Omit spaces between name and colon in header syntax verify error
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
1ab95fa6 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.302 2006/02/14 15:56:43 ph10 Exp $
495ae4b0
PH
2
3Change log file for Exim from version 4.21
4-------------------------------------------
5
214e2000
PH
6Exim version 4.61
7-----------------
8
9PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
15 addresses as local.
16
f9daeae0
PH
17PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
19
7e66e54d
PH
20PH/03 Added disable_ipv6.
21
c8ea1597
PH
22PH/04 Changed $reply_address to use the raw form of the headers instead of the
23 decoded form, because it is most often used to construct To: headers
24 lines in autoreplies, and the decoded form may well be syntactically
25 invalid. However, $reply_address has leading white space removed, and all
26 newlines turned into spaces so that the autoreply transport does not
27 grumble.
28
911f6fde
PH
29PH/05 If group was specified without a user on a router, and no group or user
30 was specified on a transport, the group from the router was ignored.
31
47ca6d6c
PH
32PH/06 Increased the number of ACL variables to 20 of each type, and arranged
33 for visible compile-time settings that can be used to change these
34 numbers, for those that want even more. Backwards compatibility with old
35 spool files has been maintained. However, going back to a previous Exim
36 release will lost any variables that are in spool files.
37
ed0e9820
PH
38PH/07 Two small changes when running in the test harness: increase delay when
39 passing a TCP/IP connection to a new process, in case the original
40 process has to generate a bounce, and remove special handling of
41 127.0.0.2 (sic), which is no longer necessary.
42
eff37e47
PH
43PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
44 be the same on different OS.
45
1921d2ea
PH
46PH/09 Moved a debug statement in filter processing to avoid a race problem when
47 testing.
48
b3f69ca8
JJ
49JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
50 whether --show-vars was specified or not
51
52JJ/02 exipick: Added support for new ACL variable spool format introduced
53 in 4.61-PH/06
54
424a1c63
PH
55PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
56 syntactically invalid From: or Reply-to: line, and a filter used this to
57 generate an autoreply, and therefore failed to obtain an address for the
58 autoreply, Exim could try to deliver to a non-existent relative file
59 name, causing unrelated and misleading errors. What now happens is that
60 it logs this as a hard delivery error, but does not attempt to create a
61 bounce message.
62
7a100415
PH
63PH/11 The exinext utility has a -C option for testing purposes, but although
64 the given file was scanned by exinext itself; it wasn't being passed on
65 when Exim was called.
66
19b9dc85
PH
67PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
68 an end-of-file indication when reading a command response.
69
309bd837
PH
70PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
71 compiled. In many other places in Exim, IPv6 addresses are always
72 recognized, so I have changed this. It also means that IPv4 domain
73 literals of the form [IPV4:n.n.n.n] are now always recognized.
74
59e82a2a
PH
75PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
76 used if the router is not running as root, for example, when verifying at
77 ACL time, or when using -bh. The debugging output from this situation was
78 non-existent - all you got was a failure to exec. I have made two
79 changes:
80
81 (a) Failures to set uid/gid, the current directory, or a process leader
82 in a subprocess such as that created by queryprogram now generate
83 suitable debugging ouput when -d is set.
84
85 (b) The queryprogram router detects when it is not running as root,
86 outputs suitable debugging information if -d is set, and then runs
87 the subprocess without attempting to change uid/gid.
88
9edc04ce
PH
89PH/15 Minor change to Makefile for building test_host (undocumented testing
90 feature).
91
1349e1e5
PH
92PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
93 additional section of a DNS packet that returns MX or SRV records.
94 Instead, it always explicitly searches for A/AAAA records. This avoids
95 major problems that occur when a DNS server includes only records of one
96 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
97 fixed another bug: if SRV records were looked up and the corresponding
98 address records were *not* found in the additional section, the port
99 values from the SRV records were lost.
100
ea49d0e1
PH
101PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
102 using the correct key (the original address) when searching the retry
103 rules in order to find which one to use for generating the retry hint.
104
064a94c9
PH
105PH/18 If quota_warn_message contains a From: header, Exim now refrains from
106 adding the default one. Similarly, if it contains a Reply-To: header, the
107 errors_reply_to option, if set, is not used.
108
727071f8
PH
109PH/19 When calculating a retry time, Exim used to measure the "time since
110 failure" by looking at the "first failed" field in the retry record. Now
111 it does not use this if it is later than than the arrival time of the
112 message. Instead it uses the arrival time. This makes for better
113 behaviour in cases where some deliveries succeed, thus re-setting the
114 "first failed" field. An example is a quota failure for a huge message
115 when small messages continue to be delivered. Without this change, the
116 "time since failure" will always be short, possible causing more frequent
117 delivery attempts for the huge message than are intended.
118
f78eb7c6
PH
119PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
120 $1, $2, $3) because the numerical variables can be reset during some
121 expansion items (e.g. "match"), thereby losing the authentication data.
122
21c28500
PH
123PH/21 Make -bV show the size of off_t variables so that the test suite can
124 decide whether to run tests for quotas > 2G.
125
126PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
127 mailbox_size, and mailbox_filecount in the appendfile transport. If a
128 filecount value is greater than 2G or if a quota value is greater than 2G
129 on a system where the size of off_t is not greater than 4, a panic error
130 is given.
131
1688f43b
PH
132PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
133 never match. The debug and -bh output now contains an explicit error
134 message indicating a malformed IPv4 address or mask.
135
136PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
137 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
138 PH/23 above applies.
139
9675b384
PH
140PH/25 Do not write to syslog when running in the test harness. The only
141 occasion when this arises is a failure to open the main or panic logs
142 (for which there is an explicit test).
143
6a3f1455
PH
144PH/26 Added the /no_tell option to "control=freeze".
145
dac79d3e
PH
146PH/27 If a host name lookup failed very early in a connection, for example, if
147 the IP address matched host_lookup and the reverse lookup yielded a name
148 that did not have a forward lookup, an error message of the form "no IP
149 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
150 could be logged. Now it outputs the IP address instead of "NULL".
1349e1e5 151
5977a0b3
PH
152PH/28 An enabling patch from MH: add new function child_open_exim2() which
153 allows the sender and the authenticated sender to be set when
154 submitting a message from within Exim. Since child_open_exim() is
155 documented for local_scan(), the new function should be too.
156
c91535f3
PH
157PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
158 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
159 results in an empty string is now treated as unset.
160
0d46a8c8
PH
161PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
162
278c6e6c
PH
163PH/31 Added sender_verify_fail logging option.
164
2cbb4081
PH
165PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
166 needed by RFC 822 but not by RFC 2822 was commented out. I have now
167 tidied the source and removed it altogether.
168
3eef829e
PH
169PH/33 When a queue run was abandoned because the load average was too high, a
170 log line was always written; now it is written only if the queue_run log
171 selector is set. In addition, the log line for abandonment now contains
172 information about the queue run such as the pid. This is always present
173 in "start" and "stop" lines but was omitted from the "abandon" line.
174
1ab95fa6
PH
175PH/34 Omit spaces between a header name and the colon in the error message that
176 is given when verify = headers_syntax fails (if there are lots of them,
177 the message gets confusing).
178
214e2000 179
5de37277
PH
180Exim version 4.60
181-----------------
182
cc38ddbf
PH
183PH/01 Two changes to the default runtime configuration:
184
185 (1) Move the checks for relay_from_hosts and authenticated clients from
186 after to before the (commented out) DNS black list checks.
187
188 (2) Add control=submission to the relay_from_hosts and authenticated
189 clients checks, on the grounds that messages accepted by these
190 statements are most likely to be submissions.
5de37277 191
72fdd6ae
PH
192PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
193
194 (1) Generate an error if the third argument for the ${prvs expansion is
195 not a single digit.
196
197 (2) Treat a missing third argument of ${prvscheck as if it were an empty
198 string.
199
200 (3) Reset the variables that are obtained from the first argument of
201 ${prvscheck and used in the second argument before leaving the code,
202 because their memory is reclaimed, so using them afterwards may do
203 silly things.
204
205 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
206 one (it's much easier than Tom thought :-).
207
208 (5) Because of (4), we can now allow for the use of $prvscheck_result
209 inside the third argument.
cb9328de 210
cb741023
PH
211PH/03 For some reason, the default setting of PATH when running a command from
212 a pipe transport was just "/usr/bin". I have changed it to
213 "/bin:/usr/bin".
214
f174f16e
PH
215PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
216 anything to be listed in the output from -bV.
b2f5a032 217
c25242d7
PH
218PH/05 When a filter generated an autoreply, the entire To: header line was
219 quoted in the delivery log line, like this:
220
221 => >A.N.Other <ano@some.domain> <original@ddress> ...
222
223 This has been changed so that it extracts the operative address. There
224 may be more than one such address. If so, they are comma-separated, like
225 this:
226
227 => >ano@some.domain,ona@other.domain <original@ddress> ...
228
82c19f95
PH
229PH/06 When a client host used a correct literal IP address in a HELO or EHLO
230 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
231 not being looked up in the rDNS to get a host name, Exim was showing the
232 IP address twice in Received: lines, even though the IP addresses were
233 identical. For example:
234
235 Received: from [1.2.3.4] (helo=[1.2.3.4])
236
237 However, if the real host name was known, it was omitting the HELO data
238 if it matched the actual IP address. This has been tidied up so that it
239 doesn't show the same IP address twice.
240
d7ffbc12
PH
241PH/07 When both +timestamp and +memory debugging was on, the value given by
242 $tod_xxx expansions could be wrong, because the tod_stamp() function was
243 called by the debug printing, thereby overwriting the timestamp buffer.
244 Debugging no longer uses the tod_stamp() function when +timestamp is set.
245
9f526266
PH
246PH/08 When the original message was included in an autoreply transport, it
247 always said "this is a copy of the message, including all the headers",
248 even if body_only or headers_only was set. It now gives an appropriate
249 message.
250
87fcc8b9
PH
251PH/09 Applied a patch from the Sieve maintainer which:
252
253 o fixes some comments
254 o adds the (disabled) notify extension core
255 o adds some debug output for the result of if/elsif tests
256 o points to the current vacation draft in the documentation
257 and documents the missing references header update
258
259 and most important:
260
261 o fixes a bug in processing the envelope test (when testing
262 multiple envelope elements, the last element determinted the
263 result)
264
456682f5
PH
265PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
266 Electronic Mail") by including:
267
268 Auto-submitted: auto-generated
269
270 in the messages that it generates (bounce messages and others, such as
271 warnings). In the case of bounce messages for non-SMTP mesages, there was
272 also a typo: it was using "Auto_submitted" (underscore instead of
273 hyphen). Since every message generated by Exim is necessarily in response
274 to another message, thes have all been changed to:
275
276 Auto-Submitted: auto-replied
277
278 in accordance with these statements in the RFC:
279
280 The auto-replied keyword:
281
282 - SHOULD be used on messages sent in direct response to another
283 message by an automatic process,
284
285 - MUST NOT be used on manually-generated messages,
286
287 - MAY be used on Delivery Status Notifications (DSNs) and Message
288 Disposition Notifications (MDNs),
289
290 - MUST NOT be used on messages generated by automatic or periodic
291 processes, except for messages which are automatic responses to
292 other messages.
293
3e46c1aa
PH
294PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
295 to the default Received: header definition.
456682f5 296
49826d12
PH
297PH/12 Added log selector acl_warn_skipped (default on).
298
eba0c039
PH
299PH/13 After a successful wildlsearch lookup, discard the values of numeric
300 variables because (a) they are in the wrong storage pool and (b) even if
301 they were copied, it wouldn't work properly because of the caching.
302
a0d6ba8a
PH
303PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
304 checking when decoding. Apparently there are clients that generate
305 overlong encoded strings. Why am I not surprised?
306
f0917727
PH
307PH/15 If the first argument of "${if match_address" was not empty, but did not
308 contain an "@" character, Exim crashed. Now it writes a panic log message
309 and treats the condition as false.
310
096fee00
PH
311PH/16 In autoreply, treat an empty string for "once" the same as unset.
312
024bd3c2
PH
313PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
314 extension "envelope-auth". The code is finished and in agreement with
315 other implementations, but there is no documentation so far and in fact,
316 nobody wrote the draft yet. This extension is currently #undef'ed, thus
317 not changing the active code.
318
319 Print executed "if" and "elsif" statements when debugging is used. This
320 helps a great deal to understand what a filter does.
321
322 Document more things not specified clearly in RFC3028. I had all this
323 sorted out, when out of a sudden new issues came to my mind. Oops."
324
df199fec
PH
325PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
326 (Bugzilla #53).
327
d27f1df3
PH
328PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
329 canonical form (as documented). However, after a host name lookup from
330 the IP address, check_host() was doing a simple string comparison with
331 addresses acquired from the DNS when checking that the found name did
332 have the original IP as one of its addresses. Since any found IPv6
333 addresses are likely to be in abbreviated form, the comparison could
334 fail. Luckily, there already exists a function for doing the comparison
335 by converting both addresses to binary, so now that is used instead of
336 the text comparison.
337
96776534
PH
338PH/20 There was another similar case to PH/19, when a complete host name was
339 given in a host list; looking up its IP address could give an abbreviated
340 form, whereas the current host's name might or might not be abbreviated.
341 The same fix has been applied.
342
5de37277 343
9a799bc0
PH
344Exim version 4.54
345-----------------
346
347PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
348 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
349 It now does.
350
99a4b039
PH
351PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
352 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
353
354PH/03 Typo: missing ".o" in src/pcre/Makefile.
355
4b233853
PH
356PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
357 header line, restrict the check to what is listed in RFCs 2369 and 2929.
358 Also, for "Auto-Submitted", treat anything other than "no" as
359 non-personal, in accordance with RFC 3834. (Previously it treated
360 anything starting "auto-" as non-personal.)
361
8857ccfd
PH
362TF/01 The control=submission/name=... option had a problem with syntax
363 errors if the name included a slash character. The /name= option
364 now slurps the rest of the string, so it can include any characters
365 but it must come last in the list of options (after /sender_retain
366 or /domain=).
367
433a2980
PH
368PH/05 Some modifications to the interface to the fake nameserver for the new
369 testing suite.
370
3e46c1aa 371
9a799bc0 372
e3a311ba
TK
373Exim version 4.53
374-----------------
375
376TK/01 Added the "success_on_redirect" address verification option. See
377 NewStuff for rationale and an example.
378
13b685f9
PH
379PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
380
395ff96d
PH
381PH/02 Patch to exigrep to allow it to work on syslog lines.
382
5b68f6e4
PH
383PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
384 fread() to skip over the body file's header line, because in Cygwin the
385 header line is locked and is inaccessible.
386
1ab52c69
PH
387PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
388 co-exist for some time) to make it clear that it is the Exim ID that is
389 referenced, not the Message-ID: header line.
390
b07e6aa3
PH
391PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
392 string_format() function, because snprintf() does not exist on all
393 operating systems.
394
254e032f
PH
395PH/06 The use of forbid_filter_existstest now also locks out the use of the
396 ${stat: expansion item.
397
3af76a81
PH
398PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
399 protocol synchronization error", to keep the pedants happy.
400
2548ba04
PH
401PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
402 well as for IRIX systems, when gcc is being used. See the host.c source
403 file for comments.
404
b6c6011d
PH
405PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
406
cf39cf57
PH
407PH/10 Named domain lists were not working if used in a queue_smtp_domains
408 setting.
409
f1513293
PH
410PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
411 transport and to the smtp transport in LMTP mode.
412
727549a4
PH
413TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
414
af46795e
PH
415PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
416 run a filter in a subprocess. This could lead to confusion in subsequent
417 lookups in the parent process. There should also be a search_tidyup() at
418 the end of the subprocess.
419
d7b47fd0
PH
420PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
421 only if the host matched helo_try_verify_hosts, which caused the
422 verification to occur when the EHLO/HELO command was issued. The ACL just
423 tested the remembered result. Now, if a previous verification attempt has
424 not happened, "verify = helo" does it there and then.
425
ee744174
JJ
426JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
427
b582ab87
PH
428TK/03 Fix log output including CR from clamd.
429
41a13e0a
PH
430PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
431 exist provoked a memory error which could cause a segfault.
432
f625cc5a
PH
433PH/15 Installed PCRE 6.2
434
435PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
436
21f7af35
PH
437PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
438 of the problem. Specifically, suggested +O2 rather than +O1 for the
439 HP-UX compiler.
440
31480e42
PH
441PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
442
2d280592
PH
443PH/20 If a delivery was routed to a non-standard port by means of an SRV
444 record, the port was not correctly logged when the outgoing_port log
445 selector was set (it logged the transort's default port).
446
7cd1141b
PH
447PH/21 Added support for host-specific ports to manualroute, queryprogram,
448 fallback_hosts, and "hosts" in the smtp transport.
449
450PH/22 If the log selector "outgoing_port" is set, the port is now also given on
451 host errors such as "Connection refused".
452
750af86e
PH
453PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
454 authentication with radiusclient 0.4.9:
455
456 - Error returned from rc_read_config was caught wrongly
457 - Username/password not passed on to radius server due to wrong length.
458
459 The presumption is that some radiusclient API changes for 4.51/PH/17
460 were not taken care of correctly. The code is still untested by me (my
461 Linux distribution still has 0.3.2 of radiusclient), but it was
462 contributed by a Radius user.
463
464PH/24 When doing a callout, the value of $domain wasn't set correctly when
465 expanding the "port" option of the smtp transport.
466
4304270b
TK
467TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
468 while reading a MIME header. Thanks to Tom Hughes for a patch.
469
750af86e
PH
470PH/24 Include config.h inside local_scan.h so that configuration settings are
471 available.
472
64ffc24f
PH
473PH/25 Make $smtp_command_argument available after all SMTP commands. This means
474 that in an ACL for RCPT (for example), you can examine exactly what was
475 received.
476
5dd9625b
PH
477PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
478 commands, but it was not correctly comparing the address with the actual
479 client host address. Thus, it would show the EHLO address in Received:
480 header lines when this was not necessary.
481
5591031b
PH
482PH/27 Added the % operator to ${eval:}.
483
ba18e66a
PH
484PH/28 Exim tries to create and chdir to its spool directory when it starts;
485 it should be ignoring failures (because with -C, for example, it has lost
486 privilege). It wasn't ignoring creation failures other than "already
487 exists".
488
9cec981f
PH
489PH/29 Added "crypteq" to the list of supported features that Exim outputs when
490 -bV or -d is used.
491
aa2b5c79
PH
492PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
493 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
494 too many continuations, the temporary file was not being removed, and the
495 return code was incorrect.
aa2b5c79 496
48a53b7f
PH
497PH/31 Missing "BOOL" in function definition in filtertest.c.
498
1c59d63b
PH
499PH/32 Applied Sieve patches from the maintainer.
500
671012da
TK
501TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
502
1509d3a8
PH
503PH/33 Added "verify = not_blind".
504
505PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
506 Local/Makefile (with some defaults set). These are used in built scripts
507 such as exicyclog, but they have never been used in the exim_install
508 script (though there are many overriding facilities there). I have
509 arranged that the exim_install script now takes note of these two
510 settings.
511
512PH/35 Installed configuration files for Dragonfly.
513
2fe1a124
PH
514PH/36 When a locally submitted message by a trusted user did not contain a
515 From: header, and the sender address was obtained from -f or from an SMTP
516 MAIL command, and the trusted user did not use -F to supply a sender
517 name, $originator_name was incorrectly used when constructing a From:
518 header. Furthermore, $originator_name was used for submission mode
519 messages from external hosts without From: headers in a similar way,
520 which is clearly wrong.
521
8800895a
PH
522PH/37 Added control=suppress_local_fixups.
523
ccfdb010
PH
524PH/38 When log_selector = +received_sender was set, and the addition of the
525 sender made the log line's construction buffer exactly full, or one byte
526 less than full, an overflow happened when the terminating "\n" was
527 subsequently added.
528
1130bfb0
PH
529PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
530 when the result of a list match is failure because a DNS lookup failed.
531
ebcb507f
PH
532PH/40 RM_COMMAND is now used in the building process.
533
c35e155c
PH
534PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
535 the "build-* directories that it finds.
536
95d1f782
PH
537PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
538 address in a domain literal was a prefix of an interface address.
539
fd6de02e
PH
540PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
541 when verifying a sender address, unless rewrite_headers is false.
542
58de37c5
PH
543PH/44 Wrote a long comment about why errors_to addresses are verified as
544 recipients, not senders.
545
261cf466
TF
546TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
547 the ratelimit ACL was added.
548
3ee512ff
PH
549PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
550
e08c430f
PH
551PH/46 Added extra information about PostgreSQL errors to the error string.
552
bef5a11f
PH
553PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
554 avoiding the need to install special zones in a real server. This is
555 backwards compatible; if it can't find the fake resolver, it drops back.
556 Thus, both old and new test suites can be run.
557
7546de58
TF
558TF/02 Added util/ratelimit.pl
559
e5d5a95f
TF
560TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
561 clock is set back in time.
562
2e88a017
TF
563TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
564 Candler <B.Candler@pobox.com>.
565
a5f65aa4
TF
566TF/05 The fix for PH/43 was not completely correct; widen_domains is always
567 OK for addresses that are the result of redirections.
568
e7726cbf
PH
569PH/48 A number of further additions for the benefit of the new test suite,
570 including a fake gethostbyname() that interfaces to the fake DNS resolver
571 (see PH/47 above).
572
a7fdad5b
TF
573TF/06 The fix for widen_domains has also been applied to qualify_single and
574 search_parents which are the other dnslookup options that can cause
575 header rewrites.
576
6af56900
PH
577PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
578 type ("H").
579
0925ede6
PH
580PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
581
66afa403
TF
582TF/07 Exim produced the error message "an SRV record indicated no SMTP
583 service" if it encountered an MX record with an empty target hostname.
584 The message is now "an MX or SRV record indicated no SMTP service".
585
0154e85a
TF
586TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
587 if the DNS of the sending site is misconfigured. This is quite a
588 common situation. This change restores the behaviour of treating a
589 helo verification defer as a failure.
590
16f12c76
PH
591PH/51 If self=fail was set on a router, the bounce message did not include the
592 actual error message.
593
bbe902f0 594
e5a9dba6
PH
595Exim version 4.52
596-----------------
597
598TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
599
22c3b60b
PH
600PH/01 When a transport filter timed out in a pipe delivery, and the pipe
601 command itself ended in error, the underlying message about the transport
602 filter timeout was being overwritten with the pipe command error. Now the
603 underlying error message should be appended to the second error message.
604
06a9b4b5
PH
605TK/01 Fix poll() being unavailable on Mac OSX 10.2.
606
c1ac6996
PH
607PH/02 Reduce the amount of output that "make" produces by default. Full output
608 can still be requested.
609
9c7a242c
PH
610PH/03 The warning log line about a condition test deferring for a "warn" verb
611 was being output only once per connection, rather than after each
612 occurrence (because it was using the same function as for successful
613 "warn" verbs). This seems wrong, so I have changed it.
614
87ba3f5f
PH
615TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
616 it should not have, which might have caused a crash in the right
617 circumstances, but probably never did.
618
619PH/04 Installed a modified version of Tony Finch's patch to make submission
620 mode fix the return path as well as the Sender: header line, and to
621 add a /name= option so that you can make the user's friendly name appear
622 in the header line.
623
29aba418
TF
624TF/03 Added the control = fakedefer ACL modifier.
625
fe0dab11
TF
626TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
627 Mark Lowes for thorough testing.
870f6ba8 628
11d337a4
TK
629TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
630
631TK/03 Merged latest SRS patch from Miles Wilton.
632
415c8f3b
PH
633PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
634 with the definition in sysexits.h (which is #included earlier).
635 Fortunately, Exim does not actually use EX_OK. The code used to try to
636 preserve the sysexits.h value, by assumimg that macro definitions were
637 scanned for macro replacements. I have been disabused of this notion,
638 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 639
958541e9
PH
640PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
641 in the smtp transport. When a block could not be written in a single
642 write() function, the timeout was being re-applied to each part-write.
643 This seems wrong - if the receiver was accepting one byte at a time it
644 would take for ever. The timeout is now adjusted when this happens. It
645 doesn't have to be particularly precise.
646
c206415f
TK
647TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
648 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
649
2a4be8f9
PH
650PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
651 without a domain if the check to <postmaster@domain> fails.
652
1cba11c5
SC
653SC/01 Eximstats: added -xls and the ability to specify output files
654 (patch written by Frank Heydlauf).
655
656SC/02 Eximstats: use FileHandles for outputing results.
657
658SC/03 Eximstats: allow any combination of xls, txt, and html output.
659
660SC/04 Eximstats: fixed display of large numbers with -nvr option
661
662SC/05 Eximstats: fixed merging of reports with empty tables.
663
664SC/06 Eximstats: added the -include_original_destination flag
665
666SC/07 Eximstats: removed tabs and trailing whitespace.
667
1005d00e
TK
668TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
669
670TK/06 MBOX spool code: Add real "From " MBOX separator line
671 so the .eml file is really in mbox format (even though
672 most programs do not really care). Patch from Alex Miller.
673
674TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
675 The latter is generated from $received_to and is only set if the
676 message has one envelope recipient. SA can use these headers,
677 obviously out-of-the-box. Patch from Alex Miller.
678
9b4768fa
PH
679PH/08 The ${def test on a variable was returning false if the variable's
680 value was "0", contrary to what the specification has always said!
681 The result should be true unless the variable is empty.
682
683PH/09 The syntax error of a character other than { following "${if
684 def:variable_name" (after optional whitespace) was not being diagnosed.
685 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
686 accidental colon was present, for example, could give incorrect results.
687
0d7eb84a
PH
688PH/10 Tidied the code in a number of places where the st_size field of a stat()
689 result is used (not including appendfile, where other changes are about
690 to be made).
691
692PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
693 This involved changing a lot of size variables from int to off_t. It
694 should work with maildirs and everything.
695
40727bee
TK
696TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
697 spamd dying while we are connected to it.
698
554d2369
TF
699TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
700 <jgh@wizmail.org>
701
1f922db1
PH
702PH/12 Applied Alex Kiernan's patch for the API change for the error callback
703 function for BDB 4.3.
704
ef213c3b
PH
705PH/13 Changed auto_thaw such that it does not apply to bounce messages.
706
8ac170f3
PH
707PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
708 the sources for PCRE have been re-arranged and more files are now
709 involved.
710
b1c749bb
PH
711PH/15 The code I had for printing potentially long long variables in PH/11
712 above was not the best (it lost precision). The length of off_t variables
713 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
714 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
715 to be "long long int" or "long int". This is needed for the internal
716 formatting function string_vformat().
b1c749bb 717
4aac9b49
PH
718PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
719 the configuration file to be ":syslog", then the script "guesses" where
720 the logs files are, rather than using the compiled in default. In our
721 case the guess is not the same as the compiled default, so the script
722 suddenly stopped working when I started to use syslog. The patch checks
723 to see if log_file_path is "". If so, it attempts to read it from exim
724 with no configuration file to get the compiled in version, before it
725 falls back to the previous guessing code."
726
294520c8
TK
727TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
728 implementing BATV in an Exim configuration. See NewStuff for the gory
729 details.
730
5bd022fe
PH
731PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
732 Makefile that are specific to HP-UX.
733
90e9ce59
PH
734PH/18 If the "use_postmaster" option was set for a recipient callout together
735 with the "random" option, the postmaster address was used as the MAIL
736 FROM address for the random test, but not for the subsequent recipient
737 test. It is now used for both.
738
5ea81592
PH
739PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
740 patch removes a few documentation additions to RFC 3028, because the
741 latest draft now contains them. It adds the new en;ascii-case comparator
742 and a new error check for 8bit text in MIME parts. Comparator and
743 require names are now matched exactly. I enabled the subaddress
744 extension, but it is not well tested yet (read: it works for me)."
745
c6c2dc1d
PH
746PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
747 rework some of the code of TK/09 above to avoid the hardwired use of
748 "%lld" and "long long". Replaced the call to snprintf() with a call to
749 string_vformat().
750
fffffe4c
PH
751PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
752 records point to non-existent hosts", "retry timeout exceeded", and
753 "retry time not reached for any host after a long failure period".
ca02eafb 754
9a26b6b2
PH
755PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
756 experimental DomainKeys support:
757
758 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
759 (2) On an error such as an illegally used "control", the wrong name for
760 the control was given.
761
762 These problems did NOT occur unless DomainKeys support was compiled.
763
4aee0225
PH
764PH/23 Added daemon_startup_retries and daemon_startup_sleep.
765
32d668a5
PH
766PH/24 Added ${if match_ip condition.
767
8187c3f3
PH
768PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
769 databases so that it will be absolutely obvious if a crash occurs in the
770 DB library. This is a regular occurrence (often caused by mis-matched
771 db.h files).
772
ff790e47 773PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
774 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
775 picked up on a user's system that detects such things. There doesn't seem
776 to be a gcc warning option for this - only an attribute that has to be
777 put on the function's prototype. It seems that in Fedora Core 4 they have
778 set this on a number of new functions. No doubt there will be more in due
779 course.
ff790e47 780
5417f6d1
PH
781PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
782 specify a transport. However, if an address that was verified by such a
783 router was the subject of a callout, Exim crashed because it tried to
784 read the rcpt_include_affixes from the non-existent transport. Now it
785 just assumes that the setting of that option is false. This bug was
786 introduced by 4.51/PH/31.
787
59cf8544
PH
788PH/28 Changed -d+all to exclude +memory, because that information is very
789 rarely of interest, but it makes the output a lot bigger. People tend to
790 do -d+all out of habit.
791
e7ad8a65
PH
792PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
793 code in os-type was giving problems when libc.so lives in lib64, like on
794 x86_64 Fedora Core.
795
ade42478
PH
796PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
797 aren't the modern standard, and it seems that some systems' include files
798 don't always have them. Exim was already checking for some of the newer
799 ones like T_AAAA, and defining it itself. I've added checks for all the
800 record types that Exim uses.
801
182ad5cf
PH
802PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
803 not automatically generating a new one, as it is supposed to. This
804 prevented TLS from working. If the file did exist, but contained invalid
805 data, a new version was generated, as expected. It was only the case of a
806 non-existent file that was broken.
807
b0d9fc80
TK
808TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
809 with a change in libdomainkeys > 0.64.
810
811TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
812 from DNS. If the selector record carries the flag, it now has
813 precedence over the domain-wide flag.
814
815TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
816
47c7a64a
PH
817PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
818 the use of a port name that isn't defined in /etc/services) occurred, the
819 message was deferred as in a normal delivery, and thus remained on the
820 spool, instead of being failed because of the mua_wrapper setting. This
821 is now fixed, and I tidied up some of the mua_wrapper messages at the
822 same time.
823
a388bce4
SC
824SC/08 Eximstats: whilst parsing the mainlog(s), store information about
825 the messages in a hash of arrays rather than using individual hashes.
826 This is a bit cleaner and results in dramatic memory savings, albeit
827 at a slight CPU cost.
828
829SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
830 as requested by Marc Sherman.
831
832SC/10 Eximstats: added histograms for user specified patterns as requested
833 by Marc Sherman.
834
0793e4ed
SC
835SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
836
c58b88df
PH
837PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
838 fopen() in the content-scanning modules that did not already have it.
839
e7ad8a65 840
7982096b
PH
841Exim version 4.51
842-----------------
843
1a46a8c5
PH
844TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
845 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
846
2f079f46 847TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
848
849TK/03 Fix ACL "regex" condition not being available in MIME ACL.
850
851PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
852 to test Sieve filters that use "vacation".
853
854PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
855 that changes the way the GnuTLS parameters are stored in the cache file.
856 The new format can be generated externally. For backward compatibility,
857 if the data in the cache doesn't make sense, Exim assumes it has read an
858 old-format file, and it generates new data and writes a new file. This
859 means that you can't go back to an older release without removing the
860 file.
861
862PH/03 A redirect router that has both "unseen" and "one_time" set does not
863 work if there are any delivery delays because "one_time" forces the
864 parent to be marked "delivered", so its unseen clone is never tried
865 again. For this reason, Exim now forbids the simultaneous setting of
866 these two options.
867
868PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
869 redirected to themselves ("homonym" addresses). Read the long ChangeLog
870 entry if you want to know the details. The fix, however, neglected to
871 consider the case when local delivery batching is involved. The test for
872 "previously delivered" was not happening when checking to see if an
873 address could be batched with a previous (undelivered) one; under
874 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 875 address.
1a46a8c5
PH
876
877PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
878 in its include files, and this causes problems building Exim.
879
880PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
881 header_syntax) but Exim was just ignoring anything given after a slash.
882 In particular, this caused confusion with an attempt to use "verify =
883 reverse_host_lookup/defer_ok". An error is now given when options are
884 supplied for verify items that do not have them. (Maybe reverse_host_
885 lookup should have a defer_ok option, but that's a different point.)
886
887PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
888 defined by RFC 821) to 2048, because there were problems with some AUTH
889 commands, and RFC 1869 says the size should be increased for extended
890 SMTP commands that take arguments.
891
892PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
893 Finch).
894
895PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
896 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 897
49c2d5ea
PH
898PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
899 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
900 whatever). Otherwise the password may appear in the log. Following change
901 PH/42 below, there is no longer a chance of it appearing in a bounce
902 message.
49c2d5ea 903
bf759a8b
PH
904PH/11 Installed exipick version 20050225.0 from John Jetmore.
905
83364d30
PH
906PH/12 If the last host in a fallback_hosts list was multihomed, only the first
907 of its addresses was ever tried. (Bugzilla bug #2.)
908
7999bbd7
PH
909PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
910 the result incorrectly in the debug output. (It correctly added a newline
911 to what was transported.)
912
7dbf77c9
PH
913TF/01 Added $received_time.
914
74e0617f
PH
915PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
916 commented out examples of how to interface to a virus scanner and to
917 SpamAssassin. Also added commented examples of av_scanner and
918 spamd_address settings.
919
2f079f46
PH
920PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
921 and controls are allowed in which ACLs. There were a couple of minor
922 errors. Some of the entries in the conditions table (which is a table of
923 where they are NOT allowed) were getting very unwieldy; rewrote them as a
924 negation of where the condition IS allowed.
925
8c841523
PH
926PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
927
7766a4f0
PH
928PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
929 header file does not have a version number, so I've had to invent a new
930 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
931 API. The code is untested by me (my Linux distribution still has 0.3.2 of
932 radiusclient), but it was contributed by a Radius user.
933
8b417f2c
PH
934PH/18 Installed Lars Mainka's patch for the support of CRL collections in
935 files or directories, for OpenSSL.
936
901f42cb
PH
937PH/19 When an Exim process that is running as root has to create an Exim log
938 file, it does so in a subprocess that runs as exim:exim so as to get the
939 ownership right at creation (otherwise, other Exim processes might see
940 the file with the wrong ownership). There was no test for failure of this
941 fork() call, which would lead to the process getting stuck as it waited
942 for a non-existent subprocess. Forks do occasionally fail when resources
943 run out. I reviewed all the other calls to fork(); they all seem to check
944 for failure.
945
f9b9210e
PH
946PH/20 When checking for unexpected SMTP input at connect time (before writing
947 the banner), Exim was not dealing correctly with a non-positive return
948 from the read() function. If the client had disconnected by this time,
949 the result was a log entry for a synchronization error with an empty
950 string after "input=" when read() returned zero. If read() returned -1
951 (an event I could not check), uninitialized data bytes were printed.
952 There were reports of junk text (parts of files, etc) appearing after
953 "input=".
954
54cdb463
PH
955PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
956
cf00dad6
PH
957PH/22 Added support for macro redefinition, and (re)definition in between
958 driver and ACL definitions.
959
acb1b346
PH
960PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
961 forgetting to use the resulting value; it was using the unexpanded value.
962
c5ddb310
PH
963PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
964 hadn't been configured. The fix is from Juergen Kreileder, who
965 understands it better than I do:
966
967 "Here's what I see happening with three configured cyrus_sasl
968 authenticators configured (plain, login, cram-md5):
969
970 On startup auth_cyrus_sasl_init() gets called for each of these.
971 This means three calls to sasl_listmech() without a specified mech_list.
972 => SASL tests which mechs of all available mechs actually work
973 => three warnings about OTP not working
974 => the returned list contains: plain, login, cram-md5, digest-md5, ...
975
976 With the patch, sasl_listmech() also gets called three times. But now
977 SASL's mech_list option is set to the server_mech specified in the the
978 authenticator. Or in other words, the answer from sasl_listmech()
979 gets limited to just the mech you're testing for (which is different
980 for each call.)
981 => the return list contains just 'plain' or 'login', 'cram-md5' or
982 nothing depending on the value of ob->server_mech.
983
984 I've just tested the patch: Authentication still works fine,
985 unavailable mechs specified in the exim configuration are still
986 caught, and the auth.log warnings about OTP are gone."
987
31619da6
PH
988PH/25 When debugging is enabled, the contents of the command line are added
989 to the debugging output, even when log_selector=+arguments is not
990 specified.
991
bebaf0fc
PH
992PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
993 answer is "GNU", and only if the return is "GNU/something" is the answer
994 "Linux".
995
475fe28a
PH
996PH/27 $acl_verify_message is now set immediately after the failure of a
997 verification in an ACL, and so is available in subsequent modifiers. In
998 particular, the message can be preserved by coding like this:
999
1000 warn !verify = sender
1001 set acl_m0 = $acl_verify_message
1002
1003 Previously, $acl_verify_message was set only while expanding "message"
1004 and "log_message" when a very denied access.
1005
7e8bec7a
PH
1006PH/28 Modified OS/os.c-Linux with
1007
1008 -#ifndef OS_LOAD_AVERAGE
1009 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1010
1011 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1012 nomenclature these days.)
1013
e4a89c47
PH
1014PH/29 Installed patch from the Sieve maintainer that adds the options
1015 sieve_useraddress and sieve_subaddress to the redirect router.
1016
5ca2a9a1
PH
1017PH/30 In these circumstances:
1018 . Two addresses routed to the same list of hosts;
1019 . First host does not offer TLS;
1020 . First host accepts first address;
1021 . First host gives temporary error to second address;
1022 . Second host offers TLS and a TLS session is established;
1023 . Second host accepts second address.
1024 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1025 and peerdn, if requested) that were in fact used only for the second
1026 address.
7e8bec7a 1027
c688b954
PH
1028PH/31 When doing a callout as part of verifying an address, Exim was not paying
1029 attention to any local part prefix or suffix that was matched by the
1030 router that accepted the address. It now behaves in the same way as it
1031 does for delivery: the affixes are removed from the local part unless
1032 rcpt_include_affixes is set on the transport.
1033
fed77020
PH
1034PH/32 Add the sender address, as F=<...>, to the log line when logging a
1035 timeout during the DATA phase of an incoming message.
1036
7fe1560f
PH
1037PH/33 Sieve envelope tests were broken for match types other than :is. I have
1038 applied a patch sanctioned by the Sieve maintainer.
c688b954 1039
ebb6e6d5
PH
1040PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1041 the uid or gid is negative. A case of a negative gid caused this to be
1042 noticed. The fix allows for either to be negative.
1043
9c4e8f60
PH
1044PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1045 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1046 overlooked.
1047
1048PH/36 The change PH/12 above was broken. Fixed it.
1049
d7174846
PH
1050PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1051 the grounds that routing the same address twice would always produce the
1052 same answer. This might have been true once, but it is certainly no
1053 longer true now. Routing a child address may depend on the previous
1054 routing that produced that child. Some complicated redirection strategies
1055 went wrong when messages had multiple recipients, and made Exim's
1056 behaviour dependent on the order in which the addresses were given.
1057
1058 I have moved the duplicate checking until after the routing is complete.
1059 Exim scans the addresses that are assigned to local and remote
1060 transports, and removes any duplicates. This means that more work will be
1061 done, as duplicates will always all be routed, but duplicates are
1062 presumably rare, so I don't expect this is of any significance.
1063
1064 For deliveries to pipes, files, and autoreplies, the duplicate checking
1065 still happens during the routing process, since they are not going to be
1066 routed further.
1067
cfe75fc3
PH
1068PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1069 It corrects a timeout issue with spamd. This is Ian's comment: "The
1070 background is that sometimes spamd either never reads data from a
1071 connection it has accepted, or it never writes response data. The exiscan
1072 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1073 blindly assumes that writes won't block so it may never time out."
1074
be22d70e
PH
1075PH/39 Allow G after quota size as well as K and M.
1076
0612b098
PH
1077PH/40 The value set for $authenticated_id in an authenticator may not contain
1078 binary zeroes or newlines because the value is written to log lines and
1079 to spool files. There was no check on this. Now the value is run through
1080 the string_printing() function so that such characters are converted to
1081 printable escape sequences.
1082
2e0c1448
PH
1083PH/41 $message_linecount is a new variable that contains the total number of
1084 lines in the message. Compare $body_linecount, which is the count for the
1085 body only.
1086
447d236c
PH
1087PH/42 Exim no longer gives details of delivery errors for specific addresses in
1088 bounce and delay warning messages, except in certain special cases, which
1089 are as follows:
1090
1091 (a) An SMTP error message from a remote host;
1092 (b) A message specified in a :fail: redirection;
1093 (c) A message specified in a "fail" command in a system filter;
1094 (d) A message specified in a FAIL return from the queryprogram router;
1095 (e) A message specified by the cannot_route_message router option.
1096
1097 In these cases only, Exim does include the error details in bounce and
1098 warning messages. There are also a few cases where bland messages such
1099 as "unrouteable address" or "local delivery error" are given.
1100
d20976dc
PH
1101PH/43 $value is now also set for the "else" part of a ${run expansion.
1102
f656d135
PH
1103PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1104 being worked on, but at least Exim now implements the latest version to
1105 play with."
1106
2e2a30b4
PH
1107PH/45 In a pipe transport, although a timeout while waiting for the pipe
1108 process to complete was treated as a delivery failure, a timeout while
1109 writing the message to the pipe was logged, but erroneously treated as a
1110 successful delivery. Such timeouts include transport filter timeouts. For
1111 consistency with the overall process timeout, these timeouts are now
1112 treated as errors, giving rise to delivery failures by default. However,
1113 there is now a new Boolean option for the pipe transport called
1114 timeout_defer, which, if set TRUE, converts the failures into defers for
1115 both kinds of timeout. A transport filter timeout is now identified in
1116 the log output.
1117
9176e9f0
PH
1118PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1119 systems where "make" and "gmake" are different, calling "gmake" at top
1120 level broke things. I've arranged for the value of $(MAKE) to be passed
1121 from the Makefile to this script so that it can call the same version of
1122 "make".
1123
7982096b 1124
bbe902f0
PH
1125A note about Exim versions 4.44 and 4.50
1126----------------------------------------
1127
1128Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1129changes of various kinds. As a consequence, a big documentation update was
1130needed. This delayed the release for rather longer than seemed good, especially
1131in the light of a couple of (minor) security issues. Therefore, the changes
1132that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1133release. So 4.44 and 4.50 are in effect two different branches that both start
1134from 4.43.
1135
1136I have left the 4.50 change log unchanged; it contains all the changes since
11374.43. The change log for 4.44 is below; many of its items are identical to
1138those for 4.50. This seems to be the most sensible way to preserve the
1139historical information.
1140
1141
f7b63901 1142Exim version 4.50
495ae4b0
PH
1143-----------------
1144
5fe762f6
PH
1145 1. Minor wording change to the doc/README.SIEVE file.
1146
139059f6 1147 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1148 computation of the current number of files was incorrect.
495ae4b0 1149
7086e875
PH
1150 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1151 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1152 place.
1153
35af9f61
PH
1154 4. Give more explanation in the error message when the command for a transport
1155 filter fails to execute.
1156
b668c215
PH
1157 5. There are several places where Exim runs a non-Exim command in a
1158 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1159 was being done only for the command run by the queryprogram router. It is
1160 now done for all such subprocesses. The other cases are: ${run, transport
1161 filters, and the commands run by the lmtp and pipe transports.
1162
a494b1e1
PH
1163 6. Added CONFIGURE_GROUP build-time option.
1164
1165 7. Some older OS have a limit of 256 on the maximum number of file
1166 descriptors. Exim was using setrlimit() to set 1000 as a large value
1167 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1168 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1169
c5fcb476
PH
1170 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1171 was an oversight, and furthermore, ever since the addition of extra
1172 controls (e.g. 4.43/32), the checks on when to allow different forms of
1173 "control" were broken. There should now be diagnostics for all cases when a
1174 control that does not make sense is encountered.
1175
69358f02
PH
1176 9. Added the /retain_sender option to "control=submission".
1177
5be20824
PH
117810. $recipients is now available in the predata ACL (oversight).
1179
eb2c0248
PH
118011. Tidy the search cache before the fork to do a delivery from a message
1181 received from the command line. Otherwise the child will trigger a lookup
1182 failure and thereby defer the delivery if it tries to use (for example) a
1183 cached ldap connection that the parent has called unbind on.
1184
2a3eea10
PH
118512. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1186 of $address_data from the recipient verification was clobbered by the
1187 sender verification.
1188
118913. The value of address_data from a sender verification is now available in
1190 $sender_address_data in subsequent conditions in the ACL statement.
1191
23c7ff99
PH
119214. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1193
4deaf07d
PH
119415. Added a new option "connect=<time>" to callout options, to set a different
1195 connection timeout.
1196
926e1192
PH
119716. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1198 was its contents. (It was OK if the option was not defined at all.)
1199
650edc6f
PH
120017. A "Completed" log line is now written for messages that are removed from
1201 the spool by the -Mrm option.
1202
2c7db3f5
PH
120318. New variables $sender_verify_failure and $recipient_verify_failure contain
1204 information about exactly what failed.
1205
3d235903
PH
120619. Added -dd to debug only the daemon process.
1207
7c7ad977
PH
120820. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1209 handles timeouts, both on the server side and network timeouts. Renamed the
1210 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1211
981756db
PH
121221. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1213 It is now set to "smtps".
1214
d4eb88df
PH
121522. $host_address is now set to the target address during the checking of
1216 ignore_target_hosts.
1217
121823. When checking ignore_target_hosts for an ipliteral router, no host name was
1219 being passed; this would have caused $sender_host_name to have been used if
1220 matching the list had actually called for a host name (not very likely,
1221 since this list is usually IP addresses). A host name is now passed as
1222 "[x.x.x.x]".
1223
7d468ab8
PH
122424. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1225 code that specifies a non-restarting handler (typically sigaction() in
1226 modern systems) in an attempt to fix a rare and obscure crash bug.
1227
122825. Narrowed the window for a race in the daemon that could cause it to ignore
1229 SIGCHLD signals. This is not a major problem, because they are used only to
1230 wake it up if nothing else does.
1231
62c0818f
PH
123226. A malformed maildirsize file could cause Exim to calculate negative values
1233 for the mailbox size or file count. Odd effects could occur as a result.
1234 The maildirsize information is now recalculated if the size or filecount
1235 end up negative.
1236
26034054
PH
123727. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1238 support for a long time. Removed HAVE_SYS_VFS_H.
1239
af66f652
PH
124028. Installed the latest version of exipick from John Jetmore.
1241
90af77f4
PH
124229. In an address list, if the pattern was not a regular expression, an empty
1243 subject address (from a bounce message) matched only if the pattern was an
1244 empty string. Non-empty patterns were not even tested. This was the wrong
1245 because it is perfectly reasonable to use an empty address as part of a
1246 database query. An empty address is now tested by patterns that are
1247 lookups. However, all the other forms of pattern expect the subject to
1248 contain a local part and a domain, and therefore, for them, an empty
1249 address still always fails if the pattern is not itself empty.
1250
d8ef3577
PH
125130. Exim went into a mad DNS loop when attempting to do a callout where the
1252 host was specified on an smtp transport, and looking it up yielded more
1253 than one IP address.
1254
5cb8cbc6
PH
125531. Re-factored the code for checking spool and log partition space into a
1256 function that finds that data and another that does the check. The former
1257 is then used to implement four new variables: $spool_space, $log_space,
1258 $spool_inodes, and $log_inodes.
1259
14702f5b
PH
126032. The RFC2047 encoding function was originally intended for short strings
1261 such as real names; it was not keeping to the 75-character limit for
1262 encoded words that the RFC imposes. It now respects the limit, and
1263 generates multiple encoded words if necessary. To be on the safe side, I
1264 have increased the buffer size for the ${rfc2047: expansion operator from
1265 1024 to 2048 bytes.
1266
063b1e99
PH
126733. It is now permitted to omit both strings after an "if" condition; if the
1268 condition is true, the result is "true". As before, when the second string
1269 is omitted, a false condition yields an empty string. This makes it less
1270 cumbersome to write custom ACL and router conditions.
1271
652e1b65
PH
127234. Failure to deliver a bounce message always caused it to be frozen, even if
1273 there was an errors_to setting on the router. The errors_to setting is now
1274 respected.
1275
6f0c9a4f
PH
127635. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1277 canonical form (fully expanded) before being placed in
1278 $sender_host_address.
1279
33397d19
PH
128036. The table in the code that translates DNS record types into text (T_A to
1281 "A" for instance) was missing entries for NS and CNAME. It is just possible
1282 that this could have caused confusion if both these types were looked up
1283 for the same domain, because the text type is used as part of Exim's
1284 per-process caching. But the chance of anyone hitting this buglet seems
1285 very small.
1286
7bb56e1f
PH
128737. The dnsdb lookup has been extended in a number of ways.
1288
1289 (1) There is a new type, "zns", which walks up the domain tree until it
1290 finds some nameserver records. It should be used with care.
1291
ea3bc19b
PH
1292 (2) There is a new type, "mxh", which is like "mx" except that it returns
1293 just the host names, not the priorities.
1294
1295 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1296 looked up. The behaviour when one of the lookups defers can be
1297 controlled by a keyword.
7bb56e1f 1298
ea3bc19b 1299 (4) It is now possible to specify the separator character for use when
7bb56e1f 1300 multiple records are returned.
33397d19 1301
0bcb2a0e
PH
130238. The dnslists ACL condition has been extended: it is now possible to supply
1303 a list of IP addresses and/or domains to be looked up in a particular DNS
1304 domain.
1305
2ac0e484
PH
130639. Added log_selector=+queue_time_overall.
1307
4e1fde53
PH
130840. When running the queue in the test harness, wait just a tad after forking a
1309 delivery process, to get repeatability of debugging output.
1310
de365ded
PH
131141. Include certificate and key file names in error message when GnuTLS fails
1312 to set them up, because the GnuTLS error message doesn't include the name
1313 of the failing file when there is a problem reading it.
1314
f05da2e8
PH
131542. Allow both -bf and -bF in the same test run.
1316
d6453af2
PH
131743. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1318
f7b63901
PH
131944. The "Exiscan patch" is now merged into the mainline Exim source.
1320
132145. Sometimes the final signoff response after QUIT could fail to get
1322 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1323 < 0 before doing a fflush(). This bug looks as though it goes back to the
1324 introduction of TLS in release 3.20, but "sometimes" must have been rare
1325 because the tests only now provoked it.
1326
a444213a
PH
132746. Reset the locale to "C" after calling embedded Perl, in case it was changed
1328 (this can affect the format of dates).
1329
0ec020ea
PH
133047. exim_tidydb, when checking for the continued existence of a message for
1331 which it has found a message-specific retry record, was not finding
1332 messages that were in split spool directories. Consequently, it was
1333 deleting retry records that should have stayed in existence.
1334
b1206957
PH
133548. Steve fixed some bugs in eximstats.
1336
133749. The SPA authentication driver was not abandoning authentication and moving
1338 on to the next authenticator when an expansion was forced to fail,
1339 contradicting the general specification for all authenticators. Instead it
1340 was generating a temporary error. It now behaves as specified.
1341
26dd5a95
PH
134250. The default ordering of permitted cipher suites for GnuTLS was pessimal
1343 (the order specifies the preference for clients). The order is now AES256,
1344 AES128, 3DES, ARCFOUR128.
1345
343b2385
PH
134651. Small patch to Sieve code - explicitly set From: when generating an
1347 autoreply.
1348
1c5466b9
PH
134952. Exim crashed if a remote delivery caused a very long error message to be
1350 recorded - for instance if somebody sent an entire SpamAssassin report back
1351 as a large number of 550 error lines. This bug was coincidentally fixed by
1352 increasing the size of one of Exim's internal buffers (big_buffer) that
1353 happened as part of the Exiscan merge. However, to be on the safe side, I
1354 have made the code more robust (and fixed the comments that describe what
1355 is going on).
1356
55ee9ee3
PH
135753. Now that there can be additional text after "Completed" in log lines (if
1358 the queue_time_overall log selector is set), a one-byte patch to exigrep
1359 was needed to allow it to recognize "Completed" as not the last thing in
1360 the line.
1361
d38f8232
PH
136254. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1363 patch that reportedly fixes this has been added. I am not expert enough to
1364 create a test for it. This is what the patch creator wrote:
1365
1366 "I found a little strange behaviour of ldap code when working with
1367 Windows 2003 AD Domain, where users was placed in more than one
1368 Organization Units. When I tried to give exim partial DN, the exit code
1369 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1370 But simultaneously result of request was absolutely normal ldap result,
1371 so I produce this patch..."
1372
3295e65b
PH
1373 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1374 REFERENCE, so I have modified the code to exclude the patch when that macro
1375 is not defined.
1376
7102e136
PH
137755. Some experimental protocols are using DNS PTR records for new purposes. The
1378 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1379 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1380 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1381 CAN-2005-0021
7102e136 1382
3ca0ba97
PH
138356. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1384
c2bcbe20
PH
138557. Double the size of the debug message buffer (to 2048) so that more of very
1386 long debug lines gets shown.
1387
18ce445d
PH
138858. The exicyclog utility now does better if the number of log files to keep
1389 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1390
1f5b4c3d
PH
139159. Two changes related to the smtp_active_hostname option:
1392
1393 (1) $smtp_active_hostname is now available as a variable.
1394 (2) The default for smtp_banner uses $smtp_active_hostname instead
1395 of $primary_hostname.
1396
b975ba52
PH
139760. The host_aton() function is supposed to be passed a string that is known
1398 to be a valid IP address. However, in the case of IPv6 addresses, it was
1399 not checking this. This is a hostage to fortune. Exim now panics and dies
1400 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1401 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1402 components; fortuitously, this particular loophole had already been fixed
1403 by change 4.50/55 above.
1404
1405 If there are any other similar loopholes, the new check in host_aton()
1406 itself should stop them being exploited. The report I received stated that
1407 data on the command line could provoke the exploit when Exim was running as
1408 exim, but did not say which command line option was involved. All I could
1409 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1410 running as the user.
ea3a6f44 1411 CAN-2005-0021
85b87bc2
PH
1412
141361. There was a buffer overflow vulnerability in the SPA authentication code
1414 (which came originally from the Samba project). I have added a test to the
1415 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1416 CAN-2005-0022
b975ba52 1417
17ffcae7
PH
141862. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1419 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1420
d95f9fdb
PH
142163. The daemon start-up calls getloadavg() while still root for those OS that
1422 need the first call to be done as root, but it missed one case: when
1423 deliver_queue_load_max is set with deliver_drop_privilege. This is
1424 necessary for the benefit of the queue runner, because there is no re-exec
1425 when deliver_drop_privilege is set.
1426
86b8287f
PH
142764. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1428 This has been fixed.
1429
60dc5e56
PH
143065. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1431 was in use, was not putting the data itself into the right store pool;
1432 consequently, it could be overwritten for a subsequent message in the same
1433 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1434 the caching.)
1435
533244af
PH
143666. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1437
a5a28604
PH
143867. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1439 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1440 the function were treating the return as a boolean value, which happened to
1441 work because 0=false and not-0=true, but is not correct code.
1442
7e634d24
PH
144368. The host_aton() function was not handling scoped IPv6 addresses (those
1444 with, for example, "%eth0" on the end) correctly.
1445
3e11c26b
PH
144669. Fixed some compiler warnings in acl.c for the bitmaps specified with
1447 negated items (that is, ~something) in unsigned ints. Some compilers
1448 apparently mutter when there is no cast.
1449
6729cf78
PH
145070. If an address verification called from an ACL failed, and did not produce a
1451 user-specific message (i.e. there was only a "system" message), nothing was
1452 put in $acl_verify_message. In this situation, it now puts the system
1453 message there.
1454
00f00ca5
PH
145571. Change 4.23/11 added synchronization checking at the start of an SMTP
1456 session; change 4.31/43 added the unwanted input to the log line - except
1457 that it did not do this in the start of session case. It now does.
1458
c9bdd01c
PH
145972. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1460 This is wrong and can cause the other end to generate a synchronization
1461 error if it is another Exim or anything else that does the synchronization
1462 check. A QUIT command is no longer sent after a timeout.
1463
d43194df
PH
146473. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1465 during host lookups.
1466
fe5b5d0b
PH
146774. The defer_ok option of callout verification was not working if it was used
1468 when verifying addresses in header lines, that is, for this case:
1469
1470 verify = header_sender/callout=defer_ok
1471
76a2d7ba
PH
147275. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1473 those file descriptors could be used for SMTP connections. If anything
1474 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1475 could be sent to the SMTP client, causing chaos. The daemon now opens
1476 stdin, stdout, and stderr to /dev/null when it puts itself into the
1477 background.
1478
147976. Arrange for output from Perl's "warn" command to be written to Exim's main
1480 log by default. The user can override this with suitable Perl magic.
1481
04f7d5b9
PH
148277. The use of log_message on a "discard" ACL verb, which is supposed to add to
1483 the log message when discard triggers, was not working for the DATA ACL or
1484 for the non-SMTP ACL.
1485
bc60667e
PH
148678. Error message wording change in sieve.c.
1487
bb6e88ff
PH
148879. If smtp_accept_max_per_host was set, the number of connections could be
1489 restricted to fewer than expected, because the daemon was trying to set up
1490 a new connection before checking whether the processes handling previous
1491 connections had finished. The check for completed processes is now done
1492 earlier. On busy systems, this bug wouldn't be noticed because something
1493 else would have woken the daemon, and it would have reaped the completed
1494 process earlier.
1495
1e70f85b
PH
149680. If a message was submitted locally by a user whose login name contained one
1497 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1498 It caused a spool format error. I have fixed the spool reading code. A
1499 related problem was that the "from" clause in the Received: line became
1500 illegal because of the space(s). It is now covered by ${quote_local_part.
1501
150281. Included the latest eximstats from Steve (adds average sizes to HTML Top
1503 tables).
1504
4e01f9d6
PH
150582. Updated OS/Makefile-AIX as per message from Mike Meredith.
1506
1ee1cef2
PH
150783. Patch from Sieve maintainer to fix unterminated string problem in
1508 "vacation" handling.
1509
6e2b4ccc
PH
151084. Some minor changes to the Linux configuration files to help with other
1511 OS variants using glibc.
1512
8e669ac1
PH
151385. One more patch for Sieve to update vacation handling to latest spec.
1514
495ae4b0 1515
bbe902f0
PH
1516----------------------------------------------------
1517See the note above about the 4.44 and 4.50 releases.
1518----------------------------------------------------
1519
1520
1521Exim version 4.44
1522-----------------
1523
1524 1. Change 4.43/35 introduced a bug that caused file counts to be
1525 incorrectly computed when quota_filecount was set in an appendfile
1526 transport
1527
1528 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1529 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1530 place.
1531
1532 3. Give more explanation in the error message when the command for a transport
1533 filter fails to execute.
1534
1535 4. There are several places where Exim runs a non-Exim command in a
1536 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1537 was being done only for the command run by the queryprogram router. It is
1538 now done for all such subprocesses. The other cases are: ${run, transport
1539 filters, and the commands run by the lmtp and pipe transports.
1540
1541 5. Some older OS have a limit of 256 on the maximum number of file
1542 descriptors. Exim was using setrlimit() to set 1000 as a large value
1543 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1544 systems. I've change it so that if it can't get 1000, it tries for 256.
1545
1546 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1547 was an oversight, and furthermore, ever since the addition of extra
1548 controls (e.g. 4.43/32), the checks on when to allow different forms of
1549 "control" were broken. There should now be diagnostics for all cases when a
1550 control that does not make sense is encountered.
1551
1552 7. $recipients is now available in the predata ACL (oversight).
1553
1554 8. Tidy the search cache before the fork to do a delivery from a message
1555 received from the command line. Otherwise the child will trigger a lookup
1556 failure and thereby defer the delivery if it tries to use (for example) a
1557 cached ldap connection that the parent has called unbind on.
1558
1559 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1560 of $address_data from the recipient verification was clobbered by the
1561 sender verification.
1562
156310. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1564 was its contents. (It was OK if the option was not defined at all.)
1565
156611. A "Completed" log line is now written for messages that are removed from
1567 the spool by the -Mrm option.
1568
156912. $host_address is now set to the target address during the checking of
1570 ignore_target_hosts.
1571
157213. When checking ignore_target_hosts for an ipliteral router, no host name was
1573 being passed; this would have caused $sender_host_name to have been used if
1574 matching the list had actually called for a host name (not very likely,
1575 since this list is usually IP addresses). A host name is now passed as
1576 "[x.x.x.x]".
1577
157814. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1579 code that specifies a non-restarting handler (typically sigaction() in
1580 modern systems) in an attempt to fix a rare and obscure crash bug.
1581
158215. Narrowed the window for a race in the daemon that could cause it to ignore
1583 SIGCHLD signals. This is not a major problem, because they are used only to
1584 wake it up if nothing else does.
1585
158616. A malformed maildirsize file could cause Exim to calculate negative values
1587 for the mailbox size or file count. Odd effects could occur as a result.
1588 The maildirsize information is now recalculated if the size or filecount
1589 end up negative.
1590
159117. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1592 support for a long time. Removed HAVE_SYS_VFS_H.
1593
ea3a6f44 159418. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1595
159619. Allow an empty sender to be matched against a lookup in an address list.
1597 Previously the only cases considered were a regular expression, or an
1598 empty pattern.
1599
160020. Exim went into a mad DNS lookup loop when doing a callout where the
1601 host was specified on the transport, if the DNS lookup yielded more than
1602 one IP address.
1603
ea3a6f44
NM
160421. The RFC2047 encoding function was originally intended for short strings
1605 such as real names; it was not keeping to the 75-character limit for
1606 encoded words that the RFC imposes. It now respects the limit, and
1607 generates multiple encoded words if necessary. To be on the safe side, I
1608 have increased the buffer size for the ${rfc2047: expansion operator from
1609 1024 to 2048 bytes.
bbe902f0 1610
ea3a6f44
NM
161122. Failure to deliver a bounce message always caused it to be frozen, even if
1612 there was an errors_to setting on the router. The errors_to setting is now
1613 respected.
bbe902f0
PH
1614
161523. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1616 canonical form (fully expanded) before being placed in
1617 $sender_host_address.
1618
161924. Updated eximstats to version 1.33
1620
ea3a6f44
NM
162125. Include certificate and key file names in error message when GnuTLS fails
1622 to set them up, because the GnuTLS error message doesn't include the name
1623 of the failing file when there is a problem reading it.
bbe902f0
PH
1624
162526. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1626 As per change 25.
bbe902f0 1627
ea3a6f44
NM
162827. Reset the locale to "C" after calling embedded Perl, in case it was changed
1629 (this can affect the format of dates).
bbe902f0 1630
ea3a6f44
NM
163128. exim_tidydb, when checking for the continued existence of a message for
1632 which it has found a message-specific retry record, was not finding
1633 messages that were in split spool directories. Consequently, it was
1634 deleting retry records that should have stayed in existence.
bbe902f0
PH
1635
163629. eximstats updated to version 1.35
1637 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1638 1.35 - bugfix such that pie charts by volume are generated correctly
1639
ea3a6f44
NM
164030. The SPA authentication driver was not abandoning authentication and moving
1641 on to the next authenticator when an expansion was forced to fail,
1642 contradicting the general specification for all authenticators. Instead it
1643 was generating a temporary error. It now behaves as specified.
bbe902f0 1644
ea3a6f44
NM
164531. The default ordering of permitted cipher suites for GnuTLS was pessimal
1646 (the order specifies the preference for clients). The order is now AES256,
1647 AES128, 3DES, ARCFOUR128.
bbe902f0 1648
ea3a6f44
NM
164931. Small patch to Sieve code - explicitly set From: when generating an
1650 autoreply.
bbe902f0 1651
ea3a6f44
NM
165232. Exim crashed if a remote delivery caused a very long error message to be
1653 recorded - for instance if somebody sent an entire SpamAssassin report back
1654 as a large number of 550 error lines. This bug was coincidentally fixed by
1655 increasing the size of one of Exim's internal buffers (big_buffer) that
1656 happened as part of the Exiscan merge. However, to be on the safe side, I
1657 have made the code more robust (and fixed the comments that describe what
1658 is going on).
bbe902f0 1659
ea3a6f44
NM
166033. Some experimental protocols are using DNS PTR records for new purposes. The
1661 keys for these records are domain names, not reversed IP addresses. The
1662 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1663 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1664 CAN-2005-0021
1665
ea3a6f44
NM
166634. The host_aton() function is supposed to be passed a string that is known
1667 to be a valid IP address. However, in the case of IPv6 addresses, it was
1668 not checking this. This is a hostage to fortune. Exim now panics and dies
1669 if the condition is not met. A case was found where this could be provoked
1670 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1671 components; fortuitously, this particular loophole had already been fixed
1672 by change 4.50/55 or 4.44/33 above.
1673
1674 If there are any other similar loopholes, the new check in host_aton()
1675 itself should stop them being exploited. The report I received stated that
1676 data on the command line could provoke the exploit when Exim was running as
1677 exim, but did not say which command line option was involved. All I could
1678 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1679 running as the user.
bbe902f0
PH
1680 CAN-2005-0021
1681
ea3a6f44
NM
168235. There was a buffer overflow vulnerability in the SPA authentication code
1683 (which came originally from the Samba project). I have added a test to the
1684 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1685 CAN-2005-0022
1686
ea3a6f44
NM
168736. The daemon start-up calls getloadavg() while still root for those OS that
1688 need the first call to be done as root, but it missed one case: when
1689 deliver_queue_load_max is set with deliver_drop_privilege. This is
1690 necessary for the benefit of the queue runner, because there is no re-exec
1691 when deliver_drop_privilege is set.
bbe902f0 1692
ea3a6f44
NM
169337. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1694 was in use, was not putting the data itself into the right store pool;
1695 consequently, it could be overwritten for a subsequent message in the same
1696 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1697 the caching.)
bbe902f0 1698
ea3a6f44
NM
169938. Sometimes the final signoff response after QUIT could fail to get
1700 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1701 < 0 before doing a fflush(). This bug looks as though it goes back to the
1702 introduction of TLS in release 3.20, but "sometimes" must have been rare
1703 because the tests only now provoked it.
bbe902f0
PH
1704
1705
495ae4b0
PH
1706Exim version 4.43
1707-----------------
1708
1709 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1710 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1711 Now it returns an integer. A number of calls were still expecting a T/F
1712 return. Fortuitously, in all cases, the tests worked in OK situations,
1713 which is the norm. However, things would have gone wrong on any write
1714 failures on the smtp file descriptor. This function is used when sending
1715 messages over SMTP and also when doing verify callouts.
1716
1717 2. When Exim is called to do synchronous delivery of a locally submitted
1718 message (the -odf or -odi options), it no longer closes stderr before doing
1719 the delivery.
1720
1721 3. Implemented the mua_wrapper option.
1722
1723 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1724
1725 5. Implemented the functions header_remove(), header_testname(),
1726 header_add_at_position(), and receive_remove_recipient(), and exported them
1727 to local_scan().
1728
1729 6. If an ACL "warn" statement specified the addition of headers, Exim already
1730 inserted X-ACL-Warn: at the start if there was no header name. However, it
1731 was not making this test for the second and subsequent header lines if
1732 there were newlines in the string. This meant that an invalid header could
1733 be inserted if Exim was badly configured.
1734
1735 7. Allow an ACL "warn" statement to add header lines at the start or after all
1736 the Received: headers, as well as at the end.
1737
1738 8. Added the rcpt_4xx retry error code.
1739
1740 9. Added postmaster_mailfrom=xxx to callout verification option.
1741
174210. Added mailfrom=xxxx to the callout verification option, for verify=
1743 header_sender only.
1744
174511. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1746 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1747
174812. Inserted some casts to stop certain compilers warning when using pointer
1749 differences as field lengths or precisions in printf-type calls (mostly
1750 affecting debugging statements).
1751
175213. Added optional readline() support for -be (dynamically loaded).
1753
175414. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1755 same clock tick as a message's arrival, so that its received time was the
1756 same as the "first fail" time on the retry record, and that message
1757 remained on the queue past the ultimate address timeout, every queue runner
1758 would try a delivery (because it was past the ultimate address timeout) but
1759 after another failure, the ultimate address timeout, which should have then
1760 bounced the address, did not kick in. This was a "< instead of <=" error;
1761 in most cases the first failure would have been in the next clock tick
1762 after the received time, and all would be well.
1763
176415. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1765 being recognized when the domain list was tested by the match_domain
1766 condition in an expansion string.
1767
176816. Added the ${str2b64: operator.
1769
177017. Exim was always calling setrlimit() to set a large limit for the number of
1771 processes, without checking whether the existing limit was already
1772 adequate. (It did check for the limit on file descriptors.) Furthermore,
1773 errors from getrlimit() and setrlimit() were being ignored. Now they are
1774 logged to the main and panic logs, but Exim does carry on, to try to do its
1775 job under whatever limits there are.
1776
177718. Imported PCRE 5.0.
1778
177919. Trivial typo in log message " temporarily refused connection" (the leading
1780 space).
1781
178220. If the log selector return_path_on_delivery was set and an address was
1783 redirected to /dev/null, the delivery process crashed because it assumed
1784 that a return path would always be set for a "successful" delivery. In this
1785 case, the whole delivery is bypassed as an optimization, and therefore no
1786 return path is set.
1787
178821. Internal re-arrangement: the function for sending a challenge and reading
1789 a response while authentication was assuming a zero-terminated challenge
1790 string. It's now changed to take a pointer and a length, to allow for
1791 binary data in such strings.
1792
179322. Added the cyrus_sasl authenticator (code supplied by MBM).
1794
179523. Exim was not respecting finduser_retries when seeking the login of the
1796 uid under which it was called; it was always trying 10 times. (The default
1797 setting of finduser_retries is zero.) Also, it was sleeping after the final
1798 failure, which is pointless.
1799
180024. Implemented tls_on_connect_ports.
1801
180225. Implemented acl_smtp_predata.
1803
180426. If the domain in control=submission is set empty, Exim assumes that the
1805 authenticated id is a complete email address when it generates From: or
1806 Sender: header lines.
1807
180827. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1809 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1810 chown and chgrp in /bin and hostname in /usr/bin.
1811
181228. Exim was keeping the "process log" file open after each use, just as it
1813 does for the main log. This opens the possibility of it remaining open for
1814 long periods when the USR1 signal hits a daemon. Occasional processlog
1815 errors were reported, that could have been caused by this. Anyway, it seems
1816 much more sensible not to leave this file open at all, so that is what now
1817 happens.
1818
181929. The long-running daemon process does not normally write to the log once it
1820 has entered its main loop, and it closes the log before doing so. This is
1821 so that log files can straightforwardly be renamed and moved. However,
1822 there are a couple of unusual error situations where the daemon does write
1823 log entries, and I had neglected to close the log afterwards.
1824
182530. The text of an SMTP error response that was received during a remote
1826 delivery was being truncated at 512 bytes. This is too short for some of
1827 the long messages that one sometimes sees. I've increased the limit to
1828 1024.
1829
183031. It is now possible to make retry rules that apply only when a message has a
1831 specific sender, in particular, an empty sender.
1832
183332. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1834 it possible to be selective about when SMTP synchronization is enforced.
1835
183633. Added "control = caseful_local_part" and "control = "caselower_local_part".
1837
183832. Implemented hosts_connection_nolog.
1839
184033. Added an ACL for QUIT.
1841
184234. Setting "delay_warning=" to disable warnings was not working; it gave a
1843 syntax error.
1844
184535. Added mailbox_size and mailbox_filecount to appendfile.
1846
184736. Added control = no_multiline_responses to ACLs.
1848
184937. There was a bug in the logic of the code that waits for the clock to tick
1850 in the case where the clock went backwards by a substantial amount such
1851 that the microsecond fraction of "now" was more than the microsecond
1852 fraction of "then" (but the whole seconds number was less).
1853
185438. Added support for the libradius Radius client library this is found on
1855 FreeBSD (previously only the radiusclient library was supported).
1856
1857
1858Exim version 4.42
1859-----------------
1860
1861 1. When certain lookups returned multiple values in the form name=value, the
1862 quoting of the values was not always being done properly. Specifically:
1863 (a) If the value started with a double quote, but contained no whitespace,
1864 it was not quoted.
1865 (b) If the value contained whitespace other than a space character (i.e.
1866 tabs or newlines or carriage returns) it was not quoted.
1867 This fix has been applied to the mysql and pgsql lookups by writing a
1868 separate quoting function and calling it from the lookup code. The fix
1869 should probably also be applied to nisplus, ibase and oracle lookups, but
1870 since I cannot test any of those, I have not disturbed their existing code.
1871
1872 2. A hit in the callout cache for a specific address caused a log line with no
1873 reason for rejecting RCPT. Now it says "Previous (cached) callout
1874 verification failure".
1875
1876 3. There was an off-by-one bug in the queryprogram router. An over-long
1877 return line was truncated at 256 instead of 255 characters, thereby
1878 overflowing its buffer with the terminating zero. As well as fixing this, I
1879 have increased the buffer size to 1024 (and made a note to document this).
1880
1881 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1882 when Exim is waiting for an SMTP response from a remote server, Exim
1883 restarts its select() call on the socket, thereby resetting its timeout.
1884 This is not a problem when such interrupts are rare. Somebody set up a cron
1885 job to run exiwhat every 2 minutes, which is less than the normal select()
1886 timeout (5 or 10 minutes). This meant that the select() timeout never
1887 kicked in because it was always reset. I have fixed this by comparing the
1888 time when an interrupt arrives with the time at the start of the first call
1889 to select(). If more time than the timeout has elapsed, the interrupt is
1890 treated as a timeout.
1891
1892 5. Some internal re-factoring in preparation for the addition of Sieve
1893 extensions (by MH). In particular, the "personal" test is moved to a
1894 separate function, and given an option for scanning Cc: and Bcc: (which is
1895 not set for Exim filters).
1896
1897 6. When Exim created an email address using the login of the caller as the
1898 local part (e.g. when creating a From: or Sender: header line), it was not
1899 quoting the local part when it contained special characters such as @.
1900
1901 7. Installed new OpenBSD configuration files.
1902
1903 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1904 try to make them clearer.
1905
1906 9. Callout options, other than the timeout value, were being ignored when
1907 verifying sender addresses in header lines. For example, when using
1908
1909 verify = header_sender/callout=no_cache
1910
1911 the cache was (incorrectly) being used.
1912
191310. Added a missing instance of ${EXE} to the exim_install script; this affects
1914 only the Cygwin environment.
1915
191611. When return_path_on_delivery was set as a log selector, if different remote
1917 addresses in the same message used different return paths and parallel
1918 remote delivery occurred, the wrong values would sometimes be logged.
1919 (Whenever a remote delivery process finished, the return path value from
1920 the most recently started remote delivery process was logged.)
1921
192212. RFC 3848 specifies standard names for the "with" phrase in Received: header
1923 lines when AUTH and/or TLS are in use. This is the "received protocol"
1924 field. Exim used to use "asmtp" for authenticated SMTP, without any
1925 indication (in the protocol name) for TLS use. Now it follows the RFC and
1926 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1927 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1928 names appear in log lines as well as in Received: header lines.
1929
193013. Installed MH's patches for Sieve to add the "copy" and "vacation"
1931 extensions, and comparison tests, and to fix some bugs.
1932
193314. Changes to the "personal" filter test:
1934
1935 (1) The test was buggy in that it was just doing the equivalent of
1936 "contains" tests on header lines. For example, if a user's address was
1937 anne@some.where, the "personal" test would incorrectly be true for
1938
1939 To: susanne@some.where
1940
1941 This test is now done by extracting each address from the header in turn,
1942 and checking the entire address. Other tests that are part of "personal"
1943 are now done using regular expressions (for example, to check local parts
1944 of addresses in From: header lines).
1945
1946 (2) The list of non-personal local parts in From: addresses has been
1947 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1948 taken from the Sieve specification recommendations.
1949
1950 (3) If the message contains any header line starting with "List-" it is
1951 treated as non-personal.
1952
1953 (4) The test for "circular" in the Subject: header line has been removed
1954 because it now seems ill-conceived.
1955
195615. Minor typos in src/EDITME comments corrected.
1957
195816. Installed latest exipick from John Jetmore.
1959
196017. If headers_add on a router specified a text string that was too long for
1961 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1962 of string_sprintf() is now avoided.
1963
196418. $message_body_size was not set (it was always zero) when running the DATA
1965 ACL and the local_scan() function.
1966
196719. For the "mail" command in an Exim filter, no default was being set for
1968 the once_repeat time, causing a random time value to be used if "once" was
1969 specified. (If the value happened to be <= 0, no repeat happened.) The
1970 default is now 0s, meaning "never repeat". The "vacation" command was OK
1971 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1972 (I found it when inspecting the code).
1973
197420. There is now an overall timeout for performing a callout verification. It
1975 defaults to 4 times the callout timeout, which applies to individual SMTP
1976 commands during the callout. The overall timeout applies when there is more
1977 than one host that can be tried. The timeout is checked before trying the
1978 next host. This prevents very long delays if there are a large number of
1979 hosts and all are timing out (e.g. when the network connections are timing
1980 out). The value of the overall timeout can be changed by specifying an
1981 additional sub-option for "callout", called "maxwait". For example:
1982
1983 verify = sender/callout=5s,maxwait=20s
1984
198521. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1986 to the end before writing, but this should make it even safer).
1987
198822. Exim was forgetting that it had advertised PIPELINING for the second and
1989 subsequent messages on an SMTP connection. It was also not resetting its
1990 memory on STARTTLS and an internal HELO.
1991
199223. When Exim logs an SMTP synchronization error within a session, it now
1993 records whether PIPELINING has been advertised or not.
1994
199524. Added 3 instances of "(long int)" casts to time_t variables that were being
1996 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1997 rather than long int.
1998
199925. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2000
200126. Added the never_mail option to autoreply.
2002
2003
2004Exim version 4.41
2005-----------------
2006
2007 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2008 crash if the getsockname() call failed; this can happen if a connection is
2009 closed very soon after it is established. The problem was simply in the
2010 order in which certain operations were done, causing Exim to try to write
2011 to the SMTP stream before it had set up the file descriptor. The bug has
2012 been fixed by making things happen in the correct order.
2013
2014
2015Exim version 4.40
2016-----------------
2017
2018 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2019 before the connection was closed, thus losing the rejection response.
2020
2021 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2022 some early Solaris releases, but causes trouble in current releases where
2023 socklen_t is defined.
2024
2025 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2026 always exist.
2027
2028 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2029 configured.
2030
2031 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2032 that releasing the top of it at the end releases what was used for sub-
2033 expansions (unless the block got too big). However, discard this block if
2034 the first thing is a variable or header, so that we can use its block when
2035 it is dynamic (useful for very large $message_headers, for example).
2036
2037 6. Lookups now cache *every* query, not just the most recent. A new, separate
2038 store pool is used for this. It can be recovered when all lookup caches are
2039 flushed. Lookups now release memory at the end of their result strings.
2040 This has involved some general refactoring of the lookup sources.
2041
2042 7. Some code has been added to the store_xxx() functions to reduce the amount
2043 of flapping under certain conditions.
2044
2045 8. log_incoming_interface used to affect only the <= reception log lines. Now
2046 it causes the local interface and port to be added to several more SMTP log
2047 lines, for example "SMTP connection from", and rejection lines.
2048
2049 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2050
205110. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2052
205311. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2054 could be overwritten at the end of the current message (or the start of a
2055 new message if it was set in a HELO ACL). The value is now preserved for
2056 the duration of the SMTP connection.
2057
205812. If a transport had a headers_rewrite setting, and a matching header line
2059 contained an unqualified address, that address was qualified, even if it
2060 did not match any rewriting rules. The underlying bug was that the values
2061 of the flags that permit the existence of unqualified sender and recipient
2062 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2063 non-local messages, and by -bnq for local messages) were not being
2064 preserved with the message after it was received.
2065
206613. When Exim was logging an SMTP synchronization error, it could sometimes log
2067 "next input=" as part of the text comprising the host identity instead of
2068 the correct text. The code was using the same buffer for two different
2069 strings. However, depending on which order the printing function evaluated
2070 its arguments, the bug did not always show up. Under Linux, for example, my
2071 test suite worked just fine.
2072
207314. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2074 doesn't work with some older versions of Perl. It has been changed to "my",
2075 which in any case is probably the better facility to use.
2076
207715. A really picky compiler found some instances of statements for creating
2078 error messages that either had too many or two few arguments for the format
2079 string.
2080
208116. The size of the buffer for calls to the DNS resolver has been increased
2082 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2083 for addresses that have a lot of PTR records. This alleviates a problem; it
2084 does not fully solve it.
2085
208617. A dnsdb lookup for PTR records that receives more data than will fit in the
2087 buffer now truncates the list and logs the incident, which is the same
2088 action as happens when Exim is looking up a host name and its aliases.
2089 Previously in this situation something unpredictable would happen;
2090 sometimes it was "internal error: store_reset failed".
2091
209218. If a server dropped the connection unexpectedly when an Exim client was
2093 using GnuTLS and trying to read a response, the client delivery process
2094 crashed while trying to generate an error log message.
2095
209619. If a "warn" verb in an ACL added multiple headers to a message in a single
2097 string, for example:
2098
2099 warn message = H1: something\nH2: something
2100
2101 the text was added as a single header line from Exim's point of view
2102 though it ended up OK in the delivered message. However, searching for the
2103 second and subsequent header lines using $h_h2: did not work. This has been
2104 fixed. Similarly, if a system filter added multiple headers in this way,
2105 the routers could not see them.
2106
210720. Expanded the error message when iplsearch is called with an invalid key to
2108 suggest using net-iplsearch in a host list.
2109
211021. When running tests using -bh, any delays imposed by "delay" modifiers in
2111 ACLs are no longer actually imposed (and a message to that effect is
2112 output).
2113
211422. If a "gecos" field in a passwd entry contained escaped characters, in
2115 particular, if it contained a \" sequence, Exim got it wrong when building
2116 a From: or a Sender: header from that name. A second bug also caused
2117 incorrect handling when an unquoted " was present following a character
2118 that needed quoting.
2119
212023. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2121 was not being matched caselessly.
2122
212324. Arranged for all hyphens in the exim.8 source to be escaped with
2124 backslashes.
2125
212625. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2127 properly. Recipient callout cache records were still being keyed to include
2128 the sender, even when use_sender was set false. This led to far more
2129 callouts that were necessary. The sender is no longer included in the key
2130 when use_sender is false.
2131
213226. Added "control = submission" modifier to ACLs.
2133
213427. Added the ${base62d: operator to decode base 62 numbers.
2135
213628. dnsdb lookups can now access SRV records.
2137
213829. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2139 the configuration file.
2140
214130. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2142 (-v) mode. This makes the output for a verbose queue run more intelligible.
2143
214431. Added a use_postmaster feature to recipient callouts.
2145
214632. Added the $body_zerocount variable, containing the number of binary zero
2147 bytes in the message body.
2148
214933. The time of last modification of the "new" subdirectory is now used as the
2150 "mailbox time last read" when there is a quota error for a maildir
2151 delivery.
2152
215334. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2154
215535. Added +ignore_unknown as a special item in host lists.
2156
215736. Code for decoding IPv6 addresses in host lists is now included, even if
2158 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2159 address was recognized as an IP address, but was then not correctly decoded
2160 into binary, causing unexpected and incorrect effects when compared with
2161 another IP address.
2162
2163
2164Exim version 4.34
2165-----------------
2166
2167 1. Very minor rewording of debugging text in manualroute to say "list of
2168 hosts" instead of "hostlist".
2169
2170 2. If verify=header_syntax was set, and a header line with an unqualified
2171 address (no domain) and a large number of spaces between the end of the
2172 name and the colon was received, the reception process suffered a buffer
2173 overflow, and (when I tested it) crashed. This was caused by some obsolete
2174 code that should have been removed. The fix is to remove it!
2175
2176 3. When running in the test harness, delay a bit after writing a bounce
2177 message to get a bit more predictability in the log output.
2178
2179 4. Added a call to search_tidyup() just before forking a reception process. In
2180 theory, someone could use a lookup in the expansion of smtp_accept_max_
2181 per_host which, without the tidyup, could leave open a database connection.
2182
2183 5. Added the variables $recipient_data and $sender_data which get set from a
2184 lookup success in an ACL "recipients" or "senders" condition, or a router
2185 "senders" option, similar to $domain_data and $local_part_data.
2186
2187 6. Moved the writing of debug_print from before to after the "senders" test
2188 for routers.
2189
2190 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2191 problems for message scanning, either using a data ACL, or using
2192 local_scan() because the Received: header was not generated till after they
2193 were called (in order to set the time as the time of reception completion).
2194 I have revised the way this works. The header is now generated after the
2195 body is received, but before the ACL or local_scan() are called. After they
2196 are run, the timestamp in the header is updated.
2197
2198
2199Exim version 4.33
2200-----------------
2201
2202 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2203 before starting a queue runner without re-exec. This happened only when
2204 deliver_drop_privilege was set or when the Exim user was set to root. The
2205 effect of the bug was that timeouts during subsequent deliveries caused
2206 crashes instead of being properly handled. The handler is now left at its
2207 default (and expected) setting.
2208
2209 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2210 message, again when deliver_drop_privilege is set or Exim is run as root.
2211 The bug described in (1) was not present in this case, but the tidying up
2212 of the other signals was missing. I have made the two cases consistent.
2213
2214 3. The ignore_target_hosts setting on a manualroute router was being ignored
2215 for hosts that were looked up using the /MX notation.
2216
2217 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2218 in domain lists.
2219
2220 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2221 operated on the sender address. After changing the $sender_address to <>
2222 for the sender address verify, Exim was re-instated it as the original
2223 (before rewriting) address, but remembering that it had rewritten it, so it
2224 wasn't rewriting it again. This bug also had the effect of breaking the
2225 sender address verification caching when the sender address was rewritten.
2226
2227 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2228 This has been changed so that if the ip literal address matches
2229 ignore_target_hosts, the router declines.
2230
2231 7. Added expansion conditions match_domain, match_address, and match_local_
2232 part (NOT match_host).
2233
2234 8. The placeholder for the Received: header didn't have a length field set.
2235
2236 9. Added code to Exim itself and to exim_lock to test for a specific race
2237 condition that could lead to file corruption when using MBX delivery. The
2238 issue is with the lockfile that is created in /tmp. If this file is removed
2239 after a process has opened it but before that process has acquired a lock,
2240 there is the potential for a second process to recreate the file and also
2241 acquire a lock. This could lead to two Exim processes writing to the file
2242 at the same time. The added code performs the same test as UW imapd; it
2243 checks after acquiring the lock that its file descriptor still refers to
2244 the same named file.
2245
224610. The buffer for building added header lines was of fixed size, 8192 bytes.
2247 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2248 when Exim is built.
2249
225011. Added the smtp_active_hostname option. If used, this will typically be made
2251 to depend on the incoming interface address. Because $interface_address is
2252 not set up until the daemon has forked a reception process, error responses
2253 that can happen earlier (such as "too many connections") no longer contain
2254 a host name.
2255
225612. If an expansion in a condition on a "warn" statement fails because a lookup
2257 defers, the "warn" statement is abandoned, and the next ACL statement is
2258 processed. Previously this caused the whole ACL to be aborted.
2259
226013. Added the iplsearch lookup type.
2261
226214. Added ident_timeout as a log selector.
2263
226415. Added tls_certificate_verified as a log selector.
2265
226616. Added a global option tls_require_ciphers (compare the smtp transport
2267 option of the same name). This controls incoming TLS connections.
2268
226917. I finally figured out how to make tls_require_ciphers do a similar thing
2270 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2271 before starting the TLS session.
2272
227318. Tabs are now shown as \t in -bP output.
2274
227519. If the log selector return_path_on_delivery was set, Exim crashed when
2276 bouncing a message because it had too many Received: header lines.
2277
227820. If two routers both had headers_remove settings, and the first one included
2279 a superfluous trailing colon, the final name in the first list and the
2280 first name in the second list were incorrectly joined into one item (with a
2281 colon in the middle).
2282
2283
2284Exim version 4.32
2285-----------------
2286
2287 1. Added -C and -D options to the exinext utility, mainly to make it easier
2288 to include in the automated testing, but these could be helpful when
2289 multiple configurations are in use.
2290
2291 2. The exinext utility was not formatting the output nicely when there was
2292 an alternate port involved in the retry record key, nor when there was a
2293 message id as well (for retries that were specific to a specific message
2294 and a specific host). It was also confused by IPv6 addresses, because of
2295 the additional colons they contain. I have fixed the IPv4 problem, and
2296 patched it up to do a reasonable job for IPv6.
2297
2298 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2299 delivery, the log line now contains "pipelined" if PIPELINING was used.
2300
2301 4. An SMTP transport process used to panic and die if the bind() call to set
2302 an explicit outgoing interface failed. This has been changed; it is now
2303 treated in the same way as a connect() failure.
2304
2305 5. A reference to $sender_host_name in the part of a conditional expansion
2306 that was being skipped was still causing a DNS lookup. This no longer
2307 occurs.
2308
2309 6. The def: expansion condition was not recognizing references to header lines
2310 that used bh_ and bheader_.
2311
2312 7. Added the _cache feature to named lists.
2313
2314 8. The code for checking quota_filecount in the appendfile transport was
2315 allowing one more file than it should have been.
2316
2317 9. For compatibility with Sendmail, the command line option
2318
2319 -prval:sval
2320
2321 is equivalent to
2322
2323 -oMr rval -oMs sval
2324
2325 and sets the incoming protocol and host name (for trusted callers). The
2326 host name and its colon can be omitted when only the protocol is to be set.
2327 Note the Exim already has two private options, -pd and -ps, that refer to
2328 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2329 "s", but I don't think that's a major issue.
2330
233110. A number of refactoring changes to the code, none of which should affect
2332 Exim's behaviour:
2333
2334 (a) The number of logging options was getting close to filling up the
2335 32-bit word that was used as a bit map. I have split them into two classes:
2336 those that are passed in the argument to log_write(), and those that are
2337 only ever tested independently outside of that function. These are now in
2338 separate 32-bit words, so there is plenty of room for expansion again.
2339 There is no change in the user interface or the logging behaviour.
2340
2341 (b) When building, for example, log lines, the code previously used a
2342 macro that called string_cat() twice, in order to add two strings. This is
2343 not really sufficiently general. Furthermore, there was one instance where
2344 it was actually wrong because one of the argument was used twice, and in
2345 one call a function was used. (As it happened, calling the function twice
2346 did not affect the overall behaviour.) The macro has been replaced by a
2347 function that can join an arbitrary number of extra strings onto a growing
2348 string.
2349
2350 (c) The code for expansion conditions now uses a table and a binary chop
2351 instead of a serial search (which was left over from when there were very
2352 few conditions). Also, it now recognizes conditions like "pam" even when
2353 the relevant support is not compiled in: a suitably worded error message is
2354 given if an attempt is made to use such a condition.
2355
235611. Added ${time_interval:xxxxx}.
2357
235812. A bug was causing one of the ddress fields not to be passed back correctly
2359 from remote delivery subprocesses. The field in question was not being
2360 subsequently used, so this caused to problems in practice.
2361
236213. Added new log selectors queue_time and deliver_time.
2363
236414. Might have fixed a bug in maildirsizefile handling that threw up
2365 "unexpected character" debug warnings, and recalculated the data
2366 unnecessarily. In any case, I expanded the warning message to give more
2367 information.
2368
236915. Added the message "Restricted characters in address" to the statements in
2370 the default ACL that block characters like @ and % in local parts.
2371
237216. Change 71 for release 4.31 proved to be much less benign that I imagined.
2373 Three changes have been made:
2374
2375 (a) There was a serious bug; a negative response to MAIL caused the whole
2376 recipient domain to be cached as invalid, thereby blocking all messages
2377 to all local parts at the same domain, from all senders. This bug has
2378 been fixed. The domain is no longer cached after a negative response to
2379 MAIL if the sender used is not empty.
2380
2381 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2382 been restored.
2383
2384 (c) A new callout option, "use_sender" has been added for people who want
2385 the modified behaviour.
2386
2387
2388Exim version 4.31
2389-----------------
2390
2391 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2392 Larry Rosenman.
2393
2394 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2395 indeed breaks things for older releases.
2396
2397 3. Added additional logging to the case where there is a problem reading data
2398 from a filter that is running in a subprocess using a pipe, in order to
2399 try to track down a specific problem.
2400
2401 4. Testing facility fudge: when running in the test harness and attempting
2402 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2403 getting "No route to host". Convert this to a timeout.
2404
2405 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2406 warning.
2407
2408 6. Some OS don't have socklen_t but use size_t instead. This affects the
2409 fifth argument of getsockopt() amongst other things. This is now
2410 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2411 can be set for individual OS. I have set it for SunOS5, OSF1, and
2412 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2413 some earlier ones do not.
2414
2415 7. Change 4.30/15 was not doing the test caselessly.
2416
2417 8. The standard form for an IPv6 address literal was being rejected by address
2418 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2419 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2420 this, as well as the form without the "IPv6" on the front (but only when
2421 address literals are enabled, of course).
2422
2423 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2424
242510. Exim crashed if a message with an empty sender address specified by -f
2426 encountered a router with an errors_to setting. This could be provoked only
2427 by a command such as
2428
2429 exim -f "" ...
2430
2431 where an empty string was supplied; "<>" did not hit this bug.
2432
243311. Installed PCRE release 4.5.
2434
243512. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2436 remained set. It is now erased.
2437
243813. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2439 times from message ids (which are base 36 rather than the normal 62).
2440
244114. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2442 were being counted as actual protocol errors, and logged if the log
2443 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2444 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2445 and DATA following a set of rejected RCPTs do not count as protocol errors.
2446 In other words, Exim assumes they were pipelined, though this may not
2447 actually be the case. Of course, in all cases the client gets an
2448 appropriate error code.
2449
245015. If a lookup fails in an ACL condition, a message about the failure may
2451 be available; it is used if testing the ACL cannot continue, because most
2452 such messages specify what the cause of the deferral is. However, some
2453 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2454 that caused an old message to be retained and used if a later statement
2455 caused a defer, replacing the real cause of the deferral.
2456
245716. If an IP address had so many PTR records that the DNS lookup buffer
2458 was not large enough to hold them, Exim could crash while trying to process
2459 the truncated data. It now detects and logs this case.
2460
246117. Further to 4.21/58, another change has been made: if (and only if) the
2462 first line of a message (the first header line) ends with CRLF, a bare LF
2463 in a subsequent header line has a space inserted after it, so as not to
2464 terminate the header.
2465
246618. Refactoring: tidied an ugly bit of code in appendfile that copied data
2467 unnecessarily, used atoi() instead of strtol(), and didn't check the
2468 termination when getting file sizes from file names by regex.
2469
247019. Completely re-implemented the support for maildirsize files, in the light
2471 of a number of problems with the previous contributed implementation
2472 (4.30/29). In particular:
2473
2474 . If the quota is zero, the maildirsize file is maintained, but no quota is
2475 imposed.
2476
2477 . If the maildir directory does not exist, it is created before any attempt
2478 to write a maildirsize file.
2479
2480 . The quota value in the file is just a cache; if the quota is changed in
2481 the transport, the new value overrides.
2482
2483 . A regular expression is available for excluding directories from the
2484 count.
2485
248620. The autoreply transport checks the characters in options that define the
2487 message's headers; it allows continued headers, but it was checking with
2488 isspace() after an embedded newline instead of explicitly looking for a
2489 space or a tab.
2490
249121. If all the "regular" hosts to which an address was routed had passed their
2492 expiry times, and had not reached their retry times, the address was
2493 bounced, even if fallback hosts were defined. Now Exim should go on to try
2494 the fallback hosts.
2495
249622. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2497 equivalent code in the SMTP transport. Some hosts send humungous responses
2498 to HELO/EHLO, more than 1024 it seems.
2499
250023. Refactoring: code in filter.c used (void *) for "any old type" but this
2501 gives compiler warnings in some environments. I've now done it "properly",
2502 using a union.
2503
250424. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2505 (because of problems with the built-in one) was declared to return uschar *
2506 instead of char *, causing compiler failure.
2507
250825. Fixed a file descriptor leak when processing alias/forward files.
2509
251026. Fixed a minor format string issue in dbfn.c.
2511
251227. Typo in exim.c: ("dmbnz" for "dbmnz").
2513
251428. If a filter file refered to $h_xxx or $message_headers, and the headers
2515 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2516 become corrupted.
2517
251829. When a sender address is verified, it is cached, to save repeating the test
2519 when there is more than one recipient in a message. However, when the
2520 verification involves a callout, it is possible for different callout
2521 options to be set for different recipients. It is too complicated to keep
2522 track of this in the cache, so now Exim always runs a verification when a
2523 callout is required, relying on the callout cache for the optimization.
2524 The overhead is duplication of the address routing, but this should not be
2525 too great.
2526
252730. Fixed a bug in callout caching. If a RCPT command caused the sender address
2528 to be verified with callout=postmaster, and the main callout worked but the
2529 postmaster check failed, the verification correctly failed. However, if a
2530 subsequent RCPT command asked for sender verification *without* the
2531 postmaster check, incorrect caching caused this verification also to fail,
2532 incorrectly.
2533
253431. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2535 it was not caching the DNS options (qualify_single, search_parents) that
2536 were used when the lookup failed. A subsequent lookup with different
2537 options therefore always gave the same answer, though there were cases
2538 where it should not have. (Example: a "domains = !$mx_any" option on a
2539 dnslookup router: the "domains" option is always processed without any
2540 widening, but the router might have qualify_single set.) Now Exim uses the
2541 cached value only when the same options are set.
2542
254332. Added John Jetmore's "exipick" utility to the distribution.
2544
254533. GnuTLS: When an attempt to start a TLS session fails for any reason other
2546 than a timeout (e.g. a certificate is required, and is not provided), an
2547 Exim server now closes the connection immediately. Previously it waited for
2548 the client to close - but if the client is SSL, it seems that they each
2549 wait for each other, leading to a delay before one of them times out.
2550
255134: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2552 maintained 0.8.x compatibility because I don't think many are using it, and
2553 it is clearly obsolete.
2554
255535. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2556 transport.
2557
255836. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2559 client certificate was expired. A simple patch fixes this, though I don't
2560 understand the full logic of why the verify callback is called multiple
2561 times.
2562
256337. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2564 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2565 which causes problems with some clients (such as the Certicom SSL Plus
2566 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2567 disables the coutermeasure allowing Eudora to connect."
2568
256938. Exim was not checking that a write() to a log file succeeded. This could
2570 lead to Bad Things if a log got too big, in particular if it hit a file
2571 size limit. Exim now panics and dies if it cannot write to a log file, just
2572 as it does if it cannot open a log file.
2573
257439. Modified OS/Makefile-Linux so that it now contains
2575
2576 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2577
2578 The two -D definitions ensure that Exim is compiled with large file
2579 support, which makes it possible to handle log files that are bigger than
2580 2^31.
2581
258240. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2583 instance) a domain was checked against a named list that involved a lookup,
2584 causing $domain_data to be set, then another domain was checked against the
2585 same list, then the first domain was re-checked, the value of $domain_data
2586 after the final check could be wrong. In particular, if the second check
2587 failed, it could be set empty. This bug probably also applied to
2588 $localpart_data.
2589
259041. The strip_trailing_dot option was not being applied to the address given
2591 with the -f command-line option.
2592
259342. The code for reading a message's header from the spool was incrementing
2594 $received_count, but never initializing it. This meant that the value was
2595 incorrect (doubled) while delivering a message in the same process in which
2596 it was received. In the most common configuration of Exim, this never
2597 happens - a fresh exec is done - but it can happen when
2598 deliver_drop_privilege is set.
2599
260043. When Exim logs an SMTP synchronization error - client data sent too soon -
2601 it now includes up to 150 characters of the unexpected data in the log
2602 line.
2603
260444. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2605 and building data strings. The size of both of these buffers was 10 000
2606 bytes - far larger than anybody would *ever* want, thought I. Needless to
2607 say, somebody hit the limit. I have increased the maximum line length to
2608 20 000 and the maximum data length of concatenated lines to 100 000. I have
2609 also fixed two bugs, because there was no checking on these buffers. Tsk,
2610 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2611 buffer is too small.
2612
261345. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2614 lsearch lookups. Now it does.
2615
261646. When parsing a route_list item in a manualroute router, a fixed-length
2617 buffer was used for the list of hosts. I made this 1024 bytes long,
2618 thinking that nobody would ever have a list of hosts that long. Wrong.
2619 Somebody had a whole pile of complicated expansion conditions, and the
2620 string was silently truncated, leading to an expansion error. It turns out
2621 that it is easier to change to an unlimited length (owing to other changes
2622 that have happened since this code was originally written) than to build
2623 structure for giving a limitation error. The length of the item that
2624 expands into the list of hosts is now unlimited.
2625
262647. The lsearch lookup could not handle data where the length of text line was
2627 more than 4095 characters. Such lines were truncated, leading to shortened
2628 data being returned. It should now handle lines of any length.
2629
263048. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2631 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2632 ACL").
2633
263449. Cosmetic tidy to scripts like exicyclog that are generated by globally
2635 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2636 no longer happens in comment lines. A list of replacements is now placed
2637 at the head of all of the source files, except those whose only change is
2638 to replace PERL_COMMAND in the very first #! line.
2639
264050. Replaced the slow insertion sort in queue.c, for sorting the list of
2641 messages on the queue, with a bottom-up merge sort, using code contributed
2642 by Michael Haardt. This should make operations like -bp somewhat faster on
2643 large queues. It won't affect queue runners, except when queue_run_in_order
2644 is set.
2645
264651. Installed eximstats 1.31 in the distribution.
2647
264852. Added support for SRV lookups to the dnslookup router.
2649
265053. If an ACL referred to $message_body or $message_body_end, the value was not
2651 reset for any messages that followed in the same SMTP session.
2652
265354. The store-handling optimization for building very long strings was not
2654 differentiating between the different store pools. I don't think this
2655 actually made any difference in practice, but I've tidied it.
2656
265755. While running the routers to verify a sender address, $sender_address
2658 was still set to the sender address. This is wrong, because when routing to
2659 send a bounce to the sender, it would be empty. Therefore, I have changed
2660 it so that, while verifying a sender address, $sender_address is set to <>.
2661 (There is no change to what happens when verifying a recipient address.)
2662
266356. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2664 target A or AAAA records (if not already returned) without resetting the
2665 qualify_single or search_parents options of the DNS resolver. These are
2666 inappropriate in this case because the targets of MX and SRV records must
2667 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2668 target that, when qualified, matched something in the local domain. These
2669 two options are now turned off when doing these lookups.
2670
267157. It seems that at least some releases of Reiserfs (which does not have the
2672 concept of a fixed number of inodes) returns zero and not -1 for the
2673 number of available inodes. This interacted badly with check_spool_inodes,
2674 which assumed that -1 was the "no such thing" setting. What I have done is
2675 to check that the total number of inodes is greater than zero before doing
2676 the test of how many are available.
2677
267858. When a "warn" ACL statement has a log_message modifier, the message is
2679 remembered, and not repeated. This is to avoid a lot of repetition when a
2680 message has many recipients that cause the same warning to be written.
2681 Howewer, Exim was preserving the list of already written lines for an
2682 entire SMTP session, which doesn't seem right. The memory is now reset if a
2683 new message is started.
2684
268559. The "rewrite" debugging flag was not showing the result of rewriting in the
2686 debugging output unless log_rewrite was also set.
2687
268860. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2689 of (int)(handle) when we know that handle contains (void *)(-1).
2690
269161. The Exim daemon panic-logs an error return when it closes the incoming
2692 connection. However "connection reset by peer" seems to be common, and
2693 isn't really an error worthy of noting specially, so that particular error
2694 is no long logged.
2695
269662. When Exim is trying to find all the local interfaces, it used to panic and
2697 die if the ioctl to get the interface flags failed. However, it seems that
2698 on at least one OS (Solaris 9) it is possible to have an interface that is
2699 included in the list of interfaces, but for which you get a failure error
2700 for this call. This happens when the interface is not "plumbed" into a
2701 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2702 failure of the "get flags" call assumes that the interface is down.
2703
270463. Added a ${eval10: operator, which assumes all numbers are decimal. This
2705 makes life easier for people who are doing arithmetic on fields extracted
2706 from dates, where you often get leading zeros that should not be
2707 interpreted as octal.
2708
270964. Added qualify_domain to the redirect router, to override the global
2710 setting.
2711
271265. If a pathologically long header line contained very many addresses (the
2713 report of this problem mentioned 10 000) and each of them was rewritten,
2714 Exim could use up a very large amount of memory. (It kept on making new
2715 copies of the header line as it rewrote, and never released the old ones.)
2716 At the expense of a bit more processing, the header rewriting function has
2717 been changed so that it no longer eats memory in this way.
2718
271966. The generation of the Received: header has been moved from the time that a
2720 message starts to be received, to the time that it finishes. The timestamp
2721 in the Received: header should now be very close to that of the <= log
2722 line. There are two side-effects of this change:
2723
2724 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2725 logged header lines no longer include the local Received: line, because
2726 it has not yet been created. The same applies to a copy of the message
2727 that is returned to a non-SMTP sender when a message is rejected.
2728
2729 (b) When a filter file is tested using -bf, no additional Received: header
2730 is added to the test message. After some thought, I decided that this
2731 is a bug fix.
2732
2733 This change does not affect the value of $received_for. It is still set
2734 after address rewriting, but before local_scan() is called.
2735
273667. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2737
273868. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2739 gave an unhelpful panic error message, and a defer error. I have managed to
2740 change this behaviour so that it now rejects any supplied certificate,
2741 which seems right, as the list of acceptable certificates is empty.
2742
274369. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2744 gave an unhelpful defer error. I have not managed to make this reject any
2745 supplied certificates, but the error message it gives is "no certificate
2746 supplied", which is not helpful.
2747
274870. exigrep's output now also includes lines that are not associated with any
2749 message, but which match the given pattern. Implemented by a patch from
2750 Martin Sluka, which also tidied up the Perl a bit.
2751
275271. Recipient callout verification, like sender verification, was using <> in
2753 the MAIL FROM command. This isn't really the right thing, since the actual
2754 sender may affect whether the remote host accepts the recipient or not. I
2755 have changed it to use the actual sender in the callout; this means that
2756 the cache record is now keyed on a recipient/sender pair, not just the
2757 recipient address. There doesn't seem to be a real danger of callout loops,
2758 since a callout by the remote host to check the sender would use <>.
2759 [SEE ABOVE: changed after hitting problems.]
2760
276172. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2762 temporary errors. However, in the case of such a code being given after
2763 the end of a data transmission (i.e. after ".") Exim was failing to write
2764 a retry record for the message. (Yes, there was some broken host that was
2765 actually sending 8xx at this point.)
2766
276773. An unknown lookup type in a host list could cause Exim to panic-die when
2768 the list was checked. (An example that provoked this was putting <; in the
2769 middle of a list instead of at the start.) If this happened during a DATA
2770 ACL check, a -D file could be left lying around. This kind of configuration
2771 error no longer causes Exim to die; instead it causes a defer errror. The
2772 incident is still logged to the main and panic logs.
2773
277474. Buglet left over from Exim 3 conversion. The message "too many messages
2775 in one connection" was written to the rejectlog but not the mainlog, except
2776 when address rewriting (yes!) was being logged.
2777
277875. Added write_rejectlog option.
2779
278076. When a system filter was run not as root (that is, when system_filter_user
2781 was set), the values of the $n variables were not being returned to the
2782 main process; thus, they were not subsequently available in the $sn
2783 variables.
2784
278577. Added +return_path_on_delivery log selector.
2786
278778. A connection timeout was being treated differently from recipients deferred
2788 when testing hosts_max_try with a message that was older than the host's
2789 retry timeout. (The host should not be counted, thus allowing all hosts to
2790 be tried at least once before bouncing.) This may have been the cause of an
2791 occasionally reported bug whereby a message would remain on the queue
2792 longer than the retry timeout, but would be bounced if a delivery was
2793 forced. I say "may" because I never totally pinned down the problem;
2794 setting up timeout/retry tests is difficult. See also the next item.
2795
279679. The ultimate address timeout was not being applied to errors that involved
2797 a combination of host plus message (for example, a timeout on a MAIL
2798 command). When an address resolved to a number of possible hosts, and they
2799 were not all tried for each delivery (e.g. because of hosts_max_try), a
2800 message could remain on the queue longer than the retry timeout.
2801
280280. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2803 Haardt.
2804
280581. Fixed an obscure SMTP outgoing bug which required at least the following
2806 conditions: (a) there was another message waiting for the same server;
2807 (b) the server returned 5xx to all RCPT commands in the first message so
2808 that the message was not completed; (c) the server dropped the connection
2809 or gave a negative response to the RSET that Exim sends to abort the
2810 transaction. The observed case was a dropped connection after DATA that had
2811 been sent in pipelining mode. That is, the server had advertised PIPELINING
2812 but was not implementing it correctly. The effect of the bug was incorrect
2813 behaviour, such as trying another host, and this could lead to a crash.
2814
2815
2816Exim version 4.30
2817-----------------
2818
2819 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2820 and daemon.c were passed as pointers to ints; they should have been
2821 pointers to socklen_t variables (which are typically unsigned ints).
2822
2823 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2824 fixed.
2825
2826 3. Fixed a really odd bug that affected only the testing scheme; patching a
2827 certain fixed string in the binary changed the value of another string that
2828 happened to be identical to the end of the original first string.
2829
2830 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2831 name", it returns that address as the IP address. On some operating
2832 systems (e.g. Solaris), it also passes back the IP address string as the
2833 "host name". However, on others (e.g. Linux), it passes back an empty
2834 string. Exim wasn't checking for this, and was changing the host name to an
2835 empty string, assuming it had been canonicized.
2836
2837 5. Although rare, it is permitted to have more than one PTR record for a given
2838 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2839 all the names associated with an address, because they do in Solaris.
2840 However, it seems that they do not in Linux for data that comes from the
2841 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2842 I found this out when I moved to a new Linux workstation and tried to run
2843 the Exim test suite.
2844
2845 To get round this problem I have changed the code so that it now does its
2846 own call to the DNS to look up PTR records when searching for a host name.
2847 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2848 addresses that are only in /etc/hosts are still found.
2849
2850 This behaviour is, however, controlled by an option called host_lookup_
2851 order, which defaults to "bydns:byaddr". If people want to use the other
2852 order, or indeed, just use one or the other means of lookup, they can
2853 specify it in this variable.
2854
2855 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2856 some operating systems, this comes back from gethostbyaddr() as an empty
2857 string, and this is what Exim used to test for. However, it seems that in
2858 other systems, "." is yielded. Exim now tests for this case too.
2859
2860 7. The values of check_spool_space and check_log_space are now held internally
2861 as a number of kilobytes instead of an absolute number of bytes. If a
2862 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2863 kilobyte. This means that much larger values can be stored.
2864
2865 8. Exim monitor: an attempt to get the action menu when not actually pointing
2866 at a message produces an empty menu entitled "No message selected". This
2867 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2868 no entries in it ("Shell widget menu has zero width and/or height"). So I
2869 have added a single, blank menu entry in this case.
2870
2871 9. Added ${quote_local_part.
2872
287310. MIME decoding is now applied to the contents of Subject: header lines when
2874 they are logged.
2875
287611. Now that a reference to $sender_host_address automatically causes a reverse
2877 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2878 host lookup before query-style lookups in lists that might use this
2879 variable. This has therefore been abolished, and the "net-" prefix is no
2880 longer necessary for query-style lookups.
2881
288212. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2883 have been a typo for LFLAGS, so it has been changed.
2884
288513. The install script calls Exim with "-C /dev/null" in order to find the
2886 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2887 to be output. Howeve, since Exim outputs its version number before the
2888 error, it didn't break the script. It just looked ugly. I fixed this by
2889 always allowing "-C /dev/null" if the caller is root.
2890
289114. Ignore overlarge ACL variable number when reading spool file - insurance
2892 against a later release with more variables having written the file.
2893
289415. The standard form for an IPv6 address literal was being rejected by EHLO.
2895 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2896 this, as well as the form without the "IPv6" on the front.
2897
289816. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2899 OS/Makefile-Darwin file.
2900
290117. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2902 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2903
290418. After change 4.21/52, "%ld" was used to format the contents of the $inode
2905 variable. However, some OS use ints for inodes. I've added cast to long int
2906 to get rid of the compiler warning.
2907
290819. I had forgotten to lock out "/../" in configuration file names when
2909 ALT_CONFIG_PREFIX was set.
2910
291120. Routers used for verification do not need to specify transports. However,
2912 if such a router generated a host list, and callout was configured, Exim
2913 crashed, because it could not find a port number from the (non-existent)
2914 transport. It now assumes port 25 in this circumstance.
2915
291621. Added the -t option to exigrep.
2917
291822. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2919 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2920 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2921 them is set, LOOKUP_LSEARCH is forced.
2922
292323. "exim -bV" now outputs a list of lookups that are included in the binary.
2924
292524. Added sender and host information to the "rejected by local_scan()" log
2926 line; previously there was no indication of these.
2927
292825. Added .include_if_exists.
2929
293026. Change 3.952/11 added an explicit directory sync on top of a file sync for
2931 Linux. It turns out that not all file systems support this. Apparently some
2932 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2933 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2934 not supported on the file descriptor, is now ignored when Exim is trying to
2935 sync a directory. This applies only to Linux.
2936
293727. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2938
293928. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2940 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2941 this.
2942
294329. Added support for maildirsize files from supplied patch (modified a bit).
2944
294530. The use of :fail: followed by an empty string could lead Exim to respond to
2946 sender verification failures with (e.g.):
2947
2948 550 Verification failed for <xxx>
2949 550 Sender verify failed
2950
2951 where the first response line was missing the '-' that indicates it is not
2952 the final line of the response.
2953
295431. The loop for finding the name of the user that called Exim had a hardwired
2955 limit of 10; it now uses the value of finduser_retries, which is used for
2956 all other user lookups.
2957
295832. Added $received_count variable, available in data and not_smtp ACLs, and at
2959 delivery time.
2960
296133. Exim was neglecting to zero errno before one call of strtol() when
2962 expanding a string and expecting an integer value. On some systems this
2963 resulted in spurious "integer overflow" errors. Also, it was casting the
2964 result into an int without checking.
2965
296634. Testing for a connection timeout using "timeout_connect" in the retry rules
2967 did not work. The code looks as if it has *never* worked, though it appears
2968 to have been documented since at least releast 1.62. I have made it work.
2969
297035. The "timeout_DNS" error in retry rules, also documented since at least
2971 1.62, also never worked. As it isn't clear exactly what this means, and
2972 clearly it isn't a major issue, I have abolished the feature by treating it
2973 as "timeout", and writing a warning to the main and panic logs.
2974
297536. The display of retry rules for -brt wasn't always showing the error code
2976 correctly.
2977
297837. Added new error conditions to retry rules: timeout_A, timeout_MX,
2979 timeout_connect_A, timeout_connect_MX.
2980
298138. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2982 to the empty sender.
2983
298439. The daemon was not analysing the content of -oX till after it had closed
2985 stderr and disconnected from the controlling terminal. This meant that any
2986 syntax errors were only noted on the panic log, and the return code from
2987 the command was 0. By re-arranging the code a little, I've made the
2988 decoding happen first, so such errors now appear on stderr, and the return
2989 code is 1. However, the actual setting up of the sockets still happens in
2990 the disconnected process, so errors there are still only recorded on the
2991 panic log.
2992
299340. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2994 connections (as happens on some IP stacks) was logged at start up time as
2995 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2996 it from "IPv6 and IPv4", which means that two separate sockets are being
2997 used.
2998
299941. The debug output for gethostbyname2() or getipnodebyname() failures now
3000 says whether AF_INET or AF_INET6 was passed as an argument.
3001
300242. Exiwhat output was messed up when time zones were included in log
3003 timestamps.
3004
300543. Exiwhat now gives more information about the daemon's listening ports,
3006 and whether -tls-on-connect was used.
3007
300844. The "port" option of the smtp transport is now expanded.
3009
301045. A "message" modifier in a "warn" statement in a non-message ACL was being
3011 silently ignored. Now an error message is written to the main and panic
3012 logs.
3013
301446. There's a new ACL modifier called "logwrite" which writes to a log file
3015 as soon as it is encountered.
3016
301747. Added $local_user_uid and $local_user_gid at routing time.
3018
301948. Exim crashed when trying to verify a sender address that was being
3020 rewritten to "<>".
3021
302249. Exim was recognizing only a space character after ".include". It now also
3023 recognizes a tab character.
3024
302550. Fixed several bugs in the Perl script that creates the exim.8 man page by
3026 extracting the relevant information from the specification. The man page no
3027 longer contains scrambled data for the -d option, and I've added a section
3028 at the front about calling Exim under different names.
3029
303051. Added "extra_headers" argument to the "mail" command in filter files.
3031
303252. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3033 crash.
3034
303553. Installed eximstats 1.29.
3036
303754. Added transport_filter_timeout as a generic transport option.
3038
303955. Exim no longer adds an empty Bcc: header to messages that have no To: or
3040 Cc: header lines. This was required by RFC 822, but it not required by RFC
3041 2822.
3042
304356. Exim used to add From:, Date:, and Message-Id: header lines to any
3044 incoming messages that did not have them. Now it does so only if the
3045 message originates locally, that is, if there is no associated remote host
3046 address. When Resent- header lines are present, this applies to the Resent-
3047 lines rather than the non-Resent- lines.
3048
304957. Drop incoming SMTP connection after too many syntax or protocol errors. The
3050 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3051
305258. Messages for configuration errors now include the name of the main
3053 configuration file - useful now that there may be more than one file in a
3054 list (.included file names were always shown).
3055
305659. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3057 for those rare installations that do not start the daemon as root or run it
3058 setuid root. I've cut out the call to initgroups() if the daemon is not
3059 root at that time.
3060
306160. The Exim user and group can now be bound into the binary as text strings
3062 that are looked up at the start of Exim's processing.
3063
306461. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3065
306662. Added $mailstore_basename variable.
3067
306863. Installed patch to sieve.c from Michael Haardt.
3069
307064. When Exim failed to open the panic log after failing to open the main log,
3071 the original message it was trying to log was written to stderr and debug
3072 output, but if they were not available (the usual case in production), it
3073 was lost. Now it is written to syslog before the two lines that record the
3074 failures to open the logs.
3075
307665. Users' Exim filters run in subprocesses under the user's uid. It is
3077 possible for a "deliver" command or an alias in a "personal" command to
3078 provoke an address rewrite. If logging of address rewriting is configured,
3079 this fails because the process is not running as root or exim. There may be
3080 a better way of dealing with this, but for the moment (because 4.30 needs
3081 to be released), I have disabled address rewrite logging when running a
3082 filter in a non-root, non-exim process.
3083
3084
3085Exim version 4.24
3086-----------------
3087
3088 1. The buildconfig auxiliary program wasn't quoting the value set for
3089 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3090 not defined. This bug was masked in 4.22 by the effect that was fixed in
3091 change 4.23/1.
3092
3093 2. Some messages that were rejected after a message id was allocated were
3094 shown as "incomplete" by exigrep. It no longer does this for messages that
3095 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3096
3097 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3098 have allow_domain_literals set, the ID did not get logged in the <= line.
3099 Domain literals are now always recognized in Message-ID: header lines.
3100
3101 4. The first argument for a ${extract expansion item is the key name or field
3102 number. Leading and trailing spaces in this item were not being ignored,
3103 causing some misleading effects.
3104
3105 5. When deliver_drop_privilege was set, single queue runner processes started
3106 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3107 same command in the process it spins off) were not dropping privilege.
3108
3109 6. When the daemon running as "exim" started a queue runner, it always
3110 re-executed Exim in the spun-off process. This is a waste of effort when
3111 deliver_drop_privilege is set. The new process now just calls the
3112 queue-runner function directly.
3113
3114
3115Exim version 4.23
3116-----------------
3117
3118 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3119 HEADERS_CHARSET.
3120
3121 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3122 ignored. Though the use of -oP was forcing the writing of a pid file, it
3123 was always written to the default place.
3124
3125 3. If the message "no IP address found for host xxxx" is generated during
3126 incoming verification, it is now followed by identification of the incoming
3127 connection (so you can more easily find what provoked it).
3128
3129 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3130
3131 5. Added some features to "harden" Exim a bit more against certain attacks:
3132
3133 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3134 be put in Local/Makefile. This is like the never_users runtime option,
3135 but it cannot be overridden. The default setting is "root".
3136
3137 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3138 prefix string with which any file named in a -C command line option
3139 must start.
3140
3141 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3142 is retained for -C and -D only if the caller of Exim is root. Without
3143 it, the exim user may also use -C and -D and retain privilege.
3144
3145 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3146 command line option is disabled.
3147
3148 6. Macro names set by the -D option must start with an upper case letter, just
3149 like macro names defined in the configuration file.
3150
3151 7. Added "dereference=" facility to LDAP.
3152
3153 8. Two instances of the typo "uknown" in the source files are fixed.
3154
3155 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3156 the Configure-Makefile script screwed up while processing it.
3157
315810. Incorporated PCRE 4.4.
3159
316011. The SMTP synchronization check was not operating right at the start of an
3161 SMTP session. For example, it could not catch a HELO sent before the client
3162 waited for the greeting. There is now a check for outstanding input at the
3163 point when the greeting is written. Because of the duplex, asynchronous
3164 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3165 way, but not yet received, when the check is performed.
3166
316712. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3168 on TCP/IP sockets, because this apparently causes some broken clients to
3169 timeout.
3170
317113. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3172 unchanged) from the Cygwin maintainer.
3173
317414. The code for -bV that shows what is in the binary showed "mbx" when maildir
3175 was supported instead of testing for mbx. Effectively a typo.
3176
317715. The spa authenticator server code was not checking that the input it
3178 received was valid base64.
3179
318016. The debug output line for the "set" modifier in ACLs was not showing the
3181 name of the variable that was being set.
3182
318317. Code tidy: the variable type "vtype_string" was never used. Removed it.
3184
318518. Previously, a reference to $sender_host_name did not cause a DNS reverse
3186 lookup on its own. Something else was needed to trigger the lookup. For
3187 example, a match in host_lookup or the need for a host name in a host list.
3188 Now, if $sender_host_name is referenced and the host name has not yet been
3189 looked up, a lookup is performed. If the lookup fails, the variable remains
3190 empty, and $host_lookup_failed is set to "1".
3191
319219. Added "eqi" as a case-independent comparison operator.
3193
319420. The saslauthd authentication condition could segfault if neither service
3195 nor realm was specified.
3196
319721. If an overflowing value such as "2048M" was set for message_size_limit, the
3198 error message that was logged was misleading, and incoming SMTP
3199 connections were dropped. The message is now more accurate, and temporary
3200 errors are given to SMTP connections.
3201
320222. In some error situations (such as 21 above) Exim rejects all SMTP commands
3203 (except RSET) with a 421 error, until QUIT is received. However, it was
3204 failing to send a response to QUIT.
3205
320623. The HELO ACL was being run before the code for helo_try_verify_hosts,
3207 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3208 ACL is now run after the helo_try_verify_hosts code.
3209
321024. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3211 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3212 case-independent, so other case variants are also recognized). Apparently
3213 some systems use these upper case variants.
3214
321525. If more than two messages were waiting for the same host, and a transport
3216 filter was specified for the transport, Exim sent two messages over the
3217 same TCP/IP connection, and then failed with "socket operation on non-
3218 socket" when it tried to send the third.
3219
322026. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3221
322227. The extern definition of crypt16() in expand.c was not being excluded when
3223 the OS had its own crypt16() function.
3224
322528. Added bounce_return_body as a new option, and bounce_return_size_limit
3226 as a preferred synonym for return_size_limit, both as an option and as an
3227 expansion variable.
3228
322929. Added LIBS=-liconv to OS/Makefile-OSF1.
3230
323130. Changed the default configuration ACL to relax the local part checking rule
3232 for addresses that are not in any local domains. For these addresses,
3233 slashes and pipe symbols are allowed within local parts, but the sequence
3234 /../ is explicitly forbidden.
3235
323631. SPA server authentication was not clearing the challenge buffer before
3237 using it.
3238
323932. log_message in a "warn" ACL statement was writing to the reject log as
3240 well as to the main log, which contradicts the documentation and doesn't
3241 seem right (because no rejection is happening). So I have stopped it.
3242
324333. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3244 However, I am unable to do any testing of this.
3245
324634. Fixed an infelicity in the appendfile transport. When checking directories
3247 for a mailbox, to see if any needed to be created, it was accidentally
3248 using path names with one or more superfluous leading slashes; tracing
3249 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3250
325135. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3252 contents are added to the log line that is written for every discarded
3253 recipient. (Previously a log_message setting was ignored.)
3254
325536. The ${quote: operator now quotes the string if it is empty.
3256
325737. The install script runs exim in order to find its version number. If for
3258 some reason other than non-existence or emptiness, which it checks, it
3259 could not run './exim', it was installing it with an empty version number,
3260 i.e. as "exim-". This error state is now caught, and the installation is
3261 aborted.
3262
326338. An argument was missing from the function that creates an error message
3264 when Exim fails to connect to the socket for saslauthd authentication.
3265 This could cause Exim to crash, or give a corrupted message.
3266
326739. Added isip, isip4, and isip6 to ${if conditions.
3268
326940. The ACL variables $acl_xx are now saved with the message, and can be
3270 accessed later in routers, transports, and filters.
3271
327241. The new lookup type nwildlsearch is like wildlsearch, except that the key
3273 strings in the file are not string-expanded.
3274
327542. If a MAIL command specified a SIZE value that was too large to fit into an
3276 int variable, the check against message_size_limit failed. Such values are
3277 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3278 day this will have to be increased, but I don't think I want to be around
3279 when emails are that large.
3280
3281
3282
3283Exim version 4.22
3284-----------------
3285
3286 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3287 iconv() is not standard in FreeBSD.
3288
3289 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3290 IPv6 enabled. The observed symptom was a segmentation fault on return from
3291 the function os_common_find_running_interfaces() in src/os.c.
3292
3293 3. In the check_special_case() function in daemon.c I had used "errno" as an
3294 argument name, which causes warnings on some systems. This was basically a
3295 typo, since it was named "eno" in the comments!
3296
3297 4. The code that waits for the clock to tick (at a resolution of some fraction
3298 of a second) so as to ensure message-id uniqueness was always waiting for
3299 at least one whole tick, when it could have waited for less. [This is
3300 almost certainly not relevant at current processor speeds, where it is
3301 unlikely to ever wait at all. But we try to future-proof.]
3302
3303 5. The function that sleeps for a time interval that includes fractions of a
3304 second contained a race. It did not block SIGALRM between setting the
3305 timer, and suspending (a couple of lines later). If the interval was short
3306 and the sigsuspend() was delayed until after it had expired, the suspension
3307 never ended. On busy systems this could lead to processes getting stuck for
3308 ever.
3309
3310 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3311 process, before it forks any delivery processes. The open lookup caching
3312 mechanism meant that the open file or database connection was passed into
3313 the delivery process. The problem was that delivery processes always tidy
3314 up cached lookup data. This could cause a problem for the next delivery
3315 process started by the queue runner, because the external queue runner
3316 process does not know about the closure. So the next delivery process
3317 still has data in the lookup cache. In the case of a file lookup, there was
3318 no problem because closing a file descriptor in a subprocess doesn't affect
3319 the parent. However, if the lookup was caching a connection to a database,
3320 the connection was closed, and the second delivery process was likely to
3321 see errors such as "PGSQL: query failed: server closed the connection
3322 unexpectedly". The problem has been fixed by closing all cached lookups
3323 in a queue runner before running a delivery process.
3324
3325 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3326 seem to have the "const" qualifier which it has on other OS. I've
3327 parameterised it.
3328
3329 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3330 *of the same type* (client or server) with the same public name that an
3331 error should be diagnosed.
3332
3333 9. When Exim looked up a host name for an IP address, but failed to find the
3334 original IP address when looking up the host name (a safety check), it
3335 output the message "<ip address> does not match any IP for NULL", which was
3336 confusing, to say the least. The bug was that the host name should have
3337 appeared instead of "NULL".
3338
333910. Since release 3.03, if Exim is called by a uid other than root or the Exim
3340 user that is built into the binary, and the -C or -D options is used, root
3341 privilege is dropped before the configuration file is read. In addition,
3342 logging is switched to stderr instead of the normal log files. If the
3343 configuration then re-defines the Exim user, the unprivileged environment
3344 is probably not what is expected, so Exim logs a panic warning message (but
3345 proceeds).
3346
3347 However, if deliver_drop_privilege is set, the unprivileged state may well
3348 be exactly what is intended, so the warning has been cut out in that case,
3349 and Exim is allowed to try to write to its normal log files.
3350
3351
3352Exim version 4.21
3353-----------------
3354
3355 1. smtp_return_error_details was not giving details for temporary sender
3356 or receiver verification errors.
3357
3358 2. Diagnose a configuration error if two authenticators have the same public
3359 name.
3360
3361 3. Exim used not to create the message log file for a message until the first
3362 delivery attempt. This could be confusing when incoming messages were held
3363 for policy or load reasons. The message log file is now created at the time
3364 the message is received, and an initial "Received" line is written to it.
3365
3366 4. The automatically generated man page for command line options had a minor
3367 bug that caused no ill effects; however, a more serious problem was that
3368 the procedure for building the man page automatically didn't always
3369 operate. Consequently, release 4.20 contains an out-of-date version. This
3370 shouldn't happen again.
3371
3372 5. When building Exim with embedded Perl support, the script that builds the
3373 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3374 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3375
3376 6. The freeze_tell option was not being used for messages that were frozen on
3377 arrival, either by an ACL or by local_scan().
3378
3379 7. Added the smtp_incomplete_transaction log selector.
3380
3381 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3382 was accepting AUTH without a new EHLO.
3383
3384 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3385 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3386 RFC.
3387
338810. Logging of TCP/IP connections (when configured) now happens in the main
3389 daemon process instead of the child process, so that the TCP/IP connection
3390 count is more accurate (but it can never be perfect).
3391
339211. The use of "drop" in a nested ACL was not being handled correctly in the
3393 outer ACL. Now, if condition failure induced by the nested "drop" causes
3394 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3395 or "require"), the connection is dropped.
3396
339712. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3398 that yield "discard" can now be used with an "accept" or a "discard" verb,
3399 but an error is generated for any others (because I can't see a useful way
3400 to define what should happen).
3401
340213. When an ACL is read dynamically from a file (or anywhere else), the lines
3403 are now processed in the same way as lines in the Exim configuration file.
3404 In particular, continuation lines are supported.
3405
340614. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3407
340815. Added -ti meaning -t -i.
3409
341016. Check for letters, digits, hyphens, and dots in the names of dnslist
3411 domains, and warn by logging if others are found.
3412
341317. At least on BSD, alignment is not guarenteed for the array of ifreq's
3414 returned from GIFCONF when Exim is trying to find the list of interfaces on
3415 a host. The code in os.c has been modified to copy each ifreq to an aligned
3416 structure in all cases.
3417
3418 Also, in some cases, the returned ifreq's were being copied to a 'struct
3419 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3420 means the last couple of bytes of an IPv6 address could be chopped if the
3421 ifreq contained only a normal sockaddr (14 bytes storage).
3422
342318. Named domain lists were not supported in the hosts_treat_as_local option.
3424 An entry such as +xxxx was not recognized, and was treated as a literal
3425 domain name.
3426
342719. Ensure that header lines added by a DATA ACL are included in the reject log
3428 if the ACL subsequently rejects the message.
3429
343020. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3431 MD5 (which is deprecated).
3432
343321. When testing a filter file using -bf, Exim was writing a message when it
3434 took the sender from a "From " line in the message, but it was not doing so
3435 when it took $return_path from a Return-Path: header line. It now does.
3436
343722. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3438 with a valid header line field name (a series of printing characters
3439 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3440
344123. Changed "disc" in the source to "disk" to conform to the documentation and
3442 the book and for uniformity.
3443
344424. Ignore Sendmail's -Ooption=value command line item.
3445
344625. When execve() failed while trying to run a command in a pipe transport,
3447 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3448 could be confused with a return value of 69 from the command itself. This
3449 has been changed to 127, the value the shell returns if it is asked to run
3450 a non-existent command. The wording for the related log line suggests a
3451 non-existent command as the problem.
3452
345326. If received_header_text expands to an empty string, do not add a Received:
3454 header line to the message. (Well, it adds a token one on the spool, but
3455 marks it "old" so that it doesn't get used or transmitted.)
3456
345727. Installed eximstats 1.28 (addition of -nt option).
3458
345928. There was no check for failure on the call to getsockname() in the daemon
3460 code. This can fail if there is a shortage of resources on the system, with
3461 ENOMEM, for example. A temporary error is now given on failure.
3462
346329. Contrary to the C standard, it seems that in some environments, the
3464 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3465 program. Exim now does this explicitly; it affects the formatting of
3466 timestamps using strftime().
3467
346830. If exiqsumm was given junk data, it threw up some uninitialized variable
3469 complaints. I've now initialized all the variables, to avoid this.
3470
347132. Header lines added by a system filter were not being "seen" during
3472 transport-time rewrites.
3473
347433. The info_callback() function passed to OpenSSL is set up with type void
3475 (*)(SSL *, int, int), as described somewhere. However, when calling the
3476 function (actually a macro) that sets it up, the type void(*)() is
3477 expected. I've put in a cast to prevent warnings from picky compilers.
3478
347934. If a DNS black list lookup found a CNAME record, but there were no A
3480 records associated with the domain it pointed at, Exim crashed.
3481
348235. If a DNS black list lookup returned more than one A record, Exim ignored
3483 all but the first. It now scans all returned addresses if a particular IP
3484 value is being sought. In this situation, the contents of the
3485 $dnslist_value variable are a list of all the addresses, separated by a
3486 comma and a space.
3487
348836. Tightened up the rules for host name lookups using reverse DNS. Exim used
3489 to accept a host name and all its aliases if the forward lookup for any of
3490 them yielded the IP address of the incoming connection. Now it accepts only
3491 those names whose forward lookup yields the correct IP address. Any other
3492 names are discarded. This closes a loophole whereby a rogue DNS
3493 administrator could create reverse DNS records to break through a
3494 wildcarded host restriction in an ACL.
3495
349637. If a user filter or a system filter that ran in a subprocess used any of
3497 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3498 the wrong values were passed to the pipe command ($thisaddress had the
3499 value of $0, $0 had the value of $1, etc). This bug was introduced by
3500 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3501
350238. Improved the line breaking for long SMTP error messages from ACLs.
3503 Previously, if there was no break point between 40 and 75 characters, Exim
3504 left the rest of the message alone. Two changes have been made: (a) I've
3505 reduced the minimum length to 35 characters; (b) if it can't find a break
3506 point between 35 and 75 characters, it looks ahead and uses the first one
3507 that it finds. This may give the occasional overlong line, but at least the
3508 remaining text gets split now.
3509
351039. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3511 file descriptors might be low, and that setting 1000 would always raise it.
3512 It turns out that in some environments, the limit is already over 1000 and
3513 that lowering it causes trouble. So now Exim takes care not to decrease it.
3514
351540. When delivering a message, the value of $return_path is set to $sender_
3516 address at the start of routing (routers may change the value). By an
3517 oversight, this default was not being set up when an address was tested by
3518 -bt or -bv, which affected the outcome if any router or filter referred to
3519 $return_path.
3520
352141. The idea of the "warn" ACL verb is that it adds a header or writes to the
3522 log only when "message" or "log_message" are set. However, if one of the
3523 conditions was an address verification, or a call to a nested ACL, the
3524 messages generated by the underlying test were being passed through. This
3525 no longer happens. The underlying message is available in $acl_verify_
3526 message for both "message" and "log_message" expansions, so it can be
3527 passed through if needed.
3528
352942. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3530 new expansion $bh_ to give the encoded byte string without charset
3531 translation. Translation happens only if iconv() is available; HAVE_ICONV
3532 indicates this at build time. HEADERS_CHARSET gives the charset to
3533 translate to; headers_charset can change it in the configuration, and
3534 "headers charset" can change it in an individual filter file.
3535
353643. Now that we have a default RFC 2047 charset (see above), the code in Exim
3537 that creates RFC 2047 encoded "words" labels them as that charset instead
3538 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3539 expansion operator; (ii) when Exim creates a From: line for a local
3540 message; (iii) when a header line is rewritten to include a "phrase" part.
3541
354244. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3543 buggy, causing it to skip the first lines of messages whose message ID
3544 ended in 'D'. This would not have bitten before Exim release 4.14, because
3545 message IDs were unlikely to end in 'D' before then. The effect was to have
3546 incorrect size information for certain domains.
3547
354845. #include "config.h" was missing at the start of the crypt16.c module. This
3549 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3550 noticed.
3551
355246. If there was a timeout during a "random" callout check, Exim treated it as
3553 a failure of the random address, and carried on sending RSET and the real
3554 address. If the delay was just some slowness somewhere, the response to the
3555 original RCPT would be taken as a response to RSET and so on, causing
3556 mayhem of various kinds.
3557
355847. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3559 when I implemented it. It didn't allow for the fact that some option values
3560 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3561 the right test for positive values.
3562
356348. Domain names in DNS records are case-independent. Exim always looks them up
3564 in lower case. Some resolvers return domain names in exactly the case they
3565 appear in the zone file, that is, they may contain uppercase letters. Not
3566 all resolvers do this - some return always lower case. Exim was treating a
3567 change of case by a resolver as a change of domain, similar to a widening
3568 of a domain abbreviation. This triggered its re-routing code and so it was
3569 trying to route what was effectively the same domain again. This normally
3570 caused routing to fail (because the router wouldn't handle the domain
3571 twice). Now Exim checks for this case specially, and just changes the
3572 casing of the domain that it ultimately uses when it transmits the message
3573 envelope.
3574
357549. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3576 module.
3577
357850. If a filter generated a file delivery with a non-absolute name (possible if
3579 no home directory exists for the router), the forbid_file option was not
3580 forbidding it.
3581
358251. Added '&' feature to dnslists, to provide bit mask matching in addition to
3583 the existing equality matching.
3584
358552. Exim was using ints instead of ino_t variables in some places where it was
3586 dealing with inode numbers.
3587
358853. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3589 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3590 called TMPDIR, and if it finds it is different, it changes its value.
3591
359254. The smtp_printf() function is now made available to local_scan() so
3593 additional output lines can be written before returning. There is also an
3594 smtp_fflush() function to enable the detection of a dropped connection.
3595 The variables smtp_input and smtp_batched_input are exported to
3596 local_scan().
3597
359855. Changed the default runtime configuration: the message "Unknown user"
3599 has been removed from the ACL, and instead placed on the localuser router,
3600 using the cannot_route_message feature. This means that any verification
3601 failures that generate their own messages won't get overridden. Similarly,
3602 the "Unrouteable address" message that was in the ACL for unverifiable
3603 relay addresses has also been removed.
3604
360556. Added hosts_avoid_esmtp to the smtp transport.
3606
360757. The exicyclog script was not checking for the esoteric option
3608 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3609 will work only if exicyclog is run under the appropriate euid.
3610
361158. Following a discussion on the list, the rules by which Exim recognises line
3612 endings on incoming messages have been changed. The -dropcr and drop_cr
3613 options are now no-ops, retained only for backwards compatibility. The
3614 following line terminators are recognized: LF CRLF CR. However, special
3615 processing applies to CR:
3616
3617 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3618 nor a local message in the state where . is a terminator.
3619
3620 (ii) If a bare CR is encountered in a header line, an extra space is added
3621 after the line terminator so as not to end the header. The reasoning
3622 behind this is that bare CRs in header lines are most likely either
3623 to be mistakes, or people trying to play silly games.
3624
362559. The size of a message, as listed by "-bp" or in the Exim monitor window,
3626 was being incorrectly given as 18 bytes larger than it should have been.
3627 This is a VOB (very old bug).
3628
362960. This may never have affected anything current, but just in case it has:
3630 When the local host is found other than at the start of a list of hosts,
3631 the local host, those with the same MX, and any that follow, are discarded.
3632 When the list in question was part of a longer list of hosts, the following
3633 hosts (not currently being processed) were also being discarded. This no
3634 longer happens. I'm not sure if this situation could ever has previously
3635 arisen.
3636
363761. Added the "/MX" feature to lists of hosts in the manualroute and query
3638 program routers.
3639
364062. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3641 header. This is something that is recommended in a new Internet Draft, and
3642 is something that is documented as being done by Sendmail. There are two
3643 possible values. For messages generated by the autoreply transport, Exim
3644 adds:
3645
3646 Auto-Submitted: auto-replied
3647
3648 whereas for all other generated messages (e.g. bounces) it adds
3649
3650 Auto-Submitted: auto-generated
3651
365263. The "personal" condition in filters now includes a test for the
3653 Auto-Submitted: header. If it contains the string "auto-" the message it
3654 not considered personal.
3655
365664. Added rcpt_include_affixes as a generic transport option.
3657
365865. Added queue_only_override (default true).
3659
366066. Added the syslog_duplication option.
3661
366267. If what should have been the first header line of a message consisted of
3663 a space followed by a colon, Exim was mis-interpreting it as a header line.
3664 It isn't of course - it is syntactically invalid and should therefore be
3665 treated as the start of the message body. The misbehaviour could have
3666 caused a number of strange effects, including loss of data in subsequent
3667 header lines, and spool format errors.
3668
366968. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3670 client host had authenticated. This control can now be exercised by an ACL
3671 for more flexibility.
3672
367369. By default, callouts do not happen when testing with -bh. There is now a
3674 variant, -bhc, which does actually run the callout code, including
3675 consulting and updating the callout cache.
3676
367770. Added support for saslauthd authentication, courtesy of Alexander
3678 Sabourenkov.
3679
368071. If statvfs() failed on the spool or log directories while checking their
3681 size for availability, Exim confusingly gave the error "space shortage".
3682 Furthermore, in debugging mode it crashed with a floating point exception.
3683 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3684 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3685 serious problem, Exim now writes to the main and panic logs when this
3686 happens, with details of the failure. It then refuses to accept the
3687 incoming message, giving the message "spool directory problem" or "log