Update the test suite's runtest script to allow for ClamAV support via
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
75e0e026 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.304 2006/02/16 10:05:33 ph10 Exp $
495ae4b0
PH
2
3Change log file for Exim from version 4.21
4-------------------------------------------
5
214e2000
PH
6Exim version 4.61
7-----------------
8
9PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
15 addresses as local.
16
f9daeae0
PH
17PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
19
7e66e54d
PH
20PH/03 Added disable_ipv6.
21
c8ea1597
PH
22PH/04 Changed $reply_address to use the raw form of the headers instead of the
23 decoded form, because it is most often used to construct To: headers
24 lines in autoreplies, and the decoded form may well be syntactically
25 invalid. However, $reply_address has leading white space removed, and all
26 newlines turned into spaces so that the autoreply transport does not
27 grumble.
28
911f6fde
PH
29PH/05 If group was specified without a user on a router, and no group or user
30 was specified on a transport, the group from the router was ignored.
31
47ca6d6c
PH
32PH/06 Increased the number of ACL variables to 20 of each type, and arranged
33 for visible compile-time settings that can be used to change these
34 numbers, for those that want even more. Backwards compatibility with old
35 spool files has been maintained. However, going back to a previous Exim
36 release will lost any variables that are in spool files.
37
ed0e9820
PH
38PH/07 Two small changes when running in the test harness: increase delay when
39 passing a TCP/IP connection to a new process, in case the original
40 process has to generate a bounce, and remove special handling of
41 127.0.0.2 (sic), which is no longer necessary.
42
eff37e47
PH
43PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
44 be the same on different OS.
45
1921d2ea
PH
46PH/09 Moved a debug statement in filter processing to avoid a race problem when
47 testing.
48
b3f69ca8
JJ
49JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
50 whether --show-vars was specified or not
51
52JJ/02 exipick: Added support for new ACL variable spool format introduced
53 in 4.61-PH/06
54
424a1c63
PH
55PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
56 syntactically invalid From: or Reply-to: line, and a filter used this to
57 generate an autoreply, and therefore failed to obtain an address for the
58 autoreply, Exim could try to deliver to a non-existent relative file
59 name, causing unrelated and misleading errors. What now happens is that
60 it logs this as a hard delivery error, but does not attempt to create a
61 bounce message.
62
7a100415
PH
63PH/11 The exinext utility has a -C option for testing purposes, but although
64 the given file was scanned by exinext itself; it wasn't being passed on
65 when Exim was called.
66
19b9dc85
PH
67PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
68 an end-of-file indication when reading a command response.
69
309bd837
PH
70PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
71 compiled. In many other places in Exim, IPv6 addresses are always
72 recognized, so I have changed this. It also means that IPv4 domain
73 literals of the form [IPV4:n.n.n.n] are now always recognized.
74
59e82a2a
PH
75PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
76 used if the router is not running as root, for example, when verifying at
77 ACL time, or when using -bh. The debugging output from this situation was
78 non-existent - all you got was a failure to exec. I have made two
79 changes:
80
81 (a) Failures to set uid/gid, the current directory, or a process leader
82 in a subprocess such as that created by queryprogram now generate
83 suitable debugging ouput when -d is set.
84
85 (b) The queryprogram router detects when it is not running as root,
86 outputs suitable debugging information if -d is set, and then runs
87 the subprocess without attempting to change uid/gid.
88
9edc04ce
PH
89PH/15 Minor change to Makefile for building test_host (undocumented testing
90 feature).
91
1349e1e5
PH
92PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
93 additional section of a DNS packet that returns MX or SRV records.
94 Instead, it always explicitly searches for A/AAAA records. This avoids
95 major problems that occur when a DNS server includes only records of one
96 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
97 fixed another bug: if SRV records were looked up and the corresponding
98 address records were *not* found in the additional section, the port
99 values from the SRV records were lost.
100
ea49d0e1
PH
101PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
102 using the correct key (the original address) when searching the retry
103 rules in order to find which one to use for generating the retry hint.
104
064a94c9
PH
105PH/18 If quota_warn_message contains a From: header, Exim now refrains from
106 adding the default one. Similarly, if it contains a Reply-To: header, the
107 errors_reply_to option, if set, is not used.
108
727071f8
PH
109PH/19 When calculating a retry time, Exim used to measure the "time since
110 failure" by looking at the "first failed" field in the retry record. Now
111 it does not use this if it is later than than the arrival time of the
112 message. Instead it uses the arrival time. This makes for better
113 behaviour in cases where some deliveries succeed, thus re-setting the
114 "first failed" field. An example is a quota failure for a huge message
115 when small messages continue to be delivered. Without this change, the
116 "time since failure" will always be short, possible causing more frequent
117 delivery attempts for the huge message than are intended.
118
f78eb7c6
PH
119PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
120 $1, $2, $3) because the numerical variables can be reset during some
121 expansion items (e.g. "match"), thereby losing the authentication data.
122
21c28500
PH
123PH/21 Make -bV show the size of off_t variables so that the test suite can
124 decide whether to run tests for quotas > 2G.
125
126PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
127 mailbox_size, and mailbox_filecount in the appendfile transport. If a
128 filecount value is greater than 2G or if a quota value is greater than 2G
129 on a system where the size of off_t is not greater than 4, a panic error
130 is given.
131
1688f43b
PH
132PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
133 never match. The debug and -bh output now contains an explicit error
134 message indicating a malformed IPv4 address or mask.
135
136PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
137 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
138 PH/23 above applies.
139
9675b384
PH
140PH/25 Do not write to syslog when running in the test harness. The only
141 occasion when this arises is a failure to open the main or panic logs
142 (for which there is an explicit test).
143
6a3f1455
PH
144PH/26 Added the /no_tell option to "control=freeze".
145
dac79d3e
PH
146PH/27 If a host name lookup failed very early in a connection, for example, if
147 the IP address matched host_lookup and the reverse lookup yielded a name
148 that did not have a forward lookup, an error message of the form "no IP
149 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
150 could be logged. Now it outputs the IP address instead of "NULL".
1349e1e5 151
5977a0b3
PH
152PH/28 An enabling patch from MH: add new function child_open_exim2() which
153 allows the sender and the authenticated sender to be set when
154 submitting a message from within Exim. Since child_open_exim() is
155 documented for local_scan(), the new function should be too.
156
c91535f3
PH
157PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
158 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
159 results in an empty string is now treated as unset.
160
0d46a8c8
PH
161PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
162
278c6e6c
PH
163PH/31 Added sender_verify_fail logging option.
164
2cbb4081
PH
165PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
166 needed by RFC 822 but not by RFC 2822 was commented out. I have now
167 tidied the source and removed it altogether.
168
3eef829e
PH
169PH/33 When a queue run was abandoned because the load average was too high, a
170 log line was always written; now it is written only if the queue_run log
171 selector is set. In addition, the log line for abandonment now contains
172 information about the queue run such as the pid. This is always present
173 in "start" and "stop" lines but was omitted from the "abandon" line.
174
1ab95fa6
PH
175PH/34 Omit spaces between a header name and the colon in the error message that
176 is given when verify = headers_syntax fails (if there are lots of them,
177 the message gets confusing).
178
230205fc
PH
179PH/35 Change the default for dns_check_names_pattern to allow slashes within
180 names, as there are now some PTR records that contain slashes. This check
181 is only to protect against broken name servers that fall over on strange
182 characters, so the fact that it applies to all lookups doesn't matter.
183
75e0e026
PH
184PH/36 Now that the new test suite is complete, we can remove some of the
185 special code in Exim that was needed for the old test suite. For example,
186 sorting DNS records because real resolvers return them in an arbitrary
187 order. The new test suite's fake resolver always returns records in the
188 same order.
189
190PH/37 When running in the test harness, use -odi for submitted messages (e.g.
191 bounces) except when queue_only is set, to avoid logging races between
192 the different processes.
193
214e2000 194
5de37277
PH
195Exim version 4.60
196-----------------
197
cc38ddbf
PH
198PH/01 Two changes to the default runtime configuration:
199
200 (1) Move the checks for relay_from_hosts and authenticated clients from
201 after to before the (commented out) DNS black list checks.
202
203 (2) Add control=submission to the relay_from_hosts and authenticated
204 clients checks, on the grounds that messages accepted by these
205 statements are most likely to be submissions.
5de37277 206
72fdd6ae
PH
207PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
208
209 (1) Generate an error if the third argument for the ${prvs expansion is
210 not a single digit.
211
212 (2) Treat a missing third argument of ${prvscheck as if it were an empty
213 string.
214
215 (3) Reset the variables that are obtained from the first argument of
216 ${prvscheck and used in the second argument before leaving the code,
217 because their memory is reclaimed, so using them afterwards may do
218 silly things.
219
220 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
221 one (it's much easier than Tom thought :-).
222
223 (5) Because of (4), we can now allow for the use of $prvscheck_result
224 inside the third argument.
cb9328de 225
cb741023
PH
226PH/03 For some reason, the default setting of PATH when running a command from
227 a pipe transport was just "/usr/bin". I have changed it to
228 "/bin:/usr/bin".
229
f174f16e
PH
230PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
231 anything to be listed in the output from -bV.
b2f5a032 232
c25242d7
PH
233PH/05 When a filter generated an autoreply, the entire To: header line was
234 quoted in the delivery log line, like this:
235
236 => >A.N.Other <ano@some.domain> <original@ddress> ...
237
238 This has been changed so that it extracts the operative address. There
239 may be more than one such address. If so, they are comma-separated, like
240 this:
241
242 => >ano@some.domain,ona@other.domain <original@ddress> ...
243
82c19f95
PH
244PH/06 When a client host used a correct literal IP address in a HELO or EHLO
245 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
246 not being looked up in the rDNS to get a host name, Exim was showing the
247 IP address twice in Received: lines, even though the IP addresses were
248 identical. For example:
249
250 Received: from [1.2.3.4] (helo=[1.2.3.4])
251
252 However, if the real host name was known, it was omitting the HELO data
253 if it matched the actual IP address. This has been tidied up so that it
254 doesn't show the same IP address twice.
255
d7ffbc12
PH
256PH/07 When both +timestamp and +memory debugging was on, the value given by
257 $tod_xxx expansions could be wrong, because the tod_stamp() function was
258 called by the debug printing, thereby overwriting the timestamp buffer.
259 Debugging no longer uses the tod_stamp() function when +timestamp is set.
260
9f526266
PH
261PH/08 When the original message was included in an autoreply transport, it
262 always said "this is a copy of the message, including all the headers",
263 even if body_only or headers_only was set. It now gives an appropriate
264 message.
265
87fcc8b9
PH
266PH/09 Applied a patch from the Sieve maintainer which:
267
268 o fixes some comments
269 o adds the (disabled) notify extension core
270 o adds some debug output for the result of if/elsif tests
271 o points to the current vacation draft in the documentation
272 and documents the missing references header update
273
274 and most important:
275
276 o fixes a bug in processing the envelope test (when testing
277 multiple envelope elements, the last element determinted the
278 result)
279
456682f5
PH
280PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
281 Electronic Mail") by including:
282
283 Auto-submitted: auto-generated
284
285 in the messages that it generates (bounce messages and others, such as
286 warnings). In the case of bounce messages for non-SMTP mesages, there was
287 also a typo: it was using "Auto_submitted" (underscore instead of
288 hyphen). Since every message generated by Exim is necessarily in response
289 to another message, thes have all been changed to:
290
291 Auto-Submitted: auto-replied
292
293 in accordance with these statements in the RFC:
294
295 The auto-replied keyword:
296
297 - SHOULD be used on messages sent in direct response to another
298 message by an automatic process,
299
300 - MUST NOT be used on manually-generated messages,
301
302 - MAY be used on Delivery Status Notifications (DSNs) and Message
303 Disposition Notifications (MDNs),
304
305 - MUST NOT be used on messages generated by automatic or periodic
306 processes, except for messages which are automatic responses to
307 other messages.
308
3e46c1aa
PH
309PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
310 to the default Received: header definition.
456682f5 311
49826d12
PH
312PH/12 Added log selector acl_warn_skipped (default on).
313
eba0c039
PH
314PH/13 After a successful wildlsearch lookup, discard the values of numeric
315 variables because (a) they are in the wrong storage pool and (b) even if
316 they were copied, it wouldn't work properly because of the caching.
317
a0d6ba8a
PH
318PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
319 checking when decoding. Apparently there are clients that generate
320 overlong encoded strings. Why am I not surprised?
321
f0917727
PH
322PH/15 If the first argument of "${if match_address" was not empty, but did not
323 contain an "@" character, Exim crashed. Now it writes a panic log message
324 and treats the condition as false.
325
096fee00
PH
326PH/16 In autoreply, treat an empty string for "once" the same as unset.
327
024bd3c2
PH
328PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
329 extension "envelope-auth". The code is finished and in agreement with
330 other implementations, but there is no documentation so far and in fact,
331 nobody wrote the draft yet. This extension is currently #undef'ed, thus
332 not changing the active code.
333
334 Print executed "if" and "elsif" statements when debugging is used. This
335 helps a great deal to understand what a filter does.
336
337 Document more things not specified clearly in RFC3028. I had all this
338 sorted out, when out of a sudden new issues came to my mind. Oops."
339
df199fec
PH
340PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
341 (Bugzilla #53).
342
d27f1df3
PH
343PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
344 canonical form (as documented). However, after a host name lookup from
345 the IP address, check_host() was doing a simple string comparison with
346 addresses acquired from the DNS when checking that the found name did
347 have the original IP as one of its addresses. Since any found IPv6
348 addresses are likely to be in abbreviated form, the comparison could
349 fail. Luckily, there already exists a function for doing the comparison
350 by converting both addresses to binary, so now that is used instead of
351 the text comparison.
352
96776534
PH
353PH/20 There was another similar case to PH/19, when a complete host name was
354 given in a host list; looking up its IP address could give an abbreviated
355 form, whereas the current host's name might or might not be abbreviated.
356 The same fix has been applied.
357
5de37277 358
9a799bc0
PH
359Exim version 4.54
360-----------------
361
362PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
363 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
364 It now does.
365
99a4b039
PH
366PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
367 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
368
369PH/03 Typo: missing ".o" in src/pcre/Makefile.
370
4b233853
PH
371PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
372 header line, restrict the check to what is listed in RFCs 2369 and 2929.
373 Also, for "Auto-Submitted", treat anything other than "no" as
374 non-personal, in accordance with RFC 3834. (Previously it treated
375 anything starting "auto-" as non-personal.)
376
8857ccfd
PH
377TF/01 The control=submission/name=... option had a problem with syntax
378 errors if the name included a slash character. The /name= option
379 now slurps the rest of the string, so it can include any characters
380 but it must come last in the list of options (after /sender_retain
381 or /domain=).
382
433a2980
PH
383PH/05 Some modifications to the interface to the fake nameserver for the new
384 testing suite.
385
3e46c1aa 386
9a799bc0 387
e3a311ba
TK
388Exim version 4.53
389-----------------
390
391TK/01 Added the "success_on_redirect" address verification option. See
392 NewStuff for rationale and an example.
393
13b685f9
PH
394PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
395
395ff96d
PH
396PH/02 Patch to exigrep to allow it to work on syslog lines.
397
5b68f6e4
PH
398PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
399 fread() to skip over the body file's header line, because in Cygwin the
400 header line is locked and is inaccessible.
401
1ab52c69
PH
402PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
403 co-exist for some time) to make it clear that it is the Exim ID that is
404 referenced, not the Message-ID: header line.
405
b07e6aa3
PH
406PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
407 string_format() function, because snprintf() does not exist on all
408 operating systems.
409
254e032f
PH
410PH/06 The use of forbid_filter_existstest now also locks out the use of the
411 ${stat: expansion item.
412
3af76a81
PH
413PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
414 protocol synchronization error", to keep the pedants happy.
415
2548ba04
PH
416PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
417 well as for IRIX systems, when gcc is being used. See the host.c source
418 file for comments.
419
b6c6011d
PH
420PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
421
cf39cf57
PH
422PH/10 Named domain lists were not working if used in a queue_smtp_domains
423 setting.
424
f1513293
PH
425PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
426 transport and to the smtp transport in LMTP mode.
427
727549a4
PH
428TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
429
af46795e
PH
430PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
431 run a filter in a subprocess. This could lead to confusion in subsequent
432 lookups in the parent process. There should also be a search_tidyup() at
433 the end of the subprocess.
434
d7b47fd0
PH
435PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
436 only if the host matched helo_try_verify_hosts, which caused the
437 verification to occur when the EHLO/HELO command was issued. The ACL just
438 tested the remembered result. Now, if a previous verification attempt has
439 not happened, "verify = helo" does it there and then.
440
ee744174
JJ
441JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
442
b582ab87
PH
443TK/03 Fix log output including CR from clamd.
444
41a13e0a
PH
445PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
446 exist provoked a memory error which could cause a segfault.
447
f625cc5a
PH
448PH/15 Installed PCRE 6.2
449
450PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
451
21f7af35
PH
452PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
453 of the problem. Specifically, suggested +O2 rather than +O1 for the
454 HP-UX compiler.
455
31480e42
PH
456PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
457
2d280592
PH
458PH/20 If a delivery was routed to a non-standard port by means of an SRV
459 record, the port was not correctly logged when the outgoing_port log
460 selector was set (it logged the transort's default port).
461
7cd1141b
PH
462PH/21 Added support for host-specific ports to manualroute, queryprogram,
463 fallback_hosts, and "hosts" in the smtp transport.
464
465PH/22 If the log selector "outgoing_port" is set, the port is now also given on
466 host errors such as "Connection refused".
467
750af86e
PH
468PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
469 authentication with radiusclient 0.4.9:
470
471 - Error returned from rc_read_config was caught wrongly
472 - Username/password not passed on to radius server due to wrong length.
473
474 The presumption is that some radiusclient API changes for 4.51/PH/17
475 were not taken care of correctly. The code is still untested by me (my
476 Linux distribution still has 0.3.2 of radiusclient), but it was
477 contributed by a Radius user.
478
479PH/24 When doing a callout, the value of $domain wasn't set correctly when
480 expanding the "port" option of the smtp transport.
481
4304270b
TK
482TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
483 while reading a MIME header. Thanks to Tom Hughes for a patch.
484
750af86e
PH
485PH/24 Include config.h inside local_scan.h so that configuration settings are
486 available.
487
64ffc24f
PH
488PH/25 Make $smtp_command_argument available after all SMTP commands. This means
489 that in an ACL for RCPT (for example), you can examine exactly what was
490 received.
491
5dd9625b
PH
492PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
493 commands, but it was not correctly comparing the address with the actual
494 client host address. Thus, it would show the EHLO address in Received:
495 header lines when this was not necessary.
496
5591031b
PH
497PH/27 Added the % operator to ${eval:}.
498
ba18e66a
PH
499PH/28 Exim tries to create and chdir to its spool directory when it starts;
500 it should be ignoring failures (because with -C, for example, it has lost
501 privilege). It wasn't ignoring creation failures other than "already
502 exists".
503
9cec981f
PH
504PH/29 Added "crypteq" to the list of supported features that Exim outputs when
505 -bV or -d is used.
506
aa2b5c79
PH
507PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
508 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
509 too many continuations, the temporary file was not being removed, and the
510 return code was incorrect.
aa2b5c79 511
48a53b7f
PH
512PH/31 Missing "BOOL" in function definition in filtertest.c.
513
1c59d63b
PH
514PH/32 Applied Sieve patches from the maintainer.
515
671012da
TK
516TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
517
1509d3a8
PH
518PH/33 Added "verify = not_blind".
519
520PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
521 Local/Makefile (with some defaults set). These are used in built scripts
522 such as exicyclog, but they have never been used in the exim_install
523 script (though there are many overriding facilities there). I have
524 arranged that the exim_install script now takes note of these two
525 settings.
526
527PH/35 Installed configuration files for Dragonfly.
528
2fe1a124
PH
529PH/36 When a locally submitted message by a trusted user did not contain a
530 From: header, and the sender address was obtained from -f or from an SMTP
531 MAIL command, and the trusted user did not use -F to supply a sender
532 name, $originator_name was incorrectly used when constructing a From:
533 header. Furthermore, $originator_name was used for submission mode
534 messages from external hosts without From: headers in a similar way,
535 which is clearly wrong.
536
8800895a
PH
537PH/37 Added control=suppress_local_fixups.
538
ccfdb010
PH
539PH/38 When log_selector = +received_sender was set, and the addition of the
540 sender made the log line's construction buffer exactly full, or one byte
541 less than full, an overflow happened when the terminating "\n" was
542 subsequently added.
543
1130bfb0
PH
544PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
545 when the result of a list match is failure because a DNS lookup failed.
546
ebcb507f
PH
547PH/40 RM_COMMAND is now used in the building process.
548
c35e155c
PH
549PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
550 the "build-* directories that it finds.
551
95d1f782
PH
552PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
553 address in a domain literal was a prefix of an interface address.
554
fd6de02e
PH
555PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
556 when verifying a sender address, unless rewrite_headers is false.
557
58de37c5
PH
558PH/44 Wrote a long comment about why errors_to addresses are verified as
559 recipients, not senders.
560
261cf466
TF
561TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
562 the ratelimit ACL was added.
563
3ee512ff
PH
564PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
565
e08c430f
PH
566PH/46 Added extra information about PostgreSQL errors to the error string.
567
bef5a11f
PH
568PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
569 avoiding the need to install special zones in a real server. This is
570 backwards compatible; if it can't find the fake resolver, it drops back.
571 Thus, both old and new test suites can be run.
572
7546de58
TF
573TF/02 Added util/ratelimit.pl
574
e5d5a95f
TF
575TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
576 clock is set back in time.
577
2e88a017
TF
578TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
579 Candler <B.Candler@pobox.com>.
580
a5f65aa4
TF
581TF/05 The fix for PH/43 was not completely correct; widen_domains is always
582 OK for addresses that are the result of redirections.
583
e7726cbf
PH
584PH/48 A number of further additions for the benefit of the new test suite,
585 including a fake gethostbyname() that interfaces to the fake DNS resolver
586 (see PH/47 above).
587
a7fdad5b
TF
588TF/06 The fix for widen_domains has also been applied to qualify_single and
589 search_parents which are the other dnslookup options that can cause
590 header rewrites.
591
6af56900
PH
592PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
593 type ("H").
594
0925ede6
PH
595PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
596
66afa403
TF
597TF/07 Exim produced the error message "an SRV record indicated no SMTP
598 service" if it encountered an MX record with an empty target hostname.
599 The message is now "an MX or SRV record indicated no SMTP service".
600
0154e85a
TF
601TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
602 if the DNS of the sending site is misconfigured. This is quite a
603 common situation. This change restores the behaviour of treating a
604 helo verification defer as a failure.
605
16f12c76
PH
606PH/51 If self=fail was set on a router, the bounce message did not include the
607 actual error message.
608
bbe902f0 609
e5a9dba6
PH
610Exim version 4.52
611-----------------
612
613TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
614
22c3b60b
PH
615PH/01 When a transport filter timed out in a pipe delivery, and the pipe
616 command itself ended in error, the underlying message about the transport
617 filter timeout was being overwritten with the pipe command error. Now the
618 underlying error message should be appended to the second error message.
619
06a9b4b5
PH
620TK/01 Fix poll() being unavailable on Mac OSX 10.2.
621
c1ac6996
PH
622PH/02 Reduce the amount of output that "make" produces by default. Full output
623 can still be requested.
624
9c7a242c
PH
625PH/03 The warning log line about a condition test deferring for a "warn" verb
626 was being output only once per connection, rather than after each
627 occurrence (because it was using the same function as for successful
628 "warn" verbs). This seems wrong, so I have changed it.
629
87ba3f5f
PH
630TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
631 it should not have, which might have caused a crash in the right
632 circumstances, but probably never did.
633
634PH/04 Installed a modified version of Tony Finch's patch to make submission
635 mode fix the return path as well as the Sender: header line, and to
636 add a /name= option so that you can make the user's friendly name appear
637 in the header line.
638
29aba418
TF
639TF/03 Added the control = fakedefer ACL modifier.
640
fe0dab11
TF
641TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
642 Mark Lowes for thorough testing.
870f6ba8 643
11d337a4
TK
644TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
645
646TK/03 Merged latest SRS patch from Miles Wilton.
647
415c8f3b
PH
648PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
649 with the definition in sysexits.h (which is #included earlier).
650 Fortunately, Exim does not actually use EX_OK. The code used to try to
651 preserve the sysexits.h value, by assumimg that macro definitions were
652 scanned for macro replacements. I have been disabused of this notion,
653 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 654
958541e9
PH
655PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
656 in the smtp transport. When a block could not be written in a single
657 write() function, the timeout was being re-applied to each part-write.
658 This seems wrong - if the receiver was accepting one byte at a time it
659 would take for ever. The timeout is now adjusted when this happens. It
660 doesn't have to be particularly precise.
661
c206415f
TK
662TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
663 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
664
2a4be8f9
PH
665PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
666 without a domain if the check to <postmaster@domain> fails.
667
1cba11c5
SC
668SC/01 Eximstats: added -xls and the ability to specify output files
669 (patch written by Frank Heydlauf).
670
671SC/02 Eximstats: use FileHandles for outputing results.
672
673SC/03 Eximstats: allow any combination of xls, txt, and html output.
674
675SC/04 Eximstats: fixed display of large numbers with -nvr option
676
677SC/05 Eximstats: fixed merging of reports with empty tables.
678
679SC/06 Eximstats: added the -include_original_destination flag
680
681SC/07 Eximstats: removed tabs and trailing whitespace.
682
1005d00e
TK
683TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
684
685TK/06 MBOX spool code: Add real "From " MBOX separator line
686 so the .eml file is really in mbox format (even though
687 most programs do not really care). Patch from Alex Miller.
688
689TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
690 The latter is generated from $received_to and is only set if the
691 message has one envelope recipient. SA can use these headers,
692 obviously out-of-the-box. Patch from Alex Miller.
693
9b4768fa
PH
694PH/08 The ${def test on a variable was returning false if the variable's
695 value was "0", contrary to what the specification has always said!
696 The result should be true unless the variable is empty.
697
698PH/09 The syntax error of a character other than { following "${if
699 def:variable_name" (after optional whitespace) was not being diagnosed.
700 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
701 accidental colon was present, for example, could give incorrect results.
702
0d7eb84a
PH
703PH/10 Tidied the code in a number of places where the st_size field of a stat()
704 result is used (not including appendfile, where other changes are about
705 to be made).
706
707PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
708 This involved changing a lot of size variables from int to off_t. It
709 should work with maildirs and everything.
710
40727bee
TK
711TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
712 spamd dying while we are connected to it.
713
554d2369
TF
714TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
715 <jgh@wizmail.org>
716
1f922db1
PH
717PH/12 Applied Alex Kiernan's patch for the API change for the error callback
718 function for BDB 4.3.
719
ef213c3b
PH
720PH/13 Changed auto_thaw such that it does not apply to bounce messages.
721
8ac170f3
PH
722PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
723 the sources for PCRE have been re-arranged and more files are now
724 involved.
725
b1c749bb
PH
726PH/15 The code I had for printing potentially long long variables in PH/11
727 above was not the best (it lost precision). The length of off_t variables
728 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
729 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
730 to be "long long int" or "long int". This is needed for the internal
731 formatting function string_vformat().
b1c749bb 732
4aac9b49
PH
733PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
734 the configuration file to be ":syslog", then the script "guesses" where
735 the logs files are, rather than using the compiled in default. In our
736 case the guess is not the same as the compiled default, so the script
737 suddenly stopped working when I started to use syslog. The patch checks
738 to see if log_file_path is "". If so, it attempts to read it from exim
739 with no configuration file to get the compiled in version, before it
740 falls back to the previous guessing code."
741
294520c8
TK
742TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
743 implementing BATV in an Exim configuration. See NewStuff for the gory
744 details.
745
5bd022fe
PH
746PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
747 Makefile that are specific to HP-UX.
748
90e9ce59
PH
749PH/18 If the "use_postmaster" option was set for a recipient callout together
750 with the "random" option, the postmaster address was used as the MAIL
751 FROM address for the random test, but not for the subsequent recipient
752 test. It is now used for both.
753
5ea81592
PH
754PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
755 patch removes a few documentation additions to RFC 3028, because the
756 latest draft now contains them. It adds the new en;ascii-case comparator
757 and a new error check for 8bit text in MIME parts. Comparator and
758 require names are now matched exactly. I enabled the subaddress
759 extension, but it is not well tested yet (read: it works for me)."
760
c6c2dc1d
PH
761PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
762 rework some of the code of TK/09 above to avoid the hardwired use of
763 "%lld" and "long long". Replaced the call to snprintf() with a call to
764 string_vformat().
765
fffffe4c
PH
766PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
767 records point to non-existent hosts", "retry timeout exceeded", and
768 "retry time not reached for any host after a long failure period".
ca02eafb 769
9a26b6b2
PH
770PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
771 experimental DomainKeys support:
772
773 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
774 (2) On an error such as an illegally used "control", the wrong name for
775 the control was given.
776
777 These problems did NOT occur unless DomainKeys support was compiled.
778
4aee0225
PH
779PH/23 Added daemon_startup_retries and daemon_startup_sleep.
780
32d668a5
PH
781PH/24 Added ${if match_ip condition.
782
8187c3f3
PH
783PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
784 databases so that it will be absolutely obvious if a crash occurs in the
785 DB library. This is a regular occurrence (often caused by mis-matched
786 db.h files).
787
ff790e47 788PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
789 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
790 picked up on a user's system that detects such things. There doesn't seem
791 to be a gcc warning option for this - only an attribute that has to be
792 put on the function's prototype. It seems that in Fedora Core 4 they have
793 set this on a number of new functions. No doubt there will be more in due
794 course.
ff790e47 795
5417f6d1
PH
796PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
797 specify a transport. However, if an address that was verified by such a
798 router was the subject of a callout, Exim crashed because it tried to
799 read the rcpt_include_affixes from the non-existent transport. Now it
800 just assumes that the setting of that option is false. This bug was
801 introduced by 4.51/PH/31.
802
59cf8544
PH
803PH/28 Changed -d+all to exclude +memory, because that information is very
804 rarely of interest, but it makes the output a lot bigger. People tend to
805 do -d+all out of habit.
806
e7ad8a65
PH
807PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
808 code in os-type was giving problems when libc.so lives in lib64, like on
809 x86_64 Fedora Core.
810
ade42478
PH
811PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
812 aren't the modern standard, and it seems that some systems' include files
813 don't always have them. Exim was already checking for some of the newer
814 ones like T_AAAA, and defining it itself. I've added checks for all the
815 record types that Exim uses.
816
182ad5cf
PH
817PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
818 not automatically generating a new one, as it is supposed to. This
819 prevented TLS from working. If the file did exist, but contained invalid
820 data, a new version was generated, as expected. It was only the case of a
821 non-existent file that was broken.
822
b0d9fc80
TK
823TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
824 with a change in libdomainkeys > 0.64.
825
826TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
827 from DNS. If the selector record carries the flag, it now has
828 precedence over the domain-wide flag.
829
830TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
831
47c7a64a
PH
832PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
833 the use of a port name that isn't defined in /etc/services) occurred, the
834 message was deferred as in a normal delivery, and thus remained on the
835 spool, instead of being failed because of the mua_wrapper setting. This
836 is now fixed, and I tidied up some of the mua_wrapper messages at the
837 same time.
838
a388bce4
SC
839SC/08 Eximstats: whilst parsing the mainlog(s), store information about
840 the messages in a hash of arrays rather than using individual hashes.
841 This is a bit cleaner and results in dramatic memory savings, albeit
842 at a slight CPU cost.
843
844SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
845 as requested by Marc Sherman.
846
847SC/10 Eximstats: added histograms for user specified patterns as requested
848 by Marc Sherman.
849
0793e4ed
SC
850SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
851
c58b88df
PH
852PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
853 fopen() in the content-scanning modules that did not already have it.
854
e7ad8a65 855
7982096b
PH
856Exim version 4.51
857-----------------
858
1a46a8c5
PH
859TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
860 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
861
2f079f46 862TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
863
864TK/03 Fix ACL "regex" condition not being available in MIME ACL.
865
866PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
867 to test Sieve filters that use "vacation".
868
869PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
870 that changes the way the GnuTLS parameters are stored in the cache file.
871 The new format can be generated externally. For backward compatibility,
872 if the data in the cache doesn't make sense, Exim assumes it has read an
873 old-format file, and it generates new data and writes a new file. This
874 means that you can't go back to an older release without removing the
875 file.
876
877PH/03 A redirect router that has both "unseen" and "one_time" set does not
878 work if there are any delivery delays because "one_time" forces the
879 parent to be marked "delivered", so its unseen clone is never tried
880 again. For this reason, Exim now forbids the simultaneous setting of
881 these two options.
882
883PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
884 redirected to themselves ("homonym" addresses). Read the long ChangeLog
885 entry if you want to know the details. The fix, however, neglected to
886 consider the case when local delivery batching is involved. The test for
887 "previously delivered" was not happening when checking to see if an
888 address could be batched with a previous (undelivered) one; under
889 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 890 address.
1a46a8c5
PH
891
892PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
893 in its include files, and this causes problems building Exim.
894
895PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
896 header_syntax) but Exim was just ignoring anything given after a slash.
897 In particular, this caused confusion with an attempt to use "verify =
898 reverse_host_lookup/defer_ok". An error is now given when options are
899 supplied for verify items that do not have them. (Maybe reverse_host_
900 lookup should have a defer_ok option, but that's a different point.)
901
902PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
903 defined by RFC 821) to 2048, because there were problems with some AUTH
904 commands, and RFC 1869 says the size should be increased for extended
905 SMTP commands that take arguments.
906
907PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
908 Finch).
909
910PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
911 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 912
49c2d5ea
PH
913PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
914 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
915 whatever). Otherwise the password may appear in the log. Following change
916 PH/42 below, there is no longer a chance of it appearing in a bounce
917 message.
49c2d5ea 918
bf759a8b
PH
919PH/11 Installed exipick version 20050225.0 from John Jetmore.
920
83364d30
PH
921PH/12 If the last host in a fallback_hosts list was multihomed, only the first
922 of its addresses was ever tried. (Bugzilla bug #2.)
923
7999bbd7
PH
924PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
925 the result incorrectly in the debug output. (It correctly added a newline
926 to what was transported.)
927
7dbf77c9
PH
928TF/01 Added $received_time.
929
74e0617f
PH
930PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
931 commented out examples of how to interface to a virus scanner and to
932 SpamAssassin. Also added commented examples of av_scanner and
933 spamd_address settings.
934
2f079f46
PH
935PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
936 and controls are allowed in which ACLs. There were a couple of minor
937 errors. Some of the entries in the conditions table (which is a table of
938 where they are NOT allowed) were getting very unwieldy; rewrote them as a
939 negation of where the condition IS allowed.
940
8c841523
PH
941PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
942
7766a4f0
PH
943PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
944 header file does not have a version number, so I've had to invent a new
945 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
946 API. The code is untested by me (my Linux distribution still has 0.3.2 of
947 radiusclient), but it was contributed by a Radius user.
948
8b417f2c
PH
949PH/18 Installed Lars Mainka's patch for the support of CRL collections in
950 files or directories, for OpenSSL.
951
901f42cb
PH
952PH/19 When an Exim process that is running as root has to create an Exim log
953 file, it does so in a subprocess that runs as exim:exim so as to get the
954 ownership right at creation (otherwise, other Exim processes might see
955 the file with the wrong ownership). There was no test for failure of this
956 fork() call, which would lead to the process getting stuck as it waited
957 for a non-existent subprocess. Forks do occasionally fail when resources
958 run out. I reviewed all the other calls to fork(); they all seem to check
959 for failure.
960
f9b9210e
PH
961PH/20 When checking for unexpected SMTP input at connect time (before writing
962 the banner), Exim was not dealing correctly with a non-positive return
963 from the read() function. If the client had disconnected by this time,
964 the result was a log entry for a synchronization error with an empty
965 string after "input=" when read() returned zero. If read() returned -1
966 (an event I could not check), uninitialized data bytes were printed.
967 There were reports of junk text (parts of files, etc) appearing after
968 "input=".
969
54cdb463
PH
970PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
971
cf00dad6
PH
972PH/22 Added support for macro redefinition, and (re)definition in between
973 driver and ACL definitions.
974
acb1b346
PH
975PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
976 forgetting to use the resulting value; it was using the unexpanded value.
977
c5ddb310
PH
978PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
979 hadn't been configured. The fix is from Juergen Kreileder, who
980 understands it better than I do:
981
982 "Here's what I see happening with three configured cyrus_sasl
983 authenticators configured (plain, login, cram-md5):
984
985 On startup auth_cyrus_sasl_init() gets called for each of these.
986 This means three calls to sasl_listmech() without a specified mech_list.
987 => SASL tests which mechs of all available mechs actually work
988 => three warnings about OTP not working
989 => the returned list contains: plain, login, cram-md5, digest-md5, ...
990
991 With the patch, sasl_listmech() also gets called three times. But now
992 SASL's mech_list option is set to the server_mech specified in the the
993 authenticator. Or in other words, the answer from sasl_listmech()
994 gets limited to just the mech you're testing for (which is different
995 for each call.)
996 => the return list contains just 'plain' or 'login', 'cram-md5' or
997 nothing depending on the value of ob->server_mech.
998
999 I've just tested the patch: Authentication still works fine,
1000 unavailable mechs specified in the exim configuration are still
1001 caught, and the auth.log warnings about OTP are gone."
1002
31619da6
PH
1003PH/25 When debugging is enabled, the contents of the command line are added
1004 to the debugging output, even when log_selector=+arguments is not
1005 specified.
1006
bebaf0fc
PH
1007PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1008 answer is "GNU", and only if the return is "GNU/something" is the answer
1009 "Linux".
1010
475fe28a
PH
1011PH/27 $acl_verify_message is now set immediately after the failure of a
1012 verification in an ACL, and so is available in subsequent modifiers. In
1013 particular, the message can be preserved by coding like this:
1014
1015 warn !verify = sender
1016 set acl_m0 = $acl_verify_message
1017
1018 Previously, $acl_verify_message was set only while expanding "message"
1019 and "log_message" when a very denied access.
1020
7e8bec7a
PH
1021PH/28 Modified OS/os.c-Linux with
1022
1023 -#ifndef OS_LOAD_AVERAGE
1024 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1025
1026 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1027 nomenclature these days.)
1028
e4a89c47
PH
1029PH/29 Installed patch from the Sieve maintainer that adds the options
1030 sieve_useraddress and sieve_subaddress to the redirect router.
1031
5ca2a9a1
PH
1032PH/30 In these circumstances:
1033 . Two addresses routed to the same list of hosts;
1034 . First host does not offer TLS;
1035 . First host accepts first address;
1036 . First host gives temporary error to second address;
1037 . Second host offers TLS and a TLS session is established;
1038 . Second host accepts second address.
1039 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1040 and peerdn, if requested) that were in fact used only for the second
1041 address.
7e8bec7a 1042
c688b954
PH
1043PH/31 When doing a callout as part of verifying an address, Exim was not paying
1044 attention to any local part prefix or suffix that was matched by the
1045 router that accepted the address. It now behaves in the same way as it
1046 does for delivery: the affixes are removed from the local part unless
1047 rcpt_include_affixes is set on the transport.
1048
fed77020
PH
1049PH/32 Add the sender address, as F=<...>, to the log line when logging a
1050 timeout during the DATA phase of an incoming message.
1051
7fe1560f
PH
1052PH/33 Sieve envelope tests were broken for match types other than :is. I have
1053 applied a patch sanctioned by the Sieve maintainer.
c688b954 1054
ebb6e6d5
PH
1055PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1056 the uid or gid is negative. A case of a negative gid caused this to be
1057 noticed. The fix allows for either to be negative.
1058
9c4e8f60
PH
1059PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1060 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1061 overlooked.
1062
1063PH/36 The change PH/12 above was broken. Fixed it.
1064
d7174846
PH
1065PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1066 the grounds that routing the same address twice would always produce the
1067 same answer. This might have been true once, but it is certainly no
1068 longer true now. Routing a child address may depend on the previous
1069 routing that produced that child. Some complicated redirection strategies
1070 went wrong when messages had multiple recipients, and made Exim's
1071 behaviour dependent on the order in which the addresses were given.
1072
1073 I have moved the duplicate checking until after the routing is complete.
1074 Exim scans the addresses that are assigned to local and remote
1075 transports, and removes any duplicates. This means that more work will be
1076 done, as duplicates will always all be routed, but duplicates are
1077 presumably rare, so I don't expect this is of any significance.
1078
1079 For deliveries to pipes, files, and autoreplies, the duplicate checking
1080 still happens during the routing process, since they are not going to be
1081 routed further.
1082
cfe75fc3
PH
1083PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1084 It corrects a timeout issue with spamd. This is Ian's comment: "The
1085 background is that sometimes spamd either never reads data from a
1086 connection it has accepted, or it never writes response data. The exiscan
1087 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1088 blindly assumes that writes won't block so it may never time out."
1089
be22d70e
PH
1090PH/39 Allow G after quota size as well as K and M.
1091
0612b098
PH
1092PH/40 The value set for $authenticated_id in an authenticator may not contain
1093 binary zeroes or newlines because the value is written to log lines and
1094 to spool files. There was no check on this. Now the value is run through
1095 the string_printing() function so that such characters are converted to
1096 printable escape sequences.
1097
2e0c1448
PH
1098PH/41 $message_linecount is a new variable that contains the total number of
1099 lines in the message. Compare $body_linecount, which is the count for the
1100 body only.
1101
447d236c
PH
1102PH/42 Exim no longer gives details of delivery errors for specific addresses in
1103 bounce and delay warning messages, except in certain special cases, which
1104 are as follows:
1105
1106 (a) An SMTP error message from a remote host;
1107 (b) A message specified in a :fail: redirection;
1108 (c) A message specified in a "fail" command in a system filter;
1109 (d) A message specified in a FAIL return from the queryprogram router;
1110 (e) A message specified by the cannot_route_message router option.
1111
1112 In these cases only, Exim does include the error details in bounce and
1113 warning messages. There are also a few cases where bland messages such
1114 as "unrouteable address" or "local delivery error" are given.
1115
d20976dc
PH
1116PH/43 $value is now also set for the "else" part of a ${run expansion.
1117
f656d135
PH
1118PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1119 being worked on, but at least Exim now implements the latest version to
1120 play with."
1121
2e2a30b4
PH
1122PH/45 In a pipe transport, although a timeout while waiting for the pipe
1123 process to complete was treated as a delivery failure, a timeout while
1124 writing the message to the pipe was logged, but erroneously treated as a
1125 successful delivery. Such timeouts include transport filter timeouts. For
1126 consistency with the overall process timeout, these timeouts are now
1127 treated as errors, giving rise to delivery failures by default. However,
1128 there is now a new Boolean option for the pipe transport called
1129 timeout_defer, which, if set TRUE, converts the failures into defers for
1130 both kinds of timeout. A transport filter timeout is now identified in
1131 the log output.
1132
9176e9f0
PH
1133PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1134 systems where "make" and "gmake" are different, calling "gmake" at top
1135 level broke things. I've arranged for the value of $(MAKE) to be passed
1136 from the Makefile to this script so that it can call the same version of
1137 "make".
1138
7982096b 1139
bbe902f0
PH
1140A note about Exim versions 4.44 and 4.50
1141----------------------------------------
1142
1143Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1144changes of various kinds. As a consequence, a big documentation update was
1145needed. This delayed the release for rather longer than seemed good, especially
1146in the light of a couple of (minor) security issues. Therefore, the changes
1147that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1148release. So 4.44 and 4.50 are in effect two different branches that both start
1149from 4.43.
1150
1151I have left the 4.50 change log unchanged; it contains all the changes since
11524.43. The change log for 4.44 is below; many of its items are identical to
1153those for 4.50. This seems to be the most sensible way to preserve the
1154historical information.
1155
1156
f7b63901 1157Exim version 4.50
495ae4b0
PH
1158-----------------
1159
5fe762f6
PH
1160 1. Minor wording change to the doc/README.SIEVE file.
1161
139059f6 1162 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1163 computation of the current number of files was incorrect.
495ae4b0 1164
7086e875
PH
1165 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1166 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1167 place.
1168
35af9f61
PH
1169 4. Give more explanation in the error message when the command for a transport
1170 filter fails to execute.
1171
b668c215
PH
1172 5. There are several places where Exim runs a non-Exim command in a
1173 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1174 was being done only for the command run by the queryprogram router. It is
1175 now done for all such subprocesses. The other cases are: ${run, transport
1176 filters, and the commands run by the lmtp and pipe transports.
1177
a494b1e1
PH
1178 6. Added CONFIGURE_GROUP build-time option.
1179
1180 7. Some older OS have a limit of 256 on the maximum number of file
1181 descriptors. Exim was using setrlimit() to set 1000 as a large value
1182 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1183 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1184
c5fcb476
PH
1185 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1186 was an oversight, and furthermore, ever since the addition of extra
1187 controls (e.g. 4.43/32), the checks on when to allow different forms of
1188 "control" were broken. There should now be diagnostics for all cases when a
1189 control that does not make sense is encountered.
1190
69358f02
PH
1191 9. Added the /retain_sender option to "control=submission".
1192
5be20824
PH
119310. $recipients is now available in the predata ACL (oversight).
1194
eb2c0248
PH
119511. Tidy the search cache before the fork to do a delivery from a message
1196 received from the command line. Otherwise the child will trigger a lookup
1197 failure and thereby defer the delivery if it tries to use (for example) a
1198 cached ldap connection that the parent has called unbind on.
1199
2a3eea10
PH
120012. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1201 of $address_data from the recipient verification was clobbered by the
1202 sender verification.
1203
120413. The value of address_data from a sender verification is now available in
1205 $sender_address_data in subsequent conditions in the ACL statement.
1206
23c7ff99
PH
120714. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1208
4deaf07d
PH
120915. Added a new option "connect=<time>" to callout options, to set a different
1210 connection timeout.
1211
926e1192
PH
121216. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1213 was its contents. (It was OK if the option was not defined at all.)
1214
650edc6f
PH
121517. A "Completed" log line is now written for messages that are removed from
1216 the spool by the -Mrm option.
1217
2c7db3f5
PH
121818. New variables $sender_verify_failure and $recipient_verify_failure contain
1219 information about exactly what failed.
1220
3d235903
PH
122119. Added -dd to debug only the daemon process.
1222
7c7ad977
PH
122320. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1224 handles timeouts, both on the server side and network timeouts. Renamed the
1225 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1226
981756db
PH
122721. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1228 It is now set to "smtps".
1229
d4eb88df
PH
123022. $host_address is now set to the target address during the checking of
1231 ignore_target_hosts.
1232
123323. When checking ignore_target_hosts for an ipliteral router, no host name was
1234 being passed; this would have caused $sender_host_name to have been used if
1235 matching the list had actually called for a host name (not very likely,
1236 since this list is usually IP addresses). A host name is now passed as
1237 "[x.x.x.x]".
1238
7d468ab8
PH
123924. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1240 code that specifies a non-restarting handler (typically sigaction() in
1241 modern systems) in an attempt to fix a rare and obscure crash bug.
1242
124325. Narrowed the window for a race in the daemon that could cause it to ignore
1244 SIGCHLD signals. This is not a major problem, because they are used only to
1245 wake it up if nothing else does.
1246
62c0818f
PH
124726. A malformed maildirsize file could cause Exim to calculate negative values
1248 for the mailbox size or file count. Odd effects could occur as a result.
1249 The maildirsize information is now recalculated if the size or filecount
1250 end up negative.
1251
26034054
PH
125227. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1253 support for a long time. Removed HAVE_SYS_VFS_H.
1254
af66f652
PH
125528. Installed the latest version of exipick from John Jetmore.
1256
90af77f4
PH
125729. In an address list, if the pattern was not a regular expression, an empty
1258 subject address (from a bounce message) matched only if the pattern was an
1259 empty string. Non-empty patterns were not even tested. This was the wrong
1260 because it is perfectly reasonable to use an empty address as part of a
1261 database query. An empty address is now tested by patterns that are
1262 lookups. However, all the other forms of pattern expect the subject to
1263 contain a local part and a domain, and therefore, for them, an empty
1264 address still always fails if the pattern is not itself empty.
1265
d8ef3577
PH
126630. Exim went into a mad DNS loop when attempting to do a callout where the
1267 host was specified on an smtp transport, and looking it up yielded more
1268 than one IP address.
1269
5cb8cbc6
PH
127031. Re-factored the code for checking spool and log partition space into a
1271 function that finds that data and another that does the check. The former
1272 is then used to implement four new variables: $spool_space, $log_space,
1273 $spool_inodes, and $log_inodes.
1274
14702f5b
PH
127532. The RFC2047 encoding function was originally intended for short strings
1276 such as real names; it was not keeping to the 75-character limit for
1277 encoded words that the RFC imposes. It now respects the limit, and
1278 generates multiple encoded words if necessary. To be on the safe side, I
1279 have increased the buffer size for the ${rfc2047: expansion operator from
1280 1024 to 2048 bytes.
1281
063b1e99
PH
128233. It is now permitted to omit both strings after an "if" condition; if the
1283 condition is true, the result is "true". As before, when the second string
1284 is omitted, a false condition yields an empty string. This makes it less
1285 cumbersome to write custom ACL and router conditions.
1286
652e1b65
PH
128734. Failure to deliver a bounce message always caused it to be frozen, even if
1288 there was an errors_to setting on the router. The errors_to setting is now
1289 respected.
1290
6f0c9a4f
PH
129135. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1292 canonical form (fully expanded) before being placed in
1293 $sender_host_address.
1294
33397d19
PH
129536. The table in the code that translates DNS record types into text (T_A to
1296 "A" for instance) was missing entries for NS and CNAME. It is just possible
1297 that this could have caused confusion if both these types were looked up
1298 for the same domain, because the text type is used as part of Exim's
1299 per-process caching. But the chance of anyone hitting this buglet seems
1300 very small.
1301
7bb56e1f
PH
130237. The dnsdb lookup has been extended in a number of ways.
1303
1304 (1) There is a new type, "zns", which walks up the domain tree until it
1305 finds some nameserver records. It should be used with care.
1306
ea3bc19b
PH
1307 (2) There is a new type, "mxh", which is like "mx" except that it returns
1308 just the host names, not the priorities.
1309
1310 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1311 looked up. The behaviour when one of the lookups defers can be
1312 controlled by a keyword.
7bb56e1f 1313
ea3bc19b 1314 (4) It is now possible to specify the separator character for use when
7bb56e1f 1315 multiple records are returned.
33397d19 1316
0bcb2a0e
PH
131738. The dnslists ACL condition has been extended: it is now possible to supply
1318 a list of IP addresses and/or domains to be looked up in a particular DNS
1319 domain.
1320
2ac0e484
PH
132139. Added log_selector=+queue_time_overall.
1322
4e1fde53
PH
132340. When running the queue in the test harness, wait just a tad after forking a
1324 delivery process, to get repeatability of debugging output.
1325
de365ded
PH
132641. Include certificate and key file names in error message when GnuTLS fails
1327 to set them up, because the GnuTLS error message doesn't include the name
1328 of the failing file when there is a problem reading it.
1329
f05da2e8
PH
133042. Allow both -bf and -bF in the same test run.
1331
d6453af2
PH
133243. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1333
f7b63901
PH
133444. The "Exiscan patch" is now merged into the mainline Exim source.
1335
133645. Sometimes the final signoff response after QUIT could fail to get
1337 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1338 < 0 before doing a fflush(). This bug looks as though it goes back to the
1339 introduction of TLS in release 3.20, but "sometimes" must have been rare
1340 because the tests only now provoked it.
1341
a444213a
PH
134246. Reset the locale to "C" after calling embedded Perl, in case it was changed
1343 (this can affect the format of dates).
1344
0ec020ea
PH
134547. exim_tidydb, when checking for the continued existence of a message for
1346 which it has found a message-specific retry record, was not finding
1347 messages that were in split spool directories. Consequently, it was
1348 deleting retry records that should have stayed in existence.
1349
b1206957
PH
135048. Steve fixed some bugs in eximstats.
1351
135249. The SPA authentication driver was not abandoning authentication and moving
1353 on to the next authenticator when an expansion was forced to fail,
1354 contradicting the general specification for all authenticators. Instead it
1355 was generating a temporary error. It now behaves as specified.
1356
26dd5a95
PH
135750. The default ordering of permitted cipher suites for GnuTLS was pessimal
1358 (the order specifies the preference for clients). The order is now AES256,
1359 AES128, 3DES, ARCFOUR128.
1360
343b2385
PH
136151. Small patch to Sieve code - explicitly set From: when generating an
1362 autoreply.
1363
1c5466b9
PH
136452. Exim crashed if a remote delivery caused a very long error message to be
1365 recorded - for instance if somebody sent an entire SpamAssassin report back
1366 as a large number of 550 error lines. This bug was coincidentally fixed by
1367 increasing the size of one of Exim's internal buffers (big_buffer) that
1368 happened as part of the Exiscan merge. However, to be on the safe side, I
1369 have made the code more robust (and fixed the comments that describe what
1370 is going on).
1371
55ee9ee3
PH
137253. Now that there can be additional text after "Completed" in log lines (if
1373 the queue_time_overall log selector is set), a one-byte patch to exigrep
1374 was needed to allow it to recognize "Completed" as not the last thing in
1375 the line.
1376
d38f8232
PH
137754. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1378 patch that reportedly fixes this has been added. I am not expert enough to
1379 create a test for it. This is what the patch creator wrote:
1380
1381 "I found a little strange behaviour of ldap code when working with
1382 Windows 2003 AD Domain, where users was placed in more than one
1383 Organization Units. When I tried to give exim partial DN, the exit code
1384 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1385 But simultaneously result of request was absolutely normal ldap result,
1386 so I produce this patch..."
1387
3295e65b
PH
1388 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1389 REFERENCE, so I have modified the code to exclude the patch when that macro
1390 is not defined.
1391
7102e136
PH
139255. Some experimental protocols are using DNS PTR records for new purposes. The
1393 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1394 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1395 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1396 CAN-2005-0021
7102e136 1397
3ca0ba97
PH
139856. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1399
c2bcbe20
PH
140057. Double the size of the debug message buffer (to 2048) so that more of very
1401 long debug lines gets shown.
1402
18ce445d
PH
140358. The exicyclog utility now does better if the number of log files to keep
1404 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1405
1f5b4c3d
PH
140659. Two changes related to the smtp_active_hostname option:
1407
1408 (1) $smtp_active_hostname is now available as a variable.
1409 (2) The default for smtp_banner uses $smtp_active_hostname instead
1410 of $primary_hostname.
1411
b975ba52
PH
141260. The host_aton() function is supposed to be passed a string that is known
1413 to be a valid IP address. However, in the case of IPv6 addresses, it was
1414 not checking this. This is a hostage to fortune. Exim now panics and dies
1415 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1416 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1417 components; fortuitously, this particular loophole had already been fixed
1418 by change 4.50/55 above.
1419
1420 If there are any other similar loopholes, the new check in host_aton()
1421 itself should stop them being exploited. The report I received stated that
1422 data on the command line could provoke the exploit when Exim was running as
1423 exim, but did not say which command line option was involved. All I could
1424 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1425 running as the user.
ea3a6f44 1426 CAN-2005-0021
85b87bc2
PH
1427
142861. There was a buffer overflow vulnerability in the SPA authentication code
1429 (which came originally from the Samba project). I have added a test to the
1430 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1431 CAN-2005-0022
b975ba52 1432
17ffcae7
PH
143362. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1434 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1435
d95f9fdb
PH
143663. The daemon start-up calls getloadavg() while still root for those OS that
1437 need the first call to be done as root, but it missed one case: when
1438 deliver_queue_load_max is set with deliver_drop_privilege. This is
1439 necessary for the benefit of the queue runner, because there is no re-exec
1440 when deliver_drop_privilege is set.
1441
86b8287f
PH
144264. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1443 This has been fixed.
1444
60dc5e56
PH
144565. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1446 was in use, was not putting the data itself into the right store pool;
1447 consequently, it could be overwritten for a subsequent message in the same
1448 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1449 the caching.)
1450
533244af
PH
145166. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1452
a5a28604
PH
145367. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1454 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1455 the function were treating the return as a boolean value, which happened to
1456 work because 0=false and not-0=true, but is not correct code.
1457
7e634d24
PH
145868. The host_aton() function was not handling scoped IPv6 addresses (those
1459 with, for example, "%eth0" on the end) correctly.
1460
3e11c26b
PH
146169. Fixed some compiler warnings in acl.c for the bitmaps specified with
1462 negated items (that is, ~something) in unsigned ints. Some compilers
1463 apparently mutter when there is no cast.
1464
6729cf78
PH
146570. If an address verification called from an ACL failed, and did not produce a
1466 user-specific message (i.e. there was only a "system" message), nothing was
1467 put in $acl_verify_message. In this situation, it now puts the system
1468 message there.
1469
00f00ca5
PH
147071. Change 4.23/11 added synchronization checking at the start of an SMTP
1471 session; change 4.31/43 added the unwanted input to the log line - except
1472 that it did not do this in the start of session case. It now does.
1473
c9bdd01c
PH
147472. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1475 This is wrong and can cause the other end to generate a synchronization
1476 error if it is another Exim or anything else that does the synchronization
1477 check. A QUIT command is no longer sent after a timeout.
1478
d43194df
PH
147973. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1480 during host lookups.
1481
fe5b5d0b
PH
148274. The defer_ok option of callout verification was not working if it was used
1483 when verifying addresses in header lines, that is, for this case:
1484
1485 verify = header_sender/callout=defer_ok
1486
76a2d7ba
PH
148775. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1488 those file descriptors could be used for SMTP connections. If anything
1489 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1490 could be sent to the SMTP client, causing chaos. The daemon now opens
1491 stdin, stdout, and stderr to /dev/null when it puts itself into the
1492 background.
1493
149476. Arrange for output from Perl's "warn" command to be written to Exim's main
1495 log by default. The user can override this with suitable Perl magic.
1496
04f7d5b9
PH
149777. The use of log_message on a "discard" ACL verb, which is supposed to add to
1498 the log message when discard triggers, was not working for the DATA ACL or
1499 for the non-SMTP ACL.
1500
bc60667e
PH
150178. Error message wording change in sieve.c.
1502
bb6e88ff
PH
150379. If smtp_accept_max_per_host was set, the number of connections could be
1504 restricted to fewer than expected, because the daemon was trying to set up
1505 a new connection before checking whether the processes handling previous
1506 connections had finished. The check for completed processes is now done
1507 earlier. On busy systems, this bug wouldn't be noticed because something
1508 else would have woken the daemon, and it would have reaped the completed
1509 process earlier.
1510
1e70f85b
PH
151180. If a message was submitted locally by a user whose login name contained one
1512 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1513 It caused a spool format error. I have fixed the spool reading code. A
1514 related problem was that the "from" clause in the Received: line became
1515 illegal because of the space(s). It is now covered by ${quote_local_part.
1516
151781. Included the latest eximstats from Steve (adds average sizes to HTML Top
1518 tables).
1519
4e01f9d6
PH
152082. Updated OS/Makefile-AIX as per message from Mike Meredith.
1521
1ee1cef2
PH
152283. Patch from Sieve maintainer to fix unterminated string problem in
1523 "vacation" handling.
1524
6e2b4ccc
PH
152584. Some minor changes to the Linux configuration files to help with other
1526 OS variants using glibc.
1527
8e669ac1
PH
152885. One more patch for Sieve to update vacation handling to latest spec.
1529
495ae4b0 1530
bbe902f0
PH
1531----------------------------------------------------
1532See the note above about the 4.44 and 4.50 releases.
1533----------------------------------------------------
1534
1535
1536Exim version 4.44
1537-----------------
1538
1539 1. Change 4.43/35 introduced a bug that caused file counts to be
1540 incorrectly computed when quota_filecount was set in an appendfile
1541 transport
1542
1543 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1544 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1545 place.
1546
1547 3. Give more explanation in the error message when the command for a transport
1548 filter fails to execute.
1549
1550 4. There are several places where Exim runs a non-Exim command in a
1551 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1552 was being done only for the command run by the queryprogram router. It is
1553 now done for all such subprocesses. The other cases are: ${run, transport
1554 filters, and the commands run by the lmtp and pipe transports.
1555
1556 5. Some older OS have a limit of 256 on the maximum number of file
1557 descriptors. Exim was using setrlimit() to set 1000 as a large value
1558 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1559 systems. I've change it so that if it can't get 1000, it tries for 256.
1560
1561 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1562 was an oversight, and furthermore, ever since the addition of extra
1563 controls (e.g. 4.43/32), the checks on when to allow different forms of
1564 "control" were broken. There should now be diagnostics for all cases when a
1565 control that does not make sense is encountered.
1566
1567 7. $recipients is now available in the predata ACL (oversight).
1568
1569 8. Tidy the search cache before the fork to do a delivery from a message
1570 received from the command line. Otherwise the child will trigger a lookup
1571 failure and thereby defer the delivery if it tries to use (for example) a
1572 cached ldap connection that the parent has called unbind on.
1573
1574 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1575 of $address_data from the recipient verification was clobbered by the
1576 sender verification.
1577
157810. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1579 was its contents. (It was OK if the option was not defined at all.)
1580
158111. A "Completed" log line is now written for messages that are removed from
1582 the spool by the -Mrm option.
1583
158412. $host_address is now set to the target address during the checking of
1585 ignore_target_hosts.
1586
158713. When checking ignore_target_hosts for an ipliteral router, no host name was
1588 being passed; this would have caused $sender_host_name to have been used if
1589 matching the list had actually called for a host name (not very likely,
1590 since this list is usually IP addresses). A host name is now passed as
1591 "[x.x.x.x]".
1592
159314. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1594 code that specifies a non-restarting handler (typically sigaction() in
1595 modern systems) in an attempt to fix a rare and obscure crash bug.
1596
159715. Narrowed the window for a race in the daemon that could cause it to ignore
1598 SIGCHLD signals. This is not a major problem, because they are used only to
1599 wake it up if nothing else does.
1600
160116. A malformed maildirsize file could cause Exim to calculate negative values
1602 for the mailbox size or file count. Odd effects could occur as a result.
1603 The maildirsize information is now recalculated if the size or filecount
1604 end up negative.
1605
160617. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1607 support for a long time. Removed HAVE_SYS_VFS_H.
1608
ea3a6f44 160918. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1610
161119. Allow an empty sender to be matched against a lookup in an address list.
1612 Previously the only cases considered were a regular expression, or an
1613 empty pattern.
1614
161520. Exim went into a mad DNS lookup loop when doing a callout where the
1616 host was specified on the transport, if the DNS lookup yielded more than
1617 one IP address.
1618
ea3a6f44
NM
161921. The RFC2047 encoding function was originally intended for short strings
1620 such as real names; it was not keeping to the 75-character limit for
1621 encoded words that the RFC imposes. It now respects the limit, and
1622 generates multiple encoded words if necessary. To be on the safe side, I
1623 have increased the buffer size for the ${rfc2047: expansion operator from
1624 1024 to 2048 bytes.
bbe902f0 1625
ea3a6f44
NM
162622. Failure to deliver a bounce message always caused it to be frozen, even if
1627 there was an errors_to setting on the router. The errors_to setting is now
1628 respected.
bbe902f0
PH
1629
163023. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1631 canonical form (fully expanded) before being placed in
1632 $sender_host_address.
1633
163424. Updated eximstats to version 1.33
1635
ea3a6f44
NM
163625. Include certificate and key file names in error message when GnuTLS fails
1637 to set them up, because the GnuTLS error message doesn't include the name
1638 of the failing file when there is a problem reading it.
bbe902f0
PH
1639
164026. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1641 As per change 25.
bbe902f0 1642
ea3a6f44
NM
164327. Reset the locale to "C" after calling embedded Perl, in case it was changed
1644 (this can affect the format of dates).
bbe902f0 1645
ea3a6f44
NM
164628. exim_tidydb, when checking for the continued existence of a message for
1647 which it has found a message-specific retry record, was not finding
1648 messages that were in split spool directories. Consequently, it was
1649 deleting retry records that should have stayed in existence.
bbe902f0
PH
1650
165129. eximstats updated to version 1.35
1652 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1653 1.35 - bugfix such that pie charts by volume are generated correctly
1654
ea3a6f44
NM
165530. The SPA authentication driver was not abandoning authentication and moving
1656 on to the next authenticator when an expansion was forced to fail,
1657 contradicting the general specification for all authenticators. Instead it
1658 was generating a temporary error. It now behaves as specified.
bbe902f0 1659
ea3a6f44
NM
166031. The default ordering of permitted cipher suites for GnuTLS was pessimal
1661 (the order specifies the preference for clients). The order is now AES256,
1662 AES128, 3DES, ARCFOUR128.
bbe902f0 1663
ea3a6f44
NM
166431. Small patch to Sieve code - explicitly set From: when generating an
1665 autoreply.
bbe902f0 1666
ea3a6f44
NM
166732. Exim crashed if a remote delivery caused a very long error message to be
1668 recorded - for instance if somebody sent an entire SpamAssassin report back
1669 as a large number of 550 error lines. This bug was coincidentally fixed by
1670 increasing the size of one of Exim's internal buffers (big_buffer) that
1671 happened as part of the Exiscan merge. However, to be on the safe side, I
1672 have made the code more robust (and fixed the comments that describe what
1673 is going on).
bbe902f0 1674
ea3a6f44
NM
167533. Some experimental protocols are using DNS PTR records for new purposes. The
1676 keys for these records are domain names, not reversed IP addresses. The
1677 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1678 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1679 CAN-2005-0021
1680
ea3a6f44
NM
168134. The host_aton() function is supposed to be passed a string that is known
1682 to be a valid IP address. However, in the case of IPv6 addresses, it was
1683 not checking this. This is a hostage to fortune. Exim now panics and dies
1684 if the condition is not met. A case was found where this could be provoked
1685 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1686 components; fortuitously, this particular loophole had already been fixed
1687 by change 4.50/55 or 4.44/33 above.
1688
1689 If there are any other similar loopholes, the new check in host_aton()
1690 itself should stop them being exploited. The report I received stated that
1691 data on the command line could provoke the exploit when Exim was running as
1692 exim, but did not say which command line option was involved. All I could
1693 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1694 running as the user.
bbe902f0
PH
1695 CAN-2005-0021
1696
ea3a6f44
NM
169735. There was a buffer overflow vulnerability in the SPA authentication code
1698 (which came originally from the Samba project). I have added a test to the
1699 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1700 CAN-2005-0022
1701
ea3a6f44
NM
170236. The daemon start-up calls getloadavg() while still root for those OS that
1703 need the first call to be done as root, but it missed one case: when
1704 deliver_queue_load_max is set with deliver_drop_privilege. This is
1705 necessary for the benefit of the queue runner, because there is no re-exec
1706 when deliver_drop_privilege is set.
bbe902f0 1707
ea3a6f44
NM
170837. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1709 was in use, was not putting the data itself into the right store pool;
1710 consequently, it could be overwritten for a subsequent message in the same
1711 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1712 the caching.)
bbe902f0 1713
ea3a6f44
NM
171438. Sometimes the final signoff response after QUIT could fail to get
1715 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1716 < 0 before doing a fflush(). This bug looks as though it goes back to the
1717 introduction of TLS in release 3.20, but "sometimes" must have been rare
1718 because the tests only now provoked it.
bbe902f0
PH
1719
1720
495ae4b0
PH
1721Exim version 4.43
1722-----------------
1723
1724 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1725 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1726 Now it returns an integer. A number of calls were still expecting a T/F
1727 return. Fortuitously, in all cases, the tests worked in OK situations,
1728 which is the norm. However, things would have gone wrong on any write
1729 failures on the smtp file descriptor. This function is used when sending
1730 messages over SMTP and also when doing verify callouts.
1731
1732 2. When Exim is called to do synchronous delivery of a locally submitted
1733 message (the -odf or -odi options), it no longer closes stderr before doing
1734 the delivery.
1735
1736 3. Implemented the mua_wrapper option.
1737
1738 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1739
1740 5. Implemented the functions header_remove(), header_testname(),
1741 header_add_at_position(), and receive_remove_recipient(), and exported them
1742 to local_scan().
1743
1744 6. If an ACL "warn" statement specified the addition of headers, Exim already
1745 inserted X-ACL-Warn: at the start if there was no header name. However, it
1746 was not making this test for the second and subsequent header lines if
1747 there were newlines in the string. This meant that an invalid header could
1748 be inserted if Exim was badly configured.
1749
1750 7. Allow an ACL "warn" statement to add header lines at the start or after all
1751 the Received: headers, as well as at the end.
1752
1753 8. Added the rcpt_4xx retry error code.
1754
1755 9. Added postmaster_mailfrom=xxx to callout verification option.
1756
175710. Added mailfrom=xxxx to the callout verification option, for verify=
1758 header_sender only.
1759
176011. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1761 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1762
176312. Inserted some casts to stop certain compilers warning when using pointer
1764 differences as field lengths or precisions in printf-type calls (mostly
1765 affecting debugging statements).
1766
176713. Added optional readline() support for -be (dynamically loaded).
1768
176914. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1770 same clock tick as a message's arrival, so that its received time was the
1771 same as the "first fail" time on the retry record, and that message
1772 remained on the queue past the ultimate address timeout, every queue runner
1773 would try a delivery (because it was past the ultimate address timeout) but
1774 after another failure, the ultimate address timeout, which should have then
1775 bounced the address, did not kick in. This was a "< instead of <=" error;
1776 in most cases the first failure would have been in the next clock tick
1777 after the received time, and all would be well.
1778
177915. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1780 being recognized when the domain list was tested by the match_domain
1781 condition in an expansion string.
1782
178316. Added the ${str2b64: operator.
1784
178517. Exim was always calling setrlimit() to set a large limit for the number of
1786 processes, without checking whether the existing limit was already
1787 adequate. (It did check for the limit on file descriptors.) Furthermore,
1788 errors from getrlimit() and setrlimit() were being ignored. Now they are
1789 logged to the main and panic logs, but Exim does carry on, to try to do its
1790 job under whatever limits there are.
1791
179218. Imported PCRE 5.0.
1793
179419. Trivial typo in log message " temporarily refused connection" (the leading
1795 space).
1796
179720. If the log selector return_path_on_delivery was set and an address was
1798 redirected to /dev/null, the delivery process crashed because it assumed
1799 that a return path would always be set for a "successful" delivery. In this
1800 case, the whole delivery is bypassed as an optimization, and therefore no
1801 return path is set.
1802
180321. Internal re-arrangement: the function for sending a challenge and reading
1804 a response while authentication was assuming a zero-terminated challenge
1805 string. It's now changed to take a pointer and a length, to allow for
1806 binary data in such strings.
1807
180822. Added the cyrus_sasl authenticator (code supplied by MBM).
1809
181023. Exim was not respecting finduser_retries when seeking the login of the
1811 uid under which it was called; it was always trying 10 times. (The default
1812 setting of finduser_retries is zero.) Also, it was sleeping after the final
1813 failure, which is pointless.
1814
181524. Implemented tls_on_connect_ports.
1816
181725. Implemented acl_smtp_predata.
1818
181926. If the domain in control=submission is set empty, Exim assumes that the
1820 authenticated id is a complete email address when it generates From: or
1821 Sender: header lines.
1822
182327. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1824 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1825 chown and chgrp in /bin and hostname in /usr/bin.
1826
182728. Exim was keeping the "process log" file open after each use, just as it
1828 does for the main log. This opens the possibility of it remaining open for
1829 long periods when the USR1 signal hits a daemon. Occasional processlog
1830 errors were reported, that could have been caused by this. Anyway, it seems
1831 much more sensible not to leave this file open at all, so that is what now
1832 happens.
1833
183429. The long-running daemon process does not normally write to the log once it
1835 has entered its main loop, and it closes the log before doing so. This is
1836 so that log files can straightforwardly be renamed and moved. However,
1837 there are a couple of unusual error situations where the daemon does write
1838 log entries, and I had neglected to close the log afterwards.
1839
184030. The text of an SMTP error response that was received during a remote
1841 delivery was being truncated at 512 bytes. This is too short for some of
1842 the long messages that one sometimes sees. I've increased the limit to
1843 1024.
1844
184531. It is now possible to make retry rules that apply only when a message has a
1846 specific sender, in particular, an empty sender.
1847
184832. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1849 it possible to be selective about when SMTP synchronization is enforced.
1850
185133. Added "control = caseful_local_part" and "control = "caselower_local_part".
1852
185332. Implemented hosts_connection_nolog.
1854
185533. Added an ACL for QUIT.
1856
185734. Setting "delay_warning=" to disable warnings was not working; it gave a
1858 syntax error.
1859
186035. Added mailbox_size and mailbox_filecount to appendfile.
1861
186236. Added control = no_multiline_responses to ACLs.
1863
186437. There was a bug in the logic of the code that waits for the clock to tick
1865 in the case where the clock went backwards by a substantial amount such
1866 that the microsecond fraction of "now" was more than the microsecond
1867 fraction of "then" (but the whole seconds number was less).
1868
186938. Added support for the libradius Radius client library this is found on
1870 FreeBSD (previously only the radiusclient library was supported).
1871
1872
1873Exim version 4.42
1874-----------------
1875
1876 1. When certain lookups returned multiple values in the form name=value, the
1877 quoting of the values was not always being done properly. Specifically:
1878 (a) If the value started with a double quote, but contained no whitespace,
1879 it was not quoted.
1880 (b) If the value contained whitespace other than a space character (i.e.
1881 tabs or newlines or carriage returns) it was not quoted.
1882 This fix has been applied to the mysql and pgsql lookups by writing a
1883 separate quoting function and calling it from the lookup code. The fix
1884 should probably also be applied to nisplus, ibase and oracle lookups, but
1885 since I cannot test any of those, I have not disturbed their existing code.
1886
1887 2. A hit in the callout cache for a specific address caused a log line with no
1888 reason for rejecting RCPT. Now it says "Previous (cached) callout
1889 verification failure".
1890
1891 3. There was an off-by-one bug in the queryprogram router. An over-long
1892 return line was truncated at 256 instead of 255 characters, thereby
1893 overflowing its buffer with the terminating zero. As well as fixing this, I
1894 have increased the buffer size to 1024 (and made a note to document this).
1895
1896 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1897 when Exim is waiting for an SMTP response from a remote server, Exim
1898 restarts its select() call on the socket, thereby resetting its timeout.
1899 This is not a problem when such interrupts are rare. Somebody set up a cron
1900 job to run exiwhat every 2 minutes, which is less than the normal select()
1901 timeout (5 or 10 minutes). This meant that the select() timeout never
1902 kicked in because it was always reset. I have fixed this by comparing the
1903 time when an interrupt arrives with the time at the start of the first call
1904 to select(). If more time than the timeout has elapsed, the interrupt is
1905 treated as a timeout.
1906
1907 5. Some internal re-factoring in preparation for the addition of Sieve
1908 extensions (by MH). In particular, the "personal" test is moved to a
1909 separate function, and given an option for scanning Cc: and Bcc: (which is
1910 not set for Exim filters).
1911
1912 6. When Exim created an email address using the login of the caller as the
1913 local part (e.g. when creating a From: or Sender: header line), it was not
1914 quoting the local part when it contained special characters such as @.
1915
1916 7. Installed new OpenBSD configuration files.
1917
1918 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1919 try to make them clearer.
1920
1921 9. Callout options, other than the timeout value, were being ignored when
1922 verifying sender addresses in header lines. For example, when using
1923
1924 verify = header_sender/callout=no_cache
1925
1926 the cache was (incorrectly) being used.
1927
192810. Added a missing instance of ${EXE} to the exim_install script; this affects
1929 only the Cygwin environment.
1930
193111. When return_path_on_delivery was set as a log selector, if different remote
1932 addresses in the same message used different return paths and parallel
1933 remote delivery occurred, the wrong values would sometimes be logged.
1934 (Whenever a remote delivery process finished, the return path value from
1935 the most recently started remote delivery process was logged.)
1936
193712. RFC 3848 specifies standard names for the "with" phrase in Received: header
1938 lines when AUTH and/or TLS are in use. This is the "received protocol"
1939 field. Exim used to use "asmtp" for authenticated SMTP, without any
1940 indication (in the protocol name) for TLS use. Now it follows the RFC and
1941 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1942 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1943 names appear in log lines as well as in Received: header lines.
1944
194513. Installed MH's patches for Sieve to add the "copy" and "vacation"
1946 extensions, and comparison tests, and to fix some bugs.
1947
194814. Changes to the "personal" filter test:
1949
1950 (1) The test was buggy in that it was just doing the equivalent of
1951 "contains" tests on header lines. For example, if a user's address was
1952 anne@some.where, the "personal" test would incorrectly be true for
1953
1954 To: susanne@some.where
1955
1956 This test is now done by extracting each address from the header in turn,
1957 and checking the entire address. Other tests that are part of "personal"
1958 are now done using regular expressions (for example, to check local parts
1959 of addresses in From: header lines).
1960
1961 (2) The list of non-personal local parts in From: addresses has been
1962 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1963 taken from the Sieve specification recommendations.
1964
1965 (3) If the message contains any header line starting with "List-" it is
1966 treated as non-personal.
1967
1968 (4) The test for "circular" in the Subject: header line has been removed
1969 because it now seems ill-conceived.
1970
197115. Minor typos in src/EDITME comments corrected.
1972
197316. Installed latest exipick from John Jetmore.
1974
197517. If headers_add on a router specified a text string that was too long for
1976 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1977 of string_sprintf() is now avoided.
1978
197918. $message_body_size was not set (it was always zero) when running the DATA
1980 ACL and the local_scan() function.
1981
198219. For the "mail" command in an Exim filter, no default was being set for
1983 the once_repeat time, causing a random time value to be used if "once" was
1984 specified. (If the value happened to be <= 0, no repeat happened.) The
1985 default is now 0s, meaning "never repeat". The "vacation" command was OK
1986 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1987 (I found it when inspecting the code).
1988
198920. There is now an overall timeout for performing a callout verification. It
1990 defaults to 4 times the callout timeout, which applies to individual SMTP
1991 commands during the callout. The overall timeout applies when there is more
1992 than one host that can be tried. The timeout is checked before trying the
1993 next host. This prevents very long delays if there are a large number of
1994 hosts and all are timing out (e.g. when the network connections are timing
1995 out). The value of the overall timeout can be changed by specifying an
1996 additional sub-option for "callout", called "maxwait". For example:
1997
1998 verify = sender/callout=5s,maxwait=20s
1999
200021. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2001 to the end before writing, but this should make it even safer).
2002
200322. Exim was forgetting that it had advertised PIPELINING for the second and
2004 subsequent messages on an SMTP connection. It was also not resetting its
2005 memory on STARTTLS and an internal HELO.
2006
200723. When Exim logs an SMTP synchronization error within a session, it now
2008 records whether PIPELINING has been advertised or not.
2009
201024. Added 3 instances of "(long int)" casts to time_t variables that were being
2011 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2012 rather than long int.
2013
201425. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2015
201626. Added the never_mail option to autoreply.
2017
2018
2019Exim version 4.41
2020-----------------
2021
2022 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2023 crash if the getsockname() call failed; this can happen if a connection is
2024 closed very soon after it is established. The problem was simply in the
2025 order in which certain operations were done, causing Exim to try to write
2026 to the SMTP stream before it had set up the file descriptor. The bug has
2027 been fixed by making things happen in the correct order.
2028
2029
2030Exim version 4.40
2031-----------------
2032
2033 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2034 before the connection was closed, thus losing the rejection response.
2035
2036 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2037 some early Solaris releases, but causes trouble in current releases where
2038 socklen_t is defined.
2039
2040 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2041 always exist.
2042
2043 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2044 configured.
2045
2046 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2047 that releasing the top of it at the end releases what was used for sub-
2048 expansions (unless the block got too big). However, discard this block if
2049 the first thing is a variable or header, so that we can use its block when
2050 it is dynamic (useful for very large $message_headers, for example).
2051
2052 6. Lookups now cache *every* query, not just the most recent. A new, separate
2053 store pool is used for this. It can be recovered when all lookup caches are
2054 flushed. Lookups now release memory at the end of their result strings.
2055 This has involved some general refactoring of the lookup sources.
2056
2057 7. Some code has been added to the store_xxx() functions to reduce the amount
2058 of flapping under certain conditions.
2059
2060 8. log_incoming_interface used to affect only the <= reception log lines. Now
2061 it causes the local interface and port to be added to several more SMTP log
2062 lines, for example "SMTP connection from", and rejection lines.
2063
2064 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2065
206610. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2067
206811. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2069 could be overwritten at the end of the current message (or the start of a
2070 new message if it was set in a HELO ACL). The value is now preserved for
2071 the duration of the SMTP connection.
2072
207312. If a transport had a headers_rewrite setting, and a matching header line
2074 contained an unqualified address, that address was qualified, even if it
2075 did not match any rewriting rules. The underlying bug was that the values
2076 of the flags that permit the existence of unqualified sender and recipient
2077 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2078 non-local messages, and by -bnq for local messages) were not being
2079 preserved with the message after it was received.
2080
208113. When Exim was logging an SMTP synchronization error, it could sometimes log
2082 "next input=" as part of the text comprising the host identity instead of
2083 the correct text. The code was using the same buffer for two different
2084 strings. However, depending on which order the printing function evaluated
2085 its arguments, the bug did not always show up. Under Linux, for example, my
2086 test suite worked just fine.
2087
208814. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2089 doesn't work with some older versions of Perl. It has been changed to "my",
2090 which in any case is probably the better facility to use.
2091
209215. A really picky compiler found some instances of statements for creating
2093 error messages that either had too many or two few arguments for the format
2094 string.
2095
209616. The size of the buffer for calls to the DNS resolver has been increased
2097 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2098 for addresses that have a lot of PTR records. This alleviates a problem; it
2099 does not fully solve it.
2100
210117. A dnsdb lookup for PTR records that receives more data than will fit in the
2102 buffer now truncates the list and logs the incident, which is the same
2103 action as happens when Exim is looking up a host name and its aliases.
2104 Previously in this situation something unpredictable would happen;
2105 sometimes it was "internal error: store_reset failed".
2106
210718. If a server dropped the connection unexpectedly when an Exim client was
2108 using GnuTLS and trying to read a response, the client delivery process
2109 crashed while trying to generate an error log message.
2110
211119. If a "warn" verb in an ACL added multiple headers to a message in a single
2112 string, for example:
2113
2114 warn message = H1: something\nH2: something
2115
2116 the text was added as a single header line from Exim's point of view
2117 though it ended up OK in the delivered message. However, searching for the
2118 second and subsequent header lines using $h_h2: did not work. This has been
2119 fixed. Similarly, if a system filter added multiple headers in this way,
2120 the routers could not see them.
2121
212220. Expanded the error message when iplsearch is called with an invalid key to
2123 suggest using net-iplsearch in a host list.
2124
212521. When running tests using -bh, any delays imposed by "delay" modifiers in
2126 ACLs are no longer actually imposed (and a message to that effect is
2127 output).
2128
212922. If a "gecos" field in a passwd entry contained escaped characters, in
2130 particular, if it contained a \" sequence, Exim got it wrong when building
2131 a From: or a Sender: header from that name. A second bug also caused
2132 incorrect handling when an unquoted " was present following a character
2133 that needed quoting.
2134
213523. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2136 was not being matched caselessly.
2137
213824. Arranged for all hyphens in the exim.8 source to be escaped with
2139 backslashes.
2140
214125. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2142 properly. Recipient callout cache records were still being keyed to include
2143 the sender, even when use_sender was set false. This led to far more
2144 callouts that were necessary. The sender is no longer included in the key
2145 when use_sender is false.
2146
214726. Added "control = submission" modifier to ACLs.
2148
214927. Added the ${base62d: operator to decode base 62 numbers.
2150
215128. dnsdb lookups can now access SRV records.
2152
215329. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2154 the configuration file.
2155
215630. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2157 (-v) mode. This makes the output for a verbose queue run more intelligible.
2158
215931. Added a use_postmaster feature to recipient callouts.
2160
216132. Added the $body_zerocount variable, containing the number of binary zero
2162 bytes in the message body.
2163
216433. The time of last modification of the "new" subdirectory is now used as the
2165 "mailbox time last read" when there is a quota error for a maildir
2166 delivery.
2167
216834. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2169
217035. Added +ignore_unknown as a special item in host lists.
2171
217236. Code for decoding IPv6 addresses in host lists is now included, even if
2173 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2174 address was recognized as an IP address, but was then not correctly decoded
2175 into binary, causing unexpected and incorrect effects when compared with
2176 another IP address.
2177
2178
2179Exim version 4.34
2180-----------------
2181
2182 1. Very minor rewording of debugging text in manualroute to say "list of
2183 hosts" instead of "hostlist".
2184
2185 2. If verify=header_syntax was set, and a header line with an unqualified
2186 address (no domain) and a large number of spaces between the end of the
2187 name and the colon was received, the reception process suffered a buffer
2188 overflow, and (when I tested it) crashed. This was caused by some obsolete
2189 code that should have been removed. The fix is to remove it!
2190
2191 3. When running in the test harness, delay a bit after writing a bounce
2192 message to get a bit more predictability in the log output.
2193
2194 4. Added a call to search_tidyup() just before forking a reception process. In
2195 theory, someone could use a lookup in the expansion of smtp_accept_max_
2196 per_host which, without the tidyup, could leave open a database connection.
2197
2198 5. Added the variables $recipient_data and $sender_data which get set from a
2199 lookup success in an ACL "recipients" or "senders" condition, or a router
2200 "senders" option, similar to $domain_data and $local_part_data.
2201
2202 6. Moved the writing of debug_print from before to after the "senders" test
2203 for routers.
2204
2205 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2206 problems for message scanning, either using a data ACL, or using
2207 local_scan() because the Received: header was not generated till after they
2208 were called (in order to set the time as the time of reception completion).
2209 I have revised the way this works. The header is now generated after the
2210 body is received, but before the ACL or local_scan() are called. After they
2211 are run, the timestamp in the header is updated.
2212
2213
2214Exim version 4.33
2215-----------------
2216
2217 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2218 before starting a queue runner without re-exec. This happened only when
2219 deliver_drop_privilege was set or when the Exim user was set to root. The
2220 effect of the bug was that timeouts during subsequent deliveries caused
2221 crashes instead of being properly handled. The handler is now left at its
2222 default (and expected) setting.
2223
2224 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2225 message, again when deliver_drop_privilege is set or Exim is run as root.
2226 The bug described in (1) was not present in this case, but the tidying up
2227 of the other signals was missing. I have made the two cases consistent.
2228
2229 3. The ignore_target_hosts setting on a manualroute router was being ignored
2230 for hosts that were looked up using the /MX notation.
2231
2232 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2233 in domain lists.
2234
2235 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2236 operated on the sender address. After changing the $sender_address to <>
2237 for the sender address verify, Exim was re-instated it as the original
2238 (before rewriting) address, but remembering that it had rewritten it, so it
2239 wasn't rewriting it again. This bug also had the effect of breaking the
2240 sender address verification caching when the sender address was rewritten.
2241
2242 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2243 This has been changed so that if the ip literal address matches
2244 ignore_target_hosts, the router declines.
2245
2246 7. Added expansion conditions match_domain, match_address, and match_local_
2247 part (NOT match_host).
2248
2249 8. The placeholder for the Received: header didn't have a length field set.
2250
2251 9. Added code to Exim itself and to exim_lock to test for a specific race
2252 condition that could lead to file corruption when using MBX delivery. The
2253 issue is with the lockfile that is created in /tmp. If this file is removed
2254 after a process has opened it but before that process has acquired a lock,
2255 there is the potential for a second process to recreate the file and also
2256 acquire a lock. This could lead to two Exim processes writing to the file
2257 at the same time. The added code performs the same test as UW imapd; it
2258 checks after acquiring the lock that its file descriptor still refers to
2259 the same named file.
2260
226110. The buffer for building added header lines was of fixed size, 8192 bytes.
2262 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2263 when Exim is built.
2264
226511. Added the smtp_active_hostname option. If used, this will typically be made
2266 to depend on the incoming interface address. Because $interface_address is
2267 not set up until the daemon has forked a reception process, error responses
2268 that can happen earlier (such as "too many connections") no longer contain
2269 a host name.
2270
227112. If an expansion in a condition on a "warn" statement fails because a lookup
2272 defers, the "warn" statement is abandoned, and the next ACL statement is
2273 processed. Previously this caused the whole ACL to be aborted.
2274
227513. Added the iplsearch lookup type.
2276
227714. Added ident_timeout as a log selector.
2278
227915. Added tls_certificate_verified as a log selector.
2280
228116. Added a global option tls_require_ciphers (compare the smtp transport
2282 option of the same name). This controls incoming TLS connections.
2283
228417. I finally figured out how to make tls_require_ciphers do a similar thing
2285 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2286 before starting the TLS session.
2287
228818. Tabs are now shown as \t in -bP output.
2289
229019. If the log selector return_path_on_delivery was set, Exim crashed when
2291 bouncing a message because it had too many Received: header lines.
2292
229320. If two routers both had headers_remove settings, and the first one included
2294 a superfluous trailing colon, the final name in the first list and the
2295 first name in the second list were incorrectly joined into one item (with a
2296 colon in the middle).
2297
2298
2299Exim version 4.32
2300-----------------
2301
2302 1. Added -C and -D options to the exinext utility, mainly to make it easier
2303 to include in the automated testing, but these could be helpful when
2304 multiple configurations are in use.
2305
2306 2. The exinext utility was not formatting the output nicely when there was
2307 an alternate port involved in the retry record key, nor when there was a
2308 message id as well (for retries that were specific to a specific message
2309 and a specific host). It was also confused by IPv6 addresses, because of
2310 the additional colons they contain. I have fixed the IPv4 problem, and
2311 patched it up to do a reasonable job for IPv6.
2312
2313 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2314 delivery, the log line now contains "pipelined" if PIPELINING was used.
2315
2316 4. An SMTP transport process used to panic and die if the bind() call to set
2317 an explicit outgoing interface failed. This has been changed; it is now
2318 treated in the same way as a connect() failure.
2319
2320 5. A reference to $sender_host_name in the part of a conditional expansion
2321 that was being skipped was still causing a DNS lookup. This no longer
2322 occurs.
2323
2324 6. The def: expansion condition was not recognizing references to header lines
2325 that used bh_ and bheader_.
2326
2327 7. Added the _cache feature to named lists.
2328
2329 8. The code for checking quota_filecount in the appendfile transport was
2330 allowing one more file than it should have been.
2331
2332 9. For compatibility with Sendmail, the command line option
2333
2334 -prval:sval
2335
2336 is equivalent to
2337
2338 -oMr rval -oMs sval
2339
2340 and sets the incoming protocol and host name (for trusted callers). The
2341 host name and its colon can be omitted when only the protocol is to be set.
2342 Note the Exim already has two private options, -pd and -ps, that refer to
2343 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2344 "s", but I don't think that's a major issue.
2345
234610. A number of refactoring changes to the code, none of which should affect
2347 Exim's behaviour:
2348
2349 (a) The number of logging options was getting close to filling up the
2350 32-bit word that was used as a bit map. I have split them into two classes:
2351 those that are passed in the argument to log_write(), and those that are
2352 only ever tested independently outside of that function. These are now in
2353 separate 32-bit words, so there is plenty of room for expansion again.
2354 There is no change in the user interface or the logging behaviour.
2355
2356 (b) When building, for example, log lines, the code previously used a
2357 macro that called string_cat() twice, in order to add two strings. This is
2358 not really sufficiently general. Furthermore, there was one instance where
2359 it was actually wrong because one of the argument was used twice, and in
2360 one call a function was used. (As it happened, calling the function twice
2361 did not affect the overall behaviour.) The macro has been replaced by a
2362 function that can join an arbitrary number of extra strings onto a growing
2363 string.
2364
2365 (c) The code for expansion conditions now uses a table and a binary chop
2366 instead of a serial search (which was left over from when there were very
2367 few conditions). Also, it now recognizes conditions like "pam" even when
2368 the relevant support is not compiled in: a suitably worded error message is
2369 given if an attempt is made to use such a condition.
2370
237111. Added ${time_interval:xxxxx}.
2372
237312. A bug was causing one of the ddress fields not to be passed back correctly
2374 from remote delivery subprocesses. The field in question was not being
2375 subsequently used, so this caused to problems in practice.
2376
237713. Added new log selectors queue_time and deliver_time.
2378
237914. Might have fixed a bug in maildirsizefile handling that threw up
2380 "unexpected character" debug warnings, and recalculated the data
2381 unnecessarily. In any case, I expanded the warning message to give more
2382 information.
2383
238415. Added the message "Restricted characters in address" to the statements in
2385 the default ACL that block characters like @ and % in local parts.
2386
238716. Change 71 for release 4.31 proved to be much less benign that I imagined.
2388 Three changes have been made:
2389
2390 (a) There was a serious bug; a negative response to MAIL caused the whole
2391 recipient domain to be cached as invalid, thereby blocking all messages
2392 to all local parts at the same domain, from all senders. This bug has
2393 been fixed. The domain is no longer cached after a negative response to
2394 MAIL if the sender used is not empty.
2395
2396 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2397 been restored.
2398
2399 (c) A new callout option, "use_sender" has been added for people who want
2400 the modified behaviour.
2401
2402
2403Exim version 4.31
2404-----------------
2405
2406 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2407 Larry Rosenman.
2408
2409 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2410 indeed breaks things for older releases.
2411
2412 3. Added additional logging to the case where there is a problem reading data
2413 from a filter that is running in a subprocess using a pipe, in order to
2414 try to track down a specific problem.
2415
2416 4. Testing facility fudge: when running in the test harness and attempting
2417 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2418 getting "No route to host". Convert this to a timeout.
2419
2420 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2421 warning.
2422
2423 6. Some OS don't have socklen_t but use size_t instead. This affects the
2424 fifth argument of getsockopt() amongst other things. This is now
2425 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2426 can be set for individual OS. I have set it for SunOS5, OSF1, and
2427 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2428 some earlier ones do not.
2429
2430 7. Change 4.30/15 was not doing the test caselessly.
2431
2432 8. The standard form for an IPv6 address literal was being rejected by address
2433 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2434 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2435 this, as well as the form without the "IPv6" on the front (but only when
2436 address literals are enabled, of course).
2437
2438 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2439
244010. Exim crashed if a message with an empty sender address specified by -f
2441 encountered a router with an errors_to setting. This could be provoked only
2442 by a command such as
2443
2444 exim -f "" ...
2445
2446 where an empty string was supplied; "<>" did not hit this bug.
2447
244811. Installed PCRE release 4.5.
2449
245012. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2451 remained set. It is now erased.
2452
245313. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2454 times from message ids (which are base 36 rather than the normal 62).
2455
245614. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2457 were being counted as actual protocol errors, and logged if the log
2458 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2459 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2460 and DATA following a set of rejected RCPTs do not count as protocol errors.
2461 In other words, Exim assumes they were pipelined, though this may not
2462 actually be the case. Of course, in all cases the client gets an
2463 appropriate error code.
2464
246515. If a lookup fails in an ACL condition, a message about the failure may
2466 be available; it is used if testing the ACL cannot continue, because most
2467 such messages specify what the cause of the deferral is. However, some
2468 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2469 that caused an old message to be retained and used if a later statement
2470 caused a defer, replacing the real cause of the deferral.
2471
247216. If an IP address had so many PTR records that the DNS lookup buffer
2473 was not large enough to hold them, Exim could crash while trying to process
2474 the truncated data. It now detects and logs this case.
2475
247617. Further to 4.21/58, another change has been made: if (and only if) the
2477 first line of a message (the first header line) ends with CRLF, a bare LF
2478 in a subsequent header line has a space inserted after it, so as not to
2479 terminate the header.
2480
248118. Refactoring: tidied an ugly bit of code in appendfile that copied data
2482 unnecessarily, used atoi() instead of strtol(), and didn't check the
2483 termination when getting file sizes from file names by regex.
2484
248519. Completely re-implemented the support for maildirsize files, in the light
2486 of a number of problems with the previous contributed implementation
2487 (4.30/29). In particular:
2488
2489 . If the quota is zero, the maildirsize file is maintained, but no quota is
2490 imposed.
2491
2492 . If the maildir directory does not exist, it is created before any attempt
2493 to write a maildirsize file.
2494
2495 . The quota value in the file is just a cache; if the quota is changed in
2496 the transport, the new value overrides.
2497
2498 . A regular expression is available for excluding directories from the
2499 count.
2500
250120. The autoreply transport checks the characters in options that define the
2502 message's headers; it allows continued headers, but it was checking with
2503 isspace() after an embedded newline instead of explicitly looking for a
2504 space or a tab.
2505
250621. If all the "regular" hosts to which an address was routed had passed their
2507 expiry times, and had not reached their retry times, the address was
2508 bounced, even if fallback hosts were defined. Now Exim should go on to try
2509 the fallback hosts.
2510
251122. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2512 equivalent code in the SMTP transport. Some hosts send humungous responses
2513 to HELO/EHLO, more than 1024 it seems.
2514
251523. Refactoring: code in filter.c used (void *) for "any old type" but this
2516 gives compiler warnings in some environments. I've now done it "properly",
2517 using a union.
2518
251924. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2520 (because of problems with the built-in one) was declared to return uschar *
2521 instead of char *, causing compiler failure.
2522
252325. Fixed a file descriptor leak when processing alias/forward files.
2524
252526. Fixed a minor format string issue in dbfn.c.
2526
252727. Typo in exim.c: ("dmbnz" for "dbmnz").
2528
252928. If a filter file refered to $h_xxx or $message_headers, and the headers
2530 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2531 become corrupted.
2532
253329. When a sender address is verified, it is cached, to save repeating the test
2534 when there is more than one recipient in a message. However, when the
2535 verification involves a callout, it is possible for different callout
2536 options to be set for different recipients. It is too complicated to keep
2537 track of this in the cache, so now Exim always runs a verification when a
2538 callout is required, relying on the callout cache for the optimization.
2539 The overhead is duplication of the address routing, but this should not be
2540 too great.
2541
254230. Fixed a bug in callout caching. If a RCPT command caused the sender address
2543 to be verified with callout=postmaster, and the main callout worked but the
2544 postmaster check failed, the verification correctly failed. However, if a
2545 subsequent RCPT command asked for sender verification *without* the
2546 postmaster check, incorrect caching caused this verification also to fail,
2547 incorrectly.
2548
254931. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2550 it was not caching the DNS options (qualify_single, search_parents) that
2551 were used when the lookup failed. A subsequent lookup with different
2552 options therefore always gave the same answer, though there were cases
2553 where it should not have. (Example: a "domains = !$mx_any" option on a
2554 dnslookup router: the "domains" option is always processed without any
2555 widening, but the router might have qualify_single set.) Now Exim uses the
2556 cached value only when the same options are set.
2557
255832. Added John Jetmore's "exipick" utility to the distribution.
2559
256033. GnuTLS: When an attempt to start a TLS session fails for any reason other
2561 than a timeout (e.g. a certificate is required, and is not provided), an
2562 Exim server now closes the connection immediately. Previously it waited for
2563 the client to close - but if the client is SSL, it seems that they each
2564 wait for each other, leading to a delay before one of them times out.
2565
256634: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2567 maintained 0.8.x compatibility because I don't think many are using it, and
2568 it is clearly obsolete.
2569
257035. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2571 transport.
2572
257336. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2574 client certificate was expired. A simple patch fixes this, though I don't
2575 understand the full logic of why the verify callback is called multiple
2576 times.
2577
257837. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2579 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2580 which causes problems with some clients (such as the Certicom SSL Plus
2581 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2582 disables the coutermeasure allowing Eudora to connect."
2583
258438. Exim was not checking that a write() to a log file succeeded. This could
2585 lead to Bad Things if a log got too big, in particular if it hit a file
2586 size limit. Exim now panics and dies if it cannot write to a log file, just
2587 as it does if it cannot open a log file.
2588
258939. Modified OS/Makefile-Linux so that it now contains
2590
2591 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2592
2593 The two -D definitions ensure that Exim is compiled with large file
2594 support, which makes it possible to handle log files that are bigger than
2595 2^31.
2596
259740. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2598 instance) a domain was checked against a named list that involved a lookup,
2599 causing $domain_data to be set, then another domain was checked against the
2600 same list, then the first domain was re-checked, the value of $domain_data
2601 after the final check could be wrong. In particular, if the second check
2602 failed, it could be set empty. This bug probably also applied to
2603 $localpart_data.
2604
260541. The strip_trailing_dot option was not being applied to the address given
2606 with the -f command-line option.
2607
260842. The code for reading a message's header from the spool was incrementing
2609 $received_count, but never initializing it. This meant that the value was
2610 incorrect (doubled) while delivering a message in the same process in which
2611 it was received. In the most common configuration of Exim, this never
2612 happens - a fresh exec is done - but it can happen when
2613 deliver_drop_privilege is set.
2614
261543. When Exim logs an SMTP synchronization error - client data sent too soon -
2616 it now includes up to 150 characters of the unexpected data in the log
2617 line.
2618
261944. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2620 and building data strings. The size of both of these buffers was 10 000
2621 bytes - far larger than anybody would *ever* want, thought I. Needless to
2622 say, somebody hit the limit. I have increased the maximum line length to
2623 20 000 and the maximum data length of concatenated lines to 100 000. I have
2624 also fixed two bugs, because there was no checking on these buffers. Tsk,
2625 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2626 buffer is too small.
2627
262845. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2629 lsearch lookups. Now it does.
2630
263146. When parsing a route_list item in a manualroute router, a fixed-length
2632 buffer was used for the list of hosts. I made this 1024 bytes long,
2633 thinking that nobody would ever have a list of hosts that long. Wrong.
2634 Somebody had a whole pile of complicated expansion conditions, and the
2635 string was silently truncated, leading to an expansion error. It turns out
2636 that it is easier to change to an unlimited length (owing to other changes
2637 that have happened since this code was originally written) than to build
2638 structure for giving a limitation error. The length of the item that
2639 expands into the list of hosts is now unlimited.
2640
264147. The lsearch lookup could not handle data where the length of text line was
2642 more than 4095 characters. Such lines were truncated, leading to shortened
2643 data being returned. It should now handle lines of any length.
2644
264548. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2646 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2647 ACL").
2648
264949. Cosmetic tidy to scripts like exicyclog that are generated by globally
2650 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2651 no longer happens in comment lines. A list of replacements is now placed
2652 at the head of all of the source files, except those whose only change is
2653 to replace PERL_COMMAND in the very first #! line.
2654
265550. Replaced the slow insertion sort in queue.c, for sorting the list of
2656 messages on the queue, with a bottom-up merge sort, using code contributed
2657 by Michael Haardt. This should make operations like -bp somewhat faster on
2658 large queues. It won't affect queue runners, except when queue_run_in_order
2659 is set.
2660
266151. Installed eximstats 1.31 in the distribution.
2662
266352. Added support for SRV lookups to the dnslookup router.
2664
266553. If an ACL referred to $message_body or $message_body_end, the value was not
2666 reset for any messages that followed in the same SMTP session.
2667
266854. The store-handling optimization for building very long strings was not
2669 differentiating between the different store pools. I don't think this
2670 actually made any difference in practice, but I've tidied it.
2671
267255. While running the routers to verify a sender address, $sender_address
2673 was still set to the sender address. This is wrong, because when routing to
2674 send a bounce to the sender, it would be empty. Therefore, I have changed
2675 it so that, while verifying a sender address, $sender_address is set to <>.
2676 (There is no change to what happens when verifying a recipient address.)
2677
267856. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2679 target A or AAAA records (if not already returned) without resetting the
2680 qualify_single or search_parents options of the DNS resolver. These are
2681 inappropriate in this case because the targets of MX and SRV records must
2682 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2683 target that, when qualified, matched something in the local domain. These
2684 two options are now turned off when doing these lookups.
2685
268657. It seems that at least some releases of Reiserfs (which does not have the
2687 concept of a fixed number of inodes) returns zero and not -1 for the
2688 number of available inodes. This interacted badly with check_spool_inodes,
2689 which assumed that -1 was the "no such thing" setting. What I have done is
2690 to check that the total number of inodes is greater than zero before doing
2691 the test of how many are available.
2692
269358. When a "warn" ACL statement has a log_message modifier, the message is
2694 remembered, and not repeated. This is to avoid a lot of repetition when a
2695 message has many recipients that cause the same warning to be written.
2696 Howewer, Exim was preserving the list of already written lines for an
2697 entire SMTP session, which doesn't seem right. The memory is now reset if a
2698 new message is started.
2699
270059. The "rewrite" debugging flag was not showing the result of rewriting in the
2701 debugging output unless log_rewrite was also set.
2702
270360. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2704 of (int)(handle) when we know that handle contains (void *)(-1).
2705
270661. The Exim daemon panic-logs an error return when it closes the incoming
2707 connection. However "connection reset by peer" seems to be common, and
2708 isn't really an error worthy of noting specially, so that particular error
2709 is no long logged.
2710
271162. When Exim is trying to find all the local interfaces, it used to panic and
2712 die if the ioctl to get the interface flags failed. However, it seems that
2713 on at least one OS (Solaris 9) it is possible to have an interface that is
2714 included in the list of interfaces, but for which you get a failure error
2715 for this call. This happens when the interface is not "plumbed" into a
2716 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2717 failure of the "get flags" call assumes that the interface is down.
2718
271963. Added a ${eval10: operator, which assumes all numbers are decimal. This
2720 makes life easier for people who are doing arithmetic on fields extracted
2721 from dates, where you often get leading zeros that should not be
2722 interpreted as octal.
2723
272464. Added qualify_domain to the redirect router, to override the global
2725 setting.
2726
272765. If a pathologically long header line contained very many addresses (the
2728 report of this problem mentioned 10 000) and each of them was rewritten,
2729 Exim could use up a very large amount of memory. (It kept on making new
2730 copies of the header line as it rewrote, and never released the old ones.)
2731 At the expense of a bit more processing, the header rewriting function has
2732 been changed so that it no longer eats memory in this way.
2733
273466. The generation of the Received: header has been moved from the time that a
2735 message starts to be received, to the time that it finishes. The timestamp
2736 in the Received: header should now be very close to that of the <= log
2737 line. There are two side-effects of this change:
2738
2739 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2740 logged header lines no longer include the local Received: line, because
2741 it has not yet been created. The same applies to a copy of the message
2742 that is returned to a non-SMTP sender when a message is rejected.
2743
2744 (b) When a filter file is tested using -bf, no additional Received: header
2745 is added to the test message. After some thought, I decided that this
2746 is a bug fix.
2747
2748 This change does not affect the value of $received_for. It is still set
2749 after address rewriting, but before local_scan() is called.
2750
275167. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2752
275368. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2754 gave an unhelpful panic error message, and a defer error. I have managed to
2755 change this behaviour so that it now rejects any supplied certificate,
2756 which seems right, as the list of acceptable certificates is empty.
2757
275869. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2759 gave an unhelpful defer error. I have not managed to make this reject any
2760 supplied certificates, but the error message it gives is "no certificate
2761 supplied", which is not helpful.
2762
276370. exigrep's output now also includes lines that are not associated with any
2764 message, but which match the given pattern. Implemented by a patch from
2765 Martin Sluka, which also tidied up the Perl a bit.
2766
276771. Recipient callout verification, like sender verification, was using <> in
2768 the MAIL FROM command. This isn't really the right thing, since the actual
2769 sender may affect whether the remote host accepts the recipient or not. I
2770 have changed it to use the actual sender in the callout; this means that
2771 the cache record is now keyed on a recipient/sender pair, not just the
2772 recipient address. There doesn't seem to be a real danger of callout loops,
2773 since a callout by the remote host to check the sender would use <>.
2774 [SEE ABOVE: changed after hitting problems.]
2775
277672. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2777 temporary errors. However, in the case of such a code being given after
2778 the end of a data transmission (i.e. after ".") Exim was failing to write
2779 a retry record for the message. (Yes, there was some broken host that was
2780 actually sending 8xx at this point.)
2781
278273. An unknown lookup type in a host list could cause Exim to panic-die when
2783 the list was checked. (An example that provoked this was putting <; in the
2784 middle of a list instead of at the start.) If this happened during a DATA
2785 ACL check, a -D file could be left lying around. This kind of configuration
2786 error no longer causes Exim to die; instead it causes a defer errror. The
2787 incident is still logged to the main and panic logs.
2788
278974. Buglet left over from Exim 3 conversion. The message "too many messages
2790 in one connection" was written to the rejectlog but not the mainlog, except
2791 when address rewriting (yes!) was being logged.
2792
279375. Added write_rejectlog option.
2794
279576. When a system filter was run not as root (that is, when system_filter_user
2796 was set), the values of the $n variables were not being returned to the
2797 main process; thus, they were not subsequently available in the $sn
2798 variables.
2799
280077. Added +return_path_on_delivery log selector.
2801
280278. A connection timeout was being treated differently from recipients deferred
2803 when testing hosts_max_try with a message that was older than the host's
2804 retry timeout. (The host should not be counted, thus allowing all hosts to
2805 be tried at least once before bouncing.) This may have been the cause of an
2806 occasionally reported bug whereby a message would remain on the queue
2807 longer than the retry timeout, but would be bounced if a delivery was
2808 forced. I say "may" because I never totally pinned down the problem;
2809 setting up timeout/retry tests is difficult. See also the next item.
2810
281179. The ultimate address timeout was not being applied to errors that involved
2812 a combination of host plus message (for example, a timeout on a MAIL
2813 command). When an address resolved to a number of possible hosts, and they
2814 were not all tried for each delivery (e.g. because of hosts_max_try), a
2815 message could remain on the queue longer than the retry timeout.
2816
281780. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2818 Haardt.
2819
282081. Fixed an obscure SMTP outgoing bug which required at least the following
2821 conditions: (a) there was another message waiting for the same server;
2822 (b) the server returned 5xx to all RCPT commands in the first message so
2823 that the message was not completed; (c) the server dropped the connection
2824 or gave a negative response to the RSET that Exim sends to abort the
2825 transaction. The observed case was a dropped connection after DATA that had
2826 been sent in pipelining mode. That is, the server had advertised PIPELINING
2827 but was not implementing it correctly. The effect of the bug was incorrect
2828 behaviour, such as trying another host, and this could lead to a crash.
2829
2830
2831Exim version 4.30
2832-----------------
2833
2834 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2835 and daemon.c were passed as pointers to ints; they should have been
2836 pointers to socklen_t variables (which are typically unsigned ints).
2837
2838 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2839 fixed.
2840
2841 3. Fixed a really odd bug that affected only the testing scheme; patching a
2842 certain fixed string in the binary changed the value of another string that
2843 happened to be identical to the end of the original first string.
2844
2845 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2846 name", it returns that address as the IP address. On some operating
2847 systems (e.g. Solaris), it also passes back the IP address string as the
2848 "host name". However, on others (e.g. Linux), it passes back an empty
2849 string. Exim wasn't checking for this, and was changing the host name to an
2850 empty string, assuming it had been canonicized.
2851
2852 5. Although rare, it is permitted to have more than one PTR record for a given
2853 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2854 all the names associated with an address, because they do in Solaris.
2855 However, it seems that they do not in Linux for data that comes from the
2856 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2857 I found this out when I moved to a new Linux workstation and tried to run
2858 the Exim test suite.
2859
2860 To get round this problem I have changed the code so that it now does its
2861 own call to the DNS to look up PTR records when searching for a host name.
2862 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2863 addresses that are only in /etc/hosts are still found.
2864
2865 This behaviour is, however, controlled by an option called host_lookup_
2866 order, which defaults to "bydns:byaddr". If people want to use the other
2867 order, or indeed, just use one or the other means of lookup, they can
2868 specify it in this variable.
2869
2870 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2871 some operating systems, this comes back from gethostbyaddr() as an empty
2872 string, and this is what Exim used to test for. However, it seems that in
2873 other systems, "." is yielded. Exim now tests for this case too.
2874
2875 7. The values of check_spool_space and check_log_space are now held internally
2876 as a number of kilobytes instead of an absolute number of bytes. If a
2877 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2878 kilobyte. This means that much larger values can be stored.
2879
2880 8. Exim monitor: an attempt to get the action menu when not actually pointing
2881 at a message produces an empty menu entitled "No message selected". This
2882 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2883 no entries in it ("Shell widget menu has zero width and/or height"). So I
2884 have added a single, blank menu entry in this case.
2885
2886 9. Added ${quote_local_part.
2887
288810. MIME decoding is now applied to the contents of Subject: header lines when
2889 they are logged.
2890
289111. Now that a reference to $sender_host_address automatically causes a reverse
2892 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2893 host lookup before query-style lookups in lists that might use this
2894 variable. This has therefore been abolished, and the "net-" prefix is no
2895 longer necessary for query-style lookups.
2896
289712. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2898 have been a typo for LFLAGS, so it has been changed.
2899
290013. The install script calls Exim with "-C /dev/null" in order to find the
2901 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2902 to be output. Howeve, since Exim outputs its version number before the
2903 error, it didn't break the script. It just looked ugly. I fixed this by
2904 always allowing "-C /dev/null" if the caller is root.
2905
290614. Ignore overlarge ACL variable number when reading spool file - insurance
2907 against a later release with more variables having written the file.
2908
290915. The standard form for an IPv6 address literal was being rejected by EHLO.
2910 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2911 this, as well as the form without the "IPv6" on the front.
2912
291316. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2914 OS/Makefile-Darwin file.
2915
291617. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2917 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2918
291918. After change 4.21/52, "%ld" was used to format the contents of the $inode
2920 variable. However, some OS use ints for inodes. I've added cast to long int
2921 to get rid of the compiler warning.
2922
292319. I had forgotten to lock out "/../" in configuration file names when
2924 ALT_CONFIG_PREFIX was set.
2925
292620. Routers used for verification do not need to specify transports. However,
2927 if such a router generated a host list, and callout was configured, Exim
2928 crashed, because it could not find a port number from the (non-existent)
2929 transport. It now assumes port 25 in this circumstance.
2930
293121. Added the -t option to exigrep.
2932
293322. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2934 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2935 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2936 them is set, LOOKUP_LSEARCH is forced.
2937
293823. "exim -bV" now outputs a list of lookups that are included in the binary.
2939
294024. Added sender and host information to the "rejected by local_scan()" log
2941 line; previously there was no indication of these.
2942
294325. Added .include_if_exists.
2944
294526. Change 3.952/11 added an explicit directory sync on top of a file sync for
2946 Linux. It turns out that not all file systems support this. Apparently some
2947 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2948 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2949 not supported on the file descriptor, is now ignored when Exim is trying to
2950 sync a directory. This applies only to Linux.
2951
295227. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2953
295428. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2955 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2956 this.
2957
295829. Added support for maildirsize files from supplied patch (modified a bit).
2959
296030. The use of :fail: followed by an empty string could lead Exim to respond to
2961 sender verification failures with (e.g.):
2962
2963 550 Verification failed for <xxx>
2964 550 Sender verify failed
2965
2966 where the first response line was missing the '-' that indicates it is not
2967 the final line of the response.
2968
296931. The loop for finding the name of the user that called Exim had a hardwired
2970 limit of 10; it now uses the value of finduser_retries, which is used for
2971 all other user lookups.
2972
297332. Added $received_count variable, available in data and not_smtp ACLs, and at
2974 delivery time.
2975
297633. Exim was neglecting to zero errno before one call of strtol() when
2977 expanding a string and expecting an integer value. On some systems this
2978 resulted in spurious "integer overflow" errors. Also, it was casting the
2979 result into an int without checking.
2980
298134. Testing for a connection timeout using "timeout_connect" in the retry rules
2982 did not work. The code looks as if it has *never* worked, though it appears
2983 to have been documented since at least releast 1.62. I have made it work.
2984
298535. The "timeout_DNS" error in retry rules, also documented since at least
2986 1.62, also never worked. As it isn't clear exactly what this means, and
2987 clearly it isn't a major issue, I have abolished the feature by treating it
2988 as "timeout", and writing a warning to the main and panic logs.
2989
299036. The display of retry rules for -brt wasn't always showing the error code
2991 correctly.
2992
299337. Added new error conditions to retry rules: timeout_A, timeout_MX,
2994 timeout_connect_A, timeout_connect_MX.
2995
299638. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2997 to the empty sender.
2998
299939. The daemon was not analysing the content of -oX till after it had closed
3000 stderr and disconnected from the controlling terminal. This meant that any
3001 syntax errors were only noted on the panic log, and the return code from
3002 the command was 0. By re-arranging the code a little, I've made the
3003 decoding happen first, so such errors now appear on stderr, and the return
3004 code is 1. However, the actual setting up of the sockets still happens in
3005 the disconnected process, so errors there are still only recorded on the
3006 panic log.
3007
300840. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3009 connections (as happens on some IP stacks) was logged at start up time as
3010 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3011 it from "IPv6 and IPv4", which means that two separate sockets are being
3012 used.
3013
301441. The debug output for gethostbyname2() or getipnodebyname() failures now
3015 says whether AF_INET or AF_INET6 was passed as an argument.
3016
301742. Exiwhat output was messed up when time zones were included in log
3018 timestamps.
3019
302043. Exiwhat now gives more information about the daemon's listening ports,
3021 and whether -tls-on-connect was used.
3022
302344. The "port" option of the smtp transport is now expanded.
3024
302545. A "message" modifier in a "warn" statement in a non-message ACL was being
3026 silently ignored. Now an error message is written to the main and panic
3027 logs.
3028
302946. There's a new ACL modifier called "logwrite" which writes to a log file
3030 as soon as it is encountered.
3031
303247. Added $local_user_uid and $local_user_gid at routing time.
3033
303448. Exim crashed when trying to verify a sender address that was being
3035 rewritten to "<>".
3036
303749. Exim was recognizing only a space character after ".include". It now also
3038 recognizes a tab character.
3039
304050. Fixed several bugs in the Perl script that creates the exim.8 man page by
3041 extracting the relevant information from the specification. The man page no
3042 longer contains scrambled data for the -d option, and I've added a section
3043 at the front about calling Exim under different names.
3044
304551. Added "extra_headers" argument to the "mail" command in filter files.
3046
304752. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3048 crash.
3049
305053. Installed eximstats 1.29.
3051
305254. Added transport_filter_timeout as a generic transport option.
3053
305455. Exim no longer adds an empty Bcc: header to messages that have no To: or
3055 Cc: header lines. This was required by RFC 822, but it not required by RFC
3056 2822.
3057
305856. Exim used to add From:, Date:, and Message-Id: header lines to any
3059 incoming messages that did not have them. Now it does so only if the
3060 message originates locally, that is, if there is no associated remote host
3061 address. When Resent- header lines are present, this applies to the Resent-
3062 lines rather than the non-Resent- lines.
3063
306457. Drop incoming SMTP connection after too many syntax or protocol errors. The
3065 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3066
306758. Messages for configuration errors now include the name of the main
3068 configuration file - useful now that there may be more than one file in a
3069 list (.included file names were always shown).
3070
307159. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3072 for those rare installations that do not start the daemon as root or run it
3073 setuid root. I've cut out the call to initgroups() if the daemon is not
3074 root at that time.
3075
307660. The Exim user and group can now be bound into the binary as text strings
3077 that are looked up at the start of Exim's processing.
3078
307961. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3080
308162. Added $mailstore_basename variable.
3082
308363. Installed patch to sieve.c from Michael Haardt.
3084
308564. When Exim failed to open the panic log after failing to open the main log,
3086 the original message it was trying to log was written to stderr and debug
3087 output, but if they were not available (the usual case in production), it
3088 was lost. Now it is written to syslog before the two lines that record the
3089 failures to open the logs.
3090
309165. Users' Exim filters run in subprocesses under the user's uid. It is
3092 possible for a "deliver" command or an alias in a "personal" command to
3093 provoke an address rewrite. If logging of address rewriting is configured,
3094 this fails because the process is not running as root or exim. There may be
3095 a better way of dealing with this, but for the moment (because 4.30 needs
3096 to be released), I have disabled address rewrite logging when running a
3097 filter in a non-root, non-exim process.
3098
3099
3100Exim version 4.24
3101-----------------
3102
3103 1. The buildconfig auxiliary program wasn't quoting the value set for
3104 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3105 not defined. This bug was masked in 4.22 by the effect that was fixed in
3106 change 4.23/1.
3107
3108 2. Some messages that were rejected after a message id was allocated were
3109 shown as "incomplete" by exigrep. It no longer does this for messages that
3110 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3111
3112 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3113 have allow_domain_literals set, the ID did not get logged in the <= line.
3114 Domain literals are now always recognized in Message-ID: header lines.
3115
3116 4. The first argument for a ${extract expansion item is the key name or field
3117 number. Leading and trailing spaces in this item were not being ignored,
3118 causing some misleading effects.
3119
3120 5. When deliver_drop_privilege was set, single queue runner processes started
3121 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3122 same command in the process it spins off) were not dropping privilege.
3123
3124 6. When the daemon running as "exim" started a queue runner, it always
3125 re-executed Exim in the spun-off process. This is a waste of effort when
3126 deliver_drop_privilege is set. The new process now just calls the
3127 queue-runner function directly.
3128
3129
3130Exim version 4.23
3131-----------------
3132
3133 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3134 HEADERS_CHARSET.
3135
3136 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3137 ignored. Though the use of -oP was forcing the writing of a pid file, it
3138 was always written to the default place.
3139
3140 3. If the message "no IP address found for host xxxx" is generated during
3141 incoming verification, it is now followed by identification of the incoming
3142 connection (so you can more easily find what provoked it).
3143
3144 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3145
3146 5. Added some features to "harden" Exim a bit more against certain attacks:
3147
3148 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3149 be put in Local/Makefile. This is like the never_users runtime option,
3150 but it cannot be overridden. The default setting is "root".
3151
3152 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3153 prefix string with which any file named in a -C command line option
3154 must start.
3155
3156 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3157 is retained for -C and -D only if the caller of Exim is root. Without
3158 it, the exim user may also use -C and -D and retain privilege.
3159
3160 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3161 command line option is disabled.
3162
3163 6. Macro names set by the -D option must start with an upper case letter, just
3164 like macro names defined in the configuration file.
3165
3166 7. Added "dereference=" facility to LDAP.
3167
3168 8. Two instances of the typo "uknown" in the source files are fixed.
3169
3170 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3171 the Configure-Makefile script screwed up while processing it.
3172
317310. Incorporated PCRE 4.4.
3174
317511. The SMTP synchronization check was not operating right at the start of an
3176 SMTP session. For example, it could not catch a HELO sent before the client
3177 waited for the greeting. There is now a check for outstanding input at the
3178 point when the greeting is written. Because of the duplex, asynchronous
3179 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3180 way, but not yet received, when the check is performed.
3181
318212. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3183 on TCP/IP sockets, because this apparently causes some broken clients to
3184 timeout.
3185
318613. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3187 unchanged) from the Cygwin maintainer.
3188
318914. The code for -bV that shows what is in the binary showed "mbx" when maildir
3190 was supported instead of testing for mbx. Effectively a typo.
3191
319215. The spa authenticator server code was not checking that the input it
3193 received was valid base64.
3194
319516. The debug output line for the "set" modifier in ACLs was not showing the
3196 name of the variable that was being set.
3197
319817. Code tidy: the variable type "vtype_string" was never used. Removed it.
3199
320018. Previously, a reference to $sender_host_name did not cause a DNS reverse
3201 lookup on its own. Something else was needed to trigger the lookup. For
3202 example, a match in host_lookup or the need for a host name in a host list.
3203 Now, if $sender_host_name is referenced and the host name has not yet been
3204 looked up, a lookup is performed. If the lookup fails, the variable remains
3205 empty, and $host_lookup_failed is set to "1".
3206
320719. Added "eqi" as a case-independent comparison operator.
3208
320920. The saslauthd authentication condition could segfault if neither service
3210 nor realm was specified.
3211
321221. If an overflowing value such as "2048M" was set for message_size_limit, the
3213 error message that was logged was misleading, and incoming SMTP
3214 connections were dropped. The message is now more accurate, and temporary
3215 errors are given to SMTP connections.
3216
321722. In some error situations (such as 21 above) Exim rejects all SMTP commands
3218 (except RSET) with a 421 error, until QUIT is received. However, it was
3219 failing to send a response to QUIT.
3220
322123. The HELO ACL was being run before the code for helo_try_verify_hosts,
3222 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3223 ACL is now run after the helo_try_verify_hosts code.
3224
322524. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3226 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3227 case-independent, so other case variants are also recognized). Apparently
3228 some systems use these upper case variants.
3229
323025. If more than two messages were waiting for the same host, and a transport
3231 filter was specified for the transport, Exim sent two messages over the
3232 same TCP/IP connection, and then failed with "socket operation on non-
3233 socket" when it tried to send the third.
3234
323526. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3236
323727. The extern definition of crypt16() in expand.c was not being excluded when
3238 the OS had its own crypt16() function.
3239
324028. Added bounce_return_body as a new option, and bounce_return_size_limit
3241 as a preferred synonym for return_size_limit, both as an option and as an
3242 expansion variable.
3243
324429. Added LIBS=-liconv to OS/Makefile-OSF1.
3245
324630. Changed the default configuration ACL to relax the local part checking rule
3247 for addresses that are not in any local domains. For these addresses,
3248 slashes and pipe symbols are allowed within local parts, but the sequence
3249 /../ is explicitly forbidden.
3250
325131. SPA server authentication was not clearing the challenge buffer before
3252 using it.
3253
325432. log_message in a "warn" ACL statement was writing to the reject log as
3255 well as to the main log, which contradicts the documentation and doesn't
3256 seem right (because no rejection is happening). So I have stopped it.
3257
325833. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3259 However, I am unable to do any testing of this.
3260
326134. Fixed an infelicity in the appendfile transport. When checking directories
3262 for a mailbox, to see if any needed to be created, it was accidentally
3263 using path names with one or more superfluous leading slashes; tracing
3264 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3265
326635. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3267 contents are added to the log line that is written for every discarded
3268 recipient. (Previously a log_message setting was ignored.)
3269
327036. The ${quote: operator now quotes the string if it is empty.
3271
327237. The install script runs exim in order to find its version number. If for
3273 some reason other than non-existence or emptiness, which it checks, it
3274 could not run './exim', it was installing it with an empty version number,
3275 i.e. as "exim-". This error state is now caught, and the installation is
3276 aborted.
3277
327838. An argument was missing from the function that creates an error message
3279 when Exim fails to connect to the socket for saslauthd authentication.
3280 This could cause Exim to crash, or give a corrupted message.
3281
328239. Added isip, isip4, and isip6 to ${if conditions.
3283
328440. The ACL variables $acl_xx are now saved with the message, and can be
3285 accessed later in routers, transports, and filters.
3286
328741. The new lookup type nwildlsearch is like wildlsearch, except that the key
3288 strings in the file are not string-expanded.
3289
329042. If a MAIL command specified a SIZE value that was too large to fit into an
3291 int variable, the check against message_size_limit failed. Such values are
3292 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3293 day this will have to be increased, but I don't think I want to be around
3294 when emails are that large.
3295
3296
3297
3298Exim version 4.22
3299-----------------
3300
3301 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3302 iconv() is not standard in FreeBSD.
3303
3304 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3305 IPv6 enabled. The observed symptom was a segmentation fault on return from
3306 the function os_common_find_running_interfaces() in src/os.c.
3307
3308 3. In the check_special_case() function in daemon.c I had used "errno" as an
3309 argument name, which causes warnings on some systems. This was basically a
3310 typo, since it was named "eno" in the comments!
3311
3312 4. The code that waits for the clock to tick (at a resolution of some fraction
3313 of a second) so as to ensure message-id uniqueness was always waiting for
3314 at least one whole tick, when it could have waited for less. [This is
3315 almost certainly not relevant at current processor speeds, where it is
3316 unlikely to ever wait at all. But we try to future-proof.]
3317
3318 5. The function that sleeps for a time interval that includes fractions of a
3319 second contained a race. It did not block SIGALRM between setting the
3320 timer, and suspending (a couple of lines later). If the interval was short
3321 and the sigsuspend() was delayed until after it had expired, the suspension
3322 never ended. On busy systems this could lead to processes getting stuck for
3323 ever.
3324
3325 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3326 process, before it forks any delivery processes. The open lookup caching
3327 mechanism meant that the open file or database connection was passed into
3328 the delivery process. The problem was that delivery processes always tidy
3329 up cached lookup data. This could cause a problem for the next delivery
3330 process started by the queue runner, because the external queue runner
3331 process does not know about the closure. So the next delivery process
3332 still has data in the lookup cache. In the case of a file lookup, there was
3333 no problem because closing a file descriptor in a subprocess doesn't affect
3334 the parent. However, if the lookup was caching a connection to a database,
3335 the connection was closed, and the second delivery process was likely to
3336 see errors such as "PGSQL: query failed: server closed the connection
3337 unexpectedly". The problem has been fixed by closing all cached lookups
3338 in a queue runner before running a delivery process.
3339
3340 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3341 seem to have the "const" qualifier which it has on other OS. I've
3342 parameterised it.
3343
3344 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3345 *of the same type* (client or server) with the same public name that an
3346 error should be diagnosed.
3347
3348 9. When Exim looked up a host name for an IP address, but failed to find the
3349 original IP address when looking up the host name (a safety check), it
3350 output the message "<ip address> does not match any IP for NULL", which was
3351 confusing, to say the least. The bug was that the host name should have
3352 appeared instead of "NULL".
3353
335410. Since release 3.03, if Exim is called by a uid other than root or the Exim
3355 user that is built into the binary, and the -C or -D options is used, root
3356 privilege is dropped before the configuration file is read. In addition,
3357 logging is switched to stderr instead of the normal log files. If the
3358 configuration then re-defines the Exim user, the unprivileged environment
3359 is probably not what is expected, so Exim logs a panic warning message (but
3360 proceeds).
3361
3362 However, if deliver_drop_privilege is set, the unprivileged state may well
3363 be exactly what is intended, so the warning has been cut out in that case,
3364 and Exim is allowed to try to write to its normal log files.
3365
3366
3367Exim version 4.21
3368-----------------
3369
3370 1. smtp_return_error_details was not giving details for temporary sender
3371 or receiver verification errors.
3372
3373 2. Diagnose a configuration error if two authenticators have the same public
3374 name.
3375
3376 3. Exim used not to create the message log file for a message until the first
3377 delivery attempt. This could be confusing when incoming messages were held
3378 for policy or load reasons. The message log file is now created at the time
3379 the message is received, and an initial "Received" line is written to it.
3380
3381 4. The automatically generated man page for command line options had a minor
3382 bug that caused no ill effects; however, a more serious problem was that
3383 the procedure for building the man page automatically didn't always
3384 operate. Consequently, release 4.20 contains an out-of-date version. This
3385 shouldn't happen again.
3386
3387 5. When building Exim with embedded Perl support, the script that builds the
3388 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3389 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3390
3391 6. The freeze_tell option was not being used for messages that were frozen on
3392 arrival, either by an ACL or by local_scan().
3393
3394 7. Added the smtp_incomplete_transaction log selector.
3395
3396 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3397 was accepting AUTH without a new EHLO.
3398
3399 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3400 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3401 RFC.
3402
340310. Logging of TCP/IP connections (when configured) now happens in the main
3404 daemon process instead of the child process, so that the TCP/IP connection
3405 count is more accurate (but it can never be perfect).
3406
340711. The use of "drop" in a nested ACL was not being handled correctly in the
3408 outer ACL. Now, if condition failure induced by the nested "drop" causes
3409 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3410 or "require"), the connection is dropped.
3411
341212. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3413 that yield "discard" can now be used with an "accept" or a "discard" verb,
3414 but an error is generated for any others (because I can't see a useful way
3415 to define what should happen).
3416
341713. When an ACL is read dynamically from a file (or anywhere else), the lines
3418 are now processed in the same way as lines in the Exim configuration file.
3419 In particular, continuation lines are supported.
3420
342114. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3422
342315. Added -ti meaning -t -i.
3424
342516. Check for letters, digits, hyphens, and dots in the names of dnslist
3426 domains, and warn by logging if others are found.
3427
342817. At least on BSD, alignment is not guarenteed for the array of ifreq's
3429 returned from GIFCONF when Exim is trying to find the list of interfaces on
3430 a host. The code in os.c has been modified to copy each ifreq to an aligned
3431 structure in all cases.
3432
3433 Also, in some cases, the returned ifreq's were being copied to a 'struct
3434 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3435 means the last couple of bytes of an IPv6 address could be chopped if the
3436 ifreq contained only a normal sockaddr (14 bytes storage).
3437
343818. Named domain lists were not supported in the hosts_treat_as_local option.
3439 An entry such as +xxxx was not recognized, and was treated as a literal
3440 domain name.
3441
344219. Ensure that header lines added by a DATA ACL are included in the reject log
3443 if the ACL subsequently rejects the message.
3444
344520. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3446 MD5 (which is deprecated).
3447
344821. When testing a filter file using -bf, Exim was writing a message when it
3449 took the sender from a "From " line in the message, but it was not doing so
3450 when it took $return_path from a Return-Path: header line. It now does.
3451
345222. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3453 with a valid header line field name (a series of printing characters
3454 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3455
345623. Changed "disc" in the source to "disk" to conform to the documentation and
3457 the book and for uniformity.
3458
345924. Ignore Sendmail's -Ooption=value command line item.
3460
346125. When execve() failed while trying to run a command in a pipe transport,
3462 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3463 could be confused with a return value of 69 from the command itself. This
3464 has been changed to 127, the value the shell returns if it is asked to run
3465 a non-existent command. The wording for the related log line suggests a
3466 non-existent command as the problem.
3467
346826. If received_header_text expands to an empty string, do not add a Received:
3469 header line to the message. (Well, it adds a token one on the spool, but
3470 marks it "old" so that it doesn't get used or transmitted.)
3471
347227. Installed eximstats 1.28 (addition of -nt option).
3473
347428. There was no check for failure on the call to getsockname() in the daemon
3475 code. This can fail if there is a shortage of resources on the system, with
3476 ENOMEM, for example. A temporary error is now given on failure.
3477
347829. Contrary to the C standard, it seems that in some environments, the
3479 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3480 program. Exim now does this explicitly; it affects the formatting of
3481 timestamps using strftime().
3482
348330. If exiqsumm was given junk data, it threw up some uninitialized variable
3484 complaints. I've now initialized all the variables, to avoid this.
3485
348632. Header lines added by a system filter were not being "seen" during
3487 transport-time rewrites.
3488
348933. The info_callback() function passed to OpenSSL is set up with type void
3490 (*)(SSL *, int, int), as described somewhere. However, when calling the
3491 function (actually a macro) that sets it up, the type void(*)() is
3492 expected. I've put in a cast to prevent warnings from picky compilers.
3493
349434. If a DNS black list lookup found a CNAME record, but there were no A
3495 records associated with the domain it pointed at, Exim crashed.
3496
349735. If a DNS black list lookup returned more than one A record, Exim ignored
3498 all but the first. It now scans all returned addresses if a particular IP
3499 value is being sought. In this situation, the contents of the
3500 $dnslist_value variable are a list of all the addresses, separated by a
3501 comma and a space.
3502
350336. Tightened up the rules for host name lookups using reverse DNS. Exim used
3504 to accept a host name and all its aliases if the forward lookup for any of
3505 them yielded the IP address of the incoming connection. Now it accepts only
3506 those names whose forward lookup yields the correct IP address. Any other
3507 names are discarded. This closes a loophole whereby a rogue DNS
3508 administrator could create reverse DNS records to break through a
3509 wildcarded host restriction in an ACL.
3510
351137. If a user filter or a system filter that ran in a subprocess used any of
3512 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3513 the wrong values were passed to the pipe command ($thisaddress had the
3514 value of $0, $0 had the value of $1, etc). This bug was introduced by
3515 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3516
351738. Improved the line breaking for long SMTP error messages from ACLs.
3518 Previously, if there was no break point between 40 and 75 characters, Exim
3519 left the rest of the message alone. Two changes have been made: (a) I've
3520 reduced the minimum length to 35 characters; (b) if it can't find a break
3521 point between 35 and 75 characters, it looks ahead and uses the first one
3522 that it finds. This may give the occasional overlong line, but at least the
3523 remaining text gets split now.
3524
352539. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3526 file descriptors might be low, and that setting 1000 would always raise it.
3527 It turns out that in some environments, the limit is already over 1000 and
3528 that lowering it causes trouble. So now Exim takes care not to decrease it.
3529
353040. When delivering a message, the value of $return_path is set to $sender_
3531 address at the start of routing (routers may change the value). By an
3532 oversight, this default was not being set up when an address was tested by
3533 -bt or -bv, which affected the outcome if any router or filter referred to
3534 $return_path.
3535
353641. The idea of the "warn" ACL verb is that it adds a header or writes to the
3537 log only when "message" or "log_message" are set. However, if one of the
3538 conditions was an address verification, or a call to a nested ACL, the
3539 messages generated by the underlying test were being passed through. This
3540 no longer happens. The underlying message is available in $acl_verify_
3541 message for both "message" and "log_message" expansions, so it can be
3542 passed through if needed.
3543
354442. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3545 new expansion $bh_ to give the encoded byte string without charset
3546 translation. Translation happens only if iconv() is available; HAVE_ICONV
3547 indicates this at build time. HEADERS_CHARSET gives the charset to
3548 translate to; headers_charset can change it in the configuration, and
3549 "headers charset" can change it in an individual filter file.
3550
355143. Now that we have a default RFC 2047 charset (see above), the code in Exim
3552 that creates RFC 2047 encoded "words" labels them as that charset instead
3553 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3554 expansion operator; (ii) when Exim creates a From: line for a local
3555 message; (iii) when a header line is rewritten to include a "phrase" part.
3556
355744. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3558 buggy, causing it to skip the first lines of messages whose message ID
3559 ended in 'D'. This would not have bitten before Exim release 4.14, because
3560 message IDs were unlikely to end in 'D' before then. The effect was to have
3561 incorrect size information for certain domains.
3562
356345. #include "config.h" was missing at the start of the crypt16.c module. This
3564 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3565 noticed.
3566
356746. If there was a timeout during a "random" callout check, Exim treated it as
3568 a failure of the random address, and carried on sending RSET and the real
3569 address. If the delay was just some slowness somewhere, the response to the
3570 original RCPT would be taken as a response to RSET and so on, causing
3571 mayhem of various kinds.
3572
357347. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3574 when I implemented it. It didn't allow for the fact that some option values
3575 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3576 the right test for positive values.
3577
357848. Domain names in DNS records are case-independent. Exim always looks them up
3579 in lower case. Some resolvers return domain names in exactly the case they
3580 appear in the zone file, that is, they may contain uppercase letters. Not
3581 all resolvers do this - some return always lower case. Exim was treating a
3582 change of case by a resolver as a change of domain, similar to a widening
3583 of a domain abbreviation. This triggered its re-routing code and so it was
3584 trying to route what was effectively the same domain again. This normally
3585 caused routing to fail (because the router wouldn't handle the domain
3586 twice). Now Exim checks for this case specially, and just changes the
3587 casing of the domain that it ultimately uses when it transmits the message
3588 envelope.
3589
359049. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3591 module.
3592
359350. If a filter generated a file delivery with a non-absolute name (possible if
3594 no home directory exists for the router), the forbid_file option was not
3595 forbidding it.
3596
359751. Added '&' feature to dnslists, to provide bit mask matching in addition to
3598 the existing equality matching.
3599
360052. Exim was using ints instead of ino_t variables in some places where it was
3601 dealing with inode numbers.
3602
360353. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3604 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3605 called TMPDIR, and if it finds it is different, it changes its value.
3606
360754. The smtp_printf() function is now made available to local_scan() so
3608 additional output lines can be written before returning. There is also an
3609 smtp_fflush() function to enable the detection of a dropped connection.
3610 The variables smtp_input and smtp_batched_input are exported to
3611 local_scan().
3612
361355. Changed the default runtime configuration: the message "Unknown user"
3614 has been removed from the ACL, and instead placed on the localuser router,
3615 using the cannot_route_message feature. This means that any verification
3616 failures that generate their own messages won't get overridden. Similarly,
3617 the "Unrouteable address" message that was in the ACL for unverifiable
3618 relay addresses has also been removed.
3619
362056. Added hosts_avoid_esmtp to the smtp transport.
3621
362257. The exicyclog script was not checking for the esoteric option
3623 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3624 will work only if exicyclog is run under the appropriate euid.
3625
362658. Following a discussion on the list, the rules by which Exim recognises line
3627 endings on incoming messages have been changed. The -dropcr and drop_cr
3628 options are now no-ops, retained only for backwards compatibility. The
3629 following line terminators are recognized: LF CRLF CR. However, special
3630 processing applies to CR:
3631
3632 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3633 nor a local message in the state where . is a terminator.
3634
3635 (ii) If a bare CR is encountered in a header line, an extra space is added
3636 after the line terminator so as not to end the header. The reasoning
3637 behind this is that bare CRs in header lines are most likely either
3638 to be mistakes, or people trying to play silly games.
3639
364059. The size of a message, as listed by "-bp" or in the Exim monitor window,
3641 was being incorrectly given as 18 bytes larger than it should have been.
3642 This is a VOB (very old bug).
3643
364460. This may never have affected anything current, but just in case it has:
3645 When the local host is found other than at the start of a list of hosts,
3646 the local host, those with the same MX, and any that follow, are discarded.
3647 When the list in question was part of a longer list of hosts, the following
3648 hosts (not currently being processed) were also being discarded. This no
3649 longer happens. I'm not sure if this situation could ever has previously
3650 arisen.
3651
365261. Added the "/MX" feature to lists of hosts in the manualroute and query
3653 program routers.
3654
365562. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3656 header. This is something that is recommended in a new Internet Draft, and
3657 is something that is documented as being done by Sendmail. There are two
3658 possible values. For messages generated by the autoreply transport, Exim
3659 adds:
3660
3661 Auto-Submitted: auto-replied
3662
3663 whereas for all other generated messages (e.g. bounces) it adds
3664
3665 Auto-Submitted: auto-generated
3666
366763. The "personal" condition in filters now includes a test for the
3668 Auto-Submitted: header. If it contains the string "auto-" the message it
3669 not considered personal.
3670
367164. Added rcpt_include_affixes as a generic transport option.
3672
367365. Added queue_only_override (default true).
3674
367566. Added the syslog_duplication option.
3676
367767. If what should have been the first header line of a message consisted of
3678 a space followed by a colon, Exim was mis-interpreting it as a header line.
3679 It isn't of course - it is syntactically invalid and should therefore be
3680 treated as the start of the message body. The misbehaviour could have
3681 caused a number of strange effects, including loss of data in subsequent
3682 header lines, and spool format errors.
3683
368468. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3685 client host had authenticated. This control can now be exercised by an ACL
3686 for more flexibility.
3687