Remove obsolete commented-out code from receive.c.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
2cbb4081 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.300 2006/02/14 15:11:43 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
214e2000 169
5de37277
PH
170Exim version 4.60
171-----------------
172
cc38ddbf
PH
173PH/01 Two changes to the default runtime configuration:
174
175 (1) Move the checks for relay_from_hosts and authenticated clients from
176 after to before the (commented out) DNS black list checks.
177
178 (2) Add control=submission to the relay_from_hosts and authenticated
179 clients checks, on the grounds that messages accepted by these
180 statements are most likely to be submissions.
5de37277 181
72fdd6ae
PH
182PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
183
184 (1) Generate an error if the third argument for the ${prvs expansion is
185 not a single digit.
186
187 (2) Treat a missing third argument of ${prvscheck as if it were an empty
188 string.
189
190 (3) Reset the variables that are obtained from the first argument of
191 ${prvscheck and used in the second argument before leaving the code,
192 because their memory is reclaimed, so using them afterwards may do
193 silly things.
194
195 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
196 one (it's much easier than Tom thought :-).
197
198 (5) Because of (4), we can now allow for the use of $prvscheck_result
199 inside the third argument.
cb9328de 200
cb741023
PH
201PH/03 For some reason, the default setting of PATH when running a command from
202 a pipe transport was just "/usr/bin". I have changed it to
203 "/bin:/usr/bin".
204
f174f16e
PH
205PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
206 anything to be listed in the output from -bV.
b2f5a032 207
c25242d7
PH
208PH/05 When a filter generated an autoreply, the entire To: header line was
209 quoted in the delivery log line, like this:
210
211 => >A.N.Other <ano@some.domain> <original@ddress> ...
212
213 This has been changed so that it extracts the operative address. There
214 may be more than one such address. If so, they are comma-separated, like
215 this:
216
217 => >ano@some.domain,ona@other.domain <original@ddress> ...
218
82c19f95
PH
219PH/06 When a client host used a correct literal IP address in a HELO or EHLO
220 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
221 not being looked up in the rDNS to get a host name, Exim was showing the
222 IP address twice in Received: lines, even though the IP addresses were
223 identical. For example:
224
225 Received: from [1.2.3.4] (helo=[1.2.3.4])
226
227 However, if the real host name was known, it was omitting the HELO data
228 if it matched the actual IP address. This has been tidied up so that it
229 doesn't show the same IP address twice.
230
d7ffbc12
PH
231PH/07 When both +timestamp and +memory debugging was on, the value given by
232 $tod_xxx expansions could be wrong, because the tod_stamp() function was
233 called by the debug printing, thereby overwriting the timestamp buffer.
234 Debugging no longer uses the tod_stamp() function when +timestamp is set.
235
9f526266
PH
236PH/08 When the original message was included in an autoreply transport, it
237 always said "this is a copy of the message, including all the headers",
238 even if body_only or headers_only was set. It now gives an appropriate
239 message.
240
87fcc8b9
PH
241PH/09 Applied a patch from the Sieve maintainer which:
242
243 o fixes some comments
244 o adds the (disabled) notify extension core
245 o adds some debug output for the result of if/elsif tests
246 o points to the current vacation draft in the documentation
247 and documents the missing references header update
248
249 and most important:
250
251 o fixes a bug in processing the envelope test (when testing
252 multiple envelope elements, the last element determinted the
253 result)
254
456682f5
PH
255PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
256 Electronic Mail") by including:
257
258 Auto-submitted: auto-generated
259
260 in the messages that it generates (bounce messages and others, such as
261 warnings). In the case of bounce messages for non-SMTP mesages, there was
262 also a typo: it was using "Auto_submitted" (underscore instead of
263 hyphen). Since every message generated by Exim is necessarily in response
264 to another message, thes have all been changed to:
265
266 Auto-Submitted: auto-replied
267
268 in accordance with these statements in the RFC:
269
270 The auto-replied keyword:
271
272 - SHOULD be used on messages sent in direct response to another
273 message by an automatic process,
274
275 - MUST NOT be used on manually-generated messages,
276
277 - MAY be used on Delivery Status Notifications (DSNs) and Message
278 Disposition Notifications (MDNs),
279
280 - MUST NOT be used on messages generated by automatic or periodic
281 processes, except for messages which are automatic responses to
282 other messages.
283
3e46c1aa
PH
284PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
285 to the default Received: header definition.
456682f5 286
49826d12
PH
287PH/12 Added log selector acl_warn_skipped (default on).
288
eba0c039
PH
289PH/13 After a successful wildlsearch lookup, discard the values of numeric
290 variables because (a) they are in the wrong storage pool and (b) even if
291 they were copied, it wouldn't work properly because of the caching.
292
a0d6ba8a
PH
293PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
294 checking when decoding. Apparently there are clients that generate
295 overlong encoded strings. Why am I not surprised?
296
f0917727
PH
297PH/15 If the first argument of "${if match_address" was not empty, but did not
298 contain an "@" character, Exim crashed. Now it writes a panic log message
299 and treats the condition as false.
300
096fee00
PH
301PH/16 In autoreply, treat an empty string for "once" the same as unset.
302
024bd3c2
PH
303PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
304 extension "envelope-auth". The code is finished and in agreement with
305 other implementations, but there is no documentation so far and in fact,
306 nobody wrote the draft yet. This extension is currently #undef'ed, thus
307 not changing the active code.
308
309 Print executed "if" and "elsif" statements when debugging is used. This
310 helps a great deal to understand what a filter does.
311
312 Document more things not specified clearly in RFC3028. I had all this
313 sorted out, when out of a sudden new issues came to my mind. Oops."
314
df199fec
PH
315PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
316 (Bugzilla #53).
317
d27f1df3
PH
318PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
319 canonical form (as documented). However, after a host name lookup from
320 the IP address, check_host() was doing a simple string comparison with
321 addresses acquired from the DNS when checking that the found name did
322 have the original IP as one of its addresses. Since any found IPv6
323 addresses are likely to be in abbreviated form, the comparison could
324 fail. Luckily, there already exists a function for doing the comparison
325 by converting both addresses to binary, so now that is used instead of
326 the text comparison.
327
96776534
PH
328PH/20 There was another similar case to PH/19, when a complete host name was
329 given in a host list; looking up its IP address could give an abbreviated
330 form, whereas the current host's name might or might not be abbreviated.
331 The same fix has been applied.
332
5de37277 333
9a799bc0
PH
334Exim version 4.54
335-----------------
336
337PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
338 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
339 It now does.
340
99a4b039
PH
341PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
342 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
343
344PH/03 Typo: missing ".o" in src/pcre/Makefile.
345
4b233853
PH
346PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
347 header line, restrict the check to what is listed in RFCs 2369 and 2929.
348 Also, for "Auto-Submitted", treat anything other than "no" as
349 non-personal, in accordance with RFC 3834. (Previously it treated
350 anything starting "auto-" as non-personal.)
351
8857ccfd
PH
352TF/01 The control=submission/name=... option had a problem with syntax
353 errors if the name included a slash character. The /name= option
354 now slurps the rest of the string, so it can include any characters
355 but it must come last in the list of options (after /sender_retain
356 or /domain=).
357
433a2980
PH
358PH/05 Some modifications to the interface to the fake nameserver for the new
359 testing suite.
360
3e46c1aa 361
9a799bc0 362
e3a311ba
TK
363Exim version 4.53
364-----------------
365
366TK/01 Added the "success_on_redirect" address verification option. See
367 NewStuff for rationale and an example.
368
13b685f9
PH
369PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
370
395ff96d
PH
371PH/02 Patch to exigrep to allow it to work on syslog lines.
372
5b68f6e4
PH
373PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
374 fread() to skip over the body file's header line, because in Cygwin the
375 header line is locked and is inaccessible.
376
1ab52c69
PH
377PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
378 co-exist for some time) to make it clear that it is the Exim ID that is
379 referenced, not the Message-ID: header line.
380
b07e6aa3
PH
381PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
382 string_format() function, because snprintf() does not exist on all
383 operating systems.
384
254e032f
PH
385PH/06 The use of forbid_filter_existstest now also locks out the use of the
386 ${stat: expansion item.
387
3af76a81
PH
388PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
389 protocol synchronization error", to keep the pedants happy.
390
2548ba04
PH
391PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
392 well as for IRIX systems, when gcc is being used. See the host.c source
393 file for comments.
394
b6c6011d
PH
395PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
396
cf39cf57
PH
397PH/10 Named domain lists were not working if used in a queue_smtp_domains
398 setting.
399
f1513293
PH
400PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
401 transport and to the smtp transport in LMTP mode.
402
727549a4
PH
403TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
404
af46795e
PH
405PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
406 run a filter in a subprocess. This could lead to confusion in subsequent
407 lookups in the parent process. There should also be a search_tidyup() at
408 the end of the subprocess.
409
d7b47fd0
PH
410PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
411 only if the host matched helo_try_verify_hosts, which caused the
412 verification to occur when the EHLO/HELO command was issued. The ACL just
413 tested the remembered result. Now, if a previous verification attempt has
414 not happened, "verify = helo" does it there and then.
415
ee744174
JJ
416JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
417
b582ab87
PH
418TK/03 Fix log output including CR from clamd.
419
41a13e0a
PH
420PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
421 exist provoked a memory error which could cause a segfault.
422
f625cc5a
PH
423PH/15 Installed PCRE 6.2
424
425PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
426
21f7af35
PH
427PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
428 of the problem. Specifically, suggested +O2 rather than +O1 for the
429 HP-UX compiler.
430
31480e42
PH
431PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
432
2d280592
PH
433PH/20 If a delivery was routed to a non-standard port by means of an SRV
434 record, the port was not correctly logged when the outgoing_port log
435 selector was set (it logged the transort's default port).
436
7cd1141b
PH
437PH/21 Added support for host-specific ports to manualroute, queryprogram,
438 fallback_hosts, and "hosts" in the smtp transport.
439
440PH/22 If the log selector "outgoing_port" is set, the port is now also given on
441 host errors such as "Connection refused".
442
750af86e
PH
443PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
444 authentication with radiusclient 0.4.9:
445
446 - Error returned from rc_read_config was caught wrongly
447 - Username/password not passed on to radius server due to wrong length.
448
449 The presumption is that some radiusclient API changes for 4.51/PH/17
450 were not taken care of correctly. The code is still untested by me (my
451 Linux distribution still has 0.3.2 of radiusclient), but it was
452 contributed by a Radius user.
453
454PH/24 When doing a callout, the value of $domain wasn't set correctly when
455 expanding the "port" option of the smtp transport.
456
4304270b
TK
457TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
458 while reading a MIME header. Thanks to Tom Hughes for a patch.
459
750af86e
PH
460PH/24 Include config.h inside local_scan.h so that configuration settings are
461 available.
462
64ffc24f
PH
463PH/25 Make $smtp_command_argument available after all SMTP commands. This means
464 that in an ACL for RCPT (for example), you can examine exactly what was
465 received.
466
5dd9625b
PH
467PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
468 commands, but it was not correctly comparing the address with the actual
469 client host address. Thus, it would show the EHLO address in Received:
470 header lines when this was not necessary.
471
5591031b
PH
472PH/27 Added the % operator to ${eval:}.
473
ba18e66a
PH
474PH/28 Exim tries to create and chdir to its spool directory when it starts;
475 it should be ignoring failures (because with -C, for example, it has lost
476 privilege). It wasn't ignoring creation failures other than "already
477 exists".
478
9cec981f
PH
479PH/29 Added "crypteq" to the list of supported features that Exim outputs when
480 -bV or -d is used.
481
aa2b5c79
PH
482PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
483 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
484 too many continuations, the temporary file was not being removed, and the
485 return code was incorrect.
aa2b5c79 486
48a53b7f
PH
487PH/31 Missing "BOOL" in function definition in filtertest.c.
488
1c59d63b
PH
489PH/32 Applied Sieve patches from the maintainer.
490
671012da
TK
491TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
492
1509d3a8
PH
493PH/33 Added "verify = not_blind".
494
495PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
496 Local/Makefile (with some defaults set). These are used in built scripts
497 such as exicyclog, but they have never been used in the exim_install
498 script (though there are many overriding facilities there). I have
499 arranged that the exim_install script now takes note of these two
500 settings.
501
502PH/35 Installed configuration files for Dragonfly.
503
2fe1a124
PH
504PH/36 When a locally submitted message by a trusted user did not contain a
505 From: header, and the sender address was obtained from -f or from an SMTP
506 MAIL command, and the trusted user did not use -F to supply a sender
507 name, $originator_name was incorrectly used when constructing a From:
508 header. Furthermore, $originator_name was used for submission mode
509 messages from external hosts without From: headers in a similar way,
510 which is clearly wrong.
511
8800895a
PH
512PH/37 Added control=suppress_local_fixups.
513
ccfdb010
PH
514PH/38 When log_selector = +received_sender was set, and the addition of the
515 sender made the log line's construction buffer exactly full, or one byte
516 less than full, an overflow happened when the terminating "\n" was
517 subsequently added.
518
1130bfb0
PH
519PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
520 when the result of a list match is failure because a DNS lookup failed.
521
ebcb507f
PH
522PH/40 RM_COMMAND is now used in the building process.
523
c35e155c
PH
524PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
525 the "build-* directories that it finds.
526
95d1f782
PH
527PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
528 address in a domain literal was a prefix of an interface address.
529
fd6de02e
PH
530PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
531 when verifying a sender address, unless rewrite_headers is false.
532
58de37c5
PH
533PH/44 Wrote a long comment about why errors_to addresses are verified as
534 recipients, not senders.
535
261cf466
TF
536TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
537 the ratelimit ACL was added.
538
3ee512ff
PH
539PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
540
e08c430f
PH
541PH/46 Added extra information about PostgreSQL errors to the error string.
542
bef5a11f
PH
543PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
544 avoiding the need to install special zones in a real server. This is
545 backwards compatible; if it can't find the fake resolver, it drops back.
546 Thus, both old and new test suites can be run.
547
7546de58
TF
548TF/02 Added util/ratelimit.pl
549
e5d5a95f
TF
550TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
551 clock is set back in time.
552
2e88a017
TF
553TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
554 Candler <B.Candler@pobox.com>.
555
a5f65aa4
TF
556TF/05 The fix for PH/43 was not completely correct; widen_domains is always
557 OK for addresses that are the result of redirections.
558
e7726cbf
PH
559PH/48 A number of further additions for the benefit of the new test suite,
560 including a fake gethostbyname() that interfaces to the fake DNS resolver
561 (see PH/47 above).
562
a7fdad5b
TF
563TF/06 The fix for widen_domains has also been applied to qualify_single and
564 search_parents which are the other dnslookup options that can cause
565 header rewrites.
566
6af56900
PH
567PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
568 type ("H").
569
0925ede6
PH
570PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
571
66afa403
TF
572TF/07 Exim produced the error message "an SRV record indicated no SMTP
573 service" if it encountered an MX record with an empty target hostname.
574 The message is now "an MX or SRV record indicated no SMTP service".
575
0154e85a
TF
576TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
577 if the DNS of the sending site is misconfigured. This is quite a
578 common situation. This change restores the behaviour of treating a
579 helo verification defer as a failure.
580
16f12c76
PH
581PH/51 If self=fail was set on a router, the bounce message did not include the
582 actual error message.
583
bbe902f0 584
e5a9dba6
PH
585Exim version 4.52
586-----------------
587
588TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
589
22c3b60b
PH
590PH/01 When a transport filter timed out in a pipe delivery, and the pipe
591 command itself ended in error, the underlying message about the transport
592 filter timeout was being overwritten with the pipe command error. Now the
593 underlying error message should be appended to the second error message.
594
06a9b4b5
PH
595TK/01 Fix poll() being unavailable on Mac OSX 10.2.
596
c1ac6996
PH
597PH/02 Reduce the amount of output that "make" produces by default. Full output
598 can still be requested.
599
9c7a242c
PH
600PH/03 The warning log line about a condition test deferring for a "warn" verb
601 was being output only once per connection, rather than after each
602 occurrence (because it was using the same function as for successful
603 "warn" verbs). This seems wrong, so I have changed it.
604
87ba3f5f
PH
605TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
606 it should not have, which might have caused a crash in the right
607 circumstances, but probably never did.
608
609PH/04 Installed a modified version of Tony Finch's patch to make submission
610 mode fix the return path as well as the Sender: header line, and to
611 add a /name= option so that you can make the user's friendly name appear
612 in the header line.
613
29aba418
TF
614TF/03 Added the control = fakedefer ACL modifier.
615
fe0dab11
TF
616TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
617 Mark Lowes for thorough testing.
870f6ba8 618
11d337a4
TK
619TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
620
621TK/03 Merged latest SRS patch from Miles Wilton.
622
415c8f3b
PH
623PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
624 with the definition in sysexits.h (which is #included earlier).
625 Fortunately, Exim does not actually use EX_OK. The code used to try to
626 preserve the sysexits.h value, by assumimg that macro definitions were
627 scanned for macro replacements. I have been disabused of this notion,
628 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 629
958541e9
PH
630PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
631 in the smtp transport. When a block could not be written in a single
632 write() function, the timeout was being re-applied to each part-write.
633 This seems wrong - if the receiver was accepting one byte at a time it
634 would take for ever. The timeout is now adjusted when this happens. It
635 doesn't have to be particularly precise.
636
c206415f
TK
637TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
638 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
639
2a4be8f9
PH
640PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
641 without a domain if the check to <postmaster@domain> fails.
642
1cba11c5
SC
643SC/01 Eximstats: added -xls and the ability to specify output files
644 (patch written by Frank Heydlauf).
645
646SC/02 Eximstats: use FileHandles for outputing results.
647
648SC/03 Eximstats: allow any combination of xls, txt, and html output.
649
650SC/04 Eximstats: fixed display of large numbers with -nvr option
651
652SC/05 Eximstats: fixed merging of reports with empty tables.
653
654SC/06 Eximstats: added the -include_original_destination flag
655
656SC/07 Eximstats: removed tabs and trailing whitespace.
657
1005d00e
TK
658TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
659
660TK/06 MBOX spool code: Add real "From " MBOX separator line
661 so the .eml file is really in mbox format (even though
662 most programs do not really care). Patch from Alex Miller.
663
664TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
665 The latter is generated from $received_to and is only set if the
666 message has one envelope recipient. SA can use these headers,
667 obviously out-of-the-box. Patch from Alex Miller.
668
9b4768fa
PH
669PH/08 The ${def test on a variable was returning false if the variable's
670 value was "0", contrary to what the specification has always said!
671 The result should be true unless the variable is empty.
672
673PH/09 The syntax error of a character other than { following "${if
674 def:variable_name" (after optional whitespace) was not being diagnosed.
675 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
676 accidental colon was present, for example, could give incorrect results.
677
0d7eb84a
PH
678PH/10 Tidied the code in a number of places where the st_size field of a stat()
679 result is used (not including appendfile, where other changes are about
680 to be made).
681
682PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
683 This involved changing a lot of size variables from int to off_t. It
684 should work with maildirs and everything.
685
40727bee
TK
686TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
687 spamd dying while we are connected to it.
688
554d2369
TF
689TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
690 <jgh@wizmail.org>
691
1f922db1
PH
692PH/12 Applied Alex Kiernan's patch for the API change for the error callback
693 function for BDB 4.3.
694
ef213c3b
PH
695PH/13 Changed auto_thaw such that it does not apply to bounce messages.
696
8ac170f3
PH
697PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
698 the sources for PCRE have been re-arranged and more files are now
699 involved.
700
b1c749bb
PH
701PH/15 The code I had for printing potentially long long variables in PH/11
702 above was not the best (it lost precision). The length of off_t variables
703 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
704 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
705 to be "long long int" or "long int". This is needed for the internal
706 formatting function string_vformat().
b1c749bb 707
4aac9b49
PH
708PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
709 the configuration file to be ":syslog", then the script "guesses" where
710 the logs files are, rather than using the compiled in default. In our
711 case the guess is not the same as the compiled default, so the script
712 suddenly stopped working when I started to use syslog. The patch checks
713 to see if log_file_path is "". If so, it attempts to read it from exim
714 with no configuration file to get the compiled in version, before it
715 falls back to the previous guessing code."
716
294520c8
TK
717TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
718 implementing BATV in an Exim configuration. See NewStuff for the gory
719 details.
720
5bd022fe
PH
721PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
722 Makefile that are specific to HP-UX.
723
90e9ce59
PH
724PH/18 If the "use_postmaster" option was set for a recipient callout together
725 with the "random" option, the postmaster address was used as the MAIL
726 FROM address for the random test, but not for the subsequent recipient
727 test. It is now used for both.
728
5ea81592
PH
729PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
730 patch removes a few documentation additions to RFC 3028, because the
731 latest draft now contains them. It adds the new en;ascii-case comparator
732 and a new error check for 8bit text in MIME parts. Comparator and
733 require names are now matched exactly. I enabled the subaddress
734 extension, but it is not well tested yet (read: it works for me)."
735
c6c2dc1d
PH
736PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
737 rework some of the code of TK/09 above to avoid the hardwired use of
738 "%lld" and "long long". Replaced the call to snprintf() with a call to
739 string_vformat().
740
fffffe4c
PH
741PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
742 records point to non-existent hosts", "retry timeout exceeded", and
743 "retry time not reached for any host after a long failure period".
ca02eafb 744
9a26b6b2
PH
745PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
746 experimental DomainKeys support:
747
748 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
749 (2) On an error such as an illegally used "control", the wrong name for
750 the control was given.
751
752 These problems did NOT occur unless DomainKeys support was compiled.
753
4aee0225
PH
754PH/23 Added daemon_startup_retries and daemon_startup_sleep.
755
32d668a5
PH
756PH/24 Added ${if match_ip condition.
757
8187c3f3
PH
758PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
759 databases so that it will be absolutely obvious if a crash occurs in the
760 DB library. This is a regular occurrence (often caused by mis-matched
761 db.h files).
762
ff790e47 763PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
764 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
765 picked up on a user's system that detects such things. There doesn't seem
766 to be a gcc warning option for this - only an attribute that has to be
767 put on the function's prototype. It seems that in Fedora Core 4 they have
768 set this on a number of new functions. No doubt there will be more in due
769 course.
ff790e47 770
5417f6d1
PH
771PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
772 specify a transport. However, if an address that was verified by such a
773 router was the subject of a callout, Exim crashed because it tried to
774 read the rcpt_include_affixes from the non-existent transport. Now it
775 just assumes that the setting of that option is false. This bug was
776 introduced by 4.51/PH/31.
777
59cf8544
PH
778PH/28 Changed -d+all to exclude +memory, because that information is very
779 rarely of interest, but it makes the output a lot bigger. People tend to
780 do -d+all out of habit.
781
e7ad8a65
PH
782PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
783 code in os-type was giving problems when libc.so lives in lib64, like on
784 x86_64 Fedora Core.
785
ade42478
PH
786PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
787 aren't the modern standard, and it seems that some systems' include files
788 don't always have them. Exim was already checking for some of the newer
789 ones like T_AAAA, and defining it itself. I've added checks for all the
790 record types that Exim uses.
791
182ad5cf
PH
792PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
793 not automatically generating a new one, as it is supposed to. This
794 prevented TLS from working. If the file did exist, but contained invalid
795 data, a new version was generated, as expected. It was only the case of a
796 non-existent file that was broken.
797
b0d9fc80
TK
798TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
799 with a change in libdomainkeys > 0.64.
800
801TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
802 from DNS. If the selector record carries the flag, it now has
803 precedence over the domain-wide flag.
804
805TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
806
47c7a64a
PH
807PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
808 the use of a port name that isn't defined in /etc/services) occurred, the
809 message was deferred as in a normal delivery, and thus remained on the
810 spool, instead of being failed because of the mua_wrapper setting. This
811 is now fixed, and I tidied up some of the mua_wrapper messages at the
812 same time.
813
a388bce4
SC
814SC/08 Eximstats: whilst parsing the mainlog(s), store information about
815 the messages in a hash of arrays rather than using individual hashes.
816 This is a bit cleaner and results in dramatic memory savings, albeit
817 at a slight CPU cost.
818
819SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
820 as requested by Marc Sherman.
821
822SC/10 Eximstats: added histograms for user specified patterns as requested
823 by Marc Sherman.
824
0793e4ed
SC
825SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
826
c58b88df
PH
827PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
828 fopen() in the content-scanning modules that did not already have it.
829
e7ad8a65 830
7982096b
PH
831Exim version 4.51
832-----------------
833
1a46a8c5
PH
834TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
835 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
836
2f079f46 837TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
838
839TK/03 Fix ACL "regex" condition not being available in MIME ACL.
840
841PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
842 to test Sieve filters that use "vacation".
843
844PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
845 that changes the way the GnuTLS parameters are stored in the cache file.
846 The new format can be generated externally. For backward compatibility,
847 if the data in the cache doesn't make sense, Exim assumes it has read an
848 old-format file, and it generates new data and writes a new file. This
849 means that you can't go back to an older release without removing the
850 file.
851
852PH/03 A redirect router that has both "unseen" and "one_time" set does not
853 work if there are any delivery delays because "one_time" forces the
854 parent to be marked "delivered", so its unseen clone is never tried
855 again. For this reason, Exim now forbids the simultaneous setting of
856 these two options.
857
858PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
859 redirected to themselves ("homonym" addresses). Read the long ChangeLog
860 entry if you want to know the details. The fix, however, neglected to
861 consider the case when local delivery batching is involved. The test for
862 "previously delivered" was not happening when checking to see if an
863 address could be batched with a previous (undelivered) one; under
864 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 865 address.
1a46a8c5
PH
866
867PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
868 in its include files, and this causes problems building Exim.
869
870PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
871 header_syntax) but Exim was just ignoring anything given after a slash.
872 In particular, this caused confusion with an attempt to use "verify =
873 reverse_host_lookup/defer_ok". An error is now given when options are
874 supplied for verify items that do not have them. (Maybe reverse_host_
875 lookup should have a defer_ok option, but that's a different point.)
876
877PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
878 defined by RFC 821) to 2048, because there were problems with some AUTH
879 commands, and RFC 1869 says the size should be increased for extended
880 SMTP commands that take arguments.
881
882PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
883 Finch).
884
885PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
886 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 887
49c2d5ea
PH
888PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
889 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
890 whatever). Otherwise the password may appear in the log. Following change
891 PH/42 below, there is no longer a chance of it appearing in a bounce
892 message.
49c2d5ea 893
bf759a8b
PH
894PH/11 Installed exipick version 20050225.0 from John Jetmore.
895
83364d30
PH
896PH/12 If the last host in a fallback_hosts list was multihomed, only the first
897 of its addresses was ever tried. (Bugzilla bug #2.)
898
7999bbd7
PH
899PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
900 the result incorrectly in the debug output. (It correctly added a newline
901 to what was transported.)
902
7dbf77c9
PH
903TF/01 Added $received_time.
904
74e0617f
PH
905PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
906 commented out examples of how to interface to a virus scanner and to
907 SpamAssassin. Also added commented examples of av_scanner and
908 spamd_address settings.
909
2f079f46
PH
910PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
911 and controls are allowed in which ACLs. There were a couple of minor
912 errors. Some of the entries in the conditions table (which is a table of
913 where they are NOT allowed) were getting very unwieldy; rewrote them as a
914 negation of where the condition IS allowed.
915
8c841523
PH
916PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
917
7766a4f0
PH
918PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
919 header file does not have a version number, so I've had to invent a new
920 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
921 API. The code is untested by me (my Linux distribution still has 0.3.2 of
922 radiusclient), but it was contributed by a Radius user.
923
8b417f2c
PH
924PH/18 Installed Lars Mainka's patch for the support of CRL collections in
925 files or directories, for OpenSSL.
926
901f42cb
PH
927PH/19 When an Exim process that is running as root has to create an Exim log
928 file, it does so in a subprocess that runs as exim:exim so as to get the
929 ownership right at creation (otherwise, other Exim processes might see
930 the file with the wrong ownership). There was no test for failure of this
931 fork() call, which would lead to the process getting stuck as it waited
932 for a non-existent subprocess. Forks do occasionally fail when resources
933 run out. I reviewed all the other calls to fork(); they all seem to check
934 for failure.
935
f9b9210e
PH
936PH/20 When checking for unexpected SMTP input at connect time (before writing
937 the banner), Exim was not dealing correctly with a non-positive return
938 from the read() function. If the client had disconnected by this time,
939 the result was a log entry for a synchronization error with an empty
940 string after "input=" when read() returned zero. If read() returned -1
941 (an event I could not check), uninitialized data bytes were printed.
942 There were reports of junk text (parts of files, etc) appearing after
943 "input=".
944
54cdb463
PH
945PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
946
cf00dad6
PH
947PH/22 Added support for macro redefinition, and (re)definition in between
948 driver and ACL definitions.
949
acb1b346
PH
950PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
951 forgetting to use the resulting value; it was using the unexpanded value.
952
c5ddb310
PH
953PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
954 hadn't been configured. The fix is from Juergen Kreileder, who
955 understands it better than I do:
956
957 "Here's what I see happening with three configured cyrus_sasl
958 authenticators configured (plain, login, cram-md5):
959
960 On startup auth_cyrus_sasl_init() gets called for each of these.
961 This means three calls to sasl_listmech() without a specified mech_list.
962 => SASL tests which mechs of all available mechs actually work
963 => three warnings about OTP not working
964 => the returned list contains: plain, login, cram-md5, digest-md5, ...
965
966 With the patch, sasl_listmech() also gets called three times. But now
967 SASL's mech_list option is set to the server_mech specified in the the
968 authenticator. Or in other words, the answer from sasl_listmech()
969 gets limited to just the mech you're testing for (which is different
970 for each call.)
971 => the return list contains just 'plain' or 'login', 'cram-md5' or
972 nothing depending on the value of ob->server_mech.
973
974 I've just tested the patch: Authentication still works fine,
975 unavailable mechs specified in the exim configuration are still
976 caught, and the auth.log warnings about OTP are gone."
977
31619da6
PH
978PH/25 When debugging is enabled, the contents of the command line are added
979 to the debugging output, even when log_selector=+arguments is not
980 specified.
981
bebaf0fc
PH
982PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
983 answer is "GNU", and only if the return is "GNU/something" is the answer
984 "Linux".
985
475fe28a
PH
986PH/27 $acl_verify_message is now set immediately after the failure of a
987 verification in an ACL, and so is available in subsequent modifiers. In
988 particular, the message can be preserved by coding like this:
989
990 warn !verify = sender
991 set acl_m0 = $acl_verify_message
992
993 Previously, $acl_verify_message was set only while expanding "message"
994 and "log_message" when a very denied access.
995
7e8bec7a
PH
996PH/28 Modified OS/os.c-Linux with
997
998 -#ifndef OS_LOAD_AVERAGE
999 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1000
1001 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1002 nomenclature these days.)
1003
e4a89c47
PH
1004PH/29 Installed patch from the Sieve maintainer that adds the options
1005 sieve_useraddress and sieve_subaddress to the redirect router.
1006
5ca2a9a1
PH
1007PH/30 In these circumstances:
1008 . Two addresses routed to the same list of hosts;
1009 . First host does not offer TLS;
1010 . First host accepts first address;
1011 . First host gives temporary error to second address;
1012 . Second host offers TLS and a TLS session is established;
1013 . Second host accepts second address.
1014 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1015 and peerdn, if requested) that were in fact used only for the second
1016 address.
7e8bec7a 1017
c688b954
PH
1018PH/31 When doing a callout as part of verifying an address, Exim was not paying
1019 attention to any local part prefix or suffix that was matched by the
1020 router that accepted the address. It now behaves in the same way as it
1021 does for delivery: the affixes are removed from the local part unless
1022 rcpt_include_affixes is set on the transport.
1023
fed77020
PH
1024PH/32 Add the sender address, as F=<...>, to the log line when logging a
1025 timeout during the DATA phase of an incoming message.
1026
7fe1560f
PH
1027PH/33 Sieve envelope tests were broken for match types other than :is. I have
1028 applied a patch sanctioned by the Sieve maintainer.
c688b954 1029
ebb6e6d5
PH
1030PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1031 the uid or gid is negative. A case of a negative gid caused this to be
1032 noticed. The fix allows for either to be negative.
1033
9c4e8f60
PH
1034PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1035 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1036 overlooked.
1037
1038PH/36 The change PH/12 above was broken. Fixed it.
1039
d7174846
PH
1040PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1041 the grounds that routing the same address twice would always produce the
1042 same answer. This might have been true once, but it is certainly no
1043 longer true now. Routing a child address may depend on the previous
1044 routing that produced that child. Some complicated redirection strategies
1045 went wrong when messages had multiple recipients, and made Exim's
1046 behaviour dependent on the order in which the addresses were given.
1047
1048 I have moved the duplicate checking until after the routing is complete.
1049 Exim scans the addresses that are assigned to local and remote
1050 transports, and removes any duplicates. This means that more work will be
1051 done, as duplicates will always all be routed, but duplicates are
1052 presumably rare, so I don't expect this is of any significance.
1053
1054 For deliveries to pipes, files, and autoreplies, the duplicate checking
1055 still happens during the routing process, since they are not going to be
1056 routed further.
1057
cfe75fc3
PH
1058PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1059 It corrects a timeout issue with spamd. This is Ian's comment: "The
1060 background is that sometimes spamd either never reads data from a
1061 connection it has accepted, or it never writes response data. The exiscan
1062 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1063 blindly assumes that writes won't block so it may never time out."
1064
be22d70e
PH
1065PH/39 Allow G after quota size as well as K and M.
1066
0612b098
PH
1067PH/40 The value set for $authenticated_id in an authenticator may not contain
1068 binary zeroes or newlines because the value is written to log lines and
1069 to spool files. There was no check on this. Now the value is run through
1070 the string_printing() function so that such characters are converted to
1071 printable escape sequences.
1072
2e0c1448
PH
1073PH/41 $message_linecount is a new variable that contains the total number of
1074 lines in the message. Compare $body_linecount, which is the count for the
1075 body only.
1076
447d236c
PH
1077PH/42 Exim no longer gives details of delivery errors for specific addresses in
1078 bounce and delay warning messages, except in certain special cases, which
1079 are as follows:
1080
1081 (a) An SMTP error message from a remote host;
1082 (b) A message specified in a :fail: redirection;
1083 (c) A message specified in a "fail" command in a system filter;
1084 (d) A message specified in a FAIL return from the queryprogram router;
1085 (e) A message specified by the cannot_route_message router option.
1086
1087 In these cases only, Exim does include the error details in bounce and
1088 warning messages. There are also a few cases where bland messages such
1089 as "unrouteable address" or "local delivery error" are given.
1090
d20976dc
PH
1091PH/43 $value is now also set for the "else" part of a ${run expansion.
1092
f656d135
PH
1093PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1094 being worked on, but at least Exim now implements the latest version to
1095 play with."
1096
2e2a30b4
PH
1097PH/45 In a pipe transport, although a timeout while waiting for the pipe
1098 process to complete was treated as a delivery failure, a timeout while
1099 writing the message to the pipe was logged, but erroneously treated as a
1100 successful delivery. Such timeouts include transport filter timeouts. For
1101 consistency with the overall process timeout, these timeouts are now
1102 treated as errors, giving rise to delivery failures by default. However,
1103 there is now a new Boolean option for the pipe transport called
1104 timeout_defer, which, if set TRUE, converts the failures into defers for
1105 both kinds of timeout. A transport filter timeout is now identified in
1106 the log output.
1107
9176e9f0
PH
1108PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1109 systems where "make" and "gmake" are different, calling "gmake" at top
1110 level broke things. I've arranged for the value of $(MAKE) to be passed
1111 from the Makefile to this script so that it can call the same version of
1112 "make".
1113
7982096b 1114
bbe902f0
PH
1115A note about Exim versions 4.44 and 4.50
1116----------------------------------------
1117
1118Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1119changes of various kinds. As a consequence, a big documentation update was
1120needed. This delayed the release for rather longer than seemed good, especially
1121in the light of a couple of (minor) security issues. Therefore, the changes
1122that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1123release. So 4.44 and 4.50 are in effect two different branches that both start
1124from 4.43.
1125
1126I have left the 4.50 change log unchanged; it contains all the changes since
11274.43. The change log for 4.44 is below; many of its items are identical to
1128those for 4.50. This seems to be the most sensible way to preserve the
1129historical information.
1130
1131
f7b63901 1132Exim version 4.50
495ae4b0
PH
1133-----------------
1134
5fe762f6
PH
1135 1. Minor wording change to the doc/README.SIEVE file.
1136
139059f6 1137 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1138 computation of the current number of files was incorrect.
495ae4b0 1139
7086e875
PH
1140 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1141 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1142 place.
1143
35af9f61
PH
1144 4. Give more explanation in the error message when the command for a transport
1145 filter fails to execute.
1146
b668c215
PH
1147 5. There are several places where Exim runs a non-Exim command in a
1148 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1149 was being done only for the command run by the queryprogram router. It is
1150 now done for all such subprocesses. The other cases are: ${run, transport
1151 filters, and the commands run by the lmtp and pipe transports.
1152
a494b1e1
PH
1153 6. Added CONFIGURE_GROUP build-time option.
1154
1155 7. Some older OS have a limit of 256 on the maximum number of file
1156 descriptors. Exim was using setrlimit() to set 1000 as a large value
1157 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1158 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1159
c5fcb476
PH
1160 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1161 was an oversight, and furthermore, ever since the addition of extra
1162 controls (e.g. 4.43/32), the checks on when to allow different forms of
1163 "control" were broken. There should now be diagnostics for all cases when a
1164 control that does not make sense is encountered.
1165
69358f02
PH
1166 9. Added the /retain_sender option to "control=submission".
1167
5be20824
PH
116810. $recipients is now available in the predata ACL (oversight).
1169
eb2c0248
PH
117011. Tidy the search cache before the fork to do a delivery from a message
1171 received from the command line. Otherwise the child will trigger a lookup
1172 failure and thereby defer the delivery if it tries to use (for example) a
1173 cached ldap connection that the parent has called unbind on.
1174
2a3eea10
PH
117512. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1176 of $address_data from the recipient verification was clobbered by the
1177 sender verification.
1178
117913. The value of address_data from a sender verification is now available in
1180 $sender_address_data in subsequent conditions in the ACL statement.
1181
23c7ff99
PH
118214. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1183
4deaf07d
PH
118415. Added a new option "connect=<time>" to callout options, to set a different
1185 connection timeout.
1186
926e1192
PH
118716. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1188 was its contents. (It was OK if the option was not defined at all.)
1189
650edc6f
PH
119017. A "Completed" log line is now written for messages that are removed from
1191 the spool by the -Mrm option.
1192
2c7db3f5
PH
119318. New variables $sender_verify_failure and $recipient_verify_failure contain
1194 information about exactly what failed.
1195
3d235903
PH
119619. Added -dd to debug only the daemon process.
1197
7c7ad977
PH
119820. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1199 handles timeouts, both on the server side and network timeouts. Renamed the
1200 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1201
981756db
PH
120221. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1203 It is now set to "smtps".
1204
d4eb88df
PH
120522. $host_address is now set to the target address during the checking of
1206 ignore_target_hosts.
1207
120823. When checking ignore_target_hosts for an ipliteral router, no host name was
1209 being passed; this would have caused $sender_host_name to have been used if
1210 matching the list had actually called for a host name (not very likely,
1211 since this list is usually IP addresses). A host name is now passed as
1212 "[x.x.x.x]".
1213
7d468ab8
PH
121424. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1215 code that specifies a non-restarting handler (typically sigaction() in
1216 modern systems) in an attempt to fix a rare and obscure crash bug.
1217
121825. Narrowed the window for a race in the daemon that could cause it to ignore
1219 SIGCHLD signals. This is not a major problem, because they are used only to
1220 wake it up if nothing else does.
1221
62c0818f
PH
122226. A malformed maildirsize file could cause Exim to calculate negative values
1223 for the mailbox size or file count. Odd effects could occur as a result.
1224 The maildirsize information is now recalculated if the size or filecount
1225 end up negative.
1226
26034054
PH
122727. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1228 support for a long time. Removed HAVE_SYS_VFS_H.
1229
af66f652
PH
123028. Installed the latest version of exipick from John Jetmore.
1231
90af77f4
PH
123229. In an address list, if the pattern was not a regular expression, an empty
1233 subject address (from a bounce message) matched only if the pattern was an
1234 empty string. Non-empty patterns were not even tested. This was the wrong
1235 because it is perfectly reasonable to use an empty address as part of a
1236 database query. An empty address is now tested by patterns that are
1237 lookups. However, all the other forms of pattern expect the subject to
1238 contain a local part and a domain, and therefore, for them, an empty
1239 address still always fails if the pattern is not itself empty.
1240
d8ef3577
PH
124130. Exim went into a mad DNS loop when attempting to do a callout where the
1242 host was specified on an smtp transport, and looking it up yielded more
1243 than one IP address.
1244
5cb8cbc6
PH
124531. Re-factored the code for checking spool and log partition space into a
1246 function that finds that data and another that does the check. The former
1247 is then used to implement four new variables: $spool_space, $log_space,
1248 $spool_inodes, and $log_inodes.
1249
14702f5b
PH
125032. The RFC2047 encoding function was originally intended for short strings
1251 such as real names; it was not keeping to the 75-character limit for
1252 encoded words that the RFC imposes. It now respects the limit, and
1253 generates multiple encoded words if necessary. To be on the safe side, I
1254 have increased the buffer size for the ${rfc2047: expansion operator from
1255 1024 to 2048 bytes.
1256
063b1e99
PH
125733. It is now permitted to omit both strings after an "if" condition; if the
1258 condition is true, the result is "true". As before, when the second string
1259 is omitted, a false condition yields an empty string. This makes it less
1260 cumbersome to write custom ACL and router conditions.
1261
652e1b65
PH
126234. Failure to deliver a bounce message always caused it to be frozen, even if
1263 there was an errors_to setting on the router. The errors_to setting is now
1264 respected.
1265
6f0c9a4f
PH
126635. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1267 canonical form (fully expanded) before being placed in
1268 $sender_host_address.
1269
33397d19
PH
127036. The table in the code that translates DNS record types into text (T_A to
1271 "A" for instance) was missing entries for NS and CNAME. It is just possible
1272 that this could have caused confusion if both these types were looked up
1273 for the same domain, because the text type is used as part of Exim's
1274 per-process caching. But the chance of anyone hitting this buglet seems
1275 very small.
1276
7bb56e1f
PH
127737. The dnsdb lookup has been extended in a number of ways.
1278
1279 (1) There is a new type, "zns", which walks up the domain tree until it
1280 finds some nameserver records. It should be used with care.
1281
ea3bc19b
PH
1282 (2) There is a new type, "mxh", which is like "mx" except that it returns
1283 just the host names, not the priorities.
1284
1285 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1286 looked up. The behaviour when one of the lookups defers can be
1287 controlled by a keyword.
7bb56e1f 1288
ea3bc19b 1289 (4) It is now possible to specify the separator character for use when
7bb56e1f 1290 multiple records are returned.
33397d19 1291
0bcb2a0e
PH
129238. The dnslists ACL condition has been extended: it is now possible to supply
1293 a list of IP addresses and/or domains to be looked up in a particular DNS
1294 domain.
1295
2ac0e484
PH
129639. Added log_selector=+queue_time_overall.
1297
4e1fde53
PH
129840. When running the queue in the test harness, wait just a tad after forking a
1299 delivery process, to get repeatability of debugging output.
1300
de365ded
PH
130141. Include certificate and key file names in error message when GnuTLS fails
1302 to set them up, because the GnuTLS error message doesn't include the name
1303 of the failing file when there is a problem reading it.
1304
f05da2e8
PH
130542. Allow both -bf and -bF in the same test run.
1306
d6453af2
PH
130743. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1308
f7b63901
PH
130944. The "Exiscan patch" is now merged into the mainline Exim source.
1310
131145. Sometimes the final signoff response after QUIT could fail to get
1312 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1313 < 0 before doing a fflush(). This bug looks as though it goes back to the
1314 introduction of TLS in release 3.20, but "sometimes" must have been rare
1315 because the tests only now provoked it.
1316
a444213a
PH
131746. Reset the locale to "C" after calling embedded Perl, in case it was changed
1318 (this can affect the format of dates).
1319
0ec020ea
PH
132047. exim_tidydb, when checking for the continued existence of a message for
1321 which it has found a message-specific retry record, was not finding
1322 messages that were in split spool directories. Consequently, it was
1323 deleting retry records that should have stayed in existence.
1324
b1206957
PH
132548. Steve fixed some bugs in eximstats.
1326
132749. The SPA authentication driver was not abandoning authentication and moving
1328 on to the next authenticator when an expansion was forced to fail,
1329 contradicting the general specification for all authenticators. Instead it
1330 was generating a temporary error. It now behaves as specified.
1331
26dd5a95
PH
133250. The default ordering of permitted cipher suites for GnuTLS was pessimal
1333 (the order specifies the preference for clients). The order is now AES256,
1334 AES128, 3DES, ARCFOUR128.
1335
343b2385
PH
133651. Small patch to Sieve code - explicitly set From: when generating an
1337 autoreply.
1338
1c5466b9
PH
133952. Exim crashed if a remote delivery caused a very long error message to be
1340 recorded - for instance if somebody sent an entire SpamAssassin report back
1341 as a large number of 550 error lines. This bug was coincidentally fixed by
1342 increasing the size of one of Exim's internal buffers (big_buffer) that
1343 happened as part of the Exiscan merge. However, to be on the safe side, I
1344 have made the code more robust (and fixed the comments that describe what
1345 is going on).
1346
55ee9ee3
PH
134753. Now that there can be additional text after "Completed" in log lines (if
1348 the queue_time_overall log selector is set), a one-byte patch to exigrep
1349 was needed to allow it to recognize "Completed" as not the last thing in
1350 the line.
1351
d38f8232
PH
135254. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1353 patch that reportedly fixes this has been added. I am not expert enough to
1354 create a test for it. This is what the patch creator wrote:
1355
1356 "I found a little strange behaviour of ldap code when working with
1357 Windows 2003 AD Domain, where users was placed in more than one
1358 Organization Units. When I tried to give exim partial DN, the exit code
1359 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1360 But simultaneously result of request was absolutely normal ldap result,
1361 so I produce this patch..."
1362
3295e65b
PH
1363 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1364 REFERENCE, so I have modified the code to exclude the patch when that macro
1365 is not defined.
1366
7102e136
PH
136755. Some experimental protocols are using DNS PTR records for new purposes. The
1368 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1369 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1370 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1371 CAN-2005-0021
7102e136 1372
3ca0ba97
PH
137356. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1374
c2bcbe20
PH
137557. Double the size of the debug message buffer (to 2048) so that more of very
1376 long debug lines gets shown.
1377
18ce445d
PH
137858. The exicyclog utility now does better if the number of log files to keep
1379 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1380
1f5b4c3d
PH
138159. Two changes related to the smtp_active_hostname option:
1382
1383 (1) $smtp_active_hostname is now available as a variable.
1384 (2) The default for smtp_banner uses $smtp_active_hostname instead
1385 of $primary_hostname.
1386
b975ba52
PH
138760. The host_aton() function is supposed to be passed a string that is known
1388 to be a valid IP address. However, in the case of IPv6 addresses, it was
1389 not checking this. This is a hostage to fortune. Exim now panics and dies
1390 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1391 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1392 components; fortuitously, this particular loophole had already been fixed
1393 by change 4.50/55 above.
1394
1395 If there are any other similar loopholes, the new check in host_aton()
1396 itself should stop them being exploited. The report I received stated that
1397 data on the command line could provoke the exploit when Exim was running as
1398 exim, but did not say which command line option was involved. All I could
1399 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1400 running as the user.
ea3a6f44 1401 CAN-2005-0021
85b87bc2
PH
1402
140361. There was a buffer overflow vulnerability in the SPA authentication code
1404 (which came originally from the Samba project). I have added a test to the
1405 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1406 CAN-2005-0022
b975ba52 1407
17ffcae7
PH
140862. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1409 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1410
d95f9fdb
PH
141163. The daemon start-up calls getloadavg() while still root for those OS that
1412 need the first call to be done as root, but it missed one case: when
1413 deliver_queue_load_max is set with deliver_drop_privilege. This is
1414 necessary for the benefit of the queue runner, because there is no re-exec
1415 when deliver_drop_privilege is set.
1416
86b8287f
PH
141764. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1418 This has been fixed.
1419
60dc5e56
PH
142065. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1421 was in use, was not putting the data itself into the right store pool;
1422 consequently, it could be overwritten for a subsequent message in the same
1423 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1424 the caching.)
1425
533244af
PH
142666. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1427
a5a28604
PH
142867. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1429 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1430 the function were treating the return as a boolean value, which happened to
1431 work because 0=false and not-0=true, but is not correct code.
1432
7e634d24
PH
143368. The host_aton() function was not handling scoped IPv6 addresses (those
1434 with, for example, "%eth0" on the end) correctly.
1435
3e11c26b
PH
143669. Fixed some compiler warnings in acl.c for the bitmaps specified with
1437 negated items (that is, ~something) in unsigned ints. Some compilers
1438 apparently mutter when there is no cast.
1439
6729cf78
PH
144070. If an address verification called from an ACL failed, and did not produce a
1441 user-specific message (i.e. there was only a "system" message), nothing was
1442 put in $acl_verify_message. In this situation, it now puts the system
1443 message there.
1444
00f00ca5
PH
144571. Change 4.23/11 added synchronization checking at the start of an SMTP
1446 session; change 4.31/43 added the unwanted input to the log line - except
1447 that it did not do this in the start of session case. It now does.
1448
c9bdd01c
PH
144972. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1450 This is wrong and can cause the other end to generate a synchronization
1451 error if it is another Exim or anything else that does the synchronization
1452 check. A QUIT command is no longer sent after a timeout.
1453
d43194df
PH
145473. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1455 during host lookups.
1456
fe5b5d0b
PH
145774. The defer_ok option of callout verification was not working if it was used
1458 when verifying addresses in header lines, that is, for this case:
1459
1460 verify = header_sender/callout=defer_ok
1461
76a2d7ba
PH
146275. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1463 those file descriptors could be used for SMTP connections. If anything
1464 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1465 could be sent to the SMTP client, causing chaos. The daemon now opens
1466 stdin, stdout, and stderr to /dev/null when it puts itself into the
1467 background.
1468
146976. Arrange for output from Perl's "warn" command to be written to Exim's main
1470 log by default. The user can override this with suitable Perl magic.
1471
04f7d5b9
PH
147277. The use of log_message on a "discard" ACL verb, which is supposed to add to
1473 the log message when discard triggers, was not working for the DATA ACL or
1474 for the non-SMTP ACL.
1475
bc60667e
PH
147678. Error message wording change in sieve.c.
1477
bb6e88ff
PH
147879. If smtp_accept_max_per_host was set, the number of connections could be
1479 restricted to fewer than expected, because the daemon was trying to set up
1480 a new connection before checking whether the processes handling previous
1481 connections had finished. The check for completed processes is now done
1482 earlier. On busy systems, this bug wouldn't be noticed because something
1483 else would have woken the daemon, and it would have reaped the completed
1484 process earlier.
1485
1e70f85b
PH
148680. If a message was submitted locally by a user whose login name contained one
1487 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1488 It caused a spool format error. I have fixed the spool reading code. A
1489 related problem was that the "from" clause in the Received: line became
1490 illegal because of the space(s). It is now covered by ${quote_local_part.
1491
149281. Included the latest eximstats from Steve (adds average sizes to HTML Top
1493 tables).
1494
4e01f9d6
PH
149582. Updated OS/Makefile-AIX as per message from Mike Meredith.
1496
1ee1cef2
PH
149783. Patch from Sieve maintainer to fix unterminated string problem in
1498 "vacation" handling.
1499
6e2b4ccc
PH
150084. Some minor changes to the Linux configuration files to help with other
1501 OS variants using glibc.
1502
8e669ac1
PH
150385. One more patch for Sieve to update vacation handling to latest spec.
1504
495ae4b0 1505
bbe902f0
PH
1506----------------------------------------------------
1507See the note above about the 4.44 and 4.50 releases.
1508----------------------------------------------------
1509
1510
1511Exim version 4.44
1512-----------------
1513
1514 1. Change 4.43/35 introduced a bug that caused file counts to be
1515 incorrectly computed when quota_filecount was set in an appendfile
1516 transport
1517
1518 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1519 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1520 place.
1521
1522 3. Give more explanation in the error message when the command for a transport
1523 filter fails to execute.
1524
1525 4. There are several places where Exim runs a non-Exim command in a
1526 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1527 was being done only for the command run by the queryprogram router. It is
1528 now done for all such subprocesses. The other cases are: ${run, transport
1529 filters, and the commands run by the lmtp and pipe transports.
1530
1531 5. Some older OS have a limit of 256 on the maximum number of file
1532 descriptors. Exim was using setrlimit() to set 1000 as a large value
1533 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1534 systems. I've change it so that if it can't get 1000, it tries for 256.
1535
1536 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1537 was an oversight, and furthermore, ever since the addition of extra
1538 controls (e.g. 4.43/32), the checks on when to allow different forms of
1539 "control" were broken. There should now be diagnostics for all cases when a
1540 control that does not make sense is encountered.
1541
1542 7. $recipients is now available in the predata ACL (oversight).
1543
1544 8. Tidy the search cache before the fork to do a delivery from a message
1545 received from the command line. Otherwise the child will trigger a lookup
1546 failure and thereby defer the delivery if it tries to use (for example) a
1547 cached ldap connection that the parent has called unbind on.
1548
1549 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1550 of $address_data from the recipient verification was clobbered by the
1551 sender verification.
1552
155310. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1554 was its contents. (It was OK if the option was not defined at all.)
1555
155611. A "Completed" log line is now written for messages that are removed from
1557 the spool by the -Mrm option.
1558
155912. $host_address is now set to the target address during the checking of
1560 ignore_target_hosts.
1561
156213. When checking ignore_target_hosts for an ipliteral router, no host name was
1563 being passed; this would have caused $sender_host_name to have been used if
1564 matching the list had actually called for a host name (not very likely,
1565 since this list is usually IP addresses). A host name is now passed as
1566 "[x.x.x.x]".
1567
156814. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1569 code that specifies a non-restarting handler (typically sigaction() in
1570 modern systems) in an attempt to fix a rare and obscure crash bug.
1571
157215. Narrowed the window for a race in the daemon that could cause it to ignore
1573 SIGCHLD signals. This is not a major problem, because they are used only to
1574 wake it up if nothing else does.
1575
157616. A malformed maildirsize file could cause Exim to calculate negative values
1577 for the mailbox size or file count. Odd effects could occur as a result.
1578 The maildirsize information is now recalculated if the size or filecount
1579 end up negative.
1580
158117. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1582 support for a long time. Removed HAVE_SYS_VFS_H.
1583
ea3a6f44 158418. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1585
158619. Allow an empty sender to be matched against a lookup in an address list.
1587 Previously the only cases considered were a regular expression, or an
1588 empty pattern.
1589
159020. Exim went into a mad DNS lookup loop when doing a callout where the
1591 host was specified on the transport, if the DNS lookup yielded more than
1592 one IP address.
1593
ea3a6f44
NM
159421. The RFC2047 encoding function was originally intended for short strings
1595 such as real names; it was not keeping to the 75-character limit for
1596 encoded words that the RFC imposes. It now respects the limit, and
1597 generates multiple encoded words if necessary. To be on the safe side, I
1598 have increased the buffer size for the ${rfc2047: expansion operator from
1599 1024 to 2048 bytes.
bbe902f0 1600
ea3a6f44
NM
160122. Failure to deliver a bounce message always caused it to be frozen, even if
1602 there was an errors_to setting on the router. The errors_to setting is now
1603 respected.
bbe902f0
PH
1604
160523. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1606 canonical form (fully expanded) before being placed in
1607 $sender_host_address.
1608
160924. Updated eximstats to version 1.33
1610
ea3a6f44
NM
161125. Include certificate and key file names in error message when GnuTLS fails
1612 to set them up, because the GnuTLS error message doesn't include the name
1613 of the failing file when there is a problem reading it.
bbe902f0
PH
1614
161526. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1616 As per change 25.
bbe902f0 1617
ea3a6f44
NM
161827. Reset the locale to "C" after calling embedded Perl, in case it was changed
1619 (this can affect the format of dates).
bbe902f0 1620
ea3a6f44
NM
162128. exim_tidydb, when checking for the continued existence of a message for
1622 which it has found a message-specific retry record, was not finding
1623 messages that were in split spool directories. Consequently, it was
1624 deleting retry records that should have stayed in existence.
bbe902f0
PH
1625
162629. eximstats updated to version 1.35
1627 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1628 1.35 - bugfix such that pie charts by volume are generated correctly
1629
ea3a6f44
NM
163030. The SPA authentication driver was not abandoning authentication and moving
1631 on to the next authenticator when an expansion was forced to fail,
1632 contradicting the general specification for all authenticators. Instead it
1633 was generating a temporary error. It now behaves as specified.
bbe902f0 1634
ea3a6f44
NM
163531. The default ordering of permitted cipher suites for GnuTLS was pessimal
1636 (the order specifies the preference for clients). The order is now AES256,
1637 AES128, 3DES, ARCFOUR128.
bbe902f0 1638
ea3a6f44
NM
163931. Small patch to Sieve code - explicitly set From: when generating an
1640 autoreply.
bbe902f0 1641
ea3a6f44
NM
164232. Exim crashed if a remote delivery caused a very long error message to be
1643 recorded - for instance if somebody sent an entire SpamAssassin report back
1644 as a large number of 550 error lines. This bug was coincidentally fixed by
1645 increasing the size of one of Exim's internal buffers (big_buffer) that
1646 happened as part of the Exiscan merge. However, to be on the safe side, I
1647 have made the code more robust (and fixed the comments that describe what
1648 is going on).
bbe902f0 1649
ea3a6f44
NM
165033. Some experimental protocols are using DNS PTR records for new purposes. The
1651 keys for these records are domain names, not reversed IP addresses. The
1652 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1653 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1654 CAN-2005-0021
1655
ea3a6f44
NM
165634. The host_aton() function is supposed to be passed a string that is known
1657 to be a valid IP address. However, in the case of IPv6 addresses, it was
1658 not checking this. This is a hostage to fortune. Exim now panics and dies
1659 if the condition is not met. A case was found where this could be provoked
1660 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1661 components; fortuitously, this particular loophole had already been fixed
1662 by change 4.50/55 or 4.44/33 above.
1663
1664 If there are any other similar loopholes, the new check in host_aton()
1665 itself should stop them being exploited. The report I received stated that
1666 data on the command line could provoke the exploit when Exim was running as
1667 exim, but did not say which command line option was involved. All I could
1668 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1669 running as the user.
bbe902f0
PH
1670 CAN-2005-0021
1671
ea3a6f44
NM
167235. There was a buffer overflow vulnerability in the SPA authentication code
1673 (which came originally from the Samba project). I have added a test to the
1674 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1675 CAN-2005-0022
1676
ea3a6f44
NM
167736. The daemon start-up calls getloadavg() while still root for those OS that
1678 need the first call to be done as root, but it missed one case: when
1679 deliver_queue_load_max is set with deliver_drop_privilege. This is
1680 necessary for the benefit of the queue runner, because there is no re-exec
1681 when deliver_drop_privilege is set.
bbe902f0 1682
ea3a6f44
NM
168337. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1684 was in use, was not putting the data itself into the right store pool;
1685 consequently, it could be overwritten for a subsequent message in the same
1686 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1687 the caching.)
bbe902f0 1688
ea3a6f44
NM
168938. Sometimes the final signoff response after QUIT could fail to get
1690 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1691 < 0 before doing a fflush(). This bug looks as though it goes back to the
1692 introduction of TLS in release 3.20, but "sometimes" must have been rare
1693 because the tests only now provoked it.
bbe902f0
PH
1694
1695
495ae4b0
PH
1696Exim version 4.43
1697-----------------
1698
1699 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1700 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1701 Now it returns an integer. A number of calls were still expecting a T/F
1702 return. Fortuitously, in all cases, the tests worked in OK situations,
1703 which is the norm. However, things would have gone wrong on any write
1704 failures on the smtp file descriptor. This function is used when sending
1705 messages over SMTP and also when doing verify callouts.
1706
1707 2. When Exim is called to do synchronous delivery of a locally submitted
1708 message (the -odf or -odi options), it no longer closes stderr before doing
1709 the delivery.
1710
1711 3. Implemented the mua_wrapper option.
1712
1713 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1714
1715 5. Implemented the functions header_remove(), header_testname(),
1716 header_add_at_position(), and receive_remove_recipient(), and exported them
1717 to local_scan().
1718
1719 6. If an ACL "warn" statement specified the addition of headers, Exim already
1720 inserted X-ACL-Warn: at the start if there was no header name. However, it
1721 was not making this test for the second and subsequent header lines if
1722 there were newlines in the string. This meant that an invalid header could
1723 be inserted if Exim was badly configured.
1724
1725 7. Allow an ACL "warn" statement to add header lines at the start or after all
1726 the Received: headers, as well as at the end.
1727
1728 8. Added the rcpt_4xx retry error code.
1729
1730 9. Added postmaster_mailfrom=xxx to callout verification option.
1731
173210. Added mailfrom=xxxx to the callout verification option, for verify=
1733 header_sender only.
1734
173511. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1736 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1737
173812. Inserted some casts to stop certain compilers warning when using pointer
1739 differences as field lengths or precisions in printf-type calls (mostly
1740 affecting debugging statements).
1741
174213. Added optional readline() support for -be (dynamically loaded).
1743
174414. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1745 same clock tick as a message's arrival, so that its received time was the
1746 same as the "first fail" time on the retry record, and that message
1747 remained on the queue past the ultimate address timeout, every queue runner
1748 would try a delivery (because it was past the ultimate address timeout) but
1749 after another failure, the ultimate address timeout, which should have then
1750 bounced the address, did not kick in. This was a "< instead of <=" error;
1751 in most cases the first failure would have been in the next clock tick
1752 after the received time, and all would be well.
1753
175415. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1755 being recognized when the domain list was tested by the match_domain
1756 condition in an expansion string.
1757
175816. Added the ${str2b64: operator.
1759
176017. Exim was always calling setrlimit() to set a large limit for the number of
1761 processes, without checking whether the existing limit was already
1762 adequate. (It did check for the limit on file descriptors.) Furthermore,
1763 errors from getrlimit() and setrlimit() were being ignored. Now they are
1764 logged to the main and panic logs, but Exim does carry on, to try to do its
1765 job under whatever limits there are.
1766
176718. Imported PCRE 5.0.
1768
176919. Trivial typo in log message " temporarily refused connection" (the leading
1770 space).
1771
177220. If the log selector return_path_on_delivery was set and an address was
1773 redirected to /dev/null, the delivery process crashed because it assumed
1774 that a return path would always be set for a "successful" delivery. In this
1775 case, the whole delivery is bypassed as an optimization, and therefore no
1776 return path is set.
1777
177821. Internal re-arrangement: the function for sending a challenge and reading
1779 a response while authentication was assuming a zero-terminated challenge
1780 string. It's now changed to take a pointer and a length, to allow for
1781 binary data in such strings.
1782
178322. Added the cyrus_sasl authenticator (code supplied by MBM).
1784
178523. Exim was not respecting finduser_retries when seeking the login of the
1786 uid under which it was called; it was always trying 10 times. (The default
1787 setting of finduser_retries is zero.) Also, it was sleeping after the final
1788 failure, which is pointless.
1789
179024. Implemented tls_on_connect_ports.
1791
179225. Implemented acl_smtp_predata.
1793
179426. If the domain in control=submission is set empty, Exim assumes that the
1795 authenticated id is a complete email address when it generates From: or
1796 Sender: header lines.
1797
179827. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1799 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1800 chown and chgrp in /bin and hostname in /usr/bin.
1801
180228. Exim was keeping the "process log" file open after each use, just as it
1803 does for the main log. This opens the possibility of it remaining open for
1804 long periods when the USR1 signal hits a daemon. Occasional processlog
1805 errors were reported, that could have been caused by this. Anyway, it seems
1806 much more sensible not to leave this file open at all, so that is what now
1807 happens.
1808
180929. The long-running daemon process does not normally write to the log once it
1810 has entered its main loop, and it closes the log before doing so. This is
1811 so that log files can straightforwardly be renamed and moved. However,
1812 there are a couple of unusual error situations where the daemon does write
1813 log entries, and I had neglected to close the log afterwards.
1814
181530. The text of an SMTP error response that was received during a remote
1816 delivery was being truncated at 512 bytes. This is too short for some of
1817 the long messages that one sometimes sees. I've increased the limit to
1818 1024.
1819
182031. It is now possible to make retry rules that apply only when a message has a
1821 specific sender, in particular, an empty sender.
1822
182332. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1824 it possible to be selective about when SMTP synchronization is enforced.
1825
182633. Added "control = caseful_local_part" and "control = "caselower_local_part".
1827
182832. Implemented hosts_connection_nolog.
1829
183033. Added an ACL for QUIT.
1831
183234. Setting "delay_warning=" to disable warnings was not working; it gave a
1833 syntax error.
1834
183535. Added mailbox_size and mailbox_filecount to appendfile.
1836
183736. Added control = no_multiline_responses to ACLs.
1838
183937. There was a bug in the logic of the code that waits for the clock to tick
1840 in the case where the clock went backwards by a substantial amount such
1841 that the microsecond fraction of "now" was more than the microsecond
1842 fraction of "then" (but the whole seconds number was less).
1843
184438. Added support for the libradius Radius client library this is found on
1845 FreeBSD (previously only the radiusclient library was supported).
1846
1847
1848Exim version 4.42
1849-----------------
1850
1851 1. When certain lookups returned multiple values in the form name=value, the
1852 quoting of the values was not always being done properly. Specifically:
1853 (a) If the value started with a double quote, but contained no whitespace,
1854 it was not quoted.
1855 (b) If the value contained whitespace other than a space character (i.e.
1856 tabs or newlines or carriage returns) it was not quoted.
1857 This fix has been applied to the mysql and pgsql lookups by writing a
1858 separate quoting function and calling it from the lookup code. The fix
1859 should probably also be applied to nisplus, ibase and oracle lookups, but
1860 since I cannot test any of those, I have not disturbed their existing code.
1861
1862 2. A hit in the callout cache for a specific address caused a log line with no
1863 reason for rejecting RCPT. Now it says "Previous (cached) callout
1864 verification failure".
1865
1866 3. There was an off-by-one bug in the queryprogram router. An over-long
1867 return line was truncated at 256 instead of 255 characters, thereby
1868 overflowing its buffer with the terminating zero. As well as fixing this, I
1869 have increased the buffer size to 1024 (and made a note to document this).
1870
1871 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1872 when Exim is waiting for an SMTP response from a remote server, Exim
1873 restarts its select() call on the socket, thereby resetting its timeout.
1874 This is not a problem when such interrupts are rare. Somebody set up a cron
1875 job to run exiwhat every 2 minutes, which is less than the normal select()
1876 timeout (5 or 10 minutes). This meant that the select() timeout never
1877 kicked in because it was always reset. I have fixed this by comparing the
1878 time when an interrupt arrives with the time at the start of the first call
1879 to select(). If more time than the timeout has elapsed, the interrupt is
1880 treated as a timeout.
1881
1882 5. Some internal re-factoring in preparation for the addition of Sieve
1883 extensions (by MH). In particular, the "personal" test is moved to a
1884 separate function, and given an option for scanning Cc: and Bcc: (which is
1885 not set for Exim filters).
1886
1887 6. When Exim created an email address using the login of the caller as the
1888 local part (e.g. when creating a From: or Sender: header line), it was not
1889 quoting the local part when it contained special characters such as @.
1890
1891 7. Installed new OpenBSD configuration files.
1892
1893 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1894 try to make them clearer.
1895
1896 9. Callout options, other than the timeout value, were being ignored when
1897 verifying sender addresses in header lines. For example, when using
1898
1899 verify = header_sender/callout=no_cache
1900
1901 the cache was (incorrectly) being used.
1902
190310. Added a missing instance of ${EXE} to the exim_install script; this affects
1904 only the Cygwin environment.
1905
190611. When return_path_on_delivery was set as a log selector, if different remote
1907 addresses in the same message used different return paths and parallel
1908 remote delivery occurred, the wrong values would sometimes be logged.
1909 (Whenever a remote delivery process finished, the return path value from
1910 the most recently started remote delivery process was logged.)
1911
191212. RFC 3848 specifies standard names for the "with" phrase in Received: header
1913 lines when AUTH and/or TLS are in use. This is the "received protocol"
1914 field. Exim used to use "asmtp" for authenticated SMTP, without any
1915 indication (in the protocol name) for TLS use. Now it follows the RFC and
1916 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1917 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1918 names appear in log lines as well as in Received: header lines.
1919
192013. Installed MH's patches for Sieve to add the "copy" and "vacation"
1921 extensions, and comparison tests, and to fix some bugs.
1922
192314. Changes to the "personal" filter test:
1924
1925 (1) The test was buggy in that it was just doing the equivalent of
1926 "contains" tests on header lines. For example, if a user's address was
1927 anne@some.where, the "personal" test would incorrectly be true for
1928
1929 To: susanne@some.where
1930
1931 This test is now done by extracting each address from the header in turn,
1932 and checking the entire address. Other tests that are part of "personal"
1933 are now done using regular expressions (for example, to check local parts
1934 of addresses in From: header lines).
1935
1936 (2) The list of non-personal local parts in From: addresses has been
1937 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1938 taken from the Sieve specification recommendations.
1939
1940 (3) If the message contains any header line starting with "List-" it is
1941 treated as non-personal.
1942
1943 (4) The test for "circular" in the Subject: header line has been removed
1944 because it now seems ill-conceived.
1945
194615. Minor typos in src/EDITME comments corrected.
1947
194816. Installed latest exipick from John Jetmore.
1949
195017. If headers_add on a router specified a text string that was too long for
1951 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1952 of string_sprintf() is now avoided.
1953
195418. $message_body_size was not set (it was always zero) when running the DATA
1955 ACL and the local_scan() function.
1956
195719. For the "mail" command in an Exim filter, no default was being set for
1958 the once_repeat time, causing a random time value to be used if "once" was
1959 specified. (If the value happened to be <= 0, no repeat happened.) The
1960 default is now 0s, meaning "never repeat". The "vacation" command was OK
1961 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1962 (I found it when inspecting the code).
1963
196420. There is now an overall timeout for performing a callout verification. It
1965 defaults to 4 times the callout timeout, which applies to individual SMTP
1966 commands during the callout. The overall timeout applies when there is more
1967 than one host that can be tried. The timeout is checked before trying the
1968 next host. This prevents very long delays if there are a large number of
1969 hosts and all are timing out (e.g. when the network connections are timing
1970 out). The value of the overall timeout can be changed by specifying an
1971 additional sub-option for "callout", called "maxwait". For example:
1972
1973 verify = sender/callout=5s,maxwait=20s
1974
197521. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1976 to the end before writing, but this should make it even safer).
1977
197822. Exim was forgetting that it had advertised PIPELINING for the second and
1979 subsequent messages on an SMTP connection. It was also not resetting its
1980 memory on STARTTLS and an internal HELO.
1981
198223. When Exim logs an SMTP synchronization error within a session, it now
1983 records whether PIPELINING has been advertised or not.
1984
198524. Added 3 instances of "(long int)" casts to time_t variables that were being
1986 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1987 rather than long int.
1988
198925. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1990
199126. Added the never_mail option to autoreply.
1992
1993
1994Exim version 4.41
1995-----------------
1996
1997 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1998 crash if the getsockname() call failed; this can happen if a connection is
1999 closed very soon after it is established. The problem was simply in the
2000 order in which certain operations were done, causing Exim to try to write
2001 to the SMTP stream before it had set up the file descriptor. The bug has
2002 been fixed by making things happen in the correct order.
2003
2004
2005Exim version 4.40
2006-----------------
2007
2008 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2009 before the connection was closed, thus losing the rejection response.
2010
2011 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2012 some early Solaris releases, but causes trouble in current releases where
2013 socklen_t is defined.
2014
2015 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2016 always exist.
2017
2018 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2019 configured.
2020
2021 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2022 that releasing the top of it at the end releases what was used for sub-
2023 expansions (unless the block got too big). However, discard this block if
2024 the first thing is a variable or header, so that we can use its block when
2025 it is dynamic (useful for very large $message_headers, for example).
2026
2027 6. Lookups now cache *every* query, not just the most recent. A new, separate
2028 store pool is used for this. It can be recovered when all lookup caches are
2029 flushed. Lookups now release memory at the end of their result strings.
2030 This has involved some general refactoring of the lookup sources.
2031
2032 7. Some code has been added to the store_xxx() functions to reduce the amount
2033 of flapping under certain conditions.
2034
2035 8. log_incoming_interface used to affect only the <= reception log lines. Now
2036 it causes the local interface and port to be added to several more SMTP log
2037 lines, for example "SMTP connection from", and rejection lines.
2038
2039 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2040
204110. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2042
204311. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2044 could be overwritten at the end of the current message (or the start of a
2045 new message if it was set in a HELO ACL). The value is now preserved for
2046 the duration of the SMTP connection.
2047
204812. If a transport had a headers_rewrite setting, and a matching header line
2049 contained an unqualified address, that address was qualified, even if it
2050 did not match any rewriting rules. The underlying bug was that the values
2051 of the flags that permit the existence of unqualified sender and recipient
2052 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2053 non-local messages, and by -bnq for local messages) were not being
2054 preserved with the message after it was received.
2055
205613. When Exim was logging an SMTP synchronization error, it could sometimes log
2057 "next input=" as part of the text comprising the host identity instead of
2058 the correct text. The code was using the same buffer for two different
2059 strings. However, depending on which order the printing function evaluated
2060 its arguments, the bug did not always show up. Under Linux, for example, my
2061 test suite worked just fine.
2062
206314. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2064 doesn't work with some older versions of Perl. It has been changed to "my",
2065 which in any case is probably the better facility to use.
2066
206715. A really picky compiler found some instances of statements for creating
2068 error messages that either had too many or two few arguments for the format
2069 string.
2070
207116. The size of the buffer for calls to the DNS resolver has been increased
2072 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2073 for addresses that have a lot of PTR records. This alleviates a problem; it
2074 does not fully solve it.
2075
207617. A dnsdb lookup for PTR records that receives more data than will fit in the
2077 buffer now truncates the list and logs the incident, which is the same
2078 action as happens when Exim is looking up a host name and its aliases.
2079 Previously in this situation something unpredictable would happen;
2080 sometimes it was "internal error: store_reset failed".
2081
208218. If a server dropped the connection unexpectedly when an Exim client was
2083 using GnuTLS and trying to read a response, the client delivery process
2084 crashed while trying to generate an error log message.
2085
208619. If a "warn" verb in an ACL added multiple headers to a message in a single
2087 string, for example:
2088
2089 warn message = H1: something\nH2: something
2090
2091 the text was added as a single header line from Exim's point of view
2092 though it ended up OK in the delivered message. However, searching for the
2093 second and subsequent header lines using $h_h2: did not work. This has been
2094 fixed. Similarly, if a system filter added multiple headers in this way,
2095 the routers could not see them.
2096
209720. Expanded the error message when iplsearch is called with an invalid key to
2098 suggest using net-iplsearch in a host list.
2099
210021. When running tests using -bh, any delays imposed by "delay" modifiers in
2101 ACLs are no longer actually imposed (and a message to that effect is
2102 output).
2103
210422. If a "gecos" field in a passwd entry contained escaped characters, in
2105 particular, if it contained a \" sequence, Exim got it wrong when building
2106 a From: or a Sender: header from that name. A second bug also caused
2107 incorrect handling when an unquoted " was present following a character
2108 that needed quoting.
2109
211023. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2111 was not being matched caselessly.
2112
211324. Arranged for all hyphens in the exim.8 source to be escaped with
2114 backslashes.
2115
211625. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2117 properly. Recipient callout cache records were still being keyed to include
2118 the sender, even when use_sender was set false. This led to far more
2119 callouts that were necessary. The sender is no longer included in the key
2120 when use_sender is false.
2121
212226. Added "control = submission" modifier to ACLs.
2123
212427. Added the ${base62d: operator to decode base 62 numbers.
2125
212628. dnsdb lookups can now access SRV records.
2127
212829. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2129 the configuration file.
2130
213130. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2132 (-v) mode. This makes the output for a verbose queue run more intelligible.
2133
213431. Added a use_postmaster feature to recipient callouts.
2135
213632. Added the $body_zerocount variable, containing the number of binary zero
2137 bytes in the message body.
2138
213933. The time of last modification of the "new" subdirectory is now used as the
2140 "mailbox time last read" when there is a quota error for a maildir
2141 delivery.
2142
214334. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2144
214535. Added +ignore_unknown as a special item in host lists.
2146
214736. Code for decoding IPv6 addresses in host lists is now included, even if
2148 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2149 address was recognized as an IP address, but was then not correctly decoded
2150 into binary, causing unexpected and incorrect effects when compared with
2151 another IP address.
2152
2153
2154Exim version 4.34
2155-----------------
2156
2157 1. Very minor rewording of debugging text in manualroute to say "list of
2158 hosts" instead of "hostlist".
2159
2160 2. If verify=header_syntax was set, and a header line with an unqualified
2161 address (no domain) and a large number of spaces between the end of the
2162 name and the colon was received, the reception process suffered a buffer
2163 overflow, and (when I tested it) crashed. This was caused by some obsolete
2164 code that should have been removed. The fix is to remove it!
2165
2166 3. When running in the test harness, delay a bit after writing a bounce
2167 message to get a bit more predictability in the log output.
2168
2169 4. Added a call to search_tidyup() just before forking a reception process. In
2170 theory, someone could use a lookup in the expansion of smtp_accept_max_
2171 per_host which, without the tidyup, could leave open a database connection.
2172
2173 5. Added the variables $recipient_data and $sender_data which get set from a
2174 lookup success in an ACL "recipients" or "senders" condition, or a router
2175 "senders" option, similar to $domain_data and $local_part_data.
2176
2177 6. Moved the writing of debug_print from before to after the "senders" test
2178 for routers.
2179
2180 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2181 problems for message scanning, either using a data ACL, or using
2182 local_scan() because the Received: header was not generated till after they
2183 were called (in order to set the time as the time of reception completion).
2184 I have revised the way this works. The header is now generated after the
2185 body is received, but before the ACL or local_scan() are called. After they
2186 are run, the timestamp in the header is updated.
2187
2188
2189Exim version 4.33
2190-----------------
2191
2192 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2193 before starting a queue runner without re-exec. This happened only when
2194 deliver_drop_privilege was set or when the Exim user was set to root. The
2195 effect of the bug was that timeouts during subsequent deliveries caused
2196 crashes instead of being properly handled. The handler is now left at its
2197 default (and expected) setting.
2198
2199 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2200 message, again when deliver_drop_privilege is set or Exim is run as root.
2201 The bug described in (1) was not present in this case, but the tidying up
2202 of the other signals was missing. I have made the two cases consistent.
2203
2204 3. The ignore_target_hosts setting on a manualroute router was being ignored
2205 for hosts that were looked up using the /MX notation.
2206
2207 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2208 in domain lists.
2209
2210 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2211 operated on the sender address. After changing the $sender_address to <>
2212 for the sender address verify, Exim was re-instated it as the original
2213 (before rewriting) address, but remembering that it had rewritten it, so it
2214 wasn't rewriting it again. This bug also had the effect of breaking the
2215 sender address verification caching when the sender address was rewritten.
2216
2217 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2218 This has been changed so that if the ip literal address matches
2219 ignore_target_hosts, the router declines.
2220
2221 7. Added expansion conditions match_domain, match_address, and match_local_
2222 part (NOT match_host).
2223
2224 8. The placeholder for the Received: header didn't have a length field set.
2225
2226 9. Added code to Exim itself and to exim_lock to test for a specific race
2227 condition that could lead to file corruption when using MBX delivery. The
2228 issue is with the lockfile that is created in /tmp. If this file is removed
2229 after a process has opened it but before that process has acquired a lock,
2230 there is the potential for a second process to recreate the file and also
2231 acquire a lock. This could lead to two Exim processes writing to the file
2232 at the same time. The added code performs the same test as UW imapd; it
2233 checks after acquiring the lock that its file descriptor still refers to
2234 the same named file.
2235
223610. The buffer for building added header lines was of fixed size, 8192 bytes.
2237 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2238 when Exim is built.
2239
224011. Added the smtp_active_hostname option. If used, this will typically be made
2241 to depend on the incoming interface address. Because $interface_address is
2242 not set up until the daemon has forked a reception process, error responses
2243 that can happen earlier (such as "too many connections") no longer contain
2244 a host name.
2245
224612. If an expansion in a condition on a "warn" statement fails because a lookup
2247 defers, the "warn" statement is abandoned, and the next ACL statement is
2248 processed. Previously this caused the whole ACL to be aborted.
2249
225013. Added the iplsearch lookup type.
2251
225214. Added ident_timeout as a log selector.
2253
225415. Added tls_certificate_verified as a log selector.
2255
225616. Added a global option tls_require_ciphers (compare the smtp transport
2257 option of the same name). This controls incoming TLS connections.
2258
225917. I finally figured out how to make tls_require_ciphers do a similar thing
2260 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2261 before starting the TLS session.
2262
226318. Tabs are now shown as \t in -bP output.
2264
226519. If the log selector return_path_on_delivery was set, Exim crashed when
2266 bouncing a message because it had too many Received: header lines.
2267
226820. If two routers both had headers_remove settings, and the first one included
2269 a superfluous trailing colon, the final name in the first list and the
2270 first name in the second list were incorrectly joined into one item (with a
2271 colon in the middle).
2272
2273
2274Exim version 4.32
2275-----------------
2276
2277 1. Added -C and -D options to the exinext utility, mainly to make it easier
2278 to include in the automated testing, but these could be helpful when
2279 multiple configurations are in use.
2280
2281 2. The exinext utility was not formatting the output nicely when there was
2282 an alternate port involved in the retry record key, nor when there was a
2283 message id as well (for retries that were specific to a specific message
2284 and a specific host). It was also confused by IPv6 addresses, because of
2285 the additional colons they contain. I have fixed the IPv4 problem, and
2286 patched it up to do a reasonable job for IPv6.
2287
2288 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2289 delivery, the log line now contains "pipelined" if PIPELINING was used.
2290
2291 4. An SMTP transport process used to panic and die if the bind() call to set
2292 an explicit outgoing interface failed. This has been changed; it is now
2293 treated in the same way as a connect() failure.
2294
2295 5. A reference to $sender_host_name in the part of a conditional expansion
2296 that was being skipped was still causing a DNS lookup. This no longer
2297 occurs.
2298
2299 6. The def: expansion condition was not recognizing references to header lines
2300 that used bh_ and bheader_.
2301
2302 7. Added the _cache feature to named lists.
2303
2304 8. The code for checking quota_filecount in the appendfile transport was
2305 allowing one more file than it should have been.
2306
2307 9. For compatibility with Sendmail, the command line option
2308
2309 -prval:sval
2310
2311 is equivalent to
2312
2313 -oMr rval -oMs sval
2314
2315 and sets the incoming protocol and host name (for trusted callers). The
2316 host name and its colon can be omitted when only the protocol is to be set.
2317 Note the Exim already has two private options, -pd and -ps, that refer to
2318 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2319 "s", but I don't think that's a major issue.
2320
232110. A number of refactoring changes to the code, none of which should affect
2322 Exim's behaviour:
2323
2324 (a) The number of logging options was getting close to filling up the
2325 32-bit word that was used as a bit map. I have split them into two classes:
2326 those that are passed in the argument to log_write(), and those that are
2327 only ever tested independently outside of that function. These are now in
2328 separate 32-bit words, so there is plenty of room for expansion again.
2329 There is no change in the user interface or the logging behaviour.
2330
2331 (b) When building, for example, log lines, the code previously used a
2332 macro that called string_cat() twice, in order to add two strings. This is
2333 not really sufficiently general. Furthermore, there was one instance where
2334 it was actually wrong because one of the argument was used twice, and in
2335 one call a function was used. (As it happened, calling the function twice
2336 did not affect the overall behaviour.) The macro has been replaced by a
2337 function that can join an arbitrary number of extra strings onto a growing
2338 string.
2339
2340 (c) The code for expansion conditions now uses a table and a binary chop
2341 instead of a serial search (which was left over from when there were very
2342 few conditions). Also, it now recognizes conditions like "pam" even when
2343 the relevant support is not compiled in: a suitably worded error message is
2344 given if an attempt is made to use such a condition.
2345
234611. Added ${time_interval:xxxxx}.
2347
234812. A bug was causing one of the ddress fields not to be passed back correctly
2349 from remote delivery subprocesses. The field in question was not being
2350 subsequently used, so this caused to problems in practice.
2351
235213. Added new log selectors queue_time and deliver_time.
2353
235414. Might have fixed a bug in maildirsizefile handling that threw up
2355 "unexpected character" debug warnings, and recalculated the data
2356 unnecessarily. In any case, I expanded the warning message to give more
2357 information.
2358
235915. Added the message "Restricted characters in address" to the statements in
2360 the default ACL that block characters like @ and % in local parts.
2361
236216. Change 71 for release 4.31 proved to be much less benign that I imagined.
2363 Three changes have been made:
2364
2365 (a) There was a serious bug; a negative response to MAIL caused the whole
2366 recipient domain to be cached as invalid, thereby blocking all messages
2367 to all local parts at the same domain, from all senders. This bug has
2368 been fixed. The domain is no longer cached after a negative response to
2369 MAIL if the sender used is not empty.
2370
2371 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2372 been restored.
2373
2374 (c) A new callout option, "use_sender" has been added for people who want
2375 the modified behaviour.
2376
2377
2378Exim version 4.31
2379-----------------
2380
2381 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2382 Larry Rosenman.
2383
2384 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2385 indeed breaks things for older releases.
2386
2387 3. Added additional logging to the case where there is a problem reading data
2388 from a filter that is running in a subprocess using a pipe, in order to
2389 try to track down a specific problem.
2390
2391 4. Testing facility fudge: when running in the test harness and attempting
2392 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2393 getting "No route to host". Convert this to a timeout.
2394
2395 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2396 warning.
2397
2398 6. Some OS don't have socklen_t but use size_t instead. This affects the
2399 fifth argument of getsockopt() amongst other things. This is now
2400 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2401 can be set for individual OS. I have set it for SunOS5, OSF1, and
2402 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2403 some earlier ones do not.
2404
2405 7. Change 4.30/15 was not doing the test caselessly.
2406
2407 8. The standard form for an IPv6 address literal was being rejected by address
2408 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2409 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2410 this, as well as the form without the "IPv6" on the front (but only when
2411 address literals are enabled, of course).
2412
2413 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2414
241510. Exim crashed if a message with an empty sender address specified by -f
2416 encountered a router with an errors_to setting. This could be provoked only
2417 by a command such as
2418
2419 exim -f "" ...
2420
2421 where an empty string was supplied; "<>" did not hit this bug.
2422
242311. Installed PCRE release 4.5.
2424
242512. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2426 remained set. It is now erased.
2427
242813. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2429 times from message ids (which are base 36 rather than the normal 62).
2430
243114. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2432 were being counted as actual protocol errors, and logged if the log
2433 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2434 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2435 and DATA following a set of rejected RCPTs do not count as protocol errors.
2436 In other words, Exim assumes they were pipelined, though this may not
2437 actually be the case. Of course, in all cases the client gets an
2438 appropriate error code.
2439
244015. If a lookup fails in an ACL condition, a message about the failure may
2441 be available; it is used if testing the ACL cannot continue, because most
2442 such messages specify what the cause of the deferral is. However, some
2443 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2444 that caused an old message to be retained and used if a later statement
2445 caused a defer, replacing the real cause of the deferral.
2446
244716. If an IP address had so many PTR records that the DNS lookup buffer
2448 was not large enough to hold them, Exim could crash while trying to process
2449 the truncated data. It now detects and logs this case.
2450
245117. Further to 4.21/58, another change has been made: if (and only if) the
2452 first line of a message (the first header line) ends with CRLF, a bare LF
2453 in a subsequent header line has a space inserted after it, so as not to
2454 terminate the header.
2455
245618. Refactoring: tidied an ugly bit of code in appendfile that copied data
2457 unnecessarily, used atoi() instead of strtol(), and didn't check the
2458 termination when getting file sizes from file names by regex.
2459
246019. Completely re-implemented the support for maildirsize files, in the light
2461 of a number of problems with the previous contributed implementation
2462 (4.30/29). In particular:
2463
2464 . If the quota is zero, the maildirsize file is maintained, but no quota is
2465 imposed.
2466
2467 . If the maildir directory does not exist, it is created before any attempt
2468 to write a maildirsize file.
2469
2470 . The quota value in the file is just a cache; if the quota is changed in
2471 the transport, the new value overrides.
2472
2473 . A regular expression is available for excluding directories from the
2474 count.
2475
247620. The autoreply transport checks the characters in options that define the
2477 message's headers; it allows continued headers, but it was checking with
2478 isspace() after an embedded newline instead of explicitly looking for a
2479 space or a tab.
2480
248121. If all the "regular" hosts to which an address was routed had passed their
2482 expiry times, and had not reached their retry times, the address was
2483 bounced, even if fallback hosts were defined. Now Exim should go on to try
2484 the fallback hosts.
2485
248622. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2487 equivalent code in the SMTP transport. Some hosts send humungous responses
2488 to HELO/EHLO, more than 1024 it seems.
2489
249023. Refactoring: code in filter.c used (void *) for "any old type" but this
2491 gives compiler warnings in some environments. I've now done it "properly",
2492 using a union.
2493
249424. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2495 (because of problems with the built-in one) was declared to return uschar *
2496 instead of char *, causing compiler failure.
2497
249825. Fixed a file descriptor leak when processing alias/forward files.
2499
250026. Fixed a minor format string issue in dbfn.c.
2501
250227. Typo in exim.c: ("dmbnz" for "dbmnz").
2503
250428. If a filter file refered to $h_xxx or $message_headers, and the headers
2505 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2506 become corrupted.
2507
250829. When a sender address is verified, it is cached, to save repeating the test
2509 when there is more than one recipient in a message. However, when the
2510 verification involves a callout, it is possible for different callout
2511 options to be set for different recipients. It is too complicated to keep
2512 track of this in the cache, so now Exim always runs a verification when a
2513 callout is required, relying on the callout cache for the optimization.
2514 The overhead is duplication of the address routing, but this should not be
2515 too great.
2516
251730. Fixed a bug in callout caching. If a RCPT command caused the sender address
2518 to be verified with callout=postmaster, and the main callout worked but the
2519 postmaster check failed, the verification correctly failed. However, if a
2520 subsequent RCPT command asked for sender verification *without* the
2521 postmaster check, incorrect caching caused this verification also to fail,
2522 incorrectly.
2523
252431. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2525 it was not caching the DNS options (qualify_single, search_parents) that
2526 were used when the lookup failed. A subsequent lookup with different
2527 options therefore always gave the same answer, though there were cases
2528 where it should not have. (Example: a "domains = !$mx_any" option on a
2529 dnslookup router: the "domains" option is always processed without any
2530 widening, but the router might have qualify_single set.) Now Exim uses the
2531 cached value only when the same options are set.
2532
253332. Added John Jetmore's "exipick" utility to the distribution.
2534
253533. GnuTLS: When an attempt to start a TLS session fails for any reason other
2536 than a timeout (e.g. a certificate is required, and is not provided), an
2537 Exim server now closes the connection immediately. Previously it waited for
2538 the client to close - but if the client is SSL, it seems that they each
2539 wait for each other, leading to a delay before one of them times out.
2540
254134: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2542 maintained 0.8.x compatibility because I don't think many are using it, and
2543 it is clearly obsolete.
2544
254535. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2546 transport.
2547
254836. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2549 client certificate was expired. A simple patch fixes this, though I don't
2550 understand the full logic of why the verify callback is called multiple
2551 times.
2552
255337. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2554 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2555 which causes problems with some clients (such as the Certicom SSL Plus
2556 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2557 disables the coutermeasure allowing Eudora to connect."
2558
255938. Exim was not checking that a write() to a log file succeeded. This could
2560 lead to Bad Things if a log got too big, in particular if it hit a file
2561 size limit. Exim now panics and dies if it cannot write to a log file, just
2562 as it does if it cannot open a log file.
2563
256439. Modified OS/Makefile-Linux so that it now contains
2565
2566 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2567
2568 The two -D definitions ensure that Exim is compiled with large file
2569 support, which makes it possible to handle log files that are bigger than
2570 2^31.
2571
257240. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2573 instance) a domain was checked against a named list that involved a lookup,
2574 causing $domain_data to be set, then another domain was checked against the
2575 same list, then the first domain was re-checked, the value of $domain_data
2576 after the final check could be wrong. In particular, if the second check
2577 failed, it could be set empty. This bug probably also applied to
2578 $localpart_data.
2579
258041. The strip_trailing_dot option was not being applied to the address given
2581 with the -f command-line option.
2582
258342. The code for reading a message's header from the spool was incrementing
2584 $received_count, but never initializing it. This meant that the value was
2585 incorrect (doubled) while delivering a message in the same process in which
2586 it was received. In the most common configuration of Exim, this never
2587 happens - a fresh exec is done - but it can happen when
2588 deliver_drop_privilege is set.
2589
259043. When Exim logs an SMTP synchronization error - client data sent too soon -
2591 it now includes up to 150 characters of the unexpected data in the log
2592 line.
2593
259444. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2595 and building data strings. The size of both of these buffers was 10 000
2596 bytes - far larger than anybody would *ever* want, thought I. Needless to
2597 say, somebody hit the limit. I have increased the maximum line length to
2598 20 000 and the maximum data length of concatenated lines to 100 000. I have
2599 also fixed two bugs, because there was no checking on these buffers. Tsk,
2600 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2601 buffer is too small.
2602
260345. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2604 lsearch lookups. Now it does.
2605
260646. When parsing a route_list item in a manualroute router, a fixed-length
2607 buffer was used for the list of hosts. I made this 1024 bytes long,
2608 thinking that nobody would ever have a list of hosts that long. Wrong.
2609 Somebody had a whole pile of complicated expansion conditions, and the
2610 string was silently truncated, leading to an expansion error. It turns out
2611 that it is easier to change to an unlimited length (owing to other changes
2612 that have happened since this code was originally written) than to build
2613 structure for giving a limitation error. The length of the item that
2614 expands into the list of hosts is now unlimited.
2615
261647. The lsearch lookup could not handle data where the length of text line was
2617 more than 4095 characters. Such lines were truncated, leading to shortened
2618 data being returned. It should now handle lines of any length.
2619
262048. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2621 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2622 ACL").
2623
262449. Cosmetic tidy to scripts like exicyclog that are generated by globally
2625 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2626 no longer happens in comment lines. A list of replacements is now placed
2627 at the head of all of the source files, except those whose only change is
2628 to replace PERL_COMMAND in the very first #! line.
2629
263050. Replaced the slow insertion sort in queue.c, for sorting the list of
2631 messages on the queue, with a bottom-up merge sort, using code contributed
2632 by Michael Haardt. This should make operations like -bp somewhat faster on
2633 large queues. It won't affect queue runners, except when queue_run_in_order
2634 is set.
2635
263651. Installed eximstats 1.31 in the distribution.
2637
263852. Added support for SRV lookups to the dnslookup router.
2639
264053. If an ACL referred to $message_body or $message_body_end, the value was not
2641 reset for any messages that followed in the same SMTP session.
2642
264354. The store-handling optimization for building very long strings was not
2644 differentiating between the different store pools. I don't think this
2645 actually made any difference in practice, but I've tidied it.
2646
264755. While running the routers to verify a sender address, $sender_address
2648 was still set to the sender address. This is wrong, because when routing to
2649 send a bounce to the sender, it would be empty. Therefore, I have changed
2650 it so that, while verifying a sender address, $sender_address is set to <>.
2651 (There is no change to what happens when verifying a recipient address.)
2652
265356. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2654 target A or AAAA records (if not already returned) without resetting the
2655 qualify_single or search_parents options of the DNS resolver. These are
2656 inappropriate in this case because the targets of MX and SRV records must
2657 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2658 target that, when qualified, matched something in the local domain. These
2659 two options are now turned off when doing these lookups.
2660
266157. It seems that at least some releases of Reiserfs (which does not have the
2662 concept of a fixed number of inodes) returns zero and not -1 for the
2663 number of available inodes. This interacted badly with check_spool_inodes,
2664 which assumed that -1 was the "no such thing" setting. What I have done is
2665 to check that the total number of inodes is greater than zero before doing
2666 the test of how many are available.
2667
266858. When a "warn" ACL statement has a log_message modifier, the message is
2669 remembered, and not repeated. This is to avoid a lot of repetition when a
2670 message has many recipients that cause the same warning to be written.
2671 Howewer, Exim was preserving the list of already written lines for an
2672 entire SMTP session, which doesn't seem right. The memory is now reset if a
2673 new message is started.
2674
267559. The "rewrite" debugging flag was not showing the result of rewriting in the
2676 debugging output unless log_rewrite was also set.
2677
267860. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2679 of (int)(handle) when we know that handle contains (void *)(-1).
2680
268161. The Exim daemon panic-logs an error return when it closes the incoming
2682 connection. However "connection reset by peer" seems to be common, and
2683 isn't really an error worthy of noting specially, so that particular error
2684 is no long logged.
2685
268662. When Exim is trying to find all the local interfaces, it used to panic and
2687 die if the ioctl to get the interface flags failed. However, it seems that
2688 on at least one OS (Solaris 9) it is possible to have an interface that is
2689 included in the list of interfaces, but for which you get a failure error
2690 for this call. This happens when the interface is not "plumbed" into a
2691 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2692 failure of the "get flags" call assumes that the interface is down.
2693
269463. Added a ${eval10: operator, which assumes all numbers are decimal. This
2695 makes life easier for people who are doing arithmetic on fields extracted
2696 from dates, where you often get leading zeros that should not be
2697 interpreted as octal.
2698
269964. Added qualify_domain to the redirect router, to override the global
2700 setting.
2701
270265. If a pathologically long header line contained very many addresses (the
2703 report of this problem mentioned 10 000) and each of them was rewritten,
2704 Exim could use up a very large amount of memory. (It kept on making new
2705 copies of the header line as it rewrote, and never released the old ones.)
2706 At the expense of a bit more processing, the header rewriting function has
2707 been changed so that it no longer eats memory in this way.
2708
270966. The generation of the Received: header has been moved from the time that a
2710 message starts to be received, to the time that it finishes. The timestamp
2711 in the Received: header should now be very close to that of the <= log
2712 line. There are two side-effects of this change:
2713
2714 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2715 logged header lines no longer include the local Received: line, because
2716 it has not yet been created. The same applies to a copy of the message
2717 that is returned to a non-SMTP sender when a message is rejected.
2718
2719 (b) When a filter file is tested using -bf, no additional Received: header
2720 is added to the test message. After some thought, I decided that this
2721 is a bug fix.
2722
2723 This change does not affect the value of $received_for. It is still set
2724 after address rewriting, but before local_scan() is called.
2725
272667. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2727
272868. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2729 gave an unhelpful panic error message, and a defer error. I have managed to
2730 change this behaviour so that it now rejects any supplied certificate,
2731 which seems right, as the list of acceptable certificates is empty.
2732
273369. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2734 gave an unhelpful defer error. I have not managed to make this reject any
2735 supplied certificates, but the error message it gives is "no certificate
2736 supplied", which is not helpful.
2737
273870. exigrep's output now also includes lines that are not associated with any
2739 message, but which match the given pattern. Implemented by a patch from
2740 Martin Sluka, which also tidied up the Perl a bit.
2741
274271. Recipient callout verification, like sender verification, was using <> in
2743 the MAIL FROM command. This isn't really the right thing, since the actual
2744 sender may affect whether the remote host accepts the recipient or not. I
2745 have changed it to use the actual sender in the callout; this means that
2746 the cache record is now keyed on a recipient/sender pair, not just the
2747 recipient address. There doesn't seem to be a real danger of callout loops,
2748 since a callout by the remote host to check the sender would use <>.
2749 [SEE ABOVE: changed after hitting problems.]
2750
275172. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2752 temporary errors. However, in the case of such a code being given after
2753 the end of a data transmission (i.e. after ".") Exim was failing to write
2754 a retry record for the message. (Yes, there was some broken host that was
2755 actually sending 8xx at this point.)
2756
275773. An unknown lookup type in a host list could cause Exim to panic-die when
2758 the list was checked. (An example that provoked this was putting <; in the
2759 middle of a list instead of at the start.) If this happened during a DATA
2760 ACL check, a -D file could be left lying around. This kind of configuration
2761 error no longer causes Exim to die; instead it causes a defer errror. The
2762 incident is still logged to the main and panic logs.
2763
276474. Buglet left over from Exim 3 conversion. The message "too many messages
2765 in one connection" was written to the rejectlog but not the mainlog, except
2766 when address rewriting (yes!) was being logged.
2767
276875. Added write_rejectlog option.
2769
277076. When a system filter was run not as root (that is, when system_filter_user
2771 was set), the values of the $n variables were not being returned to the
2772 main process; thus, they were not subsequently available in the $sn
2773 variables.
2774
277577. Added +return_path_on_delivery log selector.
2776
277778. A connection timeout was being treated differently from recipients deferred
2778 when testing hosts_max_try with a message that was older than the host's
2779 retry timeout. (The host should not be counted, thus allowing all hosts to
2780 be tried at least once before bouncing.) This may have been the cause of an
2781 occasionally reported bug whereby a message would remain on the queue
2782 longer than the retry timeout, but would be bounced if a delivery was
2783 forced. I say "may" because I never totally pinned down the problem;
2784 setting up timeout/retry tests is difficult. See also the next item.
2785
278679. The ultimate address timeout was not being applied to errors that involved
2787 a combination of host plus message (for example, a timeout on a MAIL
2788 command). When an address resolved to a number of possible hosts, and they
2789 were not all tried for each delivery (e.g. because of hosts_max_try), a
2790 message could remain on the queue longer than the retry timeout.
2791
279280. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2793 Haardt.
2794
279581. Fixed an obscure SMTP outgoing bug which required at least the following
2796 conditions: (a) there was another message waiting for the same server;
2797 (b) the server returned 5xx to all RCPT commands in the first message so
2798 that the message was not completed; (c) the server dropped the connection
2799 or gave a negative response to the RSET that Exim sends to abort the
2800 transaction. The observed case was a dropped connection after DATA that had
2801 been sent in pipelining mode. That is, the server had advertised PIPELINING
2802 but was not implementing it correctly. The effect of the bug was incorrect
2803 behaviour, such as trying another host, and this could lead to a crash.
2804
2805
2806Exim version 4.30
2807-----------------
2808
2809 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2810 and daemon.c were passed as pointers to ints; they should have been
2811 pointers to socklen_t variables (which are typically unsigned ints).
2812
2813 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2814 fixed.
2815
2816 3. Fixed a really odd bug that affected only the testing scheme; patching a
2817 certain fixed string in the binary changed the value of another string that
2818 happened to be identical to the end of the original first string.
2819
2820 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2821 name", it returns that address as the IP address. On some operating
2822 systems (e.g. Solaris), it also passes back the IP address string as the
2823 "host name". However, on others (e.g. Linux), it passes back an empty
2824 string. Exim wasn't checking for this, and was changing the host name to an
2825 empty string, assuming it had been canonicized.
2826
2827 5. Although rare, it is permitted to have more than one PTR record for a given
2828 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2829 all the names associated with an address, because they do in Solaris.
2830 However, it seems that they do not in Linux for data that comes from the
2831 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2832 I found this out when I moved to a new Linux workstation and tried to run
2833 the Exim test suite.
2834
2835 To get round this problem I have changed the code so that it now does its
2836 own call to the DNS to look up PTR records when searching for a host name.
2837 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2838 addresses that are only in /etc/hosts are still found.
2839
2840 This behaviour is, however, controlled by an option called host_lookup_
2841 order, which defaults to "bydns:byaddr". If people want to use the other
2842 order, or indeed, just use one or the other means of lookup, they can
2843 specify it in this variable.
2844
2845 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2846 some operating systems, this comes back from gethostbyaddr() as an empty
2847 string, and this is what Exim used to test for. However, it seems that in
2848 other systems, "." is yielded. Exim now tests for this case too.
2849
2850 7. The values of check_spool_space and check_log_space are now held internally
2851 as a number of kilobytes instead of an absolute number of bytes. If a
2852 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2853 kilobyte. This means that much larger values can be stored.
2854
2855 8. Exim monitor: an attempt to get the action menu when not actually pointing
2856 at a message produces an empty menu entitled "No message selected". This
2857 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2858 no entries in it ("Shell widget menu has zero width and/or height"). So I
2859 have added a single, blank menu entry in this case.
2860
2861 9. Added ${quote_local_part.
2862
286310. MIME decoding is now applied to the contents of Subject: header lines when
2864 they are logged.
2865
286611. Now that a reference to $sender_host_address automatically causes a reverse
2867 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2868 host lookup before query-style lookups in lists that might use this
2869 variable. This has therefore been abolished, and the "net-" prefix is no
2870 longer necessary for query-style lookups.
2871
287212. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2873 have been a typo for LFLAGS, so it has been changed.
2874
287513. The install script calls Exim with "-C /dev/null" in order to find the
2876 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2877 to be output. Howeve, since Exim outputs its version number before the
2878 error, it didn't break the script. It just looked ugly. I fixed this by
2879 always allowing "-C /dev/null" if the caller is root.
2880
288114. Ignore overlarge ACL variable number when reading spool file - insurance
2882 against a later release with more variables having written the file.
2883
288415. The standard form for an IPv6 address literal was being rejected by EHLO.
2885 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2886 this, as well as the form without the "IPv6" on the front.
2887
288816. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2889 OS/Makefile-Darwin file.
2890
289117. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2892 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2893
289418. After change 4.21/52, "%ld" was used to format the contents of the $inode
2895 variable. However, some OS use ints for inodes. I've added cast to long int
2896 to get rid of the compiler warning.
2897
289819. I had forgotten to lock out "/../" in configuration file names when
2899 ALT_CONFIG_PREFIX was set.
2900
290120. Routers used for verification do not need to specify transports. However,
2902 if such a router generated a host list, and callout was configured, Exim
2903 crashed, because it could not find a port number from the (non-existent)
2904 transport. It now assumes port 25 in this circumstance.
2905
290621. Added the -t option to exigrep.
2907
290822. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2909 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2910 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2911 them is set, LOOKUP_LSEARCH is forced.
2912
291323. "exim -bV" now outputs a list of lookups that are included in the binary.
2914
291524. Added sender and host information to the "rejected by local_scan()" log
2916 line; previously there was no indication of these.
2917
291825. Added .include_if_exists.
2919
292026. Change 3.952/11 added an explicit directory sync on top of a file sync for
2921 Linux. It turns out that not all file systems support this. Apparently some
2922 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2923 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2924 not supported on the file descriptor, is now ignored when Exim is trying to
2925 sync a directory. This applies only to Linux.
2926
292727. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2928
292928. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2930 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2931 this.
2932
293329. Added support for maildirsize files from supplied patch (modified a bit).
2934
293530. The use of :fail: followed by an empty string could lead Exim to respond to
2936 sender verification failures with (e.g.):
2937
2938 550 Verification failed for <xxx>
2939 550 Sender verify failed
2940
2941 where the first response line was missing the '-' that indicates it is not
2942 the final line of the response.
2943
294431. The loop for finding the name of the user that called Exim had a hardwired
2945 limit of 10; it now uses the value of finduser_retries, which is used for
2946 all other user lookups.
2947
294832. Added $received_count variable, available in data and not_smtp ACLs, and at
2949 delivery time.
2950
295133. Exim was neglecting to zero errno before one call of strtol() when
2952 expanding a string and expecting an integer value. On some systems this
2953 resulted in spurious "integer overflow" errors. Also, it was casting the
2954 result into an int without checking.
2955
295634. Testing for a connection timeout using "timeout_connect" in the retry rules
2957 did not work. The code looks as if it has *never* worked, though it appears
2958 to have been documented since at least releast 1.62. I have made it work.
2959
296035. The "timeout_DNS" error in retry rules, also documented since at least
2961 1.62, also never worked. As it isn't clear exactly what this means, and
2962 clearly it isn't a major issue, I have abolished the feature by treating it
2963 as "timeout", and writing a warning to the main and panic logs.
2964
296536. The display of retry rules for -brt wasn't always showing the error code
2966 correctly.
2967
296837. Added new error conditions to retry rules: timeout_A, timeout_MX,
2969 timeout_connect_A, timeout_connect_MX.
2970
297138. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2972 to the empty sender.
2973
297439. The daemon was not analysing the content of -oX till after it had closed
2975 stderr and disconnected from the controlling terminal. This meant that any
2976 syntax errors were only noted on the panic log, and the return code from
2977 the command was 0. By re-arranging the code a little, I've made the
2978 decoding happen first, so such errors now appear on stderr, and the return
2979 code is 1. However, the actual setting up of the sockets still happens in
2980 the disconnected process, so errors there are still only recorded on the
2981 panic log.
2982
298340. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2984 connections (as happens on some IP stacks) was logged at start up time as
2985 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2986 it from "IPv6 and IPv4", which means that two separate sockets are being
2987 used.
2988
298941. The debug output for gethostbyname2() or getipnodebyname() failures now
2990 says whether AF_INET or AF_INET6 was passed as an argument.
2991
299242. Exiwhat output was messed up when time zones were included in log
2993 timestamps.
2994
299543. Exiwhat now gives more information about the daemon's listening ports,
2996 and whether -tls-on-connect was used.
2997
299844. The "port" option of the smtp transport is now expanded.
2999
300045. A "message" modifier in a "warn" statement in a non-message ACL was being
3001 silently ignored. Now an error message is written to the main and panic
3002 logs.
3003
300446. There's a new ACL modifier called "logwrite" which writes to a log file
3005 as soon as it is encountered.
3006
300747. Added $local_user_uid and $local_user_gid at routing time.
3008
300948. Exim crashed when trying to verify a sender address that was being
3010 rewritten to "<>".
3011
301249. Exim was recognizing only a space character after ".include". It now also
3013 recognizes a tab character.
3014
301550. Fixed several bugs in the Perl script that creates the exim.8 man page by
3016 extracting the relevant information from the specification. The man page no
3017 longer contains scrambled data for the -d option, and I've added a section
3018 at the front about calling Exim under different names.
3019
302051. Added "extra_headers" argument to the "mail" command in filter files.
3021
302252. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3023 crash.
3024
302553. Installed eximstats 1.29.
3026
302754. Added transport_filter_timeout as a generic transport option.
3028
302955. Exim no longer adds an empty Bcc: header to messages that have no To: or
3030 Cc: header lines. This was required by RFC 822, but it not required by RFC
3031 2822.
3032
303356. Exim used to add From:, Date:, and Message-Id: header lines to any
3034 incoming messages that did not have them. Now it does so only if the
3035 message originates locally, that is, if there is no associated remote host
3036 address. When Resent- header lines are present, this applies to the Resent-
3037 lines rather than the non-Resent- lines.
3038
303957. Drop incoming SMTP connection after too many syntax or protocol errors. The
3040 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3041
304258. Messages for configuration errors now include the name of the main
3043 configuration file - useful now that there may be more than one file in a
3044 list (.included file names were always shown).
3045
304659. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3047 for those rare installations that do not start the daemon as root or run it
3048 setuid root. I've cut out the call to initgroups() if the daemon is not
3049 root at that time.
3050
305160. The Exim user and group can now be bound into the binary as text strings
3052 that are looked up at the start of Exim's processing.
3053
305461. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3055
305662. Added $mailstore_basename variable.
3057
305863. Installed patch to sieve.c from Michael Haardt.
3059
306064. When Exim failed to open the panic log after failing to open the main log,
3061 the original message it was trying to log was written to stderr and debug
3062 output, but if they were not available (the usual case in production), it
3063 was lost. Now it is written to syslog before the two lines that record the
3064 failures to open the logs.
3065
306665. Users' Exim filters run in subprocesses under the user's uid. It is
3067 possible for a "deliver" command or an alias in a "personal" command to
3068 provoke an address rewrite. If logging of address rewriting is configured,
3069 this fails because the process is not running as root or exim. There may be
3070 a better way of dealing with this, but for the moment (because 4.30 needs
3071 to be released), I have disabled address rewrite logging when running a
3072 filter in a non-root, non-exim process.
3073
3074
3075Exim version 4.24
3076-----------------
3077
3078 1. The buildconfig auxiliary program wasn't quoting the value set for
3079 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3080 not defined. This bug was masked in 4.22 by the effect that was fixed in
3081 change 4.23/1.
3082
3083 2. Some messages that were rejected after a message id was allocated were
3084 shown as "incomplete" by exigrep. It no longer does this for messages that
3085 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3086
3087 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3088 have allow_domain_literals set, the ID did not get logged in the <= line.
3089 Domain literals are now always recognized in Message-ID: header lines.
3090
3091 4. The first argument for a ${extract expansion item is the key name or field
3092 number. Leading and trailing spaces in this item were not being ignored,
3093 causing some misleading effects.
3094
3095 5. When deliver_drop_privilege was set, single queue runner processes started
3096 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3097 same command in the process it spins off) were not dropping privilege.
3098
3099 6. When the daemon running as "exim" started a queue runner, it always
3100 re-executed Exim in the spun-off process. This is a waste of effort when
3101 deliver_drop_privilege is set. The new process now just calls the
3102 queue-runner function directly.
3103
3104
3105Exim version 4.23
3106-----------------
3107
3108 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3109 HEADERS_CHARSET.
3110
3111 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3112 ignored. Though the use of -oP was forcing the writing of a pid file, it
3113 was always written to the default place.
3114
3115 3. If the message "no IP address found for host xxxx" is generated during
3116 incoming verification, it is now followed by identification of the incoming
3117 connection (so you can more easily find what provoked it).
3118
3119 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3120
3121 5. Added some features to "harden" Exim a bit more against certain attacks:
3122
3123 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3124 be put in Local/Makefile. This is like the never_users runtime option,
3125 but it cannot be overridden. The default setting is "root".
3126
3127 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3128 prefix string with which any file named in a -C command line option
3129 must start.
3130
3131 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3132 is retained for -C and -D only if the caller of Exim is root. Without
3133 it, the exim user may also use -C and -D and retain privilege.
3134
3135 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3136 command line option is disabled.
3137
3138 6. Macro names set by the -D option must start with an upper case letter, just
3139 like macro names defined in the configuration file.
3140
3141 7. Added "dereference=" facility to LDAP.
3142
3143 8. Two instances of the typo "uknown" in the source files are fixed.
3144
3145 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3146 the Configure-Makefile script screwed up while processing it.
3147
314810. Incorporated PCRE 4.4.
3149
315011. The SMTP synchronization check was not operating right at the start of an
3151 SMTP session. For example, it could not catch a HELO sent before the client
3152 waited for the greeting. There is now a check for outstanding input at the
3153 point when the greeting is written. Because of the duplex, asynchronous
3154 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3155 way, but not yet received, when the check is performed.
3156
315712. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3158 on TCP/IP sockets, because this apparently causes some broken clients to
3159 timeout.
3160
316113. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3162 unchanged) from the Cygwin maintainer.
3163
316414. The code for -bV that shows what is in the binary showed "mbx" when maildir
3165 was supported instead of testing for mbx. Effectively a typo.
3166
316715. The spa authenticator server code was not checking that the input it
3168 received was valid base64.
3169
317016. The debug output line for the "set" modifier in ACLs was not showing the
3171 name of the variable that was being set.
3172
317317. Code tidy: the variable type "vtype_string" was never used. Removed it.
3174
317518. Previously, a reference to $sender_host_name did not cause a DNS reverse
3176 lookup on its own. Something else was needed to trigger the lookup. For
3177 example, a match in host_lookup or the need for a host name in a host list.
3178 Now, if $sender_host_name is referenced and the host name has not yet been
3179 looked up, a lookup is performed. If the lookup fails, the variable remains
3180 empty, and $host_lookup_failed is set to "1".
3181
318219. Added "eqi" as a case-independent comparison operator.
3183
318420. The saslauthd authentication condition could segfault if neither service
3185 nor realm was specified.
3186
318721. If an overflowing value such as "2048M" was set for message_size_limit, the
3188 error message that was logged was misleading, and incoming SMTP
3189 connections were dropped. The message is now more accurate, and temporary
3190 errors are given to SMTP connections.
3191
319222. In some error situations (such as 21 above) Exim rejects all SMTP commands
3193 (except RSET) with a 421 error, until QUIT is received. However, it was
3194 failing to send a response to QUIT.
3195
319623. The HELO ACL was being run before the code for helo_try_verify_hosts,
3197 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3198 ACL is now run after the helo_try_verify_hosts code.
3199
320024. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3201 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3202 case-independent, so other case variants are also recognized). Apparently
3203 some systems use these upper case variants.
3204
320525. If more than two messages were waiting for the same host, and a transport
3206 filter was specified for the transport, Exim sent two messages over the
3207 same TCP/IP connection, and then failed with "socket operation on non-
3208 socket" when it tried to send the third.
3209
321026. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3211
321227. The extern definition of crypt16() in expand.c was not being excluded when
3213 the OS had its own crypt16() function.
3214
321528. Added bounce_return_body as a new option, and bounce_return_size_limit
3216 as a preferred synonym for return_size_limit, both as an option and as an
3217 expansion variable.
3218
321929. Added LIBS=-liconv to OS/Makefile-OSF1.
3220
322130. Changed the default configuration ACL to relax the local part checking rule
3222 for addresses that are not in any local domains. For these addresses,
3223 slashes and pipe symbols are allowed within local parts, but the sequence
3224 /../ is explicitly forbidden.
3225
322631. SPA server authentication was not clearing the challenge buffer before
3227 using it.
3228
322932. log_message in a "warn" ACL statement was writing to the reject log as
3230 well as to the main log, which contradicts the documentation and doesn't
3231 seem right (because no rejection is happening). So I have stopped it.
3232
323333. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3234 However, I am unable to do any testing of this.
3235
323634. Fixed an infelicity in the appendfile transport. When checking directories
3237 for a mailbox, to see if any needed to be created, it was accidentally
3238 using path names with one or more superfluous leading slashes; tracing
3239 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3240
324135. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3242 contents are added to the log line that is written for every discarded
3243 recipient. (Previously a log_message setting was ignored.)
3244
324536. The ${quote: operator now quotes the string if it is empty.
3246
324737. The install script runs exim in order to find its version number. If for
3248 some reason other than non-existence or emptiness, which it checks, it
3249 could not run './exim', it was installing it with an empty version number,
3250 i.e. as "exim-". This error state is now caught, and the installation is
3251 aborted.
3252
325338. An argument was missing from the function that creates an error message
3254 when Exim fails to connect to the socket for saslauthd authentication.
3255 This could cause Exim to crash, or give a corrupted message.
3256
325739. Added isip, isip4, and isip6 to ${if conditions.
3258
325940. The ACL variables $acl_xx are now saved with the message, and can be
3260 accessed later in routers, transports, and filters.
3261
326241. The new lookup type nwildlsearch is like wildlsearch, except that the key
3263 strings in the file are not string-expanded.
3264
326542. If a MAIL command specified a SIZE value that was too large to fit into an
3266 int variable, the check against message_size_limit failed. Such values are
3267 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3268 day this will have to be increased, but I don't think I want to be around
3269 when emails are that large.
3270
3271
3272
3273Exim version 4.22
3274-----------------
3275
3276 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3277 iconv() is not standard in FreeBSD.
3278
3279 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3280 IPv6 enabled. The observed symptom was a segmentation fault on return from
3281 the function os_common_find_running_interfaces() in src/os.c.
3282
3283 3. In the check_special_case() function in daemon.c I had used "errno" as an
3284 argument name, which causes warnings on some systems. This was basically a
3285 typo, since it was named "eno" in the comments!
3286
3287 4. The code that waits for the clock to tick (at a resolution of some fraction
3288 of a second) so as to ensure message-id uniqueness was always waiting for
3289 at least one whole tick, when it could have waited for less. [This is
3290 almost certainly not relevant at current processor speeds, where it is
3291 unlikely to ever wait at all. But we try to future-proof.]
3292
3293 5. The function that sleeps for a time interval that includes fractions of a
3294 second contained a race. It did not block SIGALRM between setting the
3295 timer, and suspending (a couple of lines later). If the interval was short
3296 and the sigsuspend() was delayed until after it had expired, the suspension
3297 never ended. On busy systems this could lead to processes getting stuck for
3298 ever.
3299
3300 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3301 process, before it forks any delivery processes. The open lookup caching
3302 mechanism meant that the open file or database connection was passed into
3303 the delivery process. The problem was that delivery processes always tidy
3304 up cached lookup data. This could cause a problem for the next delivery
3305 process started by the queue runner, because the external queue runner
3306 process does not know about the closure. So the next delivery process
3307 still has data in the lookup cache. In the case of a file lookup, there was
3308 no problem because closing a file descriptor in a subprocess doesn't affect
3309 the parent. However, if the lookup was caching a connection to a database,
3310 the connection was closed, and the second delivery process was likely to
3311 see errors such as "PGSQL: query failed: server closed the connection
3312 unexpectedly". The problem has been fixed by closing all cached lookups
3313 in a queue runner before running a delivery process.
3314
3315 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3316 seem to have the "const" qualifier which it has on other OS. I've
3317 parameterised it.
3318
3319 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3320 *of the same type* (client or server) with the same public name that an
3321 error should be diagnosed.
3322
3323 9. When Exim looked up a host name for an IP address, but failed to find the
3324 original IP address when looking up the host name (a safety check), it
3325 output the message "<ip address> does not match any IP for NULL", which was
3326 confusing, to say the least. The bug was that the host name should have
3327 appeared instead of "NULL".
3328
332910. Since release 3.03, if Exim is called by a uid other than root or the Exim
3330 user that is built into the binary, and the -C or -D options is used, root
3331 privilege is dropped before the configuration file is read. In addition,
3332 logging is switched to stderr instead of the normal log files. If the
3333 configuration then re-defines the Exim user, the unprivileged environment
3334 is probably not what is expected, so Exim logs a panic warning message (but
3335 proceeds).
3336
3337 However, if deliver_drop_privilege is set, the unprivileged state may well
3338 be exactly what is intended, so the warning has been cut out in that case,
3339 and Exim is allowed to try to write to its normal log files.
3340
3341
3342Exim version 4.21
3343-----------------
3344
3345 1. smtp_return_error_details was not giving details for temporary sender
3346 or receiver verification errors.
3347
3348 2. Diagnose a configuration error if two authenticators have the same public
3349 name.
3350
3351 3. Exim used not to create the message log file for a message until the first
3352 delivery attempt. This could be confusing when incoming messages were held
3353 for policy or load reasons. The message log file is now created at the time
3354 the message is received, and an initial "Received" line is written to it.
3355
3356 4. The automatically generated man page for command line options had a minor
3357 bug that caused no ill effects; however, a more serious problem was that
3358 the procedure for building the man page automatically didn't always
3359 operate. Consequently, release 4.20 contains an out-of-date version. This
3360 shouldn't happen again.
3361
3362 5. When building Exim with embedded Perl support, the script that builds the
3363 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3364 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3365
3366 6. The freeze_tell option was not being used for messages that were frozen on
3367 arrival, either by an ACL or by local_scan().
3368
3369 7. Added the smtp_incomplete_transaction log selector.
3370
3371 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3372 was accepting AUTH without a new EHLO.
3373
3374 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3375 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3376 RFC.
3377
337810. Logging of TCP/IP connections (when configured) now happens in the main
3379 daemon process instead of the child process, so that the TCP/IP connection
3380 count is more accurate (but it can never be perfect).
3381
338211. The use of "drop" in a nested ACL was not being handled correctly in the
3383 outer ACL. Now, if condition failure induced by the nested "drop" causes
3384 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3385 or "require"), the connection is dropped.
3386
338712. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3388 that yield "discard" can now be used with an "accept" or a "discard" verb,
3389 but an error is generated for any others (because I can't see a useful way
3390 to define what should happen).
3391
339213. When an ACL is read dynamically from a file (or anywhere else), the lines
3393 are now processed in the same way as lines in the Exim configuration file.
3394 In particular, continuation lines are supported.
3395
339614. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3397
339815. Added -ti meaning -t -i.
3399
340016. Check for letters, digits, hyphens, and dots in the names of dnslist
3401 domains, and warn by logging if others are found.
3402
340317. At least on BSD, alignment is not guarenteed for the array of ifreq's
3404 returned from GIFCONF when Exim is trying to find the list of interfaces on
3405 a host. The code in os.c has been modified to copy each ifreq to an aligned
3406 structure in all cases.
3407
3408 Also, in some cases, the returned ifreq's were being copied to a 'struct
3409 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3410 means the last couple of bytes of an IPv6 address could be chopped if the
3411 ifreq contained only a normal sockaddr (14 bytes storage).
3412
341318. Named domain lists were not supported in the hosts_treat_as_local option.
3414 An entry such as +xxxx was not recognized, and was treated as a literal
3415 domain name.
3416
341719. Ensure that header lines added by a DATA ACL are included in the reject log
3418 if the ACL subsequently rejects the message.
3419
342020. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3421 MD5 (which is deprecated).
3422
342321. When testing a filter file using -bf, Exim was writing a message when it
3424 took the sender from a "From " line in the message, but it was not doing so
3425 when it took $return_path from a Return-Path: header line. It now does.
3426
342722. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3428 with a valid header line field name (a series of printing characters
3429 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3430
343123. Changed "disc" in the source to "disk" to conform to the documentation and
3432 the book and for uniformity.
3433
343424. Ignore Sendmail's -Ooption=value command line item.
3435
343625. When execve() failed while trying to run a command in a pipe transport,
3437 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3438 could be confused with a return value of 69 from the command itself. This
3439 has been changed to 127, the value the shell returns if it is asked to run
3440 a non-existent command. The wording for the related log line suggests a
3441 non-existent command as the problem.
3442
344326. If received_header_text expands to an empty string, do not add a Received:
3444 header line to the message. (Well, it adds a token one on the spool, but
3445 marks it "old" so that it doesn't get used or transmitted.)
3446
344727. Installed eximstats 1.28 (addition of -nt option).
3448
344928. There was no check for failure on the call to getsockname() in the daemon
3450 code. This can fail if there is a shortage of resources on the system, with
3451 ENOMEM, for example. A temporary error is now given on failure.
3452
345329. Contrary to the C standard, it seems that in some environments, the
3454 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3455 program. Exim now does this explicitly; it affects the formatting of
3456 timestamps using strftime().
3457
345830. If exiqsumm was given junk data, it threw up some uninitialized variable
3459 complaints. I've now initialized all the variables, to avoid this.
3460
346132. Header lines added by a system filter were not being "seen" during
3462 transport-time rewrites.
3463
346433. The info_callback() function passed to OpenSSL is set up with type void
3465 (*)(SSL *, int, int), as described somewhere. However, when calling the
3466 function (actually a macro) that sets it up, the type void(*)() is
3467 expected. I've put in a cast to prevent warnings from picky compilers.
3468
346934. If a DNS black list lookup found a CNAME record, but there were no A
3470 records associated with the domain it pointed at, Exim crashed.
3471
347235. If a DNS black list lookup returned more than one A record, Exim ignored
3473 all but the first. It now scans all returned addresses if a particular IP
3474 value is being sought. In this situation, the contents of the
3475 $dnslist_value variable are a list of all the addresses, separated by a
3476 comma and a space.
3477
347836. Tightened up the rules for host name lookups using reverse DNS. Exim used
3479 to accept a host name and all its aliases if the forward lookup for any of
3480 them yielded the IP address of the incoming connection. Now it accepts only
3481 those names whose forward lookup yields the correct IP address. Any other
3482 names are discarded. This closes a loophole whereby a rogue DNS
3483 administrator could create reverse DNS records to break through a
3484 wildcarded host restriction in an ACL.
3485
348637. If a user filter or a system filter that ran in a subprocess used any of
3487 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3488 the wrong values were passed to the pipe command ($thisaddress had the
3489 value of $0, $0 had the value of $1, etc). This bug was introduced by
3490 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3491
349238. Improved the line breaking for long SMTP error messages from ACLs.
3493 Previously, if there was no break point between 40 and 75 characters, Exim
3494 left the rest of the message alone. Two changes have been made: (a) I've
3495 reduced the minimum length to 35 characters; (b) if it can't find a break
3496 point between 35 and 75 characters, it looks ahead and uses the first one
3497 that it finds. This may give the occasional overlong line, but at least the
3498 remaining text gets split now.
3499
350039. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3501 file descriptors might be low, and that setting 1000 would always raise it.
3502 It turns out that in some environments, the limit is already over 1000 and
3503 that lowering it causes trouble. So now Exim takes care not to decrease it.
3504
350540. When delivering a message, the value of $return_path is set to $sender_
3506 address at the start of routing (routers may change the value). By an
3507 oversight, this default was not being set up when an address was tested by
3508 -bt or -bv, which affected the outcome if any router or filter referred to
3509 $return_path.
3510
351141. The idea of the "warn" ACL verb is that it adds a header or writes to the
3512 log only when "message" or "log_message" are set. However, if one of the
3513 conditions was an address verification, or a call to a nested ACL, the
3514 messages generated by the underlying test were being passed through. This
3515 no longer happens. The underlying message is available in $acl_verify_
3516 message for both "message" and "log_message" expansions, so it can be
3517 passed through if needed.
3518
351942. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3520 new expansion $bh_ to give the encoded byte string without charset
3521 translation. Translation happens only if iconv() is available; HAVE_ICONV
3522 indicates this at build time. HEADERS_CHARSET gives the charset to
3523 translate to; headers_charset can change it in the configuration, and
3524 "headers charset" can change it in an individual filter file.
3525
352643. Now that we have a default RFC 2047 charset (see above), the code in Exim
3527 that creates RFC 2047 encoded "words" labels them as that charset instead
3528 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3529 expansion operator; (ii) when Exim creates a From: line for a local
3530 message; (iii) when a header line is rewritten to include a "phrase" part.
3531
353244. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3533 buggy, causing it to skip the first lines of messages whose message ID
3534 ended in 'D'. This would not have bitten before Exim release 4.14, because
3535 message IDs were unlikely to end in 'D' before then. The effect was to have
3536 incorrect size information for certain domains.
3537
353845. #include "config.h" was missing at the start of the crypt16.c module. This
3539 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3540 noticed.
3541
354246. If there was a timeout during a "random" callout check, Exim treated it as
3543 a failure of the random address, and carried on sending RSET and the real
3544 address. If the delay was just some slowness somewhere, the response to the
3545 original RCPT would be taken as a response to RSET and so on, causing
3546 mayhem of various kinds.
3547
354847. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3549 when I implemented it. It didn't allow for the fact that some option values
3550 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3551 the right test for positive values.
3552
355348. Domain names in DNS records are case-independent. Exim always looks them up
3554 in lower case. Some resolvers return domain names in exactly the case they
3555 appear in the zone file, that is, they may contain uppercase letters. Not
3556 all resolvers do this - some return always lower case. Exim was treating a
3557 change of case by a resolver as a change of domain, similar to a widening
3558 of a domain abbreviation. This triggered its re-routing code and so it was
3559 trying to route what was effectively the same domain again. This normally
3560 caused routing to fail (because the router wouldn't handle the domain
3561 twice). Now Exim checks for this case specially, and just changes the
3562 casing of the domain that it ultimately uses when it transmits the message
3563 envelope.
3564
356549. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3566 module.
3567
356850. If a filter generated a file delivery with a non-absolute name (possible if
3569 no home directory exists for the router), the forbid_file option was not
3570 forbidding it.
3571
357251. Added '&' feature to dnslists, to provide bit mask matching in addition to
3573 the existing equality matching.
3574
357552. Exim was using ints instead of ino_t variables in some places where it was
3576 dealing with inode numbers.
3577
357853. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3579 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3580 called TMPDIR, and if it finds it is different, it changes its value.
3581
358254. The smtp_printf() function is now made available to local_scan() so
3583 additional output lines can be written before returning. There is also an
3584 smtp_fflush() function to enable the detection of a dropped connection.
3585 The variables smtp_input and smtp_batched_input are exported to
3586 local_scan().
3587
358855. Changed the default runtime configuration: the message "Unknown user"
3589 has been removed from the ACL, and instead placed on the localuser router,
3590 using the cannot_route_message feature. This means that any verification
3591 failures that generate their own messages won't get overridden. Similarly,
3592 the "Unrouteable address" message that was in the ACL for unverifiable
3593 relay addresses has also been removed.
3594
359556. Added hosts_avoid_esmtp to the smtp transport.
3596
359757. The exicyclog script was not checking for the esoteric option
3598 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3599 will work only if exicyclog is run under the appropriate euid.
3600
360158. Following a discussion on the list, the rules by which Exim recognises line
3602 endings on incoming messages have been changed. The -dropcr and drop_cr
3603 options are now no-ops, retained only for backwards compatibility. The
3604 following line terminators are recognized: LF CRLF CR. However, special
3605 processing applies to CR:
3606
3607 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3608 nor a local message in the state where . is a terminator.
3609
3610 (ii) If a bare CR is encountered in a header line, an extra space is added
3611 after the line terminator so as not to end the header. The reasoning
3612 behind this is that bare CRs in header lines are most likely either
3613 to be mistakes, or people trying to play silly games.
3614
361559. The size of a message, as listed by "-bp" or in the Exim monitor window,
3616 was being incorrectly given as 18 bytes larger than it should have been.
3617 This is a VOB (very old bug).
3618
361960. This may never have affected anything current, but just in case it has:
3620 When the local host is found other than at the start of a list of hosts,
3621 the local host, those with the same MX, and any that follow, are discarded.
3622 When the list in question was part of a longer list of hosts, the following
3623 hosts (not currently being processed) were also being discarded. This no
3624 longer happens. I'm not sure if this situation could ever has previously
3625 arisen.
3626
362761. Added the "/MX" feature to lists of hosts in the manualroute and query
3628 program routers.
3629
363062. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3631 header. This is something that is recommended in a new Internet Draft, and
3632 is something that is documented as being done by Sendmail. There are two
3633 possible values. For messages generated by the autoreply transport, Exim
3634 adds:
3635
3636 Auto-Submitted: auto-replied
3637
3638 whereas for all other generated messages (e.g. bounces) it adds
3639
3640 Auto-Submitted: auto-generated
3641
364263. The "personal" condition in filters now includes a test for the
3643 Auto-Submitted: header. If it contains the string "auto-" the message it
3644 not considered personal.
3645
364664. Added rcpt_include_affixes as a generic transport option.
3647
364865. Added queue_only_override (default true).
3649
365066. Added the syslog_duplication option.
3651
365267. If what should have been the first header line of a message consisted of
3653 a space followed by a colon, Exim was mis-interpreting it as a header line.
3654 It isn't of course - it is syntactically invalid and should therefore be
3655 treated as the start of the message body. The misbehaviour could have
3656 caused a number of strange effects, including loss of data in subsequent
3657 header lines, and spool format errors.
3658
365968. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3660 client host had authenticated. This control can now be exercised by an ACL
3661 for more flexibility.
3662
366369. By default, callouts do not happen when testing with -bh. There is now a
3664 variant, -bhc, which does actually run the callout code, including
3665 consulting and updating the callout cache.
3666
366770. Added support for saslauthd authentication, courtesy of Alexander
3668 Sabourenkov.
3669
367071. If statvfs() failed on the spool or log directories while checking their
3671 size for availability, Exim confusingly gave the error "space shortage".
3672 Furthermore, in debugging mode it crashed with a floating point exception.
3673 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3674 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3675 serious problem, Exim now writes to the main and panic logs when this
3676 happens, with details of the failure. It then refuses to accept the
3677 incoming message, giving the message "spool directory problem" or "log
3678 directory problem" with a 421 code for SMTP messages.
3679
368072. When Exim is about to re-exec itself, it ensures that the file descriptors
3681 0, 1, and 2 exist, because some OS complain for execs without them (see
3682 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3683 descriptors. However, the code omitted to check that the open succeeded,
3684 causing mysterious errors if for some reason the permissions on /dev/null
3685 got screwed. Now Exim writes a message to the main and panic logs, and
3686 bombs out if it can't open /dev/null.
3687
368873. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and