Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
4a23603b 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.312 2006/02/22 15:10:28 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
11121d3d
JJ
198JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
199 a label. They prevented compilation on older perls.
200
201JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
202 a warning to be raised on newish perls.
203
204JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
205 on queue. Changes to match documented behaviour of showing count of
206 messages matching specified criteria.
207
8def5aaf
PH
208PH/40 Changed the default ident timeout from 30s to 5s.
209
929ba01c
PH
210PH/41 Added support for the use of login_cap features, on those BSD systems
211 that have them, for controlling the resources used by pipe deliveries.
212
2632889e
PH
213PH/42 The content-scanning code uses fopen() to create files in which to put
214 message data. Previously it was not paying any attention to the mode of
215 the files. Exim runs with umask(0) because the rest of the code creates
216 files with open(), and sets the required mode explicitly. Thus, these
217 files were ending up world-writeable. This was not a big issue, because,
218 being within the spool directory, they were not world-accessible. I have
219 created a function called modefopen, which takes an additional mode
220 argument. It sets umask(777), creates the file, chmods it to the required
221 mode, then resets the umask. All the relevant calls to fopen() in the
222 content scanning code have been changed to use this function.
223
944a9c55
PH
224PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
225 to 24 hours. This avoids potential overflow problems when processing G
226 and H retry rules. I suspect nobody ever tinkers with this value.
227
4a23603b
PH
228PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
229
214e2000 230
5de37277
PH
231Exim version 4.60
232-----------------
233
cc38ddbf
PH
234PH/01 Two changes to the default runtime configuration:
235
236 (1) Move the checks for relay_from_hosts and authenticated clients from
237 after to before the (commented out) DNS black list checks.
238
239 (2) Add control=submission to the relay_from_hosts and authenticated
240 clients checks, on the grounds that messages accepted by these
241 statements are most likely to be submissions.
5de37277 242
72fdd6ae
PH
243PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
244
245 (1) Generate an error if the third argument for the ${prvs expansion is
246 not a single digit.
247
248 (2) Treat a missing third argument of ${prvscheck as if it were an empty
249 string.
250
251 (3) Reset the variables that are obtained from the first argument of
252 ${prvscheck and used in the second argument before leaving the code,
253 because their memory is reclaimed, so using them afterwards may do
254 silly things.
255
256 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
257 one (it's much easier than Tom thought :-).
258
259 (5) Because of (4), we can now allow for the use of $prvscheck_result
260 inside the third argument.
cb9328de 261
cb741023
PH
262PH/03 For some reason, the default setting of PATH when running a command from
263 a pipe transport was just "/usr/bin". I have changed it to
264 "/bin:/usr/bin".
265
f174f16e
PH
266PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
267 anything to be listed in the output from -bV.
b2f5a032 268
c25242d7
PH
269PH/05 When a filter generated an autoreply, the entire To: header line was
270 quoted in the delivery log line, like this:
271
272 => >A.N.Other <ano@some.domain> <original@ddress> ...
273
274 This has been changed so that it extracts the operative address. There
275 may be more than one such address. If so, they are comma-separated, like
276 this:
277
278 => >ano@some.domain,ona@other.domain <original@ddress> ...
279
82c19f95
PH
280PH/06 When a client host used a correct literal IP address in a HELO or EHLO
281 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
282 not being looked up in the rDNS to get a host name, Exim was showing the
283 IP address twice in Received: lines, even though the IP addresses were
284 identical. For example:
285
286 Received: from [1.2.3.4] (helo=[1.2.3.4])
287
288 However, if the real host name was known, it was omitting the HELO data
289 if it matched the actual IP address. This has been tidied up so that it
290 doesn't show the same IP address twice.
291
d7ffbc12
PH
292PH/07 When both +timestamp and +memory debugging was on, the value given by
293 $tod_xxx expansions could be wrong, because the tod_stamp() function was
294 called by the debug printing, thereby overwriting the timestamp buffer.
295 Debugging no longer uses the tod_stamp() function when +timestamp is set.
296
9f526266
PH
297PH/08 When the original message was included in an autoreply transport, it
298 always said "this is a copy of the message, including all the headers",
299 even if body_only or headers_only was set. It now gives an appropriate
300 message.
301
87fcc8b9
PH
302PH/09 Applied a patch from the Sieve maintainer which:
303
304 o fixes some comments
305 o adds the (disabled) notify extension core
306 o adds some debug output for the result of if/elsif tests
307 o points to the current vacation draft in the documentation
308 and documents the missing references header update
309
310 and most important:
311
312 o fixes a bug in processing the envelope test (when testing
313 multiple envelope elements, the last element determinted the
314 result)
315
456682f5
PH
316PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
317 Electronic Mail") by including:
318
319 Auto-submitted: auto-generated
320
321 in the messages that it generates (bounce messages and others, such as
322 warnings). In the case of bounce messages for non-SMTP mesages, there was
323 also a typo: it was using "Auto_submitted" (underscore instead of
324 hyphen). Since every message generated by Exim is necessarily in response
325 to another message, thes have all been changed to:
326
327 Auto-Submitted: auto-replied
328
329 in accordance with these statements in the RFC:
330
331 The auto-replied keyword:
332
333 - SHOULD be used on messages sent in direct response to another
334 message by an automatic process,
335
336 - MUST NOT be used on manually-generated messages,
337
338 - MAY be used on Delivery Status Notifications (DSNs) and Message
339 Disposition Notifications (MDNs),
340
341 - MUST NOT be used on messages generated by automatic or periodic
342 processes, except for messages which are automatic responses to
343 other messages.
344
3e46c1aa
PH
345PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
346 to the default Received: header definition.
456682f5 347
49826d12
PH
348PH/12 Added log selector acl_warn_skipped (default on).
349
eba0c039
PH
350PH/13 After a successful wildlsearch lookup, discard the values of numeric
351 variables because (a) they are in the wrong storage pool and (b) even if
352 they were copied, it wouldn't work properly because of the caching.
353
a0d6ba8a
PH
354PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
355 checking when decoding. Apparently there are clients that generate
356 overlong encoded strings. Why am I not surprised?
357
f0917727
PH
358PH/15 If the first argument of "${if match_address" was not empty, but did not
359 contain an "@" character, Exim crashed. Now it writes a panic log message
360 and treats the condition as false.
361
096fee00
PH
362PH/16 In autoreply, treat an empty string for "once" the same as unset.
363
024bd3c2
PH
364PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
365 extension "envelope-auth". The code is finished and in agreement with
366 other implementations, but there is no documentation so far and in fact,
367 nobody wrote the draft yet. This extension is currently #undef'ed, thus
368 not changing the active code.
369
370 Print executed "if" and "elsif" statements when debugging is used. This
371 helps a great deal to understand what a filter does.
372
373 Document more things not specified clearly in RFC3028. I had all this
374 sorted out, when out of a sudden new issues came to my mind. Oops."
375
df199fec
PH
376PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
377 (Bugzilla #53).
378
d27f1df3
PH
379PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
380 canonical form (as documented). However, after a host name lookup from
381 the IP address, check_host() was doing a simple string comparison with
382 addresses acquired from the DNS when checking that the found name did
383 have the original IP as one of its addresses. Since any found IPv6
384 addresses are likely to be in abbreviated form, the comparison could
385 fail. Luckily, there already exists a function for doing the comparison
386 by converting both addresses to binary, so now that is used instead of
387 the text comparison.
388
96776534
PH
389PH/20 There was another similar case to PH/19, when a complete host name was
390 given in a host list; looking up its IP address could give an abbreviated
391 form, whereas the current host's name might or might not be abbreviated.
392 The same fix has been applied.
393
5de37277 394
9a799bc0
PH
395Exim version 4.54
396-----------------
397
398PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
399 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
400 It now does.
401
99a4b039
PH
402PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
403 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
404
405PH/03 Typo: missing ".o" in src/pcre/Makefile.
406
4b233853
PH
407PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
408 header line, restrict the check to what is listed in RFCs 2369 and 2929.
409 Also, for "Auto-Submitted", treat anything other than "no" as
410 non-personal, in accordance with RFC 3834. (Previously it treated
411 anything starting "auto-" as non-personal.)
412
8857ccfd
PH
413TF/01 The control=submission/name=... option had a problem with syntax
414 errors if the name included a slash character. The /name= option
415 now slurps the rest of the string, so it can include any characters
416 but it must come last in the list of options (after /sender_retain
417 or /domain=).
418
433a2980
PH
419PH/05 Some modifications to the interface to the fake nameserver for the new
420 testing suite.
421
3e46c1aa 422
9a799bc0 423
e3a311ba
TK
424Exim version 4.53
425-----------------
426
427TK/01 Added the "success_on_redirect" address verification option. See
428 NewStuff for rationale and an example.
429
13b685f9
PH
430PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
431
395ff96d
PH
432PH/02 Patch to exigrep to allow it to work on syslog lines.
433
5b68f6e4
PH
434PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
435 fread() to skip over the body file's header line, because in Cygwin the
436 header line is locked and is inaccessible.
437
1ab52c69
PH
438PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
439 co-exist for some time) to make it clear that it is the Exim ID that is
440 referenced, not the Message-ID: header line.
441
b07e6aa3
PH
442PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
443 string_format() function, because snprintf() does not exist on all
444 operating systems.
445
254e032f
PH
446PH/06 The use of forbid_filter_existstest now also locks out the use of the
447 ${stat: expansion item.
448
3af76a81
PH
449PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
450 protocol synchronization error", to keep the pedants happy.
451
2548ba04
PH
452PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
453 well as for IRIX systems, when gcc is being used. See the host.c source
454 file for comments.
455
b6c6011d
PH
456PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
457
cf39cf57
PH
458PH/10 Named domain lists were not working if used in a queue_smtp_domains
459 setting.
460
f1513293
PH
461PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
462 transport and to the smtp transport in LMTP mode.
463
727549a4
PH
464TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
465
af46795e
PH
466PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
467 run a filter in a subprocess. This could lead to confusion in subsequent
468 lookups in the parent process. There should also be a search_tidyup() at
469 the end of the subprocess.
470
d7b47fd0
PH
471PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
472 only if the host matched helo_try_verify_hosts, which caused the
473 verification to occur when the EHLO/HELO command was issued. The ACL just
474 tested the remembered result. Now, if a previous verification attempt has
475 not happened, "verify = helo" does it there and then.
476
ee744174
JJ
477JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
478
b582ab87
PH
479TK/03 Fix log output including CR from clamd.
480
41a13e0a
PH
481PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
482 exist provoked a memory error which could cause a segfault.
483
f625cc5a
PH
484PH/15 Installed PCRE 6.2
485
486PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
487
21f7af35
PH
488PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
489 of the problem. Specifically, suggested +O2 rather than +O1 for the
490 HP-UX compiler.
491
31480e42
PH
492PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
493
2d280592
PH
494PH/20 If a delivery was routed to a non-standard port by means of an SRV
495 record, the port was not correctly logged when the outgoing_port log
496 selector was set (it logged the transort's default port).
497
7cd1141b
PH
498PH/21 Added support for host-specific ports to manualroute, queryprogram,
499 fallback_hosts, and "hosts" in the smtp transport.
500
501PH/22 If the log selector "outgoing_port" is set, the port is now also given on
502 host errors such as "Connection refused".
503
750af86e
PH
504PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
505 authentication with radiusclient 0.4.9:
506
507 - Error returned from rc_read_config was caught wrongly
508 - Username/password not passed on to radius server due to wrong length.
509
510 The presumption is that some radiusclient API changes for 4.51/PH/17
511 were not taken care of correctly. The code is still untested by me (my
512 Linux distribution still has 0.3.2 of radiusclient), but it was
513 contributed by a Radius user.
514
515PH/24 When doing a callout, the value of $domain wasn't set correctly when
516 expanding the "port" option of the smtp transport.
517
4304270b
TK
518TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
519 while reading a MIME header. Thanks to Tom Hughes for a patch.
520
750af86e
PH
521PH/24 Include config.h inside local_scan.h so that configuration settings are
522 available.
523
64ffc24f
PH
524PH/25 Make $smtp_command_argument available after all SMTP commands. This means
525 that in an ACL for RCPT (for example), you can examine exactly what was
526 received.
527
5dd9625b
PH
528PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
529 commands, but it was not correctly comparing the address with the actual
530 client host address. Thus, it would show the EHLO address in Received:
531 header lines when this was not necessary.
532
5591031b
PH
533PH/27 Added the % operator to ${eval:}.
534
ba18e66a
PH
535PH/28 Exim tries to create and chdir to its spool directory when it starts;
536 it should be ignoring failures (because with -C, for example, it has lost
537 privilege). It wasn't ignoring creation failures other than "already
538 exists".
539
9cec981f
PH
540PH/29 Added "crypteq" to the list of supported features that Exim outputs when
541 -bV or -d is used.
542
aa2b5c79
PH
543PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
544 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
545 too many continuations, the temporary file was not being removed, and the
546 return code was incorrect.
aa2b5c79 547
48a53b7f
PH
548PH/31 Missing "BOOL" in function definition in filtertest.c.
549
1c59d63b
PH
550PH/32 Applied Sieve patches from the maintainer.
551
671012da
TK
552TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
553
1509d3a8
PH
554PH/33 Added "verify = not_blind".
555
556PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
557 Local/Makefile (with some defaults set). These are used in built scripts
558 such as exicyclog, but they have never been used in the exim_install
559 script (though there are many overriding facilities there). I have
560 arranged that the exim_install script now takes note of these two
561 settings.
562
563PH/35 Installed configuration files for Dragonfly.
564
2fe1a124
PH
565PH/36 When a locally submitted message by a trusted user did not contain a
566 From: header, and the sender address was obtained from -f or from an SMTP
567 MAIL command, and the trusted user did not use -F to supply a sender
568 name, $originator_name was incorrectly used when constructing a From:
569 header. Furthermore, $originator_name was used for submission mode
570 messages from external hosts without From: headers in a similar way,
571 which is clearly wrong.
572
8800895a
PH
573PH/37 Added control=suppress_local_fixups.
574
ccfdb010
PH
575PH/38 When log_selector = +received_sender was set, and the addition of the
576 sender made the log line's construction buffer exactly full, or one byte
577 less than full, an overflow happened when the terminating "\n" was
578 subsequently added.
579
1130bfb0
PH
580PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
581 when the result of a list match is failure because a DNS lookup failed.
582
ebcb507f
PH
583PH/40 RM_COMMAND is now used in the building process.
584
c35e155c
PH
585PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
586 the "build-* directories that it finds.
587
95d1f782
PH
588PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
589 address in a domain literal was a prefix of an interface address.
590
fd6de02e
PH
591PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
592 when verifying a sender address, unless rewrite_headers is false.
593
58de37c5
PH
594PH/44 Wrote a long comment about why errors_to addresses are verified as
595 recipients, not senders.
596
261cf466
TF
597TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
598 the ratelimit ACL was added.
599
3ee512ff
PH
600PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
601
e08c430f
PH
602PH/46 Added extra information about PostgreSQL errors to the error string.
603
bef5a11f
PH
604PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
605 avoiding the need to install special zones in a real server. This is
606 backwards compatible; if it can't find the fake resolver, it drops back.
607 Thus, both old and new test suites can be run.
608
7546de58
TF
609TF/02 Added util/ratelimit.pl
610
e5d5a95f
TF
611TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
612 clock is set back in time.
613
2e88a017
TF
614TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
615 Candler <B.Candler@pobox.com>.
616
a5f65aa4
TF
617TF/05 The fix for PH/43 was not completely correct; widen_domains is always
618 OK for addresses that are the result of redirections.
619
e7726cbf
PH
620PH/48 A number of further additions for the benefit of the new test suite,
621 including a fake gethostbyname() that interfaces to the fake DNS resolver
622 (see PH/47 above).
623
a7fdad5b
TF
624TF/06 The fix for widen_domains has also been applied to qualify_single and
625 search_parents which are the other dnslookup options that can cause
626 header rewrites.
627
6af56900
PH
628PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
629 type ("H").
630
0925ede6
PH
631PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
632
66afa403
TF
633TF/07 Exim produced the error message "an SRV record indicated no SMTP
634 service" if it encountered an MX record with an empty target hostname.
635 The message is now "an MX or SRV record indicated no SMTP service".
636
0154e85a
TF
637TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
638 if the DNS of the sending site is misconfigured. This is quite a
639 common situation. This change restores the behaviour of treating a
640 helo verification defer as a failure.
641
16f12c76
PH
642PH/51 If self=fail was set on a router, the bounce message did not include the
643 actual error message.
644
bbe902f0 645
e5a9dba6
PH
646Exim version 4.52
647-----------------
648
649TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
650
22c3b60b
PH
651PH/01 When a transport filter timed out in a pipe delivery, and the pipe
652 command itself ended in error, the underlying message about the transport
653 filter timeout was being overwritten with the pipe command error. Now the
654 underlying error message should be appended to the second error message.
655
06a9b4b5
PH
656TK/01 Fix poll() being unavailable on Mac OSX 10.2.
657
c1ac6996
PH
658PH/02 Reduce the amount of output that "make" produces by default. Full output
659 can still be requested.
660
9c7a242c
PH
661PH/03 The warning log line about a condition test deferring for a "warn" verb
662 was being output only once per connection, rather than after each
663 occurrence (because it was using the same function as for successful
664 "warn" verbs). This seems wrong, so I have changed it.
665
87ba3f5f
PH
666TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
667 it should not have, which might have caused a crash in the right
668 circumstances, but probably never did.
669
670PH/04 Installed a modified version of Tony Finch's patch to make submission
671 mode fix the return path as well as the Sender: header line, and to
672 add a /name= option so that you can make the user's friendly name appear
673 in the header line.
674
29aba418
TF
675TF/03 Added the control = fakedefer ACL modifier.
676
fe0dab11
TF
677TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
678 Mark Lowes for thorough testing.
870f6ba8 679
11d337a4
TK
680TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
681
682TK/03 Merged latest SRS patch from Miles Wilton.
683
415c8f3b
PH
684PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
685 with the definition in sysexits.h (which is #included earlier).
686 Fortunately, Exim does not actually use EX_OK. The code used to try to
687 preserve the sysexits.h value, by assumimg that macro definitions were
688 scanned for macro replacements. I have been disabused of this notion,
689 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 690
958541e9
PH
691PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
692 in the smtp transport. When a block could not be written in a single
693 write() function, the timeout was being re-applied to each part-write.
694 This seems wrong - if the receiver was accepting one byte at a time it
695 would take for ever. The timeout is now adjusted when this happens. It
696 doesn't have to be particularly precise.
697
c206415f
TK
698TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
699 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
700
2a4be8f9
PH
701PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
702 without a domain if the check to <postmaster@domain> fails.
703
1cba11c5
SC
704SC/01 Eximstats: added -xls and the ability to specify output files
705 (patch written by Frank Heydlauf).
706
707SC/02 Eximstats: use FileHandles for outputing results.
708
709SC/03 Eximstats: allow any combination of xls, txt, and html output.
710
711SC/04 Eximstats: fixed display of large numbers with -nvr option
712
713SC/05 Eximstats: fixed merging of reports with empty tables.
714
715SC/06 Eximstats: added the -include_original_destination flag
716
717SC/07 Eximstats: removed tabs and trailing whitespace.
718
1005d00e
TK
719TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
720
721TK/06 MBOX spool code: Add real "From " MBOX separator line
722 so the .eml file is really in mbox format (even though
723 most programs do not really care). Patch from Alex Miller.
724
725TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
726 The latter is generated from $received_to and is only set if the
727 message has one envelope recipient. SA can use these headers,
728 obviously out-of-the-box. Patch from Alex Miller.
729
9b4768fa
PH
730PH/08 The ${def test on a variable was returning false if the variable's
731 value was "0", contrary to what the specification has always said!
732 The result should be true unless the variable is empty.
733
734PH/09 The syntax error of a character other than { following "${if
735 def:variable_name" (after optional whitespace) was not being diagnosed.
736 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
737 accidental colon was present, for example, could give incorrect results.
738
0d7eb84a
PH
739PH/10 Tidied the code in a number of places where the st_size field of a stat()
740 result is used (not including appendfile, where other changes are about
741 to be made).
742
743PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
744 This involved changing a lot of size variables from int to off_t. It
745 should work with maildirs and everything.
746
40727bee
TK
747TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
748 spamd dying while we are connected to it.
749
554d2369
TF
750TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
751 <jgh@wizmail.org>
752
1f922db1
PH
753PH/12 Applied Alex Kiernan's patch for the API change for the error callback
754 function for BDB 4.3.
755
ef213c3b
PH
756PH/13 Changed auto_thaw such that it does not apply to bounce messages.
757
8ac170f3
PH
758PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
759 the sources for PCRE have been re-arranged and more files are now
760 involved.
761
b1c749bb
PH
762PH/15 The code I had for printing potentially long long variables in PH/11
763 above was not the best (it lost precision). The length of off_t variables
764 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
765 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
766 to be "long long int" or "long int". This is needed for the internal
767 formatting function string_vformat().
b1c749bb 768
4aac9b49
PH
769PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
770 the configuration file to be ":syslog", then the script "guesses" where
771 the logs files are, rather than using the compiled in default. In our
772 case the guess is not the same as the compiled default, so the script
773 suddenly stopped working when I started to use syslog. The patch checks
774 to see if log_file_path is "". If so, it attempts to read it from exim
775 with no configuration file to get the compiled in version, before it
776 falls back to the previous guessing code."
777
294520c8
TK
778TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
779 implementing BATV in an Exim configuration. See NewStuff for the gory
780 details.
781
5bd022fe
PH
782PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
783 Makefile that are specific to HP-UX.
784
90e9ce59
PH
785PH/18 If the "use_postmaster" option was set for a recipient callout together
786 with the "random" option, the postmaster address was used as the MAIL
787 FROM address for the random test, but not for the subsequent recipient
788 test. It is now used for both.
789
5ea81592
PH
790PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
791 patch removes a few documentation additions to RFC 3028, because the
792 latest draft now contains them. It adds the new en;ascii-case comparator
793 and a new error check for 8bit text in MIME parts. Comparator and
794 require names are now matched exactly. I enabled the subaddress
795 extension, but it is not well tested yet (read: it works for me)."
796
c6c2dc1d
PH
797PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
798 rework some of the code of TK/09 above to avoid the hardwired use of
799 "%lld" and "long long". Replaced the call to snprintf() with a call to
800 string_vformat().
801
fffffe4c
PH
802PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
803 records point to non-existent hosts", "retry timeout exceeded", and
804 "retry time not reached for any host after a long failure period".
ca02eafb 805
9a26b6b2
PH
806PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
807 experimental DomainKeys support:
808
809 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
810 (2) On an error such as an illegally used "control", the wrong name for
811 the control was given.
812
813 These problems did NOT occur unless DomainKeys support was compiled.
814
4aee0225
PH
815PH/23 Added daemon_startup_retries and daemon_startup_sleep.
816
32d668a5
PH
817PH/24 Added ${if match_ip condition.
818
8187c3f3
PH
819PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
820 databases so that it will be absolutely obvious if a crash occurs in the
821 DB library. This is a regular occurrence (often caused by mis-matched
822 db.h files).
823
ff790e47 824PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
825 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
826 picked up on a user's system that detects such things. There doesn't seem
827 to be a gcc warning option for this - only an attribute that has to be
828 put on the function's prototype. It seems that in Fedora Core 4 they have
829 set this on a number of new functions. No doubt there will be more in due
830 course.
ff790e47 831
5417f6d1
PH
832PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
833 specify a transport. However, if an address that was verified by such a
834 router was the subject of a callout, Exim crashed because it tried to
835 read the rcpt_include_affixes from the non-existent transport. Now it
836 just assumes that the setting of that option is false. This bug was
837 introduced by 4.51/PH/31.
838
59cf8544
PH
839PH/28 Changed -d+all to exclude +memory, because that information is very
840 rarely of interest, but it makes the output a lot bigger. People tend to
841 do -d+all out of habit.
842
e7ad8a65
PH
843PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
844 code in os-type was giving problems when libc.so lives in lib64, like on
845 x86_64 Fedora Core.
846
ade42478
PH
847PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
848 aren't the modern standard, and it seems that some systems' include files
849 don't always have them. Exim was already checking for some of the newer
850 ones like T_AAAA, and defining it itself. I've added checks for all the
851 record types that Exim uses.
852
182ad5cf
PH
853PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
854 not automatically generating a new one, as it is supposed to. This
855 prevented TLS from working. If the file did exist, but contained invalid
856 data, a new version was generated, as expected. It was only the case of a
857 non-existent file that was broken.
858
b0d9fc80
TK
859TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
860 with a change in libdomainkeys > 0.64.
861
862TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
863 from DNS. If the selector record carries the flag, it now has
864 precedence over the domain-wide flag.
865
866TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
867
47c7a64a
PH
868PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
869 the use of a port name that isn't defined in /etc/services) occurred, the
870 message was deferred as in a normal delivery, and thus remained on the
871 spool, instead of being failed because of the mua_wrapper setting. This
872 is now fixed, and I tidied up some of the mua_wrapper messages at the
873 same time.
874
a388bce4
SC
875SC/08 Eximstats: whilst parsing the mainlog(s), store information about
876 the messages in a hash of arrays rather than using individual hashes.
877 This is a bit cleaner and results in dramatic memory savings, albeit
878 at a slight CPU cost.
879
880SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
881 as requested by Marc Sherman.
882
883SC/10 Eximstats: added histograms for user specified patterns as requested
884 by Marc Sherman.
885
0793e4ed
SC
886SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
887
c58b88df
PH
888PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
889 fopen() in the content-scanning modules that did not already have it.
890
e7ad8a65 891
7982096b
PH
892Exim version 4.51
893-----------------
894
1a46a8c5
PH
895TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
896 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
897
2f079f46 898TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
899
900TK/03 Fix ACL "regex" condition not being available in MIME ACL.
901
902PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
903 to test Sieve filters that use "vacation".
904
905PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
906 that changes the way the GnuTLS parameters are stored in the cache file.
907 The new format can be generated externally. For backward compatibility,
908 if the data in the cache doesn't make sense, Exim assumes it has read an
909 old-format file, and it generates new data and writes a new file. This
910 means that you can't go back to an older release without removing the
911 file.
912
913PH/03 A redirect router that has both "unseen" and "one_time" set does not
914 work if there are any delivery delays because "one_time" forces the
915 parent to be marked "delivered", so its unseen clone is never tried
916 again. For this reason, Exim now forbids the simultaneous setting of
917 these two options.
918
919PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
920 redirected to themselves ("homonym" addresses). Read the long ChangeLog
921 entry if you want to know the details. The fix, however, neglected to
922 consider the case when local delivery batching is involved. The test for
923 "previously delivered" was not happening when checking to see if an
924 address could be batched with a previous (undelivered) one; under
925 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 926 address.
1a46a8c5
PH
927
928PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
929 in its include files, and this causes problems building Exim.
930
931PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
932 header_syntax) but Exim was just ignoring anything given after a slash.
933 In particular, this caused confusion with an attempt to use "verify =
934 reverse_host_lookup/defer_ok". An error is now given when options are
935 supplied for verify items that do not have them. (Maybe reverse_host_
936 lookup should have a defer_ok option, but that's a different point.)
937
938PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
939 defined by RFC 821) to 2048, because there were problems with some AUTH
940 commands, and RFC 1869 says the size should be increased for extended
941 SMTP commands that take arguments.
942
943PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
944 Finch).
945
946PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
947 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 948
49c2d5ea
PH
949PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
950 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
951 whatever). Otherwise the password may appear in the log. Following change
952 PH/42 below, there is no longer a chance of it appearing in a bounce
953 message.
49c2d5ea 954
bf759a8b
PH
955PH/11 Installed exipick version 20050225.0 from John Jetmore.
956
83364d30
PH
957PH/12 If the last host in a fallback_hosts list was multihomed, only the first
958 of its addresses was ever tried. (Bugzilla bug #2.)
959
7999bbd7
PH
960PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
961 the result incorrectly in the debug output. (It correctly added a newline
962 to what was transported.)
963
7dbf77c9
PH
964TF/01 Added $received_time.
965
74e0617f
PH
966PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
967 commented out examples of how to interface to a virus scanner and to
968 SpamAssassin. Also added commented examples of av_scanner and
969 spamd_address settings.
970
2f079f46
PH
971PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
972 and controls are allowed in which ACLs. There were a couple of minor
973 errors. Some of the entries in the conditions table (which is a table of
974 where they are NOT allowed) were getting very unwieldy; rewrote them as a
975 negation of where the condition IS allowed.
976
8c841523
PH
977PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
978
7766a4f0
PH
979PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
980 header file does not have a version number, so I've had to invent a new
981 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
982 API. The code is untested by me (my Linux distribution still has 0.3.2 of
983 radiusclient), but it was contributed by a Radius user.
984
8b417f2c
PH
985PH/18 Installed Lars Mainka's patch for the support of CRL collections in
986 files or directories, for OpenSSL.
987
901f42cb
PH
988PH/19 When an Exim process that is running as root has to create an Exim log
989 file, it does so in a subprocess that runs as exim:exim so as to get the
990 ownership right at creation (otherwise, other Exim processes might see
991 the file with the wrong ownership). There was no test for failure of this
992 fork() call, which would lead to the process getting stuck as it waited
993 for a non-existent subprocess. Forks do occasionally fail when resources
994 run out. I reviewed all the other calls to fork(); they all seem to check
995 for failure.
996
f9b9210e
PH
997PH/20 When checking for unexpected SMTP input at connect time (before writing
998 the banner), Exim was not dealing correctly with a non-positive return
999 from the read() function. If the client had disconnected by this time,
1000 the result was a log entry for a synchronization error with an empty
1001 string after "input=" when read() returned zero. If read() returned -1
1002 (an event I could not check), uninitialized data bytes were printed.
1003 There were reports of junk text (parts of files, etc) appearing after
1004 "input=".
1005
54cdb463
PH
1006PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1007
cf00dad6
PH
1008PH/22 Added support for macro redefinition, and (re)definition in between
1009 driver and ACL definitions.
1010
acb1b346
PH
1011PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1012 forgetting to use the resulting value; it was using the unexpanded value.
1013
c5ddb310
PH
1014PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1015 hadn't been configured. The fix is from Juergen Kreileder, who
1016 understands it better than I do:
1017
1018 "Here's what I see happening with three configured cyrus_sasl
1019 authenticators configured (plain, login, cram-md5):
1020
1021 On startup auth_cyrus_sasl_init() gets called for each of these.
1022 This means three calls to sasl_listmech() without a specified mech_list.
1023 => SASL tests which mechs of all available mechs actually work
1024 => three warnings about OTP not working
1025 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1026
1027 With the patch, sasl_listmech() also gets called three times. But now
1028 SASL's mech_list option is set to the server_mech specified in the the
1029 authenticator. Or in other words, the answer from sasl_listmech()
1030 gets limited to just the mech you're testing for (which is different
1031 for each call.)
1032 => the return list contains just 'plain' or 'login', 'cram-md5' or
1033 nothing depending on the value of ob->server_mech.
1034
1035 I've just tested the patch: Authentication still works fine,
1036 unavailable mechs specified in the exim configuration are still
1037 caught, and the auth.log warnings about OTP are gone."
1038
31619da6
PH
1039PH/25 When debugging is enabled, the contents of the command line are added
1040 to the debugging output, even when log_selector=+arguments is not
1041 specified.
1042
bebaf0fc
PH
1043PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1044 answer is "GNU", and only if the return is "GNU/something" is the answer
1045 "Linux".
1046
475fe28a
PH
1047PH/27 $acl_verify_message is now set immediately after the failure of a
1048 verification in an ACL, and so is available in subsequent modifiers. In
1049 particular, the message can be preserved by coding like this:
1050
1051 warn !verify = sender
1052 set acl_m0 = $acl_verify_message
1053
1054 Previously, $acl_verify_message was set only while expanding "message"
1055 and "log_message" when a very denied access.
1056
7e8bec7a
PH
1057PH/28 Modified OS/os.c-Linux with
1058
1059 -#ifndef OS_LOAD_AVERAGE
1060 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1061
1062 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1063 nomenclature these days.)
1064
e4a89c47
PH
1065PH/29 Installed patch from the Sieve maintainer that adds the options
1066 sieve_useraddress and sieve_subaddress to the redirect router.
1067
5ca2a9a1
PH
1068PH/30 In these circumstances:
1069 . Two addresses routed to the same list of hosts;
1070 . First host does not offer TLS;
1071 . First host accepts first address;
1072 . First host gives temporary error to second address;
1073 . Second host offers TLS and a TLS session is established;
1074 . Second host accepts second address.
1075 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1076 and peerdn, if requested) that were in fact used only for the second
1077 address.
7e8bec7a 1078
c688b954
PH
1079PH/31 When doing a callout as part of verifying an address, Exim was not paying
1080 attention to any local part prefix or suffix that was matched by the
1081 router that accepted the address. It now behaves in the same way as it
1082 does for delivery: the affixes are removed from the local part unless
1083 rcpt_include_affixes is set on the transport.
1084
fed77020
PH
1085PH/32 Add the sender address, as F=<...>, to the log line when logging a
1086 timeout during the DATA phase of an incoming message.
1087
7fe1560f
PH
1088PH/33 Sieve envelope tests were broken for match types other than :is. I have
1089 applied a patch sanctioned by the Sieve maintainer.
c688b954 1090
ebb6e6d5
PH
1091PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1092 the uid or gid is negative. A case of a negative gid caused this to be
1093 noticed. The fix allows for either to be negative.
1094
9c4e8f60
PH
1095PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1096 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1097 overlooked.
1098
1099PH/36 The change PH/12 above was broken. Fixed it.
1100
d7174846
PH
1101PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1102 the grounds that routing the same address twice would always produce the
1103 same answer. This might have been true once, but it is certainly no
1104 longer true now. Routing a child address may depend on the previous
1105 routing that produced that child. Some complicated redirection strategies
1106 went wrong when messages had multiple recipients, and made Exim's
1107 behaviour dependent on the order in which the addresses were given.
1108
1109 I have moved the duplicate checking until after the routing is complete.
1110 Exim scans the addresses that are assigned to local and remote
1111 transports, and removes any duplicates. This means that more work will be
1112 done, as duplicates will always all be routed, but duplicates are
1113 presumably rare, so I don't expect this is of any significance.
1114
1115 For deliveries to pipes, files, and autoreplies, the duplicate checking
1116 still happens during the routing process, since they are not going to be
1117 routed further.
1118
cfe75fc3
PH
1119PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1120 It corrects a timeout issue with spamd. This is Ian's comment: "The
1121 background is that sometimes spamd either never reads data from a
1122 connection it has accepted, or it never writes response data. The exiscan
1123 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1124 blindly assumes that writes won't block so it may never time out."
1125
be22d70e
PH
1126PH/39 Allow G after quota size as well as K and M.
1127
0612b098
PH
1128PH/40 The value set for $authenticated_id in an authenticator may not contain
1129 binary zeroes or newlines because the value is written to log lines and
1130 to spool files. There was no check on this. Now the value is run through
1131 the string_printing() function so that such characters are converted to
1132 printable escape sequences.
1133
2e0c1448
PH
1134PH/41 $message_linecount is a new variable that contains the total number of
1135 lines in the message. Compare $body_linecount, which is the count for the
1136 body only.
1137
447d236c
PH
1138PH/42 Exim no longer gives details of delivery errors for specific addresses in
1139 bounce and delay warning messages, except in certain special cases, which
1140 are as follows:
1141
1142 (a) An SMTP error message from a remote host;
1143 (b) A message specified in a :fail: redirection;
1144 (c) A message specified in a "fail" command in a system filter;
1145 (d) A message specified in a FAIL return from the queryprogram router;
1146 (e) A message specified by the cannot_route_message router option.
1147
1148 In these cases only, Exim does include the error details in bounce and
1149 warning messages. There are also a few cases where bland messages such
1150 as "unrouteable address" or "local delivery error" are given.
1151
d20976dc
PH
1152PH/43 $value is now also set for the "else" part of a ${run expansion.
1153
f656d135
PH
1154PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1155 being worked on, but at least Exim now implements the latest version to
1156 play with."
1157
2e2a30b4
PH
1158PH/45 In a pipe transport, although a timeout while waiting for the pipe
1159 process to complete was treated as a delivery failure, a timeout while
1160 writing the message to the pipe was logged, but erroneously treated as a
1161 successful delivery. Such timeouts include transport filter timeouts. For
1162 consistency with the overall process timeout, these timeouts are now
1163 treated as errors, giving rise to delivery failures by default. However,
1164 there is now a new Boolean option for the pipe transport called
1165 timeout_defer, which, if set TRUE, converts the failures into defers for
1166 both kinds of timeout. A transport filter timeout is now identified in
1167 the log output.
1168
9176e9f0
PH
1169PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1170 systems where "make" and "gmake" are different, calling "gmake" at top
1171 level broke things. I've arranged for the value of $(MAKE) to be passed
1172 from the Makefile to this script so that it can call the same version of
1173 "make".
1174
7982096b 1175
bbe902f0
PH
1176A note about Exim versions 4.44 and 4.50
1177----------------------------------------
1178
1179Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1180changes of various kinds. As a consequence, a big documentation update was
1181needed. This delayed the release for rather longer than seemed good, especially
1182in the light of a couple of (minor) security issues. Therefore, the changes
1183that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1184release. So 4.44 and 4.50 are in effect two different branches that both start
1185from 4.43.
1186
1187I have left the 4.50 change log unchanged; it contains all the changes since
11884.43. The change log for 4.44 is below; many of its items are identical to
1189those for 4.50. This seems to be the most sensible way to preserve the
1190historical information.
1191
1192
f7b63901 1193Exim version 4.50
495ae4b0
PH
1194-----------------
1195
5fe762f6
PH
1196 1. Minor wording change to the doc/README.SIEVE file.
1197
139059f6 1198 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1199 computation of the current number of files was incorrect.
495ae4b0 1200
7086e875
PH
1201 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1202 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1203 place.
1204
35af9f61
PH
1205 4. Give more explanation in the error message when the command for a transport
1206 filter fails to execute.
1207
b668c215
PH
1208 5. There are several places where Exim runs a non-Exim command in a
1209 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1210 was being done only for the command run by the queryprogram router. It is
1211 now done for all such subprocesses. The other cases are: ${run, transport
1212 filters, and the commands run by the lmtp and pipe transports.
1213
a494b1e1
PH
1214 6. Added CONFIGURE_GROUP build-time option.
1215
1216 7. Some older OS have a limit of 256 on the maximum number of file
1217 descriptors. Exim was using setrlimit() to set 1000 as a large value
1218 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1219 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1220
c5fcb476
PH
1221 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1222 was an oversight, and furthermore, ever since the addition of extra
1223 controls (e.g. 4.43/32), the checks on when to allow different forms of
1224 "control" were broken. There should now be diagnostics for all cases when a
1225 control that does not make sense is encountered.
1226
69358f02
PH
1227 9. Added the /retain_sender option to "control=submission".
1228
5be20824
PH
122910. $recipients is now available in the predata ACL (oversight).
1230
eb2c0248
PH
123111. Tidy the search cache before the fork to do a delivery from a message
1232 received from the command line. Otherwise the child will trigger a lookup
1233 failure and thereby defer the delivery if it tries to use (for example) a
1234 cached ldap connection that the parent has called unbind on.
1235
2a3eea10
PH
123612. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1237 of $address_data from the recipient verification was clobbered by the
1238 sender verification.
1239
124013. The value of address_data from a sender verification is now available in
1241 $sender_address_data in subsequent conditions in the ACL statement.
1242
23c7ff99
PH
124314. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1244
4deaf07d
PH
124515. Added a new option "connect=<time>" to callout options, to set a different
1246 connection timeout.
1247
926e1192
PH
124816. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1249 was its contents. (It was OK if the option was not defined at all.)
1250
650edc6f
PH
125117. A "Completed" log line is now written for messages that are removed from
1252 the spool by the -Mrm option.
1253
2c7db3f5
PH
125418. New variables $sender_verify_failure and $recipient_verify_failure contain
1255 information about exactly what failed.
1256
3d235903
PH
125719. Added -dd to debug only the daemon process.
1258
7c7ad977
PH
125920. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1260 handles timeouts, both on the server side and network timeouts. Renamed the
1261 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1262
981756db
PH
126321. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1264 It is now set to "smtps".
1265
d4eb88df
PH
126622. $host_address is now set to the target address during the checking of
1267 ignore_target_hosts.
1268
126923. When checking ignore_target_hosts for an ipliteral router, no host name was
1270 being passed; this would have caused $sender_host_name to have been used if
1271 matching the list had actually called for a host name (not very likely,
1272 since this list is usually IP addresses). A host name is now passed as
1273 "[x.x.x.x]".
1274
7d468ab8
PH
127524. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1276 code that specifies a non-restarting handler (typically sigaction() in
1277 modern systems) in an attempt to fix a rare and obscure crash bug.
1278
127925. Narrowed the window for a race in the daemon that could cause it to ignore
1280 SIGCHLD signals. This is not a major problem, because they are used only to
1281 wake it up if nothing else does.
1282
62c0818f
PH
128326. A malformed maildirsize file could cause Exim to calculate negative values
1284 for the mailbox size or file count. Odd effects could occur as a result.
1285 The maildirsize information is now recalculated if the size or filecount
1286 end up negative.
1287
26034054
PH
128827. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1289 support for a long time. Removed HAVE_SYS_VFS_H.
1290
af66f652
PH
129128. Installed the latest version of exipick from John Jetmore.
1292
90af77f4
PH
129329. In an address list, if the pattern was not a regular expression, an empty
1294 subject address (from a bounce message) matched only if the pattern was an
1295 empty string. Non-empty patterns were not even tested. This was the wrong
1296 because it is perfectly reasonable to use an empty address as part of a
1297 database query. An empty address is now tested by patterns that are
1298 lookups. However, all the other forms of pattern expect the subject to
1299 contain a local part and a domain, and therefore, for them, an empty
1300 address still always fails if the pattern is not itself empty.
1301
d8ef3577
PH
130230. Exim went into a mad DNS loop when attempting to do a callout where the
1303 host was specified on an smtp transport, and looking it up yielded more
1304 than one IP address.
1305
5cb8cbc6
PH
130631. Re-factored the code for checking spool and log partition space into a
1307 function that finds that data and another that does the check. The former
1308 is then used to implement four new variables: $spool_space, $log_space,
1309 $spool_inodes, and $log_inodes.
1310
14702f5b
PH
131132. The RFC2047 encoding function was originally intended for short strings
1312 such as real names; it was not keeping to the 75-character limit for
1313 encoded words that the RFC imposes. It now respects the limit, and
1314 generates multiple encoded words if necessary. To be on the safe side, I
1315 have increased the buffer size for the ${rfc2047: expansion operator from
1316 1024 to 2048 bytes.
1317
063b1e99
PH
131833. It is now permitted to omit both strings after an "if" condition; if the
1319 condition is true, the result is "true". As before, when the second string
1320 is omitted, a false condition yields an empty string. This makes it less
1321 cumbersome to write custom ACL and router conditions.
1322
652e1b65
PH
132334. Failure to deliver a bounce message always caused it to be frozen, even if
1324 there was an errors_to setting on the router. The errors_to setting is now
1325 respected.
1326
6f0c9a4f
PH
132735. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1328 canonical form (fully expanded) before being placed in
1329 $sender_host_address.
1330
33397d19
PH
133136. The table in the code that translates DNS record types into text (T_A to
1332 "A" for instance) was missing entries for NS and CNAME. It is just possible
1333 that this could have caused confusion if both these types were looked up
1334 for the same domain, because the text type is used as part of Exim's
1335 per-process caching. But the chance of anyone hitting this buglet seems
1336 very small.
1337
7bb56e1f
PH
133837. The dnsdb lookup has been extended in a number of ways.
1339
1340 (1) There is a new type, "zns", which walks up the domain tree until it
1341 finds some nameserver records. It should be used with care.
1342
ea3bc19b
PH
1343 (2) There is a new type, "mxh", which is like "mx" except that it returns
1344 just the host names, not the priorities.
1345
1346 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1347 looked up. The behaviour when one of the lookups defers can be
1348 controlled by a keyword.
7bb56e1f 1349
ea3bc19b 1350 (4) It is now possible to specify the separator character for use when
7bb56e1f 1351 multiple records are returned.
33397d19 1352
0bcb2a0e
PH
135338. The dnslists ACL condition has been extended: it is now possible to supply
1354 a list of IP addresses and/or domains to be looked up in a particular DNS
1355 domain.
1356
2ac0e484
PH
135739. Added log_selector=+queue_time_overall.
1358
4e1fde53
PH
135940. When running the queue in the test harness, wait just a tad after forking a
1360 delivery process, to get repeatability of debugging output.
1361
de365ded
PH
136241. Include certificate and key file names in error message when GnuTLS fails
1363 to set them up, because the GnuTLS error message doesn't include the name
1364 of the failing file when there is a problem reading it.
1365
f05da2e8
PH
136642. Allow both -bf and -bF in the same test run.
1367
d6453af2
PH
136843. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1369
f7b63901
PH
137044. The "Exiscan patch" is now merged into the mainline Exim source.
1371
137245. Sometimes the final signoff response after QUIT could fail to get
1373 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1374 < 0 before doing a fflush(). This bug looks as though it goes back to the
1375 introduction of TLS in release 3.20, but "sometimes" must have been rare
1376 because the tests only now provoked it.
1377
a444213a
PH
137846. Reset the locale to "C" after calling embedded Perl, in case it was changed
1379 (this can affect the format of dates).
1380
0ec020ea
PH
138147. exim_tidydb, when checking for the continued existence of a message for
1382 which it has found a message-specific retry record, was not finding
1383 messages that were in split spool directories. Consequently, it was
1384 deleting retry records that should have stayed in existence.
1385
b1206957
PH
138648. Steve fixed some bugs in eximstats.
1387
138849. The SPA authentication driver was not abandoning authentication and moving
1389 on to the next authenticator when an expansion was forced to fail,
1390 contradicting the general specification for all authenticators. Instead it
1391 was generating a temporary error. It now behaves as specified.
1392
26dd5a95
PH
139350. The default ordering of permitted cipher suites for GnuTLS was pessimal
1394 (the order specifies the preference for clients). The order is now AES256,
1395 AES128, 3DES, ARCFOUR128.
1396
343b2385
PH
139751. Small patch to Sieve code - explicitly set From: when generating an
1398 autoreply.
1399
1c5466b9
PH
140052. Exim crashed if a remote delivery caused a very long error message to be
1401 recorded - for instance if somebody sent an entire SpamAssassin report back
1402 as a large number of 550 error lines. This bug was coincidentally fixed by
1403 increasing the size of one of Exim's internal buffers (big_buffer) that
1404 happened as part of the Exiscan merge. However, to be on the safe side, I
1405 have made the code more robust (and fixed the comments that describe what
1406 is going on).
1407
55ee9ee3
PH
140853. Now that there can be additional text after "Completed" in log lines (if
1409 the queue_time_overall log selector is set), a one-byte patch to exigrep
1410 was needed to allow it to recognize "Completed" as not the last thing in
1411 the line.
1412
d38f8232
PH
141354. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1414 patch that reportedly fixes this has been added. I am not expert enough to
1415 create a test for it. This is what the patch creator wrote:
1416
1417 "I found a little strange behaviour of ldap code when working with
1418 Windows 2003 AD Domain, where users was placed in more than one
1419 Organization Units. When I tried to give exim partial DN, the exit code
1420 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1421 But simultaneously result of request was absolutely normal ldap result,
1422 so I produce this patch..."
1423
3295e65b
PH
1424 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1425 REFERENCE, so I have modified the code to exclude the patch when that macro
1426 is not defined.
1427
7102e136
PH
142855. Some experimental protocols are using DNS PTR records for new purposes. The
1429 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1430 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1431 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1432 CAN-2005-0021
7102e136 1433
3ca0ba97
PH
143456. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1435
c2bcbe20
PH
143657. Double the size of the debug message buffer (to 2048) so that more of very
1437 long debug lines gets shown.
1438
18ce445d
PH
143958. The exicyclog utility now does better if the number of log files to keep
1440 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1441
1f5b4c3d
PH
144259. Two changes related to the smtp_active_hostname option:
1443
1444 (1) $smtp_active_hostname is now available as a variable.
1445 (2) The default for smtp_banner uses $smtp_active_hostname instead
1446 of $primary_hostname.
1447
b975ba52
PH
144860. The host_aton() function is supposed to be passed a string that is known
1449 to be a valid IP address. However, in the case of IPv6 addresses, it was
1450 not checking this. This is a hostage to fortune. Exim now panics and dies
1451 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1452 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1453 components; fortuitously, this particular loophole had already been fixed
1454 by change 4.50/55 above.
1455
1456 If there are any other similar loopholes, the new check in host_aton()
1457 itself should stop them being exploited. The report I received stated that
1458 data on the command line could provoke the exploit when Exim was running as
1459 exim, but did not say which command line option was involved. All I could
1460 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1461 running as the user.
ea3a6f44 1462 CAN-2005-0021
85b87bc2
PH
1463
146461. There was a buffer overflow vulnerability in the SPA authentication code
1465 (which came originally from the Samba project). I have added a test to the
1466 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1467 CAN-2005-0022
b975ba52 1468
17ffcae7
PH
146962. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1470 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1471
d95f9fdb
PH
147263. The daemon start-up calls getloadavg() while still root for those OS that
1473 need the first call to be done as root, but it missed one case: when
1474 deliver_queue_load_max is set with deliver_drop_privilege. This is
1475 necessary for the benefit of the queue runner, because there is no re-exec
1476 when deliver_drop_privilege is set.
1477
86b8287f
PH
147864. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1479 This has been fixed.
1480
60dc5e56
PH
148165. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1482 was in use, was not putting the data itself into the right store pool;
1483 consequently, it could be overwritten for a subsequent message in the same
1484 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1485 the caching.)
1486
533244af
PH
148766. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1488
a5a28604
PH
148967. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1490 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1491 the function were treating the return as a boolean value, which happened to
1492 work because 0=false and not-0=true, but is not correct code.
1493
7e634d24
PH
149468. The host_aton() function was not handling scoped IPv6 addresses (those
1495 with, for example, "%eth0" on the end) correctly.
1496
3e11c26b
PH
149769. Fixed some compiler warnings in acl.c for the bitmaps specified with
1498 negated items (that is, ~something) in unsigned ints. Some compilers
1499 apparently mutter when there is no cast.
1500
6729cf78
PH
150170. If an address verification called from an ACL failed, and did not produce a
1502 user-specific message (i.e. there was only a "system" message), nothing was
1503 put in $acl_verify_message. In this situation, it now puts the system
1504 message there.
1505
00f00ca5
PH
150671. Change 4.23/11 added synchronization checking at the start of an SMTP
1507 session; change 4.31/43 added the unwanted input to the log line - except
1508 that it did not do this in the start of session case. It now does.
1509
c9bdd01c
PH
151072. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1511 This is wrong and can cause the other end to generate a synchronization
1512 error if it is another Exim or anything else that does the synchronization
1513 check. A QUIT command is no longer sent after a timeout.
1514
d43194df
PH
151573. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1516 during host lookups.
1517
fe5b5d0b
PH
151874. The defer_ok option of callout verification was not working if it was used
1519 when verifying addresses in header lines, that is, for this case:
1520
1521 verify = header_sender/callout=defer_ok
1522
76a2d7ba
PH
152375. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1524 those file descriptors could be used for SMTP connections. If anything
1525 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1526 could be sent to the SMTP client, causing chaos. The daemon now opens
1527 stdin, stdout, and stderr to /dev/null when it puts itself into the
1528 background.
1529
153076. Arrange for output from Perl's "warn" command to be written to Exim's main
1531 log by default. The user can override this with suitable Perl magic.
1532
04f7d5b9
PH
153377. The use of log_message on a "discard" ACL verb, which is supposed to add to
1534 the log message when discard triggers, was not working for the DATA ACL or
1535 for the non-SMTP ACL.
1536
bc60667e
PH
153778. Error message wording change in sieve.c.
1538
bb6e88ff
PH
153979. If smtp_accept_max_per_host was set, the number of connections could be
1540 restricted to fewer than expected, because the daemon was trying to set up
1541 a new connection before checking whether the processes handling previous
1542 connections had finished. The check for completed processes is now done
1543 earlier. On busy systems, this bug wouldn't be noticed because something
1544 else would have woken the daemon, and it would have reaped the completed
1545 process earlier.
1546
1e70f85b
PH
154780. If a message was submitted locally by a user whose login name contained one
1548 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1549 It caused a spool format error. I have fixed the spool reading code. A
1550 related problem was that the "from" clause in the Received: line became
1551 illegal because of the space(s). It is now covered by ${quote_local_part.
1552
155381. Included the latest eximstats from Steve (adds average sizes to HTML Top
1554 tables).
1555
4e01f9d6
PH
155682. Updated OS/Makefile-AIX as per message from Mike Meredith.
1557
1ee1cef2
PH
155883. Patch from Sieve maintainer to fix unterminated string problem in
1559 "vacation" handling.
1560
6e2b4ccc
PH
156184. Some minor changes to the Linux configuration files to help with other
1562 OS variants using glibc.
1563
8e669ac1
PH
156485. One more patch for Sieve to update vacation handling to latest spec.
1565
495ae4b0 1566
bbe902f0
PH
1567----------------------------------------------------
1568See the note above about the 4.44 and 4.50 releases.
1569----------------------------------------------------
1570
1571
1572Exim version 4.44
1573-----------------
1574
1575 1. Change 4.43/35 introduced a bug that caused file counts to be
1576 incorrectly computed when quota_filecount was set in an appendfile
1577 transport
1578
1579 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1580 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1581 place.
1582
1583 3. Give more explanation in the error message when the command for a transport
1584 filter fails to execute.
1585
1586 4. There are several places where Exim runs a non-Exim command in a
1587 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1588 was being done only for the command run by the queryprogram router. It is
1589 now done for all such subprocesses. The other cases are: ${run, transport
1590 filters, and the commands run by the lmtp and pipe transports.
1591
1592 5. Some older OS have a limit of 256 on the maximum number of file
1593 descriptors. Exim was using setrlimit() to set 1000 as a large value
1594 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1595 systems. I've change it so that if it can't get 1000, it tries for 256.
1596
1597 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1598 was an oversight, and furthermore, ever since the addition of extra
1599 controls (e.g. 4.43/32), the checks on when to allow different forms of
1600 "control" were broken. There should now be diagnostics for all cases when a
1601 control that does not make sense is encountered.
1602
1603 7. $recipients is now available in the predata ACL (oversight).
1604
1605 8. Tidy the search cache before the fork to do a delivery from a message
1606 received from the command line. Otherwise the child will trigger a lookup
1607 failure and thereby defer the delivery if it tries to use (for example) a
1608 cached ldap connection that the parent has called unbind on.
1609
1610 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1611 of $address_data from the recipient verification was clobbered by the
1612 sender verification.
1613
161410. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1615 was its contents. (It was OK if the option was not defined at all.)
1616
161711. A "Completed" log line is now written for messages that are removed from
1618 the spool by the -Mrm option.
1619
162012. $host_address is now set to the target address during the checking of
1621 ignore_target_hosts.
1622
162313. When checking ignore_target_hosts for an ipliteral router, no host name was
1624 being passed; this would have caused $sender_host_name to have been used if
1625 matching the list had actually called for a host name (not very likely,
1626 since this list is usually IP addresses). A host name is now passed as
1627 "[x.x.x.x]".
1628
162914. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1630 code that specifies a non-restarting handler (typically sigaction() in
1631 modern systems) in an attempt to fix a rare and obscure crash bug.
1632
163315. Narrowed the window for a race in the daemon that could cause it to ignore
1634 SIGCHLD signals. This is not a major problem, because they are used only to
1635 wake it up if nothing else does.
1636
163716. A malformed maildirsize file could cause Exim to calculate negative values
1638 for the mailbox size or file count. Odd effects could occur as a result.
1639 The maildirsize information is now recalculated if the size or filecount
1640 end up negative.
1641
164217. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1643 support for a long time. Removed HAVE_SYS_VFS_H.
1644
ea3a6f44 164518. Updated exipick to current release from John Jetmore.
bbe902f0
PH
1646
164719. Allow an empty sender to be matched against a lookup in an address list.
1648 Previously the only cases considered were a regular expression, or an
1649 empty pattern.
1650
165120. Exim went into a mad DNS lookup loop when doing a callout where the
1652 host was specified on the transport, if the DNS lookup yielded more than
1653 one IP address.
1654
ea3a6f44
NM
165521. The RFC2047 encoding function was originally intended for short strings
1656 such as real names; it was not keeping to the 75-character limit for
1657 encoded words that the RFC imposes. It now respects the limit, and
1658 generates multiple encoded words if necessary. To be on the safe side, I
1659 have increased the buffer size for the ${rfc2047: expansion operator from
1660 1024 to 2048 bytes.
bbe902f0 1661
ea3a6f44
NM
166222. Failure to deliver a bounce message always caused it to be frozen, even if
1663 there was an errors_to setting on the router. The errors_to setting is now
1664 respected.
bbe902f0
PH
1665
166623. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1667 canonical form (fully expanded) before being placed in
1668 $sender_host_address.
1669
167024. Updated eximstats to version 1.33
1671
ea3a6f44
NM
167225. Include certificate and key file names in error message when GnuTLS fails
1673 to set them up, because the GnuTLS error message doesn't include the name
1674 of the failing file when there is a problem reading it.
bbe902f0
PH
1675
167626. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1677 As per change 25.
bbe902f0 1678
ea3a6f44
NM
167927. Reset the locale to "C" after calling embedded Perl, in case it was changed
1680 (this can affect the format of dates).
bbe902f0 1681
ea3a6f44
NM
168228. exim_tidydb, when checking for the continued existence of a message for
1683 which it has found a message-specific retry record, was not finding
1684 messages that were in split spool directories. Consequently, it was
1685 deleting retry records that should have stayed in existence.
bbe902f0
PH
1686
168729. eximstats updated to version 1.35
1688 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1689 1.35 - bugfix such that pie charts by volume are generated correctly
1690
ea3a6f44
NM
169130. The SPA authentication driver was not abandoning authentication and moving
1692 on to the next authenticator when an expansion was forced to fail,
1693 contradicting the general specification for all authenticators. Instead it
1694 was generating a temporary error. It now behaves as specified.
bbe902f0 1695
ea3a6f44
NM
169631. The default ordering of permitted cipher suites for GnuTLS was pessimal
1697 (the order specifies the preference for clients). The order is now AES256,
1698 AES128, 3DES, ARCFOUR128.
bbe902f0 1699
ea3a6f44
NM
170031. Small patch to Sieve code - explicitly set From: when generating an
1701 autoreply.
bbe902f0 1702
ea3a6f44
NM
170332. Exim crashed if a remote delivery caused a very long error message to be
1704 recorded - for instance if somebody sent an entire SpamAssassin report back
1705 as a large number of 550 error lines. This bug was coincidentally fixed by
1706 increasing the size of one of Exim's internal buffers (big_buffer) that
1707 happened as part of the Exiscan merge. However, to be on the safe side, I
1708 have made the code more robust (and fixed the comments that describe what
1709 is going on).
bbe902f0 1710
ea3a6f44
NM
171133. Some experimental protocols are using DNS PTR records for new purposes. The
1712 keys for these records are domain names, not reversed IP addresses. The
1713 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1714 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
1715 CAN-2005-0021
1716
ea3a6f44
NM
171734. The host_aton() function is supposed to be passed a string that is known
1718 to be a valid IP address. However, in the case of IPv6 addresses, it was
1719 not checking this. This is a hostage to fortune. Exim now panics and dies
1720 if the condition is not met. A case was found where this could be provoked
1721 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1722 components; fortuitously, this particular loophole had already been fixed
1723 by change 4.50/55 or 4.44/33 above.
1724
1725 If there are any other similar loopholes, the new check in host_aton()
1726 itself should stop them being exploited. The report I received stated that
1727 data on the command line could provoke the exploit when Exim was running as
1728 exim, but did not say which command line option was involved. All I could
1729 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1730 running as the user.
bbe902f0
PH
1731 CAN-2005-0021
1732
ea3a6f44
NM
173335. There was a buffer overflow vulnerability in the SPA authentication code
1734 (which came originally from the Samba project). I have added a test to the
1735 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1736 CAN-2005-0022
1737
ea3a6f44
NM
173836. The daemon start-up calls getloadavg() while still root for those OS that
1739 need the first call to be done as root, but it missed one case: when
1740 deliver_queue_load_max is set with deliver_drop_privilege. This is
1741 necessary for the benefit of the queue runner, because there is no re-exec
1742 when deliver_drop_privilege is set.
bbe902f0 1743
ea3a6f44
NM
174437. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1745 was in use, was not putting the data itself into the right store pool;
1746 consequently, it could be overwritten for a subsequent message in the same
1747 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1748 the caching.)
bbe902f0 1749
ea3a6f44
NM
175038. Sometimes the final signoff response after QUIT could fail to get
1751 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1752 < 0 before doing a fflush(). This bug looks as though it goes back to the
1753 introduction of TLS in release 3.20, but "sometimes" must have been rare
1754 because the tests only now provoked it.
bbe902f0
PH
1755
1756
495ae4b0
PH
1757Exim version 4.43
1758-----------------
1759
1760 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1761 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1762 Now it returns an integer. A number of calls were still expecting a T/F
1763 return. Fortuitously, in all cases, the tests worked in OK situations,
1764 which is the norm. However, things would have gone wrong on any write
1765 failures on the smtp file descriptor. This function is used when sending
1766 messages over SMTP and also when doing verify callouts.
1767
1768 2. When Exim is called to do synchronous delivery of a locally submitted
1769 message (the -odf or -odi options), it no longer closes stderr before doing
1770 the delivery.
1771
1772 3. Implemented the mua_wrapper option.
1773
1774 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1775
1776 5. Implemented the functions header_remove(), header_testname(),
1777 header_add_at_position(), and receive_remove_recipient(), and exported them
1778 to local_scan().
1779
1780 6. If an ACL "warn" statement specified the addition of headers, Exim already
1781 inserted X-ACL-Warn: at the start if there was no header name. However, it
1782 was not making this test for the second and subsequent header lines if
1783 there were newlines in the string. This meant that an invalid header could
1784 be inserted if Exim was badly configured.
1785
1786 7. Allow an ACL "warn" statement to add header lines at the start or after all
1787 the Received: headers, as well as at the end.
1788
1789 8. Added the rcpt_4xx retry error code.
1790
1791 9. Added postmaster_mailfrom=xxx to callout verification option.
1792
179310. Added mailfrom=xxxx to the callout verification option, for verify=
1794 header_sender only.
1795
179611. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1797 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1798
179912. Inserted some casts to stop certain compilers warning when using pointer
1800 differences as field lengths or precisions in printf-type calls (mostly
1801 affecting debugging statements).
1802
180313. Added optional readline() support for -be (dynamically loaded).
1804
180514. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1806 same clock tick as a message's arrival, so that its received time was the
1807 same as the "first fail" time on the retry record, and that message
1808 remained on the queue past the ultimate address timeout, every queue runner
1809 would try a delivery (because it was past the ultimate address timeout) but
1810 after another failure, the ultimate address timeout, which should have then
1811 bounced the address, did not kick in. This was a "< instead of <=" error;
1812 in most cases the first failure would have been in the next clock tick
1813 after the received time, and all would be well.
1814
181515. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1816 being recognized when the domain list was tested by the match_domain
1817 condition in an expansion string.
1818
181916. Added the ${str2b64: operator.
1820
182117. Exim was always calling setrlimit() to set a large limit for the number of
1822 processes, without checking whether the existing limit was already
1823 adequate. (It did check for the limit on file descriptors.) Furthermore,
1824 errors from getrlimit() and setrlimit() were being ignored. Now they are
1825 logged to the main and panic logs, but Exim does carry on, to try to do its
1826 job under whatever limits there are.
1827
182818. Imported PCRE 5.0.
1829
183019. Trivial typo in log message " temporarily refused connection" (the leading
1831 space).
1832
183320. If the log selector return_path_on_delivery was set and an address was
1834 redirected to /dev/null, the delivery process crashed because it assumed
1835 that a return path would always be set for a "successful" delivery. In this
1836 case, the whole delivery is bypassed as an optimization, and therefore no
1837 return path is set.
1838
183921. Internal re-arrangement: the function for sending a challenge and reading
1840 a response while authentication was assuming a zero-terminated challenge
1841 string. It's now changed to take a pointer and a length, to allow for
1842 binary data in such strings.
1843
184422. Added the cyrus_sasl authenticator (code supplied by MBM).
1845
184623. Exim was not respecting finduser_retries when seeking the login of the
1847 uid under which it was called; it was always trying 10 times. (The default
1848 setting of finduser_retries is zero.) Also, it was sleeping after the final
1849 failure, which is pointless.
1850
185124. Implemented tls_on_connect_ports.
1852
185325. Implemented acl_smtp_predata.
1854
185526. If the domain in control=submission is set empty, Exim assumes that the
1856 authenticated id is a complete email address when it generates From: or
1857 Sender: header lines.
1858
185927. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1860 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1861 chown and chgrp in /bin and hostname in /usr/bin.
1862
186328. Exim was keeping the "process log" file open after each use, just as it
1864 does for the main log. This opens the possibility of it remaining open for
1865 long periods when the USR1 signal hits a daemon. Occasional processlog
1866 errors were reported, that could have been caused by this. Anyway, it seems
1867 much more sensible not to leave this file open at all, so that is what now
1868 happens.
1869
187029. The long-running daemon process does not normally write to the log once it
1871 has entered its main loop, and it closes the log before doing so. This is
1872 so that log files can straightforwardly be renamed and moved. However,
1873 there are a couple of unusual error situations where the daemon does write
1874 log entries, and I had neglected to close the log afterwards.
1875
187630. The text of an SMTP error response that was received during a remote
1877 delivery was being truncated at 512 bytes. This is too short for some of
1878 the long messages that one sometimes sees. I've increased the limit to
1879 1024.
1880
188131. It is now possible to make retry rules that apply only when a message has a
1882 specific sender, in particular, an empty sender.
1883
188432. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1885 it possible to be selective about when SMTP synchronization is enforced.
1886
188733. Added "control = caseful_local_part" and "control = "caselower_local_part".
1888
188932. Implemented hosts_connection_nolog.
1890
189133. Added an ACL for QUIT.
1892
189334. Setting "delay_warning=" to disable warnings was not working; it gave a
1894 syntax error.
1895
189635. Added mailbox_size and mailbox_filecount to appendfile.
1897
189836. Added control = no_multiline_responses to ACLs.
1899
190037. There was a bug in the logic of the code that waits for the clock to tick
1901 in the case where the clock went backwards by a substantial amount such
1902 that the microsecond fraction of "now" was more than the microsecond
1903 fraction of "then" (but the whole seconds number was less).
1904
190538. Added support for the libradius Radius client library this is found on
1906 FreeBSD (previously only the radiusclient library was supported).
1907
1908
1909Exim version 4.42
1910-----------------
1911
1912 1. When certain lookups returned multiple values in the form name=value, the
1913 quoting of the values was not always being done properly. Specifically:
1914 (a) If the value started with a double quote, but contained no whitespace,
1915 it was not quoted.
1916 (b) If the value contained whitespace other than a space character (i.e.
1917 tabs or newlines or carriage returns) it was not quoted.
1918 This fix has been applied to the mysql and pgsql lookups by writing a
1919 separate quoting function and calling it from the lookup code. The fix
1920 should probably also be applied to nisplus, ibase and oracle lookups, but
1921 since I cannot test any of those, I have not disturbed their existing code.
1922
1923 2. A hit in the callout cache for a specific address caused a log line with no
1924 reason for rejecting RCPT. Now it says "Previous (cached) callout
1925 verification failure".
1926
1927 3. There was an off-by-one bug in the queryprogram router. An over-long
1928 return line was truncated at 256 instead of 255 characters, thereby
1929 overflowing its buffer with the terminating zero. As well as fixing this, I
1930 have increased the buffer size to 1024 (and made a note to document this).
1931
1932 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1933 when Exim is waiting for an SMTP response from a remote server, Exim
1934 restarts its select() call on the socket, thereby resetting its timeout.
1935 This is not a problem when such interrupts are rare. Somebody set up a cron
1936 job to run exiwhat every 2 minutes, which is less than the normal select()
1937 timeout (5 or 10 minutes). This meant that the select() timeout never
1938 kicked in because it was always reset. I have fixed this by comparing the
1939 time when an interrupt arrives with the time at the start of the first call
1940 to select(). If more time than the timeout has elapsed, the interrupt is
1941 treated as a timeout.
1942
1943 5. Some internal re-factoring in preparation for the addition of Sieve
1944 extensions (by MH). In particular, the "personal" test is moved to a
1945 separate function, and given an option for scanning Cc: and Bcc: (which is
1946 not set for Exim filters).
1947
1948 6. When Exim created an email address using the login of the caller as the
1949 local part (e.g. when creating a From: or Sender: header line), it was not
1950 quoting the local part when it contained special characters such as @.
1951
1952 7. Installed new OpenBSD configuration files.
1953
1954 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1955 try to make them clearer.
1956
1957 9. Callout options, other than the timeout value, were being ignored when
1958 verifying sender addresses in header lines. For example, when using
1959
1960 verify = header_sender/callout=no_cache
1961
1962 the cache was (incorrectly) being used.
1963
196410. Added a missing instance of ${EXE} to the exim_install script; this affects
1965 only the Cygwin environment.
1966
196711. When return_path_on_delivery was set as a log selector, if different remote
1968 addresses in the same message used different return paths and parallel
1969 remote delivery occurred, the wrong values would sometimes be logged.
1970 (Whenever a remote delivery process finished, the return path value from
1971 the most recently started remote delivery process was logged.)
1972
197312. RFC 3848 specifies standard names for the "with" phrase in Received: header
1974 lines when AUTH and/or TLS are in use. This is the "received protocol"
1975 field. Exim used to use "asmtp" for authenticated SMTP, without any
1976 indication (in the protocol name) for TLS use. Now it follows the RFC and
1977 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1978 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1979 names appear in log lines as well as in Received: header lines.
1980
198113. Installed MH's patches for Sieve to add the "copy" and "vacation"
1982 extensions, and comparison tests, and to fix some bugs.
1983
198414. Changes to the "personal" filter test:
1985
1986 (1) The test was buggy in that it was just doing the equivalent of
1987 "contains" tests on header lines. For example, if a user's address was
1988 anne@some.where, the "personal" test would incorrectly be true for
1989
1990 To: susanne@some.where
1991
1992 This test is now done by extracting each address from the header in turn,
1993 and checking the entire address. Other tests that are part of "personal"
1994 are now done using regular expressions (for example, to check local parts
1995 of addresses in From: header lines).
1996
1997 (2) The list of non-personal local parts in From: addresses has been
1998 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1999 taken from the Sieve specification recommendations.
2000
2001 (3) If the message contains any header line starting with "List-" it is
2002 treated as non-personal.
2003
2004 (4) The test for "circular" in the Subject: header line has been removed
2005 because it now seems ill-conceived.
2006
200715. Minor typos in src/EDITME comments corrected.
2008
200916. Installed latest exipick from John Jetmore.
2010
201117. If headers_add on a router specified a text string that was too long for
2012 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2013 of string_sprintf() is now avoided.
2014
201518. $message_body_size was not set (it was always zero) when running the DATA
2016 ACL and the local_scan() function.
2017
201819. For the "mail" command in an Exim filter, no default was being set for
2019 the once_repeat time, causing a random time value to be used if "once" was
2020 specified. (If the value happened to be <= 0, no repeat happened.) The
2021 default is now 0s, meaning "never repeat". The "vacation" command was OK
2022 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2023 (I found it when inspecting the code).
2024
202520. There is now an overall timeout for performing a callout verification. It
2026 defaults to 4 times the callout timeout, which applies to individual SMTP
2027 commands during the callout. The overall timeout applies when there is more
2028 than one host that can be tried. The timeout is checked before trying the
2029 next host. This prevents very long delays if there are a large number of
2030 hosts and all are timing out (e.g. when the network connections are timing
2031 out). The value of the overall timeout can be changed by specifying an
2032 additional sub-option for "callout", called "maxwait". For example:
2033
2034 verify = sender/callout=5s,maxwait=20s
2035
203621. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2037 to the end before writing, but this should make it even safer).
2038
203922. Exim was forgetting that it had advertised PIPELINING for the second and
2040 subsequent messages on an SMTP connection. It was also not resetting its
2041 memory on STARTTLS and an internal HELO.
2042
204323. When Exim logs an SMTP synchronization error within a session, it now
2044 records whether PIPELINING has been advertised or not.
2045
204624. Added 3 instances of "(long int)" casts to time_t variables that were being
2047 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2048 rather than long int.
2049
205025. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2051
205226. Added the never_mail option to autoreply.
2053
2054
2055Exim version 4.41
2056-----------------
2057
2058 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2059 crash if the getsockname() call failed; this can happen if a connection is
2060 closed very soon after it is established. The problem was simply in the
2061 order in which certain operations were done, causing Exim to try to write
2062 to the SMTP stream before it had set up the file descriptor. The bug has
2063 been fixed by making things happen in the correct order.
2064
2065
2066Exim version 4.40
2067-----------------
2068
2069 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2070 before the connection was closed, thus losing the rejection response.
2071
2072 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2073 some early Solaris releases, but causes trouble in current releases where
2074 socklen_t is defined.
2075
2076 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2077 always exist.
2078
2079 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2080 configured.
2081
2082 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2083 that releasing the top of it at the end releases what was used for sub-
2084 expansions (unless the block got too big). However, discard this block if
2085 the first thing is a variable or header, so that we can use its block when
2086 it is dynamic (useful for very large $message_headers, for example).
2087
2088 6. Lookups now cache *every* query, not just the most recent. A new, separate
2089 store pool is used for this. It can be recovered when all lookup caches are
2090 flushed. Lookups now release memory at the end of their result strings.
2091 This has involved some general refactoring of the lookup sources.
2092
2093 7. Some code has been added to the store_xxx() functions to reduce the amount
2094 of flapping under certain conditions.
2095
2096 8. log_incoming_interface used to affect only the <= reception log lines. Now
2097 it causes the local interface and port to be added to several more SMTP log
2098 lines, for example "SMTP connection from", and rejection lines.
2099
2100 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2101
210210. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2103
210411. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2105 could be overwritten at the end of the current message (or the start of a
2106 new message if it was set in a HELO ACL). The value is now preserved for
2107 the duration of the SMTP connection.
2108
210912. If a transport had a headers_rewrite setting, and a matching header line
2110 contained an unqualified address, that address was qualified, even if it
2111 did not match any rewriting rules. The underlying bug was that the values
2112 of the flags that permit the existence of unqualified sender and recipient
2113 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2114 non-local messages, and by -bnq for local messages) were not being
2115 preserved with the message after it was received.
2116
211713. When Exim was logging an SMTP synchronization error, it could sometimes log
2118 "next input=" as part of the text comprising the host identity instead of
2119 the correct text. The code was using the same buffer for two different
2120 strings. However, depending on which order the printing function evaluated
2121 its arguments, the bug did not always show up. Under Linux, for example, my
2122 test suite worked just fine.
2123
212414. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2125 doesn't work with some older versions of Perl. It has been changed to "my",
2126 which in any case is probably the better facility to use.
2127
212815. A really picky compiler found some instances of statements for creating
2129 error messages that either had too many or two few arguments for the format
2130 string.
2131
213216. The size of the buffer for calls to the DNS resolver has been increased
2133 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2134 for addresses that have a lot of PTR records. This alleviates a problem; it
2135 does not fully solve it.
2136
213717. A dnsdb lookup for PTR records that receives more data than will fit in the
2138 buffer now truncates the list and logs the incident, which is the same
2139 action as happens when Exim is looking up a host name and its aliases.
2140 Previously in this situation something unpredictable would happen;
2141 sometimes it was "internal error: store_reset failed".
2142
214318. If a server dropped the connection unexpectedly when an Exim client was
2144 using GnuTLS and trying to read a response, the client delivery process
2145 crashed while trying to generate an error log message.
2146
214719. If a "warn" verb in an ACL added multiple headers to a message in a single
2148 string, for example:
2149
2150 warn message = H1: something\nH2: something
2151
2152 the text was added as a single header line from Exim's point of view
2153 though it ended up OK in the delivered message. However, searching for the
2154 second and subsequent header lines using $h_h2: did not work. This has been
2155 fixed. Similarly, if a system filter added multiple headers in this way,
2156 the routers could not see them.
2157
215820. Expanded the error message when iplsearch is called with an invalid key to
2159 suggest using net-iplsearch in a host list.
2160
216121. When running tests using -bh, any delays imposed by "delay" modifiers in
2162 ACLs are no longer actually imposed (and a message to that effect is
2163 output).
2164
216522. If a "gecos" field in a passwd entry contained escaped characters, in
2166 particular, if it contained a \" sequence, Exim got it wrong when building
2167 a From: or a Sender: header from that name. A second bug also caused
2168 incorrect handling when an unquoted " was present following a character
2169 that needed quoting.
2170
217123. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2172 was not being matched caselessly.
2173
217424. Arranged for all hyphens in the exim.8 source to be escaped with
2175 backslashes.
2176
217725. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2178 properly. Recipient callout cache records were still being keyed to include
2179 the sender, even when use_sender was set false. This led to far more
2180 callouts that were necessary. The sender is no longer included in the key
2181 when use_sender is false.
2182
218326. Added "control = submission" modifier to ACLs.
2184
218527. Added the ${base62d: operator to decode base 62 numbers.
2186
218728. dnsdb lookups can now access SRV records.
2188
218929. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2190 the configuration file.
2191
219230. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2193 (-v) mode. This makes the output for a verbose queue run more intelligible.
2194
219531. Added a use_postmaster feature to recipient callouts.
2196
219732. Added the $body_zerocount variable, containing the number of binary zero
2198 bytes in the message body.
2199
220033. The time of last modification of the "new" subdirectory is now used as the
2201 "mailbox time last read" when there is a quota error for a maildir
2202 delivery.
2203
220434. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2205
220635. Added +ignore_unknown as a special item in host lists.
2207
220836. Code for decoding IPv6 addresses in host lists is now included, even if
2209 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2210 address was recognized as an IP address, but was then not correctly decoded
2211 into binary, causing unexpected and incorrect effects when compared with
2212 another IP address.
2213
2214
2215Exim version 4.34
2216-----------------
2217
2218 1. Very minor rewording of debugging text in manualroute to say "list of
2219 hosts" instead of "hostlist".
2220
2221 2. If verify=header_syntax was set, and a header line with an unqualified
2222 address (no domain) and a large number of spaces between the end of the
2223 name and the colon was received, the reception process suffered a buffer
2224 overflow, and (when I tested it) crashed. This was caused by some obsolete
2225 code that should have been removed. The fix is to remove it!
2226
2227 3. When running in the test harness, delay a bit after writing a bounce
2228 message to get a bit more predictability in the log output.
2229
2230 4. Added a call to search_tidyup() just before forking a reception process. In
2231 theory, someone could use a lookup in the expansion of smtp_accept_max_
2232 per_host which, without the tidyup, could leave open a database connection.
2233
2234 5. Added the variables $recipient_data and $sender_data which get set from a
2235 lookup success in an ACL "recipients" or "senders" condition, or a router
2236 "senders" option, similar to $domain_data and $local_part_data.
2237
2238 6. Moved the writing of debug_print from before to after the "senders" test
2239 for routers.
2240
2241 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2242 problems for message scanning, either using a data ACL, or using
2243 local_scan() because the Received: header was not generated till after they
2244 were called (in order to set the time as the time of reception completion).
2245 I have revised the way this works. The header is now generated after the
2246 body is received, but before the ACL or local_scan() are called. After they
2247 are run, the timestamp in the header is updated.
2248
2249
2250Exim version 4.33
2251-----------------
2252
2253 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2254 before starting a queue runner without re-exec. This happened only when
2255 deliver_drop_privilege was set or when the Exim user was set to root. The
2256 effect of the bug was that timeouts during subsequent deliveries caused
2257 crashes instead of being properly handled. The handler is now left at its
2258 default (and expected) setting.
2259
2260 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2261 message, again when deliver_drop_privilege is set or Exim is run as root.
2262 The bug described in (1) was not present in this case, but the tidying up
2263 of the other signals was missing. I have made the two cases consistent.
2264
2265 3. The ignore_target_hosts setting on a manualroute router was being ignored
2266 for hosts that were looked up using the /MX notation.
2267
2268 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2269 in domain lists.
2270
2271 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2272 operated on the sender address. After changing the $sender_address to <>
2273 for the sender address verify, Exim was re-instated it as the original
2274 (before rewriting) address, but remembering that it had rewritten it, so it
2275 wasn't rewriting it again. This bug also had the effect of breaking the
2276 sender address verification caching when the sender address was rewritten.
2277
2278 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2279 This has been changed so that if the ip literal address matches
2280 ignore_target_hosts, the router declines.
2281
2282 7. Added expansion conditions match_domain, match_address, and match_local_
2283 part (NOT match_host).
2284
2285 8. The placeholder for the Received: header didn't have a length field set.
2286
2287 9. Added code to Exim itself and to exim_lock to test for a specific race
2288 condition that could lead to file corruption when using MBX delivery. The
2289 issue is with the lockfile that is created in /tmp. If this file is removed
2290 after a process has opened it but before that process has acquired a lock,
2291 there is the potential for a second process to recreate the file and also
2292 acquire a lock. This could lead to two Exim processes writing to the file
2293 at the same time. The added code performs the same test as UW imapd; it
2294 checks after acquiring the lock that its file descriptor still refers to
2295 the same named file.
2296
229710. The buffer for building added header lines was of fixed size, 8192 bytes.
2298 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2299 when Exim is built.
2300
230111. Added the smtp_active_hostname option. If used, this will typically be made
2302 to depend on the incoming interface address. Because $interface_address is
2303 not set up until the daemon has forked a reception process, error responses
2304 that can happen earlier (such as "too many connections") no longer contain
2305 a host name.
2306
230712. If an expansion in a condition on a "warn" statement fails because a lookup
2308 defers, the "warn" statement is abandoned, and the next ACL statement is
2309 processed. Previously this caused the whole ACL to be aborted.
2310
231113. Added the iplsearch lookup type.
2312
231314. Added ident_timeout as a log selector.
2314
231515. Added tls_certificate_verified as a log selector.
2316
231716. Added a global option tls_require_ciphers (compare the smtp transport
2318 option of the same name). This controls incoming TLS connections.
2319
232017. I finally figured out how to make tls_require_ciphers do a similar thing
2321 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2322 before starting the TLS session.
2323
232418. Tabs are now shown as \t in -bP output.
2325
232619. If the log selector return_path_on_delivery was set, Exim crashed when
2327 bouncing a message because it had too many Received: header lines.
2328
232920. If two routers both had headers_remove settings, and the first one included
2330 a superfluous trailing colon, the final name in the first list and the
2331 first name in the second list were incorrectly joined into one item (with a
2332 colon in the middle).
2333
2334
2335Exim version 4.32
2336-----------------
2337
2338 1. Added -C and -D options to the exinext utility, mainly to make it easier
2339 to include in the automated testing, but these could be helpful when
2340 multiple configurations are in use.
2341
2342 2. The exinext utility was not formatting the output nicely when there was
2343 an alternate port involved in the retry record key, nor when there was a
2344 message id as well (for retries that were specific to a specific message
2345 and a specific host). It was also confused by IPv6 addresses, because of
2346 the additional colons they contain. I have fixed the IPv4 problem, and
2347 patched it up to do a reasonable job for IPv6.
2348
2349 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2350 delivery, the log line now contains "pipelined" if PIPELINING was used.
2351
2352 4. An SMTP transport process used to panic and die if the bind() call to set
2353 an explicit outgoing interface failed. This has been changed; it is now
2354 treated in the same way as a connect() failure.
2355
2356 5. A reference to $sender_host_name in the part of a conditional expansion
2357 that was being skipped was still causing a DNS lookup. This no longer
2358 occurs.
2359
2360 6. The def: expansion condition was not recognizing references to header lines
2361 that used bh_ and bheader_.
2362
2363 7. Added the _cache feature to named lists.
2364
2365 8. The code for checking quota_filecount in the appendfile transport was
2366 allowing one more file than it should have been.
2367
2368 9. For compatibility with Sendmail, the command line option
2369
2370 -prval:sval
2371
2372 is equivalent to
2373
2374 -oMr rval -oMs sval
2375
2376 and sets the incoming protocol and host name (for trusted callers). The
2377 host name and its colon can be omitted when only the protocol is to be set.
2378 Note the Exim already has two private options, -pd and -ps, that refer to
2379 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2380 "s", but I don't think that's a major issue.
2381
238210. A number of refactoring changes to the code, none of which should affect
2383 Exim's behaviour:
2384
2385 (a) The number of logging options was getting close to filling up the
2386 32-bit word that was used as a bit map. I have split them into two classes:
2387 those that are passed in the argument to log_write(), and those that are
2388 only ever tested independently outside of that function. These are now in
2389 separate 32-bit words, so there is plenty of room for expansion again.
2390 There is no change in the user interface or the logging behaviour.
2391
2392 (b) When building, for example, log lines, the code previously used a
2393 macro that called string_cat() twice, in order to add two strings. This is
2394 not really sufficiently general. Furthermore, there was one instance where
2395 it was actually wrong because one of the argument was used twice, and in
2396 one call a function was used. (As it happened, calling the function twice
2397 did not affect the overall behaviour.) The macro has been replaced by a
2398 function that can join an arbitrary number of extra strings onto a growing
2399 string.
2400
2401 (c) The code for expansion conditions now uses a table and a binary chop
2402 instead of a serial search (which was left over from when there were very
2403 few conditions). Also, it now recognizes conditions like "pam" even when
2404 the relevant support is not compiled in: a suitably worded error message is
2405 given if an attempt is made to use such a condition.
2406
240711. Added ${time_interval:xxxxx}.
2408
240912. A bug was causing one of the ddress fields not to be passed back correctly
2410 from remote delivery subprocesses. The field in question was not being
2411 subsequently used, so this caused to problems in practice.
2412
241313. Added new log selectors queue_time and deliver_time.
2414
241514. Might have fixed a bug in maildirsizefile handling that threw up
2416 "unexpected character" debug warnings, and recalculated the data
2417 unnecessarily. In any case, I expanded the warning message to give more
2418 information.
2419
242015. Added the message "Restricted characters in address" to the statements in
2421 the default ACL that block characters like @ and % in local parts.
2422
242316. Change 71 for release 4.31 proved to be much less benign that I imagined.
2424 Three changes have been made:
2425
2426 (a) There was a serious bug; a negative response to MAIL caused the whole
2427 recipient domain to be cached as invalid, thereby blocking all messages
2428 to all local parts at the same domain, from all senders. This bug has
2429 been fixed. The domain is no longer cached after a negative response to
2430 MAIL if the sender used is not empty.
2431
2432 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2433 been restored.
2434
2435 (c) A new callout option, "use_sender" has been added for people who want
2436 the modified behaviour.
2437
2438
2439Exim version 4.31
2440-----------------
2441
2442 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2443 Larry Rosenman.
2444
2445 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2446 indeed breaks things for older releases.
2447
2448 3. Added additional logging to the case where there is a problem reading data
2449 from a filter that is running in a subprocess using a pipe, in order to
2450 try to track down a specific problem.
2451
2452 4. Testing facility fudge: when running in the test harness and attempting
2453 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2454 getting "No route to host". Convert this to a timeout.
2455
2456 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2457 warning.
2458
2459 6. Some OS don't have socklen_t but use size_t instead. This affects the
2460 fifth argument of getsockopt() amongst other things. This is now
2461 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2462 can be set for individual OS. I have set it for SunOS5, OSF1, and
2463 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2464 some earlier ones do not.
2465
2466 7. Change 4.30/15 was not doing the test caselessly.
2467
2468 8. The standard form for an IPv6 address literal was being rejected by address
2469 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2470 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2471 this, as well as the form without the "IPv6" on the front (but only when
2472 address literals are enabled, of course).
2473
2474 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2475
247610. Exim crashed if a message with an empty sender address specified by -f
2477 encountered a router with an errors_to setting. This could be provoked only
2478 by a command such as
2479
2480 exim -f "" ...
2481
2482 where an empty string was supplied; "<>" did not hit this bug.
2483
248411. Installed PCRE release 4.5.
2485
248612. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2487 remained set. It is now erased.
2488
248913. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2490 times from message ids (which are base 36 rather than the normal 62).
2491
249214. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2493 were being counted as actual protocol errors, and logged if the log
2494 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2495 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2496 and DATA following a set of rejected RCPTs do not count as protocol errors.
2497 In other words, Exim assumes they were pipelined, though this may not
2498 actually be the case. Of course, in all cases the client gets an
2499 appropriate error code.
2500
250115. If a lookup fails in an ACL condition, a message about the failure may
2502 be available; it is used if testing the ACL cannot continue, because most
2503 such messages specify what the cause of the deferral is. However, some
2504 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2505 that caused an old message to be retained and used if a later statement
2506 caused a defer, replacing the real cause of the deferral.
2507
250816. If an IP address had so many PTR records that the DNS lookup buffer
2509 was not large enough to hold them, Exim could crash while trying to process
2510 the truncated data. It now detects and logs this case.
2511
251217. Further to 4.21/58, another change has been made: if (and only if) the
2513 first line of a message (the first header line) ends with CRLF, a bare LF
2514 in a subsequent header line has a space inserted after it, so as not to
2515 terminate the header.
2516
251718. Refactoring: tidied an ugly bit of code in appendfile that copied data
2518 unnecessarily, used atoi() instead of strtol(), and didn't check the
2519 termination when getting file sizes from file names by regex.
2520
252119. Completely re-implemented the support for maildirsize files, in the light
2522 of a number of problems with the previous contributed implementation
2523 (4.30/29). In particular:
2524
2525 . If the quota is zero, the maildirsize file is maintained, but no quota is
2526 imposed.
2527
2528 . If the maildir directory does not exist, it is created before any attempt
2529 to write a maildirsize file.
2530
2531 . The quota value in the file is just a cache; if the quota is changed in
2532 the transport, the new value overrides.
2533
2534 . A regular expression is available for excluding directories from the
2535 count.
2536
253720. The autoreply transport checks the characters in options that define the
2538 message's headers; it allows continued headers, but it was checking with
2539 isspace() after an embedded newline instead of explicitly looking for a
2540 space or a tab.
2541
254221. If all the "regular" hosts to which an address was routed had passed their
2543 expiry times, and had not reached their retry times, the address was
2544 bounced, even if fallback hosts were defined. Now Exim should go on to try
2545 the fallback hosts.
2546
254722. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2548 equivalent code in the SMTP transport. Some hosts send humungous responses
2549 to HELO/EHLO, more than 1024 it seems.
2550
255123. Refactoring: code in filter.c used (void *) for "any old type" but this
2552 gives compiler warnings in some environments. I've now done it "properly",
2553 using a union.
2554
255524. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2556 (because of problems with the built-in one) was declared to return uschar *
2557 instead of char *, causing compiler failure.
2558
255925. Fixed a file descriptor leak when processing alias/forward files.
2560
256126. Fixed a minor format string issue in dbfn.c.
2562
256327. Typo in exim.c: ("dmbnz" for "dbmnz").
2564
256528. If a filter file refered to $h_xxx or $message_headers, and the headers
2566 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2567 become corrupted.
2568
256929. When a sender address is verified, it is cached, to save repeating the test
2570 when there is more than one recipient in a message. However, when the
2571 verification involves a callout, it is possible for different callout
2572 options to be set for different recipients. It is too complicated to keep
2573 track of this in the cache, so now Exim always runs a verification when a
2574 callout is required, relying on the callout cache for the optimization.
2575 The overhead is duplication of the address routing, but this should not be
2576 too great.
2577
257830. Fixed a bug in callout caching. If a RCPT command caused the sender address
2579 to be verified with callout=postmaster, and the main callout worked but the
2580 postmaster check failed, the verification correctly failed. However, if a
2581 subsequent RCPT command asked for sender verification *without* the
2582 postmaster check, incorrect caching caused this verification also to fail,
2583 incorrectly.
2584
258531. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2586 it was not caching the DNS options (qualify_single, search_parents) that
2587 were used when the lookup failed. A subsequent lookup with different
2588 options therefore always gave the same answer, though there were cases
2589 where it should not have. (Example: a "domains = !$mx_any" option on a
2590 dnslookup router: the "domains" option is always processed without any
2591 widening, but the router might have qualify_single set.) Now Exim uses the
2592 cached value only when the same options are set.
2593
259432. Added John Jetmore's "exipick" utility to the distribution.
2595
259633. GnuTLS: When an attempt to start a TLS session fails for any reason other
2597 than a timeout (e.g. a certificate is required, and is not provided), an
2598 Exim server now closes the connection immediately. Previously it waited for
2599 the client to close - but if the client is SSL, it seems that they each
2600 wait for each other, leading to a delay before one of them times out.
2601
260234: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2603 maintained 0.8.x compatibility because I don't think many are using it, and
2604 it is clearly obsolete.
2605
260635. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2607 transport.
2608
260936. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2610 client certificate was expired. A simple patch fixes this, though I don't
2611 understand the full logic of why the verify callback is called multiple
2612 times.
2613
261437. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2615 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2616 which causes problems with some clients (such as the Certicom SSL Plus
2617 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2618 disables the coutermeasure allowing Eudora to connect."
2619
262038. Exim was not checking that a write() to a log file succeeded. This could
2621 lead to Bad Things if a log got too big, in particular if it hit a file
2622 size limit. Exim now panics and dies if it cannot write to a log file, just
2623 as it does if it cannot open a log file.
2624
262539. Modified OS/Makefile-Linux so that it now contains
2626
2627 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2628
2629 The two -D definitions ensure that Exim is compiled with large file
2630 support, which makes it possible to handle log files that are bigger than
2631 2^31.
2632
263340. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2634 instance) a domain was checked against a named list that involved a lookup,
2635 causing $domain_data to be set, then another domain was checked against the
2636 same list, then the first domain was re-checked, the value of $domain_data
2637 after the final check could be wrong. In particular, if the second check
2638 failed, it could be set empty. This bug probably also applied to
2639 $localpart_data.
2640
264141. The strip_trailing_dot option was not being applied to the address given
2642 with the -f command-line option.
2643
264442. The code for reading a message's header from the spool was incrementing
2645 $received_count, but never initializing it. This meant that the value was
2646 incorrect (doubled) while delivering a message in the same process in which
2647 it was received. In the most common configuration of Exim, this never
2648 happens - a fresh exec is done - but it can happen when
2649 deliver_drop_privilege is set.
2650
265143. When Exim logs an SMTP synchronization error - client data sent too soon -
2652 it now includes up to 150 characters of the unexpected data in the log
2653 line.
2654
265544. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2656 and building data strings. The size of both of these buffers was 10 000
2657 bytes - far larger than anybody would *ever* want, thought I. Needless to
2658 say, somebody hit the limit. I have increased the maximum line length to
2659 20 000 and the maximum data length of concatenated lines to 100 000. I have
2660 also fixed two bugs, because there was no checking on these buffers. Tsk,
2661 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2662 buffer is too small.
2663
266445. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2665 lsearch lookups. Now it does.
2666
266746. When parsing a route_list item in a manualroute router, a fixed-length
2668 buffer was used for the list of hosts. I made this 1024 bytes long,
2669 thinking that nobody would ever have a list of hosts that long. Wrong.
2670 Somebody had a whole pile of complicated expansion conditions, and the
2671 string was silently truncated, leading to an expansion error. It turns out
2672 that it is easier to change to an unlimited length (owing to other changes
2673 that have happened since this code was originally written) than to build
2674 structure for giving a limitation error. The length of the item that
2675 expands into the list of hosts is now unlimited.
2676
267747. The lsearch lookup could not handle data where the length of text line was
2678 more than 4095 characters. Such lines were truncated, leading to shortened
2679 data being returned. It should now handle lines of any length.
2680
268148. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2682 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2683 ACL").
2684
268549. Cosmetic tidy to scripts like exicyclog that are generated by globally
2686 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2687 no longer happens in comment lines. A list of replacements is now placed
2688 at the head of all of the source files, except those whose only change is
2689 to replace PERL_COMMAND in the very first #! line.
2690
269150. Replaced the slow insertion sort in queue.c, for sorting the list of
2692 messages on the queue, with a bottom-up merge sort, using code contributed
2693 by Michael Haardt. This should make operations like -bp somewhat faster on
2694 large queues. It won't affect queue runners, except when queue_run_in_order
2695 is set.
2696
269751. Installed eximstats 1.31 in the distribution.
2698
269952. Added support for SRV lookups to the dnslookup router.
2700
270153. If an ACL referred to $message_body or $message_body_end, the value was not
2702 reset for any messages that followed in the same SMTP session.
2703
270454. The store-handling optimization for building very long strings was not
2705 differentiating between the different store pools. I don't think this
2706 actually made any difference in practice, but I've tidied it.
2707
270855. While running the routers to verify a sender address, $sender_address
2709 was still set to the sender address. This is wrong, because when routing to
2710 send a bounce to the sender, it would be empty. Therefore, I have changed
2711 it so that, while verifying a sender address, $sender_address is set to <>.
2712 (There is no change to what happens when verifying a recipient address.)
2713
271456. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2715 target A or AAAA records (if not already returned) without resetting the
2716 qualify_single or search_parents options of the DNS resolver. These are
2717 inappropriate in this case because the targets of MX and SRV records must
2718 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2719 target that, when qualified, matched something in the local domain. These
2720 two options are now turned off when doing these lookups.
2721
272257. It seems that at least some releases of Reiserfs (which does not have the
2723 concept of a fixed number of inodes) returns zero and not -1 for the
2724 number of available inodes. This interacted badly with check_spool_inodes,
2725 which assumed that -1 was the "no such thing" setting. What I have done is
2726 to check that the total number of inodes is greater than zero before doing
2727 the test of how many are available.
2728
272958. When a "warn" ACL statement has a log_message modifier, the message is
2730 remembered, and not repeated. This is to avoid a lot of repetition when a
2731 message has many recipients that cause the same warning to be written.
2732 Howewer, Exim was preserving the list of already written lines for an
2733 entire SMTP session, which doesn't seem right. The memory is now reset if a
2734 new message is started.
2735
273659. The "rewrite" debugging flag was not showing the result of rewriting in the
2737 debugging output unless log_rewrite was also set.
2738
273960. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2740 of (int)(handle) when we know that handle contains (void *)(-1).
2741
274261. The Exim daemon panic-logs an error return when it closes the incoming
2743 connection. However "connection reset by peer" seems to be common, and
2744 isn't really an error worthy of noting specially, so that particular error
2745 is no long logged.
2746
274762. When Exim is trying to find all the local interfaces, it used to panic and
2748 die if the ioctl to get the interface flags failed. However, it seems that
2749 on at least one OS (Solaris 9) it is possible to have an interface that is
2750 included in the list of interfaces, but for which you get a failure error
2751 for this call. This happens when the interface is not "plumbed" into a
2752 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2753 failure of the "get flags" call assumes that the interface is down.
2754
275563. Added a ${eval10: operator, which assumes all numbers are decimal. This
2756 makes life easier for people who are doing arithmetic on fields extracted
2757 from dates, where you often get leading zeros that should not be
2758 interpreted as octal.
2759
276064. Added qualify_domain to the redirect router, to override the global
2761 setting.
2762
276365. If a pathologically long header line contained very many addresses (the
2764 report of this problem mentioned 10 000) and each of them was rewritten,
2765 Exim could use up a very large amount of memory. (It kept on making new
2766 copies of the header line as it rewrote, and never released the old ones.)
2767 At the expense of a bit more processing, the header rewriting function has
2768 been changed so that it no longer eats memory in this way.
2769
277066. The generation of the Received: header has been moved from the time that a
2771 message starts to be received, to the time that it finishes. The timestamp
2772 in the Received: header should now be very close to that of the <= log
2773 line. There are two side-effects of this change:
2774
2775 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2776 logged header lines no longer include the local Received: line, because
2777 it has not yet been created. The same applies to a copy of the message
2778 that is returned to a non-SMTP sender when a message is rejected.
2779
2780 (b) When a filter file is tested using -bf, no additional Received: header
2781 is added to the test message. After some thought, I decided that this
2782 is a bug fix.
2783
2784 This change does not affect the value of $received_for. It is still set
2785 after address rewriting, but before local_scan() is called.
2786
278767. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2788
278968. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2790 gave an unhelpful panic error message, and a defer error. I have managed to
2791 change this behaviour so that it now rejects any supplied certificate,
2792 which seems right, as the list of acceptable certificates is empty.
2793
279469. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2795 gave an unhelpful defer error. I have not managed to make this reject any
2796 supplied certificates, but the error message it gives is "no certificate
2797 supplied", which is not helpful.
2798
279970. exigrep's output now also includes lines that are not associated with any
2800 message, but which match the given pattern. Implemented by a patch from
2801 Martin Sluka, which also tidied up the Perl a bit.
2802
280371. Recipient callout verification, like sender verification, was using <> in
2804 the MAIL FROM command. This isn't really the right thing, since the actual
2805 sender may affect whether the remote host accepts the recipient or not. I
2806 have changed it to use the actual sender in the callout; this means that
2807 the cache record is now keyed on a recipient/sender pair, not just the
2808 recipient address. There doesn't seem to be a real danger of callout loops,
2809 since a callout by the remote host to check the sender would use <>.
2810 [SEE ABOVE: changed after hitting problems.]
2811
281272. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2813 temporary errors. However, in the case of such a code being given after
2814 the end of a data transmission (i.e. after ".") Exim was failing to write
2815 a retry record for the message. (Yes, there was some broken host that was
2816 actually sending 8xx at this point.)
2817
281873. An unknown lookup type in a host list could cause Exim to panic-die when
2819 the list was checked. (An example that provoked this was putting <; in the
2820 middle of a list instead of at the start.) If this happened during a DATA
2821 ACL check, a -D file could be left lying around. This kind of configuration
2822 error no longer causes Exim to die; instead it causes a defer errror. The
2823 incident is still logged to the main and panic logs.
2824
282574. Buglet left over from Exim 3 conversion. The message "too many messages
2826 in one connection" was written to the rejectlog but not the mainlog, except
2827 when address rewriting (yes!) was being logged.
2828
282975. Added write_rejectlog option.
2830
283176. When a system filter was run not as root (that is, when system_filter_user
2832 was set), the values of the $n variables were not being returned to the
2833 main process; thus, they were not subsequently available in the $sn
2834 variables.
2835
283677. Added +return_path_on_delivery log selector.
2837
283878. A connection timeout was being treated differently from recipients deferred
2839 when testing hosts_max_try with a message that was older than the host's
2840 retry timeout. (The host should not be counted, thus allowing all hosts to
2841 be tried at least once before bouncing.) This may have been the cause of an
2842 occasionally reported bug whereby a message would remain on the queue
2843 longer than the retry timeout, but would be bounced if a delivery was
2844 forced. I say "may" because I never totally pinned down the problem;
2845 setting up timeout/retry tests is difficult. See also the next item.
2846
284779. The ultimate address timeout was not being applied to errors that involved
2848 a combination of host plus message (for example, a timeout on a MAIL
2849 command). When an address resolved to a number of possible hosts, and they
2850 were not all tried for each delivery (e.g. because of hosts_max_try), a
2851 message could remain on the queue longer than the retry timeout.
2852
285380. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2854 Haardt.
2855
285681. Fixed an obscure SMTP outgoing bug which required at least the following
2857 conditions: (a) there was another message waiting for the same server;
2858 (b) the server returned 5xx to all RCPT commands in the first message so
2859 that the message was not completed; (c) the server dropped the connection
2860 or gave a negative response to the RSET that Exim sends to abort the
2861 transaction. The observed case was a dropped connection after DATA that had
2862 been sent in pipelining mode. That is, the server had advertised PIPELINING
2863 but was not implementing it correctly. The effect of the bug was incorrect
2864 behaviour, such as trying another host, and this could lead to a crash.
2865
2866
2867Exim version 4.30
2868-----------------
2869
2870 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2871 and daemon.c were passed as pointers to ints; they should have been
2872 pointers to socklen_t variables (which are typically unsigned ints).
2873
2874 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2875 fixed.
2876
2877 3. Fixed a really odd bug that affected only the testing scheme; patching a
2878 certain fixed string in the binary changed the value of another string that
2879 happened to be identical to the end of the original first string.
2880
2881 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2882 name", it returns that address as the IP address. On some operating
2883 systems (e.g. Solaris), it also passes back the IP address string as the
2884 "host name". However, on others (e.g. Linux), it passes back an empty
2885 string. Exim wasn't checking for this, and was changing the host name to an
2886 empty string, assuming it had been canonicized.
2887
2888 5. Although rare, it is permitted to have more than one PTR record for a given
2889 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2890 all the names associated with an address, because they do in Solaris.
2891 However, it seems that they do not in Linux for data that comes from the
2892 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2893 I found this out when I moved to a new Linux workstation and tried to run
2894 the Exim test suite.
2895
2896 To get round this problem I have changed the code so that it now does its
2897 own call to the DNS to look up PTR records when searching for a host name.
2898 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2899 addresses that are only in /etc/hosts are still found.
2900
2901 This behaviour is, however, controlled by an option called host_lookup_
2902 order, which defaults to "bydns:byaddr". If people want to use the other
2903 order, or indeed, just use one or the other means of lookup, they can
2904 specify it in this variable.
2905
2906 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2907 some operating systems, this comes back from gethostbyaddr() as an empty
2908 string, and this is what Exim used to test for. However, it seems that in
2909 other systems, "." is yielded. Exim now tests for this case too.
2910
2911 7. The values of check_spool_space and check_log_space are now held internally
2912 as a number of kilobytes instead of an absolute number of bytes. If a
2913 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2914 kilobyte. This means that much larger values can be stored.
2915
2916 8. Exim monitor: an attempt to get the action menu when not actually pointing
2917 at a message produces an empty menu entitled "No message selected". This
2918 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2919 no entries in it ("Shell widget menu has zero width and/or height"). So I
2920 have added a single, blank menu entry in this case.
2921
2922 9. Added ${quote_local_part.
2923
292410. MIME decoding is now applied to the contents of Subject: header lines when
2925 they are logged.
2926
292711. Now that a reference to $sender_host_address automatically causes a reverse
2928 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2929 host lookup before query-style lookups in lists that might use this
2930 variable. This has therefore been abolished, and the "net-" prefix is no
2931 longer necessary for query-style lookups.
2932
293312. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2934 have been a typo for LFLAGS, so it has been changed.
2935
293613. The install script calls Exim with "-C /dev/null" in order to find the
2937 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2938 to be output. Howeve, since Exim outputs its version number before the
2939 error, it didn't break the script. It just looked ugly. I fixed this by
2940 always allowing "-C /dev/null" if the caller is root.
2941
294214. Ignore overlarge ACL variable number when reading spool file - insurance
2943 against a later release with more variables having written the file.
2944
294515. The standard form for an IPv6 address literal was being rejected by EHLO.
2946 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2947 this, as well as the form without the "IPv6" on the front.
2948
294916. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2950 OS/Makefile-Darwin file.
2951
295217. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2953 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2954
295518. After change 4.21/52, "%ld" was used to format the contents of the $inode
2956 variable. However, some OS use ints for inodes. I've added cast to long int
2957 to get rid of the compiler warning.
2958
295919. I had forgotten to lock out "/../" in configuration file names when
2960 ALT_CONFIG_PREFIX was set.
2961
296220. Routers used for verification do not need to specify transports. However,
2963 if such a router generated a host list, and callout was configured, Exim
2964 crashed, because it could not find a port number from the (non-existent)
2965 transport. It now assumes port 25 in this circumstance.
2966
296721. Added the -t option to exigrep.
2968
296922. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2970 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2971 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2972 them is set, LOOKUP_LSEARCH is forced.
2973
297423. "exim -bV" now outputs a list of lookups that are included in the binary.
2975
297624. Added sender and host information to the "rejected by local_scan()" log
2977 line; previously there was no indication of these.
2978
297925. Added .include_if_exists.
2980
298126. Change 3.952/11 added an explicit directory sync on top of a file sync for
2982 Linux. It turns out that not all file systems support this. Apparently some
2983 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2984 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2985 not supported on the file descriptor, is now ignored when Exim is trying to
2986 sync a directory. This applies only to Linux.
2987
298827. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2989
299028. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2991 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2992 this.
2993
299429. Added support for maildirsize files from supplied patch (modified a bit).
2995
299630. The use of :fail: followed by an empty string could lead Exim to respond to
2997 sender verification failures with (e.g.):
2998
2999 550 Verification failed for <xxx>
3000 550 Sender verify failed
3001
3002 where the first response line was missing the '-' that indicates it is not
3003 the final line of the response.
3004
300531. The loop for finding the name of the user that called Exim had a hardwired
3006 limit of 10; it now uses the value of finduser_retries, which is used for
3007 all other user lookups.
3008
300932. Added $received_count variable, available in data and not_smtp ACLs, and at
3010 delivery time.
3011
301233. Exim was neglecting to zero errno before one call of strtol() when
3013 expanding a string and expecting an integer value. On some systems this
3014 resulted in spurious "integer overflow" errors. Also, it was casting the
3015 result into an int without checking.
3016
301734. Testing for a connection timeout using "timeout_connect" in the retry rules
3018 did not work. The code looks as if it has *never* worked, though it appears
3019 to have been documented since at least releast 1.62. I have made it work.
3020
302135. The "timeout_DNS" error in retry rules, also documented since at least
3022 1.62, also never worked. As it isn't clear exactly what this means, and
3023 clearly it isn't a major issue, I have abolished the feature by treating it
3024 as "timeout", and writing a warning to the main and panic logs.
3025
302636. The display of retry rules for -brt wasn't always showing the error code
3027 correctly.
3028
302937. Added new error conditions to retry rules: timeout_A, timeout_MX,
3030 timeout_connect_A, timeout_connect_MX.
3031
303238. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3033 to the empty sender.
3034
303539. The daemon was not analysing the content of -oX till after it had closed
3036 stderr and disconnected from the controlling terminal. This meant that any
3037 syntax errors were only noted on the panic log, and the return code from
3038 the command was 0. By re-arranging the code a little, I've made the
3039 decoding happen first, so such errors now appear on stderr, and the return
3040 code is 1. However, the actual setting up of the sockets still happens in
3041 the disconnected process, so errors there are still only recorded on the
3042 panic log.
3043
304440. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3045 connections (as happens on some IP stacks) was logged at start up time as
3046 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3047 it from "IPv6 and IPv4", which means that two separate sockets are being
3048 used.
3049
305041. The debug output for gethostbyname2() or getipnodebyname() failures now
3051 says whether AF_INET or AF_INET6 was passed as an argument.
3052
305342. Exiwhat output was messed up when time zones were included in log
3054 timestamps.
3055
305643. Exiwhat now gives more information about the daemon's listening ports,
3057 and whether -tls-on-connect was used.
3058
305944. The "port" option of the smtp transport is now expanded.
3060
306145. A "message" modifier in a "warn" statement in a non-message ACL was being
3062 silently ignored. Now an error message is written to the main and panic
3063 logs.
3064
306546. There's a new ACL modifier called "logwrite" which writes to a log file
3066 as soon as it is encountered.
3067
306847. Added $local_user_uid and $local_user_gid at routing time.
3069
307048. Exim crashed when trying to verify a sender address that was being
3071 rewritten to "<>".
3072
307349. Exim was recognizing only a space character after ".include". It now also
3074 recognizes a tab character.
3075
307650. Fixed several bugs in the Perl script that creates the exim.8 man page by
3077 extracting the relevant information from the specification. The man page no
3078 longer contains scrambled data for the -d option, and I've added a section
3079 at the front about calling Exim under different names.
3080
308151. Added "extra_headers" argument to the "mail" command in filter files.
3082
308352. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3084 crash.
3085
308653. Installed eximstats 1.29.
3087
308854. Added transport_filter_timeout as a generic transport option.
3089
309055. Exim no longer adds an empty Bcc: header to messages that have no To: or
3091 Cc: header lines. This was required by RFC 822, but it not required by RFC
3092 2822.
3093
309456. Exim used to add From:, Date:, and Message-Id: header lines to any
3095 incoming messages that did not have them. Now it does so only if the
3096 message originates locally, that is, if there is no associated remote host
3097 address. When Resent- header lines are present, this applies to the Resent-
3098 lines rather than the non-Resent- lines.
3099
310057. Drop incoming SMTP connection after too many syntax or protocol errors. The
3101 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3102
310358. Messages for configuration errors now include the name of the main
3104 configuration file - useful now that there may be more than one file in a
3105 list (.included file names were always shown).
3106
310759. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3108 for those rare installations that do not start the daemon as root or run it
3109 setuid root. I've cut out the call to initgroups() if the daemon is not
3110 root at that time.
3111
311260. The Exim user and group can now be bound into the binary as text strings
3113 that are looked up at the start of Exim's processing.
3114
311561. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3116
311762. Added $mailstore_basename variable.
3118
311963. Installed patch to sieve.c from Michael Haardt.
3120
312164. When Exim failed to open the panic log after failing to open the main log,
3122 the original message it was trying to log was written to stderr and debug
3123 output, but if they were not available (the usual case in production), it
3124 was lost. Now it is written to syslog before the two lines that record the
3125 failures to open the logs.
3126
312765. Users' Exim filters run in subprocesses under the user's uid. It is
3128 possible for a "deliver" command or an alias in a "personal" command to
3129 provoke an address rewrite. If logging of address rewriting is configured,
3130 this fails because the process is not running as root or exim. There may be
3131 a better way of dealing with this, but for the moment (because 4.30 needs
3132 to be released), I have disabled address rewrite logging when running a
3133 filter in a non-root, non-exim process.
3134
3135
3136Exim version 4.24
3137-----------------
3138
3139 1. The buildconfig auxiliary program wasn't quoting the value set for
3140 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3141 not defined. This bug was masked in 4.22 by the effect that was fixed in
3142 change 4.23/1.
3143
3144 2. Some messages that were rejected after a message id was allocated were
3145 shown as "incomplete" by exigrep. It no longer does this for messages that
3146 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3147
3148 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3149 have allow_domain_literals set, the ID did not get logged in the <= line.
3150 Domain literals are now always recognized in Message-ID: header lines.
3151
3152 4. The first argument for a ${extract expansion item is the key name or field
3153 number. Leading and trailing spaces in this item were not being ignored,
3154 causing some misleading effects.
3155
3156 5. When deliver_drop_privilege was set, single queue runner processes started
3157 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3158 same command in the process it spins off) were not dropping privilege.
3159
3160 6. When the daemon running as "exim" started a queue runner, it always
3161 re-executed Exim in the spun-off process. This is a waste of effort when
3162 deliver_drop_privilege is set. The new process now just calls the
3163 queue-runner function directly.
3164
3165
3166Exim version 4.23
3167-----------------
3168
3169 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3170 HEADERS_CHARSET.
3171
3172 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3173 ignored. Though the use of -oP was forcing the writing of a pid file, it
3174 was always written to the default place.
3175
3176 3. If the message "no IP address found for host xxxx" is generated during
3177 incoming verification, it is now followed by identification of the incoming
3178 connection (so you can more easily find what provoked it).
3179
3180 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3181
3182 5. Added some features to "harden" Exim a bit more against certain attacks:
3183
3184 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3185 be put in Local/Makefile. This is like the never_users runtime option,
3186 but it cannot be overridden. The default setting is "root".
3187
3188 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3189 prefix string with which any file named in a -C command line option
3190 must start.
3191
3192 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3193 is retained for -C and -D only if the caller of Exim is root. Without
3194 it, the exim user may also use -C and -D and retain privilege.
3195
3196 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3197 command line option is disabled.
3198
3199 6. Macro names set by the -D option must start with an upper case letter, just
3200 like macro names defined in the configuration file.
3201
3202 7. Added "dereference=" facility to LDAP.
3203
3204 8. Two instances of the typo "uknown" in the source files are fixed.
3205
3206 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3207 the Configure-Makefile script screwed up while processing it.
3208
320910. Incorporated PCRE 4.4.
3210
321111. The SMTP synchronization check was not operating right at the start of an
3212 SMTP session. For example, it could not catch a HELO sent before the client
3213 waited for the greeting. There is now a check for outstanding input at the
3214 point when the greeting is written. Because of the duplex, asynchronous
3215 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3216 way, but not yet received, when the check is performed.
3217
321812. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3219 on TCP/IP sockets, because this apparently causes some broken clients to
3220 timeout.
3221
322213. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3223 unchanged) from the Cygwin maintainer.
3224
322514. The code for -bV that shows what is in the binary showed "mbx" when maildir
3226 was supported instead of testing for mbx. Effectively a typo.
3227
322815. The spa authenticator server code was not checking that the input it
3229 received was valid base64.
3230
323116. The debug output line for the "set" modifier in ACLs was not showing the
3232 name of the variable that was being set.
3233
323417. Code tidy: the variable type "vtype_string" was never used. Removed it.
3235
323618. Previously, a reference to $sender_host_name did not cause a DNS reverse
3237 lookup on its own. Something else was needed to trigger the lookup. For
3238 example, a match in host_lookup or the need for a host name in a host list.
3239 Now, if $sender_host_name is referenced and the host name has not yet been
3240 looked up, a lookup is performed. If the lookup fails, the variable remains
3241 empty, and $host_lookup_failed is set to "1".
3242
324319. Added "eqi" as a case-independent comparison operator.
3244
324520. The saslauthd authentication condition could segfault if neither service
3246 nor realm was specified.
3247
324821. If an overflowing value such as "2048M" was set for message_size_limit, the
3249 error message that was logged was misleading, and incoming SMTP
3250 connections were dropped. The message is now more accurate, and temporary
3251 errors are given to SMTP connections.
3252
325322. In some error situations (such as 21 above) Exim rejects all SMTP commands
3254 (except RSET) with a 421 error, until QUIT is received. However, it was
3255 failing to send a response to QUIT.
3256
325723. The HELO ACL was being run before the code for helo_try_verify_hosts,
3258 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3259 ACL is now run after the helo_try_verify_hosts code.
3260
326124. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3262 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3263 case-independent, so other case variants are also recognized). Apparently
3264 some systems use these upper case variants.
3265
326625. If more than two messages were waiting for the same host, and a transport
3267 filter was specified for the transport, Exim sent two messages over the
3268 same TCP/IP connection, and then failed with "socket operation on non-
3269 socket" when it tried to send the third.
3270
327126. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3272
327327. The extern definition of crypt16() in expand.c was not being excluded when
3274 the OS had its own crypt16() function.
3275
327628. Added bounce_return_body as a new option, and bounce_return_size_limit
3277 as a preferred synonym for return_size_limit, both as an option and as an
3278 expansion variable.
3279
328029. Added LIBS=-liconv to OS/Makefile-OSF1.
3281
328230. Changed the default configuration ACL to relax the local part checking rule
3283 for addresses that are not in any local domains. For these addresses,
3284 slashes and pipe symbols are allowed within local parts, but the sequence
3285 /../ is explicitly forbidden.
3286
328731. SPA server authentication was not clearing the challenge buffer before
3288 using it.
3289
329032. log_message in a "warn" ACL statement was writing to the reject log as
3291 well as to the main log, which contradicts the documentation and doesn't
3292 seem right (because no rejection is happening). So I have stopped it.
3293
329433. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3295 However, I am unable to do any testing of this.
3296
329734. Fixed an infelicity in the appendfile transport. When checking directories
3298 for a mailbox, to see if any needed to be created, it was accidentally
3299 using path names with one or more superfluous leading slashes; tracing
3300 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3301
330235. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3303 contents are added to the log line that is written for every discarded
3304 recipient. (Previously a log_message setting was ignored.)
3305
330636. The ${quote: operator now quotes the string if it is empty.
3307
330837. The install script runs exim in order to find its version number. If for
3309 some reason other than non-existence or emptiness, which it checks, it
3310 could not run './exim', it was installing it with an empty version number,
3311 i.e. as "exim-". This error state is now caught, and the installation is
3312 aborted.
3313
331438. An argument was missing from the function that creates an error message
3315 when Exim fails to connect to the socket for saslauthd authentication.
3316 This could cause Exim to crash, or give a corrupted message.
3317
331839. Added isip, isip4, and isip6 to ${if conditions.
3319
332040. The ACL variables $acl_xx are now saved with the message, and can be
3321 accessed later in routers, transports, and filters.
3322
332341. The new lookup type nwildlsearch is like wildlsearch, except that the key
3324 strings in the file are not string-expanded.
3325
332642. If a MAIL command specified a SIZE value that was too large to fit into an
3327 int variable, the check against message_size_limit failed. Such values are
3328 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3329 day this will have to be increased, but I don't think I want to be around
3330 when emails are that large.
3331
3332
3333
3334Exim version 4.22
3335-----------------
3336
3337 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3338 iconv() is not standard in FreeBSD.
3339
3340 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3341 IPv6 enabled. The observed symptom was a segmentation fault on return from
3342 the function os_common_find_running_interfaces() in src/os.c.
3343
3344 3. In the check_special_case() function in daemon.c I had used "errno" as an
3345 argument name, which causes warnings on some systems. This was basically a
3346 typo, since it was named "eno" in the comments!
3347
3348 4. The code that waits for the clock to tick (at a resolution of some fraction
3349 of a second) so as to ensure message-id uniqueness was always waiting for
3350 at least one whole tick, when it could have waited for less. [This is
3351 almost certainly not relevant at current processor speeds, where it is
3352 unlikely to ever wait at all. But we try to future-proof.]
3353
3354 5. The function that sleeps for a time interval that includes fractions of a
3355 second contained a race. It did not block SIGALRM between setting the
3356 timer, and suspending (a couple of lines later). If the interval was short
3357 and the sigsuspend() was delayed until after it had expired, the suspension
3358 never ended. On busy systems this could lead to processes getting stuck for
3359 ever.
3360
3361 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3362 process, before it forks any delivery processes. The open lookup caching
3363 mechanism meant that the open file or database connection was passed into
3364 the delivery process. The problem was that delivery processes always tidy
3365 up cached lookup data. This could cause a problem for the next delivery
3366 process started by the queue runner, because the external queue runner
3367 process does not know about the closure. So the next delivery process
3368 still has data in the lookup cache. In the case of a file lookup, there was
3369 no problem because closing a file descriptor in a subprocess doesn't affect
3370 the parent. However, if the lookup was caching a connection to a database,
3371 the connection was closed, and the second delivery process was likely to
3372 see errors such as "PGSQL: query failed: server closed the connection
3373 unexpectedly". The problem has been fixed by closing all cached lookups
3374 in a queue runner before running a delivery process.
3375
3376 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3377 seem to have the "const" qualifier which it has on other OS. I've
3378 parameterised it.
3379
3380 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3381 *of the same type* (client or server) with the same public name that an
3382 error should be diagnosed.
3383
3384 9. When Exim looked up a host name for an IP address, but failed to find the
3385 original IP address when looking up the host name (a safety check), it
3386 output the message "<ip address> does not match any IP for NULL", which was
3387 confusing, to say the least. The bug was that the host name should have
3388 appeared instead of "NULL".
3389
339010. Since release 3.03, if Exim is called by a uid other than root or the Exim
3391 user that is built into the binary, and the -C or -D options is used, root
3392 privilege is dropped before the configuration file is read. In addition,
3393 logging is switched to stderr instead of the normal log files. If the
3394 configuration then re-defines the Exim user, the unprivileged environment
3395 is probably not what is expected, so Exim logs a panic warning message (but
3396 proceeds).
3397
3398 However, if deliver_drop_privilege is set, the unprivileged state may well
3399 be exactly what is intended, so the warning has been cut out in that case,
3400 and Exim is allowed to try to write to its normal log files.
3401
3402
3403Exim version 4.21
3404-----------------
3405
3406 1. smtp_return_error_details was not giving details for temporary sender
3407 or receiver verification errors.
3408
3409 2. Diagnose a configuration error if two authenticators have the same public
3410 name.
3411
3412 3. Exim used not to create the message log file for a message until the first
3413 delivery attempt. This could be confusing when incoming messages were held
3414 for policy or load reasons. The message log file is now created at the time
3415 the message is received, and an initial "Received" line is written to it.
3416
3417 4. The automatically generated man page for command line options had a minor
3418 bug that caused no ill effects; however, a more serious problem was that
3419 the procedure for building the man page automatically didn't always
3420 operate. Consequently, release 4.20 contains an out-of-date version. This
3421 shouldn't happen again.
3422
3423 5. When building Exim with embedded Perl support, the script that builds the
3424 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3425 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3426
3427 6. The freeze_tell option was not being used for messages that were frozen on
3428 arrival, either by an ACL or by local_scan().
3429
3430 7. Added the smtp_incomplete_transaction log selector.
3431
3432 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3433 was accepting AUTH without a new EHLO.
3434
3435 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3436 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3437 RFC.
3438
343910. Logging of TCP/IP connections (when configured) now happens in the main
3440 daemon process instead of the child process, so that the TCP/IP connection
3441 count is more accurate (but it can never be perfect).
3442
344311. The use of "drop" in a nested ACL was not being handled correctly in the
3444 outer ACL. Now, if condition failure induced by the nested "drop" causes
3445 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3446 or "require"), the connection is dropped.
3447
344812. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3449 that yield "discard" can now be used with an "accept" or a "discard" verb,
3450 but an error is generated for any others (because I can't see a useful way
3451 to define what should happen).
3452
345313. When an ACL is read dynamically from a file (or anywhere else), the lines
3454 are now processed in the same way as lines in the Exim configuration file.
3455 In particular, continuation lines are supported.
3456
345714. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3458
345915. Added -ti meaning -t -i.
3460
346116. Check for letters, digits, hyphens, and dots in the names of dnslist
3462 domains, and warn by logging if others are found.
3463
346417. At least on BSD, alignment is not guarenteed for the array of ifreq's
3465 returned from GIFCONF when Exim is trying to find the list of interfaces on
3466 a host. The code in os.c has been modified to copy each ifreq to an aligned
3467 structure in all cases.
3468
3469 Also, in some cases, the returned ifreq's were being copied to a 'struct
3470 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3471 means the last couple of bytes of an IPv6 address could be chopped if the
3472 ifreq contained only a normal sockaddr (14 bytes storage).
3473
347418. Named domain lists were not supported in the hosts_treat_as_local option.
3475 An entry such as +xxxx was not recognized, and was treated as a literal
3476 domain name.
3477
347819. Ensure that header lines added by a DATA ACL are included in the reject log
3479 if the ACL subsequently rejects the message.
3480
348120. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3482 MD5 (which is deprecated).
3483
348421. When testing a filter file using -bf, Exim was writing a message when it
3485 took the sender from a "From " line in the message, but it was not doing so
3486 when it took $return_path from a Return-Path: header line. It now does.
3487
348822. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3489 with a valid header line field name (a series of printing characters
3490 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3491
349223. Changed "disc" in the source to "disk" to conform to the documentation and
3493 the book and for uniformity.
3494
349524. Ignore Sendmail's -Ooption=value command line item.
3496
349725. When execve() failed while trying to run a command in a pipe transport,
3498 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3499 could be confused with a return value of 69 from the command itself. This
3500 has been changed to 127, the value the shell returns if it is asked to run
3501 a non-existent command. The wording for the related log line suggests a
3502 non-existent command as the problem.
3503
350426. If received_header_text expands to an empty string, do not add a Received:
3505 header line to the message. (Well, it adds a token one on the spool, but
3506 marks it "old" so that it doesn't get used or transmitted.)
3507
350827. Installed eximstats 1.28 (addition of -nt option).
3509
351028. There was no check for failure on the call to getsockname() in the daemon
3511 code. This can fail if there is a shortage of resources on the system, with
3512 ENOMEM, for example. A temporary error is now given on failure.
3513
351429. Contrary to the C standard, it seems that in some environments, the
3515 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3516 program. Exim now does this explicitly; it affects the formatting of
3517 timestamps using strftime().
3518
351930. If exiqsumm was given junk data, it threw up some uninitialized variable
3520 complaints. I've now initialized all the variables, to avoid this.
3521
352232. Header lines added by a system filter were not being "seen" during
3523 transport-time rewrites.
3524
352533. The info_callback() function passed to OpenSSL is set up with type void
3526 (*)(SSL *, int, int), as described somewhere. However, when calling the
3527 function (actually a macro) that sets it up, the type void(*)() is
3528 expected. I've put in a cast to prevent warnings from picky compilers.
3529
353034. If a DNS black list lookup found a CNAME record, but there were no A
3531 records associated with the domain it pointed at, Exim crashed.
3532
353335. If a DNS black list lookup returned more than one A record, Exim ignored
3534 all but the first. It now scans all returned addresses if a particular IP
3535 value is being sought. In this situation, the contents of the
3536 $dnslist_value variable are a list of all the addresses, separated by a
3537 comma and a space.
3538
353936. Tightened up the rules for host name lookups using reverse DNS. Exim used
3540 to accept a host name and all its aliases if the forward lookup for any of
3541 them yielded the IP address of the incoming connection. Now it accepts only
3542 those names whose forward lookup yields the correct IP address. Any other
3543 names are discarded. This closes a loophole whereby a rogue DNS
3544 administrator could create reverse DNS records to break through a
3545 wildcarded host restriction in an ACL.
3546
354737. If a user filter or a system filter that ran in a subprocess used any of
3548 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3549 the wrong values were passed to the pipe command ($thisaddress had the
3550 value of $0, $0 had the value of $1, etc). This bug was introduced by
3551 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3552
355338. Improved the line breaking for long SMTP error messages from ACLs.
3554 Previously, if there was no break point between 40 and 75 characters, Exim
3555 left the rest of the message alone. Two changes have been made: (a) I've
3556 reduced the minimum length to 35 characters; (b) if it can't find a break
3557 point between 35 and 75 characters, it looks ahead and uses the first one
3558 that it finds. This may give the occasional overlong line, but at least the
3559 remaining text gets split now.
3560
356139. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3562 file descriptors might be low, and that setting 1000 would always raise it.
3563 It turns out that in some environments, the limit is already over 1000 and
3564 that lowering it causes trouble. So now Exim takes care not to decrease it.
3565
356640. When delivering a message, the value of $return_path is set to $sender_
3567 address at the start of routing (routers may change the value). By an
3568 oversight, this default was not being set up when an address was tested by
3569 -bt or -bv, which affected the outcome if any router or filter referred to
3570 $return_path.
3571
357241. The idea of the "warn" ACL verb is that it adds a header or writes to the
3573 log only when "message" or "log_message" are set. However, if one of the
3574 conditions was an address verification, or a call to a nested ACL, the
3575 messages generated by the underlying test were being passed through. This
3576 no longer happens. The underlying message is available in $acl_verify_
3577 message for both "message" and "log_message" expansions, so it can be
3578 passed through if needed.
3579
358042. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3581 new expansion $bh_ to give the encoded byte string without charset
3582 translation. Translation happens only if iconv() is available; HAVE_ICONV
3583 indicates this at build time. HEADERS_CHARSET gives the charset to
3584 translate to; headers_charset can change it in the configuration, and
3585 "headers charset" can change it in an individual filter file.
3586
358743. Now that we have a default RFC 2047 charset (see above), the code in Exim
3588 that creates RFC 2047 encoded "words" labels them as that charset instead
3589 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3590 expansion operator; (ii) when Exim creates a From: line for a local
3591 message; (iii) when a header line is rewritten to include a "phrase" part.
3592
359344. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3594 buggy, causing it to skip the first lines of messages whose message ID
3595 ended in 'D'. This would not have bitten before Exim release 4.14, because
3596 message IDs were unlikely to end in 'D' before then. The effect was to have
3597 incorrect size information for certain domains.
3598
359945. #include "config.h" was missing at the start of the crypt16.c module. This
3600 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3601 noticed.
3602
360346. If there was a timeout during a "random" callout check, Exim treated it as
3604 a failure of the random address, and carried on sending RSET and the real
3605 address. If the delay was just some slowness somewhere, the response to the
3606 original RCPT would be taken as a response to RSET and so on, causing
3607 mayhem of various kinds.
3608
360947. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3610 when I implemented it. It didn't allow for the fact that some option values
3611 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3612 the right test for positive values.
3613
361448. Domain names in DNS records are case-independent. Exim always looks them up
3615 in lower case. Some resolvers return domain names in exactly the case they
3616 appear in the zone file, that is, they may contain uppercase letters. Not
3617 all resolvers do this - some return always lower case. Exim was treating a
3618 change of case by a resolver as a change of domain, similar to a widening
3619 of a domain abbreviation. This triggered its re-routing code and so it was
3620 trying to route what was effectively the same domain again. This normally
3621 caused routing to fail (because the router wouldn't handle the domain
3622 twice). Now Exim checks for this case specially, and just changes the
3623 casing of the domain that it ultimately uses when it transmits the message
3624 envelope.
3625
362649. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3627 module.
3628
362950. If a filter generated a file delivery with a non-absolute name (possible if
3630 no home directory exists for the router), the forbid_file option was not
3631 forbidding it.
3632
363351. Added '&' feature to dnslists, to provide bit mask matching in addition to
3634 the existing equality matching.
3635
363652. Exim was using ints instead of ino_t variables in some places where it was
3637 dealing with inode numbers.
3638
363953. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3640 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3641 called TMPDIR, and if it finds it is different, it changes its value.
3642
364354. The smtp_printf() function is now made available to local_scan() so
3644 additional output lines can be written before returning. There is also an
3645 smtp_fflush() function to enable the detection of a dropped connection.
3646 The variables smtp_input and smtp_batched_input are exported to
3647 local_scan().
3648
364955. Changed the default runtime configuration: the message "Unknown user"
3650 has been removed from the ACL, and instead placed on the localuser router,
3651 using the cannot_route_message feature. This means that any verification
3652 failures that generate their own messages won't get overridden. Similarly,
3653 the "Unrouteable address" message that was in the ACL for unverifiable
3654 relay addresses has also been removed.
3655
365656. Added hosts_avoid_esmtp to the smtp transport.
3657
365857. The exicyclog script was not checking for the esoteric option
3659 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3660 will work only if exicyclog is run under the appropriate euid.
3661
366258. Following a discussion on the list, the rules by which Exim recognises line
3663 endings on incoming messages have been changed. The -dropcr and drop_cr
3664 options are now no-ops, retained only for backwards compatibility. The
3665 following line terminators are recognized: LF CRLF CR. However, special
3666 processing applies to CR:
3667
3668 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3669 nor a local message in the state where . is a terminator.
3670
3671 (ii) If a bare CR is encountered in a header line, an extra space is added
3672 after the line terminator so as not to end the header. The reasoning
3673 behind this is that bare CRs in header lines are most likely either
3674 to be mistakes, or people trying to play silly games.
3675
367659. The size of a message, as listed by "-bp" or in the Exim monitor window,
3677 was being incorrectly given as 18 bytes larger than it should have been.
3678 This is a VOB (very old bug).
3679
368060. This may never have affected anything current, but just in case it has:
3681 When the local host is found other than at the start of a list of hosts,