A tweak to the "H" retry rule from Michael Haardt.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
3cd34f13 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.306 2006/02/16 16:37:57 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
145396a6
PH
194PH/38 Panic-die if .include specifies a non-absolute path.
195
3cd34f13
PH
196PH/39 A tweak to the "H" retry rule from its user.
197
214e2000 198
5de37277
PH
199Exim version 4.60
200-----------------
201
cc38ddbf
PH
202PH/01 Two changes to the default runtime configuration:
203
204 (1) Move the checks for relay_from_hosts and authenticated clients from
205 after to before the (commented out) DNS black list checks.
206
207 (2) Add control=submission to the relay_from_hosts and authenticated
208 clients checks, on the grounds that messages accepted by these
209 statements are most likely to be submissions.
5de37277 210
72fdd6ae
PH
211PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
212
213 (1) Generate an error if the third argument for the ${prvs expansion is
214 not a single digit.
215
216 (2) Treat a missing third argument of ${prvscheck as if it were an empty
217 string.
218
219 (3) Reset the variables that are obtained from the first argument of
220 ${prvscheck and used in the second argument before leaving the code,
221 because their memory is reclaimed, so using them afterwards may do
222 silly things.
223
224 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
225 one (it's much easier than Tom thought :-).
226
227 (5) Because of (4), we can now allow for the use of $prvscheck_result
228 inside the third argument.
cb9328de 229
cb741023
PH
230PH/03 For some reason, the default setting of PATH when running a command from
231 a pipe transport was just "/usr/bin". I have changed it to
232 "/bin:/usr/bin".
233
f174f16e
PH
234PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
235 anything to be listed in the output from -bV.
b2f5a032 236
c25242d7
PH
237PH/05 When a filter generated an autoreply, the entire To: header line was
238 quoted in the delivery log line, like this:
239
240 => >A.N.Other <ano@some.domain> <original@ddress> ...
241
242 This has been changed so that it extracts the operative address. There
243 may be more than one such address. If so, they are comma-separated, like
244 this:
245
246 => >ano@some.domain,ona@other.domain <original@ddress> ...
247
82c19f95
PH
248PH/06 When a client host used a correct literal IP address in a HELO or EHLO
249 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
250 not being looked up in the rDNS to get a host name, Exim was showing the
251 IP address twice in Received: lines, even though the IP addresses were
252 identical. For example:
253
254 Received: from [1.2.3.4] (helo=[1.2.3.4])
255
256 However, if the real host name was known, it was omitting the HELO data
257 if it matched the actual IP address. This has been tidied up so that it
258 doesn't show the same IP address twice.
259
d7ffbc12
PH
260PH/07 When both +timestamp and +memory debugging was on, the value given by
261 $tod_xxx expansions could be wrong, because the tod_stamp() function was
262 called by the debug printing, thereby overwriting the timestamp buffer.
263 Debugging no longer uses the tod_stamp() function when +timestamp is set.
264
9f526266
PH
265PH/08 When the original message was included in an autoreply transport, it
266 always said "this is a copy of the message, including all the headers",
267 even if body_only or headers_only was set. It now gives an appropriate
268 message.
269
87fcc8b9
PH
270PH/09 Applied a patch from the Sieve maintainer which:
271
272 o fixes some comments
273 o adds the (disabled) notify extension core
274 o adds some debug output for the result of if/elsif tests
275 o points to the current vacation draft in the documentation
276 and documents the missing references header update
277
278 and most important:
279
280 o fixes a bug in processing the envelope test (when testing
281 multiple envelope elements, the last element determinted the
282 result)
283
456682f5
PH
284PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
285 Electronic Mail") by including:
286
287 Auto-submitted: auto-generated
288
289 in the messages that it generates (bounce messages and others, such as
290 warnings). In the case of bounce messages for non-SMTP mesages, there was
291 also a typo: it was using "Auto_submitted" (underscore instead of
292 hyphen). Since every message generated by Exim is necessarily in response
293 to another message, thes have all been changed to:
294
295 Auto-Submitted: auto-replied
296
297 in accordance with these statements in the RFC:
298
299 The auto-replied keyword:
300
301 - SHOULD be used on messages sent in direct response to another
302 message by an automatic process,
303
304 - MUST NOT be used on manually-generated messages,
305
306 - MAY be used on Delivery Status Notifications (DSNs) and Message
307 Disposition Notifications (MDNs),
308
309 - MUST NOT be used on messages generated by automatic or periodic
310 processes, except for messages which are automatic responses to
311 other messages.
312
3e46c1aa
PH
313PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
314 to the default Received: header definition.
456682f5 315
49826d12
PH
316PH/12 Added log selector acl_warn_skipped (default on).
317
eba0c039
PH
318PH/13 After a successful wildlsearch lookup, discard the values of numeric
319 variables because (a) they are in the wrong storage pool and (b) even if
320 they were copied, it wouldn't work properly because of the caching.
321
a0d6ba8a
PH
322PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
323 checking when decoding. Apparently there are clients that generate
324 overlong encoded strings. Why am I not surprised?
325
f0917727
PH
326PH/15 If the first argument of "${if match_address" was not empty, but did not
327 contain an "@" character, Exim crashed. Now it writes a panic log message
328 and treats the condition as false.
329
096fee00
PH
330PH/16 In autoreply, treat an empty string for "once" the same as unset.
331
024bd3c2
PH
332PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
333 extension "envelope-auth". The code is finished and in agreement with
334 other implementations, but there is no documentation so far and in fact,
335 nobody wrote the draft yet. This extension is currently #undef'ed, thus
336 not changing the active code.
337
338 Print executed "if" and "elsif" statements when debugging is used. This
339 helps a great deal to understand what a filter does.
340
341 Document more things not specified clearly in RFC3028. I had all this
342 sorted out, when out of a sudden new issues came to my mind. Oops."
343
df199fec
PH
344PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
345 (Bugzilla #53).
346
d27f1df3
PH
347PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
348 canonical form (as documented). However, after a host name lookup from
349 the IP address, check_host() was doing a simple string comparison with
350 addresses acquired from the DNS when checking that the found name did
351 have the original IP as one of its addresses. Since any found IPv6
352 addresses are likely to be in abbreviated form, the comparison could
353 fail. Luckily, there already exists a function for doing the comparison
354 by converting both addresses to binary, so now that is used instead of
355 the text comparison.
356
96776534
PH
357PH/20 There was another similar case to PH/19, when a complete host name was
358 given in a host list; looking up its IP address could give an abbreviated
359 form, whereas the current host's name might or might not be abbreviated.
360 The same fix has been applied.
361
5de37277 362
9a799bc0
PH
363Exim version 4.54
364-----------------
365
366PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
367 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
368 It now does.
369
99a4b039
PH
370PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
371 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
372
373PH/03 Typo: missing ".o" in src/pcre/Makefile.
374
4b233853
PH
375PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
376 header line, restrict the check to what is listed in RFCs 2369 and 2929.
377 Also, for "Auto-Submitted", treat anything other than "no" as
378 non-personal, in accordance with RFC 3834. (Previously it treated
379 anything starting "auto-" as non-personal.)
380
8857ccfd
PH
381TF/01 The control=submission/name=... option had a problem with syntax
382 errors if the name included a slash character. The /name= option
383 now slurps the rest of the string, so it can include any characters
384 but it must come last in the list of options (after /sender_retain
385 or /domain=).
386
433a2980
PH
387PH/05 Some modifications to the interface to the fake nameserver for the new
388 testing suite.
389
3e46c1aa 390
9a799bc0 391
e3a311ba
TK
392Exim version 4.53
393-----------------
394
395TK/01 Added the "success_on_redirect" address verification option. See
396 NewStuff for rationale and an example.
397
13b685f9
PH
398PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
399
395ff96d
PH
400PH/02 Patch to exigrep to allow it to work on syslog lines.
401
5b68f6e4
PH
402PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
403 fread() to skip over the body file's header line, because in Cygwin the
404 header line is locked and is inaccessible.
405
1ab52c69
PH
406PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
407 co-exist for some time) to make it clear that it is the Exim ID that is
408 referenced, not the Message-ID: header line.
409
b07e6aa3
PH
410PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
411 string_format() function, because snprintf() does not exist on all
412 operating systems.
413
254e032f
PH
414PH/06 The use of forbid_filter_existstest now also locks out the use of the
415 ${stat: expansion item.
416
3af76a81
PH
417PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
418 protocol synchronization error", to keep the pedants happy.
419
2548ba04
PH
420PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
421 well as for IRIX systems, when gcc is being used. See the host.c source
422 file for comments.
423
b6c6011d
PH
424PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
425
cf39cf57
PH
426PH/10 Named domain lists were not working if used in a queue_smtp_domains
427 setting.
428
f1513293
PH
429PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
430 transport and to the smtp transport in LMTP mode.
431
727549a4
PH
432TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
433
af46795e
PH
434PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
435 run a filter in a subprocess. This could lead to confusion in subsequent
436 lookups in the parent process. There should also be a search_tidyup() at
437 the end of the subprocess.
438
d7b47fd0
PH
439PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
440 only if the host matched helo_try_verify_hosts, which caused the
441 verification to occur when the EHLO/HELO command was issued. The ACL just
442 tested the remembered result. Now, if a previous verification attempt has
443 not happened, "verify = helo" does it there and then.
444
ee744174
JJ
445JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
446
b582ab87
PH
447TK/03 Fix log output including CR from clamd.
448
41a13e0a
PH
449PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
450 exist provoked a memory error which could cause a segfault.
451
f625cc5a
PH
452PH/15 Installed PCRE 6.2
453
454PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
455
21f7af35
PH
456PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
457 of the problem. Specifically, suggested +O2 rather than +O1 for the
458 HP-UX compiler.
459
31480e42
PH
460PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
461
2d280592
PH
462PH/20 If a delivery was routed to a non-standard port by means of an SRV
463 record, the port was not correctly logged when the outgoing_port log
464 selector was set (it logged the transort's default port).
465
7cd1141b
PH
466PH/21 Added support for host-specific ports to manualroute, queryprogram,
467 fallback_hosts, and "hosts" in the smtp transport.
468
469PH/22 If the log selector "outgoing_port" is set, the port is now also given on
470 host errors such as "Connection refused".
471
750af86e
PH
472PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
473 authentication with radiusclient 0.4.9:
474
475 - Error returned from rc_read_config was caught wrongly
476 - Username/password not passed on to radius server due to wrong length.
477
478 The presumption is that some radiusclient API changes for 4.51/PH/17
479 were not taken care of correctly. The code is still untested by me (my
480 Linux distribution still has 0.3.2 of radiusclient), but it was
481 contributed by a Radius user.
482
483PH/24 When doing a callout, the value of $domain wasn't set correctly when
484 expanding the "port" option of the smtp transport.
485
4304270b
TK
486TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
487 while reading a MIME header. Thanks to Tom Hughes for a patch.
488
750af86e
PH
489PH/24 Include config.h inside local_scan.h so that configuration settings are
490 available.
491
64ffc24f
PH
492PH/25 Make $smtp_command_argument available after all SMTP commands. This means
493 that in an ACL for RCPT (for example), you can examine exactly what was
494 received.
495
5dd9625b
PH
496PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
497 commands, but it was not correctly comparing the address with the actual
498 client host address. Thus, it would show the EHLO address in Received:
499 header lines when this was not necessary.
500
5591031b
PH
501PH/27 Added the % operator to ${eval:}.
502
ba18e66a
PH
503PH/28 Exim tries to create and chdir to its spool directory when it starts;
504 it should be ignoring failures (because with -C, for example, it has lost
505 privilege). It wasn't ignoring creation failures other than "already
506 exists".
507
9cec981f
PH
508PH/29 Added "crypteq" to the list of supported features that Exim outputs when
509 -bV or -d is used.
510
aa2b5c79
PH
511PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
512 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
513 too many continuations, the temporary file was not being removed, and the
514 return code was incorrect.
aa2b5c79 515
48a53b7f
PH
516PH/31 Missing "BOOL" in function definition in filtertest.c.
517
1c59d63b
PH
518PH/32 Applied Sieve patches from the maintainer.
519
671012da
TK
520TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
521
1509d3a8
PH
522PH/33 Added "verify = not_blind".
523
524PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
525 Local/Makefile (with some defaults set). These are used in built scripts
526 such as exicyclog, but they have never been used in the exim_install
527 script (though there are many overriding facilities there). I have
528 arranged that the exim_install script now takes note of these two
529 settings.
530
531PH/35 Installed configuration files for Dragonfly.
532
2fe1a124
PH
533PH/36 When a locally submitted message by a trusted user did not contain a
534 From: header, and the sender address was obtained from -f or from an SMTP
535 MAIL command, and the trusted user did not use -F to supply a sender
536 name, $originator_name was incorrectly used when constructing a From:
537 header. Furthermore, $originator_name was used for submission mode
538 messages from external hosts without From: headers in a similar way,
539 which is clearly wrong.
540
8800895a
PH
541PH/37 Added control=suppress_local_fixups.
542
ccfdb010
PH
543PH/38 When log_selector = +received_sender was set, and the addition of the
544 sender made the log line's construction buffer exactly full, or one byte
545 less than full, an overflow happened when the terminating "\n" was
546 subsequently added.
547
1130bfb0
PH
548PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
549 when the result of a list match is failure because a DNS lookup failed.
550
ebcb507f
PH
551PH/40 RM_COMMAND is now used in the building process.
552
c35e155c
PH
553PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
554 the "build-* directories that it finds.
555
95d1f782
PH
556PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
557 address in a domain literal was a prefix of an interface address.
558
fd6de02e
PH
559PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
560 when verifying a sender address, unless rewrite_headers is false.
561
58de37c5
PH
562PH/44 Wrote a long comment about why errors_to addresses are verified as
563 recipients, not senders.
564
261cf466
TF
565TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
566 the ratelimit ACL was added.
567
3ee512ff
PH
568PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
569
e08c430f
PH
570PH/46 Added extra information about PostgreSQL errors to the error string.
571
bef5a11f
PH
572PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
573 avoiding the need to install special zones in a real server. This is
574 backwards compatible; if it can't find the fake resolver, it drops back.
575 Thus, both old and new test suites can be run.
576
7546de58
TF
577TF/02 Added util/ratelimit.pl
578
e5d5a95f
TF
579TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
580 clock is set back in time.
581
2e88a017
TF
582TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
583 Candler <B.Candler@pobox.com>.
584
a5f65aa4
TF
585TF/05 The fix for PH/43 was not completely correct; widen_domains is always
586 OK for addresses that are the result of redirections.
587
e7726cbf
PH
588PH/48 A number of further additions for the benefit of the new test suite,
589 including a fake gethostbyname() that interfaces to the fake DNS resolver
590 (see PH/47 above).
591
a7fdad5b
TF
592TF/06 The fix for widen_domains has also been applied to qualify_single and
593 search_parents which are the other dnslookup options that can cause
594 header rewrites.
595
6af56900
PH
596PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
597 type ("H").
598
0925ede6
PH
599PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
600
66afa403
TF
601TF/07 Exim produced the error message "an SRV record indicated no SMTP
602 service" if it encountered an MX record with an empty target hostname.
603 The message is now "an MX or SRV record indicated no SMTP service".
604
0154e85a
TF
605TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
606 if the DNS of the sending site is misconfigured. This is quite a
607 common situation. This change restores the behaviour of treating a
608 helo verification defer as a failure.
609
16f12c76
PH
610PH/51 If self=fail was set on a router, the bounce message did not include the
611 actual error message.
612
bbe902f0 613
e5a9dba6
PH
614Exim version 4.52
615-----------------
616
617TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
618
22c3b60b
PH
619PH/01 When a transport filter timed out in a pipe delivery, and the pipe
620 command itself ended in error, the underlying message about the transport
621 filter timeout was being overwritten with the pipe command error. Now the
622 underlying error message should be appended to the second error message.
623
06a9b4b5
PH
624TK/01 Fix poll() being unavailable on Mac OSX 10.2.
625
c1ac6996
PH
626PH/02 Reduce the amount of output that "make" produces by default. Full output
627 can still be requested.
628
9c7a242c
PH
629PH/03 The warning log line about a condition test deferring for a "warn" verb
630 was being output only once per connection, rather than after each
631 occurrence (because it was using the same function as for successful
632 "warn" verbs). This seems wrong, so I have changed it.
633
87ba3f5f
PH
634TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
635 it should not have, which might have caused a crash in the right
636 circumstances, but probably never did.
637
638PH/04 Installed a modified version of Tony Finch's patch to make submission
639 mode fix the return path as well as the Sender: header line, and to
640 add a /name= option so that you can make the user's friendly name appear
641 in the header line.
642
29aba418
TF
643TF/03 Added the control = fakedefer ACL modifier.
644
fe0dab11
TF
645TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
646 Mark Lowes for thorough testing.
870f6ba8 647
11d337a4
TK
648TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
649
650TK/03 Merged latest SRS patch from Miles Wilton.
651
415c8f3b
PH
652PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
653 with the definition in sysexits.h (which is #included earlier).
654 Fortunately, Exim does not actually use EX_OK. The code used to try to
655 preserve the sysexits.h value, by assumimg that macro definitions were
656 scanned for macro replacements. I have been disabused of this notion,
657 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 658
958541e9
PH
659PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
660 in the smtp transport. When a block could not be written in a single
661 write() function, the timeout was being re-applied to each part-write.
662 This seems wrong - if the receiver was accepting one byte at a time it
663 would take for ever. The timeout is now adjusted when this happens. It
664 doesn't have to be particularly precise.
665
c206415f
TK
666TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
667 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
668
2a4be8f9
PH
669PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
670 without a domain if the check to <postmaster@domain> fails.
671
1cba11c5
SC
672SC/01 Eximstats: added -xls and the ability to specify output files
673 (patch written by Frank Heydlauf).
674
675SC/02 Eximstats: use FileHandles for outputing results.
676
677SC/03 Eximstats: allow any combination of xls, txt, and html output.
678
679SC/04 Eximstats: fixed display of large numbers with -nvr option
680
681SC/05 Eximstats: fixed merging of reports with empty tables.
682
683SC/06 Eximstats: added the -include_original_destination flag
684
685SC/07 Eximstats: removed tabs and trailing whitespace.
686
1005d00e
TK
687TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
688
689TK/06 MBOX spool code: Add real "From " MBOX separator line
690 so the .eml file is really in mbox format (even though
691 most programs do not really care). Patch from Alex Miller.
692
693TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
694 The latter is generated from $received_to and is only set if the
695 message has one envelope recipient. SA can use these headers,
696 obviously out-of-the-box. Patch from Alex Miller.
697
9b4768fa
PH
698PH/08 The ${def test on a variable was returning false if the variable's
699 value was "0", contrary to what the specification has always said!
700 The result should be true unless the variable is empty.
701
702PH/09 The syntax error of a character other than { following "${if
703 def:variable_name" (after optional whitespace) was not being diagnosed.
704 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
705 accidental colon was present, for example, could give incorrect results.
706
0d7eb84a
PH
707PH/10 Tidied the code in a number of places where the st_size field of a stat()
708 result is used (not including appendfile, where other changes are about
709 to be made).
710
711PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
712 This involved changing a lot of size variables from int to off_t. It
713 should work with maildirs and everything.
714
40727bee
TK
715TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
716 spamd dying while we are connected to it.
717
554d2369
TF
718TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
719 <jgh@wizmail.org>
720
1f922db1
PH
721PH/12 Applied Alex Kiernan's patch for the API change for the error callback
722 function for BDB 4.3.
723
ef213c3b
PH
724PH/13 Changed auto_thaw such that it does not apply to bounce messages.
725
8ac170f3
PH
726PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
727 the sources for PCRE have been re-arranged and more files are now
728 involved.
729
b1c749bb
PH
730PH/15 The code I had for printing potentially long long variables in PH/11
731 above was not the best (it lost precision). The length of off_t variables
732 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
733 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
734 to be "long long int" or "long int". This is needed for the internal
735 formatting function string_vformat().
b1c749bb 736
4aac9b49
PH
737PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
738 the configuration file to be ":syslog", then the script "guesses" where
739 the logs files are, rather than using the compiled in default. In our
740 case the guess is not the same as the compiled default, so the script
741 suddenly stopped working when I started to use syslog. The patch checks
742 to see if log_file_path is "". If so, it attempts to read it from exim
743 with no configuration file to get the compiled in version, before it
744 falls back to the previous guessing code."
745
294520c8
TK
746TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
747 implementing BATV in an Exim configuration. See NewStuff for the gory
748 details.
749
5bd022fe
PH
750PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
751 Makefile that are specific to HP-UX.
752
90e9ce59
PH
753PH/18 If the "use_postmaster" option was set for a recipient callout together
754 with the "random" option, the postmaster address was used as the MAIL
755 FROM address for the random test, but not for the subsequent recipient
756 test. It is now used for both.
757
5ea81592
PH
758PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
759 patch removes a few documentation additions to RFC 3028, because the
760 latest draft now contains them. It adds the new en;ascii-case comparator
761 and a new error check for 8bit text in MIME parts. Comparator and
762 require names are now matched exactly. I enabled the subaddress
763 extension, but it is not well tested yet (read: it works for me)."
764
c6c2dc1d
PH
765PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
766 rework some of the code of TK/09 above to avoid the hardwired use of
767 "%lld" and "long long". Replaced the call to snprintf() with a call to
768 string_vformat().
769
fffffe4c
PH
770PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
771 records point to non-existent hosts", "retry timeout exceeded", and
772 "retry time not reached for any host after a long failure period".
ca02eafb 773
9a26b6b2
PH
774PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
775 experimental DomainKeys support:
776
777 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
778 (2) On an error such as an illegally used "control", the wrong name for
779 the control was given.
780
781 These problems did NOT occur unless DomainKeys support was compiled.
782
4aee0225
PH
783PH/23 Added daemon_startup_retries and daemon_startup_sleep.
784
32d668a5
PH
785PH/24 Added ${if match_ip condition.
786
8187c3f3
PH
787PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
788 databases so that it will be absolutely obvious if a crash occurs in the
789 DB library. This is a regular occurrence (often caused by mis-matched
790 db.h files).
791
ff790e47 792PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
793 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
794 picked up on a user's system that detects such things. There doesn't seem
795 to be a gcc warning option for this - only an attribute that has to be
796 put on the function's prototype. It seems that in Fedora Core 4 they have
797 set this on a number of new functions. No doubt there will be more in due
798 course.
ff790e47 799
5417f6d1
PH
800PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
801 specify a transport. However, if an address that was verified by such a
802 router was the subject of a callout, Exim crashed because it tried to
803 read the rcpt_include_affixes from the non-existent transport. Now it
804 just assumes that the setting of that option is false. This bug was
805 introduced by 4.51/PH/31.
806
59cf8544
PH
807PH/28 Changed -d+all to exclude +memory, because that information is very
808 rarely of interest, but it makes the output a lot bigger. People tend to
809 do -d+all out of habit.
810
e7ad8a65
PH
811PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
812 code in os-type was giving problems when libc.so lives in lib64, like on
813 x86_64 Fedora Core.
814
ade42478
PH
815PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
816 aren't the modern standard, and it seems that some systems' include files
817 don't always have them. Exim was already checking for some of the newer
818 ones like T_AAAA, and defining it itself. I've added checks for all the
819 record types that Exim uses.
820
182ad5cf
PH
821PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
822 not automatically generating a new one, as it is supposed to. This
823 prevented TLS from working. If the file did exist, but contained invalid
824 data, a new version was generated, as expected. It was only the case of a
825 non-existent file that was broken.
826
b0d9fc80
TK
827TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
828 with a change in libdomainkeys > 0.64.
829
830TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
831 from DNS. If the selector record carries the flag, it now has
832 precedence over the domain-wide flag.
833
834TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
835
47c7a64a
PH
836PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
837 the use of a port name that isn't defined in /etc/services) occurred, the
838 message was deferred as in a normal delivery, and thus remained on the
839 spool, instead of being failed because of the mua_wrapper setting. This
840 is now fixed, and I tidied up some of the mua_wrapper messages at the
841 same time.
842
a388bce4
SC
843SC/08 Eximstats: whilst parsing the mainlog(s), store information about
844 the messages in a hash of arrays rather than using individual hashes.
845 This is a bit cleaner and results in dramatic memory savings, albeit
846 at a slight CPU cost.
847
848SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
849 as requested by Marc Sherman.
850
851SC/10 Eximstats: added histograms for user specified patterns as requested
852 by Marc Sherman.
853
0793e4ed
SC
854SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
855
c58b88df
PH
856PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
857 fopen() in the content-scanning modules that did not already have it.
858
e7ad8a65 859
7982096b
PH
860Exim version 4.51
861-----------------
862
1a46a8c5
PH
863TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
864 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
865
2f079f46 866TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
867
868TK/03 Fix ACL "regex" condition not being available in MIME ACL.
869
870PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
871 to test Sieve filters that use "vacation".
872
873PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
874 that changes the way the GnuTLS parameters are stored in the cache file.
875 The new format can be generated externally. For backward compatibility,
876 if the data in the cache doesn't make sense, Exim assumes it has read an
877 old-format file, and it generates new data and writes a new file. This
878 means that you can't go back to an older release without removing the
879 file.
880
881PH/03 A redirect router that has both "unseen" and "one_time" set does not
882 work if there are any delivery delays because "one_time" forces the
883 parent to be marked "delivered", so its unseen clone is never tried
884 again. For this reason, Exim now forbids the simultaneous setting of
885 these two options.
886
887PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
888 redirected to themselves ("homonym" addresses). Read the long ChangeLog
889 entry if you want to know the details. The fix, however, neglected to
890 consider the case when local delivery batching is involved. The test for
891 "previously delivered" was not happening when checking to see if an
892 address could be batched with a previous (undelivered) one; under
893 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 894 address.
1a46a8c5
PH
895
896PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
897 in its include files, and this causes problems building Exim.
898
899PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
900 header_syntax) but Exim was just ignoring anything given after a slash.
901 In particular, this caused confusion with an attempt to use "verify =
902 reverse_host_lookup/defer_ok". An error is now given when options are
903 supplied for verify items that do not have them. (Maybe reverse_host_
904 lookup should have a defer_ok option, but that's a different point.)
905
906PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
907 defined by RFC 821) to 2048, because there were problems with some AUTH
908 commands, and RFC 1869 says the size should be increased for extended
909 SMTP commands that take arguments.
910
911PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
912 Finch).
913
914PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
915 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 916
49c2d5ea
PH
917PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
918 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
919 whatever). Otherwise the password may appear in the log. Following change
920 PH/42 below, there is no longer a chance of it appearing in a bounce
921 message.
49c2d5ea 922
bf759a8b
PH
923PH/11 Installed exipick version 20050225.0 from John Jetmore.
924
83364d30
PH
925PH/12 If the last host in a fallback_hosts list was multihomed, only the first
926 of its addresses was ever tried. (Bugzilla bug #2.)
927
7999bbd7
PH
928PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
929 the result incorrectly in the debug output. (It correctly added a newline
930 to what was transported.)
931
7dbf77c9
PH
932TF/01 Added $received_time.
933
74e0617f
PH
934PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
935 commented out examples of how to interface to a virus scanner and to
936 SpamAssassin. Also added commented examples of av_scanner and
937 spamd_address settings.
938
2f079f46
PH
939PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
940 and controls are allowed in which ACLs. There were a couple of minor
941 errors. Some of the entries in the conditions table (which is a table of
942 where they are NOT allowed) were getting very unwieldy; rewrote them as a
943 negation of where the condition IS allowed.
944
8c841523
PH
945PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
946
7766a4f0
PH
947PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
948 header file does not have a version number, so I've had to invent a new
949 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
950 API. The code is untested by me (my Linux distribution still has 0.3.2 of
951 radiusclient), but it was contributed by a Radius user.
952
8b417f2c
PH
953PH/18 Installed Lars Mainka's patch for the support of CRL collections in
954 files or directories, for OpenSSL.
955
901f42cb
PH
956PH/19 When an Exim process that is running as root has to create an Exim log
957 file, it does so in a subprocess that runs as exim:exim so as to get the
958 ownership right at creation (otherwise, other Exim processes might see
959 the file with the wrong ownership). There was no test for failure of this
960 fork() call, which would lead to the process getting stuck as it waited
961 for a non-existent subprocess. Forks do occasionally fail when resources
962 run out. I reviewed all the other calls to fork(); they all seem to check
963 for failure.
964
f9b9210e
PH
965PH/20 When checking for unexpected SMTP input at connect time (before writing
966 the banner), Exim was not dealing correctly with a non-positive return
967 from the read() function. If the client had disconnected by this time,
968 the result was a log entry for a synchronization error with an empty
969 string after "input=" when read() returned zero. If read() returned -1
970 (an event I could not check), uninitialized data bytes were printed.
971 There were reports of junk text (parts of files, etc) appearing after
972 "input=".
973
54cdb463
PH
974PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
975
cf00dad6
PH
976PH/22 Added support for macro redefinition, and (re)definition in between
977 driver and ACL definitions.
978
acb1b346
PH
979PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
980 forgetting to use the resulting value; it was using the unexpanded value.
981
c5ddb310
PH
982PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
983 hadn't been configured. The fix is from Juergen Kreileder, who
984 understands it better than I do:
985
986 "Here's what I see happening with three configured cyrus_sasl
987 authenticators configured (plain, login, cram-md5):
988
989 On startup auth_cyrus_sasl_init() gets called for each of these.
990 This means three calls to sasl_listmech() without a specified mech_list.
991 => SASL tests which mechs of all available mechs actually work
992 => three warnings about OTP not working
993 => the returned list contains: plain, login, cram-md5, digest-md5, ...
994
995 With the patch, sasl_listmech() also gets called three times. But now
996 SASL's mech_list option is set to the server_mech specified in the the
997 authenticator. Or in other words, the answer from sasl_listmech()
998 gets limited to just the mech you're testing for (which is different
999 for each call.)
1000 => the return list contains just 'plain' or 'login', 'cram-md5' or
1001 nothing depending on the value of ob->server_mech.
1002
1003 I've just tested the patch: Authentication still works fine,
1004 unavailable mechs specified in the exim configuration are still
1005 caught, and the auth.log warnings about OTP are gone."
1006
31619da6
PH
1007PH/25 When debugging is enabled, the contents of the command line are added
1008 to the debugging output, even when log_selector=+arguments is not
1009 specified.
1010
bebaf0fc
PH
1011PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1012 answer is "GNU", and only if the return is "GNU/something" is the answer
1013 "Linux".
1014
475fe28a
PH
1015PH/27 $acl_verify_message is now set immediately after the failure of a
1016 verification in an ACL, and so is available in subsequent modifiers. In
1017 particular, the message can be preserved by coding like this:
1018
1019 warn !verify = sender
1020 set acl_m0 = $acl_verify_message
1021
1022 Previously, $acl_verify_message was set only while expanding "message"
1023 and "log_message" when a very denied access.
1024
7e8bec7a
PH
1025PH/28 Modified OS/os.c-Linux with
1026
1027 -#ifndef OS_LOAD_AVERAGE
1028 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1029
1030 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1031 nomenclature these days.)
1032
e4a89c47
PH
1033PH/29 Installed patch from the Sieve maintainer that adds the options
1034 sieve_useraddress and sieve_subaddress to the redirect router.
1035
5ca2a9a1
PH
1036PH/30 In these circumstances:
1037 . Two addresses routed to the same list of hosts;
1038 . First host does not offer TLS;
1039 . First host accepts first address;
1040 . First host gives temporary error to second address;
1041 . Second host offers TLS and a TLS session is established;
1042 . Second host accepts second address.
1043 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1044 and peerdn, if requested) that were in fact used only for the second
1045 address.
7e8bec7a 1046
c688b954
PH
1047PH/31 When doing a callout as part of verifying an address, Exim was not paying
1048 attention to any local part prefix or suffix that was matched by the
1049 router that accepted the address. It now behaves in the same way as it
1050 does for delivery: the affixes are removed from the local part unless
1051 rcpt_include_affixes is set on the transport.
1052
fed77020
PH
1053PH/32 Add the sender address, as F=<...>, to the log line when logging a
1054 timeout during the DATA phase of an incoming message.
1055
7fe1560f
PH
1056PH/33 Sieve envelope tests were broken for match types other than :is. I have
1057 applied a patch sanctioned by the Sieve maintainer.
c688b954 1058
ebb6e6d5
PH
1059PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1060 the uid or gid is negative. A case of a negative gid caused this to be
1061 noticed. The fix allows for either to be negative.
1062
9c4e8f60
PH
1063PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1064 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1065 overlooked.
1066
1067PH/36 The change PH/12 above was broken. Fixed it.
1068
d7174846
PH
1069PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1070 the grounds that routing the same address twice would always produce the
1071 same answer. This might have been true once, but it is certainly no
1072 longer true now. Routing a child address may depend on the previous
1073 routing that produced that child. Some complicated redirection strategies
1074 went wrong when messages had multiple recipients, and made Exim's
1075 behaviour dependent on the order in which the addresses were given.
1076
1077 I have moved the duplicate checking until after the routing is complete.
1078 Exim scans the addresses that are assigned to local and remote
1079 transports, and removes any duplicates. This means that more work will be
1080 done, as duplicates will always all be routed, but duplicates are
1081 presumably rare, so I don't expect this is of any significance.
1082
1083 For deliveries to pipes, files, and autoreplies, the duplicate checking
1084 still happens during the routing process, since they are not going to be
1085 routed further.
1086
cfe75fc3
PH
1087PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1088 It corrects a timeout issue with spamd. This is Ian's comment: "The
1089 background is that sometimes spamd either never reads data from a
1090 connection it has accepted, or it never writes response data. The exiscan
1091 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1092 blindly assumes that writes won't block so it may never time out."
1093
be22d70e
PH
1094PH/39 Allow G after quota size as well as K and M.
1095
0612b098
PH
1096PH/40 The value set for $authenticated_id in an authenticator may not contain
1097 binary zeroes or newlines because the value is written to log lines and
1098 to spool files. There was no check on this. Now the value is run through
1099 the string_printing() function so that such characters are converted to
1100 printable escape sequences.
1101
2e0c1448
PH
1102PH/41 $message_linecount is a new variable that contains the total number of
1103 lines in the message. Compare $body_linecount, which is the count for the
1104 body only.
1105
447d236c
PH
1106PH/42 Exim no longer gives details of delivery errors for specific addresses in
1107 bounce and delay warning messages, except in certain special cases, which
1108 are as follows:
1109
1110 (a) An SMTP error message from a remote host;
1111 (b) A message specified in a :fail: redirection;
1112 (c) A message specified in a "fail" command in a system filter;
1113 (d) A message specified in a FAIL return from the queryprogram router;
1114 (e) A message specified by the cannot_route_message router option.
1115
1116 In these cases only, Exim does include the error details in bounce and
1117 warning messages. There are also a few cases where bland messages such
1118 as "unrouteable address" or "local delivery error" are given.
1119
d20976dc
PH
1120PH/43 $value is now also set for the "else" part of a ${run expansion.
1121
f656d135
PH
1122PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1123 being worked on, but at least Exim now implements the latest version to
1124 play with."
1125
2e2a30b4
PH
1126PH/45 In a pipe transport, although a timeout while waiting for the pipe
1127 process to complete was treated as a delivery failure, a timeout while
1128 writing the message to the pipe was logged, but erroneously treated as a
1129 successful delivery. Such timeouts include transport filter timeouts. For
1130 consistency with the overall process timeout, these timeouts are now
1131 treated as errors, giving rise to delivery failures by default. However,
1132 there is now a new Boolean option for the pipe transport called
1133 timeout_defer, which, if set TRUE, converts the failures into defers for
1134 both kinds of timeout. A transport filter timeout is now identified in
1135 the log output.
1136
9176e9f0
PH
1137PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1138 systems where "make" and "gmake" are different, calling "gmake" at top
1139 level broke things. I've arranged for the value of $(MAKE) to be passed
1140 from the Makefile to this script so that it can call the same version of
1141 "make".
1142
7982096b 1143
bbe902f0
PH
1144A note about Exim versions 4.44 and 4.50
1145----------------------------------------
1146
1147Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1148changes of various kinds. As a consequence, a big documentation update was
1149needed. This delayed the release for rather longer than seemed good, especially
1150in the light of a couple of (minor) security issues. Therefore, the changes
1151that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1152release. So 4.44 and 4.50 are in effect two different branches that both start
1153from 4.43.
1154
1155I have left the 4.50 change log unchanged; it contains all the changes since
11564.43. The change log for 4.44 is below; many of its items are identical to
1157those for 4.50. This seems to be the most sensible way to preserve the
1158historical information.
1159
1160
f7b63901 1161Exim version 4.50
495ae4b0
PH
1162-----------------
1163
5fe762f6
PH
1164 1. Minor wording change to the doc/README.SIEVE file.
1165
139059f6 1166 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1167 computation of the current number of files was incorrect.
495ae4b0 1168
7086e875
PH
1169 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1170 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1171 place.
1172
35af9f61
PH
1173 4. Give more explanation in the error message when the command for a transport
1174 filter fails to execute.
1175
b668c215
PH
1176 5. There are several places where Exim runs a non-Exim command in a
1177 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1178 was being done only for the command run by the queryprogram router. It is
1179 now done for all such subprocesses. The other cases are: ${run, transport
1180 filters, and the commands run by the lmtp and pipe transports.
1181
a494b1e1
PH
1182 6. Added CONFIGURE_GROUP build-time option.
1183
1184 7. Some older OS have a limit of 256 on the maximum number of file
1185 descriptors. Exim was using setrlimit() to set 1000 as a large value
1186 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1187 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1188
c5fcb476
PH
1189 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1190 was an oversight, and furthermore, ever since the addition of extra
1191 controls (e.g. 4.43/32), the checks on when to allow different forms of
1192 "control" were broken. There should now be diagnostics for all cases when a
1193 control that does not make sense is encountered.
1194
69358f02
PH
1195 9. Added the /retain_sender option to "control=submission".
1196
5be20824
PH
119710. $recipients is now available in the predata ACL (oversight).
1198
eb2c0248
PH
119911. Tidy the search cache before the fork to do a delivery from a message
1200 received from the command line. Otherwise the child will trigger a lookup
1201 failure and thereby defer the delivery if it tries to use (for example) a
1202 cached ldap connection that the parent has called unbind on.
1203
2a3eea10
PH
120412. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1205 of $address_data from the recipient verification was clobbered by the
1206 sender verification.
1207
120813. The value of address_data from a sender verification is now available in
1209 $sender_address_data in subsequent conditions in the ACL statement.
1210
23c7ff99
PH
121114. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1212
4deaf07d
PH
121315. Added a new option "connect=<time>" to callout options, to set a different
1214 connection timeout.
1215
926e1192
PH
121616. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1217 was its contents. (It was OK if the option was not defined at all.)
1218
650edc6f
PH
121917. A "Completed" log line is now written for messages that are removed from
1220 the spool by the -Mrm option.
1221
2c7db3f5
PH
122218. New variables $sender_verify_failure and $recipient_verify_failure contain
1223 information about exactly what failed.
1224
3d235903
PH
122519. Added -dd to debug only the daemon process.
1226
7c7ad977
PH
122720. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1228 handles timeouts, both on the server side and network timeouts. Renamed the
1229 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1230
981756db
PH
123121. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1232 It is now set to "smtps".
1233
d4eb88df
PH
123422. $host_address is now set to the target address during the checking of
1235 ignore_target_hosts.
1236
123723. When checking ignore_target_hosts for an ipliteral router, no host name was
1238 being passed; this would have caused $sender_host_name to have been used if
1239 matching the list had actually called for a host name (not very likely,
1240 since this list is usually IP addresses). A host name is now passed as
1241 "[x.x.x.x]".
1242
7d468ab8
PH
124324. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1244 code that specifies a non-restarting handler (typically sigaction() in
1245 modern systems) in an attempt to fix a rare and obscure crash bug.
1246
124725. Narrowed the window for a race in the daemon that could cause it to ignore
1248 SIGCHLD signals. This is not a major problem, because they are used only to
1249 wake it up if nothing else does.
1250
62c0818f
PH
125126. A malformed maildirsize file could cause Exim to calculate negative values
1252 for the mailbox size or file count. Odd effects could occur as a result.
1253 The maildirsize information is now recalculated if the size or filecount
1254 end up negative.
1255
26034054
PH
125627. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1257 support for a long time. Removed HAVE_SYS_VFS_H.
1258
af66f652
PH
125928. Installed the latest version of exipick from John Jetmore.
1260
90af77f4
PH
126129. In an address list, if the pattern was not a regular expression, an empty
1262 subject address (from a bounce message) matched only if the pattern was an
1263 empty string. Non-empty patterns were not even tested. This was the wrong
1264 because it is perfectly reasonable to use an empty address as part of a
1265 database query. An empty address is now tested by patterns that are
1266 lookups. However, all the other forms of pattern expect the subject to
1267 contain a local part and a domain, and therefore, for them, an empty
1268 address still always fails if the pattern is not itself empty.
1269
d8ef3577
PH
127030. Exim went into a mad DNS loop when attempting to do a callout where the
1271 host was specified on an smtp transport, and looking it up yielded more
1272 than one IP address.
1273
5cb8cbc6
PH
127431. Re-factored the code for checking spool and log partition space into a
1275 function that finds that data and another that does the check. The former
1276 is then used to implement four new variables: $spool_space, $log_space,
1277 $spool_inodes, and $log_inodes.
1278
14702f5b
PH
127932. The RFC2047 encoding function was originally intended for short strings
1280 such as real names; it was not keeping to the 75-character limit for
1281 encoded words that the RFC imposes. It now respects the limit, and
1282 generates multiple encoded words if necessary. To be on the safe side, I
1283 have increased the buffer size for the ${rfc2047: expansion operator from
1284 1024 to 2048 bytes.
1285
063b1e99
PH
128633. It is now permitted to omit both strings after an "if" condition; if the
1287 condition is true, the result is "true". As before, when the second string
1288 is omitted, a false condition yields an empty string. This makes it less
1289 cumbersome to write custom ACL and router conditions.
1290
652e1b65
PH
129134. Failure to deliver a bounce message always caused it to be frozen, even if
1292 there was an errors_to setting on the router. The errors_to setting is now
1293 respected.
1294
6f0c9a4f
PH
129535. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1296 canonical form (fully expanded) before being placed in
1297 $sender_host_address.
1298
33397d19
PH
129936. The table in the code that translates DNS record types into text (T_A to
1300 "A" for instance) was missing entries for NS and CNAME. It is just possible
1301 that this could have caused confusion if both these types were looked up
1302 for the same domain, because the text type is used as part of Exim's
1303 per-process caching. But the chance of anyone hitting this buglet seems
1304 very small.
1305
7bb56e1f
PH
130637. The dnsdb lookup has been extended in a number of ways.
1307
1308 (1) There is a new type, "zns", which walks up the domain tree until it
1309 finds some nameserver records. It should be used with care.
1310
ea3bc19b
PH
1311 (2) There is a new type, "mxh", which is like "mx" except that it returns
1312 just the host names, not the priorities.
1313
1314 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1315 looked up. The behaviour when one of the lookups defers can be
1316 controlled by a keyword.
7bb56e1f 1317
ea3bc19b 1318 (4) It is now possible to specify the separator character for use when
7bb56e1f 1319 multiple records are returned.
33397d19 1320
0bcb2a0e
PH
132138. The dnslists ACL condition has been extended: it is now possible to supply
1322 a list of IP addresses and/or domains to be looked up in a particular DNS
1323 domain.
1324
2ac0e484
PH
132539. Added log_selector=+queue_time_overall.
1326
4e1fde53
PH
132740. When running the queue in the test harness, wait just a tad after forking a
1328 delivery process, to get repeatability of debugging output.
1329
de365ded
PH
133041. Include certificate and key file names in error message when GnuTLS fails
1331 to set them up, because the GnuTLS error message doesn't include the name
1332 of the failing file when there is a problem reading it.
1333
f05da2e8
PH
133442. Allow both -bf and -bF in the same test run.
1335
d6453af2
PH
133643. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1337
f7b63901
PH
133844. The "Exiscan patch" is now merged into the mainline Exim source.
1339
134045. Sometimes the final signoff response after QUIT could fail to get
1341 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1342 < 0 before doing a fflush(). This bug looks as though it goes back to the
1343 introduction of TLS in release 3.20, but "sometimes" must have been rare
1344 because the tests only now provoked it.
1345
a444213a
PH
134646. Reset the locale to "C" after calling embedded Perl, in case it was changed
1347 (this can affect the format of dates).
1348
0ec020ea
PH
134947. exim_tidydb, when checking for the continued existence of a message for
1350 which it has found a message-specific retry record, was not finding
1351 messages that were in split spool directories. Consequently, it was
1352 deleting retry records that should have stayed in existence.
1353
b1206957
PH
135448. Steve fixed some bugs in eximstats.
1355
135649. The SPA authentication driver was not abandoning authentication and moving
1357 on to the next authenticator when an expansion was forced to fail,
1358 contradicting the general specification for all authenticators. Instead it
1359 was generating a temporary error. It now behaves as specified.
1360
26dd5a95
PH
136150. The default ordering of permitted cipher suites for GnuTLS was pessimal
1362 (the order specifies the preference for clients). The order is now AES256,
1363 AES128, 3DES, ARCFOUR128.
1364
343b2385
PH
136551. Small patch to Sieve code - explicitly set From: when generating an
1366 autoreply.
1367
1c5466b9
PH
136852. Exim crashed if a remote delivery caused a very long error message to be
1369 recorded - for instance if somebody sent an entire SpamAssassin report back
1370 as a large number of 550 error lines. This bug was coincidentally fixed by
1371 increasing the size of one of Exim's internal buffers (big_buffer) that
1372 happened as part of the Exiscan merge. However, to be on the safe side, I
1373 have made the code more robust (and fixed the comments that describe what
1374 is going on).
1375
55ee9ee3
PH
137653. Now that there can be additional text after "Completed" in log lines (if
1377 the queue_time_overall log selector is set), a one-byte patch to exigrep
1378 was needed to allow it to recognize "Completed" as not the last thing in
1379 the line.
1380
d38f8232
PH
138154. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1382 patch that reportedly fixes this has been added. I am not expert enough to
1383 create a test for it. This is what the patch creator wrote:
1384
1385 "I found a little strange behaviour of ldap code when working with
1386 Windows 2003 AD Domain, where users was placed in more than one
1387 Organization Units. When I tried to give exim partial DN, the exit code
1388 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1389 But simultaneously result of request was absolutely normal ldap result,
1390 so I produce this patch..."
1391
3295e65b
PH
1392 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1393 REFERENCE, so I have modified the code to exclude the patch when that macro
1394 is not defined.
1395
7102e136
PH
139655. Some experimental protocols are using DNS PTR records for new purposes. The
1397 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1398 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1399 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1400 CAN-2005-0021
7102e136 1401
3ca0ba97
PH
140256. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1403
c2bcbe20
PH
140457. Double the size of the debug message buffer (to 2048) so that more of very
1405 long debug lines gets shown.
1406
18ce445d
PH
140758. The exicyclog utility now does better if the number of log files to keep
1408 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1409
1f5b4c3d
PH
141059. Two changes related to the smtp_active_hostname option:
1411
1412 (1) $smtp_active_hostname is now available as a variable.
1413 (2) The default for smtp_banner uses $smtp_active_hostname instead
1414 of $primary_hostname.
1415
b975ba52
PH
141660. The host_aton() function is supposed to be passed a string that is known
1417 to be a valid IP address. However, in the case of IPv6 addresses, it was
1418 not checking this. This is a hostage to fortune. Exim now panics and dies
1419 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1420 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1421 components; fortuitously, this particular loophole had already been fixed
1422 by change 4.50/55 above.
1423
1424 If there are any other similar loopholes, the new check in host_aton()
1425 itself should stop them being exploited. The report I received stated that
1426 data on the command line could provoke the exploit when Exim was running as
1427 exim, but did not say which command line option was involved. All I could
1428 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1429 running as the user.
ea3a6f44 1430 CAN-2005-0021
85b87bc2
PH
1431
143261. There was a buffer overflow vulnerability in the SPA authentication code
1433 (which came originally from the Samba project). I have added a test to the
1434 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1435 CAN-2005-0022
b975ba52 1436
17ffcae7
PH
143762. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1438 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1439
d95f9fdb
PH
144063. The daemon start-up calls getloadavg() while still root for those OS that
1441 need the first call to be done as root, but it missed one case: when
1442 deliver_queue_load_max is set with deliver_drop_privilege. This is
1443 necessary for the benefit of the queue runner, because there is no re-exec
1444 when deliver_drop_privilege is set.
1445
86b8287f
PH
144664. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1447 This has been fixed.
1448
60dc5e56
PH
144965. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1450 was in use, was not putting the data itself into the right store pool;
1451 consequently, it could be overwritten for a subsequent message in the same
1452 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1453 the caching.)
1454
533244af
PH
145566. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1456
a5a28604
PH
145767. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1458 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1459 the function were treating the return as a boolean value, which happened to
1460 work because 0=false and not-0=true, but is not correct code.
1461
7e634d24
PH
146268. The host_aton() function was not handling scoped IPv6 addresses (those
1463 with, for example, "%eth0" on the end) correctly.
1464
3e11c26b
PH
146569. Fixed some compiler warnings in acl.c for the bitmaps specified with
1466 negated items (that is, ~something) in unsigned ints. Some compilers
1467 apparently mutter when there is no cast.
1468
6729cf78
PH
146970. If an address verification called from an ACL failed, and did not produce a
1470 user-specific message (i.e. there was only a "system" message), nothing was
1471 put in $acl_verify_message. In this situation, it now puts the system
1472 message there.
1473
00f00ca5
PH
147471. Change 4.23/11 added synchronization checking at the start of an SMTP
1475 session; change 4.31/43 added the unwanted input to the log line - except
1476 that it did not do this in the start of session case. It now does.
1477
c9bdd01c
PH
147872. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1479 This is wrong and can cause the other end to generate a synchronization
1480 error if it is another Exim or anything else that does the synchronization
1481 check. A QUIT command is no longer sent after a timeout.
1482
d43194df
PH
148373. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1484 during host lookups.
1485
fe5b5d0b
PH
148674. The defer_ok option of callout verification was not working if it was used
1487 when verifying addresses in header lines, that is, for this case:
1488
1489 verify = header_sender/callout=defer_ok
1490
76a2d7ba
PH
149175. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1492 those file descriptors could be used for SMTP connections. If anything
1493 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1494 could be sent to the SMTP client, causing chaos. The daemon now opens
1495 stdin, stdout, and stderr to /dev/null when it puts itself into the
1496 background.
1497
149876. Arrange for output from Perl's "warn" command to be written to Exim's main
1499 log by default. The user can override this with suitable Perl magic.
1500
04f7d5b9
PH
150177. The use of log_message on a "discard" ACL verb, which is supposed to add to
1502 the log message when discard triggers, was not working for the DATA ACL or
1503 for the non-SMTP ACL.
1504
bc60667e
PH
150578. Error message wording change in sieve.c.
1506
bb6e88ff
PH
150779. If smtp_accept_max_per_host was set, the number of connections could be
1508 restricted to fewer than expected, because the daemon was trying to set up
1509 a new connection before checking whether the processes handling previous
1510 connections had finished. The check for completed processes is now done
1511 earlier. On busy systems, this bug wouldn't be noticed because something
1512 else would have woken the daemon, and it would have reaped the completed
1513 process earlier.
1514
1e70f85b
PH
151580. If a message was submitted locally by a user whose login name contained one
1516 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1517 It caused a spool format error. I have fixed the spool reading code. A
1518 related problem was that the "from" clause in the Received: line became
1519 illegal because of the space(s). It is now covered by ${quote_local_part.
1520
152181. Included the latest eximstats from Steve (adds average sizes to HTML Top
1522 tables).
1523
4e01f9d6
PH
152482. Updated OS/Makefile-AIX as per message from Mike Meredith.
1525
1ee1cef2
PH
152683. Patch from Sieve maintainer to fix unterminated string problem in
1527 "vacation" handling.
1528
6e2b4ccc
PH
152984. Some minor changes to the Linux configuration files to help with other
1530 OS variants using glibc.
1531
8e669ac1
PH
153285. One more patch for Sieve to update vacation handling to latest spec.
1533
495ae4b0 1534
bbe902f0
PH
1535----------------------------------------------------
1536See the note above about the 4.44 and 4.50 releases.
1537----------------------------------------------------
1538
1539
1540Exim version 4.44
1541-----------------
1542
1543 1. Change 4.43/35 introduced a bug that caused file counts to be
1544 incorrectly computed when quota_filecount was set in an appendfile
1545 transport
1546
1547 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1548 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1549 place.
1550
1551 3. Give more explanation in the error message when the command for a transport
1552 filter fails to execute.
1553
1554 4. There are several places where Exim runs a non-Exim command in a
1555 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1556 was being done only for the command run by the queryprogram router. It is
1557 now done for all such subprocesses. The other cases are: ${run, transport
1558 filters, and the commands run by the lmtp and pipe transports.
1559
1560 5. Some older OS have a limit of 256 on the maximum number of file
1561 descriptors. Exim was using setrlimit() to set 1000 as a large value
1562 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1563 systems. I've change it so that if it can't get 1000, it tries for 256.
1564
1565 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1566 was an oversight, and furthermore, ever since the addition of extra
1567 controls (e.g. 4.43/32), the checks on when to allow different forms of
1568 "control" were broken. There should now be diagnostics for all cases when a
1569 control that does not make sense is encountered.
1570
1571 7. $recipients is now available in the predata ACL (oversight).
1572
1573 8. Tidy the search cache before the fork to do a delivery from a message
1574 received from the command line. Otherwise the child will trigger a lookup
1575 failure and thereby defer the delivery if it tries to use (for example) a
1576 cached ldap connection that the parent has called unbind on.
1577
1578 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1579 of $address_data from the recipient verification was clobbered by the
1580 sender verification.
1581
158210. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1583 was its contents. (It was OK if the option was not defined at all.)
1584
158511. A "Completed" log line is now written for messages that are removed from
1586 the spool by the -Mrm option.
1587
158812. $host_address is now set to the target address during the checking of
1589 ignore_target_hosts.
1590
159113. When checking ignore_target_hosts for an ipliteral router, no host name was
1592 being passed; this would have caused $sender_host_name to have been used if
1593 matching the list had actually called for a host name (not very likely,
1594 since this list is usually IP addresses). A host name is now passed as
1595 "[x.x.x.x]".
1596
159714. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1598 code that specifies a non-restarting handler (typically sigaction() in
1599 modern systems) in an attempt to fix a rare and obscure crash bug.
1600
160115. Narrowed the window for a race in the daemon that could cause it to ignore
1602 SIGCHLD signals. This is not a major problem, because they are used only to
1603 wake it up if nothing else does.
1604
160516. A malformed maildirsize file could cause Exim to calculate negative values
1606 for the mailbox size or file count. Odd effects could occur as a result.
1607 The maildirsize information is now recalculated if the size or filecount
1608 end up negative.
1609
161017. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1611 support for a long time. Removed HAVE_SYS_VFS_H.
1612
ea3a6f44 161318. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1614
161519. Allow an empty sender to be matched against a lookup in an address list.
1616 Previously the only cases considered were a regular expression, or an
1617 empty pattern.
1618
161920. Exim went into a mad DNS lookup loop when doing a callout where the
1620 host was specified on the transport, if the DNS lookup yielded more than
1621 one IP address.
1622
ea3a6f44
NM
162321. The RFC2047 encoding function was originally intended for short strings
1624 such as real names; it was not keeping to the 75-character limit for
1625 encoded words that the RFC imposes. It now respects the limit, and
1626 generates multiple encoded words if necessary. To be on the safe side, I
1627 have increased the buffer size for the ${rfc2047: expansion operator from
1628 1024 to 2048 bytes.
bbe902f0 1629
ea3a6f44
NM
163022. Failure to deliver a bounce message always caused it to be frozen, even if
1631 there was an errors_to setting on the router. The errors_to setting is now
1632 respected.
bbe902f0
PH
1633
163423. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1635 canonical form (fully expanded) before being placed in
1636 $sender_host_address.
1637
163824. Updated eximstats to version 1.33
1639
ea3a6f44
NM
164025. Include certificate and key file names in error message when GnuTLS fails
1641 to set them up, because the GnuTLS error message doesn't include the name
1642 of the failing file when there is a problem reading it.
bbe902f0
PH
1643
164426. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1645 As per change 25.
bbe902f0 1646
ea3a6f44
NM
164727. Reset the locale to "C" after calling embedded Perl, in case it was changed
1648 (this can affect the format of dates).
bbe902f0 1649
ea3a6f44
NM
165028. exim_tidydb, when checking for the continued existence of a message for
1651 which it has found a message-specific retry record, was not finding
1652 messages that were in split spool directories. Consequently, it was
1653 deleting retry records that should have stayed in existence.
bbe902f0
PH
1654
165529. eximstats updated to version 1.35
1656 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1657 1.35 - bugfix such that pie charts by volume are generated correctly
1658
ea3a6f44
NM
165930. The SPA authentication driver was not abandoning authentication and moving
1660 on to the next authenticator when an expansion was forced to fail,
1661 contradicting the general specification for all authenticators. Instead it
1662 was generating a temporary error. It now behaves as specified.
bbe902f0 1663
ea3a6f44
NM
166431. The default ordering of permitted cipher suites for GnuTLS was pessimal
1665 (the order specifies the preference for clients). The order is now AES256,
1666 AES128, 3DES, ARCFOUR128.
bbe902f0 1667
ea3a6f44
NM
166831. Small patch to Sieve code - explicitly set From: when generating an
1669 autoreply.
bbe902f0 1670
ea3a6f44
NM
167132. Exim crashed if a remote delivery caused a very long error message to be
1672 recorded - for instance if somebody sent an entire SpamAssassin report back
1673 as a large number of 550 error lines. This bug was coincidentally fixed by
1674 increasing the size of one of Exim's internal buffers (big_buffer) that
1675 happened as part of the Exiscan merge. However, to be on the safe side, I
1676 have made the code more robust (and fixed the comments that describe what
1677 is going on).
bbe902f0 1678
ea3a6f44
NM
167933. Some experimental protocols are using DNS PTR records for new purposes. The
1680 keys for these records are domain names, not reversed IP addresses. The
1681 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1682 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1683 CAN-2005-0021
1684
ea3a6f44
NM
168534. The host_aton() function is supposed to be passed a string that is known
1686 to be a valid IP address. However, in the case of IPv6 addresses, it was
1687 not checking this. This is a hostage to fortune. Exim now panics and dies
1688 if the condition is not met. A case was found where this could be provoked
1689 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1690 components; fortuitously, this particular loophole had already been fixed
1691 by change 4.50/55 or 4.44/33 above.
1692
1693 If there are any other similar loopholes, the new check in host_aton()
1694 itself should stop them being exploited. The report I received stated that
1695 data on the command line could provoke the exploit when Exim was running as
1696 exim, but did not say which command line option was involved. All I could
1697 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1698 running as the user.
bbe902f0
PH
1699 CAN-2005-0021
1700
ea3a6f44
NM
170135. There was a buffer overflow vulnerability in the SPA authentication code
1702 (which came originally from the Samba project). I have added a test to the
1703 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1704 CAN-2005-0022
1705
ea3a6f44
NM
170636. The daemon start-up calls getloadavg() while still root for those OS that
1707 need the first call to be done as root, but it missed one case: when
1708 deliver_queue_load_max is set with deliver_drop_privilege. This is
1709 necessary for the benefit of the queue runner, because there is no re-exec
1710 when deliver_drop_privilege is set.
bbe902f0 1711
ea3a6f44
NM
171237. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1713 was in use, was not putting the data itself into the right store pool;
1714 consequently, it could be overwritten for a subsequent message in the same
1715 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1716 the caching.)
bbe902f0 1717
ea3a6f44
NM
171838. Sometimes the final signoff response after QUIT could fail to get
1719 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1720 < 0 before doing a fflush(). This bug looks as though it goes back to the
1721 introduction of TLS in release 3.20, but "sometimes" must have been rare
1722 because the tests only now provoked it.
bbe902f0
PH
1723
1724
495ae4b0
PH
1725Exim version 4.43
1726-----------------
1727
1728 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1729 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1730 Now it returns an integer. A number of calls were still expecting a T/F
1731 return. Fortuitously, in all cases, the tests worked in OK situations,
1732 which is the norm. However, things would have gone wrong on any write
1733 failures on the smtp file descriptor. This function is used when sending
1734 messages over SMTP and also when doing verify callouts.
1735
1736 2. When Exim is called to do synchronous delivery of a locally submitted
1737 message (the -odf or -odi options), it no longer closes stderr before doing
1738 the delivery.
1739
1740 3. Implemented the mua_wrapper option.
1741
1742 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1743
1744 5. Implemented the functions header_remove(), header_testname(),
1745 header_add_at_position(), and receive_remove_recipient(), and exported them
1746 to local_scan().
1747
1748 6. If an ACL "warn" statement specified the addition of headers, Exim already
1749 inserted X-ACL-Warn: at the start if there was no header name. However, it
1750 was not making this test for the second and subsequent header lines if
1751 there were newlines in the string. This meant that an invalid header could
1752 be inserted if Exim was badly configured.
1753
1754 7. Allow an ACL "warn" statement to add header lines at the start or after all
1755 the Received: headers, as well as at the end.
1756
1757 8. Added the rcpt_4xx retry error code.
1758
1759 9. Added postmaster_mailfrom=xxx to callout verification option.
1760
176110. Added mailfrom=xxxx to the callout verification option, for verify=
1762 header_sender only.
1763
176411. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1765 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1766
176712. Inserted some casts to stop certain compilers warning when using pointer
1768 differences as field lengths or precisions in printf-type calls (mostly
1769 affecting debugging statements).
1770
177113. Added optional readline() support for -be (dynamically loaded).
1772
177314. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1774 same clock tick as a message's arrival, so that its received time was the
1775 same as the "first fail" time on the retry record, and that message
1776 remained on the queue past the ultimate address timeout, every queue runner
1777 would try a delivery (because it was past the ultimate address timeout) but
1778 after another failure, the ultimate address timeout, which should have then
1779 bounced the address, did not kick in. This was a "< instead of <=" error;
1780 in most cases the first failure would have been in the next clock tick
1781 after the received time, and all would be well.
1782
178315. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1784 being recognized when the domain list was tested by the match_domain
1785 condition in an expansion string.
1786
178716. Added the ${str2b64: operator.
1788
178917. Exim was always calling setrlimit() to set a large limit for the number of
1790 processes, without checking whether the existing limit was already
1791 adequate. (It did check for the limit on file descriptors.) Furthermore,
1792 errors from getrlimit() and setrlimit() were being ignored. Now they are
1793 logged to the main and panic logs, but Exim does carry on, to try to do its
1794 job under whatever limits there are.
1795
179618. Imported PCRE 5.0.
1797
179819. Trivial typo in log message " temporarily refused connection" (the leading
1799 space).
1800
180120. If the log selector return_path_on_delivery was set and an address was
1802 redirected to /dev/null, the delivery process crashed because it assumed
1803 that a return path would always be set for a "successful" delivery. In this
1804 case, the whole delivery is bypassed as an optimization, and therefore no
1805 return path is set.
1806
180721. Internal re-arrangement: the function for sending a challenge and reading
1808 a response while authentication was assuming a zero-terminated challenge
1809 string. It's now changed to take a pointer and a length, to allow for
1810 binary data in such strings.
1811
181222. Added the cyrus_sasl authenticator (code supplied by MBM).
1813
181423. Exim was not respecting finduser_retries when seeking the login of the
1815 uid under which it was called; it was always trying 10 times. (The default
1816 setting of finduser_retries is zero.) Also, it was sleeping after the final
1817 failure, which is pointless.
1818
181924. Implemented tls_on_connect_ports.
1820
182125. Implemented acl_smtp_predata.
1822
182326. If the domain in control=submission is set empty, Exim assumes that the
1824 authenticated id is a complete email address when it generates From: or
1825 Sender: header lines.
1826
182727. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1828 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1829 chown and chgrp in /bin and hostname in /usr/bin.
1830
183128. Exim was keeping the "process log" file open after each use, just as it
1832 does for the main log. This opens the possibility of it remaining open for
1833 long periods when the USR1 signal hits a daemon. Occasional processlog
1834 errors were reported, that could have been caused by this. Anyway, it seems
1835 much more sensible not to leave this file open at all, so that is what now
1836 happens.
1837
183829. The long-running daemon process does not normally write to the log once it
1839 has entered its main loop, and it closes the log before doing so. This is
1840 so that log files can straightforwardly be renamed and moved. However,
1841 there are a couple of unusual error situations where the daemon does write
1842 log entries, and I had neglected to close the log afterwards.
1843
184430. The text of an SMTP error response that was received during a remote
1845 delivery was being truncated at 512 bytes. This is too short for some of
1846 the long messages that one sometimes sees. I've increased the limit to
1847 1024.
1848
184931. It is now possible to make retry rules that apply only when a message has a
1850 specific sender, in particular, an empty sender.
1851
185232. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1853 it possible to be selective about when SMTP synchronization is enforced.
1854
185533. Added "control = caseful_local_part" and "control = "caselower_local_part".
1856
185732. Implemented hosts_connection_nolog.
1858
185933. Added an ACL for QUIT.
1860
186134. Setting "delay_warning=" to disable warnings was not working; it gave a
1862 syntax error.
1863
186435. Added mailbox_size and mailbox_filecount to appendfile.
1865
186636. Added control = no_multiline_responses to ACLs.
1867
186837. There was a bug in the logic of the code that waits for the clock to tick
1869 in the case where the clock went backwards by a substantial amount such
1870 that the microsecond fraction of "now" was more than the microsecond
1871 fraction of "then" (but the whole seconds number was less).
1872
187338. Added support for the libradius Radius client library this is found on
1874 FreeBSD (previously only the radiusclient library was supported).
1875
1876
1877Exim version 4.42
1878-----------------
1879
1880 1. When certain lookups returned multiple values in the form name=value, the
1881 quoting of the values was not always being done properly. Specifically:
1882 (a) If the value started with a double quote, but contained no whitespace,
1883 it was not quoted.
1884 (b) If the value contained whitespace other than a space character (i.e.
1885 tabs or newlines or carriage returns) it was not quoted.
1886 This fix has been applied to the mysql and pgsql lookups by writing a
1887 separate quoting function and calling it from the lookup code. The fix
1888 should probably also be applied to nisplus, ibase and oracle lookups, but
1889 since I cannot test any of those, I have not disturbed their existing code.
1890
1891 2. A hit in the callout cache for a specific address caused a log line with no
1892 reason for rejecting RCPT. Now it says "Previous (cached) callout
1893 verification failure".
1894
1895 3. There was an off-by-one bug in the queryprogram router. An over-long
1896 return line was truncated at 256 instead of 255 characters, thereby
1897 overflowing its buffer with the terminating zero. As well as fixing this, I
1898 have increased the buffer size to 1024 (and made a note to document this).
1899
1900 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1901 when Exim is waiting for an SMTP response from a remote server, Exim
1902 restarts its select() call on the socket, thereby resetting its timeout.
1903 This is not a problem when such interrupts are rare. Somebody set up a cron
1904 job to run exiwhat every 2 minutes, which is less than the normal select()
1905 timeout (5 or 10 minutes). This meant that the select() timeout never
1906 kicked in because it was always reset. I have fixed this by comparing the
1907 time when an interrupt arrives with the time at the start of the first call
1908 to select(). If more time than the timeout has elapsed, the interrupt is
1909 treated as a timeout.
1910
1911 5. Some internal re-factoring in preparation for the addition of Sieve
1912 extensions (by MH). In particular, the "personal" test is moved to a
1913 separate function, and given an option for scanning Cc: and Bcc: (which is
1914 not set for Exim filters).
1915
1916 6. When Exim created an email address using the login of the caller as the
1917 local part (e.g. when creating a From: or Sender: header line), it was not
1918 quoting the local part when it contained special characters such as @.
1919
1920 7. Installed new OpenBSD configuration files.
1921
1922 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1923 try to make them clearer.
1924
1925 9. Callout options, other than the timeout value, were being ignored when
1926 verifying sender addresses in header lines. For example, when using
1927
1928 verify = header_sender/callout=no_cache
1929
1930 the cache was (incorrectly) being used.
1931
193210. Added a missing instance of ${EXE} to the exim_install script; this affects
1933 only the Cygwin environment.
1934
193511. When return_path_on_delivery was set as a log selector, if different remote
1936 addresses in the same message used different return paths and parallel
1937 remote delivery occurred, the wrong values would sometimes be logged.
1938 (Whenever a remote delivery process finished, the return path value from
1939 the most recently started remote delivery process was logged.)
1940
194112. RFC 3848 specifies standard names for the "with" phrase in Received: header
1942 lines when AUTH and/or TLS are in use. This is the "received protocol"
1943 field. Exim used to use "asmtp" for authenticated SMTP, without any
1944 indication (in the protocol name) for TLS use. Now it follows the RFC and
1945 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1946 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1947 names appear in log lines as well as in Received: header lines.
1948
194913. Installed MH's patches for Sieve to add the "copy" and "vacation"
1950 extensions, and comparison tests, and to fix some bugs.
1951
195214. Changes to the "personal" filter test:
1953
1954 (1) The test was buggy in that it was just doing the equivalent of
1955 "contains" tests on header lines. For example, if a user's address was
1956 anne@some.where, the "personal" test would incorrectly be true for
1957
1958 To: susanne@some.where
1959
1960 This test is now done by extracting each address from the header in turn,
1961 and checking the entire address. Other tests that are part of "personal"
1962 are now done using regular expressions (for example, to check local parts
1963 of addresses in From: header lines).
1964
1965 (2) The list of non-personal local parts in From: addresses has been
1966 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1967 taken from the Sieve specification recommendations.
1968
1969 (3) If the message contains any header line starting with "List-" it is
1970 treated as non-personal.
1971
1972 (4) The test for "circular" in the Subject: header line has been removed
1973 because it now seems ill-conceived.
1974
197515. Minor typos in src/EDITME comments corrected.
1976
197716. Installed latest exipick from John Jetmore.
1978
197917. If headers_add on a router specified a text string that was too long for
1980 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1981 of string_sprintf() is now avoided.
1982
198318. $message_body_size was not set (it was always zero) when running the DATA
1984 ACL and the local_scan() function.
1985
198619. For the "mail" command in an Exim filter, no default was being set for
1987 the once_repeat time, causing a random time value to be used if "once" was
1988 specified. (If the value happened to be <= 0, no repeat happened.) The
1989 default is now 0s, meaning "never repeat". The "vacation" command was OK
1990 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1991 (I found it when inspecting the code).
1992
199320. There is now an overall timeout for performing a callout verification. It
1994 defaults to 4 times the callout timeout, which applies to individual SMTP
1995 commands during the callout. The overall timeout applies when there is more
1996 than one host that can be tried. The timeout is checked before trying the
1997 next host. This prevents very long delays if there are a large number of
1998 hosts and all are timing out (e.g. when the network connections are timing
1999 out). The value of the overall timeout can be changed by specifying an
2000 additional sub-option for "callout", called "maxwait". For example:
2001
2002 verify = sender/callout=5s,maxwait=20s
2003
200421. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2005 to the end before writing, but this should make it even safer).
2006
200722. Exim was forgetting that it had advertised PIPELINING for the second and
2008 subsequent messages on an SMTP connection. It was also not resetting its
2009 memory on STARTTLS and an internal HELO.
2010
201123. When Exim logs an SMTP synchronization error within a session, it now
2012 records whether PIPELINING has been advertised or not.
2013
201424. Added 3 instances of "(long int)" casts to time_t variables that were being
2015 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2016 rather than long int.
2017
201825. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2019
202026. Added the never_mail option to autoreply.
2021
2022
2023Exim version 4.41
2024-----------------
2025
2026 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2027 crash if the getsockname() call failed; this can happen if a connection is
2028 closed very soon after it is established. The problem was simply in the
2029 order in which certain operations were done, causing Exim to try to write
2030 to the SMTP stream before it had set up the file descriptor. The bug has
2031 been fixed by making things happen in the correct order.
2032
2033
2034Exim version 4.40
2035-----------------
2036
2037 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2038 before the connection was closed, thus losing the rejection response.
2039
2040 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2041 some early Solaris releases, but causes trouble in current releases where
2042 socklen_t is defined.
2043
2044 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2045 always exist.
2046
2047 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2048 configured.
2049
2050 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2051 that releasing the top of it at the end releases what was used for sub-
2052 expansions (unless the block got too big). However, discard this block if
2053 the first thing is a variable or header, so that we can use its block when
2054 it is dynamic (useful for very large $message_headers, for example).
2055
2056 6. Lookups now cache *every* query, not just the most recent. A new, separate
2057 store pool is used for this. It can be recovered when all lookup caches are
2058 flushed. Lookups now release memory at the end of their result strings.
2059 This has involved some general refactoring of the lookup sources.
2060
2061 7. Some code has been added to the store_xxx() functions to reduce the amount
2062 of flapping under certain conditions.
2063
2064 8. log_incoming_interface used to affect only the <= reception log lines. Now
2065 it causes the local interface and port to be added to several more SMTP log
2066 lines, for example "SMTP connection from", and rejection lines.
2067
2068 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2069
207010. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2071
207211. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2073 could be overwritten at the end of the current message (or the start of a
2074 new message if it was set in a HELO ACL). The value is now preserved for
2075 the duration of the SMTP connection.
2076
207712. If a transport had a headers_rewrite setting, and a matching header line
2078 contained an unqualified address, that address was qualified, even if it
2079 did not match any rewriting rules. The underlying bug was that the values
2080 of the flags that permit the existence of unqualified sender and recipient
2081 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2082 non-local messages, and by -bnq for local messages) were not being
2083 preserved with the message after it was received.
2084
208513. When Exim was logging an SMTP synchronization error, it could sometimes log
2086 "next input=" as part of the text comprising the host identity instead of
2087 the correct text. The code was using the same buffer for two different
2088 strings. However, depending on which order the printing function evaluated
2089 its arguments, the bug did not always show up. Under Linux, for example, my
2090 test suite worked just fine.
2091
209214. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2093 doesn't work with some older versions of Perl. It has been changed to "my",
2094 which in any case is probably the better facility to use.
2095
209615. A really picky compiler found some instances of statements for creating
2097 error messages that either had too many or two few arguments for the format
2098 string.
2099
210016. The size of the buffer for calls to the DNS resolver has been increased
2101 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2102 for addresses that have a lot of PTR records. This alleviates a problem; it
2103 does not fully solve it.
2104
210517. A dnsdb lookup for PTR records that receives more data than will fit in the
2106 buffer now truncates the list and logs the incident, which is the same
2107 action as happens when Exim is looking up a host name and its aliases.
2108 Previously in this situation something unpredictable would happen;
2109 sometimes it was "internal error: store_reset failed".
2110
211118. If a server dropped the connection unexpectedly when an Exim client was
2112 using GnuTLS and trying to read a response, the client delivery process
2113 crashed while trying to generate an error log message.
2114
211519. If a "warn" verb in an ACL added multiple headers to a message in a single
2116 string, for example:
2117
2118 warn message = H1: something\nH2: something
2119
2120 the text was added as a single header line from Exim's point of view
2121 though it ended up OK in the delivered message. However, searching for the
2122 second and subsequent header lines using $h_h2: did not work. This has been
2123 fixed. Similarly, if a system filter added multiple headers in this way,
2124 the routers could not see them.
2125
212620. Expanded the error message when iplsearch is called with an invalid key to
2127 suggest using net-iplsearch in a host list.
2128
212921. When running tests using -bh, any delays imposed by "delay" modifiers in
2130 ACLs are no longer actually imposed (and a message to that effect is
2131 output).
2132
213322. If a "gecos" field in a passwd entry contained escaped characters, in
2134 particular, if it contained a \" sequence, Exim got it wrong when building
2135 a From: or a Sender: header from that name. A second bug also caused
2136 incorrect handling when an unquoted " was present following a character
2137 that needed quoting.
2138
213923. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2140 was not being matched caselessly.
2141
214224. Arranged for all hyphens in the exim.8 source to be escaped with
2143 backslashes.
2144
214525. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2146 properly. Recipient callout cache records were still being keyed to include
2147 the sender, even when use_sender was set false. This led to far more
2148 callouts that were necessary. The sender is no longer included in the key
2149 when use_sender is false.
2150
215126. Added "control = submission" modifier to ACLs.
2152
215327. Added the ${base62d: operator to decode base 62 numbers.
2154
215528. dnsdb lookups can now access SRV records.
2156
215729. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2158 the configuration file.
2159
216030. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2161 (-v) mode. This makes the output for a verbose queue run more intelligible.
2162
216331. Added a use_postmaster feature to recipient callouts.
2164
216532. Added the $body_zerocount variable, containing the number of binary zero
2166 bytes in the message body.
2167
216833. The time of last modification of the "new" subdirectory is now used as the
2169 "mailbox time last read" when there is a quota error for a maildir
2170 delivery.
2171
217234. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2173
217435. Added +ignore_unknown as a special item in host lists.
2175
217636. Code for decoding IPv6 addresses in host lists is now included, even if
2177 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2178 address was recognized as an IP address, but was then not correctly decoded
2179 into binary, causing unexpected and incorrect effects when compared with
2180 another IP address.
2181
2182
2183Exim version 4.34
2184-----------------
2185
2186 1. Very minor rewording of debugging text in manualroute to say "list of
2187 hosts" instead of "hostlist".
2188
2189 2. If verify=header_syntax was set, and a header line with an unqualified
2190 address (no domain) and a large number of spaces between the end of the
2191 name and the colon was received, the reception process suffered a buffer
2192 overflow, and (when I tested it) crashed. This was caused by some obsolete
2193 code that should have been removed. The fix is to remove it!
2194
2195 3. When running in the test harness, delay a bit after writing a bounce
2196 message to get a bit more predictability in the log output.
2197
2198 4. Added a call to search_tidyup() just before forking a reception process. In
2199 theory, someone could use a lookup in the expansion of smtp_accept_max_
2200 per_host which, without the tidyup, could leave open a database connection.
2201
2202 5. Added the variables $recipient_data and $sender_data which get set from a
2203 lookup success in an ACL "recipients" or "senders" condition, or a router
2204 "senders" option, similar to $domain_data and $local_part_data.
2205
2206 6. Moved the writing of debug_print from before to after the "senders" test
2207 for routers.
2208
2209 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2210 problems for message scanning, either using a data ACL, or using
2211 local_scan() because the Received: header was not generated till after they
2212 were called (in order to set the time as the time of reception completion).
2213 I have revised the way this works. The header is now generated after the
2214 body is received, but before the ACL or local_scan() are called. After they
2215 are run, the timestamp in the header is updated.
2216
2217
2218Exim version 4.33
2219-----------------
2220
2221 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2222 before starting a queue runner without re-exec. This happened only when
2223 deliver_drop_privilege was set or when the Exim user was set to root. The
2224 effect of the bug was that timeouts during subsequent deliveries caused
2225 crashes instead of being properly handled. The handler is now left at its
2226 default (and expected) setting.
2227
2228 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2229 message, again when deliver_drop_privilege is set or Exim is run as root.
2230 The bug described in (1) was not present in this case, but the tidying up
2231 of the other signals was missing. I have made the two cases consistent.
2232
2233 3. The ignore_target_hosts setting on a manualroute router was being ignored
2234 for hosts that were looked up using the /MX notation.
2235
2236 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2237 in domain lists.
2238
2239 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2240 operated on the sender address. After changing the $sender_address to <>
2241 for the sender address verify, Exim was re-instated it as the original
2242 (before rewriting) address, but remembering that it had rewritten it, so it
2243 wasn't rewriting it again. This bug also had the effect of breaking the
2244 sender address verification caching when the sender address was rewritten.
2245
2246 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2247 This has been changed so that if the ip literal address matches
2248 ignore_target_hosts, the router declines.
2249
2250 7. Added expansion conditions match_domain, match_address, and match_local_
2251 part (NOT match_host).
2252
2253 8. The placeholder for the Received: header didn't have a length field set.
2254
2255 9. Added code to Exim itself and to exim_lock to test for a specific race
2256 condition that could lead to file corruption when using MBX delivery. The
2257 issue is with the lockfile that is created in /tmp. If this file is removed
2258 after a process has opened it but before that process has acquired a lock,
2259 there is the potential for a second process to recreate the file and also
2260 acquire a lock. This could lead to two Exim processes writing to the file
2261 at the same time. The added code performs the same test as UW imapd; it
2262 checks after acquiring the lock that its file descriptor still refers to
2263 the same named file.
2264
226510. The buffer for building added header lines was of fixed size, 8192 bytes.
2266 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2267 when Exim is built.
2268
226911. Added the smtp_active_hostname option. If used, this will typically be made
2270 to depend on the incoming interface address. Because $interface_address is
2271 not set up until the daemon has forked a reception process, error responses
2272 that can happen earlier (such as "too many connections") no longer contain
2273 a host name.
2274
227512. If an expansion in a condition on a "warn" statement fails because a lookup
2276 defers, the "warn" statement is abandoned, and the next ACL statement is
2277 processed. Previously this caused the whole ACL to be aborted.
2278
227913. Added the iplsearch lookup type.
2280
228114. Added ident_timeout as a log selector.
2282
228315. Added tls_certificate_verified as a log selector.
2284
228516. Added a global option tls_require_ciphers (compare the smtp transport
2286 option of the same name). This controls incoming TLS connections.
2287
228817. I finally figured out how to make tls_require_ciphers do a similar thing
2289 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2290 before starting the TLS session.
2291
229218. Tabs are now shown as \t in -bP output.
2293
229419. If the log selector return_path_on_delivery was set, Exim crashed when
2295 bouncing a message because it had too many Received: header lines.
2296
229720. If two routers both had headers_remove settings, and the first one included
2298 a superfluous trailing colon, the final name in the first list and the
2299 first name in the second list were incorrectly joined into one item (with a
2300 colon in the middle).
2301
2302
2303Exim version 4.32
2304-----------------
2305
2306 1. Added -C and -D options to the exinext utility, mainly to make it easier
2307 to include in the automated testing, but these could be helpful when
2308 multiple configurations are in use.
2309
2310 2. The exinext utility was not formatting the output nicely when there was
2311 an alternate port involved in the retry record key, nor when there was a
2312 message id as well (for retries that were specific to a specific message
2313 and a specific host). It was also confused by IPv6 addresses, because of
2314 the additional colons they contain. I have fixed the IPv4 problem, and
2315 patched it up to do a reasonable job for IPv6.
2316
2317 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2318 delivery, the log line now contains "pipelined" if PIPELINING was used.
2319
2320 4. An SMTP transport process used to panic and die if the bind() call to set
2321 an explicit outgoing interface failed. This has been changed; it is now
2322 treated in the same way as a connect() failure.
2323
2324 5. A reference to $sender_host_name in the part of a conditional expansion
2325 that was being skipped was still causing a DNS lookup. This no longer
2326 occurs.
2327
2328 6. The def: expansion condition was not recognizing references to header lines
2329 that used bh_ and bheader_.
2330
2331 7. Added the _cache feature to named lists.
2332
2333 8. The code for checking quota_filecount in the appendfile transport was
2334 allowing one more file than it should have been.
2335
2336 9. For compatibility with Sendmail, the command line option
2337
2338 -prval:sval
2339
2340 is equivalent to
2341
2342 -oMr rval -oMs sval
2343
2344 and sets the incoming protocol and host name (for trusted callers). The
2345 host name and its colon can be omitted when only the protocol is to be set.
2346 Note the Exim already has two private options, -pd and -ps, that refer to
2347 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2348 "s", but I don't think that's a major issue.
2349
235010. A number of refactoring changes to the code, none of which should affect
2351 Exim's behaviour:
2352
2353 (a) The number of logging options was getting close to filling up the
2354 32-bit word that was used as a bit map. I have split them into two classes:
2355 those that are passed in the argument to log_write(), and those that are
2356 only ever tested independently outside of that function. These are now in
2357 separate 32-bit words, so there is plenty of room for expansion again.
2358 There is no change in the user interface or the logging behaviour.
2359
2360 (b) When building, for example, log lines, the code previously used a
2361 macro that called string_cat() twice, in order to add two strings. This is
2362 not really sufficiently general. Furthermore, there was one instance where
2363 it was actually wrong because one of the argument was used twice, and in
2364 one call a function was used. (As it happened, calling the function twice
2365 did not affect the overall behaviour.) The macro has been replaced by a
2366 function that can join an arbitrary number of extra strings onto a growing
2367 string.
2368
2369 (c) The code for expansion conditions now uses a table and a binary chop
2370 instead of a serial search (which was left over from when there were very
2371 few conditions). Also, it now recognizes conditions like "pam" even when
2372 the relevant support is not compiled in: a suitably worded error message is
2373 given if an attempt is made to use such a condition.
2374
237511. Added ${time_interval:xxxxx}.
2376
237712. A bug was causing one of the ddress fields not to be passed back correctly
2378 from remote delivery subprocesses. The field in question was not being
2379 subsequently used, so this caused to problems in practice.
2380
238113. Added new log selectors queue_time and deliver_time.
2382
238314. Might have fixed a bug in maildirsizefile handling that threw up
2384 "unexpected character" debug warnings, and recalculated the data
2385 unnecessarily. In any case, I expanded the warning message to give more
2386 information.
2387
238815. Added the message "Restricted characters in address" to the statements in
2389 the default ACL that block characters like @ and % in local parts.
2390
239116. Change 71 for release 4.31 proved to be much less benign that I imagined.
2392 Three changes have been made:
2393
2394 (a) There was a serious bug; a negative response to MAIL caused the whole
2395 recipient domain to be cached as invalid, thereby blocking all messages
2396 to all local parts at the same domain, from all senders. This bug has
2397 been fixed. The domain is no longer cached after a negative response to
2398 MAIL if the sender used is not empty.
2399
2400 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2401 been restored.
2402
2403 (c) A new callout option, "use_sender" has been added for people who want
2404 the modified behaviour.
2405
2406
2407Exim version 4.31
2408-----------------
2409
2410 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2411 Larry Rosenman.
2412
2413 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2414 indeed breaks things for older releases.
2415
2416 3. Added additional logging to the case where there is a problem reading data
2417 from a filter that is running in a subprocess using a pipe, in order to
2418 try to track down a specific problem.
2419
2420 4. Testing facility fudge: when running in the test harness and attempting
2421 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2422 getting "No route to host". Convert this to a timeout.
2423
2424 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2425 warning.
2426
2427 6. Some OS don't have socklen_t but use size_t instead. This affects the
2428 fifth argument of getsockopt() amongst other things. This is now
2429 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2430 can be set for individual OS. I have set it for SunOS5, OSF1, and
2431 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2432 some earlier ones do not.
2433
2434 7. Change 4.30/15 was not doing the test caselessly.
2435
2436 8. The standard form for an IPv6 address literal was being rejected by address
2437 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2438 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2439 this, as well as the form without the "IPv6" on the front (but only when
2440 address literals are enabled, of course).
2441
2442 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2443
244410. Exim crashed if a message with an empty sender address specified by -f
2445 encountered a router with an errors_to setting. This could be provoked only
2446 by a command such as
2447
2448 exim -f "" ...
2449
2450 where an empty string was supplied; "<>" did not hit this bug.
2451
245211. Installed PCRE release 4.5.
2453
245412. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2455 remained set. It is now erased.
2456
245713. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2458 times from message ids (which are base 36 rather than the normal 62).
2459
246014. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2461 were being counted as actual protocol errors, and logged if the log
2462 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2463 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2464 and DATA following a set of rejected RCPTs do not count as protocol errors.
2465 In other words, Exim assumes they were pipelined, though this may not
2466 actually be the case. Of course, in all cases the client gets an
2467 appropriate error code.
2468
246915. If a lookup fails in an ACL condition, a message about the failure may
2470 be available; it is used if testing the ACL cannot continue, because most
2471 such messages specify what the cause of the deferral is. However, some
2472 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2473 that caused an old message to be retained and used if a later statement
2474 caused a defer, replacing the real cause of the deferral.
2475
247616. If an IP address had so many PTR records that the DNS lookup buffer
2477 was not large enough to hold them, Exim could crash while trying to process
2478 the truncated data. It now detects and logs this case.
2479
248017. Further to 4.21/58, another change has been made: if (and only if) the
2481 first line of a message (the first header line) ends with CRLF, a bare LF
2482 in a subsequent header line has a space inserted after it, so as not to
2483 terminate the header.
2484
248518. Refactoring: tidied an ugly bit of code in appendfile that copied data
2486 unnecessarily, used atoi() instead of strtol(), and didn't check the
2487 termination when getting file sizes from file names by regex.
2488
248919. Completely re-implemented the support for maildirsize files, in the light
2490 of a number of problems with the previous contributed implementation
2491 (4.30/29). In particular:
2492
2493 . If the quota is zero, the maildirsize file is maintained, but no quota is
2494 imposed.
2495
2496 . If the maildir directory does not exist, it is created before any attempt
2497 to write a maildirsize file.
2498
2499 . The quota value in the file is just a cache; if the quota is changed in
2500 the transport, the new value overrides.
2501
2502 . A regular expression is available for excluding directories from the
2503 count.
2504
250520. The autoreply transport checks the characters in options that define the
2506 message's headers; it allows continued headers, but it was checking with
2507 isspace() after an embedded newline instead of explicitly looking for a
2508 space or a tab.
2509
251021. If all the "regular" hosts to which an address was routed had passed their
2511 expiry times, and had not reached their retry times, the address was
2512 bounced, even if fallback hosts were defined. Now Exim should go on to try
2513 the fallback hosts.
2514
251522. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2516 equivalent code in the SMTP transport. Some hosts send humungous responses
2517 to HELO/EHLO, more than 1024 it seems.
2518
251923. Refactoring: code in filter.c used (void *) for "any old type" but this
2520 gives compiler warnings in some environments. I've now done it "properly",
2521 using a union.
2522
252324. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2524 (because of problems with the built-in one) was declared to return uschar *
2525 instead of char *, causing compiler failure.
2526
252725. Fixed a file descriptor leak when processing alias/forward files.
2528
252926. Fixed a minor format string issue in dbfn.c.
2530
253127. Typo in exim.c: ("dmbnz" for "dbmnz").
2532
253328. If a filter file refered to $h_xxx or $message_headers, and the headers
2534 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2535 become corrupted.
2536
253729. When a sender address is verified, it is cached, to save repeating the test
2538 when there is more than one recipient in a message. However, when the
2539 verification involves a callout, it is possible for different callout
2540 options to be set for different recipients. It is too complicated to keep
2541 track of this in the cache, so now Exim always runs a verification when a
2542 callout is required, relying on the callout cache for the optimization.
2543 The overhead is duplication of the address routing, but this should not be
2544 too great.
2545
254630. Fixed a bug in callout caching. If a RCPT command caused the sender address
2547 to be verified with callout=postmaster, and the main callout worked but the
2548 postmaster check failed, the verification correctly failed. However, if a
2549 subsequent RCPT command asked for sender verification *without* the
2550 postmaster check, incorrect caching caused this verification also to fail,
2551 incorrectly.
2552
255331. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2554 it was not caching the DNS options (qualify_single, search_parents) that
2555 were used when the lookup failed. A subsequent lookup with different
2556 options therefore always gave the same answer, though there were cases
2557 where it should not have. (Example: a "domains = !$mx_any" option on a
2558 dnslookup router: the "domains" option is always processed without any
2559 widening, but the router might have qualify_single set.) Now Exim uses the
2560 cached value only when the same options are set.
2561
256232. Added John Jetmore's "exipick" utility to the distribution.
2563
256433. GnuTLS: When an attempt to start a TLS session fails for any reason other
2565 than a timeout (e.g. a certificate is required, and is not provided), an
2566 Exim server now closes the connection immediately. Previously it waited for
2567 the client to close - but if the client is SSL, it seems that they each
2568 wait for each other, leading to a delay before one of them times out.
2569
257034: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2571 maintained 0.8.x compatibility because I don't think many are using it, and
2572 it is clearly obsolete.
2573
257435. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2575 transport.
2576
257736. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2578 client certificate was expired. A simple patch fixes this, though I don't
2579 understand the full logic of why the verify callback is called multiple
2580 times.
2581
258237. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2583 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2584 which causes problems with some clients (such as the Certicom SSL Plus
2585 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2586 disables the coutermeasure allowing Eudora to connect."
2587
258838. Exim was not checking that a write() to a log file succeeded. This could
2589 lead to Bad Things if a log got too big, in particular if it hit a file
2590 size limit. Exim now panics and dies if it cannot write to a log file, just
2591 as it does if it cannot open a log file.
2592
259339. Modified OS/Makefile-Linux so that it now contains
2594
2595 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2596
2597 The two -D definitions ensure that Exim is compiled with large file
2598 support, which makes it possible to handle log files that are bigger than
2599 2^31.
2600
260140. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2602 instance) a domain was checked against a named list that involved a lookup,
2603 causing $domain_data to be set, then another domain was checked against the
2604 same list, then the first domain was re-checked, the value of $domain_data
2605 after the final check could be wrong. In particular, if the second check
2606 failed, it could be set empty. This bug probably also applied to
2607 $localpart_data.
2608
260941. The strip_trailing_dot option was not being applied to the address given
2610 with the -f command-line option.
2611
261242. The code for reading a message's header from the spool was incrementing
2613 $received_count, but never initializing it. This meant that the value was
2614 incorrect (doubled) while delivering a message in the same process in which
2615 it was received. In the most common configuration of Exim, this never
2616 happens - a fresh exec is done - but it can happen when
2617 deliver_drop_privilege is set.
2618
261943. When Exim logs an SMTP synchronization error - client data sent too soon -
2620 it now includes up to 150 characters of the unexpected data in the log
2621 line.
2622
262344. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2624 and building data strings. The size of both of these buffers was 10 000
2625 bytes - far larger than anybody would *ever* want, thought I. Needless to
2626 say, somebody hit the limit. I have increased the maximum line length to
2627 20 000 and the maximum data length of concatenated lines to 100 000. I have
2628 also fixed two bugs, because there was no checking on these buffers. Tsk,
2629 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2630 buffer is too small.
2631
263245. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2633 lsearch lookups. Now it does.
2634
263546. When parsing a route_list item in a manualroute router, a fixed-length
2636 buffer was used for the list of hosts. I made this 1024 bytes long,
2637 thinking that nobody would ever have a list of hosts that long. Wrong.
2638 Somebody had a whole pile of complicated expansion conditions, and the
2639 string was silently truncated, leading to an expansion error. It turns out
2640 that it is easier to change to an unlimited length (owing to other changes
2641 that have happened since this code was originally written) than to build
2642 structure for giving a limitation error. The length of the item that
2643 expands into the list of hosts is now unlimited.
2644
264547. The lsearch lookup could not handle data where the length of text line was
2646 more than 4095 characters. Such lines were truncated, leading to shortened
2647 data being returned. It should now handle lines of any length.
2648
264948. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2650 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2651 ACL").
2652
265349. Cosmetic tidy to scripts like exicyclog that are generated by globally
2654 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2655 no longer happens in comment lines. A list of replacements is now placed
2656 at the head of all of the source files, except those whose only change is
2657 to replace PERL_COMMAND in the very first #! line.
2658
265950. Replaced the slow insertion sort in queue.c, for sorting the list of
2660 messages on the queue, with a bottom-up merge sort, using code contributed
2661 by Michael Haardt. This should make operations like -bp somewhat faster on
2662 large queues. It won't affect queue runners, except when queue_run_in_order
2663 is set.
2664
266551. Installed eximstats 1.31 in the distribution.
2666
266752. Added support for SRV lookups to the dnslookup router.
2668
266953. If an ACL referred to $message_body or $message_body_end, the value was not
2670 reset for any messages that followed in the same SMTP session.
2671
267254. The store-handling optimization for building very long strings was not
2673 differentiating between the different store pools. I don't think this
2674 actually made any difference in practice, but I've tidied it.
2675
267655. While running the routers to verify a sender address, $sender_address
2677 was still set to the sender address. This is wrong, because when routing to
2678 send a bounce to the sender, it would be empty. Therefore, I have changed
2679 it so that, while verifying a sender address, $sender_address is set to <>.
2680 (There is no change to what happens when verifying a recipient address.)
2681
268256. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2683 target A or AAAA records (if not already returned) without resetting the
2684 qualify_single or search_parents options of the DNS resolver. These are
2685 inappropriate in this case because the targets of MX and SRV records must
2686 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2687 target that, when qualified, matched something in the local domain. These
2688 two options are now turned off when doing these lookups.
2689
269057. It seems that at least some releases of Reiserfs (which does not have the
2691 concept of a fixed number of inodes) returns zero and not -1 for the
2692 number of available inodes. This interacted badly with check_spool_inodes,
2693 which assumed that -1 was the "no such thing" setting. What I have done is
2694 to check that the total number of inodes is greater than zero before doing
2695 the test of how many are available.
2696
269758. When a "warn" ACL statement has a log_message modifier, the message is
2698 remembered, and not repeated. This is to avoid a lot of repetition when a
2699 message has many recipients that cause the same warning to be written.
2700 Howewer, Exim was preserving the list of already written lines for an
2701 entire SMTP session, which doesn't seem right. The memory is now reset if a
2702 new message is started.
2703
270459. The "rewrite" debugging flag was not showing the result of rewriting in the
2705 debugging output unless log_rewrite was also set.
2706
270760. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2708 of (int)(handle) when we know that handle contains (void *)(-1).
2709
271061. The Exim daemon panic-logs an error return when it closes the incoming
2711 connection. However "connection reset by peer" seems to be common, and
2712 isn't really an error worthy of noting specially, so that particular error
2713 is no long logged.
2714
271562. When Exim is trying to find all the local interfaces, it used to panic and
2716 die if the ioctl to get the interface flags failed. However, it seems that
2717 on at least one OS (Solaris 9) it is possible to have an interface that is
2718 included in the list of interfaces, but for which you get a failure error
2719 for this call. This happens when the interface is not "plumbed" into a
2720 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2721 failure of the "get flags" call assumes that the interface is down.
2722
272363. Added a ${eval10: operator, which assumes all numbers are decimal. This
2724 makes life easier for people who are doing arithmetic on fields extracted
2725 from dates, where you often get leading zeros that should not be
2726 interpreted as octal.
2727
272864. Added qualify_domain to the redirect router, to override the global
2729 setting.
2730
273165. If a pathologically long header line contained very many addresses (the
2732 report of this problem mentioned 10 000) and each of them was rewritten,
2733 Exim could use up a very large amount of memory. (It kept on making new
2734 copies of the header line as it rewrote, and never released the old ones.)
2735 At the expense of a bit more processing, the header rewriting function has
2736 been changed so that it no longer eats memory in this way.
2737
273866. The generation of the Received: header has been moved from the time that a
2739 message starts to be received, to the time that it finishes. The timestamp
2740 in the Received: header should now be very close to that of the <= log
2741 line. There are two side-effects of this change:
2742
2743 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2744 logged header lines no longer include the local Received: line, because
2745 it has not yet been created. The same applies to a copy of the message
2746 that is returned to a non-SMTP sender when a message is rejected.
2747
2748 (b) When a filter file is tested using -bf, no additional Received: header
2749 is added to the test message. After some thought, I decided that this
2750 is a bug fix.
2751
2752 This change does not affect the value of $received_for. It is still set
2753 after address rewriting, but before local_scan() is called.
2754
275567. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2756
275768. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2758 gave an unhelpful panic error message, and a defer error. I have managed to
2759 change this behaviour so that it now rejects any supplied certificate,
2760 which seems right, as the list of acceptable certificates is empty.
2761
276269. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2763 gave an unhelpful defer error. I have not managed to make this reject any
2764 supplied certificates, but the error message it gives is "no certificate
2765 supplied", which is not helpful.
2766
276770. exigrep's output now also includes lines that are not associated with any
2768 message, but which match the given pattern. Implemented by a patch from
2769 Martin Sluka, which also tidied up the Perl a bit.
2770
277171. Recipient callout verification, like sender verification, was using <> in
2772 the MAIL FROM command. This isn't really the right thing, since the actual
2773 sender may affect whether the remote host accepts the recipient or not. I
2774 have changed it to use the actual sender in the callout; this means that
2775 the cache record is now keyed on a recipient/sender pair, not just the
2776 recipient address. There doesn't seem to be a real danger of callout loops,
2777 since a callout by the remote host to check the sender would use <>.
2778 [SEE ABOVE: changed after hitting problems.]
2779
278072. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2781 temporary errors. However, in the case of such a code being given after
2782 the end of a data transmission (i.e. after ".") Exim was failing to write
2783 a retry record for the message. (Yes, there was some broken host that was
2784 actually sending 8xx at this point.)
2785
278673. An unknown lookup type in a host list could cause Exim to panic-die when
2787 the list was checked. (An example that provoked this was putting <; in the
2788 middle of a list instead of at the start.) If this happened during a DATA
2789 ACL check, a -D file could be left lying around. This kind of configuration
2790 error no longer causes Exim to die; instead it causes a defer errror. The
2791 incident is still logged to the main and panic logs.
2792
279374. Buglet left over from Exim 3 conversion. The message "too many messages
2794 in one connection" was written to the rejectlog but not the mainlog, except
2795 when address rewriting (yes!) was being logged.
2796
279775. Added write_rejectlog option.
2798
279976. When a system filter was run not as root (that is, when system_filter_user
2800 was set), the values of the $n variables were not being returned to the
2801 main process; thus, they were not subsequently available in the $sn
2802 variables.
2803
280477. Added +return_path_on_delivery log selector.
2805
280678. A connection timeout was being treated differently from recipients deferred
2807 when testing hosts_max_try with a message that was older than the host's
2808 retry timeout. (The host should not be counted, thus allowing all hosts to
2809 be tried at least once before bouncing.) This may have been the cause of an
2810 occasionally reported bug whereby a message would remain on the queue
2811 longer than the retry timeout, but would be bounced if a delivery was
2812 forced. I say "may" because I never totally pinned down the problem;
2813 setting up timeout/retry tests is difficult. See also the next item.
2814
281579. The ultimate address timeout was not being applied to errors that involved
2816 a combination of host plus message (for example, a timeout on a MAIL
2817 command). When an address resolved to a number of possible hosts, and they
2818 were not all tried for each delivery (e.g. because of hosts_max_try), a
2819 message could remain on the queue longer than the retry timeout.
2820
282180. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2822 Haardt.
2823
282481. Fixed an obscure SMTP outgoing bug which required at least the following
2825 conditions: (a) there was another message waiting for the same server;
2826 (b) the server returned 5xx to all RCPT commands in the first message so
2827 that the message was not completed; (c) the server dropped the connection
2828 or gave a negative response to the RSET that Exim sends to abort the
2829 transaction. The observed case was a dropped connection after DATA that had
2830 been sent in pipelining mode. That is, the server had advertised PIPELINING
2831 but was not implementing it correctly. The effect of the bug was incorrect
2832 behaviour, such as trying another host, and this could lead to a crash.
2833
2834
2835Exim version 4.30
2836-----------------
2837
2838 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2839 and daemon.c were passed as pointers to ints; they should have been
2840 pointers to socklen_t variables (which are typically unsigned ints).
2841
2842 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2843 fixed.
2844
2845 3. Fixed a really odd bug that affected only the testing scheme; patching a
2846 certain fixed string in the binary changed the value of another string that
2847 happened to be identical to the end of the original first string.
2848
2849 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2850 name", it returns that address as the IP address. On some operating
2851 systems (e.g. Solaris), it also passes back the IP address string as the
2852 "host name". However, on others (e.g. Linux), it passes back an empty
2853 string. Exim wasn't checking for this, and was changing the host name to an
2854 empty string, assuming it had been canonicized.
2855
2856 5. Although rare, it is permitted to have more than one PTR record for a given
2857 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2858 all the names associated with an address, because they do in Solaris.
2859 However, it seems that they do not in Linux for data that comes from the
2860 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2861 I found this out when I moved to a new Linux workstation and tried to run
2862 the Exim test suite.
2863
2864 To get round this problem I have changed the code so that it now does its
2865 own call to the DNS to look up PTR records when searching for a host name.
2866 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2867 addresses that are only in /etc/hosts are still found.
2868
2869 This behaviour is, however, controlled by an option called host_lookup_
2870 order, which defaults to "bydns:byaddr". If people want to use the other
2871 order, or indeed, just use one or the other means of lookup, they can
2872 specify it in this variable.
2873
2874 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2875 some operating systems, this comes back from gethostbyaddr() as an empty
2876 string, and this is what Exim used to test for. However, it seems that in
2877 other systems, "." is yielded. Exim now tests for this case too.
2878
2879 7. The values of check_spool_space and check_log_space are now held internally
2880 as a number of kilobytes instead of an absolute number of bytes. If a
2881 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2882 kilobyte. This means that much larger values can be stored.
2883
2884 8. Exim monitor: an attempt to get the action menu when not actually pointing
2885 at a message produces an empty menu entitled "No message selected". This
2886 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2887 no entries in it ("Shell widget menu has zero width and/or height"). So I
2888 have added a single, blank menu entry in this case.
2889
2890 9. Added ${quote_local_part.
2891
289210. MIME decoding is now applied to the contents of Subject: header lines when
2893 they are logged.
2894
289511. Now that a reference to $sender_host_address automatically causes a reverse
2896 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2897 host lookup before query-style lookups in lists that might use this
2898 variable. This has therefore been abolished, and the "net-" prefix is no
2899 longer necessary for query-style lookups.
2900
290112. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2902 have been a typo for LFLAGS, so it has been changed.
2903
290413. The install script calls Exim with "-C /dev/null" in order to find the
2905 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2906 to be output. Howeve, since Exim outputs its version number before the
2907 error, it didn't break the script. It just looked ugly. I fixed this by
2908 always allowing "-C /dev/null" if the caller is root.
2909
291014. Ignore overlarge ACL variable number when reading spool file - insurance
2911 against a later release with more variables having written the file.
2912
291315. The standard form for an IPv6 address literal was being rejected by EHLO.
2914 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2915 this, as well as the form without the "IPv6" on the front.
2916
291716. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2918 OS/Makefile-Darwin file.
2919
292017. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2921 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2922
292318. After change 4.21/52, "%ld" was used to format the contents of the $inode
2924 variable. However, some OS use ints for inodes. I've added cast to long int
2925 to get rid of the compiler warning.
2926
292719. I had forgotten to lock out "/../" in configuration file names when
2928 ALT_CONFIG_PREFIX was set.
2929
293020. Routers used for verification do not need to specify transports. However,
2931 if such a router generated a host list, and callout was configured, Exim
2932 crashed, because it could not find a port number from the (non-existent)
2933 transport. It now assumes port 25 in this circumstance.
2934
293521. Added the -t option to exigrep.
2936
293722. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2938 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2939 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2940 them is set, LOOKUP_LSEARCH is forced.
2941
294223. "exim -bV" now outputs a list of lookups that are included in the binary.
2943
294424. Added sender and host information to the "rejected by local_scan()" log
2945 line; previously there was no indication of these.
2946
294725. Added .include_if_exists.
2948
294926. Change 3.952/11 added an explicit directory sync on top of a file sync for
2950 Linux. It turns out that not all file systems support this. Apparently some
2951 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2952 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2953 not supported on the file descriptor, is now ignored when Exim is trying to
2954 sync a directory. This applies only to Linux.
2955
295627. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2957
295828. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2959 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2960 this.
2961
296229. Added support for maildirsize files from supplied patch (modified a bit).
2963
296430. The use of :fail: followed by an empty string could lead Exim to respond to
2965 sender verification failures with (e.g.):
2966
2967 550 Verification failed for <xxx>
2968 550 Sender verify failed
2969
2970 where the first response line was missing the '-' that indicates it is not
2971 the final line of the response.
2972
297331. The loop for finding the name of the user that called Exim had a hardwired
2974 limit of 10; it now uses the value of finduser_retries, which is used for
2975 all other user lookups.
2976
297732. Added $received_count variable, available in data and not_smtp ACLs, and at
2978 delivery time.
2979
298033. Exim was neglecting to zero errno before one call of strtol() when
2981 expanding a string and expecting an integer value. On some systems this
2982 resulted in spurious "integer overflow" errors. Also, it was casting the
2983 result into an int without checking.
2984
298534. Testing for a connection timeout using "timeout_connect" in the retry rules
2986 did not work. The code looks as if it has *never* worked, though it appears
2987 to have been documented since at least releast 1.62. I have made it work.
2988
298935. The "timeout_DNS" error in retry rules, also documented since at least
2990 1.62, also never worked. As it isn't clear exactly what this means, and
2991 clearly it isn't a major issue, I have abolished the feature by treating it
2992 as "timeout", and writing a warning to the main and panic logs.
2993
299436. The display of retry rules for -brt wasn't always showing the error code
2995 correctly.
2996
299737. Added new error conditions to retry rules: timeout_A, timeout_MX,
2998 timeout_connect_A, timeout_connect_MX.
2999
300038. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3001 to the empty sender.
3002
300339. The daemon was not analysing the content of -oX till after it had closed
3004 stderr and disconnected from the controlling terminal. This meant that any
3005 syntax errors were only noted on the panic log, and the return code from
3006 the command was 0. By re-arranging the code a little, I've made the
3007 decoding happen first, so such errors now appear on stderr, and the return
3008 code is 1. However, the actual setting up of the sockets still happens in
3009 the disconnected process, so errors there are still only recorded on the
3010 panic log.
3011
301240. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3013 connections (as happens on some IP stacks) was logged at start up time as
3014 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3015 it from "IPv6 and IPv4", which means that two separate sockets are being
3016 used.
3017
301841. The debug output for gethostbyname2() or getipnodebyname() failures now
3019 says whether AF_INET or AF_INET6 was passed as an argument.
3020
302142. Exiwhat output was messed up when time zones were included in log
3022 timestamps.
3023
302443. Exiwhat now gives more information about the daemon's listening ports,
3025 and whether -tls-on-connect was used.
3026
302744. The "port" option of the smtp transport is now expanded.
3028
302945. A "message" modifier in a "warn" statement in a non-message ACL was being
3030 silently ignored. Now an error message is written to the main and panic
3031 logs.
3032
303346. There's a new ACL modifier called "logwrite" which writes to a log file
3034 as soon as it is encountered.
3035
303647. Added $local_user_uid and $local_user_gid at routing time.
3037
303848. Exim crashed when trying to verify a sender address that was being
3039 rewritten to "<>".
3040
304149. Exim was recognizing only a space character after ".include". It now also
3042 recognizes a tab character.
3043
304450. Fixed several bugs in the Perl script that creates the exim.8 man page by
3045 extracting the relevant information from the specification. The man page no
3046 longer contains scrambled data for the -d option, and I've added a section
3047 at the front about calling Exim under different names.
3048
304951. Added "extra_headers" argument to the "mail" command in filter files.
3050
305152. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3052 crash.
3053
305453. Installed eximstats 1.29.
3055
305654. Added transport_filter_timeout as a generic transport option.
3057
305855. Exim no longer adds an empty Bcc: header to messages that have no To: or
3059 Cc: header lines. This was required by RFC 822, but it not required by RFC
3060 2822.
3061
306256. Exim used to add From:, Date:, and Message-Id: header lines to any
3063 incoming messages that did not have them. Now it does so only if the
3064 message originates locally, that is, if there is no associated remote host
3065 address. When Resent- header lines are present, this applies to the Resent-
3066 lines rather than the non-Resent- lines.
3067
306857. Drop incoming SMTP connection after too many syntax or protocol errors. The
3069 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3070
307158. Messages for configuration errors now include the name of the main
3072 configuration file - useful now that there may be more than one file in a
3073 list (.included file names were always shown).
3074
307559. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3076 for those rare installations that do not start the daemon as root or run it
3077 setuid root. I've cut out the call to initgroups() if the daemon is not
3078 root at that time.
3079
308060. The Exim user and group can now be bound into the binary as text strings
3081 that are looked up at the start of Exim's processing.
3082
308361. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3084
308562. Added $mailstore_basename variable.
3086
308763. Installed patch to sieve.c from Michael Haardt.
3088
308964. When Exim failed to open the panic log after failing to open the main log,
3090 the original message it was trying to log was written to stderr and debug
3091 output, but if they were not available (the usual case in production), it
3092 was lost. Now it is written to syslog before the two lines that record the
3093 failures to open the logs.
3094
309565. Users' Exim filters run in subprocesses under the user's uid. It is
3096 possible for a "deliver" command or an alias in a "personal" command to
3097 provoke an address rewrite. If logging of address rewriting is configured,
3098 this fails because the process is not running as root or exim. There may be
3099 a better way of dealing with this, but for the moment (because 4.30 needs
3100 to be released), I have disabled address rewrite logging when running a
3101 filter in a non-root, non-exim process.
3102
3103
3104Exim version 4.24
3105-----------------
3106
3107 1. The buildconfig auxiliary program wasn't quoting the value set for
3108 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3109 not defined. This bug was masked in 4.22 by the effect that was fixed in
3110 change 4.23/1.
3111
3112 2. Some messages that were rejected after a message id was allocated were
3113 shown as "incomplete" by exigrep. It no longer does this for messages that
3114 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3115
3116 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3117 have allow_domain_literals set, the ID did not get logged in the <= line.
3118 Domain literals are now always recognized in Message-ID: header lines.
3119
3120 4. The first argument for a ${extract expansion item is the key name or field
3121 number. Leading and trailing spaces in this item were not being ignored,
3122 causing some misleading effects.
3123
3124 5. When deliver_drop_privilege was set, single queue runner processes started
3125 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3126 same command in the process it spins off) were not dropping privilege.
3127
3128 6. When the daemon running as "exim" started a queue runner, it always
3129 re-executed Exim in the spun-off process. This is a waste of effort when
3130 deliver_drop_privilege is set. The new process now just calls the
3131 queue-runner function directly.
3132
3133
3134Exim version 4.23
3135-----------------
3136
3137 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3138 HEADERS_CHARSET.
3139
3140 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3141 ignored. Though the use of -oP was forcing the writing of a pid file, it
3142 was always written to the default place.
3143
3144 3. If the message "no IP address found for host xxxx" is generated during
3145 incoming verification, it is now followed by identification of the incoming
3146 connection (so you can more easily find what provoked it).
3147
3148 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3149
3150 5. Added some features to "harden" Exim a bit more against certain attacks:
3151
3152 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3153 be put in Local/Makefile. This is like the never_users runtime option,
3154 but it cannot be overridden. The default setting is "root".
3155
3156 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3157 prefix string with which any file named in a -C command line option
3158 must start.
3159
3160 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3161 is retained for -C and -D only if the caller of Exim is root. Without
3162 it, the exim user may also use -C and -D and retain privilege.
3163
3164 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3165 command line option is disabled.
3166
3167 6. Macro names set by the -D option must start with an upper case letter, just
3168 like macro names defined in the configuration file.
3169
3170 7. Added "dereference=" facility to LDAP.
3171
3172 8. Two instances of the typo "uknown" in the source files are fixed.
3173
3174 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3175 the Configure-Makefile script screwed up while processing it.
3176
317710. Incorporated PCRE 4.4.
3178
317911. The SMTP synchronization check was not operating right at the start of an
3180 SMTP session. For example, it could not catch a HELO sent before the client
3181 waited for the greeting. There is now a check for outstanding input at the
3182 point when the greeting is written. Because of the duplex, asynchronous
3183 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3184 way, but not yet received, when the check is performed.
3185
318612. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3187 on TCP/IP sockets, because this apparently causes some broken clients to
3188 timeout.
3189
319013. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3191 unchanged) from the Cygwin maintainer.
3192
319314. The code for -bV that shows what is in the binary showed "mbx" when maildir
3194 was supported instead of testing for mbx. Effectively a typo.
3195
319615. The spa authenticator server code was not checking that the input it
3197 received was valid base64.
3198
319916. The debug output line for the "set" modifier in ACLs was not showing the
3200 name of the variable that was being set.
3201
320217. Code tidy: the variable type "vtype_string" was never used. Removed it.
3203
320418. Previously, a reference to $sender_host_name did not cause a DNS reverse
3205 lookup on its own. Something else was needed to trigger the lookup. For
3206 example, a match in host_lookup or the need for a host name in a host list.
3207 Now, if $sender_host_name is referenced and the host name has not yet been
3208 looked up, a lookup is performed. If the lookup fails, the variable remains
3209 empty, and $host_lookup_failed is set to "1".
3210
321119. Added "eqi" as a case-independent comparison operator.
3212
321320. The saslauthd authentication condition could segfault if neither service
3214 nor realm was specified.
3215
321621. If an overflowing value such as "2048M" was set for message_size_limit, the
3217 error message that was logged was misleading, and incoming SMTP
3218 connections were dropped. The message is now more accurate, and temporary
3219 errors are given to SMTP connections.
3220
322122. In some error situations (such as 21 above) Exim rejects all SMTP commands
3222 (except RSET) with a 421 error, until QUIT is received. However, it was
3223 failing to send a response to QUIT.
3224
322523. The HELO ACL was being run before the code for helo_try_verify_hosts,
3226 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3227 ACL is now run after the helo_try_verify_hosts code.
3228
322924. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3230 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3231 case-independent, so other case variants are also recognized). Apparently
3232 some systems use these upper case variants.
3233
323425. If more than two messages were waiting for the same host, and a transport
3235 filter was specified for the transport, Exim sent two messages over the
3236 same TCP/IP connection, and then failed with "socket operation on non-
3237 socket" when it tried to send the third.
3238
323926. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3240
324127. The extern definition of crypt16() in expand.c was not being excluded when
3242 the OS had its own crypt16() function.
3243
324428. Added bounce_return_body as a new option, and bounce_return_size_limit
3245 as a preferred synonym for return_size_limit, both as an option and as an
3246 expansion variable.
3247
324829. Added LIBS=-liconv to OS/Makefile-OSF1.
3249
325030. Changed the default configuration ACL to relax the local part checking rule
3251 for addresses that are not in any local domains. For these addresses,
3252 slashes and pipe symbols are allowed within local parts, but the sequence
3253 /../ is explicitly forbidden.
3254
325531. SPA server authentication was not clearing the challenge buffer before
3256 using it.
3257
325832. log_message in a "warn" ACL statement was writing to the reject log as
3259 well as to the main log, which contradicts the documentation and doesn't
3260 seem right (because no rejection is happening). So I have stopped it.
3261
326233. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3263 However, I am unable to do any testing of this.
3264
326534. Fixed an infelicity in the appendfile transport. When checking directories
3266 for a mailbox, to see if any needed to be created, it was accidentally
3267 using path names with one or more superfluous leading slashes; tracing
3268 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3269
327035. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3271 contents are added to the log line that is written for every discarded
3272 recipient. (Previously a log_message setting was ignored.)
3273
327436. The ${quote: operator now quotes the string if it is empty.
3275
327637. The install script runs exim in order to find its version number. If for
3277 some reason other than non-existence or emptiness, which it checks, it
3278 could not run './exim', it was installing it with an empty version number,
3279 i.e. as "exim-". This error state is now caught, and the installation is
3280 aborted.
3281
328238. An argument was missing from the function that creates an error message
3283 when Exim fails to connect to the socket for saslauthd authentication.
3284 This could cause Exim to crash, or give a corrupted message.
3285
328639. Added isip, isip4, and isip6 to ${if conditions.
3287
328840. The ACL variables $acl_xx are now saved with the message, and can be
3289 accessed later in routers, transports, and filters.
3290
329141. The new lookup type nwildlsearch is like wildlsearch, except that the key
3292 strings in the file are not string-expanded.
3293
329442. If a MAIL command specified a SIZE value that was too large to fit into an
3295 int variable, the check against message_size_limit failed. Such values are
3296 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3297 day this will have to be increased, but I don't think I want to be around
3298 when emails are that large.
3299
3300
3301
3302Exim version 4.22
3303-----------------
3304
3305 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3306 iconv() is not standard in FreeBSD.
3307
3308 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3309 IPv6 enabled. The observed symptom was a segmentation fault on return from
3310 the function os_common_find_running_interfaces() in src/os.c.
3311
3312 3. In the check_special_case() function in daemon.c I had used "errno" as an
3313 argument name, which causes warnings on some systems. This was basically a
3314 typo, since it was named "eno" in the comments!
3315
3316 4. The code that waits for the clock to tick (at a resolution of some fraction
3317 of a second) so as to ensure message-id uniqueness was always waiting for
3318 at least one whole tick, when it could have waited for less. [This is
3319 almost certainly not relevant at current processor speeds, where it is
3320 unlikely to ever wait at all. But we try to future-proof.]
3321
3322 5. The function that sleeps for a time interval that includes fractions of a
3323 second contained a race. It did not block SIGALRM between setting the
3324 timer, and suspending (a couple of lines later). If the interval was short
3325 and the sigsuspend() was delayed until after it had expired, the suspension
3326 never ended. On busy systems this could lead to processes getting stuck for
3327 ever.
3328
3329 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3330 process, before it forks any delivery processes. The open lookup caching
3331 mechanism meant that the open file or database connection was passed into
3332 the delivery process. The problem was that delivery processes always tidy
3333 up cached lookup data. This could cause a problem for the next delivery
3334 process started by the queue runner, because the external queue runner
3335 process does not know about the closure. So the next delivery process
3336 still has data in the lookup cache. In the case of a file lookup, there was
3337 no problem because closing a file descriptor in a subprocess doesn't affect
3338 the parent. However, if the lookup was caching a connection to a database,
3339 the connection was closed, and the second delivery process was likely to
3340 see errors such as "PGSQL: query failed: server closed the connection
3341 unexpectedly". The problem has been fixed by closing all cached lookups
3342 in a queue runner before running a delivery process.
3343
3344 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3345 seem to have the "const" qualifier which it has on other OS. I've
3346 parameterised it.
3347
3348 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3349 *of the same type* (client or server) with the same public name that an
3350 error should be diagnosed.
3351
3352 9. When Exim looked up a host name for an IP address, but failed to find the
3353 original IP address when looking up the host name (a safety check), it
3354 output the message "<ip address> does not match any IP for NULL", which was
3355 confusing, to say the least. The bug was that the host name should have
3356 appeared instead of "NULL".
3357
335810. Since release 3.03, if Exim is called by a uid other than root or the Exim
3359 user that is built into the binary, and the -C or -D options is used, root
3360 privilege is dropped before the configuration file is read. In addition,
3361 logging is switched to stderr instead of the normal log files. If the
3362 configuration then re-defines the Exim user, the unprivileged environment
3363 is probably not what is expected, so Exim logs a panic warning message (but
3364 proceeds).
3365
3366 However, if deliver_drop_privilege is set, the unprivileged state may well
3367 be exactly what is intended, so the warning has been cut out in that case,
3368 and Exim is allowed to try to write to its normal log files.
3369
3370
3371Exim version 4.21
3372-----------------
3373
3374 1. smtp_return_error_details was not giving details for temporary sender
3375 or receiver verification errors.
3376
3377 2. Diagnose a configuration error if two authenticators have the same public
3378 name.
3379
3380 3. Exim used not to create the message log file for a message until the first
3381 delivery attempt. This could be confusing when incoming messages were held
3382 for policy or load reasons. The message log file is now created at the time
3383 the message is received, and an initial "Received" line is written to it.
3384
3385 4. The automatically generated man page for command line options had a minor
3386 bug that caused no ill effects; however, a more serious problem was that
3387 the procedure for building the man page automatically didn't always
3388 operate. Consequently, release 4.20 contains an out-of-date version. This
3389 shouldn't happen again.
3390
3391 5. When building Exim with embedded Perl support, the script that builds the
3392 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3393 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3394
3395 6. The freeze_tell option was not being used for messages that were frozen on
3396 arrival, either by an ACL or by local_scan().
3397
3398 7. Added the smtp_incomplete_transaction log selector.
3399
3400 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3401 was accepting AUTH without a new EHLO.
3402
3403 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3404 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3405 RFC.
3406
340710. Logging of TCP/IP connections (when configured) now happens in the main
3408 daemon process instead of the child process, so that the TCP/IP connection
3409 count is more accurate (but it can never be perfect).
3410
341111. The use of "drop" in a nested ACL was not being handled correctly in the
3412 outer ACL. Now, if condition failure induced by the nested "drop" causes
3413 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3414 or "require"), the connection is dropped.
3415
341612. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3417 that yield "discard" can now be used with an "accept" or a "discard" verb,
3418 but an error is generated for any others (because I can't see a useful way
3419 to define what should happen).
3420
342113. When an ACL is read dynamically from a file (or anywhere else), the lines
3422 are now processed in the same way as lines in the Exim configuration file.
3423 In particular, continuation lines are supported.
3424
342514. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3426
342715. Added -ti meaning -t -i.
3428
342916. Check for letters, digits, hyphens, and dots in the names of dnslist
3430 domains, and warn by logging if others are found.
3431
343217. At least on BSD, alignment is not guarenteed for the array of ifreq's
3433 returned from GIFCONF when Exim is trying to find the list of interfaces on
3434 a host. The code in os.c has been modified to copy each ifreq to an aligned
3435 structure in all cases.
3436
3437 Also, in some cases, the returned ifreq's were being copied to a 'struct
3438 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3439 means the last couple of bytes of an IPv6 address could be chopped if the
3440 ifreq contained only a normal sockaddr (14 bytes storage).
3441
344218. Named domain lists were not supported in the hosts_treat_as_local option.
3443 An entry such as +xxxx was not recognized, and was treated as a literal
3444 domain name.
3445
344619. Ensure that header lines added by a DATA ACL are included in the reject log
3447 if the ACL subsequently rejects the message.
3448
344920. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3450 MD5 (which is deprecated).
3451
345221. When testing a filter file using -bf, Exim was writing a message when it
3453 took the sender from a "From " line in the message, but it was not doing so
3454 when it took $return_path from a Return-Path: header line. It now does.
3455
345622. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3457 with a valid header line field name (a series of printing characters
3458 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3459
346023. Changed "disc" in the source to "disk" to conform to the documentation and
3461 the book and for uniformity.
3462
346324. Ignore Sendmail's -Ooption=value command line item.
3464
346525. When execve() failed while trying to run a command in a pipe transport,
3466 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3467 could be confused with a return value of 69 from the command itself. This
3468 has been changed to 127, the value the shell returns if it is asked to run
3469 a non-existent command. The wording for the related log line suggests a
3470 non-existent command as the problem.
3471
347226. If received_header_text expands to an empty string, do not add a Received:
3473 header line to the message. (Well, it adds a token one on the spool, but
3474 marks it "old" so that it doesn't get used or transmitted.)
3475
347627. Installed eximstats 1.28 (addition of -nt option).
3477
347828. There was no check for failure on the call to getsockname() in the daemon
3479 code. This can fail if there is a shortage of resources on the system, with
3480 ENOMEM, for example. A temporary error is now given on failure.
3481
348229. Contrary to the C standard, it seems that in some environments, the
3483 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3484 program. Exim now does this explicitly; it affects the formatting of
3485 timestamps using strftime().
3486
348730. If exiqsumm was given junk data, it threw up some uninitialized variable
3488 complaints. I've now initialized all the variables, to avoid this.
3489
349032. Header lines added by a system filter were not being "seen" during
3491 transport-time rewrites.
3492
349333. The info_callback() function passed to OpenSSL is set up with type void
3494 (*)(SSL *, int, int), as described somewhere. However, when calling the
3495 function (actually a macro) that sets it up, the type void(*)() is
3496 expected. I've put in a cast to prevent warnings from picky compilers.
3497
349834. If a DNS black list lookup found a CNAME record, but there were no A
3499 records associated with the domain it pointed at, Exim crashed.
3500
350135. If a DNS black list lookup returned more than one A record, Exim ignored
3502 all but the first. It now scans all returned addresses if a particular IP
3503 value is being sought. In this situation, the contents of the
3504 $dnslist_value variable are a list of all the addresses, separated by a
3505 comma and a space.
3506
350736. Tightened up the rules for host name lookups using reverse DNS. Exim used
3508 to accept a host name and all its aliases if the forward lookup for any of
3509 them yielded the IP address of the incoming connection. Now it accepts only
3510 those names whose forward lookup yields the correct IP address. Any other
3511 names are discarded. This closes a loophole whereby a rogue DNS
3512 administrator could create reverse DNS records to break through a
3513 wildcarded host restriction in an ACL.
3514
351537. If a user filter or a system filter that ran in a subprocess used any of
3516 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3517 the wrong values were passed to the pipe command ($thisaddress had the
3518 value of $0, $0 had the value of $1, etc). This bug was introduced by
3519 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3520
352138. Improved the line breaking for long SMTP error messages from ACLs.
3522 Previously, if there was no break point between 40 and 75 characters, Exim
3523 left the rest of the message alone. Two changes have been made: (a) I've
3524 reduced the minimum length to 35 characters; (b) if it can't find a break
3525 point between 35 and 75 characters, it looks ahead and uses the first one
3526 that it finds. This may give the occasional overlong line, but at least the
3527 remaining text gets split now.
3528
352939. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3530 file descriptors might be low, and that setting 1000 would always raise it.
3531 It turns out that in some environments, the limit is already over 1000 and
3532 that lowering it causes trouble. So now Exim takes care not to decrease it.
3533
353440. When delivering a message, the value of $return_path is set to $sender_
3535 address at the start of routing (routers may change the value). By an
3536 oversight, this default was not being set up when an address was tested by
3537 -bt or -bv, which affected the outcome if any router or filter referred to
3538 $return_path.
3539
354041. The idea of the "warn" ACL verb is that it adds a header or writes to the
3541 log only when "message" or "log_message" are set. However, if one of the
3542 conditions was an address verification, or a call to a nested ACL, the
3543 messages generated by the underlying test were being passed through. This
3544 no longer happens. The underlying message is available in $acl_verify_
3545 message for both "message" and "log_message" expansions, so it can be
3546 passed through if needed.
3547
354842. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3549 new expansion $bh_ to give the encoded byte string without charset
3550 translation. Translation happens only if iconv() is available; HAVE_ICONV
3551 indicates this at build time. HEADERS_CHARSET gives the charset to
3552 translate to; headers_charset can change it in the configuration, and
3553 "headers charset" can change it in an individual filter file.
3554
355543. Now that we have a default RFC 2047 charset (see above), the code in Exim
3556 that creates RFC 2047 encoded "words" labels them as that charset instead
3557 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3558 expansion operator; (ii) when Exim creates a From: line for a local
3559 message; (iii) when a header line is rewritten to include a "phrase" part.
3560
356144. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3562 buggy, causing it to skip the first lines of messages whose message ID
3563 ended in 'D'. This would not have bitten before Exim release 4.14, because
3564 message IDs were unlikely to end in 'D' before then. The effect was to have
3565 incorrect size information for certain domains.
3566
356745. #include "config.h" was missing at the start of the crypt16.c module. This
3568 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3569 noticed.
3570
357146. If there was a timeout during a "random" callout check, Exim treated it as
3572 a failure of the random address, and carried on sending RSET and the real
3573 address. If the delay was just some slowness somewhere, the response to the
3574 original RCPT would be taken as a response to RSET and so on, causing
3575 mayhem of various kinds.
3576
357747. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3578 when I implemented it. It didn't allow for the fact that some option values
3579 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3580 the right test for positive values.
3581
358248. Domain names in DNS records are case-independent. Exim always looks them up
3583 in lower case. Some resolvers return domain names in exactly the case they
3584 appear in the zone file, that is, they may contain uppercase letters. Not
3585 all resolvers do this - some return always lower case. Exim was treating a
3586 change of case by a resolver as a change of domain, similar to a widening
3587 of a domain abbreviation. This triggered its re-routing code and so it was
3588 trying to route what was effectively the same domain again. This normally
3589 caused routing to fail (because the router wouldn't handle the domain
3590 twice). Now Exim checks for this case specially, and just changes the
3591 casing of the domain that it ultimately uses when it transmits the message
3592 envelope.
3593
359449. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3595 module.
3596
359750. If a filter generated a file delivery with a non-absolute name (possible if
3598 no home directory exists for the router), the forbid_file option was not
3599 forbidding it.
3600
360151. Added '&' feature to dnslists, to provide bit mask matching in addition to
3602 the existing equality matching.
3603
360452. Exim was using ints instead of ino_t variables in some places where it was
3605 dealing with inode numbers.
3606
360753. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3608 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3609 called TMPDIR, and if it finds it is different, it changes its value.
3610
361154. The smtp_printf() function is now made available to local_scan() so
3612 additional output lines can be written before returning. There is also an
3613 smtp_fflush() function to enable the detection of a dropped connection.
3614 The variables smtp_input and smtp_batched_input are exported to
3615 local_scan().
3616
361755. Changed the default runtime configuration: the message "Unknown user"
3618 has been removed from the ACL, and instead placed on the localuser router,
3619 using the cannot_route_message feature. This means that any verification
3620 failures that generate their own messages won't get overridden. Similarly,
3621 the "Unrouteable address" message that was in the ACL for unverifiable
3622 relay addresses has also been removed.
3623
362456. Added hosts_avoid_esmtp to the smtp transport.
3625
362657. The exicyclog script was not checking for the esoteric option
3627 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3628 will work only if exicyclog is run under the appropriate euid.
3629
363058. Following a discussion on the list, the rules by which Exim recognises line
3631 endings on incoming messages have been changed. The -dropcr and drop_cr
3632 options are now no-ops, retained only for backwards compatibility. The
3633 following line terminators are recognized: LF CRLF CR. However, special
3634 processing applies to CR:
3635
3636 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3637 nor a local message in the state where . is a terminator.
3638
3639 (ii) If a bare CR is encountered in a header line, an extra space is added
3640 after the line terminator so as not to end the header. The reasoning
3641 behind this is that bare CRs in header lines are most likely either
3642 to be mistakes, or people trying to play silly games.
3643
364459. The size of a message, as listed by "-bp" or in the Exim monitor window,
3645 was being incorrectly given as 18 bytes larger than it should have been.
3646 This is a VOB (very old bug).
3647
364860. This may never have affected anything current, but just in case it has:
3649 When the local host is found other than at the start of a list of hosts,
3650 the local host, those with the same MX, and any that follow, are discarded.
3651 When the list in question was part of a longer list of hosts, the following
3652 hosts (not currently being processed) were also being discarded. This no
3653 longer happens. I'm not sure if this situation could ever has previously
3654 arisen.
3655
365661. Added the "/MX" feature to lists of hosts in the manualroute and query
3657 program routers.
3658
365962. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3660 header. This is something that is recommended in a new Internet Draft, and
3661 is something that is documented as being done by Sendmail. There are two
3662 possible values. For messages generated by the autoreply transport, Exim
3663 adds:
3664
3665 Auto-Submitted: auto-replied
3666
3667 whereas for all other generated messages (e.g. bounces) it adds
3668
3669 Auto-Submitted: auto-generated
3670
367163. The "personal" condition in filters now includes a test for the
3672 Auto-Submitted: header. If it contains the string "auto-" the message it
3673 not considered personal.
3674
367564. Added rcpt_include_affixes as a generic transport option.
3676
367765. Added queue_only_override (default true).
3678
367966. Added the syslog_duplication option.
3680
368167. If what should have been the first header line of a message consisted of
3682 a space followed by a colon, Exim was mis-interpreting it as a header line.
3683 It isn't of course - it is syntactically invalid and should therefore be
3684 treated as the start of the message body. The misbehaviour could have
3685 caused a number of strange effects, including loss of data in subsequent