(1) New Wish. (2) Typo and tabs in FAQ source.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
9176e9f0 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.133 2005/05/04 10:17:28 ph10 Exp $
495ae4b0
PH
2
3Change log file for Exim from version 4.21
4-------------------------------------------
5
bbe902f0 6
7982096b
PH
7Exim version 4.51
8-----------------
9
1a46a8c5
PH
10TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
11 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
12
2f079f46 13TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
14
15TK/03 Fix ACL "regex" condition not being available in MIME ACL.
16
17PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
18 to test Sieve filters that use "vacation".
19
20PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
21 that changes the way the GnuTLS parameters are stored in the cache file.
22 The new format can be generated externally. For backward compatibility,
23 if the data in the cache doesn't make sense, Exim assumes it has read an
24 old-format file, and it generates new data and writes a new file. This
25 means that you can't go back to an older release without removing the
26 file.
27
28PH/03 A redirect router that has both "unseen" and "one_time" set does not
29 work if there are any delivery delays because "one_time" forces the
30 parent to be marked "delivered", so its unseen clone is never tried
31 again. For this reason, Exim now forbids the simultaneous setting of
32 these two options.
33
34PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
35 redirected to themselves ("homonym" addresses). Read the long ChangeLog
36 entry if you want to know the details. The fix, however, neglected to
37 consider the case when local delivery batching is involved. The test for
38 "previously delivered" was not happening when checking to see if an
39 address could be batched with a previous (undelivered) one; under
40 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 41 address.
1a46a8c5
PH
42
43PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
44 in its include files, and this causes problems building Exim.
45
46PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
47 header_syntax) but Exim was just ignoring anything given after a slash.
48 In particular, this caused confusion with an attempt to use "verify =
49 reverse_host_lookup/defer_ok". An error is now given when options are
50 supplied for verify items that do not have them. (Maybe reverse_host_
51 lookup should have a defer_ok option, but that's a different point.)
52
53PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
54 defined by RFC 821) to 2048, because there were problems with some AUTH
55 commands, and RFC 1869 says the size should be increased for extended
56 SMTP commands that take arguments.
57
58PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
59 Finch).
60
61PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
62 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 63
49c2d5ea
PH
64PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
65 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
66 whatever). Otherwise the password may appear in the log. Following change
67 PH/42 below, there is no longer a chance of it appearing in a bounce
68 message.
49c2d5ea 69
bf759a8b
PH
70PH/11 Installed exipick version 20050225.0 from John Jetmore.
71
83364d30
PH
72PH/12 If the last host in a fallback_hosts list was multihomed, only the first
73 of its addresses was ever tried. (Bugzilla bug #2.)
74
7999bbd7
PH
75PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
76 the result incorrectly in the debug output. (It correctly added a newline
77 to what was transported.)
78
7dbf77c9
PH
79TF/01 Added $received_time.
80
74e0617f
PH
81PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
82 commented out examples of how to interface to a virus scanner and to
83 SpamAssassin. Also added commented examples of av_scanner and
84 spamd_address settings.
85
2f079f46
PH
86PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
87 and controls are allowed in which ACLs. There were a couple of minor
88 errors. Some of the entries in the conditions table (which is a table of
89 where they are NOT allowed) were getting very unwieldy; rewrote them as a
90 negation of where the condition IS allowed.
91
8c841523
PH
92PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
93
7766a4f0
PH
94PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
95 header file does not have a version number, so I've had to invent a new
96 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
97 API. The code is untested by me (my Linux distribution still has 0.3.2 of
98 radiusclient), but it was contributed by a Radius user.
99
8b417f2c
PH
100PH/18 Installed Lars Mainka's patch for the support of CRL collections in
101 files or directories, for OpenSSL.
102
901f42cb
PH
103PH/19 When an Exim process that is running as root has to create an Exim log
104 file, it does so in a subprocess that runs as exim:exim so as to get the
105 ownership right at creation (otherwise, other Exim processes might see
106 the file with the wrong ownership). There was no test for failure of this
107 fork() call, which would lead to the process getting stuck as it waited
108 for a non-existent subprocess. Forks do occasionally fail when resources
109 run out. I reviewed all the other calls to fork(); they all seem to check
110 for failure.
111
f9b9210e
PH
112PH/20 When checking for unexpected SMTP input at connect time (before writing
113 the banner), Exim was not dealing correctly with a non-positive return
114 from the read() function. If the client had disconnected by this time,
115 the result was a log entry for a synchronization error with an empty
116 string after "input=" when read() returned zero. If read() returned -1
117 (an event I could not check), uninitialized data bytes were printed.
118 There were reports of junk text (parts of files, etc) appearing after
119 "input=".
120
54cdb463
PH
121PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
122
cf00dad6
PH
123PH/22 Added support for macro redefinition, and (re)definition in between
124 driver and ACL definitions.
125
acb1b346
PH
126PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
127 forgetting to use the resulting value; it was using the unexpanded value.
128
c5ddb310
PH
129PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
130 hadn't been configured. The fix is from Juergen Kreileder, who
131 understands it better than I do:
132
133 "Here's what I see happening with three configured cyrus_sasl
134 authenticators configured (plain, login, cram-md5):
135
136 On startup auth_cyrus_sasl_init() gets called for each of these.
137 This means three calls to sasl_listmech() without a specified mech_list.
138 => SASL tests which mechs of all available mechs actually work
139 => three warnings about OTP not working
140 => the returned list contains: plain, login, cram-md5, digest-md5, ...
141
142 With the patch, sasl_listmech() also gets called three times. But now
143 SASL's mech_list option is set to the server_mech specified in the the
144 authenticator. Or in other words, the answer from sasl_listmech()
145 gets limited to just the mech you're testing for (which is different
146 for each call.)
147 => the return list contains just 'plain' or 'login', 'cram-md5' or
148 nothing depending on the value of ob->server_mech.
149
150 I've just tested the patch: Authentication still works fine,
151 unavailable mechs specified in the exim configuration are still
152 caught, and the auth.log warnings about OTP are gone."
153
31619da6
PH
154PH/25 When debugging is enabled, the contents of the command line are added
155 to the debugging output, even when log_selector=+arguments is not
156 specified.
157
bebaf0fc
PH
158PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
159 answer is "GNU", and only if the return is "GNU/something" is the answer
160 "Linux".
161
475fe28a
PH
162PH/27 $acl_verify_message is now set immediately after the failure of a
163 verification in an ACL, and so is available in subsequent modifiers. In
164 particular, the message can be preserved by coding like this:
165
166 warn !verify = sender
167 set acl_m0 = $acl_verify_message
168
169 Previously, $acl_verify_message was set only while expanding "message"
170 and "log_message" when a very denied access.
171
7e8bec7a
PH
172PH/28 Modified OS/os.c-Linux with
173
174 -#ifndef OS_LOAD_AVERAGE
175 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
176
177 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
178 nomenclature these days.)
179
e4a89c47
PH
180PH/29 Installed patch from the Sieve maintainer that adds the options
181 sieve_useraddress and sieve_subaddress to the redirect router.
182
5ca2a9a1
PH
183PH/30 In these circumstances:
184 . Two addresses routed to the same list of hosts;
185 . First host does not offer TLS;
186 . First host accepts first address;
187 . First host gives temporary error to second address;
188 . Second host offers TLS and a TLS session is established;
189 . Second host accepts second address.
190 Exim incorrectly logged both deliveries with the TLS parameters (cipher
191 and peerdn, if requested) that were in fact used only for the second
192 address.
7e8bec7a 193
c688b954
PH
194PH/31 When doing a callout as part of verifying an address, Exim was not paying
195 attention to any local part prefix or suffix that was matched by the
196 router that accepted the address. It now behaves in the same way as it
197 does for delivery: the affixes are removed from the local part unless
198 rcpt_include_affixes is set on the transport.
199
fed77020
PH
200PH/32 Add the sender address, as F=<...>, to the log line when logging a
201 timeout during the DATA phase of an incoming message.
202
7fe1560f
PH
203PH/33 Sieve envelope tests were broken for match types other than :is. I have
204 applied a patch sanctioned by the Sieve maintainer.
c688b954 205
ebb6e6d5
PH
206PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
207 the uid or gid is negative. A case of a negative gid caused this to be
208 noticed. The fix allows for either to be negative.
209
9c4e8f60
PH
210PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
211 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
212 overlooked.
213
214PH/36 The change PH/12 above was broken. Fixed it.
215
d7174846
PH
216PH/37 Exim used to check for duplicate addresses in the middle of routing, on
217 the grounds that routing the same address twice would always produce the
218 same answer. This might have been true once, but it is certainly no
219 longer true now. Routing a child address may depend on the previous
220 routing that produced that child. Some complicated redirection strategies
221 went wrong when messages had multiple recipients, and made Exim's
222 behaviour dependent on the order in which the addresses were given.
223
224 I have moved the duplicate checking until after the routing is complete.
225 Exim scans the addresses that are assigned to local and remote
226 transports, and removes any duplicates. This means that more work will be
227 done, as duplicates will always all be routed, but duplicates are
228 presumably rare, so I don't expect this is of any significance.
229
230 For deliveries to pipes, files, and autoreplies, the duplicate checking
231 still happens during the routing process, since they are not going to be
232 routed further.
233
cfe75fc3
PH
234PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
235 It corrects a timeout issue with spamd. This is Ian's comment: "The
236 background is that sometimes spamd either never reads data from a
237 connection it has accepted, or it never writes response data. The exiscan
238 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
239 blindly assumes that writes won't block so it may never time out."
240
be22d70e
PH
241PH/39 Allow G after quota size as well as K and M.
242
0612b098
PH
243PH/40 The value set for $authenticated_id in an authenticator may not contain
244 binary zeroes or newlines because the value is written to log lines and
245 to spool files. There was no check on this. Now the value is run through
246 the string_printing() function so that such characters are converted to
247 printable escape sequences.
248
2e0c1448
PH
249PH/41 $message_linecount is a new variable that contains the total number of
250 lines in the message. Compare $body_linecount, which is the count for the
251 body only.
252
447d236c
PH
253PH/42 Exim no longer gives details of delivery errors for specific addresses in
254 bounce and delay warning messages, except in certain special cases, which
255 are as follows:
256
257 (a) An SMTP error message from a remote host;
258 (b) A message specified in a :fail: redirection;
259 (c) A message specified in a "fail" command in a system filter;
260 (d) A message specified in a FAIL return from the queryprogram router;
261 (e) A message specified by the cannot_route_message router option.
262
263 In these cases only, Exim does include the error details in bounce and
264 warning messages. There are also a few cases where bland messages such
265 as "unrouteable address" or "local delivery error" are given.
266
d20976dc
PH
267PH/43 $value is now also set for the "else" part of a ${run expansion.
268
f656d135
PH
269PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
270 being worked on, but at least Exim now implements the latest version to
271 play with."
272
2e2a30b4
PH
273PH/45 In a pipe transport, although a timeout while waiting for the pipe
274 process to complete was treated as a delivery failure, a timeout while
275 writing the message to the pipe was logged, but erroneously treated as a
276 successful delivery. Such timeouts include transport filter timeouts. For
277 consistency with the overall process timeout, these timeouts are now
278 treated as errors, giving rise to delivery failures by default. However,
279 there is now a new Boolean option for the pipe transport called
280 timeout_defer, which, if set TRUE, converts the failures into defers for
281 both kinds of timeout. A transport filter timeout is now identified in
282 the log output.
283
9176e9f0
PH
284PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
285 systems where "make" and "gmake" are different, calling "gmake" at top
286 level broke things. I've arranged for the value of $(MAKE) to be passed
287 from the Makefile to this script so that it can call the same version of
288 "make".
289
7982096b 290
bbe902f0
PH
291A note about Exim versions 4.44 and 4.50
292----------------------------------------
293
294Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
295changes of various kinds. As a consequence, a big documentation update was
296needed. This delayed the release for rather longer than seemed good, especially
297in the light of a couple of (minor) security issues. Therefore, the changes
298that fixed bugs were backported into 4.43, to create a 4.44 maintenance
299release. So 4.44 and 4.50 are in effect two different branches that both start
300from 4.43.
301
302I have left the 4.50 change log unchanged; it contains all the changes since
3034.43. The change log for 4.44 is below; many of its items are identical to
304those for 4.50. This seems to be the most sensible way to preserve the
305historical information.
306
307
f7b63901 308Exim version 4.50
495ae4b0
PH
309-----------------
310
5fe762f6
PH
311 1. Minor wording change to the doc/README.SIEVE file.
312
139059f6 313 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 314 computation of the current number of files was incorrect.
495ae4b0 315
7086e875
PH
316 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
317 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
318 place.
319
35af9f61
PH
320 4. Give more explanation in the error message when the command for a transport
321 filter fails to execute.
322
b668c215
PH
323 5. There are several places where Exim runs a non-Exim command in a
324 subprocess. The SIGUSR1 signal should be disabled for these processes. This
325 was being done only for the command run by the queryprogram router. It is
326 now done for all such subprocesses. The other cases are: ${run, transport
327 filters, and the commands run by the lmtp and pipe transports.
328
a494b1e1
PH
329 6. Added CONFIGURE_GROUP build-time option.
330
331 7. Some older OS have a limit of 256 on the maximum number of file
332 descriptors. Exim was using setrlimit() to set 1000 as a large value
333 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
334 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 335
c5fcb476
PH
336 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
337 was an oversight, and furthermore, ever since the addition of extra
338 controls (e.g. 4.43/32), the checks on when to allow different forms of
339 "control" were broken. There should now be diagnostics for all cases when a
340 control that does not make sense is encountered.
341
69358f02
PH
342 9. Added the /retain_sender option to "control=submission".
343
5be20824
PH
34410. $recipients is now available in the predata ACL (oversight).
345
eb2c0248
PH
34611. Tidy the search cache before the fork to do a delivery from a message
347 received from the command line. Otherwise the child will trigger a lookup
348 failure and thereby defer the delivery if it tries to use (for example) a
349 cached ldap connection that the parent has called unbind on.
350
2a3eea10
PH
35112. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
352 of $address_data from the recipient verification was clobbered by the
353 sender verification.
354
35513. The value of address_data from a sender verification is now available in
356 $sender_address_data in subsequent conditions in the ACL statement.
357
23c7ff99
PH
35814. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
359
4deaf07d
PH
36015. Added a new option "connect=<time>" to callout options, to set a different
361 connection timeout.
362
926e1192
PH
36316. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
364 was its contents. (It was OK if the option was not defined at all.)
365
650edc6f
PH
36617. A "Completed" log line is now written for messages that are removed from
367 the spool by the -Mrm option.
368
2c7db3f5
PH
36918. New variables $sender_verify_failure and $recipient_verify_failure contain
370 information about exactly what failed.
371
3d235903
PH
37219. Added -dd to debug only the daemon process.
373
7c7ad977
PH
37420. Incorporated Michael Haardt's patch to ldap.c for improving the way it
375 handles timeouts, both on the server side and network timeouts. Renamed the
376 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
377
981756db
PH
37821. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
379 It is now set to "smtps".
380
d4eb88df
PH
38122. $host_address is now set to the target address during the checking of
382 ignore_target_hosts.
383
38423. When checking ignore_target_hosts for an ipliteral router, no host name was
385 being passed; this would have caused $sender_host_name to have been used if
386 matching the list had actually called for a host name (not very likely,
387 since this list is usually IP addresses). A host name is now passed as
388 "[x.x.x.x]".
389
7d468ab8
PH
39024. Changed the calls that set up the SIGCHLD handler in the daemon to use the
391 code that specifies a non-restarting handler (typically sigaction() in
392 modern systems) in an attempt to fix a rare and obscure crash bug.
393
39425. Narrowed the window for a race in the daemon that could cause it to ignore
395 SIGCHLD signals. This is not a major problem, because they are used only to
396 wake it up if nothing else does.
397
62c0818f
PH
39826. A malformed maildirsize file could cause Exim to calculate negative values
399 for the mailbox size or file count. Odd effects could occur as a result.
400 The maildirsize information is now recalculated if the size or filecount
401 end up negative.
402
26034054
PH
40327. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
404 support for a long time. Removed HAVE_SYS_VFS_H.
405
af66f652
PH
40628. Installed the latest version of exipick from John Jetmore.
407
90af77f4
PH
40829. In an address list, if the pattern was not a regular expression, an empty
409 subject address (from a bounce message) matched only if the pattern was an
410 empty string. Non-empty patterns were not even tested. This was the wrong
411 because it is perfectly reasonable to use an empty address as part of a
412 database query. An empty address is now tested by patterns that are
413 lookups. However, all the other forms of pattern expect the subject to
414 contain a local part and a domain, and therefore, for them, an empty
415 address still always fails if the pattern is not itself empty.
416
d8ef3577
PH
41730. Exim went into a mad DNS loop when attempting to do a callout where the
418 host was specified on an smtp transport, and looking it up yielded more
419 than one IP address.
420
5cb8cbc6
PH
42131. Re-factored the code for checking spool and log partition space into a
422 function that finds that data and another that does the check. The former
423 is then used to implement four new variables: $spool_space, $log_space,
424 $spool_inodes, and $log_inodes.
425
14702f5b
PH
42632. The RFC2047 encoding function was originally intended for short strings
427 such as real names; it was not keeping to the 75-character limit for
428 encoded words that the RFC imposes. It now respects the limit, and
429 generates multiple encoded words if necessary. To be on the safe side, I
430 have increased the buffer size for the ${rfc2047: expansion operator from
431 1024 to 2048 bytes.
432
063b1e99
PH
43333. It is now permitted to omit both strings after an "if" condition; if the
434 condition is true, the result is "true". As before, when the second string
435 is omitted, a false condition yields an empty string. This makes it less
436 cumbersome to write custom ACL and router conditions.
437
652e1b65
PH
43834. Failure to deliver a bounce message always caused it to be frozen, even if
439 there was an errors_to setting on the router. The errors_to setting is now
440 respected.
441
6f0c9a4f
PH
44235. If an IPv6 address is given for -bh or -bhc, it is now converted to the
443 canonical form (fully expanded) before being placed in
444 $sender_host_address.
445
33397d19
PH
44636. The table in the code that translates DNS record types into text (T_A to
447 "A" for instance) was missing entries for NS and CNAME. It is just possible
448 that this could have caused confusion if both these types were looked up
449 for the same domain, because the text type is used as part of Exim's
450 per-process caching. But the chance of anyone hitting this buglet seems
451 very small.
452
7bb56e1f
PH
45337. The dnsdb lookup has been extended in a number of ways.
454
455 (1) There is a new type, "zns", which walks up the domain tree until it
456 finds some nameserver records. It should be used with care.
457
ea3bc19b
PH
458 (2) There is a new type, "mxh", which is like "mx" except that it returns
459 just the host names, not the priorities.
460
461 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
462 looked up. The behaviour when one of the lookups defers can be
463 controlled by a keyword.
7bb56e1f 464
ea3bc19b 465 (4) It is now possible to specify the separator character for use when
7bb56e1f 466 multiple records are returned.
33397d19 467
0bcb2a0e
PH
46838. The dnslists ACL condition has been extended: it is now possible to supply
469 a list of IP addresses and/or domains to be looked up in a particular DNS
470 domain.
471
2ac0e484
PH
47239. Added log_selector=+queue_time_overall.
473
4e1fde53
PH
47440. When running the queue in the test harness, wait just a tad after forking a
475 delivery process, to get repeatability of debugging output.
476
de365ded
PH
47741. Include certificate and key file names in error message when GnuTLS fails
478 to set them up, because the GnuTLS error message doesn't include the name
479 of the failing file when there is a problem reading it.
480
f05da2e8
PH
48142. Allow both -bf and -bF in the same test run.
482
d6453af2
PH
48343. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
484
f7b63901
PH
48544. The "Exiscan patch" is now merged into the mainline Exim source.
486
48745. Sometimes the final signoff response after QUIT could fail to get
488 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
489 < 0 before doing a fflush(). This bug looks as though it goes back to the
490 introduction of TLS in release 3.20, but "sometimes" must have been rare
491 because the tests only now provoked it.
492
a444213a
PH
49346. Reset the locale to "C" after calling embedded Perl, in case it was changed
494 (this can affect the format of dates).
495
0ec020ea
PH
49647. exim_tidydb, when checking for the continued existence of a message for
497 which it has found a message-specific retry record, was not finding
498 messages that were in split spool directories. Consequently, it was
499 deleting retry records that should have stayed in existence.
500
b1206957
PH
50148. Steve fixed some bugs in eximstats.
502
50349. The SPA authentication driver was not abandoning authentication and moving
504 on to the next authenticator when an expansion was forced to fail,
505 contradicting the general specification for all authenticators. Instead it
506 was generating a temporary error. It now behaves as specified.
507
26dd5a95
PH
50850. The default ordering of permitted cipher suites for GnuTLS was pessimal
509 (the order specifies the preference for clients). The order is now AES256,
510 AES128, 3DES, ARCFOUR128.
511
343b2385
PH
51251. Small patch to Sieve code - explicitly set From: when generating an
513 autoreply.
514
1c5466b9
PH
51552. Exim crashed if a remote delivery caused a very long error message to be
516 recorded - for instance if somebody sent an entire SpamAssassin report back
517 as a large number of 550 error lines. This bug was coincidentally fixed by
518 increasing the size of one of Exim's internal buffers (big_buffer) that
519 happened as part of the Exiscan merge. However, to be on the safe side, I
520 have made the code more robust (and fixed the comments that describe what
521 is going on).
522
55ee9ee3
PH
52353. Now that there can be additional text after "Completed" in log lines (if
524 the queue_time_overall log selector is set), a one-byte patch to exigrep
525 was needed to allow it to recognize "Completed" as not the last thing in
526 the line.
527
d38f8232
PH
52854. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
529 patch that reportedly fixes this has been added. I am not expert enough to
530 create a test for it. This is what the patch creator wrote:
531
532 "I found a little strange behaviour of ldap code when working with
533 Windows 2003 AD Domain, where users was placed in more than one
534 Organization Units. When I tried to give exim partial DN, the exit code
535 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
536 But simultaneously result of request was absolutely normal ldap result,
537 so I produce this patch..."
538
3295e65b
PH
539 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
540 REFERENCE, so I have modified the code to exclude the patch when that macro
541 is not defined.
542
7102e136
PH
54355. Some experimental protocols are using DNS PTR records for new purposes. The
544 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
545 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
546 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 547 CAN-2005-0021
7102e136 548
3ca0ba97
PH
54956. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
550
c2bcbe20
PH
55157. Double the size of the debug message buffer (to 2048) so that more of very
552 long debug lines gets shown.
553
18ce445d
PH
55458. The exicyclog utility now does better if the number of log files to keep
555 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
556
1f5b4c3d
PH
55759. Two changes related to the smtp_active_hostname option:
558
559 (1) $smtp_active_hostname is now available as a variable.
560 (2) The default for smtp_banner uses $smtp_active_hostname instead
561 of $primary_hostname.
562
b975ba52
PH
56360. The host_aton() function is supposed to be passed a string that is known
564 to be a valid IP address. However, in the case of IPv6 addresses, it was
565 not checking this. This is a hostage to fortune. Exim now panics and dies
566 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
567 from a dnsdb PTR lookup with an IPv6 address that had more than 8
568 components; fortuitously, this particular loophole had already been fixed
569 by change 4.50/55 above.
570
571 If there are any other similar loopholes, the new check in host_aton()
572 itself should stop them being exploited. The report I received stated that
573 data on the command line could provoke the exploit when Exim was running as
574 exim, but did not say which command line option was involved. All I could
575 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
576 running as the user.
ea3a6f44 577 CAN-2005-0021
85b87bc2
PH
578
57961. There was a buffer overflow vulnerability in the SPA authentication code
580 (which came originally from the Samba project). I have added a test to the
581 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 582 CAN-2005-0022
b975ba52 583
17ffcae7
PH
58462. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
585 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
586
d95f9fdb
PH
58763. The daemon start-up calls getloadavg() while still root for those OS that
588 need the first call to be done as root, but it missed one case: when
589 deliver_queue_load_max is set with deliver_drop_privilege. This is
590 necessary for the benefit of the queue runner, because there is no re-exec
591 when deliver_drop_privilege is set.
592
86b8287f
PH
59364. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
594 This has been fixed.
595
60dc5e56
PH
59665. Caching of lookup data for "hosts =" ACL conditions, when a named host list
597 was in use, was not putting the data itself into the right store pool;
598 consequently, it could be overwritten for a subsequent message in the same
599 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
600 the caching.)
601
533244af
PH
60266. Added hosts_max_try_hardlimit to the smtp transport, default 50.
603
a5a28604
PH
60467. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
605 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
606 the function were treating the return as a boolean value, which happened to
607 work because 0=false and not-0=true, but is not correct code.
608
7e634d24
PH
60968. The host_aton() function was not handling scoped IPv6 addresses (those
610 with, for example, "%eth0" on the end) correctly.
611
3e11c26b
PH
61269. Fixed some compiler warnings in acl.c for the bitmaps specified with
613 negated items (that is, ~something) in unsigned ints. Some compilers
614 apparently mutter when there is no cast.
615
6729cf78
PH
61670. If an address verification called from an ACL failed, and did not produce a
617 user-specific message (i.e. there was only a "system" message), nothing was
618 put in $acl_verify_message. In this situation, it now puts the system
619 message there.
620
00f00ca5
PH
62171. Change 4.23/11 added synchronization checking at the start of an SMTP
622 session; change 4.31/43 added the unwanted input to the log line - except
623 that it did not do this in the start of session case. It now does.
624
c9bdd01c
PH
62572. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
626 This is wrong and can cause the other end to generate a synchronization
627 error if it is another Exim or anything else that does the synchronization
628 check. A QUIT command is no longer sent after a timeout.
629
d43194df
PH
63073. $host_lookup_deferred has been added, to make it easier to detect DEFERs
631 during host lookups.
632
fe5b5d0b
PH
63374. The defer_ok option of callout verification was not working if it was used
634 when verifying addresses in header lines, that is, for this case:
635
636 verify = header_sender/callout=defer_ok
637
76a2d7ba
PH
63875. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
639 those file descriptors could be used for SMTP connections. If anything
640 wrote to stderr (the example that came up was "warn" in embedded Perl), it
641 could be sent to the SMTP client, causing chaos. The daemon now opens
642 stdin, stdout, and stderr to /dev/null when it puts itself into the
643 background.
644
64576. Arrange for output from Perl's "warn" command to be written to Exim's main
646 log by default. The user can override this with suitable Perl magic.
647
04f7d5b9
PH
64877. The use of log_message on a "discard" ACL verb, which is supposed to add to
649 the log message when discard triggers, was not working for the DATA ACL or
650 for the non-SMTP ACL.
651
bc60667e
PH
65278. Error message wording change in sieve.c.
653
bb6e88ff
PH
65479. If smtp_accept_max_per_host was set, the number of connections could be
655 restricted to fewer than expected, because the daemon was trying to set up
656 a new connection before checking whether the processes handling previous
657 connections had finished. The check for completed processes is now done
658 earlier. On busy systems, this bug wouldn't be noticed because something
659 else would have woken the daemon, and it would have reaped the completed
660 process earlier.
661
1e70f85b
PH
66280. If a message was submitted locally by a user whose login name contained one
663 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
664 It caused a spool format error. I have fixed the spool reading code. A
665 related problem was that the "from" clause in the Received: line became
666 illegal because of the space(s). It is now covered by ${quote_local_part.
667
66881. Included the latest eximstats from Steve (adds average sizes to HTML Top
669 tables).
670
4e01f9d6
PH
67182. Updated OS/Makefile-AIX as per message from Mike Meredith.
672
1ee1cef2
PH
67383. Patch from Sieve maintainer to fix unterminated string problem in
674 "vacation" handling.
675
6e2b4ccc
PH
67684. Some minor changes to the Linux configuration files to help with other
677 OS variants using glibc.
678
8e669ac1
PH
67985. One more patch for Sieve to update vacation handling to latest spec.
680
495ae4b0 681
bbe902f0
PH
682----------------------------------------------------
683See the note above about the 4.44 and 4.50 releases.
684----------------------------------------------------
685
686
687Exim version 4.44
688-----------------
689
690 1. Change 4.43/35 introduced a bug that caused file counts to be
691 incorrectly computed when quota_filecount was set in an appendfile
692 transport
693
694 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
695 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
696 place.
697
698 3. Give more explanation in the error message when the command for a transport
699 filter fails to execute.
700
701 4. There are several places where Exim runs a non-Exim command in a
702 subprocess. The SIGUSR1 signal should be disabled for these processes. This
703 was being done only for the command run by the queryprogram router. It is
704 now done for all such subprocesses. The other cases are: ${run, transport
705 filters, and the commands run by the lmtp and pipe transports.
706
707 5. Some older OS have a limit of 256 on the maximum number of file
708 descriptors. Exim was using setrlimit() to set 1000 as a large value
709 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
710 systems. I've change it so that if it can't get 1000, it tries for 256.
711
712 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
713 was an oversight, and furthermore, ever since the addition of extra
714 controls (e.g. 4.43/32), the checks on when to allow different forms of
715 "control" were broken. There should now be diagnostics for all cases when a
716 control that does not make sense is encountered.
717
718 7. $recipients is now available in the predata ACL (oversight).
719
720 8. Tidy the search cache before the fork to do a delivery from a message
721 received from the command line. Otherwise the child will trigger a lookup
722 failure and thereby defer the delivery if it tries to use (for example) a
723 cached ldap connection that the parent has called unbind on.
724
725 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
726 of $address_data from the recipient verification was clobbered by the
727 sender verification.
728
72910. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
730 was its contents. (It was OK if the option was not defined at all.)
731
73211. A "Completed" log line is now written for messages that are removed from
733 the spool by the -Mrm option.
734
73512. $host_address is now set to the target address during the checking of
736 ignore_target_hosts.
737
73813. When checking ignore_target_hosts for an ipliteral router, no host name was
739 being passed; this would have caused $sender_host_name to have been used if
740 matching the list had actually called for a host name (not very likely,
741 since this list is usually IP addresses). A host name is now passed as
742 "[x.x.x.x]".
743
74414. Changed the calls that set up the SIGCHLD handler in the daemon to use the
745 code that specifies a non-restarting handler (typically sigaction() in
746 modern systems) in an attempt to fix a rare and obscure crash bug.
747
74815. Narrowed the window for a race in the daemon that could cause it to ignore
749 SIGCHLD signals. This is not a major problem, because they are used only to
750 wake it up if nothing else does.
751
75216. A malformed maildirsize file could cause Exim to calculate negative values
753 for the mailbox size or file count. Odd effects could occur as a result.
754 The maildirsize information is now recalculated if the size or filecount
755 end up negative.
756
75717. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
758 support for a long time. Removed HAVE_SYS_VFS_H.
759
ea3a6f44 76018. Updated exipick to current release from John Jetmore.
bbe902f0
PH
761
76219. Allow an empty sender to be matched against a lookup in an address list.
763 Previously the only cases considered were a regular expression, or an
764 empty pattern.
765
76620. Exim went into a mad DNS lookup loop when doing a callout where the
767 host was specified on the transport, if the DNS lookup yielded more than
768 one IP address.
769
ea3a6f44
NM
77021. The RFC2047 encoding function was originally intended for short strings
771 such as real names; it was not keeping to the 75-character limit for
772 encoded words that the RFC imposes. It now respects the limit, and
773 generates multiple encoded words if necessary. To be on the safe side, I
774 have increased the buffer size for the ${rfc2047: expansion operator from
775 1024 to 2048 bytes.
bbe902f0 776
ea3a6f44
NM
77722. Failure to deliver a bounce message always caused it to be frozen, even if
778 there was an errors_to setting on the router. The errors_to setting is now
779 respected.
bbe902f0
PH
780
78123. If an IPv6 address is given for -bh or -bhc, it is now converted to the
782 canonical form (fully expanded) before being placed in
783 $sender_host_address.
784
78524. Updated eximstats to version 1.33
786
ea3a6f44
NM
78725. Include certificate and key file names in error message when GnuTLS fails
788 to set them up, because the GnuTLS error message doesn't include the name
789 of the failing file when there is a problem reading it.
bbe902f0
PH
790
79126. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 792 As per change 25.
bbe902f0 793
ea3a6f44
NM
79427. Reset the locale to "C" after calling embedded Perl, in case it was changed
795 (this can affect the format of dates).
bbe902f0 796
ea3a6f44
NM
79728. exim_tidydb, when checking for the continued existence of a message for
798 which it has found a message-specific retry record, was not finding
799 messages that were in split spool directories. Consequently, it was
800 deleting retry records that should have stayed in existence.
bbe902f0
PH
801
80229. eximstats updated to version 1.35
803 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
804 1.35 - bugfix such that pie charts by volume are generated correctly
805
ea3a6f44
NM
80630. The SPA authentication driver was not abandoning authentication and moving
807 on to the next authenticator when an expansion was forced to fail,
808 contradicting the general specification for all authenticators. Instead it
809 was generating a temporary error. It now behaves as specified.
bbe902f0 810
ea3a6f44
NM
81131. The default ordering of permitted cipher suites for GnuTLS was pessimal
812 (the order specifies the preference for clients). The order is now AES256,
813 AES128, 3DES, ARCFOUR128.
bbe902f0 814
ea3a6f44
NM
81531. Small patch to Sieve code - explicitly set From: when generating an
816 autoreply.
bbe902f0 817
ea3a6f44
NM
81832. Exim crashed if a remote delivery caused a very long error message to be
819 recorded - for instance if somebody sent an entire SpamAssassin report back
820 as a large number of 550 error lines. This bug was coincidentally fixed by
821 increasing the size of one of Exim's internal buffers (big_buffer) that
822 happened as part of the Exiscan merge. However, to be on the safe side, I
823 have made the code more robust (and fixed the comments that describe what
824 is going on).
bbe902f0 825
ea3a6f44
NM
82633. Some experimental protocols are using DNS PTR records for new purposes. The
827 keys for these records are domain names, not reversed IP addresses. The
828 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
829 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
830 CAN-2005-0021
831
ea3a6f44
NM
83234. The host_aton() function is supposed to be passed a string that is known
833 to be a valid IP address. However, in the case of IPv6 addresses, it was
834 not checking this. This is a hostage to fortune. Exim now panics and dies
835 if the condition is not met. A case was found where this could be provoked
836 from a dnsdb PTR lookup with an IPv6 address that had more than 8
837 components; fortuitously, this particular loophole had already been fixed
838 by change 4.50/55 or 4.44/33 above.
839
840 If there are any other similar loopholes, the new check in host_aton()
841 itself should stop them being exploited. The report I received stated that
842 data on the command line could provoke the exploit when Exim was running as
843 exim, but did not say which command line option was involved. All I could
844 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
845 running as the user.
bbe902f0
PH
846 CAN-2005-0021
847
ea3a6f44
NM
84835. There was a buffer overflow vulnerability in the SPA authentication code
849 (which came originally from the Samba project). I have added a test to the
850 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
851 CAN-2005-0022
852
ea3a6f44
NM
85336. The daemon start-up calls getloadavg() while still root for those OS that
854 need the first call to be done as root, but it missed one case: when
855 deliver_queue_load_max is set with deliver_drop_privilege. This is
856 necessary for the benefit of the queue runner, because there is no re-exec
857 when deliver_drop_privilege is set.
bbe902f0 858
ea3a6f44
NM
85937. Caching of lookup data for "hosts =" ACL conditions, when a named host list
860 was in use, was not putting the data itself into the right store pool;
861 consequently, it could be overwritten for a subsequent message in the same
862 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
863 the caching.)
bbe902f0 864
ea3a6f44
NM
86538. Sometimes the final signoff response after QUIT could fail to get
866 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
867 < 0 before doing a fflush(). This bug looks as though it goes back to the
868 introduction of TLS in release 3.20, but "sometimes" must have been rare
869 because the tests only now provoked it.
bbe902f0
PH
870
871
495ae4b0
PH
872Exim version 4.43
873-----------------
874
875 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
876 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
877 Now it returns an integer. A number of calls were still expecting a T/F
878 return. Fortuitously, in all cases, the tests worked in OK situations,
879 which is the norm. However, things would have gone wrong on any write
880 failures on the smtp file descriptor. This function is used when sending
881 messages over SMTP and also when doing verify callouts.
882
883 2. When Exim is called to do synchronous delivery of a locally submitted
884 message (the -odf or -odi options), it no longer closes stderr before doing
885 the delivery.
886
887 3. Implemented the mua_wrapper option.
888
889 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
890
891 5. Implemented the functions header_remove(), header_testname(),
892 header_add_at_position(), and receive_remove_recipient(), and exported them
893 to local_scan().
894
895 6. If an ACL "warn" statement specified the addition of headers, Exim already
896 inserted X-ACL-Warn: at the start if there was no header name. However, it
897 was not making this test for the second and subsequent header lines if
898 there were newlines in the string. This meant that an invalid header could
899 be inserted if Exim was badly configured.
900
901 7. Allow an ACL "warn" statement to add header lines at the start or after all
902 the Received: headers, as well as at the end.
903
904 8. Added the rcpt_4xx retry error code.
905
906 9. Added postmaster_mailfrom=xxx to callout verification option.
907
90810. Added mailfrom=xxxx to the callout verification option, for verify=
909 header_sender only.
910
91111. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
912 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
913
91412. Inserted some casts to stop certain compilers warning when using pointer
915 differences as field lengths or precisions in printf-type calls (mostly
916 affecting debugging statements).
917
91813. Added optional readline() support for -be (dynamically loaded).
919
92014. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
921 same clock tick as a message's arrival, so that its received time was the
922 same as the "first fail" time on the retry record, and that message
923 remained on the queue past the ultimate address timeout, every queue runner
924 would try a delivery (because it was past the ultimate address timeout) but
925 after another failure, the ultimate address timeout, which should have then
926 bounced the address, did not kick in. This was a "< instead of <=" error;
927 in most cases the first failure would have been in the next clock tick
928 after the received time, and all would be well.
929
93015. The special items beginning with @ in domain lists (e.g. @mx_any) were not
931 being recognized when the domain list was tested by the match_domain
932 condition in an expansion string.
933
93416. Added the ${str2b64: operator.
935
93617. Exim was always calling setrlimit() to set a large limit for the number of
937 processes, without checking whether the existing limit was already
938 adequate. (It did check for the limit on file descriptors.) Furthermore,
939 errors from getrlimit() and setrlimit() were being ignored. Now they are
940 logged to the main and panic logs, but Exim does carry on, to try to do its
941 job under whatever limits there are.
942
94318. Imported PCRE 5.0.
944
94519. Trivial typo in log message " temporarily refused connection" (the leading
946 space).
947
94820. If the log selector return_path_on_delivery was set and an address was
949 redirected to /dev/null, the delivery process crashed because it assumed
950 that a return path would always be set for a "successful" delivery. In this
951 case, the whole delivery is bypassed as an optimization, and therefore no
952 return path is set.
953
95421. Internal re-arrangement: the function for sending a challenge and reading
955 a response while authentication was assuming a zero-terminated challenge
956 string. It's now changed to take a pointer and a length, to allow for
957 binary data in such strings.
958
95922. Added the cyrus_sasl authenticator (code supplied by MBM).
960
96123. Exim was not respecting finduser_retries when seeking the login of the
962 uid under which it was called; it was always trying 10 times. (The default
963 setting of finduser_retries is zero.) Also, it was sleeping after the final
964 failure, which is pointless.
965
96624. Implemented tls_on_connect_ports.
967
96825. Implemented acl_smtp_predata.
969
97026. If the domain in control=submission is set empty, Exim assumes that the
971 authenticated id is a complete email address when it generates From: or
972 Sender: header lines.
973
97427. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
975 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
976 chown and chgrp in /bin and hostname in /usr/bin.
977
97828. Exim was keeping the "process log" file open after each use, just as it
979 does for the main log. This opens the possibility of it remaining open for
980 long periods when the USR1 signal hits a daemon. Occasional processlog
981 errors were reported, that could have been caused by this. Anyway, it seems
982 much more sensible not to leave this file open at all, so that is what now
983 happens.
984
98529. The long-running daemon process does not normally write to the log once it
986 has entered its main loop, and it closes the log before doing so. This is
987 so that log files can straightforwardly be renamed and moved. However,
988 there are a couple of unusual error situations where the daemon does write
989 log entries, and I had neglected to close the log afterwards.
990
99130. The text of an SMTP error response that was received during a remote
992 delivery was being truncated at 512 bytes. This is too short for some of
993 the long messages that one sometimes sees. I've increased the limit to
994 1024.
995
99631. It is now possible to make retry rules that apply only when a message has a
997 specific sender, in particular, an empty sender.
998
99932. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1000 it possible to be selective about when SMTP synchronization is enforced.
1001
100233. Added "control = caseful_local_part" and "control = "caselower_local_part".
1003
100432. Implemented hosts_connection_nolog.
1005
100633. Added an ACL for QUIT.
1007
100834. Setting "delay_warning=" to disable warnings was not working; it gave a
1009 syntax error.
1010
101135. Added mailbox_size and mailbox_filecount to appendfile.
1012
101336. Added control = no_multiline_responses to ACLs.
1014
101537. There was a bug in the logic of the code that waits for the clock to tick
1016 in the case where the clock went backwards by a substantial amount such
1017 that the microsecond fraction of "now" was more than the microsecond
1018 fraction of "then" (but the whole seconds number was less).
1019
102038. Added support for the libradius Radius client library this is found on
1021 FreeBSD (previously only the radiusclient library was supported).
1022
1023
1024Exim version 4.42
1025-----------------
1026
1027 1. When certain lookups returned multiple values in the form name=value, the
1028 quoting of the values was not always being done properly. Specifically:
1029 (a) If the value started with a double quote, but contained no whitespace,
1030 it was not quoted.
1031 (b) If the value contained whitespace other than a space character (i.e.
1032 tabs or newlines or carriage returns) it was not quoted.
1033 This fix has been applied to the mysql and pgsql lookups by writing a
1034 separate quoting function and calling it from the lookup code. The fix
1035 should probably also be applied to nisplus, ibase and oracle lookups, but
1036 since I cannot test any of those, I have not disturbed their existing code.
1037
1038 2. A hit in the callout cache for a specific address caused a log line with no
1039 reason for rejecting RCPT. Now it says "Previous (cached) callout
1040 verification failure".
1041
1042 3. There was an off-by-one bug in the queryprogram router. An over-long
1043 return line was truncated at 256 instead of 255 characters, thereby
1044 overflowing its buffer with the terminating zero. As well as fixing this, I
1045 have increased the buffer size to 1024 (and made a note to document this).
1046
1047 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1048 when Exim is waiting for an SMTP response from a remote server, Exim
1049 restarts its select() call on the socket, thereby resetting its timeout.
1050 This is not a problem when such interrupts are rare. Somebody set up a cron
1051 job to run exiwhat every 2 minutes, which is less than the normal select()
1052 timeout (5 or 10 minutes). This meant that the select() timeout never
1053 kicked in because it was always reset. I have fixed this by comparing the
1054 time when an interrupt arrives with the time at the start of the first call
1055 to select(). If more time than the timeout has elapsed, the interrupt is
1056 treated as a timeout.
1057
1058 5. Some internal re-factoring in preparation for the addition of Sieve
1059 extensions (by MH). In particular, the "personal" test is moved to a
1060 separate function, and given an option for scanning Cc: and Bcc: (which is
1061 not set for Exim filters).
1062
1063 6. When Exim created an email address using the login of the caller as the
1064 local part (e.g. when creating a From: or Sender: header line), it was not
1065 quoting the local part when it contained special characters such as @.
1066
1067 7. Installed new OpenBSD configuration files.
1068
1069 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1070 try to make them clearer.
1071
1072 9. Callout options, other than the timeout value, were being ignored when
1073 verifying sender addresses in header lines. For example, when using
1074
1075 verify = header_sender/callout=no_cache
1076
1077 the cache was (incorrectly) being used.
1078
107910. Added a missing instance of ${EXE} to the exim_install script; this affects
1080 only the Cygwin environment.
1081
108211. When return_path_on_delivery was set as a log selector, if different remote
1083 addresses in the same message used different return paths and parallel
1084 remote delivery occurred, the wrong values would sometimes be logged.
1085 (Whenever a remote delivery process finished, the return path value from
1086 the most recently started remote delivery process was logged.)
1087
108812. RFC 3848 specifies standard names for the "with" phrase in Received: header
1089 lines when AUTH and/or TLS are in use. This is the "received protocol"
1090 field. Exim used to use "asmtp" for authenticated SMTP, without any
1091 indication (in the protocol name) for TLS use. Now it follows the RFC and
1092 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1093 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1094 names appear in log lines as well as in Received: header lines.
1095
109613. Installed MH's patches for Sieve to add the "copy" and "vacation"
1097 extensions, and comparison tests, and to fix some bugs.
1098
109914. Changes to the "personal" filter test:
1100
1101 (1) The test was buggy in that it was just doing the equivalent of
1102 "contains" tests on header lines. For example, if a user's address was
1103 anne@some.where, the "personal" test would incorrectly be true for
1104
1105 To: susanne@some.where
1106
1107 This test is now done by extracting each address from the header in turn,
1108 and checking the entire address. Other tests that are part of "personal"
1109 are now done using regular expressions (for example, to check local parts
1110 of addresses in From: header lines).
1111
1112 (2) The list of non-personal local parts in From: addresses has been
1113 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1114 taken from the Sieve specification recommendations.
1115
1116 (3) If the message contains any header line starting with "List-" it is
1117 treated as non-personal.
1118
1119 (4) The test for "circular" in the Subject: header line has been removed
1120 because it now seems ill-conceived.
1121
112215. Minor typos in src/EDITME comments corrected.
1123
112416. Installed latest exipick from John Jetmore.
1125
112617. If headers_add on a router specified a text string that was too long for
1127 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1128 of string_sprintf() is now avoided.
1129
113018. $message_body_size was not set (it was always zero) when running the DATA
1131 ACL and the local_scan() function.
1132
113319. For the "mail" command in an Exim filter, no default was being set for
1134 the once_repeat time, causing a random time value to be used if "once" was
1135 specified. (If the value happened to be <= 0, no repeat happened.) The
1136 default is now 0s, meaning "never repeat". The "vacation" command was OK
1137 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1138 (I found it when inspecting the code).
1139
114020. There is now an overall timeout for performing a callout verification. It
1141 defaults to 4 times the callout timeout, which applies to individual SMTP
1142 commands during the callout. The overall timeout applies when there is more
1143 than one host that can be tried. The timeout is checked before trying the
1144 next host. This prevents very long delays if there are a large number of
1145 hosts and all are timing out (e.g. when the network connections are timing
1146 out). The value of the overall timeout can be changed by specifying an
1147 additional sub-option for "callout", called "maxwait". For example:
1148
1149 verify = sender/callout=5s,maxwait=20s
1150
115121. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1152 to the end before writing, but this should make it even safer).
1153
115422. Exim was forgetting that it had advertised PIPELINING for the second and
1155 subsequent messages on an SMTP connection. It was also not resetting its
1156 memory on STARTTLS and an internal HELO.
1157
115823. When Exim logs an SMTP synchronization error within a session, it now
1159 records whether PIPELINING has been advertised or not.
1160
116124. Added 3 instances of "(long int)" casts to time_t variables that were being
1162 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1163 rather than long int.
1164
116525. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1166
116726. Added the never_mail option to autoreply.
1168
1169
1170Exim version 4.41
1171-----------------
1172
1173 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1174 crash if the getsockname() call failed; this can happen if a connection is
1175 closed very soon after it is established. The problem was simply in the
1176 order in which certain operations were done, causing Exim to try to write
1177 to the SMTP stream before it had set up the file descriptor. The bug has
1178 been fixed by making things happen in the correct order.
1179
1180
1181Exim version 4.40
1182-----------------
1183
1184 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1185 before the connection was closed, thus losing the rejection response.
1186
1187 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1188 some early Solaris releases, but causes trouble in current releases where
1189 socklen_t is defined.
1190
1191 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1192 always exist.
1193
1194 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1195 configured.
1196
1197 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1198 that releasing the top of it at the end releases what was used for sub-
1199 expansions (unless the block got too big). However, discard this block if
1200 the first thing is a variable or header, so that we can use its block when
1201 it is dynamic (useful for very large $message_headers, for example).
1202
1203 6. Lookups now cache *every* query, not just the most recent. A new, separate
1204 store pool is used for this. It can be recovered when all lookup caches are
1205 flushed. Lookups now release memory at the end of their result strings.
1206 This has involved some general refactoring of the lookup sources.
1207
1208 7. Some code has been added to the store_xxx() functions to reduce the amount
1209 of flapping under certain conditions.
1210
1211 8. log_incoming_interface used to affect only the <= reception log lines. Now
1212 it causes the local interface and port to be added to several more SMTP log
1213 lines, for example "SMTP connection from", and rejection lines.
1214
1215 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1216
121710. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1218
121911. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1220 could be overwritten at the end of the current message (or the start of a
1221 new message if it was set in a HELO ACL). The value is now preserved for
1222 the duration of the SMTP connection.
1223
122412. If a transport had a headers_rewrite setting, and a matching header line
1225 contained an unqualified address, that address was qualified, even if it
1226 did not match any rewriting rules. The underlying bug was that the values
1227 of the flags that permit the existence of unqualified sender and recipient
1228 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1229 non-local messages, and by -bnq for local messages) were not being
1230 preserved with the message after it was received.
1231
123213. When Exim was logging an SMTP synchronization error, it could sometimes log
1233 "next input=" as part of the text comprising the host identity instead of
1234 the correct text. The code was using the same buffer for two different
1235 strings. However, depending on which order the printing function evaluated
1236 its arguments, the bug did not always show up. Under Linux, for example, my
1237 test suite worked just fine.
1238
123914. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1240 doesn't work with some older versions of Perl. It has been changed to "my",
1241 which in any case is probably the better facility to use.
1242
124315. A really picky compiler found some instances of statements for creating
1244 error messages that either had too many or two few arguments for the format
1245 string.
1246
124716. The size of the buffer for calls to the DNS resolver has been increased
1248 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1249 for addresses that have a lot of PTR records. This alleviates a problem; it
1250 does not fully solve it.
1251
125217. A dnsdb lookup for PTR records that receives more data than will fit in the
1253 buffer now truncates the list and logs the incident, which is the same
1254 action as happens when Exim is looking up a host name and its aliases.
1255 Previously in this situation something unpredictable would happen;
1256 sometimes it was "internal error: store_reset failed".
1257
125818. If a server dropped the connection unexpectedly when an Exim client was
1259 using GnuTLS and trying to read a response, the client delivery process
1260 crashed while trying to generate an error log message.
1261
126219. If a "warn" verb in an ACL added multiple headers to a message in a single
1263 string, for example:
1264
1265 warn message = H1: something\nH2: something
1266
1267 the text was added as a single header line from Exim's point of view
1268 though it ended up OK in the delivered message. However, searching for the
1269 second and subsequent header lines using $h_h2: did not work. This has been
1270 fixed. Similarly, if a system filter added multiple headers in this way,
1271 the routers could not see them.
1272
127320. Expanded the error message when iplsearch is called with an invalid key to
1274 suggest using net-iplsearch in a host list.
1275
127621. When running tests using -bh, any delays imposed by "delay" modifiers in
1277 ACLs are no longer actually imposed (and a message to that effect is
1278 output).
1279
128022. If a "gecos" field in a passwd entry contained escaped characters, in
1281 particular, if it contained a \" sequence, Exim got it wrong when building
1282 a From: or a Sender: header from that name. A second bug also caused
1283 incorrect handling when an unquoted " was present following a character
1284 that needed quoting.
1285
128623. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1287 was not being matched caselessly.
1288
128924. Arranged for all hyphens in the exim.8 source to be escaped with
1290 backslashes.
1291
129225. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1293 properly. Recipient callout cache records were still being keyed to include
1294 the sender, even when use_sender was set false. This led to far more
1295 callouts that were necessary. The sender is no longer included in the key
1296 when use_sender is false.
1297
129826. Added "control = submission" modifier to ACLs.
1299
130027. Added the ${base62d: operator to decode base 62 numbers.
1301
130228. dnsdb lookups can now access SRV records.
1303
130429. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1305 the configuration file.
1306
130730. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1308 (-v) mode. This makes the output for a verbose queue run more intelligible.
1309
131031. Added a use_postmaster feature to recipient callouts.
1311
131232. Added the $body_zerocount variable, containing the number of binary zero
1313 bytes in the message body.
1314
131533. The time of last modification of the "new" subdirectory is now used as the
1316 "mailbox time last read" when there is a quota error for a maildir
1317 delivery.
1318
131934. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1320
132135. Added +ignore_unknown as a special item in host lists.
1322
132336. Code for decoding IPv6 addresses in host lists is now included, even if
1324 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1325 address was recognized as an IP address, but was then not correctly decoded
1326 into binary, causing unexpected and incorrect effects when compared with
1327 another IP address.
1328
1329
1330Exim version 4.34
1331-----------------
1332
1333 1. Very minor rewording of debugging text in manualroute to say "list of
1334 hosts" instead of "hostlist".
1335
1336 2. If verify=header_syntax was set, and a header line with an unqualified
1337 address (no domain) and a large number of spaces between the end of the
1338 name and the colon was received, the reception process suffered a buffer
1339 overflow, and (when I tested it) crashed. This was caused by some obsolete
1340 code that should have been removed. The fix is to remove it!
1341
1342 3. When running in the test harness, delay a bit after writing a bounce
1343 message to get a bit more predictability in the log output.
1344
1345 4. Added a call to search_tidyup() just before forking a reception process. In
1346 theory, someone could use a lookup in the expansion of smtp_accept_max_
1347 per_host which, without the tidyup, could leave open a database connection.
1348
1349 5. Added the variables $recipient_data and $sender_data which get set from a
1350 lookup success in an ACL "recipients" or "senders" condition, or a router
1351 "senders" option, similar to $domain_data and $local_part_data.
1352
1353 6. Moved the writing of debug_print from before to after the "senders" test
1354 for routers.
1355
1356 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1357 problems for message scanning, either using a data ACL, or using
1358 local_scan() because the Received: header was not generated till after they
1359 were called (in order to set the time as the time of reception completion).
1360 I have revised the way this works. The header is now generated after the
1361 body is received, but before the ACL or local_scan() are called. After they
1362 are run, the timestamp in the header is updated.
1363
1364
1365Exim version 4.33
1366-----------------
1367
1368 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1369 before starting a queue runner without re-exec. This happened only when
1370 deliver_drop_privilege was set or when the Exim user was set to root. The
1371 effect of the bug was that timeouts during subsequent deliveries caused
1372 crashes instead of being properly handled. The handler is now left at its
1373 default (and expected) setting.
1374
1375 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1376 message, again when deliver_drop_privilege is set or Exim is run as root.
1377 The bug described in (1) was not present in this case, but the tidying up
1378 of the other signals was missing. I have made the two cases consistent.
1379
1380 3. The ignore_target_hosts setting on a manualroute router was being ignored
1381 for hosts that were looked up using the /MX notation.
1382
1383 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1384 in domain lists.
1385
1386 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1387 operated on the sender address. After changing the $sender_address to <>
1388 for the sender address verify, Exim was re-instated it as the original
1389 (before rewriting) address, but remembering that it had rewritten it, so it
1390 wasn't rewriting it again. This bug also had the effect of breaking the
1391 sender address verification caching when the sender address was rewritten.
1392
1393 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1394 This has been changed so that if the ip literal address matches
1395 ignore_target_hosts, the router declines.
1396
1397 7. Added expansion conditions match_domain, match_address, and match_local_
1398 part (NOT match_host).
1399
1400 8. The placeholder for the Received: header didn't have a length field set.
1401
1402 9. Added code to Exim itself and to exim_lock to test for a specific race
1403 condition that could lead to file corruption when using MBX delivery. The
1404 issue is with the lockfile that is created in /tmp. If this file is removed
1405 after a process has opened it but before that process has acquired a lock,
1406 there is the potential for a second process to recreate the file and also
1407 acquire a lock. This could lead to two Exim processes writing to the file
1408 at the same time. The added code performs the same test as UW imapd; it
1409 checks after acquiring the lock that its file descriptor still refers to
1410 the same named file.
1411
141210. The buffer for building added header lines was of fixed size, 8192 bytes.
1413 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1414 when Exim is built.
1415
141611. Added the smtp_active_hostname option. If used, this will typically be made
1417 to depend on the incoming interface address. Because $interface_address is
1418 not set up until the daemon has forked a reception process, error responses
1419 that can happen earlier (such as "too many connections") no longer contain
1420 a host name.
1421
142212. If an expansion in a condition on a "warn" statement fails because a lookup
1423 defers, the "warn" statement is abandoned, and the next ACL statement is
1424 processed. Previously this caused the whole ACL to be aborted.
1425
142613. Added the iplsearch lookup type.
1427
142814. Added ident_timeout as a log selector.
1429
143015. Added tls_certificate_verified as a log selector.
1431
143216. Added a global option tls_require_ciphers (compare the smtp transport
1433 option of the same name). This controls incoming TLS connections.
1434
143517. I finally figured out how to make tls_require_ciphers do a similar thing
1436 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1437 before starting the TLS session.
1438
143918. Tabs are now shown as \t in -bP output.
1440
144119. If the log selector return_path_on_delivery was set, Exim crashed when
1442 bouncing a message because it had too many Received: header lines.
1443
144420. If two routers both had headers_remove settings, and the first one included
1445 a superfluous trailing colon, the final name in the first list and the
1446 first name in the second list were incorrectly joined into one item (with a
1447 colon in the middle).
1448
1449
1450Exim version 4.32
1451-----------------
1452
1453 1. Added -C and -D options to the exinext utility, mainly to make it easier
1454 to include in the automated testing, but these could be helpful when
1455 multiple configurations are in use.
1456
1457 2. The exinext utility was not formatting the output nicely when there was
1458 an alternate port involved in the retry record key, nor when there was a
1459 message id as well (for retries that were specific to a specific message
1460 and a specific host). It was also confused by IPv6 addresses, because of
1461 the additional colons they contain. I have fixed the IPv4 problem, and
1462 patched it up to do a reasonable job for IPv6.
1463
1464 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1465 delivery, the log line now contains "pipelined" if PIPELINING was used.
1466
1467 4. An SMTP transport process used to panic and die if the bind() call to set
1468 an explicit outgoing interface failed. This has been changed; it is now
1469 treated in the same way as a connect() failure.
1470
1471 5. A reference to $sender_host_name in the part of a conditional expansion
1472 that was being skipped was still causing a DNS lookup. This no longer
1473 occurs.
1474
1475 6. The def: expansion condition was not recognizing references to header lines
1476 that used bh_ and bheader_.
1477
1478 7. Added the _cache feature to named lists.
1479
1480 8. The code for checking quota_filecount in the appendfile transport was
1481 allowing one more file than it should have been.
1482
1483 9. For compatibility with Sendmail, the command line option
1484
1485 -prval:sval
1486
1487 is equivalent to
1488
1489 -oMr rval -oMs sval
1490
1491 and sets the incoming protocol and host name (for trusted callers). The
1492 host name and its colon can be omitted when only the protocol is to be set.
1493 Note the Exim already has two private options, -pd and -ps, that refer to
1494 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1495 "s", but I don't think that's a major issue.
1496
149710. A number of refactoring changes to the code, none of which should affect
1498 Exim's behaviour:
1499
1500 (a) The number of logging options was getting close to filling up the
1501 32-bit word that was used as a bit map. I have split them into two classes:
1502 those that are passed in the argument to log_write(), and those that are
1503 only ever tested independently outside of that function. These are now in
1504 separate 32-bit words, so there is plenty of room for expansion again.
1505 There is no change in the user interface or the logging behaviour.
1506
1507 (b) When building, for example, log lines, the code previously used a
1508 macro that called string_cat() twice, in order to add two strings. This is
1509 not really sufficiently general. Furthermore, there was one instance where
1510 it was actually wrong because one of the argument was used twice, and in
1511 one call a function was used. (As it happened, calling the function twice
1512 did not affect the overall behaviour.) The macro has been replaced by a
1513 function that can join an arbitrary number of extra strings onto a growing
1514 string.
1515
1516 (c) The code for expansion conditions now uses a table and a binary chop
1517 instead of a serial search (which was left over from when there were very
1518 few conditions). Also, it now recognizes conditions like "pam" even when
1519 the relevant support is not compiled in: a suitably worded error message is
1520 given if an attempt is made to use such a condition.
1521
152211. Added ${time_interval:xxxxx}.
1523
152412. A bug was causing one of the ddress fields not to be passed back correctly
1525 from remote delivery subprocesses. The field in question was not being
1526 subsequently used, so this caused to problems in practice.
1527
152813. Added new log selectors queue_time and deliver_time.
1529
153014. Might have fixed a bug in maildirsizefile handling that threw up
1531 "unexpected character" debug warnings, and recalculated the data
1532 unnecessarily. In any case, I expanded the warning message to give more
1533 information.
1534
153515. Added the message "Restricted characters in address" to the statements in
1536 the default ACL that block characters like @ and % in local parts.
1537
153816. Change 71 for release 4.31 proved to be much less benign that I imagined.
1539 Three changes have been made:
1540
1541 (a) There was a serious bug; a negative response to MAIL caused the whole
1542 recipient domain to be cached as invalid, thereby blocking all messages
1543 to all local parts at the same domain, from all senders. This bug has
1544 been fixed. The domain is no longer cached after a negative response to
1545 MAIL if the sender used is not empty.
1546
1547 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1548 been restored.
1549
1550 (c) A new callout option, "use_sender" has been added for people who want
1551 the modified behaviour.
1552
1553
1554Exim version 4.31
1555-----------------
1556
1557 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1558 Larry Rosenman.
1559
1560 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1561 indeed breaks things for older releases.
1562
1563 3. Added additional logging to the case where there is a problem reading data
1564 from a filter that is running in a subprocess using a pipe, in order to
1565 try to track down a specific problem.
1566
1567 4. Testing facility fudge: when running in the test harness and attempting
1568 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1569 getting "No route to host". Convert this to a timeout.
1570
1571 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1572 warning.
1573
1574 6. Some OS don't have socklen_t but use size_t instead. This affects the
1575 fifth argument of getsockopt() amongst other things. This is now
1576 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1577 can be set for individual OS. I have set it for SunOS5, OSF1, and
1578 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1579 some earlier ones do not.
1580
1581 7. Change 4.30/15 was not doing the test caselessly.
1582
1583 8. The standard form for an IPv6 address literal was being rejected by address
1584 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1585 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1586 this, as well as the form without the "IPv6" on the front (but only when
1587 address literals are enabled, of course).
1588
1589 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1590
159110. Exim crashed if a message with an empty sender address specified by -f
1592 encountered a router with an errors_to setting. This could be provoked only
1593 by a command such as
1594
1595 exim -f "" ...
1596
1597 where an empty string was supplied; "<>" did not hit this bug.
1598
159911. Installed PCRE release 4.5.
1600
160112. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1602 remained set. It is now erased.
1603
160413. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1605 times from message ids (which are base 36 rather than the normal 62).
1606
160714. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1608 were being counted as actual protocol errors, and logged if the log
1609 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1610 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1611 and DATA following a set of rejected RCPTs do not count as protocol errors.
1612 In other words, Exim assumes they were pipelined, though this may not
1613 actually be the case. Of course, in all cases the client gets an
1614 appropriate error code.
1615
161615. If a lookup fails in an ACL condition, a message about the failure may
1617 be available; it is used if testing the ACL cannot continue, because most
1618 such messages specify what the cause of the deferral is. However, some
1619 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1620 that caused an old message to be retained and used if a later statement
1621 caused a defer, replacing the real cause of the deferral.
1622
162316. If an IP address had so many PTR records that the DNS lookup buffer
1624 was not large enough to hold them, Exim could crash while trying to process
1625 the truncated data. It now detects and logs this case.
1626
162717. Further to 4.21/58, another change has been made: if (and only if) the
1628 first line of a message (the first header line) ends with CRLF, a bare LF
1629 in a subsequent header line has a space inserted after it, so as not to
1630 terminate the header.
1631
163218. Refactoring: tidied an ugly bit of code in appendfile that copied data
1633 unnecessarily, used atoi() instead of strtol(), and didn't check the
1634 termination when getting file sizes from file names by regex.
1635
163619. Completely re-implemented the support for maildirsize files, in the light
1637 of a number of problems with the previous contributed implementation
1638 (4.30/29). In particular:
1639
1640 . If the quota is zero, the maildirsize file is maintained, but no quota is
1641 imposed.
1642
1643 . If the maildir directory does not exist, it is created before any attempt
1644 to write a maildirsize file.
1645
1646 . The quota value in the file is just a cache; if the quota is changed in
1647 the transport, the new value overrides.
1648
1649 . A regular expression is available for excluding directories from the
1650 count.
1651
165220. The autoreply transport checks the characters in options that define the
1653 message's headers; it allows continued headers, but it was checking with
1654 isspace() after an embedded newline instead of explicitly looking for a
1655 space or a tab.
1656
165721. If all the "regular" hosts to which an address was routed had passed their
1658 expiry times, and had not reached their retry times, the address was
1659 bounced, even if fallback hosts were defined. Now Exim should go on to try
1660 the fallback hosts.
1661
166222. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1663 equivalent code in the SMTP transport. Some hosts send humungous responses
1664 to HELO/EHLO, more than 1024 it seems.
1665
166623. Refactoring: code in filter.c used (void *) for "any old type" but this
1667 gives compiler warnings in some environments. I've now done it "properly",
1668 using a union.
1669
167024. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1671 (because of problems with the built-in one) was declared to return uschar *
1672 instead of char *, causing compiler failure.
1673
167425. Fixed a file descriptor leak when processing alias/forward files.
1675
167626. Fixed a minor format string issue in dbfn.c.
1677
167827. Typo in exim.c: ("dmbnz" for "dbmnz").
1679
168028. If a filter file refered to $h_xxx or $message_headers, and the headers
1681 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1682 become corrupted.
1683
168429. When a sender address is verified, it is cached, to save repeating the test
1685 when there is more than one recipient in a message. However, when the
1686 verification involves a callout, it is possible for different callout
1687 options to be set for different recipients. It is too complicated to keep
1688 track of this in the cache, so now Exim always runs a verification when a
1689 callout is required, relying on the callout cache for the optimization.
1690 The overhead is duplication of the address routing, but this should not be
1691 too great.
1692
169330. Fixed a bug in callout caching. If a RCPT command caused the sender address
1694 to be verified with callout=postmaster, and the main callout worked but the
1695 postmaster check failed, the verification correctly failed. However, if a
1696 subsequent RCPT command asked for sender verification *without* the
1697 postmaster check, incorrect caching caused this verification also to fail,
1698 incorrectly.
1699
170031. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1701 it was not caching the DNS options (qualify_single, search_parents) that
1702 were used when the lookup failed. A subsequent lookup with different
1703 options therefore always gave the same answer, though there were cases
1704 where it should not have. (Example: a "domains = !$mx_any" option on a
1705 dnslookup router: the "domains" option is always processed without any
1706 widening, but the router might have qualify_single set.) Now Exim uses the
1707 cached value only when the same options are set.
1708
170932. Added John Jetmore's "exipick" utility to the distribution.
1710
171133. GnuTLS: When an attempt to start a TLS session fails for any reason other
1712 than a timeout (e.g. a certificate is required, and is not provided), an
1713 Exim server now closes the connection immediately. Previously it waited for
1714 the client to close - but if the client is SSL, it seems that they each
1715 wait for each other, leading to a delay before one of them times out.
1716
171734: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1718 maintained 0.8.x compatibility because I don't think many are using it, and
1719 it is clearly obsolete.
1720
172135. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1722 transport.
1723
172436. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1725 client certificate was expired. A simple patch fixes this, though I don't
1726 understand the full logic of why the verify callback is called multiple
1727 times.
1728
172937. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1730 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1731 which causes problems with some clients (such as the Certicom SSL Plus
1732 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1733 disables the coutermeasure allowing Eudora to connect."
1734
173538. Exim was not checking that a write() to a log file succeeded. This could
1736 lead to Bad Things if a log got too big, in particular if it hit a file
1737 size limit. Exim now panics and dies if it cannot write to a log file, just
1738 as it does if it cannot open a log file.
1739
174039. Modified OS/Makefile-Linux so that it now contains
1741
1742 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1743
1744 The two -D definitions ensure that Exim is compiled with large file
1745 support, which makes it possible to handle log files that are bigger than
1746 2^31.
1747
174840. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1749 instance) a domain was checked against a named list that involved a lookup,
1750 causing $domain_data to be set, then another domain was checked against the
1751 same list, then the first domain was re-checked, the value of $domain_data
1752 after the final check could be wrong. In particular, if the second check
1753 failed, it could be set empty. This bug probably also applied to
1754 $localpart_data.
1755
175641. The strip_trailing_dot option was not being applied to the address given
1757 with the -f command-line option.
1758
175942. The code for reading a message's header from the spool was incrementing
1760 $received_count, but never initializing it. This meant that the value was
1761 incorrect (doubled) while delivering a message in the same process in which
1762 it was received. In the most common configuration of Exim, this never
1763 happens - a fresh exec is done - but it can happen when
1764 deliver_drop_privilege is set.
1765
176643. When Exim logs an SMTP synchronization error - client data sent too soon -
1767 it now includes up to 150 characters of the unexpected data in the log
1768 line.
1769
177044. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1771 and building data strings. The size of both of these buffers was 10 000
1772 bytes - far larger than anybody would *ever* want, thought I. Needless to
1773 say, somebody hit the limit. I have increased the maximum line length to
1774 20 000 and the maximum data length of concatenated lines to 100 000. I have
1775 also fixed two bugs, because there was no checking on these buffers. Tsk,
1776 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1777 buffer is too small.
1778
177945. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1780 lsearch lookups. Now it does.
1781
178246. When parsing a route_list item in a manualroute router, a fixed-length
1783 buffer was used for the list of hosts. I made this 1024 bytes long,
1784 thinking that nobody would ever have a list of hosts that long. Wrong.
1785 Somebody had a whole pile of complicated expansion conditions, and the
1786 string was silently truncated, leading to an expansion error. It turns out
1787 that it is easier to change to an unlimited length (owing to other changes
1788 that have happened since this code was originally written) than to build
1789 structure for giving a limitation error. The length of the item that
1790 expands into the list of hosts is now unlimited.
1791
179247. The lsearch lookup could not handle data where the length of text line was
1793 more than 4095 characters. Such lines were truncated, leading to shortened
1794 data being returned. It should now handle lines of any length.
1795
179648. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1797 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1798 ACL").
1799
180049. Cosmetic tidy to scripts like exicyclog that are generated by globally
1801 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1802 no longer happens in comment lines. A list of replacements is now placed
1803 at the head of all of the source files, except those whose only change is
1804 to replace PERL_COMMAND in the very first #! line.
1805
180650. Replaced the slow insertion sort in queue.c, for sorting the list of
1807 messages on the queue, with a bottom-up merge sort, using code contributed
1808 by Michael Haardt. This should make operations like -bp somewhat faster on
1809 large queues. It won't affect queue runners, except when queue_run_in_order
1810 is set.
1811
181251. Installed eximstats 1.31 in the distribution.
1813
181452. Added support for SRV lookups to the dnslookup router.
1815
181653. If an ACL referred to $message_body or $message_body_end, the value was not
1817 reset for any messages that followed in the same SMTP session.
1818
181954. The store-handling optimization for building very long strings was not
1820 differentiating between the different store pools. I don't think this
1821 actually made any difference in practice, but I've tidied it.
1822
182355. While running the routers to verify a sender address, $sender_address
1824 was still set to the sender address. This is wrong, because when routing to
1825 send a bounce to the sender, it would be empty. Therefore, I have changed
1826 it so that, while verifying a sender address, $sender_address is set to <>.
1827 (There is no change to what happens when verifying a recipient address.)
1828
182956. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1830 target A or AAAA records (if not already returned) without resetting the
1831 qualify_single or search_parents options of the DNS resolver. These are
1832 inappropriate in this case because the targets of MX and SRV records must
1833 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1834 target that, when qualified, matched something in the local domain. These
1835 two options are now turned off when doing these lookups.
1836
183757. It seems that at least some releases of Reiserfs (which does not have the
1838 concept of a fixed number of inodes) returns zero and not -1 for the
1839 number of available inodes. This interacted badly with check_spool_inodes,
1840 which assumed that -1 was the "no such thing" setting. What I have done is
1841 to check that the total number of inodes is greater than zero before doing
1842 the test of how many are available.
1843
184458. When a "warn" ACL statement has a log_message modifier, the message is
1845 remembered, and not repeated. This is to avoid a lot of repetition when a
1846 message has many recipients that cause the same warning to be written.
1847 Howewer, Exim was preserving the list of already written lines for an
1848 entire SMTP session, which doesn't seem right. The memory is now reset if a
1849 new message is started.
1850
185159. The "rewrite" debugging flag was not showing the result of rewriting in the
1852 debugging output unless log_rewrite was also set.
1853
185460. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1855 of (int)(handle) when we know that handle contains (void *)(-1).
1856
185761. The Exim daemon panic-logs an error return when it closes the incoming
1858 connection. However "connection reset by peer" seems to be common, and
1859 isn't really an error worthy of noting specially, so that particular error
1860 is no long logged.
1861
186262. When Exim is trying to find all the local interfaces, it used to panic and
1863 die if the ioctl to get the interface flags failed. However, it seems that
1864 on at least one OS (Solaris 9) it is possible to have an interface that is
1865 included in the list of interfaces, but for which you get a failure error
1866 for this call. This happens when the interface is not "plumbed" into a
1867 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1868 failure of the "get flags" call assumes that the interface is down.
1869
187063. Added a ${eval10: operator, which assumes all numbers are decimal. This
1871 makes life easier for people who are doing arithmetic on fields extracted
1872 from dates, where you often get leading zeros that should not be
1873 interpreted as octal.
1874
187564. Added qualify_domain to the redirect router, to override the global
1876 setting.
1877
187865. If a pathologically long header line contained very many addresses (the
1879 report of this problem mentioned 10 000) and each of them was rewritten,
1880 Exim could use up a very large amount of memory. (It kept on making new
1881 copies of the header line as it rewrote, and never released the old ones.)
1882 At the expense of a bit more processing, the header rewriting function has
1883 been changed so that it no longer eats memory in this way.
1884
188566. The generation of the Received: header has been moved from the time that a
1886 message starts to be received, to the time that it finishes. The timestamp
1887 in the Received: header should now be very close to that of the <= log
1888 line. There are two side-effects of this change:
1889
1890 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1891 logged header lines no longer include the local Received: line, because
1892 it has not yet been created. The same applies to a copy of the message
1893 that is returned to a non-SMTP sender when a message is rejected.
1894
1895 (b) When a filter file is tested using -bf, no additional Received: header
1896 is added to the test message. After some thought, I decided that this
1897 is a bug fix.
1898
1899 This change does not affect the value of $received_for. It is still set
1900 after address rewriting, but before local_scan() is called.
1901
190267. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1903
190468. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1905 gave an unhelpful panic error message, and a defer error. I have managed to
1906 change this behaviour so that it now rejects any supplied certificate,
1907 which seems right, as the list of acceptable certificates is empty.
1908
190969. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1910 gave an unhelpful defer error. I have not managed to make this reject any
1911 supplied certificates, but the error message it gives is "no certificate
1912 supplied", which is not helpful.
1913
191470. exigrep's output now also includes lines that are not associated with any
1915 message, but which match the given pattern. Implemented by a patch from
1916 Martin Sluka, which also tidied up the Perl a bit.
1917
191871. Recipient callout verification, like sender verification, was using <> in
1919 the MAIL FROM command. This isn't really the right thing, since the actual
1920 sender may affect whether the remote host accepts the recipient or not. I
1921 have changed it to use the actual sender in the callout; this means that
1922 the cache record is now keyed on a recipient/sender pair, not just the
1923 recipient address. There doesn't seem to be a real danger of callout loops,
1924 since a callout by the remote host to check the sender would use <>.
1925 [SEE ABOVE: changed after hitting problems.]
1926
192772. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1928 temporary errors. However, in the case of such a code being given after
1929 the end of a data transmission (i.e. after ".") Exim was failing to write
1930 a retry record for the message. (Yes, there was some broken host that was
1931 actually sending 8xx at this point.)
1932
193373. An unknown lookup type in a host list could cause Exim to panic-die when
1934 the list was checked. (An example that provoked this was putting <; in the
1935 middle of a list instead of at the start.) If this happened during a DATA
1936 ACL check, a -D file could be left lying around. This kind of configuration
1937 error no longer causes Exim to die; instead it causes a defer errror. The
1938 incident is still logged to the main and panic logs.
1939
194074. Buglet left over from Exim 3 conversion. The message "too many messages
1941 in one connection" was written to the rejectlog but not the mainlog, except
1942 when address rewriting (yes!) was being logged.
1943
194475. Added write_rejectlog option.
1945
194676. When a system filter was run not as root (that is, when system_filter_user
1947 was set), the values of the $n variables were not being returned to the
1948 main process; thus, they were not subsequently available in the $sn
1949 variables.
1950
195177. Added +return_path_on_delivery log selector.
1952
195378. A connection timeout was being treated differently from recipients deferred
1954 when testing hosts_max_try with a message that was older than the host's
1955 retry timeout. (The host should not be counted, thus allowing all hosts to
1956 be tried at least once before bouncing.) This may have been the cause of an
1957 occasionally reported bug whereby a message would remain on the queue
1958 longer than the retry timeout, but would be bounced if a delivery was
1959 forced. I say "may" because I never totally pinned down the problem;
1960 setting up timeout/retry tests is difficult. See also the next item.
1961
196279. The ultimate address timeout was not being applied to errors that involved
1963 a combination of host plus message (for example, a timeout on a MAIL
1964 command). When an address resolved to a number of possible hosts, and they
1965 were not all tried for each delivery (e.g. because of hosts_max_try), a
1966 message could remain on the queue longer than the retry timeout.
1967
196880. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1969 Haardt.
1970
197181. Fixed an obscure SMTP outgoing bug which required at least the following
1972 conditions: (a) there was another message waiting for the same server;
1973 (b) the server returned 5xx to all RCPT commands in the first message so
1974 that the message was not completed; (c) the server dropped the connection
1975 or gave a negative response to the RSET that Exim sends to abort the
1976 transaction. The observed case was a dropped connection after DATA that had
1977 been sent in pipelining mode. That is, the server had advertised PIPELINING
1978 but was not implementing it correctly. The effect of the bug was incorrect
1979 behaviour, such as trying another host, and this could lead to a crash.
1980
1981
1982Exim version 4.30
1983-----------------
1984
1985 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1986 and daemon.c were passed as pointers to ints; they should have been
1987 pointers to socklen_t variables (which are typically unsigned ints).
1988
1989 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1990 fixed.
1991
1992 3. Fixed a really odd bug that affected only the testing scheme; patching a
1993 certain fixed string in the binary changed the value of another string that
1994 happened to be identical to the end of the original first string.
1995
1996 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1997 name", it returns that address as the IP address. On some operating
1998 systems (e.g. Solaris), it also passes back the IP address string as the
1999 "host name". However, on others (e.g. Linux), it passes back an empty
2000 string. Exim wasn't checking for this, and was changing the host name to an
2001 empty string, assuming it had been canonicized.
2002
2003 5. Although rare, it is permitted to have more than one PTR record for a given
2004 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2005 all the names associated with an address, because they do in Solaris.
2006 However, it seems that they do not in Linux for data that comes from the
2007 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2008 I found this out when I moved to a new Linux workstation and tried to run
2009 the Exim test suite.
2010
2011 To get round this problem I have changed the code so that it now does its
2012 own call to the DNS to look up PTR records when searching for a host name.
2013 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2014 addresses that are only in /etc/hosts are still found.
2015
2016 This behaviour is, however, controlled by an option called host_lookup_
2017 order, which defaults to "bydns:byaddr". If people want to use the other
2018 order, or indeed, just use one or the other means of lookup, they can
2019 specify it in this variable.
2020
2021 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2022 some operating systems, this comes back from gethostbyaddr() as an empty
2023 string, and this is what Exim used to test for. However, it seems that in
2024 other systems, "." is yielded. Exim now tests for this case too.
2025
2026 7. The values of check_spool_space and check_log_space are now held internally
2027 as a number of kilobytes instead of an absolute number of bytes. If a
2028 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2029 kilobyte. This means that much larger values can be stored.
2030
2031 8. Exim monitor: an attempt to get the action menu when not actually pointing
2032 at a message produces an empty menu entitled "No message selected". This
2033 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2034 no entries in it ("Shell widget menu has zero width and/or height"). So I
2035 have added a single, blank menu entry in this case.
2036
2037 9. Added ${quote_local_part.
2038
203910. MIME decoding is now applied to the contents of Subject: header lines when
2040 they are logged.
2041
204211. Now that a reference to $sender_host_address automatically causes a reverse
2043 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2044 host lookup before query-style lookups in lists that might use this
2045 variable. This has therefore been abolished, and the "net-" prefix is no
2046 longer necessary for query-style lookups.
2047
204812. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2049 have been a typo for LFLAGS, so it has been changed.
2050
205113. The install script calls Exim with "-C /dev/null" in order to find the
2052 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2053 to be output. Howeve, since Exim outputs its version number before the
2054 error, it didn't break the script. It just looked ugly. I fixed this by
2055 always allowing "-C /dev/null" if the caller is root.
2056
205714. Ignore overlarge ACL variable number when reading spool file - insurance
2058 against a later release with more variables having written the file.
2059
206015. The standard form for an IPv6 address literal was being rejected by EHLO.
2061 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2062 this, as well as the form without the "IPv6" on the front.
2063
206416. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2065 OS/Makefile-Darwin file.
2066
206717. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2068 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2069
207018. After change 4.21/52, "%ld" was used to format the contents of the $inode
2071 variable. However, some OS use ints for inodes. I've added cast to long int
2072 to get rid of the compiler warning.
2073
207419. I had forgotten to lock out "/../" in configuration file names when
2075 ALT_CONFIG_PREFIX was set.
2076
207720. Routers used for verification do not need to specify transports. However,
2078 if such a router generated a host list, and callout was configured, Exim
2079 crashed, because it could not find a port number from the (non-existent)
2080 transport. It now assumes port 25 in this circumstance.
2081
208221. Added the -t option to exigrep.
2083
208422. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2085 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2086 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2087 them is set, LOOKUP_LSEARCH is forced.
2088
208923. "exim -bV" now outputs a list of lookups that are included in the binary.
2090
209124. Added sender and host information to the "rejected by local_scan()" log
2092 line; previously there was no indication of these.
2093
209425. Added .include_if_exists.
2095
209626. Change 3.952/11 added an explicit directory sync on top of a file sync for
2097 Linux. It turns out that not all file systems support this. Apparently some
2098 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2099 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2100 not supported on the file descriptor, is now ignored when Exim is trying to
2101 sync a directory. This applies only to Linux.
2102
210327. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2104
210528. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2106 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2107 this.
2108
210929. Added support for maildirsize files from supplied patch (modified a bit).
2110
211130. The use of :fail: followed by an empty string could lead Exim to respond to
2112 sender verification failures with (e.g.):
2113
2114 550 Verification failed for <xxx>
2115 550 Sender verify failed
2116
2117 where the first response line was missing the '-' that indicates it is not
2118 the final line of the response.
2119
212031. The loop for finding the name of the user that called Exim had a hardwired
2121 limit of 10; it now uses the value of finduser_retries, which is used for
2122 all other user lookups.
2123
212432. Added $received_count variable, available in data and not_smtp ACLs, and at
2125 delivery time.
2126
212733. Exim was neglecting to zero errno before one call of strtol() when
2128 expanding a string and expecting an integer value. On some systems this
2129 resulted in spurious "integer overflow" errors. Also, it was casting the
2130 result into an int without checking.
2131
213234. Testing for a connection timeout using "timeout_connect" in the retry rules
2133 did not work. The code looks as if it has *never* worked, though it appears
2134 to have been documented since at least releast 1.62. I have made it work.
2135
213635. The "timeout_DNS" error in retry rules, also documented since at least
2137 1.62, also never worked. As it isn't clear exactly what this means, and
2138 clearly it isn't a major issue, I have abolished the feature by treating it
2139 as "timeout", and writing a warning to the main and panic logs.
2140
214136. The display of retry rules for -brt wasn't always showing the error code
2142 correctly.
2143
214437. Added new error conditions to retry rules: timeout_A, timeout_MX,
2145 timeout_connect_A, timeout_connect_MX.
2146
214738. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2148 to the empty sender.
2149
215039. The daemon was not analysing the content of -oX till after it had closed
2151 stderr and disconnected from the controlling terminal. This meant that any
2152 syntax errors were only noted on the panic log, and the return code from
2153 the command was 0. By re-arranging the code a little, I've made the
2154 decoding happen first, so such errors now appear on stderr, and the return
2155 code is 1. However, the actual setting up of the sockets still happens in
2156 the disconnected process, so errors there are still only recorded on the
2157 panic log.
2158
215940. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2160 connections (as happens on some IP stacks) was logged at start up time as
2161 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2162 it from "IPv6 and IPv4", which means that two separate sockets are being
2163 used.
2164
216541. The debug output for gethostbyname2() or getipnodebyname() failures now
2166 says whether AF_INET or AF_INET6 was passed as an argument.
2167
216842. Exiwhat output was messed up when time zones were included in log
2169 timestamps.
2170
217143. Exiwhat now gives more information about the daemon's listening ports,
2172 and whether -tls-on-connect was used.
2173
217444. The "port" option of the smtp transport is now expanded.
2175
217645. A "message" modifier in a "warn" statement in a non-message ACL was being
2177 silently ignored. Now an error message is written to the main and panic
2178 logs.
2179
218046. There's a new ACL modifier called "logwrite" which writes to a log file
2181 as soon as it is encountered.
2182
218347. Added $local_user_uid and $local_user_gid at routing time.
2184
218548. Exim crashed when trying to verify a sender address that was being
2186 rewritten to "<>".
2187
218849. Exim was recognizing only a space character after ".include". It now also
2189 recognizes a tab character.
2190
219150. Fixed several bugs in the Perl script that creates the exim.8 man page by
2192 extracting the relevant information from the specification. The man page no
2193 longer contains scrambled data for the -d option, and I've added a section
2194 at the front about calling Exim under different names.
2195
219651. Added "extra_headers" argument to the "mail" command in filter files.
2197
219852. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2199 crash.
2200
220153. Installed eximstats 1.29.
2202
220354. Added transport_filter_timeout as a generic transport option.
2204
220555. Exim no longer adds an empty Bcc: header to messages that have no To: or
2206 Cc: header lines. This was required by RFC 822, but it not required by RFC
2207 2822.
2208
220956. Exim used to add From:, Date:, and Message-Id: header lines to any
2210 incoming messages that did not have them. Now it does so only if the
2211 message originates locally, that is, if there is no associated remote host
2212 address. When Resent- header lines are present, this applies to the Resent-
2213 lines rather than the non-Resent- lines.
2214
221557. Drop incoming SMTP connection after too many syntax or protocol errors. The
2216 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2217
221858. Messages for configuration errors now include the name of the main
2219 configuration file - useful now that there may be more than one file in a
2220 list (.included file names were always shown).
2221
222259. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2223 for those rare installations that do not start the daemon as root or run it
2224 setuid root. I've cut out the call to initgroups() if the daemon is not
2225 root at that time.
2226
222760. The Exim user and group can now be bound into the binary as text strings
2228 that are looked up at the start of Exim's processing.
2229
223061. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2231
223262. Added $mailstore_basename variable.
2233
223463. Installed patch to sieve.c from Michael Haardt.
2235
223664. When Exim failed to open the panic log after failing to open the main log,
2237 the original message it was trying to log was written to stderr and debug
2238 output, but if they were not available (the usual case in production), it
2239 was lost. Now it is written to syslog before the two lines that record the
2240 failures to open the logs.
2241
224265. Users' Exim filters run in subprocesses under the user's uid. It is
2243 possible for a "deliver" command or an alias in a "personal" command to
2244 provoke an address rewrite. If logging of address rewriting is configured,
2245 this fails because the process is not running as root or exim. There may be
2246 a better way of dealing with this, but for the moment (because 4.30 needs
2247 to be released), I have disabled address rewrite logging when running a
2248 filter in a non-root, non-exim process.
2249
2250
2251Exim version 4.24
2252-----------------
2253
2254 1. The buildconfig auxiliary program wasn't quoting the value set for
2255 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2256 not defined. This bug was masked in 4.22 by the effect that was fixed in
2257 change 4.23/1.
2258
2259 2. Some messages that were rejected after a message id was allocated were
2260 shown as "incomplete" by exigrep. It no longer does this for messages that
2261 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2262
2263 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2264 have allow_domain_literals set, the ID did not get logged in the <= line.
2265 Domain literals are now always recognized in Message-ID: header lines.
2266
2267 4. The first argument for a ${extract expansion item is the key name or field
2268 number. Leading and trailing spaces in this item were not being ignored,
2269 causing some misleading effects.
2270
2271 5. When deliver_drop_privilege was set, single queue runner processes started
2272 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2273 same command in the process it spins off) were not dropping privilege.
2274
2275 6. When the daemon running as "exim" started a queue runner, it always
2276 re-executed Exim in the spun-off process. This is a waste of effort when
2277 deliver_drop_privilege is set. The new process now just calls the
2278 queue-runner function directly.
2279
2280
2281Exim version 4.23
2282-----------------
2283
2284 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2285 HEADERS_CHARSET.
2286
2287 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2288 ignored. Though the use of -oP was forcing the writing of a pid file, it
2289 was always written to the default place.
2290
2291 3. If the message "no IP address found for host xxxx" is generated during
2292 incoming verification, it is now followed by identification of the incoming
2293 connection (so you can more easily find what provoked it).
2294
2295 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2296
2297 5. Added some features to "harden" Exim a bit more against certain attacks:
2298
2299 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2300 be put in Local/Makefile. This is like the never_users runtime option,
2301 but it cannot be overridden. The default setting is "root".
2302
2303 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2304 prefix string with which any file named in a -C command line option
2305 must start.
2306
2307 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2308 is retained for -C and -D only if the caller of Exim is root. Without
2309 it, the exim user may also use -C and -D and retain privilege.
2310
2311 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2312 command line option is disabled.
2313
2314 6. Macro names set by the -D option must start with an upper case letter, just
2315 like macro names defined in the configuration file.
2316
2317 7. Added "dereference=" facility to LDAP.
2318
2319 8. Two instances of the typo "uknown" in the source files are fixed.
2320
2321 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2322 the Configure-Makefile script screwed up while processing it.
2323
232410. Incorporated PCRE 4.4.
2325
232611. The SMTP synchronization check was not operating right at the start of an
2327 SMTP session. For example, it could not catch a HELO sent before the client
2328 waited for the greeting. There is now a check for outstanding input at the
2329 point when the greeting is written. Because of the duplex, asynchronous
2330 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2331 way, but not yet received, when the check is performed.
2332
233312. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2334 on TCP/IP sockets, because this apparently causes some broken clients to
2335 timeout.
2336
233713. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2338 unchanged) from the Cygwin maintainer.
2339
234014. The code for -bV that shows what is in the binary showed "mbx" when maildir
2341 was supported instead of testing for mbx. Effectively a typo.
2342
234315. The spa authenticator server code was not checking that the input it
2344 received was valid base64.
2345
234616. The debug output line for the "set" modifier in ACLs was not showing the
2347 name of the variable that was being set.
2348
234917. Code tidy: the variable type "vtype_string" was never used. Removed it.
2350
235118. Previously, a reference to $sender_host_name did not cause a DNS reverse
2352 lookup on its own. Something else was needed to trigger the lookup. For
2353 example, a match in host_lookup or the need for a host name in a host list.
2354 Now, if $sender_host_name is referenced and the host name has not yet been
2355 looked up, a lookup is performed. If the lookup fails, the variable remains
2356 empty, and $host_lookup_failed is set to "1".
2357
235819. Added "eqi" as a case-independent comparison operator.
2359
236020. The saslauthd authentication condition could segfault if neither service
2361 nor realm was specified.
2362
236321. If an overflowing value such as "2048M" was set for message_size_limit, the
2364 error message that was logged was misleading, and incoming SMTP
2365 connections were dropped. The message is now more accurate, and temporary
2366 errors are given to SMTP connections.
2367
236822. In some error situations (such as 21 above) Exim rejects all SMTP commands
2369 (except RSET) with a 421 error, until QUIT is received. However, it was
2370 failing to send a response to QUIT.
2371
237223. The HELO ACL was being run before the code for helo_try_verify_hosts,
2373 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2374 ACL is now run after the helo_try_verify_hosts code.
2375
237624. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2377 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2378 case-independent, so other case variants are also recognized). Apparently
2379 some systems use these upper case variants.
2380
238125. If more than two messages were waiting for the same host, and a transport
2382 filter was specified for the transport, Exim sent two messages over the
2383 same TCP/IP connection, and then failed with "socket operation on non-
2384 socket" when it tried to send the third.
2385
238626. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2387
238827. The extern definition of crypt16() in expand.c was not being excluded when
2389 the OS had its own crypt16() function.
2390
239128. Added bounce_return_body as a new option, and bounce_return_size_limit
2392 as a preferred synonym for return_size_limit, both as an option and as an
2393 expansion variable.
2394
239529. Added LIBS=-liconv to OS/Makefile-OSF1.
2396
239730. Changed the default configuration ACL to relax the local part checking rule
2398 for addresses that are not in any local domains. For these addresses,
2399 slashes and pipe symbols are allowed within local parts, but the sequence
2400 /../ is explicitly forbidden.
2401
240231. SPA server authentication was not clearing the challenge buffer before
2403 using it.
2404
240532. log_message in a "warn" ACL statement was writing to the reject log as
2406 well as to the main log, which contradicts the documentation and doesn't
2407 seem right (because no rejection is happening). So I have stopped it.
2408
240933. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2410 However, I am unable to do any testing of this.
2411
241234. Fixed an infelicity in the appendfile transport. When checking directories
2413 for a mailbox, to see if any needed to be created, it was accidentally
2414 using path names with one or more superfluous leading slashes; tracing
2415 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2416
241735. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2418 contents are added to the log line that is written for every discarded
2419 recipient. (Previously a log_message setting was ignored.)
2420
242136. The ${quote: operator now quotes the string if it is empty.
2422
242337. The install script runs exim in order to find its version number. If for
2424 some reason other than non-existence or emptiness, which it checks, it
2425 could not run './exim', it was installing it with an empty version number,
2426 i.e. as "exim-". This error state is now caught, and the installation is
2427 aborted.
2428
242938. An argument was missing from the function that creates an error message
2430 when Exim fails to connect to the socket for saslauthd authentication.
2431 This could cause Exim to crash, or give a corrupted message.
2432
243339. Added isip, isip4, and isip6 to ${if conditions.
2434
243540. The ACL variables $acl_xx are now saved with the message, and can be
2436 accessed later in routers, transports, and filters.
2437
243841. The new lookup type nwildlsearch is like wildlsearch, except that the key
2439 strings in the file are not string-expanded.
2440
244142. If a MAIL command specified a SIZE value that was too large to fit into an
2442 int variable, the check against message_size_limit failed. Such values are
2443 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2444 day this will have to be increased, but I don't think I want to be around
2445 when emails are that large.
2446
2447
2448
2449Exim version 4.22
2450-----------------
2451
2452 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2453 iconv() is not standard in FreeBSD.
2454
2455 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2456 IPv6 enabled. The observed symptom was a segmentation fault on return from
2457 the function os_common_find_running_interfaces() in src/os.c.
2458
2459 3. In the check_special_case() function in daemon.c I had used "errno" as an
2460 argument name, which causes warnings on some systems. This was basically a
2461 typo, since it was named "eno" in the comments!
2462
2463 4. The code that waits for the clock to tick (at a resolution of some fraction
2464 of a second) so as to ensure message-id uniqueness was always waiting for
2465 at least one whole tick, when it could have waited for less. [This is
2466 almost certainly not relevant at current processor speeds, where it is
2467 unlikely to ever wait at all. But we try to future-proof.]
2468
2469 5. The function that sleeps for a time interval that includes fractions of a
2470 second contained a race. It did not block SIGALRM between setting the
2471 timer, and suspending (a couple of lines later). If the interval was short
2472 and the sigsuspend() was delayed until after it had expired, the suspension
2473 never ended. On busy systems this could lead to processes getting stuck for
2474 ever.
2475
2476 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2477 process, before it forks any delivery processes. The open lookup caching
2478 mechanism meant that the open file or database connection was passed into
2479 the delivery process. The problem was that delivery processes always tidy
2480 up cached lookup data. This could cause a problem for the next delivery
2481 process started by the queue runner, because the external queue runner
2482 process does not know about the closure. So the next delivery process
2483 still has data in the lookup cache. In the case of a file lookup, there was
2484 no problem because closing a file descriptor in a subprocess doesn't affect
2485 the parent. However, if the lookup was caching a connection to a database,
2486 the connection was closed, and the second delivery process was likely to
2487 see errors such as "PGSQL: query failed: server closed the connection
2488 unexpectedly". The problem has been fixed by closing all cached lookups
2489 in a queue runner before running a delivery process.
2490
2491 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2492 seem to have the "const" qualifier which it has on other OS. I've
2493 parameterised it.
2494
2495 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2496 *of the same type* (client or server) with the same public name that an
2497 error should be diagnosed.
2498
2499 9. When Exim looked up a host name for an IP address, but failed to find the
2500 original IP address when looking up the host name (a safety check), it
2501 output the message "<ip address> does not match any IP for NULL", which was
2502 confusing, to say the least. The bug was that the host name should have
2503 appeared instead of "NULL".
2504
250510. Since release 3.03, if Exim is called by a uid other than root or the Exim
2506 user that is built into the binary, and the -C or -D options is used, root
2507 privilege is dropped before the configuration file is read. In addition,
2508 logging is switched to stderr instead of the normal log files. If the
2509 configuration then re-defines the Exim user, the unprivileged environment
2510 is probably not what is expected, so Exim logs a panic warning message (but
2511 proceeds).
2512
2513 However, if deliver_drop_privilege is set, the unprivileged state may well
2514 be exactly what is intended, so the warning has been cut out in that case,
2515 and Exim is allowed to try to write to its normal log files.
2516
2517
2518Exim version 4.21
2519-----------------
2520
2521 1. smtp_return_error_details was not giving details for temporary sender
2522 or receiver verification errors.
2523
2524 2. Diagnose a configuration error if two authenticators have the same public
2525 name.
2526
2527 3. Exim used not to create the message log file for a message until the first
2528 delivery attempt. This could be confusing when incoming messages were held
2529 for policy or load reasons. The message log file is now created at the time
2530 the message is received, and an initial "Received" line is written to it.
2531
2532 4. The automatically generated man page for command line options had a minor
2533 bug that caused no ill effects; however, a more serious problem was that
2534 the procedure for building the man page automatically didn't always
2535 operate. Consequently, release 4.20 contains an out-of-date version. This
2536 shouldn't happen again.
2537
2538 5. When building Exim with embedded Perl support, the script that builds the
2539 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2540 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2541
2542 6. The freeze_tell option was not being used for messages that were frozen on
2543 arrival, either by an ACL or by local_scan().
2544
2545 7. Added the smtp_incomplete_transaction log selector.
2546
2547 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2548 was accepting AUTH without a new EHLO.
2549
2550 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2551 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2552 RFC.
2553
255410. Logging of TCP/IP connections (when configured) now happens in the main
2555 daemon process instead of the child process, so that the TCP/IP connection
2556 count is more accurate (but it can never be perfect).
2557
255811. The use of "drop" in a nested ACL was not being handled correctly in the
2559 outer ACL. Now, if condition failure induced by the nested "drop" causes
2560 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2561 or "require"), the connection is dropped.
2562
256312. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2564 that yield "discard" can now be used with an "accept" or a "discard" verb,
2565 but an error is generated for any others (because I can't see a useful way
2566 to define what should happen).
2567
256813. When an ACL is read dynamically from a file (or anywhere else), the lines
2569 are now processed in the same way as lines in the Exim configuration file.
2570 In particular, continuation lines are supported.
2571
257214. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2573
257415. Added -ti meaning -t -i.
2575
257616. Check for letters, digits, hyphens, and dots in the names of dnslist
2577 domains, and warn by logging if others are found.
2578
257917. At least on BSD, alignment is not guarenteed for the array of ifreq's
2580 returned from GIFCONF when Exim is trying to find the list of interfaces on
2581 a host. The code in os.c has been modified to copy each ifreq to an aligned
2582 structure in all cases.
2583
2584 Also, in some cases, the returned ifreq's were being copied to a 'struct
2585 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2586 means the last couple of bytes of an IPv6 address could be chopped if the
2587 ifreq contained only a normal sockaddr (14 bytes storage).
2588
258918. Named domain lists were not supported in the hosts_treat_as_local option.
2590 An entry such as +xxxx was not recognized, and was treated as a literal
2591 domain name.
2592
259319. Ensure that header lines added by a DATA ACL are included in the reject log
2594 if the ACL subsequently rejects the message.
2595
259620. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2597 MD5 (which is deprecated).
2598
259921. When testing a filter file using -bf, Exim was writing a message when it
2600 took the sender from a "From " line in the message, but it was not doing so
2601 when it took $return_path from a Return-Path: header line. It now does.
2602
260322. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2604 with a valid header line field name (a series of printing characters
2605 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2606
260723. Changed "disc" in the source to "disk" to conform to the documentation and
2608 the book and for uniformity.
2609
261024. Ignore Sendmail's -Ooption=value command line item.
2611
261225. When execve() failed while trying to run a command in a pipe transport,
2613 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2614 could be confused with a return value of 69 from the command itself. This
2615 has been changed to 127, the value the shell returns if it is asked to run
2616 a non-existent command. The wording for the related log line suggests a
2617 non-existent command as the problem.
2618
261926. If received_header_text expands to an empty string, do not add a Received:
2620 header line to the message. (Well, it adds a token one on the spool, but
2621 marks it "old" so that it doesn't get used or transmitted.)
2622
262327. Installed eximstats 1.28 (addition of -nt option).
2624
262528. There was no check for failure on the call to getsockname() in the daemon
2626 code. This can fail if there is a shortage of resources on the system, with
2627 ENOMEM, for example. A temporary error is now given on failure.
2628
262929. Contrary to the C standard, it seems that in some environments, the
2630 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2631 program. Exim now does this explicitly; it affects the formatting of
2632 timestamps using strftime().
2633
263430. If exiqsumm was given junk data, it threw up some uninitialized variable
2635 complaints. I've now initialized all the variables, to avoid this.
2636
263732. Header lines added by a system filter were not being "seen" during
2638 transport-time rewrites.
2639
264033. The info_callback() function passed to OpenSSL is set up with type void
2641 (*)(SSL *, int, int), as described somewhere. However, when calling the
2642 function (actually a macro) that sets it up, the type void(*)() is
2643 expected. I've put in a cast to prevent warnings from picky compilers.
2644
264534. If a DNS black list lookup found a CNAME record, but there were no A
2646 records associated with the domain it pointed at, Exim crashed.
2647
264835. If a DNS black list lookup returned more than one A record, Exim ignored
2649 all but the first. It now scans all returned addresses if a particular IP
2650 value is being sought. In this situation, the contents of the
2651 $dnslist_value variable are a list of all the addresses, separated by a
2652 comma and a space.
2653
265436. Tightened up the rules for host name lookups using reverse DNS. Exim used
2655 to accept a host name and all its aliases if the forward lookup for any of
2656 them yielded the IP address of the incoming connection. Now it accepts only
2657 those names whose forward lookup yields the correct IP address. Any other
2658 names are discarded. This closes a loophole whereby a rogue DNS
2659 administrator could create reverse DNS records to break through a
2660 wildcarded host restriction in an ACL.
2661
266237. If a user filter or a system filter that ran in a subprocess used any of
2663 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2664 the wrong values were passed to the pipe command ($thisaddress had the
2665 value of $0, $0 had the value of $1, etc). This bug was introduced by
2666 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2667
266838. Improved the line breaking for long SMTP error messages from ACLs.
2669 Previously, if there was no break point between 40 and 75 characters, Exim
2670 left the rest of the message alone. Two changes have been made: (a) I've
2671 reduced the minimum length to 35 characters; (b) if it can't find a break
2672 point between 35 and 75 characters, it looks ahead and uses the first one
2673 that it finds. This may give the occasional overlong line, but at least the
2674 remaining text gets split now.
2675
267639. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2677 file descriptors might be low, and that setting 1000 would always raise it.
2678 It turns out that in some environments, the limit is already over 1000 and
2679 that lowering it causes trouble. So now Exim takes care not to decrease it.
2680
268140. When delivering a message, the value of $return_path is set to $sender_
2682 address at the start of routing (routers may change the value). By an
2683 oversight, this default was not being set up when an address was tested by
2684 -bt or -bv, which affected the outcome if any router or filter referred to
2685 $return_path.
2686
268741. The idea of the "warn" ACL verb is that it adds a header or writes to the
2688 log only when "message" or "log_message" are set. However, if one of the
2689 conditions was an address verification, or a call to a nested ACL, the
2690 messages generated by the underlying test were being passed through. This
2691 no longer happens. The underlying message is available in $acl_verify_
2692 message for both "message" and "log_message" expansions, so it can be
2693 passed through if needed.
2694
269542. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2696 new expansion $bh_ to give the encoded byte string without charset
2697 translation. Translation happens only if iconv() is available; HAVE_ICONV
2698 indicates this at build time. HEADERS_CHARSET gives the charset to
2699 translate to; headers_charset can change it in the configuration, and
2700 "headers charset" can change it in an individual filter file.
2701
270243. Now that we have a default RFC 2047 charset (see above), the code in Exim
2703 that creates RFC 2047 encoded "words" labels them as that charset instead
2704 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2705 expansion operator; (ii) when Exim creates a From: line for a local
2706 message; (iii) when a header line is rewritten to include a "phrase" part.
2707
270844. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2709 buggy, causing it to skip the first lines of messages whose message ID
2710 ended in 'D'. This would not have bitten before Exim release 4.14, because
2711 message IDs were unlikely to end in 'D' before then. The effect was to have
2712 incorrect size information for certain domains.
2713
271445. #include "config.h" was missing at the start of the crypt16.c module. This
2715 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2716 noticed.
2717
271846. If there was a timeout during a "random" callout check, Exim treated it as
2719 a failure of the random address, and carried on sending RSET and the real
2720 address. If the delay was just some slowness somewhere, the response to the
2721 original RCPT would be taken as a response to RSET and so on, causing
2722 mayhem of various kinds.
2723
272447. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2725 when I implemented it. It didn't allow for the fact that some option values
2726 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2727 the right test for positive values.
2728
272948. Domain names in DNS records are case-independent. Exim always looks them up
2730 in lower case. Some resolvers return domain names in exactly the case they
2731 appear in the zone file, that is, they may contain uppercase letters. Not
2732 all resolvers do this - some return always lower case. Exim was treating a
2733 change of case by a resolver as a change of domain, similar to a widening
2734 of a domain abbreviation. This triggered its re-routing code and so it was
2735 trying to route what was effectively the same domain again. This normally
2736 caused routing to fail (because the router wouldn't handle the domain
2737 twice). Now Exim checks for this case specially, and just changes the
2738 casing of the domain that it ultimately uses when it transmits the message
2739 envelope.
2740
274149. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2742 module.
2743
274450. If a filter generated a file delivery with a non-absolute name (possible if
2745 no home directory exists for the router), the forbid_file option was not
2746 forbidding it.
2747
274851. Added '&' feature to dnslists, to provide bit mask matching in addition to
2749 the existing equality matching.
2750
275152. Exim was using ints instead of ino_t variables in some places where it was
2752 dealing with inode numbers.
2753
275453. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2755 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2756 called TMPDIR, and if it finds it is different, it changes its value.
2757
275854. The smtp_printf() function is now made available to local_scan() so
2759 additional output lines can be written before returning. There is also an
2760 smtp_fflush() function to enable the detection of a dropped connection.
2761 The variables smtp_input and smtp_batched_input are exported to
2762 local_scan().
2763
276455. Changed the default runtime configuration: the message "Unknown user"
2765 has been removed from the ACL, and instead placed on the localuser router,
2766 using the cannot_route_message feature. This means that any verification
2767 failures that generate their own messages won't get overridden. Similarly,
2768 the "Unrouteable address" message that was in the ACL for unverifiable
2769 relay addresses has also been removed.
2770
277156. Added hosts_avoid_esmtp to the smtp transport.
2772
277357. The exicyclog script was not checking for the esoteric option
2774 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2775 will work only if exicyclog is run under the appropriate euid.
2776
277758. Following a discussion on the list, the rules by which Exim recognises line
2778 endings on incoming messages have been changed. The -dropcr and drop_cr
2779 options are now no-ops, retained only for backwards compatibility. The
2780 following line terminators are recognized: LF CRLF CR. However, special
2781 processing applies to CR:
2782
2783 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2784 nor a local message in the state where . is a terminator.
2785
2786 (ii) If a bare CR is encountered in a header line, an extra space is added
2787 after the line terminator so as not to end the header. The reasoning
2788 behind this is that bare CRs in header lines are most likely either
2789 to be mistakes, or people trying to play silly games.
2790
279159. The size of a message, as listed by "-bp" or in the Exim monitor window,
2792 was being incorrectly given as 18 bytes larger than it should have been.
2793 This is a VOB (very old bug).
2794
279560. This may never have affected anything current, but just in case it has:
2796 When the local host is found other than at the start of a list of hosts,
2797 the local host, those with the same MX, and any that follow, are discarded.
2798 When the list in question was part of a longer list of hosts, the following
2799 hosts (not currently being processed) were also being discarded. This no
2800 longer happens. I'm not sure if this situation could ever has previously
2801 arisen.
2802
280361. Added the "/MX" feature to lists of hosts in the manualroute and query
2804 program routers.
2805
280662. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2807 header. This is something that is recommended in a new Internet Draft, and
2808 is something that is documented as being done by Sendmail. There are two
2809 possible values. For messages generated by the autoreply transport, Exim
2810 adds:
2811
2812 Auto-Submitted: auto-replied
2813
2814 whereas for all other generated messages (e.g. bounces) it adds
2815
2816 Auto-Submitted: auto-generated
2817
281863. The "personal" condition in filters now includes a test for the
2819 Auto-Submitted: header. If it contains the string "auto-" the message it
2820 not considered personal.
2821
282264. Added rcpt_include_affixes as a generic transport option.
2823
282465. Added queue_only_override (default true).
2825
282666. Added the syslog_duplication option.
2827
282867. If what should have been the first header line of a message consisted of
2829 a space followed by a colon, Exim was mis-interpreting it as a header line.
2830 It isn't of course - it is syntactically invalid and should therefore be
2831 treated as the start of the message body. The misbehaviour could have
2832 caused a number of strange effects, including loss of data in subsequent
2833 header lines, and spool format errors.
2834
283568. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2836 client host had authenticated. This control can now be exercised by an ACL
2837 for more flexibility.
2838
283969. By default, callouts do not happen when testing with -bh. There is now a
2840 variant, -bhc, which does actually run the callout code, including
2841 consulting and updating the callout cache.
2842
284370. Added support for saslauthd authentication, courtesy of Alexander
2844 Sabourenkov.
2845
284671. If statvfs() failed on the spool or log directories while checking their
2847 size for availability, Exim confusingly gave the error "space shortage".
2848 Furthermore, in debugging mode it crashed with a floating point exception.
2849 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2850 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2851 serious problem, Exim now writes to the main and panic logs when this
2852 happens, with details of the failure. It then refuses to accept the
2853 incoming message, giving the message "spool directory problem" or "log
2854 directory problem" with a 421 code for SMTP messages.
2855
285672. When Exim is about to re-exec itself, it ensures that the file descriptors
2857 0, 1, and 2 exist, because some OS complain for execs without them (see
2858 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2859 descriptors. However, the code omitted to check that the open succeeded,
2860 causing mysterious errors if for some reason the permissions on /dev/null
2861 got screwed. Now Exim writes a message to the main and panic logs, and
2862 bombs out if it can't open /dev/null.
2863
286473. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2865 interact so that it is all more flexible. It is supposed to remain
2866 backwards compatible. Also added extra_local_interfaces.
2867
286874. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2869 to bomb out with an assertion failure - to the client this appears as a
2870 connection drop. This problem occurs in the part of the code that was taken
2871 from the Samba project. Fortunately, the assertion is in a very simple
2872 function, so I have fixed this by reproducing the function inline in the
2873 one place where it is called, and arranging for authentication to fail
2874 instead of killing the process with assert().
2875
287675. The SPA client code was not working when the server requested OEM rather
2877 than Unicode encoding.
2878
287976. Added code to make require_files with a specific uid setting more usable in
2880 the case where statting the file as root fails - usually a non-root-mounted
2881 NFS file system. When this happens and the failure is EACCES, Exim now
2882 forks a subprocess and does the per-uid checking as the relevant uid.
2883
288477. Added process_log_path.
2885
288678. If log_file_path was not explicitly set, a setting of check_log_space or
2887 check_log_inodes was ignored.
2888
288979. If a space check for the spool or log partitions fails, the incident is now
2890 logged. Of course, in the latter case the data may get lost...
2891
289280. Added the %p formatting code to string_format() so that it can be used to
2893 print addresses in debug_print(). Adjusted all the address printing in the
2894 debugging in store.c to use %p rather than %d.
2895
289681. There was a concern that a line of code in smtp_in.c could overflow a
2897 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2898 initially expressed, the concern was not well-founded, because trailing
2899 spaces are removed early. However, if the trailing spaces were followed by
2900 a NULL, they did not get removed, so the overflow was possible. Two fixes
2901 were applied:
2902
2903 (a) I re-wrote the offending code in a cleaner fashion.
2904 (b) If an incoming SMTP command contains a NULL character, it is rejected
2905 as invalid.
2906
290782. When Exim changes uid/gid to the Exim user at daemon start time, it now
2908 runs initgroups(), so that if the Exim user is in any additional groups,
2909 they will be used during message reception.
2910
2911
2912Exim version 4.20
2913-----------------
2914
2915The change log for 4.20 and earlier releases has been archived.
2916
2917****