setting. This
805 is now fixed, and I tidied up some of the mua_wrapper messages at the
806 same time.
807
a388bce4
SC
808SC/08 Eximstats: whilst parsing the mainlog(s), store information about
809 the messages in a hash of arrays rather than using individual hashes.
810 This is a bit cleaner and results in dramatic memory savings, albeit
811 at a slight CPU cost.
812
813SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
814 as requested by Marc Sherman.
815
816SC/10 Eximstats: added histograms for user specified patterns as requested
817 by Marc Sherman.
818
0793e4ed
SC
819SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
820
c58b88df
PH
821PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
822 fopen() in the content-scanning modules that did not already have it.
823
e7ad8a65 824
7982096b
PH
825Exim version 4.51
826-----------------
827
1a46a8c5
PH
828TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
829 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
830
2f079f46 831TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
832
833TK/03 Fix ACL "regex" condition not being available in MIME ACL.
834
835PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
836 to test Sieve filters that use "vacation".
837
838PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
839 that changes the way the GnuTLS parameters are stored in the cache file.
840 The new format can be generated externally. For backward compatibility,
841 if the data in the cache doesn't make sense, Exim assumes it has read an
842 old-format file, and it generates new data and writes a new file. This
843 means that you can't go back to an older release without removing the
844 file.
845
846PH/03 A redirect router that has both "unseen" and "one_time" set does not
847 work if there are any delivery delays because "one_time" forces the
848 parent to be marked "delivered", so its unseen clone is never tried
849 again. For this reason, Exim now forbids the simultaneous setting of
850 these two options.
851
852PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
853 redirected to themselves ("homonym" addresses). Read the long ChangeLog
854 entry if you want to know the details. The fix, however, neglected to
855 consider the case when local delivery batching is involved. The test for
856 "previously delivered" was not happening when checking to see if an
857 address could be batched with a previous (undelivered) one; under
858 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 859 address.
1a46a8c5
PH
860
861PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
862 in its include files, and this causes problems building Exim.
863
864PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
865 header_syntax) but Exim was just ignoring anything given after a slash.
866 In particular, this caused confusion with an attempt to use "verify =
867 reverse_host_lookup/defer_ok". An error is now given when options are
868 supplied for verify items that do not have them. (Maybe reverse_host_
869 lookup should have a defer_ok option, but that's a different point.)
870
871PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
872 defined by RFC 821) to 2048, because there were problems with some AUTH
873 commands, and RFC 1869 says the size should be increased for extended
874 SMTP commands that take arguments.
875
876PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
877 Finch).
878
879PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
880 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 881
49c2d5ea
PH
882PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
883 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
884 whatever). Otherwise the password may appear in the log. Following change
885 PH/42 below, there is no longer a chance of it appearing in a bounce
886 message.
49c2d5ea 887
bf759a8b
PH
888PH/11 Installed exipick version 20050225.0 from John Jetmore.
889
83364d30
PH
890PH/12 If the last host in a fallback_hosts list was multihomed, only the first
891 of its addresses was ever tried. (Bugzilla bug #2.)
892
7999bbd7
PH
893PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
894 the result incorrectly in the debug output. (It correctly added a newline
895 to what was transported.)
896
7dbf77c9
PH
897TF/01 Added $received_time.
898
74e0617f
PH
899PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
900 commented out examples of how to interface to a virus scanner and to
901 SpamAssassin. Also added commented examples of av_scanner and
902 spamd_address settings.
903
2f079f46
PH
904PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
905 and controls are allowed in which ACLs. There were a couple of minor
906 errors. Some of the entries in the conditions table (which is a table of
907 where they are NOT allowed) were getting very unwieldy; rewrote them as a
908 negation of where the condition IS allowed.
909
8c841523
PH
910PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
911
7766a4f0
PH
912PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
913 header file does not have a version number, so I've had to invent a new
914 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
915 API. The code is untested by me (my Linux distribution still has 0.3.2 of
916 radiusclient), but it was contributed by a Radius user.
917
8b417f2c
PH
918PH/18 Installed Lars Mainka's patch for the support of CRL collections in
919 files or directories, for OpenSSL.
920
901f42cb
PH
921PH/19 When an Exim process that is running as root has to create an Exim log
922 file, it does so in a subprocess that runs as exim:exim so as to get the
923 ownership right at creation (otherwise, other Exim processes might see
924 the file with the wrong ownership). There was no test for failure of this
925 fork() call, which would lead to the process getting stuck as it waited
926 for a non-existent subprocess. Forks do occasionally fail when resources
927 run out. I reviewed all the other calls to fork(); they all seem to check
928 for failure.
929
f9b9210e
PH
930PH/20 When checking for unexpected SMTP input at connect time (before writing
931 the banner), Exim was not dealing correctly with a non-positive return
932 from the read() function. If the client had disconnected by this time,
933 the result was a log entry for a synchronization error with an empty
934 string after "input=" when read() returned zero. If read() returned -1
935 (an event I could not check), uninitialized data bytes were printed.
936 There were reports of junk text (parts of files, etc) appearing after
937 "input=".
938
54cdb463
PH
939PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
940
cf00dad6
PH
941PH/22 Added support for macro redefinition, and (re)definition in between
942 driver and ACL definitions.
943
acb1b346
PH
944PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
945 forgetting to use the resulting value; it was using the unexpanded value.
946
c5ddb310
PH
947PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
948 hadn't been configured. The fix is from Juergen Kreileder, who
949 understands it better than I do:
950
951 "Here's what I see happening with three configured cyrus_sasl
952 authenticators configured (plain, login, cram-md5):
953
954 On startup auth_cyrus_sasl_init() gets called for each of these.
955 This means three calls to sasl_listmech() without a specified mech_list.
956 => SASL tests which mechs of all available mechs actually work
957 => three warnings about OTP not working
958 => the returned list contains: plain, login, cram-md5, digest-md5, ...
959
960 With the patch, sasl_listmech() also gets called three times. But now
961 SASL's mech_list option is set to the server_mech specified in the the
962 authenticator. Or in other words, the answer from sasl_listmech()
963 gets limited to just the mech you're testing for (which is different
964 for each call.)
965 => the return list contains just 'plain' or 'login', 'cram-md5' or
966 nothing depending on the value of ob->server_mech.
967
968 I've just tested the patch: Authentication still works fine,
969 unavailable mechs specified in the exim configuration are still
970 caught, and the auth.log warnings about OTP are gone."
971
31619da6
PH
972PH/25 When debugging is enabled, the contents of the command line are added
973 to the debugging output, even when log_selector=+arguments is not
974 specified.
975
bebaf0fc
PH
976PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
977 answer is "GNU", and only if the return is "GNU/something" is the answer
978 "Linux".
979
475fe28a
PH
980PH/27 $acl_verify_message is now set immediately after the failure of a
981 verification in an ACL, and so is available in subsequent modifiers. In
982 particular, the message can be preserved by coding like this:
983
984 warn !verify = sender
985 set acl_m0 = $acl_verify_message
986
987 Previously, $acl_verify_message was set only while expanding "message"
988 and "log_message" when a very denied access.
989
7e8bec7a
PH
990PH/28 Modified OS/os.c-Linux with
991
992 -#ifndef OS_LOAD_AVERAGE
993 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
994
995 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
996 nomenclature these days.)
997
e4a89c47
PH
998PH/29 Installed patch from the Sieve maintainer that adds the options
999 sieve_useraddress and sieve_subaddress to the redirect router.
1000
5ca2a9a1
PH
1001PH/30 In these circumstances:
1002 . Two addresses routed to the same list of hosts;
1003 . First host does not offer TLS;
1004 . First host accepts first address;
1005 . First host gives temporary error to second address;
1006 . Second host offers TLS and a TLS session is established;
1007 . Second host accepts second address.
1008 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1009 and peerdn, if requested) that were in fact used only for the second
1010 address.
7e8bec7a 1011
c688b954
PH
1012PH/31 When doing a callout as part of verifying an address, Exim was not paying
1013 attention to any local part prefix or suffix that was matched by the
1014 router that accepted the address. It now behaves in the same way as it
1015 does for delivery: the affixes are removed from the local part unless
1016 rcpt_include_affixes is set on the transport.
1017
fed77020
PH
1018PH/32 Add the sender address, as F=<...>, to the log line when logging a
1019 timeout during the DATA phase of an incoming message.
1020
7fe1560f
PH
1021PH/33 Sieve envelope tests were broken for match types other than :is. I have
1022 applied a patch sanctioned by the Sieve maintainer.
c688b954 1023
ebb6e6d5
PH
1024PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1025 the uid or gid is negative. A case of a negative gid caused this to be
1026 noticed. The fix allows for either to be negative.
1027
9c4e8f60
PH
1028PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1029 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1030 overlooked.
1031
1032PH/36 The change PH/12 above was broken. Fixed it.
1033
d7174846
PH
1034PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1035 the grounds that routing the same address twice would always produce the
1036 same answer. This might have been true once, but it is certainly no
1037 longer true now. Routing a child address may depend on the previous
1038 routing that produced that child. Some complicated redirection strategies
1039 went wrong when messages had multiple recipients, and made Exim's
1040 behaviour dependent on the order in which the addresses were given.
1041
1042 I have moved the duplicate checking until after the routing is complete.
1043 Exim scans the addresses that are assigned to local and remote
1044 transports, and removes any duplicates. This means that more work will be
1045 done, as duplicates will always all be routed, but duplicates are
1046 presumably rare, so I don't expect this is of any significance.
1047
1048 For deliveries to pipes, files, and autoreplies, the duplicate checking
1049 still happens during the routing process, since they are not going to be
1050 routed further.
1051
cfe75fc3
PH
1052PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1053 It corrects a timeout issue with spamd. This is Ian's comment: "The
1054 background is that sometimes spamd either never reads data from a
1055 connection it has accepted, or it never writes response data. The exiscan
1056 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1057 blindly assumes that writes won't block so it may never time out."
1058
be22d70e
PH
1059PH/39 Allow G after quota size as well as K and M.
1060
0612b098
PH
1061PH/40 The value set for $authenticated_id in an authenticator may not contain
1062 binary zeroes or newlines because the value is written to log lines and
1063 to spool files. There was no check on this. Now the value is run through
1064 the string_printing() function so that such characters are converted to
1065 printable escape sequences.
1066
2e0c1448
PH
1067PH/41 $message_linecount is a new variable that contains the total number of
1068 lines in the message. Compare $body_linecount, which is the count for the
1069 body only.
1070
447d236c
PH
1071PH/42 Exim no longer gives details of delivery errors for specific addresses in
1072 bounce and delay warning messages, except in certain special cases, which
1073 are as follows:
1074
1075 (a) An SMTP error message from a remote host;
1076 (b) A message specified in a :fail: redirection;
1077 (c) A message specified in a "fail" command in a system filter;
1078 (d) A message specified in a FAIL return from the queryprogram router;
1079 (e) A message specified by the cannot_route_message router option.
1080
1081 In these cases only, Exim does include the error details in bounce and
1082 warning messages. There are also a few cases where bland messages such
1083 as "unrouteable address" or "local delivery error" are given.
1084
d20976dc
PH
1085PH/43 $value is now also set for the "else" part of a ${run expansion.
1086
f656d135
PH
1087PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1088 being worked on, but at least Exim now implements the latest version to
1089 play with."
1090
2e2a30b4
PH
1091PH/45 In a pipe transport, although a timeout while waiting for the pipe
1092 process to complete was treated as a delivery failure, a timeout while
1093 writing the message to the pipe was logged, but erroneously treated as a
1094 successful delivery. Such timeouts include transport filter timeouts. For
1095 consistency with the overall process timeout, these timeouts are now
1096 treated as errors, giving rise to delivery failures by default. However,
1097 there is now a new Boolean option for the pipe transport called
1098 timeout_defer, which, if set TRUE, converts the failures into defers for
1099 both kinds of timeout. A transport filter timeout is now identified in
1100 the log output.
1101
9176e9f0
PH
1102PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1103 systems where "make" and "gmake" are different, calling "gmake" at top
1104 level broke things. I've arranged for the value of $(MAKE) to be passed
1105 from the Makefile to this script so that it can call the same version of
1106 "make".
1107
7982096b 1108
bbe902f0
PH
1109A note about Exim versions 4.44 and 4.50
1110----------------------------------------
1111
1112Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1113changes of various kinds. As a consequence, a big documentation update was
1114needed. This delayed the release for rather longer than seemed good, especially
1115in the light of a couple of (minor) security issues. Therefore, the changes
1116that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1117release. So 4.44 and 4.50 are in effect two different branches that both start
1118from 4.43.
1119
1120I have left the 4.50 change log unchanged; it contains all the changes since
11214.43. The change log for 4.44 is below; many of its items are identical to
1122those for 4.50. This seems to be the most sensible way to preserve the
1123historical information.
1124
1125
f7b63901 1126Exim version 4.50
495ae4b0
PH
1127-----------------
1128
5fe762f6
PH
1129 1. Minor wording change to the doc/README.SIEVE file.
1130
139059f6 1131 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1132 computation of the current number of files was incorrect.
495ae4b0 1133
7086e875
PH
1134 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1135 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1136 place.
1137
35af9f61
PH
1138 4. Give more explanation in the error message when the command for a transport
1139 filter fails to execute.
1140
b668c215
PH
1141 5. There are several places where Exim runs a non-Exim command in a
1142 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1143 was being done only for the command run by the queryprogram router. It is
1144 now done for all such subprocesses. The other cases are: ${run, transport
1145 filters, and the commands run by the lmtp and pipe transports.
1146
a494b1e1
PH
1147 6. Added CONFIGURE_GROUP build-time option.
1148
1149 7. Some older OS have a limit of 256 on the maximum number of file
1150 descriptors. Exim was using setrlimit() to set 1000 as a large value
1151 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1152 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1153
c5fcb476
PH
1154 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1155 was an oversight, and furthermore, ever since the addition of extra
1156 controls (e.g. 4.43/32), the checks on when to allow different forms of
1157 "control" were broken. There should now be diagnostics for all cases when a
1158 control that does not make sense is encountered.
1159
69358f02
PH
1160 9. Added the /retain_sender option to "control=submission".
1161
5be20824
PH
116210. $recipients is now available in the predata ACL (oversight).
1163
eb2c0248
PH
116411. Tidy the search cache before the fork to do a delivery from a message
1165 received from the command line. Otherwise the child will trigger a lookup
1166 failure and thereby defer the delivery if it tries to use (for example) a
1167 cached ldap connection that the parent has called unbind on.
1168
2a3eea10
PH
116912. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1170 of $address_data from the recipient verification was clobbered by the
1171 sender verification.
1172
117313. The value of address_data from a sender verification is now available in
1174 $sender_address_data in subsequent conditions in the ACL statement.
1175
23c7ff99
PH
117614. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1177
4deaf07d
PH
117815. Added a new option "connect=<time>" to callout options, to set a different
1179 connection timeout.
1180
926e1192
PH
118116. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1182 was its contents. (It was OK if the option was not defined at all.)
1183
650edc6f
PH
118417. A "Completed" log line is now written for messages that are removed from
1185 the spool by the -Mrm option.
1186
2c7db3f5
PH
118718. New variables $sender_verify_failure and $recipient_verify_failure contain
1188 information about exactly what failed.
1189
3d235903
PH
119019. Added -dd to debug only the daemon process.
1191
7c7ad977
PH
119220. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1193 handles timeouts, both on the server side and network timeouts. Renamed the
1194 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1195
981756db
PH
119621. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1197 It is now set to "smtps".
1198
d4eb88df
PH
119922. $host_address is now set to the target address during the checking of
1200 ignore_target_hosts.
1201
120223. When checking ignore_target_hosts for an ipliteral router, no host name was
1203 being passed; this would have caused $sender_host_name to have been used if
1204 matching the list had actually called for a host name (not very likely,
1205 since this list is usually IP addresses). A host name is now passed as
1206 "[x.x.x.x]".
1207
7d468ab8
PH
120824. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1209 code that specifies a non-restarting handler (typically sigaction() in
1210 modern systems) in an attempt to fix a rare and obscure crash bug.
1211
121225. Narrowed the window for a race in the daemon that could cause it to ignore
1213 SIGCHLD signals. This is not a major problem, because they are used only to
1214 wake it up if nothing else does.
1215
62c0818f
PH
121626. A malformed maildirsize file could cause Exim to calculate negative values
1217 for the mailbox size or file count. Odd effects could occur as a result.
1218 The maildirsize information is now recalculated if the size or filecount
1219 end up negative.
1220
26034054
PH
122127. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1222 support for a long time. Removed HAVE_SYS_VFS_H.
1223
af66f652
PH
122428. Installed the latest version of exipick from John Jetmore.
1225
90af77f4
PH
122629. In an address list, if the pattern was not a regular expression, an empty
1227 subject address (from a bounce message) matched only if the pattern was an
1228 empty string. Non-empty patterns were not even tested. This was the wrong
1229 because it is perfectly reasonable to use an empty address as part of a
1230 database query. An empty address is now tested by patterns that are
1231 lookups. However, all the other forms of pattern expect the subject to
1232 contain a local part and a domain, and therefore, for them, an empty
1233 address still always fails if the pattern is not itself empty.
1234
d8ef3577
PH
123530. Exim went into a mad DNS loop when attempting to do a callout where the
1236 host was specified on an smtp transport, and looking it up yielded more
1237 than one IP address.
1238
5cb8cbc6
PH
123931. Re-factored the code for checking spool and log partition space into a
1240 function that finds that data and another that does the check. The former
1241 is then used to implement four new variables: $spool_space, $log_space,
1242 $spool_inodes, and $log_inodes.
1243
14702f5b
PH
124432. The RFC2047 encoding function was originally intended for short strings
1245 such as real names; it was not keeping to the 75-character limit for
1246 encoded words that the RFC imposes. It now respects the limit, and
1247 generates multiple encoded words if necessary. To be on the safe side, I
1248 have increased the buffer size for the ${rfc2047: expansion operator from
1249 1024 to 2048 bytes.
1250
063b1e99
PH
125133. It is now permitted to omit both strings after an "if" condition; if the
1252 condition is true, the result is "true". As before, when the second string
1253 is omitted, a false condition yields an empty string. This makes it less
1254 cumbersome to write custom ACL and router conditions.
1255
652e1b65
PH
125634. Failure to deliver a bounce message always caused it to be frozen, even if
1257 there was an errors_to setting on the router. The errors_to setting is now
1258 respected.
1259
6f0c9a4f
PH
126035. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1261 canonical form (fully expanded) before being placed in
1262 $sender_host_address.
1263
33397d19
PH
126436. The table in the code that translates DNS record types into text (T_A to
1265 "A" for instance) was missing entries for NS and CNAME. It is just possible
1266 that this could have caused confusion if both these types were looked up
1267 for the same domain, because the text type is used as part of Exim's
1268 per-process caching. But the chance of anyone hitting this buglet seems
1269 very small.
1270
7bb56e1f
PH
127137. The dnsdb lookup has been extended in a number of ways.
1272
1273 (1) There is a new type, "zns", which walks up the domain tree until it
1274 finds some nameserver records. It should be used with care.
1275
ea3bc19b
PH
1276 (2) There is a new type, "mxh", which is like "mx" except that it returns
1277 just the host names, not the priorities.
1278
1279 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1280 looked up. The behaviour when one of the lookups defers can be
1281 controlled by a keyword.
7bb56e1f 1282
ea3bc19b 1283 (4) It is now possible to specify the separator character for use when
7bb56e1f 1284 multiple records are returned.
33397d19 1285
0bcb2a0e
PH
128638. The dnslists ACL condition has been extended: it is now possible to supply
1287 a list of IP addresses and/or domains to be looked up in a particular DNS
1288 domain.
1289
2ac0e484
PH
129039. Added log_selector=+queue_time_overall.
1291
4e1fde53
PH
129240. When running the queue in the test harness, wait just a tad after forking a
1293 delivery process, to get repeatability of debugging output.
1294
de365ded
PH
129541. Include certificate and key file names in error message when GnuTLS fails
1296 to set them up, because the GnuTLS error message doesn't include the name
1297 of the failing file when there is a problem reading it.
1298
f05da2e8
PH
129942. Allow both -bf and -bF in the same test run.
1300
d6453af2
PH
130143. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1302
f7b63901
PH
130344. The "Exiscan patch" is now merged into the mainline Exim source.
1304
130545. Sometimes the final signoff response after QUIT could fail to get
1306 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1307 < 0 before doing a fflush(). This bug looks as though it goes back to the
1308 introduction of TLS in release 3.20, but "sometimes" must have been rare
1309 because the tests only now provoked it.
1310
a444213a
PH
131146. Reset the locale to "C" after calling embedded Perl, in case it was changed
1312 (this can affect the format of dates).
1313
0ec020ea
PH
131447. exim_tidydb, when checking for the continued existence of a message for
1315 which it has found a message-specific retry record, was not finding
1316 messages that were in split spool directories. Consequently, it was
1317 deleting retry records that should have stayed in existence.
1318
b1206957
PH
131948. Steve fixed some bugs in eximstats.
1320
132149. The SPA authentication driver was not abandoning authentication and moving
1322 on to the next authenticator when an expansion was forced to fail,
1323 contradicting the general specification for all authenticators. Instead it
1324 was generating a temporary error. It now behaves as specified.
1325
26dd5a95
PH
132650. The default ordering of permitted cipher suites for GnuTLS was pessimal
1327 (the order specifies the preference for clients). The order is now AES256,
1328 AES128, 3DES, ARCFOUR128.
1329
343b2385
PH
133051. Small patch to Sieve code - explicitly set From: when generating an
1331 autoreply.
1332
1c5466b9
PH
133352. Exim crashed if a remote delivery caused a very long error message to be
1334 recorded - for instance if somebody sent an entire SpamAssassin report back
1335 as a large number of 550 error lines. This bug was coincidentally fixed by
1336 increasing the size of one of Exim's internal buffers (big_buffer) that
1337 happened as part of the Exiscan merge. However, to be on the safe side, I
1338 have made the code more robust (and fixed the comments that describe what
1339 is going on).
1340
55ee9ee3
PH
134153. Now that there can be additional text after "Completed" in log lines (if
1342 the queue_time_overall log selector is set), a one-byte patch to exigrep
1343 was needed to allow it to recognize "Completed" as not the last thing in
1344 the line.
1345
d38f8232
PH
134654. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1347 patch that reportedly fixes this has been added. I am not expert enough to
1348 create a test for it. This is what the patch creator wrote:
1349
1350 "I found a little strange behaviour of ldap code when working with
1351 Windows 2003 AD Domain, where users was placed in more than one
1352 Organization Units. When I tried to give exim partial DN, the exit code
1353 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1354 But simultaneously result of request was absolutely normal ldap result,
1355 so I produce this patch..."
1356
3295e65b
PH
1357 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1358 REFERENCE, so I have modified the code to exclude the patch when that macro
1359 is not defined.
1360
7102e136
PH
136155. Some experimental protocols are using DNS PTR records for new purposes. The
1362 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1363 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1364 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1365 CAN-2005-0021
7102e136 1366
3ca0ba97
PH
136756. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1368
c2bcbe20
PH
136957. Double the size of the debug message buffer (to 2048) so that more of very
1370 long debug lines gets shown.
1371
18ce445d
PH
137258. The exicyclog utility now does better if the number of log files to keep
1373 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1374
1f5b4c3d
PH
137559. Two changes related to the smtp_active_hostname option:
1376
1377 (1) $smtp_active_hostname is now available as a variable.
1378 (2) The default for smtp_banner uses $smtp_active_hostname instead
1379 of $primary_hostname.
1380
b975ba52
PH
138160. The host_aton() function is supposed to be passed a string that is known
1382 to be a valid IP address. However, in the case of IPv6 addresses, it was
1383 not checking this. This is a hostage to fortune. Exim now panics and dies
1384 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1385 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1386 components; fortuitously, this particular loophole had already been fixed
1387 by change 4.50/55 above.
1388
1389 If there are any other similar loopholes, the new check in host_aton()
1390 itself should stop them being exploited. The report I received stated that
1391 data on the command line could provoke the exploit when Exim was running as
1392 exim, but did not say which command line option was involved. All I could
1393 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1394 running as the user.
ea3a6f44 1395 CAN-2005-0021
85b87bc2
PH
1396
139761. There was a buffer overflow vulnerability in the SPA authentication code
1398 (which came originally from the Samba project). I have added a test to the
1399 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1400 CAN-2005-0022
b975ba52 1401
17ffcae7
PH
140262. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1403 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1404
d95f9fdb
PH
140563. The daemon start-up calls getloadavg() while still root for those OS that
1406 need the first call to be done as root, but it missed one case: when
1407 deliver_queue_load_max is set with deliver_drop_privilege. This is
1408 necessary for the benefit of the queue runner, because there is no re-exec
1409 when deliver_drop_privilege is set.
1410
86b8287f
PH
141164. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1412 This has been fixed.
1413
60dc5e56
PH
141465. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1415 was in use, was not putting the data itself into the right store pool;
1416 consequently, it could be overwritten for a subsequent message in the same
1417 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1418 the caching.)
1419
533244af
PH
142066. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1421
a5a28604
PH
142267. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1423 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1424 the function were treating the return as a boolean value, which happened to
1425 work because 0=false and not-0=true, but is not correct code.
1426
7e634d24
PH
142768. The host_aton() function was not handling scoped IPv6 addresses (those
1428 with, for example, "%eth0" on the end) correctly.
1429
3e11c26b
PH
143069. Fixed some compiler warnings in acl.c for the bitmaps specified with
1431 negated items (that is, ~something) in unsigned ints. Some compilers
1432 apparently mutter when there is no cast.
1433
6729cf78
PH
143470. If an address verification called from an ACL failed, and did not produce a
1435 user-specific message (i.e. there was only a "system" message), nothing was
1436 put in $acl_verify_message. In this situation, it now puts the system
1437 message there.
1438
00f00ca5
PH
143971. Change 4.23/11 added synchronization checking at the start of an SMTP
1440 session; change 4.31/43 added the unwanted input to the log line - except
1441 that it did not do this in the start of session case. It now does.
1442
c9bdd01c
PH
144372. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1444 This is wrong and can cause the other end to generate a synchronization
1445 error if it is another Exim or anything else that does the synchronization
1446 check. A QUIT command is no longer sent after a timeout.
1447
d43194df
PH
144873. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1449 during host lookups.
1450
fe5b5d0b
PH
145174. The defer_ok option of callout verification was not working if it was used
1452 when verifying addresses in header lines, that is, for this case:
1453
1454 verify = header_sender/callout=defer_ok
1455
76a2d7ba
PH
145675. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1457 those file descriptors could be used for SMTP connections. If anything
1458 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1459 could be sent to the SMTP client, causing chaos. The daemon now opens
1460 stdin, stdout, and stderr to /dev/null when it puts itself into the
1461 background.
1462
146376. Arrange for output from Perl's "warn" command to be written to Exim's main
1464 log by default. The user can override this with suitable Perl magic.
1465
04f7d5b9
PH
146677. The use of log_message on a "discard" ACL verb, which is supposed to add to
1467 the log message when discard triggers, was not working for the DATA ACL or
1468 for the non-SMTP ACL.
1469
bc60667e
PH
147078. Error message wording change in sieve.c.
1471
bb6e88ff
PH
147279. If smtp_accept_max_per_host was set, the number of connections could be
1473 restricted to fewer than expected, because the daemon was trying to set up
1474 a new connection before checking whether the processes handling previous
1475 connections had finished. The check for completed processes is now done
1476 earlier. On busy systems, this bug wouldn't be noticed because something
1477 else would have woken the daemon, and it would have reaped the completed
1478 process earlier.
1479
1e70f85b
PH
148080. If a message was submitted locally by a user whose login name contained one
1481 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1482 It caused a spool format error. I have fixed the spool reading code. A
1483 related problem was that the "from" clause in the Received: line became
1484 illegal because of the space(s). It is now covered by ${quote_local_part.
1485
148681. Included the latest eximstats from Steve (adds average sizes to HTML Top
1487 tables).
1488
4e01f9d6
PH
148982. Updated OS/Makefile-AIX as per message from Mike Meredith.
1490
1ee1cef2
PH
149183. Patch from Sieve maintainer to fix unterminated string problem in
1492 "vacation" handling.
1493
6e2b4ccc
PH
149484. Some minor changes to the Linux configuration files to help with other
1495 OS variants using glibc.
1496
8e669ac1
PH
149785. One more patch for Sieve to update vacation handling to latest spec.
1498
495ae4b0 1499
bbe902f0
PH
1500----------------------------------------------------
1501See the note above about the 4.44 and 4.50 releases.
1502----------------------------------------------------
1503
1504
1505Exim version 4.44
1506-----------------
1507
1508 1. Change 4.43/35 introduced a bug that caused file counts to be
1509 incorrectly computed when quota_filecount was set in an appendfile
1510 transport
1511
1512 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1513 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1514 place.
1515
1516 3. Give more explanation in the error message when the command for a transport
1517 filter fails to execute.
1518
1519 4. There are several places where Exim runs a non-Exim command in a
1520 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1521 was being done only for the command run by the queryprogram router. It is
1522 now done for all such subprocesses. The other cases are: ${run, transport
1523 filters, and the commands run by the lmtp and pipe transports.
1524
1525 5. Some older OS have a limit of 256 on the maximum number of file
1526 descriptors. Exim was using setrlimit() to set 1000 as a large value
1527 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1528 systems. I've change it so that if it can't get 1000, it tries for 256.
1529
1530 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1531 was an oversight, and furthermore, ever since the addition of extra
1532 controls (e.g. 4.43/32), the checks on when to allow different forms of
1533 "control" were broken. There should now be diagnostics for all cases when a
1534 control that does not make sense is encountered.
1535
1536 7. $recipients is now available in the predata ACL (oversight).
1537
1538 8. Tidy the search cache before the fork to do a delivery from a message
1539 received from the command line. Otherwise the child will trigger a lookup
1540 failure and thereby defer the delivery if it tries to use (for example) a
1541 cached ldap connection that the parent has called unbind on.
1542
1543 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1544 of $address_data from the recipient verification was clobbered by the
1545 sender verification.
1546
154710. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1548 was its contents. (It was OK if the option was not defined at all.)
1549
155011. A "Completed" log line is now written for messages that are removed from
1551 the spool by the -Mrm option.
1552
155312. $host_address is now set to the target address during the checking of
1554 ignore_target_hosts.
1555
155613. When checking ignore_target_hosts for an ipliteral router, no host name was
1557 being passed; this would have caused $sender_host_name to have been used if
1558 matching the list had actually called for a host name (not very likely,
1559 since this list is usually IP addresses). A host name is now passed as
1560 "[x.x.x.x]".
1561
156214. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1563 code that specifies a non-restarting handler (typically sigaction() in
1564 modern systems) in an attempt to fix a rare and obscure crash bug.
1565
156615. Narrowed the window for a race in the daemon that could cause it to ignore
1567 SIGCHLD signals. This is not a major problem, because they are used only to
1568 wake it up if nothing else does.
1569
157016. A malformed maildirsize file could cause Exim to calculate negative values
1571 for the mailbox size or file count. Odd effects could occur as a result.
1572 The maildirsize information is now recalculated if the size or filecount
1573 end up negative.
1574
157517. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1576 support for a long time. Removed HAVE_SYS_VFS_H.
1577
ea3a6f44 157818. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1579
158019. Allow an empty sender to be matched against a lookup in an address list.
1581 Previously the only cases considered were a regular expression, or an
1582 empty pattern.
1583
158420. Exim went into a mad DNS lookup loop when doing a callout where the
1585 host was specified on the transport, if the DNS lookup yielded more than
1586 one IP address.
1587
ea3a6f44
NM
158821. The RFC2047 encoding function was originally intended for short strings
1589 such as real names; it was not keeping to the 75-character limit for
1590 encoded words that the RFC imposes. It now respects the limit, and
1591 generates multiple encoded words if necessary. To be on the safe side, I
1592 have increased the buffer size for the ${rfc2047: expansion operator from
1593 1024 to 2048 bytes.
bbe902f0 1594
ea3a6f44
NM
159522. Failure to deliver a bounce message always caused it to be frozen, even if
1596 there was an errors_to setting on the router. The errors_to setting is now
1597 respected.
bbe902f0
PH
1598
159923. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1600 canonical form (fully expanded) before being placed in
1601 $sender_host_address.
1602
160324. Updated eximstats to version 1.33
1604
ea3a6f44
NM
160525. Include certificate and key file names in error message when GnuTLS fails
1606 to set them up, because the GnuTLS error message doesn't include the name
1607 of the failing file when there is a problem reading it.
bbe902f0
PH
1608
160926. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1610 As per change 25.
bbe902f0 1611
ea3a6f44
NM
161227. Reset the locale to "C" after calling embedded Perl, in case it was changed
1613 (this can affect the format of dates).
bbe902f0 1614
ea3a6f44
NM
161528. exim_tidydb, when checking for the continued existence of a message for
1616 which it has found a message-specific retry record, was not finding
1617 messages that were in split spool directories. Consequently, it was
1618 deleting retry records that should have stayed in existence.
bbe902f0
PH
1619
162029. eximstats updated to version 1.35
1621 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1622 1.35 - bugfix such that pie charts by volume are generated correctly
1623
ea3a6f44
NM
162430. The SPA authentication driver was not abandoning authentication and moving
1625 on to the next authenticator when an expansion was forced to fail,
1626 contradicting the general specification for all authenticators. Instead it
1627 was generating a temporary error. It now behaves as specified.
bbe902f0 1628
ea3a6f44
NM
162931. The default ordering of permitted cipher suites for GnuTLS was pessimal
1630 (the order specifies the preference for clients). The order is now AES256,
1631 AES128, 3DES, ARCFOUR128.
bbe902f0 1632
ea3a6f44
NM
163331. Small patch to Sieve code - explicitly set From: when generating an
1634 autoreply.
bbe902f0 1635
ea3a6f44
NM
163632. Exim crashed if a remote delivery caused a very long error message to be
1637 recorded - for instance if somebody sent an entire SpamAssassin report back
1638 as a large number of 550 error lines. This bug was coincidentally fixed by
1639 increasing the size of one of Exim's internal buffers (big_buffer) that
1640 happened as part of the Exiscan merge. However, to be on the safe side, I
1641 have made the code more robust (and fixed the comments that describe what
1642 is going on).
bbe902f0 1643
ea3a6f44
NM
164433. Some experimental protocols are using DNS PTR records for new purposes. The
1645 keys for these records are domain names, not reversed IP addresses. The
1646 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1647 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1648 CAN-2005-0021
1649
ea3a6f44
NM
165034. The host_aton() function is supposed to be passed a string that is known
1651 to be a valid IP address. However, in the case of IPv6 addresses, it was
1652 not checking this. This is a hostage to fortune. Exim now panics and dies
1653 if the condition is not met. A case was found where this could be provoked
1654 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1655 components; fortuitously, this particular loophole had already been fixed
1656 by change 4.50/55 or 4.44/33 above.
1657
1658 If there are any other similar loopholes, the new check in host_aton()
1659 itself should stop them being exploited. The report I received stated that
1660 data on the command line could provoke the exploit when Exim was running as
1661 exim, but did not say which command line option was involved. All I could
1662 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1663 running as the user.
bbe902f0
PH
1664 CAN-2005-0021
1665
ea3a6f44
NM
166635. There was a buffer overflow vulnerability in the SPA authentication code
1667 (which came originally from the Samba project). I have added a test to the
1668 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1669 CAN-2005-0022
1670
ea3a6f44
NM
167136. The daemon start-up calls getloadavg() while still root for those OS that
1672 need the first call to be done as root, but it missed one case: when
1673 deliver_queue_load_max is set with deliver_drop_privilege. This is
1674 necessary for the benefit of the queue runner, because there is no re-exec
1675 when deliver_drop_privilege is set.
bbe902f0 1676
ea3a6f44
NM
167737. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1678 was in use, was not putting the data itself into the right store pool;
1679 consequently, it could be overwritten for a subsequent message in the same
1680 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1681 the caching.)
bbe902f0 1682
ea3a6f44
NM
168338. Sometimes the final signoff response after QUIT could fail to get
1684 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1685 < 0 before doing a fflush(). This bug looks as though it goes back to the
1686 introduction of TLS in release 3.20, but "sometimes" must have been rare
1687 because the tests only now provoked it.
bbe902f0
PH
1688
1689
495ae4b0
PH
1690Exim version 4.43
1691-----------------
1692
1693 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1694 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1695 Now it returns an integer. A number of calls were still expecting a T/F
1696 return. Fortuitously, in all cases, the tests worked in OK situations,
1697 which is the norm. However, things would have gone wrong on any write
1698 failures on the smtp file descriptor. This function is used when sending
1699 messages over SMTP and also when doing verify callouts.
1700
1701 2. When Exim is called to do synchronous delivery of a locally submitted
1702 message (the -odf or -odi options), it no longer closes stderr before doing
1703 the delivery.
1704
1705 3. Implemented the mua_wrapper option.
1706
1707 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1708
1709 5. Implemented the functions header_remove(), header_testname(),
1710 header_add_at_position(), and receive_remove_recipient(), and exported them
1711 to local_scan().
1712
1713 6. If an ACL "warn" statement specified the addition of headers, Exim already
1714 inserted X-ACL-Warn: at the start if there was no header name. However, it
1715 was not making this test for the second and subsequent header lines if
1716 there were newlines in the string. This meant that an invalid header could
1717 be inserted if Exim was badly configured.
1718
1719 7. Allow an ACL "warn" statement to add header lines at the start or after all
1720 the Received: headers, as well as at the end.
1721
1722 8. Added the rcpt_4xx retry error code.
1723
1724 9. Added postmaster_mailfrom=xxx to callout verification option.
1725
172610. Added mailfrom=xxxx to the callout verification option, for verify=
1727 header_sender only.
1728
172911. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1730 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1731
173212. Inserted some casts to stop certain compilers warning when using pointer
1733 differences as field lengths or precisions in printf-type calls (mostly
1734 affecting debugging statements).
1735
173613. Added optional readline() support for -be (dynamically loaded).
1737
173814. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1739 same clock tick as a message's arrival, so that its received time was the
1740 same as the "first fail" time on the retry record, and that message
1741 remained on the queue past the ultimate address timeout, every queue runner
1742 would try a delivery (because it was past the ultimate address timeout) but
1743 after another failure, the ultimate address timeout, which should have then
1744 bounced the address, did not kick in. This was a "< instead of <=" error;
1745 in most cases the first failure would have been in the next clock tick
1746 after the received time, and all would be well.
1747
174815. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1749 being recognized when the domain list was tested by the match_domain
1750 condition in an expansion string.
1751
175216. Added the ${str2b64: operator.
1753
175417. Exim was always calling setrlimit() to set a large limit for the number of
1755 processes, without checking whether the existing limit was already
1756 adequate. (It did check for the limit on file descriptors.) Furthermore,
1757 errors from getrlimit() and setrlimit() were being ignored. Now they are
1758 logged to the main and panic logs, but Exim does carry on, to try to do its
1759 job under whatever limits there are.
1760
176118. Imported PCRE 5.0.
1762
176319. Trivial typo in log message " temporarily refused connection" (the leading
1764 space).
1765
176620. If the log selector return_path_on_delivery was set and an address was
1767 redirected to /dev/null, the delivery process crashed because it assumed
1768 that a return path would always be set for a "successful" delivery. In this
1769 case, the whole delivery is bypassed as an optimization, and therefore no
1770 return path is set.
1771
177221. Internal re-arrangement: the function for sending a challenge and reading
1773 a response while authentication was assuming a zero-terminated challenge
1774 string. It's now changed to take a pointer and a length, to allow for
1775 binary data in such strings.
1776
177722. Added the cyrus_sasl authenticator (code supplied by MBM).
1778
177923. Exim was not respecting finduser_retries when seeking the login of the
1780 uid under which it was called; it was always trying 10 times. (The default
1781 setting of finduser_retries is zero.) Also, it was sleeping after the final
1782 failure, which is pointless.
1783
178424. Implemented tls_on_connect_ports.
1785
178625. Implemented acl_smtp_predata.
1787
178826. If the domain in control=submission is set empty, Exim assumes that the
1789 authenticated id is a complete email address when it generates From: or
1790 Sender: header lines.
1791
179227. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1793 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1794 chown and chgrp in /bin and hostname in /usr/bin.
1795
179628. Exim was keeping the "process log" file open after each use, just as it
1797 does for the main log. This opens the possibility of it remaining open for
1798 long periods when the USR1 signal hits a daemon. Occasional processlog
1799 errors were reported, that could have been caused by this. Anyway, it seems
1800 much more sensible not to leave this file open at all, so that is what now
1801 happens.
1802
180329. The long-running daemon process does not normally write to the log once it
1804 has entered its main loop, and it closes the log before doing so. This is
1805 so that log files can straightforwardly be renamed and moved. However,
1806 there are a couple of unusual error situations where the daemon does write
1807 log entries, and I had neglected to close the log afterwards.
1808
180930. The text of an SMTP error response that was received during a remote
1810 delivery was being truncated at 512 bytes. This is too short for some of
1811 the long messages that one sometimes sees. I've increased the limit to
1812 1024.
1813
181431. It is now possible to make retry rules that apply only when a message has a
1815 specific sender, in particular, an empty sender.
1816
181732. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1818 it possible to be selective about when SMTP synchronization is enforced.
1819
182033. Added "control = caseful_local_part" and "control = "caselower_local_part".
1821
182232. Implemented hosts_connection_nolog.
1823
182433. Added an ACL for QUIT.
1825
182634. Setting "delay_warning=" to disable warnings was not working; it gave a
1827 syntax error.
1828
182935. Added mailbox_size and mailbox_filecount to appendfile.
1830
183136. Added control = no_multiline_responses to ACLs.
1832
183337. There was a bug in the logic of the code that waits for the clock to tick
1834 in the case where the clock went backwards by a substantial amount such
1835 that the microsecond fraction of "now" was more than the microsecond
1836 fraction of "then" (but the whole seconds number was less).
1837
183838. Added support for the libradius Radius client library this is found on
1839 FreeBSD (previously only the radiusclient library was supported).
1840
1841
1842Exim version 4.42
1843-----------------
1844
1845 1. When certain lookups returned multiple values in the form name=value, the
1846 quoting of the values was not always being done properly. Specifically:
1847 (a) If the value started with a double quote, but contained no whitespace,
1848 it was not quoted.
1849 (b) If the value contained whitespace other than a space character (i.e.
1850 tabs or newlines or carriage returns) it was not quoted.
1851 This fix has been applied to the mysql and pgsql lookups by writing a
1852 separate quoting function and calling it from the lookup code. The fix
1853 should probably also be applied to nisplus, ibase and oracle lookups, but
1854 since I cannot test any of those, I have not disturbed their existing code.
1855
1856 2. A hit in the callout cache for a specific address caused a log line with no
1857 reason for rejecting RCPT. Now it says "Previous (cached) callout
1858 verification failure".
1859
1860 3. There was an off-by-one bug in the queryprogram router. An over-long
1861 return line was truncated at 256 instead of 255 characters, thereby
1862 overflowing its buffer with the terminating zero. As well as fixing this, I
1863 have increased the buffer size to 1024 (and made a note to document this).
1864
1865 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1866 when Exim is waiting for an SMTP response from a remote server, Exim
1867 restarts its select() call on the socket, thereby resetting its timeout.
1868 This is not a problem when such interrupts are rare. Somebody set up a cron
1869 job to run exiwhat every 2 minutes, which is less than the normal select()
1870 timeout (5 or 10 minutes). This meant that the select() timeout never
1871 kicked in because it was always reset. I have fixed this by comparing the
1872 time when an interrupt arrives with the time at the start of the first call
1873 to select(). If more time than the timeout has elapsed, the interrupt is
1874 treated as a timeout.
1875
1876 5. Some internal re-factoring in preparation for the addition of Sieve
1877 extensions (by MH). In particular, the "personal" test is moved to a
1878 separate function, and given an option for scanning Cc: and Bcc: (which is
1879 not set for Exim filters).
1880
1881 6. When Exim created an email address using the login of the caller as the
1882 local part (e.g. when creating a From: or Sender: header line), it was not
1883 quoting the local part when it contained special characters such as @.
1884
1885 7. Installed new OpenBSD configuration files.
1886
1887 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1888 try to make them clearer.
1889
1890 9. Callout options, other than the timeout value, were being ignored when
1891 verifying sender addresses in header lines. For example, when using
1892
1893 verify = header_sender/callout=no_cache
1894
1895 the cache was (incorrectly) being used.
1896
189710. Added a missing instance of ${EXE} to the exim_install script; this affects
1898 only the Cygwin environment.
1899
190011. When return_path_on_delivery was set as a log selector, if different remote
1901 addresses in the same message used different return paths and parallel
1902 remote delivery occurred, the wrong values would sometimes be logged.
1903 (Whenever a remote delivery process finished, the return path value from
1904 the most recently started remote delivery process was logged.)
1905
190612. RFC 3848 specifies standard names for the "with" phrase in Received: header
1907 lines when AUTH and/or TLS are in use. This is the "received protocol"
1908 field. Exim used to use "asmtp" for authenticated SMTP, without any
1909 indication (in the protocol name) for TLS use. Now it follows the RFC and
1910 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1911 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1912 names appear in log lines as well as in Received: header lines.
1913
191413. Installed MH's patches for Sieve to add the "copy" and "vacation"
1915 extensions, and comparison tests, and to fix some bugs.
1916
191714. Changes to the "personal" filter test:
1918
1919 (1) The test was buggy in that it was just doing the equivalent of
1920 "contains" tests on header lines. For example, if a user's address was
1921 anne@some.where, the "personal" test would incorrectly be true for
1922
1923 To: susanne@some.where
1924
1925 This test is now done by extracting each address from the header in turn,
1926 and checking the entire address. Other tests that are part of "personal"
1927 are now done using regular expressions (for example, to check local parts
1928 of addresses in From: header lines).
1929
1930 (2) The list of non-personal local parts in From: addresses has been
1931 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1932 taken from the Sieve specification recommendations.
1933
1934 (3) If the message contains any header line starting with "List-" it is
1935 treated as non-personal.
1936
1937 (4) The test for "circular" in the Subject: header line has been removed
1938 because it now seems ill-conceived.
1939
194015. Minor typos in src/EDITME comments corrected.
1941
194216. Installed latest exipick from John Jetmore.
1943
194417. If headers_add on a router specified a text string that was too long for
1945 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1946 of string_sprintf() is now avoided.
1947
194818. $message_body_size was not set (it was always zero) when running the DATA
1949 ACL and the local_scan() function.
1950
195119. For the "mail" command in an Exim filter, no default was being set for
1952 the once_repeat time, causing a random time value to be used if "once" was
1953 specified. (If the value happened to be <= 0, no repeat happened.) The
1954 default is now 0s, meaning "never repeat". The "vacation" command was OK
1955 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1956 (I found it when inspecting the code).
1957
195820. There is now an overall timeout for performing a callout verification. It
1959 defaults to 4 times the callout timeout, which applies to individual SMTP
1960 commands during the callout. The overall timeout applies when there is more
1961 than one host that can be tried. The timeout is checked before trying the
1962 next host. This prevents very long delays if there are a large number of
1963 hosts and all are timing out (e.g. when the network connections are timing
1964 out). The value of the overall timeout can be changed by specifying an
1965 additional sub-option for "callout", called "maxwait". For example:
1966
1967 verify = sender/callout=5s,maxwait=20s
1968
196921. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1970 to the end before writing, but this should make it even safer).
1971
197222. Exim was forgetting that it had advertised PIPELINING for the second and
1973 subsequent messages on an SMTP connection. It was also not resetting its
1974 memory on STARTTLS and an internal HELO.
1975
197623. When Exim logs an SMTP synchronization error within a session, it now
1977 records whether PIPELINING has been advertised or not.
1978
197924. Added 3 instances of "(long int)" casts to time_t variables that were being
1980 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1981 rather than long int.
1982
198325. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1984
198526. Added the never_mail option to autoreply.
1986
1987
1988Exim version 4.41
1989-----------------
1990
1991 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1992 crash if the getsockname() call failed; this can happen if a connection is
1993 closed very soon after it is established. The problem was simply in the
1994 order in which certain operations were done, causing Exim to try to write
1995 to the SMTP stream before it had set up the file descriptor. The bug has
1996 been fixed by making things happen in the correct order.
1997
1998
1999Exim version 4.40
2000-----------------
2001
2002 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2003 before the connection was closed, thus losing the rejection response.
2004
2005 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2006 some early Solaris releases, but causes trouble in current releases where
2007 socklen_t is defined.
2008
2009 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2010 always exist.
2011
2012 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2013 configured.
2014
2015 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2016 that releasing the top of it at the end releases what was used for sub-
2017 expansions (unless the block got too big). However, discard this block if
2018 the first thing is a variable or header, so that we can use its block when
2019 it is dynamic (useful for very large $message_headers, for example).
2020
2021 6. Lookups now cache *every* query, not just the most recent. A new, separate
2022 store pool is used for this. It can be recovered when all lookup caches are
2023 flushed. Lookups now release memory at the end of their result strings.
2024 This has involved some general refactoring of the lookup sources.
2025
2026 7. Some code has been added to the store_xxx() functions to reduce the amount
2027 of flapping under certain conditions.
2028
2029 8. log_incoming_interface used to affect only the <= reception log lines. Now
2030 it causes the local interface and port to be added to several more SMTP log
2031 lines, for example "SMTP connection from", and rejection lines.
2032
2033 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2034
203510. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2036
203711. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2038 could be overwritten at the end of the current message (or the start of a
2039 new message if it was set in a HELO ACL). The value is now preserved for
2040 the duration of the SMTP connection.
2041
204212. If a transport had a headers_rewrite setting, and a matching header line
2043 contained an unqualified address, that address was qualified, even if it
2044 did not match any rewriting rules. The underlying bug was that the values
2045 of the flags that permit the existence of unqualified sender and recipient
2046 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2047 non-local messages, and by -bnq for local messages) were not being
2048 preserved with the message after it was received.
2049
205013. When Exim was logging an SMTP synchronization error, it could sometimes log
2051 "next input=" as part of the text comprising the host identity instead of
2052 the correct text. The code was using the same buffer for two different
2053 strings. However, depending on which order the printing function evaluated
2054 its arguments, the bug did not always show up. Under Linux, for example, my
2055 test suite worked just fine.
2056
205714. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2058 doesn't work with some older versions of Perl. It has been changed to "my",
2059 which in any case is probably the better facility to use.
2060
206115. A really picky compiler found some instances of statements for creating
2062 error messages that either had too many or two few arguments for the format
2063 string.
2064
206516. The size of the buffer for calls to the DNS resolver has been increased
2066 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2067 for addresses that have a lot of PTR records. This alleviates a problem; it
2068 does not fully solve it.
2069
207017. A dnsdb lookup for PTR records that receives more data than will fit in the
2071 buffer now truncates the list and logs the incident, which is the same
2072 action as happens when Exim is looking up a host name and its aliases.
2073 Previously in this situation something unpredictable would happen;
2074 sometimes it was "internal error: store_reset failed".
2075
207618. If a server dropped the connection unexpectedly when an Exim client was
2077 using GnuTLS and trying to read a response, the client delivery process
2078 crashed while trying to generate an error log message.
2079
208019. If a "warn" verb in an ACL added multiple headers to a message in a single
2081 string, for example:
2082
2083 warn message = H1: something\nH2: something
2084
2085 the text was added as a single header line from Exim's point of view
2086 though it ended up OK in the delivered message. However, searching for the
2087 second and subsequent header lines using $h_h2: did not work. This has been
2088 fixed. Similarly, if a system filter added multiple headers in this way,
2089 the routers could not see them.
2090
209120. Expanded the error message when iplsearch is called with an invalid key to
2092 suggest using net-iplsearch in a host list.
2093
209421. When running tests using -bh, any delays imposed by "delay" modifiers in
2095 ACLs are no longer actually imposed (and a message to that effect is
2096 output).
2097
209822. If a "gecos" field in a passwd entry contained escaped characters, in
2099 particular, if it contained a \" sequence, Exim got it wrong when building
2100 a From: or a Sender: header from that name. A second bug also caused
2101 incorrect handling when an unquoted " was present following a character
2102 that needed quoting.
2103
210423. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2105 was not being matched caselessly.
2106
210724. Arranged for all hyphens in the exim.8 source to be escaped with
2108 backslashes.
2109
211025. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2111 properly. Recipient callout cache records were still being keyed to include
2112 the sender, even when use_sender was set false. This led to far more
2113 callouts that were necessary. The sender is no longer included in the key
2114 when use_sender is false.
2115
211626. Added "control = submission" modifier to ACLs.
2117
211827. Added the ${base62d: operator to decode base 62 numbers.
2119
212028. dnsdb lookups can now access SRV records.
2121
212229. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2123 the configuration file.
2124
212530. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2126 (-v) mode. This makes the output for a verbose queue run more intelligible.
2127
212831. Added a use_postmaster feature to recipient callouts.
2129
213032. Added the $body_zerocount variable, containing the number of binary zero
2131 bytes in the message body.
2132
213333. The time of last modification of the "new" subdirectory is now used as the
2134 "mailbox time last read" when there is a quota error for a maildir
2135 delivery.
2136
213734. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2138
213935. Added +ignore_unknown as a special item in host lists.
2140
214136. Code for decoding IPv6 addresses in host lists is now included, even if
2142 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2143 address was recognized as an IP address, but was then not correctly decoded
2144 into binary, causing unexpected and incorrect effects when compared with
2145 another IP address.
2146
2147
2148Exim version 4.34
2149-----------------
2150
2151 1. Very minor rewording of debugging text in manualroute to say "list of
2152 hosts" instead of "hostlist".
2153
2154 2. If verify=header_syntax was set, and a header line with an unqualified
2155 address (no domain) and a large number of spaces between the end of the
2156 name and the colon was received, the reception process suffered a buffer
2157 overflow, and (when I tested it) crashed. This was caused by some obsolete
2158 code that should have been removed. The fix is to remove it!
2159
2160 3. When running in the test harness, delay a bit after writing a bounce
2161 message to get a bit more predictability in the log output.
2162
2163 4. Added a call to search_tidyup() just before forking a reception process. In
2164 theory, someone could use a lookup in the expansion of smtp_accept_max_
2165 per_host which, without the tidyup, could leave open a database connection.
2166
2167 5. Added the variables $recipient_data and $sender_data which get set from a
2168 lookup success in an ACL "recipients" or "senders" condition, or a router
2169 "senders" option, similar to $domain_data and $local_part_data.
2170
2171 6. Moved the writing of debug_print from before to after the "senders" test
2172 for routers.
2173
2174 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2175 problems for message scanning, either using a data ACL, or using
2176 local_scan() because the Received: header was not generated till after they
2177 were called (in order to set the time as the time of reception completion).
2178 I have revised the way this works. The header is now generated after the
2179 body is received, but before the ACL or local_scan() are called. After they
2180 are run, the timestamp in the header is updated.
2181
2182
2183Exim version 4.33
2184-----------------
2185
2186 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2187 before starting a queue runner without re-exec. This happened only when
2188 deliver_drop_privilege was set or when the Exim user was set to root. The
2189 effect of the bug was that timeouts during subsequent deliveries caused
2190 crashes instead of being properly handled. The handler is now left at its
2191 default (and expected) setting.
2192
2193 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2194 message, again when deliver_drop_privilege is set or Exim is run as root.
2195 The bug described in (1) was not present in this case, but the tidying up
2196 of the other signals was missing. I have made the two cases consistent.
2197
2198 3. The ignore_target_hosts setting on a manualroute router was being ignored
2199 for hosts that were looked up using the /MX notation.
2200
2201 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2202 in domain lists.
2203
2204 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2205 operated on the sender address. After changing the $sender_address to <>
2206 for the sender address verify, Exim was re-instated it as the original
2207 (before rewriting) address, but remembering that it had rewritten it, so it
2208 wasn't rewriting it again. This bug also had the effect of breaking the
2209 sender address verification caching when the sender address was rewritten.
2210
2211 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2212 This has been changed so that if the ip literal address matches
2213 ignore_target_hosts, the router declines.
2214
2215 7. Added expansion conditions match_domain, match_address, and match_local_
2216 part (NOT match_host).
2217
2218 8. The placeholder for the Received: header didn't have a length field set.
2219
2220 9. Added code to Exim itself and to exim_lock to test for a specific race
2221 condition that could lead to file corruption when using MBX delivery. The
2222 issue is with the lockfile that is created in /tmp. If this file is removed
2223 after a process has opened it but before that process has acquired a lock,
2224 there is the potential for a second process to recreate the file and also
2225 acquire a lock. This could lead to two Exim processes writing to the file
2226 at the same time. The added code performs the same test as UW imapd; it
2227 checks after acquiring the lock that its file descriptor still refers to
2228 the same named file.
2229
223010. The buffer for building added header lines was of fixed size, 8192 bytes.
2231 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2232 when Exim is built.
2233
223411. Added the smtp_active_hostname option. If used, this will typically be made
2235 to depend on the incoming interface address. Because $interface_address is
2236 not set up until the daemon has forked a reception process, error responses
2237 that can happen earlier (such as "too many connections") no longer contain
2238 a host name.
2239
224012. If an expansion in a condition on a "warn" statement fails because a lookup
2241 defers, the "warn" statement is abandoned, and the next ACL statement is
2242 processed. Previously this caused the whole ACL to be aborted.
2243
224413. Added the iplsearch lookup type.
2245
224614. Added ident_timeout as a log selector.
2247
224815. Added tls_certificate_verified as a log selector.
2249
225016. Added a global option tls_require_ciphers (compare the smtp transport
2251 option of the same name). This controls incoming TLS connections.
2252
225317. I finally figured out how to make tls_require_ciphers do a similar thing
2254 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2255 before starting the TLS session.
2256
225718. Tabs are now shown as \t in -bP output.
2258
225919. If the log selector return_path_on_delivery was set, Exim crashed when
2260 bouncing a message because it had too many Received: header lines.
2261
226220. If two routers both had headers_remove settings, and the first one included
2263 a superfluous trailing colon, the final name in the first list and the
2264 first name in the second list were incorrectly joined into one item (with a
2265 colon in the middle).
2266
2267
2268Exim version 4.32
2269-----------------
2270
2271 1. Added -C and -D options to the exinext utility, mainly to make it easier
2272 to include in the automated testing, but these could be helpful when
2273 multiple configurations are in use.
2274
2275 2. The exinext utility was not formatting the output nicely when there was
2276 an alternate port involved in the retry record key, nor when there was a
2277 message id as well (for retries that were specific to a specific message
2278 and a specific host). It was also confused by IPv6 addresses, because of
2279 the additional colons they contain. I have fixed the IPv4 problem, and
2280 patched it up to do a reasonable job for IPv6.
2281
2282 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2283 delivery, the log line now contains "pipelined" if PIPELINING was used.
2284
2285 4. An SMTP transport process used to panic and die if the bind() call to set
2286 an explicit outgoing interface failed. This has been changed; it is now
2287 treated in the same way as a connect() failure.
2288
2289 5. A reference to $sender_host_name in the part of a conditional expansion
2290 that was being skipped was still causing a DNS lookup. This no longer
2291 occurs.
2292
2293 6. The def: expansion condition was not recognizing references to header lines
2294 that used bh_ and bheader_.
2295
2296 7. Added the _cache feature to named lists.
2297
2298 8. The code for checking quota_filecount in the appendfile transport was
2299 allowing one more file than it should have been.
2300
2301 9. For compatibility with Sendmail, the command line option
2302
2303 -prval:sval
2304
2305 is equivalent to
2306
2307 -oMr rval -oMs sval
2308
2309 and sets the incoming protocol and host name (for trusted callers). The
2310 host name and its colon can be omitted when only the protocol is to be set.
2311 Note the Exim already has two private options, -pd and -ps, that refer to
2312 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2313 "s", but I don't think that's a major issue.
2314
231510. A number of refactoring changes to the code, none of which should affect
2316 Exim's behaviour:
2317
2318 (a) The number of logging options was getting close to filling up the
2319 32-bit word that was used as a bit map. I have split them into two classes:
2320 those that are passed in the argument to log_write(), and those that are
2321 only ever tested independently outside of that function. These are now in
2322 separate 32-bit words, so there is plenty of room for expansion again.
2323 There is no change in the user interface or the logging behaviour.
2324
2325 (b) When building, for example, log lines, the code previously used a
2326 macro that called string_cat() twice, in order to add two strings. This is
2327 not really sufficiently general. Furthermore, there was one instance where
2328 it was actually wrong because one of the argument was used twice, and in
2329 one call a function was used. (As it happened, calling the function twice
2330 did not affect the overall behaviour.) The macro has been replaced by a
2331 function that can join an arbitrary number of extra strings onto a growing
2332 string.
2333
2334 (c) The code for expansion conditions now uses a table and a binary chop
2335 instead of a serial search (which was left over from when there were very
2336 few conditions). Also, it now recognizes conditions like "pam" even when
2337 the relevant support is not compiled in: a suitably worded error message is
2338 given if an attempt is made to use such a condition.
2339
234011. Added ${time_interval:xxxxx}.
2341
234212. A bug was causing one of the ddress fields not to be passed back correctly
2343 from remote delivery subprocesses. The field in question was not being
2344 subsequently used, so this caused to problems in practice.
2345
234613. Added new log selectors queue_time and deliver_time.
2347
234814. Might have fixed a bug in maildirsizefile handling that threw up
2349 "unexpected character" debug warnings, and recalculated the data
2350 unnecessarily. In any case, I expanded the warning message to give more
2351 information.
2352
235315. Added the message "Restricted characters in address" to the statements in
2354 the default ACL that block characters like @ and % in local parts.
2355
235616. Change 71 for release 4.31 proved to be much less benign that I imagined.
2357 Three changes have been made:
2358
2359 (a) There was a serious bug; a negative response to MAIL caused the whole
2360 recipient domain to be cached as invalid, thereby blocking all messages
2361 to all local parts at the same domain, from all senders. This bug has
2362 been fixed. The domain is no longer cached after a negative response to
2363 MAIL if the sender used is not empty.
2364
2365 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2366 been restored.
2367
2368 (c) A new callout option, "use_sender" has been added for people who want
2369 the modified behaviour.
2370
2371
2372Exim version 4.31
2373-----------------
2374
2375 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2376 Larry Rosenman.
2377
2378 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2379 indeed breaks things for older releases.
2380
2381 3. Added additional logging to the case where there is a problem reading data
2382 from a filter that is running in a subprocess using a pipe, in order to
2383 try to track down a specific problem.
2384
2385 4. Testing facility fudge: when running in the test harness and attempting
2386 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2387 getting "No route to host". Convert this to a timeout.
2388
2389 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2390 warning.
2391
2392 6. Some OS don't have socklen_t but use size_t instead. This affects the
2393 fifth argument of getsockopt() amongst other things. This is now
2394 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2395 can be set for individual OS. I have set it for SunOS5, OSF1, and
2396 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2397 some earlier ones do not.
2398
2399 7. Change 4.30/15 was not doing the test caselessly.
2400
2401 8. The standard form for an IPv6 address literal was being rejected by address
2402 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2403 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2404 this, as well as the form without the "IPv6" on the front (but only when
2405 address literals are enabled, of course).
2406
2407 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2408
240910. Exim crashed if a message with an empty sender address specified by -f
2410 encountered a router with an errors_to setting. This could be provoked only
2411 by a command such as
2412
2413 exim -f "" ...
2414
2415 where an empty string was supplied; "<>" did not hit this bug.
2416
241711. Installed PCRE release 4.5.
2418
241912. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2420 remained set. It is now erased.
2421
242213. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2423 times from message ids (which are base 36 rather than the normal 62).
2424
242514. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2426 were being counted as actual protocol errors, and logged if the log
2427 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2428 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2429 and DATA following a set of rejected RCPTs do not count as protocol errors.
2430 In other words, Exim assumes they were pipelined, though this may not
2431 actually be the case. Of course, in all cases the client gets an
2432 appropriate error code.
2433
243415. If a lookup fails in an ACL condition, a message about the failure may
2435 be available; it is used if testing the ACL cannot continue, because most
2436 such messages specify what the cause of the deferral is. However, some
2437 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2438 that caused an old message to be retained and used if a later statement
2439 caused a defer, replacing the real cause of the deferral.
2440
244116. If an IP address had so many PTR records that the DNS lookup buffer
2442 was not large enough to hold them, Exim could crash while trying to process
2443 the truncated data. It now detects and logs this case.
2444
244517. Further to 4.21/58, another change has been made: if (and only if) the
2446 first line of a message (the first header line) ends with CRLF, a bare LF
2447 in a subsequent header line has a space inserted after it, so as not to
2448 terminate the header.
2449
245018. Refactoring: tidied an ugly bit of code in appendfile that copied data
2451 unnecessarily, used atoi() instead of strtol(), and didn't check the
2452 termination when getting file sizes from file names by regex.
2453
245419. Completely re-implemented the support for maildirsize files, in the light
2455 of a number of problems with the previous contributed implementation
2456 (4.30/29). In particular:
2457
2458 . If the quota is zero, the maildirsize file is maintained, but no quota is
2459 imposed.
2460
2461 . If the maildir directory does not exist, it is created before any attempt
2462 to write a maildirsize file.
2463
2464 . The quota value in the file is just a cache; if the quota is changed in
2465 the transport, the new value overrides.
2466
2467 . A regular expression is available for excluding directories from the
2468 count.
2469
247020. The autoreply transport checks the characters in options that define the
2471 message's headers; it allows continued headers, but it was checking with
2472 isspace() after an embedded newline instead of explicitly looking for a
2473 space or a tab.
2474
247521. If all the "regular" hosts to which an address was routed had passed their
2476 expiry times, and had not reached their retry times, the address was
2477 bounced, even if fallback hosts were defined. Now Exim should go on to try
2478 the fallback hosts.
2479
248022. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2481 equivalent code in the SMTP transport. Some hosts send humungous responses
2482 to HELO/EHLO, more than 1024 it seems.
2483
248423. Refactoring: code in filter.c used (void *) for "any old type" but this
2485 gives compiler warnings in some environments. I've now done it "properly",
2486 using a union.
2487
248824. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2489 (because of problems with the built-in one) was declared to return uschar *
2490 instead of char *, causing compiler failure.
2491
249225. Fixed a file descriptor leak when processing alias/forward files.
2493
249426. Fixed a minor format string issue in dbfn.c.
2495
249627. Typo in exim.c: ("dmbnz" for "dbmnz").
2497
249828. If a filter file refered to $h_xxx or $message_headers, and the headers
2499 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2500 become corrupted.
2501
250229. When a sender address is verified, it is cached, to save repeating the test
2503 when there is more than one recipient in a message. However, when the
2504 verification involves a callout, it is possible for different callout
2505 options to be set for different recipients. It is too complicated to keep
2506 track of this in the cache, so now Exim always runs a verification when a
2507 callout is required, relying on the callout cache for the optimization.
2508 The overhead is duplication of the address routing, but this should not be
2509 too great.
2510
251130. Fixed a bug in callout caching. If a RCPT command caused the sender address
2512 to be verified with callout=postmaster, and the main callout worked but the
2513 postmaster check failed, the verification correctly failed. However, if a
2514 subsequent RCPT command asked for sender verification *without* the
2515 postmaster check, incorrect caching caused this verification also to fail,
2516 incorrectly.
2517
251831. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2519 it was not caching the DNS options (qualify_single, search_parents) that
2520 were used when the lookup failed. A subsequent lookup with different
2521 options therefore always gave the same answer, though there were cases
2522 where it should not have. (Example: a "domains = !$mx_any" option on a
2523 dnslookup router: the "domains" option is always processed without any
2524 widening, but the router might have qualify_single set.) Now Exim uses the
2525 cached value only when the same options are set.
2526
252732. Added John Jetmore's "exipick" utility to the distribution.
2528
252933. GnuTLS: When an attempt to start a TLS session fails for any reason other
2530 than a timeout (e.g. a certificate is required, and is not provided), an
2531 Exim server now closes the connection immediately. Previously it waited for
2532 the client to close - but if the client is SSL, it seems that they each
2533 wait for each other, leading to a delay before one of them times out.
2534
253534: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2536 maintained 0.8.x compatibility because I don't think many are using it, and
2537 it is clearly obsolete.
2538
253935. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2540 transport.
2541
254236. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2543 client certificate was expired. A simple patch fixes this, though I don't
2544 understand the full logic of why the verify callback is called multiple
2545 times.
2546
254737. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2548 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2549 which causes problems with some clients (such as the Certicom SSL Plus
2550 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2551 disables the coutermeasure allowing Eudora to connect."
2552
255338. Exim was not checking that a write() to a log file succeeded. This could
2554 lead to Bad Things if a log got too big, in particular if it hit a file
2555 size limit. Exim now panics and dies if it cannot write to a log file, just
2556 as it does if it cannot open a log file.
2557
255839. Modified OS/Makefile-Linux so that it now contains
2559
2560 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2561
2562 The two -D definitions ensure that Exim is compiled with large file
2563 support, which makes it possible to handle log files that are bigger than
2564 2^31.
2565
256640. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2567 instance) a domain was checked against a named list that involved a lookup,
2568 causing $domain_data to be set, then another domain was checked against the
2569 same list, then the first domain was re-checked, the value of $domain_data
2570 after the final check could be wrong. In particular, if the second check
2571 failed, it could be set empty. This bug probably also applied to
2572 $localpart_data.
2573
257441. The strip_trailing_dot option was not being applied to the address given
2575 with the -f command-line option.
2576
257742. The code for reading a message's header from the spool was incrementing
2578 $received_count, but never initializing it. This meant that the value was
2579 incorrect (doubled) while delivering a message in the same process in which
2580 it was received. In the most common configuration of Exim, this never
2581 happens - a fresh exec is done - but it can happen when
2582 deliver_drop_privilege is set.
2583
258443. When Exim logs an SMTP synchronization error - client data sent too soon -
2585 it now includes up to 150 characters of the unexpected data in the log
2586 line.
2587
258844. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2589 and building data strings. The size of both of these buffers was 10 000
2590 bytes - far larger than anybody would *ever* want, thought I. Needless to
2591 say, somebody hit the limit. I have increased the maximum line length to
2592 20 000 and the maximum data length of concatenated lines to 100 000. I have
2593 also fixed two bugs, because there was no checking on these buffers. Tsk,
2594 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2595 buffer is too small.
2596
259745. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2598 lsearch lookups. Now it does.
2599
260046. When parsing a route_list item in a manualroute router, a fixed-length
2601 buffer was used for the list of hosts. I made this 1024 bytes long,
2602 thinking that nobody would ever have a list of hosts that long. Wrong.
2603 Somebody had a whole pile of complicated expansion conditions, and the
2604 string was silently truncated, leading to an expansion error. It turns out
2605 that it is easier to change to an unlimited length (owing to other changes
2606 that have happened since this code was originally written) than to build
2607 structure for giving a limitation error. The length of the item that
2608 expands into the list of hosts is now unlimited.
2609
261047. The lsearch lookup could not handle data where the length of text line was
2611 more than 4095 characters. Such lines were truncated, leading to shortened
2612 data being returned. It should now handle lines of any length.
2613
261448. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2615 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2616 ACL").
2617
261849. Cosmetic tidy to scripts like exicyclog that are generated by globally
2619 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2620 no longer happens in comment lines. A list of replacements is now placed
2621 at the head of all of the source files, except those whose only change is
2622 to replace PERL_COMMAND in the very first #! line.
2623
262450. Replaced the slow insertion sort in queue.c, for sorting the list of
2625 messages on the queue, with a bottom-up merge sort, using code contributed
2626 by Michael Haardt. This should make operations like -bp somewhat faster on
2627 large queues. It won't affect queue runners, except when queue_run_in_order
2628 is set.
2629
263051. Installed eximstats 1.31 in the distribution.
2631
263252. Added support for SRV lookups to the dnslookup router.
2633
263453. If an ACL referred to $message_body or $message_body_end, the value was not
2635 reset for any messages that followed in the same SMTP session.
2636
263754. The store-handling optimization for building very long strings was not
2638 differentiating between the different store pools. I don't think this
2639 actually made any difference in practice, but I've tidied it.
2640
264155. While running the routers to verify a sender address, $sender_address
2642 was still set to the sender address. This is wrong, because when routing to
2643 send a bounce to the sender, it would be empty. Therefore, I have changed
2644 it so that, while verifying a sender address, $sender_address is set to <>.
2645 (There is no change to what happens when verifying a recipient address.)
2646
264756. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2648 target A or AAAA records (if not already returned) without resetting the
2649 qualify_single or search_parents options of the DNS resolver. These are
2650 inappropriate in this case because the targets of MX and SRV records must
2651 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2652 target that, when qualified, matched something in the local domain. These
2653 two options are now turned off when doing these lookups.
2654
265557. It seems that at least some releases of Reiserfs (which does not have the
2656 concept of a fixed number of inodes) returns zero and not -1 for the
2657 number of available inodes. This interacted badly with check_spool_inodes,
2658 which assumed that -1 was the "no such thing" setting. What I have done is
2659 to check that the total number of inodes is greater than zero before doing
2660 the test of how many are available.
2661
266258. When a "warn" ACL statement has a log_message modifier, the message is
2663 remembered, and not repeated. This is to avoid a lot of repetition when a
2664 message has many recipients that cause the same warning to be written.
2665 Howewer, Exim was preserving the list of already written lines for an
2666 entire SMTP session, which doesn't seem right. The memory is now reset if a
2667 new message is started.
2668
266959. The "rewrite" debugging flag was not showing the result of rewriting in the
2670 debugging output unless log_rewrite was also set.
2671
267260. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2673 of (int)(handle) when we know that handle contains (void *)(-1).
2674
267561. The Exim daemon panic-logs an error return when it closes the incoming
2676 connection. However "connection reset by peer" seems to be common, and
2677 isn't really an error worthy of noting specially, so that particular error
2678 is no long logged.
2679
268062. When Exim is trying to find all the local interfaces, it used to panic and
2681 die if the ioctl to get the interface flags failed. However, it seems that
2682 on at least one OS (Solaris 9) it is possible to have an interface that is
2683 included in the list of interfaces, but for which you get a failure error
2684 for this call. This happens when the interface is not "plumbed" into a
2685 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2686 failure of the "get flags" call assumes that the interface is down.
2687
268863. Added a ${eval10: operator, which assumes all numbers are decimal. This
2689 makes life easier for people who are doing arithmetic on fields extracted
2690 from dates, where you often get leading zeros that should not be
2691 interpreted as octal.
2692
269364. Added qualify_domain to the redirect router, to override the global
2694 setting.
2695
269665. If a pathologically long header line contained very many addresses (the
2697 report of this problem mentioned 10 000) and each of them was rewritten,
2698 Exim could use up a very large amount of memory. (It kept on making new
2699 copies of the header line as it rewrote, and never released the old ones.)
2700 At the expense of a bit more processing, the header rewriting function has
2701 been changed so that it no longer eats memory in this way.
2702
270366. The generation of the Received: header has been moved from the time that a
2704 message starts to be received, to the time that it finishes. The timestamp
2705 in the Received: header should now be very close to that of the <= log
2706 line. There are two side-effects of this change:
2707
2708 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2709 logged header lines no longer include the local Received: line, because
2710 it has not yet been created. The same applies to a copy of the message
2711 that is returned to a non-SMTP sender when a message is rejected.
2712
2713 (b) When a filter file is tested using -bf, no additional Received: header
2714 is added to the test message. After some thought, I decided that this
2715 is a bug fix.
2716
2717 This change does not affect the value of $received_for. It is still set
2718 after address rewriting, but before local_scan() is called.
2719
272067. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2721
272268. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2723 gave an unhelpful panic error message, and a defer error. I have managed to
2724 change this behaviour so that it now rejects any supplied certificate,
2725 which seems right, as the list of acceptable certificates is empty.
2726
272769. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2728 gave an unhelpful defer error. I have not managed to make this reject any
2729 supplied certificates, but the error message it gives is "no certificate
2730 supplied", which is not helpful.
2731
273270. exigrep's output now also includes lines that are not associated with any
2733 message, but which match the given pattern. Implemented by a patch from
2734 Martin Sluka, which also tidied up the Perl a bit.
2735
273671. Recipient callout verification, like sender verification, was using <> in
2737 the MAIL FROM command. This isn't really the right thing, since the actual
2738 sender may affect whether the remote host accepts the recipient or not. I
2739 have changed it to use the actual sender in the callout; this means that
2740 the cache record is now keyed on a recipient/sender pair, not just the
2741 recipient address. There doesn't seem to be a real danger of callout loops,
2742 since a callout by the remote host to check the sender would use <>.
2743 [SEE ABOVE: changed after hitting problems.]
2744
274572. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2746 temporary errors. However, in the case of such a code being given after
2747 the end of a data transmission (i.e. after ".") Exim was failing to write
2748 a retry record for the message. (Yes, there was some broken host that was
2749 actually sending 8xx at this point.)
2750
275173. An unknown lookup type in a host list could cause Exim to panic-die when
2752 the list was checked. (An example that provoked this was putting <; in the
2753 middle of a list instead of at the start.) If this happened during a DATA
2754 ACL check, a -D file could be left lying around. This kind of configuration
2755 error no longer causes Exim to die; instead it causes a defer errror. The
2756 incident is still logged to the main and panic logs.
2757
275874. Buglet left over from Exim 3 conversion. The message "too many messages
2759 in one connection" was written to the rejectlog but not the mainlog, except
2760 when address rewriting (yes!) was being logged.
2761
276275. Added write_rejectlog option.
2763
276476. When a system filter was run not as root (that is, when system_filter_user
2765 was set), the values of the $n variables were not being returned to the
2766 main process; thus, they were not subsequently available in the $sn
2767 variables.
2768
276977. Added +return_path_on_delivery log selector.
2770
277178. A connection timeout was being treated differently from recipients deferred
2772 when testing hosts_max_try with a message that was older than the host's
2773 retry timeout. (The host should not be counted, thus allowing all hosts to
2774 be tried at least once before bouncing.) This may have been the cause of an
2775 occasionally reported bug whereby a message would remain on the queue
2776 longer than the retry timeout, but would be bounced if a delivery was
2777 forced. I say "may" because I never totally pinned down the problem;
2778 setting up timeout/retry tests is difficult. See also the next item.
2779
278079. The ultimate address timeout was not being applied to errors that involved
2781 a combination of host plus message (for example, a timeout on a MAIL
2782 command). When an address resolved to a number of possible hosts, and they
2783 were not all tried for each delivery (e.g. because of hosts_max_try), a
2784 message could remain on the queue longer than the retry timeout.
2785
278680. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2787 Haardt.
2788
278981. Fixed an obscure SMTP outgoing bug which required at least the following
2790 conditions: (a) there was another message waiting for the same server;
2791 (b) the server returned 5xx to all RCPT commands in the first message so
2792 that the message was not completed; (c) the server dropped the connection
2793 or gave a negative response to the RSET that Exim sends to abort the
2794 transaction. The observed case was a dropped connection after DATA that had
2795 been sent in pipelining mode. That is, the server had advertised PIPELINING
2796 but was not implementing it correctly. The effect of the bug was incorrect
2797 behaviour, such as trying another host, and this could lead to a crash.
2798
2799
2800Exim version 4.30
2801-----------------
2802
2803 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2804 and daemon.c were passed as pointers to ints; they should have been
2805 pointers to socklen_t variables (which are typically unsigned ints).
2806
2807 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2808 fixed.
2809
2810 3. Fixed a really odd bug that affected only the testing scheme; patching a
2811 certain fixed string in the binary changed the value of another string that
2812 happened to be identical to the end of the original first string.
2813
2814 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2815 name", it returns that address as the IP address. On some operating
2816 systems (e.g. Solaris), it also passes back the IP address string as the
2817 "host name". However, on others (e.g. Linux), it passes back an empty
2818 string. Exim wasn't checking for this, and was changing the host name to an
2819 empty string, assuming it had been canonicized.
2820
2821 5. Although rare, it is permitted to have more than one PTR record for a given
2822 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2823 all the names associated with an address, because they do in Solaris.
2824 However, it seems that they do not in Linux for data that comes from the
2825 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2826 I found this out when I moved to a new Linux workstation and tried to run
2827 the Exim test suite.
2828
2829 To get round this problem I have changed the code so that it now does its
2830 own call to the DNS to look up PTR records when searching for a host name.
2831 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2832 addresses that are only in /etc/hosts are still found.
2833
2834 This behaviour is, however, controlled by an option called host_lookup_
2835 order, which defaults to "bydns:byaddr". If people want to use the other
2836 order, or indeed, just use one or the other means of lookup, they can
2837 specify it in this variable.
2838
2839 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2840 some operating systems, this comes back from gethostbyaddr() as an empty
2841 string, and this is what Exim used to test for. However, it seems that in
2842 other systems, "." is yielded. Exim now tests for this case too.
2843
2844 7. The values of check_spool_space and check_log_space are now held internally
2845 as a number of kilobytes instead of an absolute number of bytes. If a
2846 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2847 kilobyte. This means that much larger values can be stored.
2848
2849 8. Exim monitor: an attempt to get the action menu when not actually pointing
2850 at a message produces an empty menu entitled "No message selected". This
2851 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2852 no entries in it ("Shell widget menu has zero width and/or height"). So I
2853 have added a single, blank menu entry in this case.
2854
2855 9. Added ${quote_local_part.
2856
285710. MIME decoding is now applied to the contents of Subject: header lines when
2858 they are logged.
2859
286011. Now that a reference to $sender_host_address automatically causes a reverse
2861 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2862 host lookup before query-style lookups in lists that might use this
2863 variable. This has therefore been abolished, and the "net-" prefix is no
2864 longer necessary for query-style lookups.
2865
286612. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2867 have been a typo for LFLAGS, so it has been changed.
2868
286913. The install script calls Exim with "-C /dev/null" in order to find the
2870 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2871 to be output. Howeve, since Exim outputs its version number before the
2872 error, it didn't break the script. It just looked ugly. I fixed this by
2873 always allowing "-C /dev/null" if the caller is root.
2874
287514. Ignore overlarge ACL variable number when reading spool file - insurance
2876 against a later release with more variables having written the file.
2877
287815. The standard form for an IPv6 address literal was being rejected by EHLO.
2879 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2880 this, as well as the form without the "IPv6" on the front.
2881
288216. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2883 OS/Makefile-Darwin file.
2884
288517. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2886 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2887
288818. After change 4.21/52, "%ld" was used to format the contents of the $inode
2889 variable. However, some OS use ints for inodes. I've added cast to long int
2890 to get rid of the compiler warning.
2891
289219. I had forgotten to lock out "/../" in configuration file names when
2893 ALT_CONFIG_PREFIX was set.
2894
289520. Routers used for verification do not need to specify transports. However,
2896 if such a router generated a host list, and callout was configured, Exim
2897 crashed, because it could not find a port number from the (non-existent)
2898 transport. It now assumes port 25 in this circumstance.
2899
290021. Added the -t option to exigrep.
2901
290222. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2903 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2904 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2905 them is set, LOOKUP_LSEARCH is forced.
2906
290723. "exim -bV" now outputs a list of lookups that are included in the binary.
2908
290924. Added sender and host information to the "rejected by local_scan()" log
2910 line; previously there was no indication of these.
2911
291225. Added .include_if_exists.
2913
291426. Change 3.952/11 added an explicit directory sync on top of a file sync for
2915 Linux. It turns out that not all file systems support this. Apparently some
2916 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2917 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2918 not supported on the file descriptor, is now ignored when Exim is trying to
2919 sync a directory. This applies only to Linux.
2920
292127. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2922
292328. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2924 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2925 this.
2926
292729. Added support for maildirsize files from supplied patch (modified a bit).
2928
292930. The use of :fail: followed by an empty string could lead Exim to respond to
2930 sender verification failures with (e.g.):
2931
2932 550 Verification failed for <xxx>
2933 550 Sender verify failed
2934
2935 where the first response line was missing the '-' that indicates it is not
2936 the final line of the response.
2937
293831. The loop for finding the name of the user that called Exim had a hardwired
2939 limit of 10; it now uses the value of finduser_retries, which is used for
2940 all other user lookups.
2941
294232. Added $received_count variable, available in data and not_smtp ACLs, and at
2943 delivery time.
2944
294533. Exim was neglecting to zero errno before one call of strtol() when
2946 expanding a string and expecting an integer value. On some systems this
2947 resulted in spurious "integer overflow" errors. Also, it was casting the
2948 result into an int without checking.
2949
295034. Testing for a connection timeout using "timeout_connect" in the retry rules
2951 did not work. The code looks as if it has *never* worked, though it appears
2952 to have been documented since at least releast 1.62. I have made it work.
2953
295435. The "timeout_DNS" error in retry rules, also documented since at least
2955 1.62, also never worked. As it isn't clear exactly what this means, and
2956 clearly it isn't a major issue, I have abolished the feature by treating it
2957 as "timeout", and writing a warning to the main and panic logs.
2958
295936. The display of retry rules for -brt wasn't always showing the error code
2960 correctly.
2961
296237. Added new error conditions to retry rules: timeout_A, timeout_MX,
2963 timeout_connect_A, timeout_connect_MX.
2964
296538. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2966 to the empty sender.
2967
296839. The daemon was not analysing the content of -oX till after it had closed
2969 stderr and disconnected from the controlling terminal. This meant that any
2970 syntax errors were only noted on the panic log, and the return code from
2971 the command was 0. By re-arranging the code a little, I've made the
2972 decoding happen first, so such errors now appear on stderr, and the return
2973 code is 1. However, the actual setting up of the sockets still happens in
2974 the disconnected process, so errors there are still only recorded on the
2975 panic log.
2976
297740. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2978 connections (as happens on some IP stacks) was logged at start up time as
2979 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2980 it from "IPv6 and IPv4", which means that two separate sockets are being
2981 used.
2982
298341. The debug output for gethostbyname2() or getipnodebyname() failures now
2984 says whether AF_INET or AF_INET6 was passed as an argument.
2985
298642. Exiwhat output was messed up when time zones were included in log
2987 timestamps.
2988
298943. Exiwhat now gives more information about the daemon's listening ports,
2990 and whether -tls-on-connect was used.
2991
299244. The "port" option of the smtp transport is now expanded.
2993
299445. A "message" modifier in a "warn" statement in a non-message ACL was being
2995 silently ignored. Now an error message is written to the main and panic
2996 logs.
2997
299846. There's a new ACL modifier called "logwrite" which writes to a log file
2999 as soon as it is encountered.
3000
300147. Added $local_user_uid and $local_user_gid at routing time.
3002
300348. Exim crashed when trying to verify a sender address that was being
3004 rewritten to "<>".
3005
300649. Exim was recognizing only a space character after ".include". It now also
3007 recognizes a tab character.
3008
300950. Fixed several bugs in the Perl script that creates the exim.8 man page by
3010 extracting the relevant information from the specification. The man page no
3011 longer contains scrambled data for the -d option, and I've added a section
3012 at the front about calling Exim under different names.
3013
301451. Added "extra_headers" argument to the "mail" command in filter files.
3015
301652. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3017 crash.
3018
301953. Installed eximstats 1.29.
3020
302154. Added transport_filter_timeout as a generic transport option.
3022
302355. Exim no longer adds an empty Bcc: header to messages that have no To: or
3024 Cc: header lines. This was required by RFC 822, but it not required by RFC
3025 2822.
3026
302756. Exim used to add From:, Date:, and Message-Id: header lines to any
3028 incoming messages that did not have them. Now it does so only if the
3029 message originates locally, that is, if there is no associated remote host
3030 address. When Resent- header lines are present, this applies to the Resent-
3031 lines rather than the non-Resent- lines.
3032
303357. Drop incoming SMTP connection after too many syntax or protocol errors. The
3034 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3035
303658. Messages for configuration errors now include the name of the main
3037 configuration file - useful now that there may be more than one file in a
3038 list (.included file names were always shown).
3039
304059. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3041 for those rare installations that do not start the daemon as root or run it
3042 setuid root. I've cut out the call to initgroups() if the daemon is not
3043 root at that time.
3044
304560. The Exim user and group can now be bound into the binary as text strings
3046 that are looked up at the start of Exim's processing.
3047
304861. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3049
305062. Added $mailstore_basename variable.
3051
305263. Installed patch to sieve.c from Michael Haardt.
3053
305464. When Exim failed to open the panic log after failing to open the main log,
3055 the original message it was trying to log was written to stderr and debug
3056 output, but if they were not available (the usual case in production), it
3057 was lost. Now it is written to syslog before the two lines that record the
3058 failures to open the logs.
3059
306065. Users' Exim filters run in subprocesses under the user's uid. It is
3061 possible for a "deliver" command or an alias in a "personal" command to
3062 provoke an address rewrite. If logging of address rewriting is configured,
3063 this fails because the process is not running as root or exim. There may be
3064 a better way of dealing with this, but for the moment (because 4.30 needs
3065 to be released), I have disabled address rewrite logging when running a
3066 filter in a non-root, non-exim process.
3067
3068
3069Exim version 4.24
3070-----------------
3071
3072 1. The buildconfig auxiliary program wasn't quoting the value set for
3073 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3074 not defined. This bug was masked in 4.22 by the effect that was fixed in
3075 change 4.23/1.
3076
3077 2. Some messages that were rejected after a message id was allocated were
3078 shown as "incomplete" by exigrep. It no longer does this for messages that
3079 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3080
3081 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3082 have allow_domain_literals set, the ID did not get logged in the <= line.
3083 Domain literals are now always recognized in Message-ID: header lines.
3084
3085 4. The first argument for a ${extract expansion item is the key name or field
3086 number. Leading and trailing spaces in this item were not being ignored,
3087 causing some misleading effects.
3088
3089 5. When deliver_drop_privilege was set, single queue runner processes started
3090 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3091 same command in the process it spins off) were not dropping privilege.
3092
3093 6. When the daemon running as "exim" started a queue runner, it always
3094 re-executed Exim in the spun-off process. This is a waste of effort when
3095 deliver_drop_privilege is set. The new process now just calls the
3096 queue-runner function directly.
3097
3098
3099Exim version 4.23
3100-----------------
3101
3102 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3103 HEADERS_CHARSET.
3104
3105 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3106 ignored. Though the use of -oP was forcing the writing of a pid file, it
3107 was always written to the default place.
3108
3109 3. If the message "no IP address found for host xxxx" is generated during
3110 incoming verification, it is now followed by identification of the incoming
3111 connection (so you can more easily find what provoked it).
3112
3113 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3114
3115 5. Added some features to "harden" Exim a bit more against certain attacks:
3116
3117 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3118 be put in Local/Makefile. This is like the never_users runtime option,
3119 but it cannot be overridden. The default setting is "root".
3120
3121 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3122 prefix string with which any file named in a -C command line option
3123 must start.
3124
3125 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3126 is retained for -C and -D only if the caller of Exim is root. Without
3127 it, the exim user may also use -C and -D and retain privilege.
3128
3129 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3130 command line option is disabled.
3131
3132 6. Macro names set by the -D option must start with an upper case letter, just
3133 like macro names defined in the configuration file.
3134
3135 7. Added "dereference=" facility to LDAP.
3136
3137 8. Two instances of the typo "uknown" in the source files are fixed.
3138
3139 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3140 the Configure-Makefile script screwed up while processing it.
3141
314210. Incorporated PCRE 4.4.
3143
314411. The SMTP synchronization check was not operating right at the start of an
3145 SMTP session. For example, it could not catch a HELO sent before the client
3146 waited for the greeting. There is now a check for outstanding input at the
3147 point when the greeting is written. Because of the duplex, asynchronous
3148 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3149 way, but not yet received, when the check is performed.
3150
315112. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3152 on TCP/IP sockets, because this apparently causes some broken clients to
3153 timeout.
3154
315513. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3156 unchanged) from the Cygwin maintainer.
3157
315814. The code for -bV that shows what is in the binary showed "mbx" when maildir
3159 was supported instead of testing for mbx. Effectively a typo.
3160
316115. The spa authenticator server code was not checking that the input it
3162 received was valid base64.
3163
316416. The debug output line for the "set" modifier in ACLs was not showing the
3165 name of the variable that was being set.
3166
316717. Code tidy: the variable type "vtype_string" was never used. Removed it.
3168
316918. Previously, a reference to $sender_host_name did not cause a DNS reverse
3170 lookup on its own. Something else was needed to trigger the lookup. For
3171 example, a match in host_lookup or the need for a host name in a host list.
3172 Now, if $sender_host_name is referenced and the host name has not yet been
3173 looked up, a lookup is performed. If the lookup fails, the variable remains
3174 empty, and $host_lookup_failed is set to "1".
3175
317619. Added "eqi" as a case-independent comparison operator.
3177
317820. The saslauthd authentication condition could segfault if neither service
3179 nor realm was specified.
3180
318121. If an overflowing value such as "2048M" was set for message_size_limit, the
3182 error message that was logged was misleading, and incoming SMTP
3183 connections were dropped. The message is now more accurate, and temporary
3184 errors are given to SMTP connections.
3185
318622. In some error situations (such as 21 above) Exim rejects all SMTP commands
3187 (except RSET) with a 421 error, until QUIT is received. However, it was
3188 failing to send a response to QUIT.
3189
319023. The HELO ACL was being run before the code for helo_try_verify_hosts,
3191 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3192 ACL is now run after the helo_try_verify_hosts code.
3193
319424. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3195 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3196 case-independent, so other case variants are also recognized). Apparently
3197 some systems use these upper case variants.
3198
319925. If more than two messages were waiting for the same host, and a transport
3200 filter was specified for the transport, Exim sent two messages over the
3201 same TCP/IP connection, and then failed with "socket operation on non-
3202 socket" when it tried to send the third.
3203
320426. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3205
320627. The extern definition of crypt16() in expand.c was not being excluded when
3207 the OS had its own crypt16() function.
3208
320928. Added bounce_return_body as a new option, and bounce_return_size_limit
3210 as a preferred synonym for return_size_limit, both as an option and as an
3211 expansion variable.
3212
321329. Added LIBS=-liconv to OS/Makefile-OSF1.
3214
321530. Changed the default configuration ACL to relax the local part checking rule
3216 for addresses that are not in any local domains. For these addresses,
3217 slashes and pipe symbols are allowed within local parts, but the sequence
3218 /../ is explicitly forbidden.
3219
322031. SPA server authentication was not clearing the challenge buffer before
3221 using it.
3222
322332. log_message in a "warn" ACL statement was writing to the reject log as
3224 well as to the main log, which contradicts the documentation and doesn't
3225 seem right (because no rejection is happening). So I have stopped it.
3226
322733. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3228 However, I am unable to do any testing of this.
3229
323034. Fixed an infelicity in the appendfile transport. When checking directories
3231 for a mailbox, to see if any needed to be created, it was accidentally
3232 using path names with one or more superfluous leading slashes; tracing
3233 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3234
323535. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3236 contents are added to the log line that is written for every discarded
3237 recipient. (Previously a log_message setting was ignored.)
3238
323936. The ${quote: operator now quotes the string if it is empty.
3240
324137. The install script runs exim in order to find its version number. If for
3242 some reason other than non-existence or emptiness, which it checks, it
3243 could not run './exim', it was installing it with an empty version number,
3244 i.e. as "exim-". This error state is now caught, and the installation is
3245 aborted.
3246
324738. An argument was missing from the function that creates an error message
3248 when Exim fails to connect to the socket for saslauthd authentication.
3249 This could cause Exim to crash, or give a corrupted message.
3250
325139. Added isip, isip4, and isip6 to ${if conditions.
3252
325340. The ACL variables $acl_xx are now saved with the message, and can be
3254 accessed later in routers, transports, and filters.
3255
325641. The new lookup type nwildlsearch is like wildlsearch, except that the key
3257 strings in the file are not string-expanded.
3258
325942. If a MAIL command specified a SIZE value that was too large to fit into an
3260 int variable, the check against message_size_limit failed. Such values are
3261 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3262 day this will have to be increased, but I don't think I want to be around
3263 when emails are that large.
3264
3265
3266
3267Exim version 4.22
3268-----------------
3269
3270 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3271 iconv() is not standard in FreeBSD.
3272
3273 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3274 IPv6 enabled. The observed symptom was a segmentation fault on return from
3275 the function os_common_find_running_interfaces() in src/os.c.
3276
3277 3. In the check_special_case() function in daemon.c I had used "errno" as an
3278 argument name, which causes warnings on some systems. This was basically a
3279 typo, since it was named "eno" in the comments!
3280
3281 4. The code that waits for the clock to tick (at a resolution of some fraction
3282 of a second) so as to ensure message-id uniqueness was always waiting for
3283 at least one whole tick, when it could have waited for less. [This is
3284 almost certainly not relevant at current processor speeds, where it is
3285 unlikely to ever wait at all. But we try to future-proof.]
3286
3287 5. The function that sleeps for a time interval that includes fractions of a
3288 second contained a race. It did not block SIGALRM between setting the
3289 timer, and suspending (a couple of lines later). If the interval was short
3290 and the sigsuspend() was delayed until after it had expired, the suspension
3291 never ended. On busy systems this could lead to processes getting stuck for
3292 ever.
3293
3294 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3295 process, before it forks any delivery processes. The open lookup caching
3296 mechanism meant that the open file or database connection was passed into
3297 the delivery process. The problem was that delivery processes always tidy
3298 up cached lookup data. This could cause a problem for the next delivery
3299 process started by the queue runner, because the external queue runner
3300 process does not know about the closure. So the next delivery process
3301 still has data in the lookup cache. In the case of a file lookup, there was
3302 no problem because closing a file descriptor in a subprocess doesn't affect
3303 the parent. However, if the lookup was caching a connection to a database,
3304 the connection was closed, and the second delivery process was likely to
3305 see errors such as "PGSQL: query failed: server closed the connection
3306 unexpectedly". The problem has been fixed by closing all cached lookups
3307 in a queue runner before running a delivery process.
3308
3309 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3310 seem to have the "const" qualifier which it has on other OS. I've
3311 parameterised it.
3312
3313 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3314 *of the same type* (client or server) with the same public name that an
3315 error should be diagnosed.
3316
3317 9. When Exim looked up a host name for an IP address, but failed to find the
3318 original IP address when looking up the host name (a safety check), it
3319 output the message "<ip address> does not match any IP for NULL", which was
3320 confusing, to say the least. The bug was that the host name should have
3321 appeared instead of "NULL".
3322
332310. Since release 3.03, if Exim is called by a uid other than root or the Exim
3324 user that is built into the binary, and the -C or -D options is used, root
3325 privilege is dropped before the configuration file is read. In addition,
3326 logging is switched to stderr instead of the normal log files. If the
3327 configuration then re-defines the Exim user, the unprivileged environment
3328 is probably not what is expected, so Exim logs a panic warning message (but
3329 proceeds).
3330
3331 However, if deliver_drop_privilege is set, the unprivileged state may well
3332 be exactly what is intended, so the warning has been cut out in that case,
3333 and Exim is allowed to try to write to its normal log files.
3334
3335
3336Exim version 4.21
3337-----------------
3338
3339 1. smtp_return_error_details was not giving details for temporary sender
3340 or receiver verification errors.
3341
3342 2. Diagnose a configuration error if two authenticators have the same public
3343 name.
3344
3345 3. Exim used not to create the message log file for a message until the first
3346 delivery attempt. This could be confusing when incoming messages were held
3347 for policy or load reasons. The message log file is now created at the time
3348 the message is received, and an initial "Received" line is written to it.
3349
3350 4. The automatically generated man page for command line options had a minor
3351 bug that caused no ill effects; however, a more serious problem was that
3352 the procedure for building the man page automatically didn't always
3353 operate. Consequently, release 4.20 contains an out-of-date version. This
3354 shouldn't happen again.
3355
3356 5. When building Exim with embedded Perl support, the script that builds the
3357 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3358 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3359
3360 6. The freeze_tell option was not being used for messages that were frozen on
3361 arrival, either by an ACL or by local_scan().
3362
3363 7. Added the smtp_incomplete_transaction log selector.
3364
3365 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3366 was accepting AUTH without a new EHLO.
3367
3368 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3369 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3370 RFC.
3371
337210. Logging of TCP/IP connections (when configured) now happens in the main
3373 daemon process instead of the child process, so that the TCP/IP connection
3374 count is more accurate (but it can never be perfect).
3375
337611. The use of "drop" in a nested ACL was not being handled correctly in the
3377 outer ACL. Now, if condition failure induced by the nested "drop" causes
3378 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3379 or "require"), the connection is dropped.
3380
338112. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3382 that yield "discard" can now be used with an "accept" or a "discard" verb,
3383 but an error is generated for any others (because I can't see a useful way
3384 to define what should happen).
3385
338613. When an ACL is read dynamically from a file (or anywhere else), the lines
3387 are now processed in the same way as lines in the Exim configuration file.
3388 In particular, continuation lines are supported.
3389
339014. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3391
339215. Added -ti meaning -t -i.
3393
339416. Check for letters, digits, hyphens, and dots in the names of dnslist
3395 domains, and warn by logging if others are found.
3396
339717. At least on BSD, alignment is not guarenteed for the array of ifreq's
3398 returned from GIFCONF when Exim is trying to find the list of interfaces on
3399 a host. The code in os.c has been modified to copy each ifreq to an aligned
3400 structure in all cases.
3401
3402 Also, in some cases, the returned ifreq's were being copied to a 'struct
3403 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3404 means the last couple of bytes of an IPv6 address could be chopped if the
3405 ifreq contained only a normal sockaddr (14 bytes storage).
3406
340718. Named domain lists were not supported in the hosts_treat_as_local option.
3408 An entry such as +xxxx was not recognized, and was treated as a literal
3409 domain name.
3410
341119. Ensure that header lines added by a DATA ACL are included in the reject log
3412 if the ACL subsequently rejects the message.
3413
341420. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3415 MD5 (which is deprecated).
3416
341721. When testing a filter file using -bf, Exim was writing a message when it
3418 took the sender from a "From " line in the message, but it was not doing so
3419 when it took $return_path from a Return-Path: header line. It now does.
3420
342122. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3422 with a valid header line field name (a series of printing characters
3423 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3424
342523. Changed "disc" in the source to "disk" to conform to the documentation and
3426 the book and for uniformity.
3427
342824. Ignore Sendmail's -Ooption=value command line item.
3429
343025. When execve() failed while trying to run a command in a pipe transport,
3431 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3432 could be confused with a return value of 69 from the command itself. This
3433 has been changed to 127, the value the shell returns if it is asked to run
3434 a non-existent command. The wording for the related log line suggests a
3435 non-existent command as the problem.
3436
343726. If received_header_text expands to an empty string, do not add a Received:
3438 header line to the message. (Well, it adds a token one on the spool, but
3439 marks it "old" so that it doesn't get used or transmitted.)
3440
344127. Installed eximstats 1.28 (addition of -nt option).
3442
344328. There was no check for failure on the call to getsockname() in the daemon
3444 code. This can fail if there is a shortage of resources on the system, with
3445 ENOMEM, for example. A temporary error is now given on failure.
3446
344729. Contrary to the C standard, it seems that in some environments, the
3448 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3449 program. Exim now does this explicitly; it affects the formatting of
3450 timestamps using strftime().
3451
345230. If exiqsumm was given junk data, it threw up some uninitialized variable
3453 complaints. I've now initialized all the variables, to avoid this.
3454
345532. Header lines added by a system filter were not being "seen" during
3456 transport-time rewrites.
3457
345833. The info_callback() function passed to OpenSSL is set up with type void
3459 (*)(SSL *, int, int), as described somewhere. However, when calling the
3460 function (actually a macro) that sets it up, the type void(*)() is
3461 expected. I've put in a cast to prevent warnings from picky compilers.
3462
346334. If a DNS black list lookup found a CNAME record, but there were no A
3464 records associated with the domain it pointed at, Exim crashed.
3465
346635. If a DNS black list lookup returned more than one A record, Exim ignored
3467 all but the first. It now scans all returned addresses if a particular IP
3468 value is being sought. In this situation, the contents of the
3469 $dnslist_value variable are a list of all the addresses, separated by a
3470 comma and a space.
3471
347236. Tightened up the rules for host name lookups using reverse DNS. Exim used
3473 to accept a host name and all its aliases if the forward lookup for any of
3474 them yielded the IP address of the incoming connection. Now it accepts only
3475 those names whose forward lookup yields the correct IP address. Any other
3476 names are discarded. This closes a loophole whereby a rogue DNS
3477 administrator could create reverse DNS records to break through a
3478 wildcarded host restriction in an ACL.
3479
348037. If a user filter or a system filter that ran in a subprocess used any of
3481 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3482 the wrong values were passed to the pipe command ($thisaddress had the
3483 value of $0, $0 had the value of $1, etc). This bug was introduced by
3484 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3485
348638. Improved the line breaking for long SMTP error messages from ACLs.
3487 Previously, if there was no break point between 40 and 75 characters, Exim
3488 left the rest of the message alone. Two changes have been made: (a) I've
3489 reduced the minimum length to 35 characters; (b) if it can't find a break
3490 point between 35 and 75 characters, it looks ahead and uses the first one
3491 that it finds. This may give the occasional overlong line, but at least the
3492 remaining text gets split now.
3493
349439. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3495 file descriptors might be low, and that setting 1000 would always raise it.
3496 It turns out that in some environments, the limit is already over 1000 and
3497 that lowering it causes trouble. So now Exim takes care not to decrease it.
3498
349940. When delivering a message, the value of $return_path is set to $sender_
3500 address at the start of routing (routers may change the value). By an
3501 oversight, this default was not being set up when an address was tested by
3502 -bt or -bv, which affected the outcome if any router or filter referred to
3503 $return_path.
3504
350541. The idea of the "warn" ACL verb is that it adds a header or writes to the
3506 log only when "message" or "log_message" are set. However, if one of the
3507 conditions was an address verification, or a call to a nested ACL, the
3508 messages generated by the underlying test were being passed through. This
3509 no longer happens. The underlying message is available in $acl_verify_
3510 message for both "message" and "log_message" expansions, so it can be
3511 passed through if needed.
3512
351342. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3514 new expansion $bh_ to give the encoded byte string without charset
3515 translation. Translation happens only if iconv() is available; HAVE_ICONV
3516 indicates this at build time. HEADERS_CHARSET gives the charset to
3517 translate to; headers_charset can change it in the configuration, and
3518 "headers charset" can change it in an individual filter file.
3519
352043. Now that we have a default RFC 2047 charset (see above), the code in Exim
3521 that creates RFC 2047 encoded "words" labels them as that charset instead
3522 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3523 expansion operator; (ii) when Exim creates a From: line for a local
3524 message; (iii) when a header line is rewritten to include a "phrase" part.
3525
352644. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3527 buggy, causing it to skip the first lines of messages whose message ID
3528 ended in 'D'. This would not have bitten before Exim release 4.14, because
3529 message IDs were unlikely to end in 'D' before then. The effect was to have
3530 incorrect size information for certain domains.
3531
353245. #include "config.h" was missing at the start of the crypt16.c module. This
3533 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3534 noticed.
3535
353646. If there was a timeout during a "random" callout check, Exim treated it as
3537 a failure of the random address, and carried on sending RSET and the real
3538 address. If the delay was just some slowness somewhere, the response to the
3539 original RCPT would be taken as a response to RSET and so on, causing
3540 mayhem of various kinds.
3541
354247. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3543 when I implemented it. It didn't allow for the fact that some option values
3544 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3545 the right test for positive values.
3546
354748. Domain names in DNS records are case-independent. Exim always looks them up
3548 in lower case. Some resolvers return domain names in exactly the case they
3549 appear in the zone file, that is, they may contain uppercase letters. Not
3550 all resolvers do this - some return always lower case. Exim was treating a
3551 change of case by a resolver as a change of domain, similar to a widening
3552 of a domain abbreviation. This triggered its re-routing code and so it was
3553 trying to route what was effectively the same domain again. This normally
3554 caused routing to fail (because the router wouldn't handle the domain
3555 twice). Now Exim checks for this case specially, and just changes the
3556 casing of the domain that it ultimately uses when it transmits the message
3557 envelope.
3558
355949. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3560 module.
3561
356250. If a filter generated a file delivery with a non-absolute name (possible if
3563 no home directory exists for the router), the forbid_file option was not
3564 forbidding it.
3565
356651. Added '&' feature to dnslists, to provide bit mask matching in addition to
3567 the existing equality matching.
3568
356952. Exim was using ints instead of ino_t variables in some places where it was
3570 dealing with inode numbers.
3571
357253. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3573 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3574 called TMPDIR, and if it finds it is different, it changes its value.
3575
357654. The smtp_printf() function is now made available to local_scan() so
3577 additional output lines can be written before returning. There is also an
3578 smtp_fflush() function to enable the detection of a dropped connection.
3579 The variables smtp_input and smtp_batched_input are exported to
3580 local_scan().
3581
358255. Changed the default runtime configuration: the message "Unknown user"
3583 has been removed from the ACL, and instead placed on the localuser router,
3584 using the cannot_route_message feature. This means that any verification
3585 failures that generate their own messages won't get overridden. Similarly,
3586 the "Unrouteable address" message that was in the ACL for unverifiable
3587 relay addresses has also been removed.
3588
358956. Added hosts_avoid_esmtp to the smtp transport.
3590
359157. The exicyclog script was not checking for the esoteric option
3592 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3593 will work only if exicyclog is run under the appropriate euid.
3594
359558. Following a discussion on the list, the rules by which Exim recognises line
3596 endings on incoming messages have been changed. The -dropcr and drop_cr
3597 options are now no-ops, retained only for backwards compatibility. The
3598 following line terminators are recognized: LF CRLF CR. However, special
3599 processing applies to CR:
3600
3601 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3602 nor a local message in the state where . is a terminator.
3603
3604 (ii) If a bare CR is encountered in a header line, an extra space is added
3605 after the line terminator so as not to end the header. The reasoning
3606 behind this is that bare CRs in header lines are most likely either
3607 to be mistakes, or people trying to play silly games.
3608
360959. The size of a message, as listed by "-bp" or in the Exim monitor window,
3610 was being incorrectly given as 18 bytes larger than it should have been.
3611 This is a VOB (very old bug).
3612
361360. This may never have affected anything current, but just in case it has:
3614 When the local host is found other than at the start of a list of hosts,
3615 the local host, those with the same MX, and any that follow, are discarded.
3616 When the list in question was part of a longer list of hosts, the following
3617 hosts (not currently being processed) were also being discarded. This no
3618 longer happens. I'm not sure if this situation could ever has previously
3619 arisen.
3620
362161. Added the "/MX" feature to lists of hosts in the manualroute and query
3622 program routers.
3623
362462. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3625 header. This is something that is recommended in a new Internet Draft, and
3626 is something that is documented as being done by Sendmail. There are two
3627 possible values. For messages generated by the autoreply transport, Exim
3628 adds:
3629
3630 Auto-Submitted: auto-replied
3631
3632 whereas for all other generated messages (e.g. bounces) it adds
3633
3634 Auto-Submitted: auto-generated
3635
363663. The "personal" condition in filters now includes a test for the
3637 Auto-Submitted: header. If it contains the string "auto-" the message it
3638 not considered personal.
3639
364064. Added rcpt_include_affixes as a generic transport option.
3641
364265. Added queue_only_override (default true).
3643
364466. Added the syslog_duplication option.
3645
364667. If what should have been the first header line of a message consisted of
3647 a space followed by a colon, Exim was mis-interpreting it as a header line.
3648 It isn't of course - it is syntactically invalid and should therefore be
3649 treated as the start of the message body. The misbehaviour could have
3650 caused a number of strange effects, including loss of data in subsequent
3651 header lines, and spool format errors.
3652
365368. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3654 client host had authenticated. This control can now be exercised by an ACL
3655 for more flexibility.
3656
365769. By default, callouts do not happen when testing with -bh. There is now a
3658 variant, -bhc, which does actually run the callout code, including
3659 consulting and updating the callout cache.
3660
366170. Added support for saslauthd authentication, courtesy of Alexander
3662 Sabourenkov.
3663
366471. If statvfs() failed on the spool or log directories while checking their
3665 size for availability, Exim confusingly gave the error "space shortage".
3666 Furthermore, in debugging mode it crashed with a floating point exception.
3667 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3668 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3669 serious problem, Exim now writes to the main and panic logs when this
3670 happens, with details of the failure. It then refuses to accept the
3671 incoming message, giving the message "spool directory problem" or "log
3672 directory problem" with a 421 code for SMTP messages.
3673
367472. When Exim is about to re-exec itself, it ensures that the file descriptors
3675 0, 1, and 2 exist, because some OS complain for execs without them (see
3676 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3677 descriptors. However, the code omitted to check that the open succeeded,
3678 causing mysterious errors if for some reason the permissions on /dev/null
3679 got screwed. Now Exim writes a message to the main and panic logs, and
3680 bombs out if it can't open /dev/null.
3681
368273. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3683 interact so that it is all more flexible. It is supposed to remain
3684 backwards compatible. Also added extra_local_interfaces.
3685
368674. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3687 to bomb out with an assertion failure - to the client this appears as a
3688 connection drop. This problem occurs in the part of the code that was taken
3689 from the Samba project. Fortunately, the assertion is in a very simple