Remove RSA_EXPORT stuff from the test client.c program in the same way
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
411ef850 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.408 2006/10/16 10:58:39 ph10 Exp $
495ae4b0
PH
2
3Change log file for Exim from version 4.21
4-------------------------------------------
5
944e9e9c
TF
6Exim version 4.64
7-----------------
8
9TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
12 in the field name.
13
2b1c6e3a
PH
14PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
25
1f872c80
PH
26PH/02 Tidied a number of unused variable and signed/unsigned warnings that
27 gcc 4.1.1 threw up.
28
29PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
33
d203e649
PH
34PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
36
cc2ed8f7 37PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
21a04aa3 38 later the small subsequent patch to fix an introduced bug.
f951fd57 39
ddfcd446
PH
40PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
41
d45b1de8
PH
42PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
43
44PH/08 An error is now given if message_size_limit is specified negative.
45
38a0a95f 46PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
641cb756 47 to be given (somewhat) arbitrary names.
38a0a95f 48
a2405d83
JJ
49JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
50 in 4.64-PH/09.
51
52JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
53 miscellaneous code fixes
54
6ea85e9a
PH
55PH/10 Added the log_reject_target ACL modifier to specify where to log
56 rejections.
57
26da7e20
PH
58PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
59 hostname. This is wrong, because it relates to the incoming message (and
60 probably the interface on which it is arriving) and not to the outgoing
61 callout (which could be using a different interface). This has been
62 changed to use the value of the helo_data option from the smtp transport
63 instead - this is what is used when a message is actually being sent. If
64 there is no remote transport (possible with a router that sets up host
65 addresses), $smtp_active_hostname is used.
6ea85e9a 66
14aa5a05
PH
67PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
68 tweaks were necessary in order to get it to work:
69 (a) The code assumed that strncpy() returns a negative number on buffer
70 overflow, which isn't the case. Replaced with Exim's string_format()
71 function.
72 (b) There were several signed/unsigned issues. I just did the minimum
73 hacking in of casts. There is scope for a larger refactoring.
74 (c) The code used strcasecmp() which is not a standard C function.
75 Replaced with Exim's strcmpic() function.
76 (d) The code set only $1; it now sets $auth1 as well.
77 (e) A simple test gave the error "authentication client didn't specify
78 service in request". It would seem that Dovecot has changed its
79 interface. Fortunately there's a specification; I followed it and
80 changed what the client sends and it appears to be working now.
81
ff75a1f7
PH
82PH/13 Added $message_headers_raw to provide the headers without RFC 2047
83 decoding.
84
e6f6568e
PH
85PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
86 address A is aliased to B and C, where B exists and C does not. Without
87 -v the output is "A verified" because verification stops after a
88 successful redirection if more than one address is generated. However,
89 with -v the child addresses are also verified. Exim was outputting "A
90 failed to verify" and then showing the successful verification for C,
91 with its parentage. It now outputs "B failed to verify", showing B's
92 parentage before showing the successful verification of C.
93
d6f6e0dc
PH
94PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
95 look up a TXT record in a specific list after matching in a combined
96 list.
97
322050c2
PH
98PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
99 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
100 they consult the DNS. I had assumed they would set it the way they
101 wanted; and indeed my experiments on Linux seem to show that in some
102 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
103 To be on the safe side, however, I have now made the interface to
104 host_find_byname() similar to host_find_bydns(), with an argument
105 containing the DNS resolver options. The host_find_byname() function now
106 sets these options at its start, just as host_find_bydns() does. The smtp
107 transport options dns_qualify_single and dns_search_parents are passed to
108 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
109 of host_find_byname() use the default settings of RES_DEFNAMES
110 (qualify_single) but not RES_DNSRCH (search_parents).
111
08955dd3
PH
112PH/17 Applied (a modified version of) Nico Erfurth's patch to make
113 spool_read_header() do less string testing, by means of a preliminary
114 switch on the second character of optional "-foo" lines. (This is
115 overdue, caused by the large number of possibilities that now exist.
116 Originally there were few.) While I was there, I also converted the
117 str(n)cmp tests so they don't re-test the leading "-" and the first
118 character, in the hope this might squeeze out yet more improvement.
119
1eccaa59
PH
120PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
121 flag allowing group syntax was set by the header_syntax check but not
122 turned off, possible causing trouble later; (2) The flag was not being
123 set at all for the header_verify test, causing "group"-style headers to
124 be rejected. I have now set it in this case, and also caused header_
125 verify to ignore an empty address taken from a group. While doing this, I
126 came across some other cases where the code for allowing group syntax
127 while scanning a header line wasn't quite right (mostly, not resetting
128 the flag correctly in the right place). These bugs could have caused
129 trouble for malformed header lines. I hope it is now all correct.
130
602e59e5
PH
131PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
132 with the "reply" argument non-NULL. The code, however (which originally
133 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
134 but it didn't always do it. This confused somebody who was copying the
135 code for some other use. I have removed all the tests.
136
411ef850
PH
137PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
138 feature that was used to support insecure browsers during the U.S. crypto
139 embargo. It requires special client support, and Exim is probably the
140 only MTA that supported it -- and would never use it because real RSA is
141 always available. This code has been removed, because it had the bad
142 effect of slowing Exim down by computing (never used) parameters for the
143 RSA_EXPORT functionality.
144
944e9e9c 145
478be7b0
SC
146Exim version 4.63
147-----------------
148
149SC/01 Use a glob alias rather than an array ref in eximstats generated
150 parser. This improves both readability and performance.
151
152SC/02 Collect SpamAssassin and rejection statistics in eximstats.
153 Don't display local sender or destination tables in eximstats unless
154 there is data to show.
155 Added average volumes into the eximstats top table text output.
156
157SC/03 Collect data on the number of addresses (recipients) as well
158 as the number of messages in eximstats.
159
2b965a65
TF
160TF/01 Correct an error in the documentation for the redirect router. Exim
161 does (usually) call initgroups() when daemonizing.
478be7b0 162
45b91596
PH
163TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
164 with consistent privilege compared to when running as a daemon.
478be7b0 165
c59f5781
TF
166TF/03 Note in the spec that $authenticated_id is not set for local
167 submissions from trusted users.
168
90fc3069
TF
169TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
170 Thanks to Dean Brooks <dean@iglou.com> for the patch.
171
6083aca0
TF
172TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
173 by adding some example configuration directives to the default
174 configuration file. A little bit of work is required to uncomment the
175 directives and define how usernames and passwords are checked, but
176 there is now a framework to start from.
177
765b530f
PH
178PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
179 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
180 without this. I don't know how relevant this is to other LDAP libraries.
181
4e167a8c
PH
182PH/02 Add the verb name to the "unknown ACL verb" error.
183
4608d683
PH
184PH/03 Magnus Holmgren's patch for filter_prepend_home.
185
b8dc3e4a
PH
186PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
187
5418e93b
PH
188PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
189 directory not expanded when it should be if an expanded home directory
190 was set for the address (which is overridden by the transport).
191
b4a9bda2
PH
192PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
193 libradius.
194
45b91596
PH
195PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
196 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
197 because it is too late at that time, and has no effect.
198
5547e2c5
PH
199PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
200 security issue with \' (bugzilla #107). I could not use the
201 PQescapeStringConn() function, because it needs a PGconn value as one of
202 its arguments.
203
dbcef0ea
PH
204PH/08 When testing addresses using -bt, indicate those final addresses that
205 are duplicates that would not cause an additional delivery. At least one
206 person was confused, thinking that -bt output corresponded to deliveries.
207 (Suppressing duplicates isn't a good idea as you lose the information
208 about possibly different redirections that led to the duplicates.)
209
25257489
PH
210PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
211 systems where poll() doesn't work, in particular OS X.
212
c816d124
PH
213PH/10 Added more information to debugging output for retry time not reached.
214
a9ccd69a
PH
215PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
216 operations in malware.c.
217
75fa1910
PH
218PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
219 signatures.
220
a7d7aa58
PH
221PH/13 If write_rejectlog was set false when logging was sent to syslog with
222 syslog_duplication set false, log lines that would normally be written
223 both the the main log and to the reject log were not written to syslog at
224 all.
225
42119b09
PH
226PH/14 In the default configuration, change the use of "message" in ACL warn
227 statements to "add_header".
228
41609df5
PH
229PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
230 not followed by a command (e.g. "seen endif").
231
a5bd321b
PH
232PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
233 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
234 latter.
235
e85a7ad5 236PH/17 Added extra conditions to the default value of delay_warning_condition
5dff5817
PH
237 so that it is now:
238
e85a7ad5
PH
239 ${if or { \
240 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
241 { match{$h_precedence:}{(?i)bulk|list|junk} } \
242 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
5dff5817
PH
243 }{no}{yes}}
244
e85a7ad5
PH
245 The Auto-Submitted: and various List- headers are standardised, whereas I
246 don't think Precedence: ever was.
5dff5817 247
d8fe1c03
PH
248PH/18 Refactored debugging code in route_finduser() to show more information,
249 in particular, the error code if getpwnam() issues one.
250
16282d2b
PH
251PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
252 This is apparently needed in addition to the PH/07 change above to avoid
253 any possible encoding problems.
254
35d40a98
PH
255PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
256 but not after initializing Perl.
257
034d99ab
PH
258PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
259 output them only if debugging. By default they are written stderr,
260 apparently, which is not desirable.
261
6ec97b1b
PH
262PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
263 queries.
264
e22ca4ac
JJ
265JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
266 --not options
267
268JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
269
33d73e3b
PH
270PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
271 authenticated or an ident call has been made. Suppress the default
272 values for $authenticated_id and $authenticated_sender (but permit -oMai
273 and -oMas) when testing with -bh.
274
9ecb03f3
PH
275PH/24 Re-jigged the order of the tests in the default configuration so that the
276 tests for valid domains and recipients precede the DNS black list and CSA
277 tests, on the grounds that those ones are more expensive.
278
084efe8d
PH
279PH/25 Exim was not testing for a space following SMTP commands such as EHLO
280 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
281 This bug exists in every version of Exim that I still have, right back to
282 0.12.
283
366fc9f0
PH
284PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
285 However, an attempt to turn on case-sensitivity in a regex key by
286 including (?-i) didn't work because the subject string was already
287 lowercased, and the effects were non-intuitive. It turns out that a
288 one-line patch can be used to allow (?-i) to work as expected.
289
c59f5781 290
c887c79e
TF
291Exim version 4.62
292-----------------
293
294TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
295 other effects) broke the use of negated acl sub-conditions.
296
1cce3af8
PH
297PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
298 patch).
299
afb3eaaf
PH
300PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
301 "Deny" causes Exim to reject the incoming connection with a 554 error.
302 Unfortunately, if there is a major crisis, such as a disk failure,
303 tcp-wrappers gives "deny", whereas what one would like would be some
304 kind of temporary error. A kludge has been added to help with this.
305 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
306 554 error is used if errno is still zero or contains ENOENT (which occurs
307 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
308 451 error is used.
309
e173618b
PH
310PH/03 Add -lutil to the default FreeBSD LIBS setting.
311
dd16e114
PH
312PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
313 errors. Otherwise a message that provokes a temporary error (when other
314 messages do not) can cause a whole host to time out.
315
f7fd3850
PH
316PH/05 Batch deliveries by appendfile and pipe transports did not work when the
317 addresses were routed directly to files or pipes from a redirect router.
318 File deliveries just didn't batch; pipe deliveries might have suffered
319 odd errors.
320
d87df92c
PH
321PH/06 A failure to get a lock for a hints database would erroneously always say
322 "Failed to get write lock", even when it was really a read lock.
323
7e9f683d
PH
324PH/07 The appendfile transport was creating MBX lock files with a fixed mode
325 of 0600. This has been changed to use the value of the lockfile_mode
326 option (which defaults to 0600).
327
bfad5236
PH
328PH/08 Applied small patch from the Sieve maintainer.
329
01c490df
PH
330PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
331 folder from quota calculations, a direct delivery into this folder messed
332 up the contents of the maildirsize file. This was because the regex was
333 used only to exclude .Trash (or whatever) when the size of the mailbox
334 was calculated. There was no check that a delivery was happening into an
335 excluded directory. This bug has been fixed by ignoring all quota
336 processing for deliveries into excluded directories.
337
d6629cdc
PH
338PH/10 Added the maildirfolder_create_regex option to appendfile.
339
1cce3af8 340
214e2000
PH
341Exim version 4.61
342-----------------
343
344PH/01 The code for finding all the local interface addresses on a FreeBSD
345 system running IPv6 was broken. This may well have applied to all BSD
346 systems, as well as to others that have similar system calls. The broken
347 code found IPv4 interfaces correctly, but gave incorrect values for the
348 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
349 that it would not match correctly against @[] and not recognize the IPv6
350 addresses as local.
351
f9daeae0
PH
352PH/02 The ipliteral router was not recognizing addresses of the form user@
353 [ipv6:....] because it didn't know about the "ipv6:" prefix.
354
7e66e54d
PH
355PH/03 Added disable_ipv6.
356
c8ea1597
PH
357PH/04 Changed $reply_address to use the raw form of the headers instead of the
358 decoded form, because it is most often used to construct To: headers
359 lines in autoreplies, and the decoded form may well be syntactically
360 invalid. However, $reply_address has leading white space removed, and all
361 newlines turned into spaces so that the autoreply transport does not
362 grumble.
363
911f6fde
PH
364PH/05 If group was specified without a user on a router, and no group or user
365 was specified on a transport, the group from the router was ignored.
366
47ca6d6c
PH
367PH/06 Increased the number of ACL variables to 20 of each type, and arranged
368 for visible compile-time settings that can be used to change these
369 numbers, for those that want even more. Backwards compatibility with old
370 spool files has been maintained. However, going back to a previous Exim
371 release will lost any variables that are in spool files.
372
ed0e9820
PH
373PH/07 Two small changes when running in the test harness: increase delay when
374 passing a TCP/IP connection to a new process, in case the original
375 process has to generate a bounce, and remove special handling of
376 127.0.0.2 (sic), which is no longer necessary.
377
eff37e47
PH
378PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
379 be the same on different OS.
380
1921d2ea
PH
381PH/09 Moved a debug statement in filter processing to avoid a race problem when
382 testing.
383
b3f69ca8
JJ
384JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
385 whether --show-vars was specified or not
386
387JJ/02 exipick: Added support for new ACL variable spool format introduced
388 in 4.61-PH/06
389
424a1c63
PH
390PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
391 syntactically invalid From: or Reply-to: line, and a filter used this to
392 generate an autoreply, and therefore failed to obtain an address for the
393 autoreply, Exim could try to deliver to a non-existent relative file
394 name, causing unrelated and misleading errors. What now happens is that
395 it logs this as a hard delivery error, but does not attempt to create a
396 bounce message.
397
7a100415
PH
398PH/11 The exinext utility has a -C option for testing purposes, but although
399 the given file was scanned by exinext itself; it wasn't being passed on
400 when Exim was called.
401
19b9dc85
PH
402PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
403 an end-of-file indication when reading a command response.
404
309bd837
PH
405PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
406 compiled. In many other places in Exim, IPv6 addresses are always
407 recognized, so I have changed this. It also means that IPv4 domain
408 literals of the form [IPV4:n.n.n.n] are now always recognized.
409
59e82a2a
PH
410PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
411 used if the router is not running as root, for example, when verifying at
412 ACL time, or when using -bh. The debugging output from this situation was
413 non-existent - all you got was a failure to exec. I have made two
414 changes:
415
416 (a) Failures to set uid/gid, the current directory, or a process leader
417 in a subprocess such as that created by queryprogram now generate
418 suitable debugging ouput when -d is set.
419
420 (b) The queryprogram router detects when it is not running as root,
421 outputs suitable debugging information if -d is set, and then runs
422 the subprocess without attempting to change uid/gid.
423
9edc04ce
PH
424PH/15 Minor change to Makefile for building test_host (undocumented testing
425 feature).
426
1349e1e5
PH
427PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
428 additional section of a DNS packet that returns MX or SRV records.
429 Instead, it always explicitly searches for A/AAAA records. This avoids
430 major problems that occur when a DNS server includes only records of one
431 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
432 fixed another bug: if SRV records were looked up and the corresponding
433 address records were *not* found in the additional section, the port
434 values from the SRV records were lost.
435
ea49d0e1
PH
436PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
437 using the correct key (the original address) when searching the retry
438 rules in order to find which one to use for generating the retry hint.
439
064a94c9
PH
440PH/18 If quota_warn_message contains a From: header, Exim now refrains from
441 adding the default one. Similarly, if it contains a Reply-To: header, the
442 errors_reply_to option, if set, is not used.
443
727071f8
PH
444PH/19 When calculating a retry time, Exim used to measure the "time since
445 failure" by looking at the "first failed" field in the retry record. Now
446 it does not use this if it is later than than the arrival time of the
447 message. Instead it uses the arrival time. This makes for better
448 behaviour in cases where some deliveries succeed, thus re-setting the
449 "first failed" field. An example is a quota failure for a huge message
450 when small messages continue to be delivered. Without this change, the
451 "time since failure" will always be short, possible causing more frequent
452 delivery attempts for the huge message than are intended.
dd16e114 453 [Note: This change was subsequently modified - see PH/04 for 4.62.]
727071f8 454
f78eb7c6
PH
455PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
456 $1, $2, $3) because the numerical variables can be reset during some
457 expansion items (e.g. "match"), thereby losing the authentication data.
458
21c28500
PH
459PH/21 Make -bV show the size of off_t variables so that the test suite can
460 decide whether to run tests for quotas > 2G.
461
462PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
463 mailbox_size, and mailbox_filecount in the appendfile transport. If a
464 filecount value is greater than 2G or if a quota value is greater than 2G
465 on a system where the size of off_t is not greater than 4, a panic error
466 is given.
467
1688f43b
PH
468PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
469 never match. The debug and -bh output now contains an explicit error
470 message indicating a malformed IPv4 address or mask.
471
472PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
473 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
474 PH/23 above applies.
475
9675b384
PH
476PH/25 Do not write to syslog when running in the test harness. The only
477 occasion when this arises is a failure to open the main or panic logs
478 (for which there is an explicit test).
479
6a3f1455
PH
480PH/26 Added the /no_tell option to "control=freeze".
481
dac79d3e
PH
482PH/27 If a host name lookup failed very early in a connection, for example, if
483 the IP address matched host_lookup and the reverse lookup yielded a name
484 that did not have a forward lookup, an error message of the form "no IP
485 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
486 could be logged. Now it outputs the IP address instead of "NULL".
1349e1e5 487
5977a0b3
PH
488PH/28 An enabling patch from MH: add new function child_open_exim2() which
489 allows the sender and the authenticated sender to be set when
490 submitting a message from within Exim. Since child_open_exim() is
491 documented for local_scan(), the new function should be too.
492
c91535f3
PH
493PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
494 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
495 results in an empty string is now treated as unset.
496
0d46a8c8
PH
497PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
498
278c6e6c
PH
499PH/31 Added sender_verify_fail logging option.
500
2cbb4081
PH
501PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
502 needed by RFC 822 but not by RFC 2822 was commented out. I have now
503 tidied the source and removed it altogether.
504
3eef829e
PH
505PH/33 When a queue run was abandoned because the load average was too high, a
506 log line was always written; now it is written only if the queue_run log
507 selector is set. In addition, the log line for abandonment now contains
508 information about the queue run such as the pid. This is always present
509 in "start" and "stop" lines but was omitted from the "abandon" line.
510
1ab95fa6
PH
511PH/34 Omit spaces between a header name and the colon in the error message that
512 is given when verify = headers_syntax fails (if there are lots of them,
513 the message gets confusing).
514
230205fc
PH
515PH/35 Change the default for dns_check_names_pattern to allow slashes within
516 names, as there are now some PTR records that contain slashes. This check
517 is only to protect against broken name servers that fall over on strange
518 characters, so the fact that it applies to all lookups doesn't matter.
519
75e0e026
PH
520PH/36 Now that the new test suite is complete, we can remove some of the
521 special code in Exim that was needed for the old test suite. For example,
522 sorting DNS records because real resolvers return them in an arbitrary
523 order. The new test suite's fake resolver always returns records in the
524 same order.
525
526PH/37 When running in the test harness, use -odi for submitted messages (e.g.
527 bounces) except when queue_only is set, to avoid logging races between
528 the different processes.
529
145396a6
PH
530PH/38 Panic-die if .include specifies a non-absolute path.
531
3cd34f13
PH
532PH/39 A tweak to the "H" retry rule from its user.
533
11121d3d
JJ
534JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
535 a label. They prevented compilation on older perls.
536
537JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
538 a warning to be raised on newish perls.
539
540JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
541 on queue. Changes to match documented behaviour of showing count of
542 messages matching specified criteria.
543
8def5aaf
PH
544PH/40 Changed the default ident timeout from 30s to 5s.
545
929ba01c
PH
546PH/41 Added support for the use of login_cap features, on those BSD systems
547 that have them, for controlling the resources used by pipe deliveries.
548
2632889e
PH
549PH/42 The content-scanning code uses fopen() to create files in which to put
550 message data. Previously it was not paying any attention to the mode of
551 the files. Exim runs with umask(0) because the rest of the code creates
552 files with open(), and sets the required mode explicitly. Thus, these
553 files were ending up world-writeable. This was not a big issue, because,
554 being within the spool directory, they were not world-accessible. I have
555 created a function called modefopen, which takes an additional mode
556 argument. It sets umask(777), creates the file, chmods it to the required
557 mode, then resets the umask. All the relevant calls to fopen() in the
558 content scanning code have been changed to use this function.
559
944a9c55
PH
560PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
561 to 24 hours. This avoids potential overflow problems when processing G
562 and H retry rules. I suspect nobody ever tinkers with this value.
563
4a23603b
PH
564PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
565
4730f942
PH
566PH/45 When the plaintext authenticator is running as a client, the server's
567 challenges are checked to ensure they are valid base64 strings. By
568 default, the authentication attempt is cancelled if an invalid string is
569 received. Setting client_ignore_invalid_base64 true ignores these errors.
570 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
571 they are received. Thus, the responses can be made to depend on the
572 challenges. If an invalid string is ignored, an empty string is placed in
573 the variable.
574
30dba1e6
PH
575PH/46 Messages that are created by the autoreply transport now contains a
576 References: header, in accordance with RFCs 2822 and 3834.
577
382afc6b
PH
578PH/47 Added authenticated_sender_force to the smtp transport.
579
a86229cf
PH
580PH/48 The ${prvs expansion was broken on systems where time_t was long long.
581
50c99ba6
PH
582PH/49 Installed latest patch from the Sieve maintainer.
583
d35e429d
PH
584PH/50 When an Exim quota was set without a file count quota, and mailbox_size
585 was also set, the appendfile transport was unnecessarily scanning a
586 directory of message files (e.g. for maildir delivery) to find the count
587 of files (along with the size), even though it did not need this
588 information. It now does the scan only if it needs to find either the
589 size of the count of files.
590
f90d018c
PH
591PH/51 Added ${time_eval: to convert Exim time strings into seconds.
592
75def545
PH
593PH/52 Two bugs concerned with error handling when the smtp transport is
594 used in LMTP mode:
595
596 (i) Exim was not creating retry information for temporary errors given
597 for individual recipients after the DATA command when the smtp transport
598 was used in LMTP mode. This meant that they could be retried too
599 frequently, and not timed out correctly.
600
601 (ii) Exim was setting the flag that allows error details to be returned
602 for LMTP errors on RCPT commands, but not for LMTP errors for individual
603 recipients that were returned after the DATA command.
604
605PH/53 This is related to PH/52, but is more general: for any failing address,
606 when detailed error information was permitted to be returned to the
607 sender, but the error was temporary, then after the final timeout, only
608 "retry timeout exceeded" was returned. Now it returns the full error as
609 well as "retry timeout exceeded".
610
c46782ef
PH
611PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
612 do this, and (what is worse) MTAs that accept it.
613
71fafd95
PH
614PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
615 will now be deprecated.
616
2c5db4fd
PH
617PH/56 New os.c-cygwin from the Cygwin maintainer.
618
9cf6b11a
JJ
619JJ/06 exipick: added --unsorted option to allow unsorted output in all output
620 formats (previously only available in exim formats via -bpr, -bpru,
621 and -bpra. Now also available in native and exiqgrep formats)
622
623JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
624 with very large, slow to parse queues
625
626JJ/08 exipick: added ! as generic prefix to negate any criteria format
627
628JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
629
898d150f
PH
630PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
631 responses to authentication challenges, though it was showing the
632 challenges; (ii) I've removed the CR characters from the debug output for
633 SMTP output lines.
634
46218253
PH
635PH/58 Allow for the insertion of a newline as well as a space when a string
636 is turned into more than one encoded-word during RFC 2047 encoding. The
637 Sieve code now uses this.
638
e97957bc
PH
639PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
640 data_4xx, lost_connection, tls_required.
641
81e509d7
PH
642PH/60 When a VRFY deferred or FAILED, the log message rather than the user
643 message was being sent as an SMTP response.
644
3d240ff7
PH
645PH/61 Add -l and -k options to exicyclog.
646
b37c4101
PH
647PH/62 When verifying, if an address was redirected to one new address, so that
648 verification continued, and the new address failed or deferred after
649 having set something in $address_data, the value of $address_data was not
650 passed back to the ACL. This was different to the case when no
651 redirection occurred. The value is now passed back in both cases.
652
79378e0f
PH
653PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
654 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
655 use login_cap.h, so on its own it isn't the distinguishing feature. The
656 new name refers directly to the setclassresources() function.
657
e49c7bb4
PH
658PH/65 Added configuration files for NetBSD3.
659
d114ec46
PH
660PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
661
f3d7df6c
PH
662PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
663 is preferred over IPv4.
664
715ab376
PH
665PH/68 The bounce_return_message and bounce_return_body options were not being
666 honoured for bounces generated during the reception of non-SMTP messages.
667 In particular, this applied to messages rejected by the ACL. This bug has
668 been fixed. However, if bounce_return_message is true and bounce_return_
669 body is false, the headers that are returned for a non-SMTP message
670 include only those that have been read before the error was detected.
671 (In the case of an ACL rejection, they have all been read.)
672
6b31b150
PH
673PH/69 The HTML version of the specification is now built in a directory called
674 spec_html instead of spec.html, because the latter looks like a path with
675 a MIME-type, and this confuses some software.
676
677PH/70 Catch two compiler warnings in sieve.c.
678
d515a917
PH
679PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
680 function verify_get_ident() calls ip_connect() to connect a socket, but
681 if the "connect()" function timed out, ip_connect() used to close the
682 socket. However, verify_get_ident() also closes the socket later, and in
683 between Exim writes to the log, which may get opened at this point. When
684 the socket was closed in ip_connect(), the log could get the same file
685 descriptor number as the socket. This naturally causes chaos. The fix is
686 not to close the socket in ip_connect(); the socket should be closed by
687 the function that creates it. There was only one place in the code where
688 this was missing, in the iplookup router, which I don't think anybody now
689 uses, but I've fixed it anyway.
690
9b8fadde
PH
691PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
692 well as to direct DNS lookups. Otherwise the handling of names in host
693 lists is inconsistent and therefore confusing.
694
214e2000 695
5de37277
PH
696Exim version 4.60
697-----------------
698
cc38ddbf
PH
699PH/01 Two changes to the default runtime configuration:
700
701 (1) Move the checks for relay_from_hosts and authenticated clients from
702 after to before the (commented out) DNS black list checks.
703
704 (2) Add control=submission to the relay_from_hosts and authenticated
705 clients checks, on the grounds that messages accepted by these
706 statements are most likely to be submissions.
5de37277 707
72fdd6ae
PH
708PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
709
710 (1) Generate an error if the third argument for the ${prvs expansion is
711 not a single digit.
712
713 (2) Treat a missing third argument of ${prvscheck as if it were an empty
714 string.
715
716 (3) Reset the variables that are obtained from the first argument of
717 ${prvscheck and used in the second argument before leaving the code,
718 because their memory is reclaimed, so using them afterwards may do
719 silly things.
720
721 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
722 one (it's much easier than Tom thought :-).
723
724 (5) Because of (4), we can now allow for the use of $prvscheck_result
725 inside the third argument.
cb9328de 726
cb741023
PH
727PH/03 For some reason, the default setting of PATH when running a command from
728 a pipe transport was just "/usr/bin". I have changed it to
729 "/bin:/usr/bin".
730
f174f16e
PH
731PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
732 anything to be listed in the output from -bV.
b2f5a032 733
c25242d7
PH
734PH/05 When a filter generated an autoreply, the entire To: header line was
735 quoted in the delivery log line, like this:
736
737 => >A.N.Other <ano@some.domain> <original@ddress> ...
738
739 This has been changed so that it extracts the operative address. There
740 may be more than one such address. If so, they are comma-separated, like
741 this:
742
743 => >ano@some.domain,ona@other.domain <original@ddress> ...
744
82c19f95
PH
745PH/06 When a client host used a correct literal IP address in a HELO or EHLO
746 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
747 not being looked up in the rDNS to get a host name, Exim was showing the
748 IP address twice in Received: lines, even though the IP addresses were
749 identical. For example:
750
751 Received: from [1.2.3.4] (helo=[1.2.3.4])
752
753 However, if the real host name was known, it was omitting the HELO data
754 if it matched the actual IP address. This has been tidied up so that it
755 doesn't show the same IP address twice.
756
d7ffbc12
PH
757PH/07 When both +timestamp and +memory debugging was on, the value given by
758 $tod_xxx expansions could be wrong, because the tod_stamp() function was
759 called by the debug printing, thereby overwriting the timestamp buffer.
760 Debugging no longer uses the tod_stamp() function when +timestamp is set.
761
9f526266
PH
762PH/08 When the original message was included in an autoreply transport, it
763 always said "this is a copy of the message, including all the headers",
764 even if body_only or headers_only was set. It now gives an appropriate
765 message.
766
87fcc8b9
PH
767PH/09 Applied a patch from the Sieve maintainer which:
768
769 o fixes some comments
770 o adds the (disabled) notify extension core
771 o adds some debug output for the result of if/elsif tests
772 o points to the current vacation draft in the documentation
773 and documents the missing references header update
774
775 and most important:
776
777 o fixes a bug in processing the envelope test (when testing
778 multiple envelope elements, the last element determinted the
779 result)
780
456682f5
PH
781PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
782 Electronic Mail") by including:
783
784 Auto-submitted: auto-generated
785
786 in the messages that it generates (bounce messages and others, such as
787 warnings). In the case of bounce messages for non-SMTP mesages, there was
788 also a typo: it was using "Auto_submitted" (underscore instead of
789 hyphen). Since every message generated by Exim is necessarily in response
790 to another message, thes have all been changed to:
791
792 Auto-Submitted: auto-replied
793
794 in accordance with these statements in the RFC:
795
796 The auto-replied keyword:
797
798 - SHOULD be used on messages sent in direct response to another
799 message by an automatic process,
800
801 - MUST NOT be used on manually-generated messages,
802
803 - MAY be used on Delivery Status Notifications (DSNs) and Message
804 Disposition Notifications (MDNs),
805
806 - MUST NOT be used on messages generated by automatic or periodic
807 processes, except for messages which are automatic responses to
808 other messages.
809
3e46c1aa
PH
810PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
811 to the default Received: header definition.
456682f5 812
49826d12
PH
813PH/12 Added log selector acl_warn_skipped (default on).
814
eba0c039
PH
815PH/13 After a successful wildlsearch lookup, discard the values of numeric
816 variables because (a) they are in the wrong storage pool and (b) even if
817 they were copied, it wouldn't work properly because of the caching.
818
a0d6ba8a
PH
819PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
820 checking when decoding. Apparently there are clients that generate
821 overlong encoded strings. Why am I not surprised?
822
f0917727
PH
823PH/15 If the first argument of "${if match_address" was not empty, but did not
824 contain an "@" character, Exim crashed. Now it writes a panic log message
825 and treats the condition as false.
826
096fee00
PH
827PH/16 In autoreply, treat an empty string for "once" the same as unset.
828
024bd3c2
PH
829PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
830 extension "envelope-auth". The code is finished and in agreement with
831 other implementations, but there is no documentation so far and in fact,
832 nobody wrote the draft yet. This extension is currently #undef'ed, thus
833 not changing the active code.
834
835 Print executed "if" and "elsif" statements when debugging is used. This
836 helps a great deal to understand what a filter does.
837
838 Document more things not specified clearly in RFC3028. I had all this
839 sorted out, when out of a sudden new issues came to my mind. Oops."
840
df199fec
PH
841PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
842 (Bugzilla #53).
843
d27f1df3
PH
844PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
845 canonical form (as documented). However, after a host name lookup from
846 the IP address, check_host() was doing a simple string comparison with
847 addresses acquired from the DNS when checking that the found name did
848 have the original IP as one of its addresses. Since any found IPv6
849 addresses are likely to be in abbreviated form, the comparison could
850 fail. Luckily, there already exists a function for doing the comparison
851 by converting both addresses to binary, so now that is used instead of
852 the text comparison.
853
96776534
PH
854PH/20 There was another similar case to PH/19, when a complete host name was
855 given in a host list; looking up its IP address could give an abbreviated
856 form, whereas the current host's name might or might not be abbreviated.
857 The same fix has been applied.
858
5de37277 859
9a799bc0
PH
860Exim version 4.54
861-----------------
862
863PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
864 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
865 It now does.
866
99a4b039
PH
867PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
868 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
869
870PH/03 Typo: missing ".o" in src/pcre/Makefile.
871
4b233853
PH
872PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
873 header line, restrict the check to what is listed in RFCs 2369 and 2929.
874 Also, for "Auto-Submitted", treat anything other than "no" as
875 non-personal, in accordance with RFC 3834. (Previously it treated
876 anything starting "auto-" as non-personal.)
877
8857ccfd
PH
878TF/01 The control=submission/name=... option had a problem with syntax
879 errors if the name included a slash character. The /name= option
880 now slurps the rest of the string, so it can include any characters
881 but it must come last in the list of options (after /sender_retain
882 or /domain=).
883
433a2980
PH
884PH/05 Some modifications to the interface to the fake nameserver for the new
885 testing suite.
886
3e46c1aa 887
9a799bc0 888
e3a311ba
TK
889Exim version 4.53
890-----------------
891
892TK/01 Added the "success_on_redirect" address verification option. See
893 NewStuff for rationale and an example.
894
13b685f9
PH
895PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
896
395ff96d
PH
897PH/02 Patch to exigrep to allow it to work on syslog lines.
898
5b68f6e4
PH
899PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
900 fread() to skip over the body file's header line, because in Cygwin the
901 header line is locked and is inaccessible.
902
1ab52c69
PH
903PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
904 co-exist for some time) to make it clear that it is the Exim ID that is
905 referenced, not the Message-ID: header line.
906
b07e6aa3
PH
907PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
908 string_format() function, because snprintf() does not exist on all
909 operating systems.
910
254e032f
PH
911PH/06 The use of forbid_filter_existstest now also locks out the use of the
912 ${stat: expansion item.
913
3af76a81
PH
914PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
915 protocol synchronization error", to keep the pedants happy.
916
2548ba04
PH
917PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
918 well as for IRIX systems, when gcc is being used. See the host.c source
919 file for comments.
920
b6c6011d
PH
921PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
922
cf39cf57
PH
923PH/10 Named domain lists were not working if used in a queue_smtp_domains
924 setting.
925
f1513293
PH
926PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
927 transport and to the smtp transport in LMTP mode.
928
727549a4
PH
929TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
930
af46795e
PH
931PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
932 run a filter in a subprocess. This could lead to confusion in subsequent
933 lookups in the parent process. There should also be a search_tidyup() at
934 the end of the subprocess.
935
d7b47fd0
PH
936PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
937 only if the host matched helo_try_verify_hosts, which caused the
938 verification to occur when the EHLO/HELO command was issued. The ACL just
939 tested the remembered result. Now, if a previous verification attempt has
940 not happened, "verify = helo" does it there and then.
941
ee744174
JJ
942JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
943
b582ab87
PH
944TK/03 Fix log output including CR from clamd.
945
41a13e0a
PH
946PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
947 exist provoked a memory error which could cause a segfault.
948
f625cc5a
PH
949PH/15 Installed PCRE 6.2
950
951PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
952
21f7af35
PH
953PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
954 of the problem. Specifically, suggested +O2 rather than +O1 for the
955 HP-UX compiler.
956
31480e42
PH
957PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
958
2d280592
PH
959PH/20 If a delivery was routed to a non-standard port by means of an SRV
960 record, the port was not correctly logged when the outgoing_port log
961 selector was set (it logged the transort's default port).
962
7cd1141b
PH
963PH/21 Added support for host-specific ports to manualroute, queryprogram,
964 fallback_hosts, and "hosts" in the smtp transport.
965
966PH/22 If the log selector "outgoing_port" is set, the port is now also given on
967 host errors such as "Connection refused".
968
750af86e
PH
969PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
970 authentication with radiusclient 0.4.9:
971
972 - Error returned from rc_read_config was caught wrongly
973 - Username/password not passed on to radius server due to wrong length.
974
975 The presumption is that some radiusclient API changes for 4.51/PH/17
976 were not taken care of correctly. The code is still untested by me (my
977 Linux distribution still has 0.3.2 of radiusclient), but it was
978 contributed by a Radius user.
979
980PH/24 When doing a callout, the value of $domain wasn't set correctly when
981 expanding the "port" option of the smtp transport.
982
4304270b
TK
983TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
984 while reading a MIME header. Thanks to Tom Hughes for a patch.
985
750af86e
PH
986PH/24 Include config.h inside local_scan.h so that configuration settings are
987 available.
988
64ffc24f
PH
989PH/25 Make $smtp_command_argument available after all SMTP commands. This means
990 that in an ACL for RCPT (for example), you can examine exactly what was
991 received.
992
5dd9625b
PH
993PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
994 commands, but it was not correctly comparing the address with the actual
995 client host address. Thus, it would show the EHLO address in Received:
996 header lines when this was not necessary.
997
5591031b
PH
998PH/27 Added the % operator to ${eval:}.
999
ba18e66a
PH
1000PH/28 Exim tries to create and chdir to its spool directory when it starts;
1001 it should be ignoring failures (because with -C, for example, it has lost
1002 privilege). It wasn't ignoring creation failures other than "already
1003 exists".
1004
9cec981f
PH
1005PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1006 -bV or -d is used.
1007
aa2b5c79
PH
1008PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1009 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
1010 too many continuations, the temporary file was not being removed, and the
1011 return code was incorrect.
aa2b5c79 1012
48a53b7f
PH
1013PH/31 Missing "BOOL" in function definition in filtertest.c.
1014
1c59d63b
PH
1015PH/32 Applied Sieve patches from the maintainer.
1016
671012da
TK
1017TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1018
1509d3a8
PH
1019PH/33 Added "verify = not_blind".
1020
1021PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1022 Local/Makefile (with some defaults set). These are used in built scripts
1023 such as exicyclog, but they have never been used in the exim_install
1024 script (though there are many overriding facilities there). I have
1025 arranged that the exim_install script now takes note of these two
1026 settings.
1027
1028PH/35 Installed configuration files for Dragonfly.
1029
2fe1a124
PH
1030PH/36 When a locally submitted message by a trusted user did not contain a
1031 From: header, and the sender address was obtained from -f or from an SMTP
1032 MAIL command, and the trusted user did not use -F to supply a sender
1033 name, $originator_name was incorrectly used when constructing a From:
1034 header. Furthermore, $originator_name was used for submission mode
1035 messages from external hosts without From: headers in a similar way,
1036 which is clearly wrong.
1037
8800895a
PH
1038PH/37 Added control=suppress_local_fixups.
1039
ccfdb010
PH
1040PH/38 When log_selector = +received_sender was set, and the addition of the
1041 sender made the log line's construction buffer exactly full, or one byte
1042 less than full, an overflow happened when the terminating "\n" was
1043 subsequently added.
1044
1130bfb0
PH
1045PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1046 when the result of a list match is failure because a DNS lookup failed.
1047
ebcb507f
PH
1048PH/40 RM_COMMAND is now used in the building process.
1049
c35e155c
PH
1050PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1051 the "build-* directories that it finds.
1052
95d1f782
PH
1053PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1054 address in a domain literal was a prefix of an interface address.
1055
fd6de02e
PH
1056PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1057 when verifying a sender address, unless rewrite_headers is false.
1058
58de37c5
PH
1059PH/44 Wrote a long comment about why errors_to addresses are verified as
1060 recipients, not senders.
1061
261cf466
TF
1062TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1063 the ratelimit ACL was added.
1064
3ee512ff
PH
1065PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1066
e08c430f
PH
1067PH/46 Added extra information about PostgreSQL errors to the error string.
1068
bef5a11f
PH
1069PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1070 avoiding the need to install special zones in a real server. This is
1071 backwards compatible; if it can't find the fake resolver, it drops back.
1072 Thus, both old and new test suites can be run.
1073
7546de58
TF
1074TF/02 Added util/ratelimit.pl
1075
e5d5a95f
TF
1076TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1077 clock is set back in time.
1078
2e88a017
TF
1079TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1080 Candler <B.Candler@pobox.com>.
1081
a5f65aa4
TF
1082TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1083 OK for addresses that are the result of redirections.
1084
e7726cbf
PH
1085PH/48 A number of further additions for the benefit of the new test suite,
1086 including a fake gethostbyname() that interfaces to the fake DNS resolver
1087 (see PH/47 above).
1088
a7fdad5b
TF
1089TF/06 The fix for widen_domains has also been applied to qualify_single and
1090 search_parents which are the other dnslookup options that can cause
1091 header rewrites.
1092
6af56900
PH
1093PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1094 type ("H").
1095
0925ede6
PH
1096PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1097
66afa403
TF
1098TF/07 Exim produced the error message "an SRV record indicated no SMTP
1099 service" if it encountered an MX record with an empty target hostname.
1100 The message is now "an MX or SRV record indicated no SMTP service".
1101
0154e85a
TF
1102TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1103 if the DNS of the sending site is misconfigured. This is quite a
1104 common situation. This change restores the behaviour of treating a
1105 helo verification defer as a failure.
1106
16f12c76
PH
1107PH/51 If self=fail was set on a router, the bounce message did not include the
1108 actual error message.
1109
bbe902f0 1110
e5a9dba6
PH
1111Exim version 4.52
1112-----------------
1113
1114TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1115
22c3b60b
PH
1116PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1117 command itself ended in error, the underlying message about the transport
1118 filter timeout was being overwritten with the pipe command error. Now the
1119 underlying error message should be appended to the second error message.
1120
06a9b4b5
PH
1121TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1122
c1ac6996
PH
1123PH/02 Reduce the amount of output that "make" produces by default. Full output
1124 can still be requested.
1125
9c7a242c
PH
1126PH/03 The warning log line about a condition test deferring for a "warn" verb
1127 was being output only once per connection, rather than after each
1128 occurrence (because it was using the same function as for successful
1129 "warn" verbs). This seems wrong, so I have changed it.
1130
87ba3f5f
PH
1131TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1132 it should not have, which might have caused a crash in the right
1133 circumstances, but probably never did.
1134
1135PH/04 Installed a modified version of Tony Finch's patch to make submission
1136 mode fix the return path as well as the Sender: header line, and to
1137 add a /name= option so that you can make the user's friendly name appear
1138 in the header line.
1139
29aba418
TF
1140TF/03 Added the control = fakedefer ACL modifier.
1141
fe0dab11
TF
1142TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1143 Mark Lowes for thorough testing.
870f6ba8 1144
11d337a4
TK
1145TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1146
1147TK/03 Merged latest SRS patch from Miles Wilton.
1148
415c8f3b
PH
1149PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1150 with the definition in sysexits.h (which is #included earlier).
1151 Fortunately, Exim does not actually use EX_OK. The code used to try to
1152 preserve the sysexits.h value, by assumimg that macro definitions were
1153 scanned for macro replacements. I have been disabused of this notion,
1154 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 1155
958541e9
PH
1156PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1157 in the smtp transport. When a block could not be written in a single
1158 write() function, the timeout was being re-applied to each part-write.
1159 This seems wrong - if the receiver was accepting one byte at a time it
1160 would take for ever. The timeout is now adjusted when this happens. It
1161 doesn't have to be particularly precise.
1162
c206415f
TK
1163TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1164 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1165
2a4be8f9
PH
1166PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1167 without a domain if the check to <postmaster@domain> fails.
1168
1cba11c5
SC
1169SC/01 Eximstats: added -xls and the ability to specify output files
1170 (patch written by Frank Heydlauf).
1171
1172SC/02 Eximstats: use FileHandles for outputing results.
1173
1174SC/03 Eximstats: allow any combination of xls, txt, and html output.
1175
1176SC/04 Eximstats: fixed display of large numbers with -nvr option
1177
1178SC/05 Eximstats: fixed merging of reports with empty tables.
1179
1180SC/06 Eximstats: added the -include_original_destination flag
1181
1182SC/07 Eximstats: removed tabs and trailing whitespace.
1183
1005d00e
TK
1184TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1185
1186TK/06 MBOX spool code: Add real "From " MBOX separator line
1187 so the .eml file is really in mbox format (even though
1188 most programs do not really care). Patch from Alex Miller.
1189
1190TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1191 The latter is generated from $received_to and is only set if the
1192 message has one envelope recipient. SA can use these headers,
1193 obviously out-of-the-box. Patch from Alex Miller.
1194
9b4768fa
PH
1195PH/08 The ${def test on a variable was returning false if the variable's
1196 value was "0", contrary to what the specification has always said!
1197 The result should be true unless the variable is empty.
1198
1199PH/09 The syntax error of a character other than { following "${if
1200 def:variable_name" (after optional whitespace) was not being diagnosed.
1201 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1202 accidental colon was present, for example, could give incorrect results.
1203
0d7eb84a
PH
1204PH/10 Tidied the code in a number of places where the st_size field of a stat()
1205 result is used (not including appendfile, where other changes are about
1206 to be made).
1207
1208PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1209 This involved changing a lot of size variables from int to off_t. It
1210 should work with maildirs and everything.
1211
40727bee
TK
1212TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1213 spamd dying while we are connected to it.
1214
554d2369
TF
1215TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1216 <jgh@wizmail.org>
1217
1f922db1
PH
1218PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1219 function for BDB 4.3.
1220
ef213c3b
PH
1221PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1222
8ac170f3
PH
1223PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1224 the sources for PCRE have been re-arranged and more files are now
1225 involved.
1226
b1c749bb
PH
1227PH/15 The code I had for printing potentially long long variables in PH/11
1228 above was not the best (it lost precision). The length of off_t variables
1229 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
1230 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1231 to be "long long int" or "long int". This is needed for the internal
1232 formatting function string_vformat().
b1c749bb 1233
4aac9b49
PH
1234PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1235 the configuration file to be ":syslog", then the script "guesses" where
1236 the logs files are, rather than using the compiled in default. In our
1237 case the guess is not the same as the compiled default, so the script
1238 suddenly stopped working when I started to use syslog. The patch checks
1239 to see if log_file_path is "". If so, it attempts to read it from exim
1240 with no configuration file to get the compiled in version, before it
1241 falls back to the previous guessing code."
1242
294520c8
TK
1243TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1244 implementing BATV in an Exim configuration. See NewStuff for the gory
1245 details.
1246
5bd022fe
PH
1247PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1248 Makefile that are specific to HP-UX.
1249
90e9ce59
PH
1250PH/18 If the "use_postmaster" option was set for a recipient callout together
1251 with the "random" option, the postmaster address was used as the MAIL
1252 FROM address for the random test, but not for the subsequent recipient
1253 test. It is now used for both.
1254
5ea81592
PH
1255PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1256 patch removes a few documentation additions to RFC 3028, because the
1257 latest draft now contains them. It adds the new en;ascii-case comparator
1258 and a new error check for 8bit text in MIME parts. Comparator and
1259 require names are now matched exactly. I enabled the subaddress
1260 extension, but it is not well tested yet (read: it works for me)."
1261
c6c2dc1d
PH
1262PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1263 rework some of the code of TK/09 above to avoid the hardwired use of
1264 "%lld" and "long long". Replaced the call to snprintf() with a call to
1265 string_vformat().
1266
fffffe4c
PH
1267PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1268 records point to non-existent hosts", "retry timeout exceeded", and
1269 "retry time not reached for any host after a long failure period".
ca02eafb 1270
9a26b6b2
PH
1271PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1272 experimental DomainKeys support:
1273
1274 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1275 (2) On an error such as an illegally used "control", the wrong name for
1276 the control was given.
1277
1278 These problems did NOT occur unless DomainKeys support was compiled.
1279
4aee0225
PH
1280PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1281
32d668a5
PH
1282PH/24 Added ${if match_ip condition.
1283
8187c3f3
PH
1284PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1285 databases so that it will be absolutely obvious if a crash occurs in the
1286 DB library. This is a regular occurrence (often caused by mis-matched
1287 db.h files).
1288
ff790e47 1289PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
1290 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1291 picked up on a user's system that detects such things. There doesn't seem
1292 to be a gcc warning option for this - only an attribute that has to be
1293 put on the function's prototype. It seems that in Fedora Core 4 they have
1294 set this on a number of new functions. No doubt there will be more in due
1295 course.
ff790e47 1296
5417f6d1
PH
1297PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1298 specify a transport. However, if an address that was verified by such a
1299 router was the subject of a callout, Exim crashed because it tried to
1300 read the rcpt_include_affixes from the non-existent transport. Now it
1301 just assumes that the setting of that option is false. This bug was
1302 introduced by 4.51/PH/31.
1303
59cf8544
PH
1304PH/28 Changed -d+all to exclude +memory, because that information is very
1305 rarely of interest, but it makes the output a lot bigger. People tend to
1306 do -d+all out of habit.
1307
e7ad8a65
PH
1308PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1309 code in os-type was giving problems when libc.so lives in lib64, like on
1310 x86_64 Fedora Core.
1311
ade42478
PH
1312PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1313 aren't the modern standard, and it seems that some systems' include files
1314 don't always have them. Exim was already checking for some of the newer
1315 ones like T_AAAA, and defining it itself. I've added checks for all the
1316 record types that Exim uses.
1317
182ad5cf
PH
1318PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1319 not automatically generating a new one, as it is supposed to. This
1320 prevented TLS from working. If the file did exist, but contained invalid
1321 data, a new version was generated, as expected. It was only the case of a
1322 non-existent file that was broken.
1323
b0d9fc80
TK
1324TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1325 with a change in libdomainkeys > 0.64.
1326
1327TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1328 from DNS. If the selector record carries the flag, it now has
1329 precedence over the domain-wide flag.
1330
1331TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1332
47c7a64a
PH
1333PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1334 the use of a port name that isn't defined in /etc/services) occurred, the
1335 message was deferred as in a normal delivery, and thus remained on the
1336 spool, instead of being failed because of the mua_wrapper setting. This
1337 is now fixed, and I tidied up some of the mua_wrapper messages at the
1338 same time.
1339
a388bce4
SC
1340SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1341 the messages in a hash of arrays rather than using individual hashes.
1342 This is a bit cleaner and results in dramatic memory savings, albeit
1343 at a slight CPU cost.
1344
1345SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1346 as requested by Marc Sherman.
1347
1348SC/10 Eximstats: added histograms for user specified patterns as requested
1349 by Marc Sherman.
1350
0793e4ed
SC
1351SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1352
c58b88df
PH
1353PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1354 fopen() in the content-scanning modules that did not already have it.
1355
e7ad8a65 1356
7982096b
PH
1357Exim version 4.51
1358-----------------
1359
1a46a8c5
PH
1360TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1361 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1362
2f079f46 1363TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
1364
1365TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1366
1367PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1368 to test Sieve filters that use "vacation".
1369
1370PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1371 that changes the way the GnuTLS parameters are stored in the cache file.
1372 The new format can be generated externally. For backward compatibility,
1373 if the data in the cache doesn't make sense, Exim assumes it has read an
1374 old-format file, and it generates new data and writes a new file. This
1375 means that you can't go back to an older release without removing the
1376 file.
1377
1378PH/03 A redirect router that has both "unseen" and "one_time" set does not
1379 work if there are any delivery delays because "one_time" forces the
1380 parent to be marked "delivered", so its unseen clone is never tried
1381 again. For this reason, Exim now forbids the simultaneous setting of
1382 these two options.
1383
1384PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1385 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1386 entry if you want to know the details. The fix, however, neglected to
1387 consider the case when local delivery batching is involved. The test for
1388 "previously delivered" was not happening when checking to see if an
1389 address could be batched with a previous (undelivered) one; under
1390 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 1391 address.
1a46a8c5
PH
1392
1393PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1394 in its include files, and this causes problems building Exim.
1395
1396PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1397 header_syntax) but Exim was just ignoring anything given after a slash.
1398 In particular, this caused confusion with an attempt to use "verify =
1399 reverse_host_lookup/defer_ok". An error is now given when options are
1400 supplied for verify items that do not have them. (Maybe reverse_host_
1401 lookup should have a defer_ok option, but that's a different point.)
1402
1403PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1404 defined by RFC 821) to 2048, because there were problems with some AUTH
1405 commands, and RFC 1869 says the size should be increased for extended
1406 SMTP commands that take arguments.
1407
1408PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1409 Finch).
1410
1411PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1412 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 1413
49c2d5ea
PH
1414PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1415 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
1416 whatever). Otherwise the password may appear in the log. Following change
1417 PH/42 below, there is no longer a chance of it appearing in a bounce
1418 message.
49c2d5ea 1419
bf759a8b
PH
1420PH/11 Installed exipick version 20050225.0 from John Jetmore.
1421
83364d30
PH
1422PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1423 of its addresses was ever tried. (Bugzilla bug #2.)
1424
7999bbd7
PH
1425PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1426 the result incorrectly in the debug output. (It correctly added a newline
1427 to what was transported.)
1428
7dbf77c9
PH
1429TF/01 Added $received_time.
1430
74e0617f
PH
1431PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1432 commented out examples of how to interface to a virus scanner and to
1433 SpamAssassin. Also added commented examples of av_scanner and
1434 spamd_address settings.
1435
2f079f46
PH
1436PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1437 and controls are allowed in which ACLs. There were a couple of minor
1438 errors. Some of the entries in the conditions table (which is a table of
1439 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1440 negation of where the condition IS allowed.
1441
8c841523
PH
1442PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1443
7766a4f0
PH
1444PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1445 header file does not have a version number, so I've had to invent a new
1446 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1447 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1448 radiusclient), but it was contributed by a Radius user.
1449
8b417f2c
PH
1450PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1451 files or directories, for OpenSSL.
1452
901f42cb
PH
1453PH/19 When an Exim process that is running as root has to create an Exim log
1454 file, it does so in a subprocess that runs as exim:exim so as to get the
1455 ownership right at creation (otherwise, other Exim processes might see
1456 the file with the wrong ownership). There was no test for failure of this
1457 fork() call, which would lead to the process getting stuck as it waited
1458 for a non-existent subprocess. Forks do occasionally fail when resources
1459 run out. I reviewed all the other calls to fork(); they all seem to check
1460 for failure.
1461
f9b9210e
PH
1462PH/20 When checking for unexpected SMTP input at connect time (before writing
1463 the banner), Exim was not dealing correctly with a non-positive return
1464 from the read() function. If the client had disconnected by this time,
1465 the result was a log entry for a synchronization error with an empty
1466 string after "input=" when read() returned zero. If read() returned -1
1467 (an event I could not check), uninitialized data bytes were printed.
1468 There were reports of junk text (parts of files, etc) appearing after
1469 "input=".
1470
54cdb463
PH
1471PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1472
cf00dad6
PH
1473PH/22 Added support for macro redefinition, and (re)definition in between
1474 driver and ACL definitions.
1475
acb1b346
PH
1476PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1477 forgetting to use the resulting value; it was using the unexpanded value.
1478
c5ddb310
PH
1479PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1480 hadn't been configured. The fix is from Juergen Kreileder, who
1481 understands it better than I do:
1482
1483 "Here's what I see happening with three configured cyrus_sasl
1484 authenticators configured (plain, login, cram-md5):
1485
1486 On startup auth_cyrus_sasl_init() gets called for each of these.
1487 This means three calls to sasl_listmech() without a specified mech_list.
1488 => SASL tests which mechs of all available mechs actually work
1489 => three warnings about OTP not working
1490 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1491
1492 With the patch, sasl_listmech() also gets called three times. But now
1493 SASL's mech_list option is set to the server_mech specified in the the
1494 authenticator. Or in other words, the answer from sasl_listmech()
1495 gets limited to just the mech you're testing for (which is different
1496 for each call.)
1497 => the return list contains just 'plain' or 'login', 'cram-md5' or
1498 nothing depending on the value of ob->server_mech.
1499
1500 I've just tested the patch: Authentication still works fine,
1501 unavailable mechs specified in the exim configuration are still
1502 caught, and the auth.log warnings about OTP are gone."
1503
31619da6
PH
1504PH/25 When debugging is enabled, the contents of the command line are added
1505 to the debugging output, even when log_selector=+arguments is not
1506 specified.
1507
bebaf0fc
PH
1508PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1509 answer is "GNU", and only if the return is "GNU/something" is the answer
1510 "Linux".
1511
475fe28a
PH
1512PH/27 $acl_verify_message is now set immediately after the failure of a
1513 verification in an ACL, and so is available in subsequent modifiers. In
1514 particular, the message can be preserved by coding like this:
1515
1516 warn !verify = sender
1517 set acl_m0 = $acl_verify_message
1518
1519 Previously, $acl_verify_message was set only while expanding "message"
1520 and "log_message" when a very denied access.
1521
7e8bec7a
PH
1522PH/28 Modified OS/os.c-Linux with
1523
1524 -#ifndef OS_LOAD_AVERAGE
1525 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1526
1527 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1528 nomenclature these days.)
1529
e4a89c47
PH
1530PH/29 Installed patch from the Sieve maintainer that adds the options
1531 sieve_useraddress and sieve_subaddress to the redirect router.
1532
5ca2a9a1
PH
1533PH/30 In these circumstances:
1534 . Two addresses routed to the same list of hosts;
1535 . First host does not offer TLS;
1536 . First host accepts first address;
1537 . First host gives temporary error to second address;
1538 . Second host offers TLS and a TLS session is established;
1539 . Second host accepts second address.
1540 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1541 and peerdn, if requested) that were in fact used only for the second
1542 address.
7e8bec7a 1543
c688b954
PH
1544PH/31 When doing a callout as part of verifying an address, Exim was not paying
1545 attention to any local part prefix or suffix that was matched by the
1546 router that accepted the address. It now behaves in the same way as it
1547 does for delivery: the affixes are removed from the local part unless
1548 rcpt_include_affixes is set on the transport.
1549
fed77020
PH
1550PH/32 Add the sender address, as F=<...>, to the log line when logging a
1551 timeout during the DATA phase of an incoming message.
1552
7fe1560f
PH
1553PH/33 Sieve envelope tests were broken for match types other than :is. I have
1554 applied a patch sanctioned by the Sieve maintainer.
c688b954 1555
ebb6e6d5
PH
1556PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1557 the uid or gid is negative. A case of a negative gid caused this to be
1558 noticed. The fix allows for either to be negative.
1559
9c4e8f60
PH
1560PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1561 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1562 overlooked.
1563
1564PH/36 The change PH/12 above was broken. Fixed it.
1565
d7174846
PH
1566PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1567 the grounds that routing the same address twice would always produce the
1568 same answer. This might have been true once, but it is certainly no
1569 longer true now. Routing a child address may depend on the previous
1570 routing that produced that child. Some complicated redirection strategies
1571 went wrong when messages had multiple recipients, and made Exim's
1572 behaviour dependent on the order in which the addresses were given.
1573
1574 I have moved the duplicate checking until after the routing is complete.
1575 Exim scans the addresses that are assigned to local and remote
1576 transports, and removes any duplicates. This means that more work will be
1577 done, as duplicates will always all be routed, but duplicates are
1578 presumably rare, so I don't expect this is of any significance.
1579
1580 For deliveries to pipes, files, and autoreplies, the duplicate checking
1581 still happens during the routing process, since they are not going to be
1582 routed further.
1583
cfe75fc3
PH
1584PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1585 It corrects a timeout issue with spamd. This is Ian's comment: "The
1586 background is that sometimes spamd either never reads data from a
1587 connection it has accepted, or it never writes response data. The exiscan
1588 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1589 blindly assumes that writes won't block so it may never time out."
1590
be22d70e
PH
1591PH/39 Allow G after quota size as well as K and M.
1592
0612b098
PH
1593PH/40 The value set for $authenticated_id in an authenticator may not contain
1594 binary zeroes or newlines because the value is written to log lines and
1595 to spool files. There was no check on this. Now the value is run through
1596 the string_printing() function so that such characters are converted to
1597 printable escape sequences.
1598
2e0c1448
PH
1599PH/41 $message_linecount is a new variable that contains the total number of
1600 lines in the message. Compare $body_linecount, which is the count for the
1601 body only.
1602
447d236c
PH
1603PH/42 Exim no longer gives details of delivery errors for specific addresses in
1604 bounce and delay warning messages, except in certain special cases, which
1605 are as follows:
1606
1607 (a) An SMTP error message from a remote host;
1608 (b) A message specified in a :fail: redirection;
1609 (c) A message specified in a "fail" command in a system filter;
1610 (d) A message specified in a FAIL return from the queryprogram router;
1611 (e) A message specified by the cannot_route_message router option.
1612
1613 In these cases only, Exim does include the error details in bounce and
1614 warning messages. There are also a few cases where bland messages such
1615 as "unrouteable address" or "local delivery error" are given.
1616
d20976dc
PH
1617PH/43 $value is now also set for the "else" part of a ${run expansion.
1618
f656d135
PH
1619PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1620 being worked on, but at least Exim now implements the latest version to
1621 play with."
1622
2e2a30b4
PH
1623PH/45 In a pipe transport, although a timeout while waiting for the pipe
1624 process to complete was treated as a delivery failure, a timeout while
1625 writing the message to the pipe was logged, but erroneously treated as a
1626 successful delivery. Such timeouts include transport filter timeouts. For
1627 consistency with the overall process timeout, these timeouts are now
1628 treated as errors, giving rise to delivery failures by default. However,
1629 there is now a new Boolean option for the pipe transport called
1630 timeout_defer, which, if set TRUE, converts the failures into defers for
1631 both kinds of timeout. A transport filter timeout is now identified in
1632 the log output.
1633
9176e9f0
PH
1634PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1635 systems where "make" and "gmake" are different, calling "gmake" at top
1636 level broke things. I've arranged for the value of $(MAKE) to be passed
1637 from the Makefile to this script so that it can call the same version of
1638 "make".
1639
7982096b 1640
bbe902f0
PH
1641A note about Exim versions 4.44 and 4.50
1642----------------------------------------
1643
1644Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1645changes of various kinds. As a consequence, a big documentation update was
1646needed. This delayed the release for rather longer than seemed good, especially
1647in the light of a couple of (minor) security issues. Therefore, the changes
1648that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1649release. So 4.44 and 4.50 are in effect two different branches that both start
1650from 4.43.
1651
1652I have left the 4.50 change log unchanged; it contains all the changes since
16534.43. The change log for 4.44 is below; many of its items are identical to
1654those for 4.50. This seems to be the most sensible way to preserve the
1655historical information.
1656
1657
f7b63901 1658Exim version 4.50
495ae4b0
PH
1659-----------------
1660
5fe762f6
PH
1661 1. Minor wording change to the doc/README.SIEVE file.
1662
139059f6 1663 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 1664 computation of the current number of files was incorrect.
495ae4b0 1665
7086e875
PH
1666 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1667 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1668 place.
1669
35af9f61
PH
1670 4. Give more explanation in the error message when the command for a transport
1671 filter fails to execute.
1672
b668c215
PH
1673 5. There are several places where Exim runs a non-Exim command in a
1674 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1675 was being done only for the command run by the queryprogram router. It is
1676 now done for all such subprocesses. The other cases are: ${run, transport
1677 filters, and the commands run by the lmtp and pipe transports.
1678
a494b1e1
PH
1679 6. Added CONFIGURE_GROUP build-time option.
1680
1681 7. Some older OS have a limit of 256 on the maximum number of file
1682 descriptors. Exim was using setrlimit() to set 1000 as a large value
1683 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1684 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 1685
c5fcb476
PH
1686 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1687 was an oversight, and furthermore, ever since the addition of extra
1688 controls (e.g. 4.43/32), the checks on when to allow different forms of
1689 "control" were broken. There should now be diagnostics for all cases when a
1690 control that does not make sense is encountered.
1691
69358f02
PH
1692 9. Added the /retain_sender option to "control=submission".
1693
5be20824
PH
169410. $recipients is now available in the predata ACL (oversight).
1695
eb2c0248
PH
169611. Tidy the search cache before the fork to do a delivery from a message
1697 received from the command line. Otherwise the child will trigger a lookup
1698 failure and thereby defer the delivery if it tries to use (for example) a
1699 cached ldap connection that the parent has called unbind on.
1700
2a3eea10
PH
170112. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1702 of $address_data from the recipient verification was clobbered by the
1703 sender verification.
1704
170513. The value of address_data from a sender verification is now available in
1706 $sender_address_data in subsequent conditions in the ACL statement.
1707
23c7ff99
PH
170814. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1709
4deaf07d
PH
171015. Added a new option "connect=<time>" to callout options, to set a different
1711 connection timeout.
1712
926e1192
PH
171316. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1714 was its contents. (It was OK if the option was not defined at all.)
1715
650edc6f
PH
171617. A "Completed" log line is now written for messages that are removed from
1717 the spool by the -Mrm option.
1718
2c7db3f5
PH
171918. New variables $sender_verify_failure and $recipient_verify_failure contain
1720 information about exactly what failed.
1721
3d235903
PH
172219. Added -dd to debug only the daemon process.
1723
7c7ad977
PH
172420. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1725 handles timeouts, both on the server side and network timeouts. Renamed the
1726 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1727
981756db
PH
172821. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1729 It is now set to "smtps".
1730
d4eb88df
PH
173122. $host_address is now set to the target address during the checking of
1732 ignore_target_hosts.
1733
173423. When checking ignore_target_hosts for an ipliteral router, no host name was
1735 being passed; this would have caused $sender_host_name to have been used if
1736 matching the list had actually called for a host name (not very likely,
1737 since this list is usually IP addresses). A host name is now passed as
1738 "[x.x.x.x]".
1739
7d468ab8
PH
174024. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1741 code that specifies a non-restarting handler (typically sigaction() in
1742 modern systems) in an attempt to fix a rare and obscure crash bug.
1743
174425. Narrowed the window for a race in the daemon that could cause it to ignore
1745 SIGCHLD signals. This is not a major problem, because they are used only to
1746 wake it up if nothing else does.
1747
62c0818f
PH
174826. A malformed maildirsize file could cause Exim to calculate negative values
1749 for the mailbox size or file count. Odd effects could occur as a result.
1750 The maildirsize information is now recalculated if the size or filecount
1751 end up negative.
1752
26034054
PH
175327. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1754 support for a long time. Removed HAVE_SYS_VFS_H.
1755
af66f652
PH
175628. Installed the latest version of exipick from John Jetmore.
1757
90af77f4
PH
175829. In an address list, if the pattern was not a regular expression, an empty
1759 subject address (from a bounce message) matched only if the pattern was an
1760 empty string. Non-empty patterns were not even tested. This was the wrong
1761 because it is perfectly reasonable to use an empty address as part of a
1762 database query. An empty address is now tested by patterns that are
1763 lookups. However, all the other forms of pattern expect the subject to
1764 contain a local part and a domain, and therefore, for them, an empty
1765 address still always fails if the pattern is not itself empty.
1766
d8ef3577
PH
176730. Exim went into a mad DNS loop when attempting to do a callout where the
1768 host was specified on an smtp transport, and looking it up yielded more
1769 than one IP address.
1770
5cb8cbc6
PH
177131. Re-factored the code for checking spool and log partition space into a
1772 function that finds that data and another that does the check. The former
1773 is then used to implement four new variables: $spool_space, $log_space,
1774 $spool_inodes, and $log_inodes.
1775
14702f5b
PH
177632. The RFC2047 encoding function was originally intended for short strings
1777 such as real names; it was not keeping to the 75-character limit for
1778 encoded words that the RFC imposes. It now respects the limit, and
1779 generates multiple encoded words if necessary. To be on the safe side, I
1780 have increased the buffer size for the ${rfc2047: expansion operator from
1781 1024 to 2048 bytes.
1782
063b1e99
PH
178333. It is now permitted to omit both strings after an "if" condition; if the
1784 condition is true, the result is "true". As before, when the second string
1785 is omitted, a false condition yields an empty string. This makes it less
1786 cumbersome to write custom ACL and router conditions.
1787
652e1b65
PH
178834. Failure to deliver a bounce message always caused it to be frozen, even if
1789 there was an errors_to setting on the router. The errors_to setting is now
1790 respected.
1791
6f0c9a4f
PH
179235. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1793 canonical form (fully expanded) before being placed in
1794 $sender_host_address.
1795
33397d19
PH
179636. The table in the code that translates DNS record types into text (T_A to
1797 "A" for instance) was missing entries for NS and CNAME. It is just possible
1798 that this could have caused confusion if both these types were looked up
1799 for the same domain, because the text type is used as part of Exim's
1800 per-process caching. But the chance of anyone hitting this buglet seems
1801 very small.
1802
7bb56e1f
PH
180337. The dnsdb lookup has been extended in a number of ways.
1804
1805 (1) There is a new type, "zns", which walks up the domain tree until it
1806 finds some nameserver records. It should be used with care.
1807
ea3bc19b
PH
1808 (2) There is a new type, "mxh", which is like "mx" except that it returns
1809 just the host names, not the priorities.
1810
1811 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
1812 looked up. The behaviour when one of the lookups defers can be
1813 controlled by a keyword.
7bb56e1f 1814
ea3bc19b 1815 (4) It is now possible to specify the separator character for use when
7bb56e1f 1816 multiple records are returned.
33397d19 1817
0bcb2a0e
PH
181838. The dnslists ACL condition has been extended: it is now possible to supply
1819 a list of IP addresses and/or domains to be looked up in a particular DNS
1820 domain.
1821
2ac0e484
PH
182239. Added log_selector=+queue_time_overall.
1823
4e1fde53
PH
182440. When running the queue in the test harness, wait just a tad after forking a
1825 delivery process, to get repeatability of debugging output.
1826
de365ded
PH
182741. Include certificate and key file names in error message when GnuTLS fails
1828 to set them up, because the GnuTLS error message doesn't include the name
1829 of the failing file when there is a problem reading it.
1830
f05da2e8
PH
183142. Allow both -bf and -bF in the same test run.
1832
d6453af2
PH
183343. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1834
f7b63901
PH
183544. The "Exiscan patch" is now merged into the mainline Exim source.
1836
183745. Sometimes the final signoff response after QUIT could fail to get
1838 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1839 < 0 before doing a fflush(). This bug looks as though it goes back to the
1840 introduction of TLS in release 3.20, but "sometimes" must have been rare
1841 because the tests only now provoked it.
1842
a444213a
PH
184346. Reset the locale to "C" after calling embedded Perl, in case it was changed
1844 (this can affect the format of dates).
1845
0ec020ea
PH
184647. exim_tidydb, when checking for the continued existence of a message for
1847 which it has found a message-specific retry record, was not finding
1848 messages that were in split spool directories. Consequently, it was
1849 deleting retry records that should have stayed in existence.
1850
b1206957
PH
185148. Steve fixed some bugs in eximstats.
1852
185349. The SPA authentication driver was not abandoning authentication and moving
1854 on to the next authenticator when an expansion was forced to fail,
1855 contradicting the general specification for all authenticators. Instead it
1856 was generating a temporary error. It now behaves as specified.
1857
26dd5a95
PH
185850. The default ordering of permitted cipher suites for GnuTLS was pessimal
1859 (the order specifies the preference for clients). The order is now AES256,
1860 AES128, 3DES, ARCFOUR128.
1861
343b2385
PH
186251. Small patch to Sieve code - explicitly set From: when generating an
1863 autoreply.
1864
1c5466b9
PH
186552. Exim crashed if a remote delivery caused a very long error message to be
1866 recorded - for instance if somebody sent an entire SpamAssassin report back
1867 as a large number of 550 error lines. This bug was coincidentally fixed by
1868 increasing the size of one of Exim's internal buffers (big_buffer) that
1869 happened as part of the Exiscan merge. However, to be on the safe side, I
1870 have made the code more robust (and fixed the comments that describe what
1871 is going on).
1872
55ee9ee3
PH
187353. Now that there can be additional text after "Completed" in log lines (if
1874 the queue_time_overall log selector is set), a one-byte patch to exigrep
1875 was needed to allow it to recognize "Completed" as not the last thing in
1876 the line.
1877
d38f8232
PH
187854. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1879 patch that reportedly fixes this has been added. I am not expert enough to
1880 create a test for it. This is what the patch creator wrote:
1881
1882 "I found a little strange behaviour of ldap code when working with
1883 Windows 2003 AD Domain, where users was placed in more than one
1884 Organization Units. When I tried to give exim partial DN, the exit code
1885 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1886 But simultaneously result of request was absolutely normal ldap result,
1887 so I produce this patch..."
1888
3295e65b
PH
1889 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1890 REFERENCE, so I have modified the code to exclude the patch when that macro
1891 is not defined.
1892
7102e136
PH
189355. Some experimental protocols are using DNS PTR records for new purposes. The
1894 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
1895 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1896 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1897 CAN-2005-0021
7102e136 1898
3ca0ba97
PH
189956. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1900
c2bcbe20
PH
190157. Double the size of the debug message buffer (to 2048) so that more of very
1902 long debug lines gets shown.
1903
18ce445d
PH
190458. The exicyclog utility now does better if the number of log files to keep
1905 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1906
1f5b4c3d
PH
190759. Two changes related to the smtp_active_hostname option:
1908
1909 (1) $smtp_active_hostname is now available as a variable.
1910 (2) The default for smtp_banner uses $smtp_active_hostname instead
1911 of $primary_hostname.
1912
b975ba52
PH
191360. The host_aton() function is supposed to be passed a string that is known
1914 to be a valid IP address. However, in the case of IPv6 addresses, it was
1915 not checking this. This is a hostage to fortune. Exim now panics and dies
1916 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
1917 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1918 components; fortuitously, this particular loophole had already been fixed
1919 by change 4.50/55 above.
1920
1921 If there are any other similar loopholes, the new check in host_aton()
1922 itself should stop them being exploited. The report I received stated that
1923 data on the command line could provoke the exploit when Exim was running as
1924 exim, but did not say which command line option was involved. All I could
1925 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1926 running as the user.
ea3a6f44 1927 CAN-2005-0021
85b87bc2
PH
1928
192961. There was a buffer overflow vulnerability in the SPA authentication code
1930 (which came originally from the Samba project). I have added a test to the
1931 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1932 CAN-2005-0022
b975ba52 1933
17ffcae7
PH
193462. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1935 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1936
d95f9fdb
PH
193763. The daemon start-up calls getloadavg() while still root for those OS that
1938 need the first call to be done as root, but it missed one case: when
1939 deliver_queue_load_max is set with deliver_drop_privilege. This is
1940 necessary for the benefit of the queue runner, because there is no re-exec
1941 when deliver_drop_privilege is set.
1942
86b8287f
PH
194364. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1944 This has been fixed.
1945
60dc5e56
PH
194665. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1947 was in use, was not putting the data itself into the right store pool;
1948 consequently, it could be overwritten for a subsequent message in the same
1949 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1950 the caching.)
1951
533244af
PH
195266. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1953
a5a28604
PH
195467. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1955 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1956 the function were treating the return as a boolean value, which happened to
1957 work because 0=false and not-0=true, but is not correct code.
1958
7e634d24
PH
195968. The host_aton() function was not handling scoped IPv6 addresses (those
1960 with, for example, "%eth0" on the end) correctly.
1961
3e11c26b
PH
196269. Fixed some compiler warnings in acl.c for the bitmaps specified with
1963 negated items (that is, ~something) in unsigned ints. Some compilers
1964 apparently mutter when there is no cast.
1965
6729cf78
PH
196670. If an address verification called from an ACL failed, and did not produce a
1967 user-specific message (i.e. there was only a "system" message), nothing was
1968 put in $acl_verify_message. In this situation, it now puts the system
1969 message there.
1970
00f00ca5
PH
197171. Change 4.23/11 added synchronization checking at the start of an SMTP
1972 session; change 4.31/43 added the unwanted input to the log line - except
1973 that it did not do this in the start of session case. It now does.
1974
c9bdd01c
PH
197572. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1976 This is wrong and can cause the other end to generate a synchronization
1977 error if it is another Exim or anything else that does the synchronization
1978 check. A QUIT command is no longer sent after a timeout.
1979
d43194df
PH
198073. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1981 during host lookups.
1982
fe5b5d0b
PH
198374. The defer_ok option of callout verification was not working if it was used
1984 when verifying addresses in header lines, that is, for this case:
1985
1986 verify = header_sender/callout=defer_ok
1987
76a2d7ba
PH
198875. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1989 those file descriptors could be used for SMTP connections. If anything
1990 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1991 could be sent to the SMTP client, causing chaos. The daemon now opens
1992 stdin, stdout, and stderr to /dev/null when it puts itself into the
1993 background.
1994
199576. Arrange for output from Perl's "warn" command to be written to Exim's main
1996 log by default. The user can override this with suitable Perl magic.
1997
04f7d5b9
PH
199877. The use of log_message on a "discard" ACL verb, which is supposed to add to
1999 the log message when discard triggers, was not working for the DATA ACL or
2000 for the non-SMTP ACL.
2001
bc60667e
PH
200278. Error message wording change in sieve.c.
2003
bb6e88ff
PH
200479. If smtp_accept_max_per_host was set, the number of connections could be
2005 restricted to fewer than expected, because the daemon was trying to set up
2006 a new connection before checking whether the processes handling previous
2007 connections had finished. The check for completed processes is now done
2008 earlier. On busy systems, this bug wouldn't be noticed because something
2009 else would have woken the daemon, and it would have reaped the completed
2010 process earlier.
2011
1e70f85b
PH
201280. If a message was submitted locally by a user whose login name contained one
2013 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2014 It caused a spool format error. I have fixed the spool reading code. A
2015 related problem was that the "from" clause in the Received: line became
2016 illegal because of the space(s). It is now covered by ${quote_local_part.
2017
201881. Included the latest eximstats from Steve (adds average sizes to HTML Top
2019 tables).
2020
4e01f9d6
PH
202182. Updated OS/Makefile-AIX as per message from Mike Meredith.
2022
1ee1cef2
PH
202383. Patch from Sieve maintainer to fix unterminated string problem in
2024 "vacation" handling.
2025
6e2b4ccc
PH
202684. Some minor changes to the Linux configuration files to help with other
2027 OS variants using glibc.
2028
8e669ac1
PH
202985. One more patch for Sieve to update vacation handling to latest spec.
2030
495ae4b0 2031
bbe902f0
PH
2032----------------------------------------------------
2033See the note above about the 4.44 and 4.50 releases.
2034----------------------------------------------------
2035
2036
2037Exim version 4.44
2038-----------------
2039
2040 1. Change 4.43/35 introduced a bug that caused file counts to be
2041 incorrectly computed when quota_filecount was set in an appendfile
2042 transport
2043
2044 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2045 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2046 place.
2047
2048 3. Give more explanation in the error message when the command for a transport
2049 filter fails to execute.
2050
2051 4. There are several places where Exim runs a non-Exim command in a
2052 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2053 was being done only for the command run by the queryprogram router. It is
2054 now done for all such subprocesses. The other cases are: ${run, transport
2055 filters, and the commands run by the lmtp and pipe transports.
2056
2057 5. Some older OS have a limit of 256 on the maximum number of file
2058 descriptors. Exim was using setrlimit() to set 1000 as a large value
2059 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2060 systems. I've change it so that if it can't get 1000, it tries for 256.
2061
2062 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2063 was an oversight, and furthermore, ever since the addition of extra
2064 controls (e.g. 4.43/32), the checks on when to allow different forms of
2065 "control" were broken. There should now be diagnostics for all cases when a
2066 control that does not make sense is encountered.
2067
2068 7. $recipients is now available in the predata ACL (oversight).
2069
2070 8. Tidy the search cache before the fork to do a delivery from a message
2071 received from the command line. Otherwise the child will trigger a lookup
2072 failure and thereby defer the delivery if it tries to use (for example) a
2073 cached ldap connection that the parent has called unbind on.
2074
2075 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2076 of $address_data from the recipient verification was clobbered by the
2077 sender verification.
2078
207910. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2080 was its contents. (It was OK if the option was not defined at all.)
2081
208211. A "Completed" log line is now written for messages that are removed from
2083 the spool by the -Mrm option.
2084
208512. $host_address is now set to the target address during the checking of
2086 ignore_target_hosts.
2087
208813. When checking ignore_target_hosts for an ipliteral router, no host name was
2089 being passed; this would have caused $sender_host_name to have been used if
2090 matching the list had actually called for a host name (not very likely,
2091 since this list is usually IP addresses). A host name is now passed as
2092 "[x.x.x.x]".
2093
209414. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2095 code that specifies a non-restarting handler (typically sigaction() in
2096 modern systems) in an attempt to fix a rare and obscure crash bug.
2097
209815. Narrowed the window for a race in the daemon that could cause it to ignore
2099 SIGCHLD signals. This is not a major problem, because they are used only to
2100 wake it up if nothing else does.
2101
210216. A malformed maildirsize file could cause Exim to calculate negative values
2103 for the mailbox size or file count. Odd effects could occur as a result.
2104 The maildirsize information is now recalculated if the size or filecount
2105 end up negative.
2106
210717. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2108 support for a long time. Removed HAVE_SYS_VFS_H.
2109
ea3a6f44 211018. Updated exipick to current release from John Jetmore.
bbe902f0
PH
2111
211219. Allow an empty sender to be matched against a lookup in an address list.
2113 Previously the only cases considered were a regular expression, or an
2114 empty pattern.
2115
211620. Exim went into a mad DNS lookup loop when doing a callout where the
2117 host was specified on the transport, if the DNS lookup yielded more than
2118 one IP address.
2119
ea3a6f44
NM
212021. The RFC2047 encoding function was originally intended for short strings
2121 such as real names; it was not keeping to the 75-character limit for
2122 encoded words that the RFC imposes. It now respects the limit, and
2123 generates multiple encoded words if necessary. To be on the safe side, I
2124 have increased the buffer size for the ${rfc2047: expansion operator from
2125 1024 to 2048 bytes.
bbe902f0 2126
ea3a6f44
NM
212722. Failure to deliver a bounce message always caused it to be frozen, even if
2128 there was an errors_to setting on the router. The errors_to setting is now
2129 respected.
bbe902f0
PH
2130
213123. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2132 canonical form (fully expanded) before being placed in
2133 $sender_host_address.
2134
213524. Updated eximstats to version 1.33
2136
ea3a6f44
NM
213725. Include certificate and key file names in error message when GnuTLS fails
2138 to set them up, because the GnuTLS error message doesn't include the name
2139 of the failing file when there is a problem reading it.
bbe902f0
PH
2140
214126. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 2142 As per change 25.
bbe902f0 2143
ea3a6f44
NM
214427. Reset the locale to "C" after calling embedded Perl, in case it was changed
2145 (this can affect the format of dates).
bbe902f0 2146
ea3a6f44
NM
214728. exim_tidydb, when checking for the continued existence of a message for
2148 which it has found a message-specific retry record, was not finding
2149 messages that were in split spool directories. Consequently, it was
2150 deleting retry records that should have stayed in existence.
bbe902f0
PH
2151
215229. eximstats updated to version 1.35
2153 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2154 1.35 - bugfix such that pie charts by volume are generated correctly
2155
ea3a6f44
NM
215630. The SPA authentication driver was not abandoning authentication and moving
2157 on to the next authenticator when an expansion was forced to fail,
2158 contradicting the general specification for all authenticators. Instead it
2159 was generating a temporary error. It now behaves as specified.
bbe902f0 2160
ea3a6f44
NM
216131. The default ordering of permitted cipher suites for GnuTLS was pessimal
2162 (the order specifies the preference for clients). The order is now AES256,
2163 AES128, 3DES, ARCFOUR128.
bbe902f0 2164
ea3a6f44
NM
216531. Small patch to Sieve code - explicitly set From: when generating an
2166 autoreply.
bbe902f0 2167
ea3a6f44
NM
216832. Exim crashed if a remote delivery caused a very long error message to be
2169 recorded - for instance if somebody sent an entire SpamAssassin report back
2170 as a large number of 550 error lines. This bug was coincidentally fixed by
2171 increasing the size of one of Exim's internal buffers (big_buffer) that
2172 happened as part of the Exiscan merge. However, to be on the safe side, I
2173 have made the code more robust (and fixed the comments that describe what
2174 is going on).
bbe902f0 2175
ea3a6f44
NM
217633. Some experimental protocols are using DNS PTR records for new purposes. The
2177 keys for these records are domain names, not reversed IP addresses. The
2178 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2179 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
2180 CAN-2005-0021
2181
ea3a6f44
NM
218234. The host_aton() function is supposed to be passed a string that is known
2183 to be a valid IP address. However, in the case of IPv6 addresses, it was
2184 not checking this. This is a hostage to fortune. Exim now panics and dies
2185 if the condition is not met. A case was found where this could be provoked
2186 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2187 components; fortuitously, this particular loophole had already been fixed
2188 by change 4.50/55 or 4.44/33 above.
2189
2190 If there are any other similar loopholes, the new check in host_aton()
2191 itself should stop them being exploited. The report I received stated that
2192 data on the command line could provoke the exploit when Exim was running as
2193 exim, but did not say which command line option was involved. All I could
2194 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2195 running as the user.
bbe902f0
PH
2196 CAN-2005-0021
2197
ea3a6f44
NM
219835. There was a buffer overflow vulnerability in the SPA authentication code
2199 (which came originally from the Samba project). I have added a test to the
2200 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
2201 CAN-2005-0022
2202
ea3a6f44
NM
220336. The daemon start-up calls getloadavg() while still root for those OS that
2204 need the first call to be done as root, but it missed one case: when
2205 deliver_queue_load_max is set with deliver_drop_privilege. This is
2206 necessary for the benefit of the queue runner, because there is no re-exec
2207 when deliver_drop_privilege is set.
bbe902f0 2208
ea3a6f44
NM
220937. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2210 was in use, was not putting the data itself into the right store pool;
2211 consequently, it could be overwritten for a subsequent message in the same
2212 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2213 the caching.)
bbe902f0 2214
ea3a6f44
NM
221538. Sometimes the final signoff response after QUIT could fail to get
2216 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2217 < 0 before doing a fflush(). This bug looks as though it goes back to the
2218 introduction of TLS in release 3.20, but "sometimes" must have been rare
2219 because the tests only now provoked it.
bbe902f0
PH
2220
2221
495ae4b0
PH
2222Exim version 4.43
2223-----------------
2224
2225 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2226 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2227 Now it returns an integer. A number of calls were still expecting a T/F
2228 return. Fortuitously, in all cases, the tests worked in OK situations,
2229 which is the norm. However, things would have gone wrong on any write
2230 failures on the smtp file descriptor. This function is used when sending
2231 messages over SMTP and also when doing verify callouts.
2232
2233 2. When Exim is called to do synchronous delivery of a locally submitted
2234 message (the -odf or -odi options), it no longer closes stderr before doing
2235 the delivery.
2236
2237 3. Implemented the mua_wrapper option.
2238
2239 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2240
2241 5. Implemented the functions header_remove(), header_testname(),
2242 header_add_at_position(), and receive_remove_recipient(), and exported them
2243 to local_scan().
2244
2245 6. If an ACL "warn" statement specified the addition of headers, Exim already
2246 inserted X-ACL-Warn: at the start if there was no header name. However, it
2247 was not making this test for the second and subsequent header lines if
2248 there were newlines in the string. This meant that an invalid header could
2249 be inserted if Exim was badly configured.
2250
2251 7. Allow an ACL "warn" statement to add header lines at the start or after all
2252 the Received: headers, as well as at the end.
2253
2254 8. Added the rcpt_4xx retry error code.
2255
2256 9. Added postmaster_mailfrom=xxx to callout verification option.
2257
225810. Added mailfrom=xxxx to the callout verification option, for verify=
2259 header_sender only.
2260
226111. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2262 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2263
226412. Inserted some casts to stop certain compilers warning when using pointer
2265 differences as field lengths or precisions in printf-type calls (mostly
2266 affecting debugging statements).
2267
226813. Added optional readline() support for -be (dynamically loaded).
2269
227014. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2271 same clock tick as a message's arrival, so that its received time was the
2272 same as the "first fail" time on the retry record, and that message
2273 remained on the queue past the ultimate address timeout, every queue runner
2274 would try a delivery (because it was past the ultimate address timeout) but
2275 after another failure, the ultimate address timeout, which should have then
2276 bounced the address, did not kick in. This was a "< instead of <=" error;
2277 in most cases the first failure would have been in the next clock tick
2278 after the received time, and all would be well.
2279
228015. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2281 being recognized when the domain list was tested by the match_domain
2282 condition in an expansion string.
2283
228416. Added the ${str2b64: operator.
2285
228617. Exim was always calling setrlimit() to set a large limit for the number of
2287 processes, without checking whether the existing limit was already
2288 adequate. (It did check for the limit on file descriptors.) Furthermore,
2289 errors from getrlimit() and setrlimit() were being ignored. Now they are
2290 logged to the main and panic logs, but Exim does carry on, to try to do its
2291 job under whatever limits there are.
2292
229318. Imported PCRE 5.0.
2294
229519. Trivial typo in log message " temporarily refused connection" (the leading
2296 space).
2297
229820. If the log selector return_path_on_delivery was set and an address was
2299 redirected to /dev/null, the delivery process crashed because it assumed
2300 that a return path would always be set for a "successful" delivery. In this
2301 case, the whole delivery is bypassed as an optimization, and therefore no
2302 return path is set.
2303
230421. Internal re-arrangement: the function for sending a challenge and reading
2305 a response while authentication was assuming a zero-terminated challenge
2306 string. It's now changed to take a pointer and a length, to allow for
2307 binary data in such strings.
2308
230922. Added the cyrus_sasl authenticator (code supplied by MBM).
2310
231123. Exim was not respecting finduser_retries when seeking the login of the
2312 uid under which it was called; it was always trying 10 times. (The default
2313 setting of finduser_retries is zero.) Also, it was sleeping after the final
2314 failure, which is pointless.
2315
231624. Implemented tls_on_connect_ports.
2317
231825. Implemented acl_smtp_predata.
2319
232026. If the domain in control=submission is set empty, Exim assumes that the
2321 authenticated id is a complete email address when it generates From: or
2322 Sender: header lines.
2323
232427. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2325 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2326 chown and chgrp in /bin and hostname in /usr/bin.
2327
232828. Exim was keeping the "process log" file open after each use, just as it
2329 does for the main log. This opens the possibility of it remaining open for
2330 long periods when the USR1 signal hits a daemon. Occasional processlog
2331 errors were reported, that could have been caused by this. Anyway, it seems
2332 much more sensible not to leave this file open at all, so that is what now
2333 happens.
2334
233529. The long-running daemon process does not normally write to the log once it
2336 has entered its main loop, and it closes the log before doing so. This is
2337 so that log files can straightforwardly be renamed and moved. However,
2338 there are a couple of unusual error situations where the daemon does write
2339 log entries, and I had neglected to close the log afterwards.
2340
234130. The text of an SMTP error response that was received during a remote
2342 delivery was being truncated at 512 bytes. This is too short for some of
2343 the long messages that one sometimes sees. I've increased the limit to
2344 1024.
2345
234631. It is now possible to make retry rules that apply only when a message has a
2347 specific sender, in particular, an empty sender.
2348
234932. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2350 it possible to be selective about when SMTP synchronization is enforced.
2351
235233. Added "control = caseful_local_part" and "control = "caselower_local_part".
2353
235432. Implemented hosts_connection_nolog.
2355
235633. Added an ACL for QUIT.
2357
235834. Setting "delay_warning=" to disable warnings was not working; it gave a
2359 syntax error.
2360
236135. Added mailbox_size and mailbox_filecount to appendfile.
2362
236336. Added control = no_multiline_responses to ACLs.
2364
236537. There was a bug in the logic of the code that waits for the clock to tick
2366 in the case where the clock went backwards by a substantial amount such
2367 that the microsecond fraction of "now" was more than the microsecond
2368 fraction of "then" (but the whole seconds number was less).
2369
237038. Added support for the libradius Radius client library this is found on
2371 FreeBSD (previously only the radiusclient library was supported).
2372
2373
2374Exim version 4.42
2375-----------------
2376
2377 1. When certain lookups returned multiple values in the form name=value, the
2378 quoting of the values was not always being done properly. Specifically:
2379 (a) If the value started with a double quote, but contained no whitespace,
2380 it was not quoted.
2381 (b) If the value contained whitespace other than a space character (i.e.
2382 tabs or newlines or carriage returns) it was not quoted.
2383 This fix has been applied to the mysql and pgsql lookups by writing a
2384 separate quoting function and calling it from the lookup code. The fix
2385 should probably also be applied to nisplus, ibase and oracle lookups, but
2386 since I cannot test any of those, I have not disturbed their existing code.
2387
2388 2. A hit in the callout cache for a specific address caused a log line with no
2389 reason for rejecting RCPT. Now it says "Previous (cached) callout
2390 verification failure".
2391
2392 3. There was an off-by-one bug in the queryprogram router. An over-long
2393 return line was truncated at 256 instead of 255 characters, thereby
2394 overflowing its buffer with the terminating zero. As well as fixing this, I
2395 have increased the buffer size to 1024 (and made a note to document this).
2396
2397 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2398 when Exim is waiting for an SMTP response from a remote server, Exim
2399 restarts its select() call on the socket, thereby resetting its timeout.
2400 This is not a problem when such interrupts are rare. Somebody set up a cron
2401 job to run exiwhat every 2 minutes, which is less than the normal select()
2402 timeout (5 or 10 minutes). This meant that the select() timeout never
2403 kicked in because it was always reset. I have fixed this by comparing the
2404 time when an interrupt arrives with the time at the start of the first call
2405 to select(). If more time than the timeout has elapsed, the interrupt is
2406 treated as a timeout.
2407
2408 5. Some internal re-factoring in preparation for the addition of Sieve
2409 extensions (by MH). In particular, the "personal" test is moved to a
2410 separate function, and given an option for scanning Cc: and Bcc: (which is
2411 not set for Exim filters).
2412
2413 6. When Exim created an email address using the login of the caller as the
2414 local part (e.g. when creating a From: or Sender: header line), it was not
2415 quoting the local part when it contained special characters such as @.
2416
2417 7. Installed new OpenBSD configuration files.
2418
2419 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2420 try to make them clearer.
2421
2422 9. Callout options, other than the timeout value, were being ignored when
2423 verifying sender addresses in header lines. For example, when using
2424
2425 verify = header_sender/callout=no_cache
2426
2427 the cache was (incorrectly) being used.
2428
242910. Added a missing instance of ${EXE} to the exim_install script; this affects
2430 only the Cygwin environment.
2431
243211. When return_path_on_delivery was set as a log selector, if different remote
2433 addresses in the same message used different return paths and parallel
2434 remote delivery occurred, the wrong values would sometimes be logged.
2435 (Whenever a remote delivery process finished, the return path value from
2436 the most recently started remote delivery process was logged.)
2437
243812. RFC 3848 specifies standard names for the "with" phrase in Received: header
2439 lines when AUTH and/or TLS are in use. This is the "received protocol"
2440 field. Exim used to use "asmtp" for authenticated SMTP, without any
2441 indication (in the protocol name) for TLS use. Now it follows the RFC and
2442 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2443 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2444 names appear in log lines as well as in Received: header lines.
2445
244613. Installed MH's patches for Sieve to add the "copy" and "vacation"
2447 extensions, and comparison tests, and to fix some bugs.
2448
244914. Changes to the "personal" filter test:
2450
2451 (1) The test was buggy in that it was just doing the equivalent of
2452 "contains" tests on header lines. For example, if a user's address was
2453 anne@some.where, the "personal" test would incorrectly be true for
2454
2455 To: susanne@some.where
2456
2457 This test is now done by extracting each address from the header in turn,
2458 and checking the entire address. Other tests that are part of "personal"
2459 are now done using regular expressions (for example, to check local parts
2460 of addresses in From: header lines).
2461
2462 (2) The list of non-personal local parts in From: addresses has been
2463 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2464 taken from the Sieve specification recommendations.
2465
2466 (3) If the message contains any header line starting with "List-" it is
2467 treated as non-personal.
2468
2469 (4) The test for "circular" in the Subject: header line has been removed
2470 because it now seems ill-conceived.
2471
247215. Minor typos in src/EDITME comments corrected.
2473
247416. Installed latest exipick from John Jetmore.
2475
247617. If headers_add on a router specified a text string that was too long for
2477 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2478 of string_sprintf() is now avoided.
2479
248018. $message_body_size was not set (it was always zero) when running the DATA
2481 ACL and the local_scan() function.
2482
248319. For the "mail" command in an Exim filter, no default was being set for
2484 the once_repeat time, causing a random time value to be used if "once" was
2485 specified. (If the value happened to be <= 0, no repeat happened.) The
2486 default is now 0s, meaning "never repeat". The "vacation" command was OK
2487 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2488 (I found it when inspecting the code).
2489
249020. There is now an overall timeout for performing a callout verification. It
2491 defaults to 4 times the callout timeout, which applies to individual SMTP
2492 commands during the callout. The overall timeout applies when there is more
2493 than one host that can be tried. The timeout is checked before trying the
2494 next host. This prevents very long delays if there are a large number of
2495 hosts and all are timing out (e.g. when the network connections are timing
2496 out). The value of the overall timeout can be changed by specifying an
2497 additional sub-option for "callout", called "maxwait". For example:
2498
2499 verify = sender/callout=5s,maxwait=20s
2500
250121. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2502 to the end before writing, but this should make it even safer).
2503
250422. Exim was forgetting that it had advertised PIPELINING for the second and
2505 subsequent messages on an SMTP connection. It was also not resetting its
2506 memory on STARTTLS and an internal HELO.
2507
250823. When Exim logs an SMTP synchronization error within a session, it now
2509 records whether PIPELINING has been advertised or not.
2510
251124. Added 3 instances of "(long int)" casts to time_t variables that were being
2512 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2513 rather than long int.
2514
251525. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2516
251726. Added the never_mail option to autoreply.
2518
2519
2520Exim version 4.41
2521-----------------
2522
2523 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2524 crash if the getsockname() call failed; this can happen if a connection is
2525 closed very soon after it is established. The problem was simply in the
2526 order in which certain operations were done, causing Exim to try to write
2527 to the SMTP stream before it had set up the file descriptor. The bug has
2528 been fixed by making things happen in the correct order.
2529
2530
2531Exim version 4.40
2532-----------------
2533
2534 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2535 before the connection was closed, thus losing the rejection response.
2536
2537 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2538 some early Solaris releases, but causes trouble in current releases where
2539 socklen_t is defined.
2540
2541 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2542 always exist.
2543
2544 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2545 configured.
2546
2547 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2548 that releasing the top of it at the end releases what was used for sub-
2549 expansions (unless the block got too big). However, discard this block if
2550 the first thing is a variable or header, so that we can use its block when
2551 it is dynamic (useful for very large $message_headers, for example).
2552
2553 6. Lookups now cache *every* query, not just the most recent. A new, separate
2554 store pool is used for this. It can be recovered when all lookup caches are
2555 flushed. Lookups now release memory at the end of their result strings.
2556 This has involved some general refactoring of the lookup sources.
2557
2558 7. Some code has been added to the store_xxx() functions to reduce the amount
2559 of flapping under certain conditions.
2560
2561 8. log_incoming_interface used to affect only the <= reception log lines. Now
2562 it causes the local interface and port to be added to several more SMTP log
2563 lines, for example "SMTP connection from", and rejection lines.
2564
2565 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2566
256710. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2568
256911. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2570 could be overwritten at the end of the current message (or the start of a
2571 new message if it was set in a HELO ACL). The value is now preserved for
2572 the duration of the SMTP connection.
2573
257412. If a transport had a headers_rewrite setting, and a matching header line
2575 contained an unqualified address, that address was qualified, even if it
2576 did not match any rewriting rules. The underlying bug was that the values
2577 of the flags that permit the existence of unqualified sender and recipient
2578 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2579 non-local messages, and by -bnq for local messages) were not being
2580 preserved with the message after it was received.
2581
258213. When Exim was logging an SMTP synchronization error, it could sometimes log
2583 "next input=" as part of the text comprising the host identity instead of
2584 the correct text. The code was using the same buffer for two different
2585 strings. However, depending on which order the printing function evaluated
2586 its arguments, the bug did not always show up. Under Linux, for example, my
2587 test suite worked just fine.
2588
258914. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2590 doesn't work with some older versions of Perl. It has been changed to "my",
2591 which in any case is probably the better facility to use.
2592
259315. A really picky compiler found some instances of statements for creating
2594 error messages that either had too many or two few arguments for the format
2595 string.
2596
259716. The size of the buffer for calls to the DNS resolver has been increased
2598 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2599 for addresses that have a lot of PTR records. This alleviates a problem; it
2600 does not fully solve it.
2601
260217. A dnsdb lookup for PTR records that receives more data than will fit in the
2603 buffer now truncates the list and logs the incident, which is the same
2604 action as happens when Exim is looking up a host name and its aliases.
2605 Previously in this situation something unpredictable would happen;
2606 sometimes it was "internal error: store_reset failed".
2607
260818. If a server dropped the connection unexpectedly when an Exim client was
2609 using GnuTLS and trying to read a response, the client delivery process
2610 crashed while trying to generate an error log message.
2611
261219. If a "warn" verb in an ACL added multiple headers to a message in a single
2613 string, for example:
2614
2615 warn message = H1: something\nH2: something
2616
2617 the text was added as a single header line from Exim's point of view
2618 though it ended up OK in the delivered message. However, searching for the
2619 second and subsequent header lines using $h_h2: did not work. This has been
2620 fixed. Similarly, if a system filter added multiple headers in this way,
2621 the routers could not see them.
2622
262320. Expanded the error message when iplsearch is called with an invalid key to
2624 suggest using net-iplsearch in a host list.
2625
262621. When running tests using -bh, any delays imposed by "delay" modifiers in
2627 ACLs are no longer actually imposed (and a message to that effect is
2628 output).
2629
263022. If a "gecos" field in a passwd entry contained escaped characters, in
2631 particular, if it contained a \" sequence, Exim got it wrong when building
2632 a From: or a Sender: header from that name. A second bug also caused
2633 incorrect handling when an unquoted " was present following a character
2634 that needed quoting.
2635
263623. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2637 was not being matched caselessly.
2638
263924. Arranged for all hyphens in the exim.8 source to be escaped with
2640 backslashes.
2641
264225. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2643 properly. Recipient callout cache records were still being keyed to include
2644 the sender, even when use_sender was set false. This led to far more
2645 callouts that were necessary. The sender is no longer included in the key
2646 when use_sender is false.
2647
264826. Added "control = submission" modifier to ACLs.
2649
265027. Added the ${base62d: operator to decode base 62 numbers.
2651
265228. dnsdb lookups can now access SRV records.
2653
265429. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2655 the configuration file.
2656
265730. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2658 (-v) mode. This makes the output for a verbose queue run more intelligible.
2659
266031. Added a use_postmaster feature to recipient callouts.
2661
266232. Added the $body_zerocount variable, containing the number of binary zero
2663 bytes in the message body.
2664
266533. The time of last modification of the "new" subdirectory is now used as the
2666 "mailbox time last read" when there is a quota error for a maildir
2667 delivery.
2668
266934. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2670
267135. Added +ignore_unknown as a special item in host lists.
2672
267336. Code for decoding IPv6 addresses in host lists is now included, even if
2674 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2675 address was recognized as an IP address, but was then not correctly decoded
2676 into binary, causing unexpected and incorrect effects when compared with
2677 another IP address.
2678
2679
2680Exim version 4.34
2681-----------------
2682
2683 1. Very minor rewording of debugging text in manualroute to say "list of
2684 hosts" instead of "hostlist".
2685
2686 2. If verify=header_syntax was set, and a header line with an unqualified
2687 address (no domain) and a large number of spaces between the end of the
2688 name and the colon was received, the reception process suffered a buffer
2689 overflow, and (when I tested it) crashed. This was caused by some obsolete
2690 code that should have been removed. The fix is to remove it!
2691
2692 3. When running in the test harness, delay a bit after writing a bounce
2693 message to get a bit more predictability in the log output.
2694
2695 4. Added a call to search_tidyup() just before forking a reception process. In
2696 theory, someone could use a lookup in the expansion of smtp_accept_max_
2697 per_host which, without the tidyup, could leave open a database connection.
2698
2699 5. Added the variables $recipient_data and $sender_data which get set from a
2700 lookup success in an ACL "recipients" or "senders" condition, or a router
2701 "senders" option, similar to $domain_data and $local_part_data.
2702
2703 6. Moved the writing of debug_print from before to after the "senders" test
2704 for routers.
2705
2706 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2707 problems for message scanning, either using a data ACL, or using
2708 local_scan() because the Received: header was not generated till after they
2709 were called (in order to set the time as the time of reception completion).
2710 I have revised the way this works. The header is now generated after the
2711 body is received, but before the ACL or local_scan() are called. After they
2712 are run, the timestamp in the header is updated.
2713
2714
2715Exim version 4.33
2716-----------------
2717
2718 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2719 before starting a queue runner without re-exec. This happened only when
2720 deliver_drop_privilege was set or when the Exim user was set to root. The
2721 effect of the bug was that timeouts during subsequent deliveries caused
2722 crashes instead of being properly handled. The handler is now left at its
2723 default (and expected) setting.
2724
2725 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2726 message, again when deliver_drop_privilege is set or Exim is run as root.
2727 The bug described in (1) was not present in this case, but the tidying up
2728 of the other signals was missing. I have made the two cases consistent.
2729
2730 3. The ignore_target_hosts setting on a manualroute router was being ignored
2731 for hosts that were looked up using the /MX notation.
2732
2733 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2734 in domain lists.
2735
2736 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2737 operated on the sender address. After changing the $sender_address to <>
2738 for the sender address verify, Exim was re-instated it as the original
2739 (before rewriting) address, but remembering that it had rewritten it, so it
2740 wasn't rewriting it again. This bug also had the effect of breaking the
2741 sender address verification caching when the sender address was rewritten.
2742
2743 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2744 This has been changed so that if the ip literal address matches
2745 ignore_target_hosts, the router declines.
2746
2747 7. Added expansion conditions match_domain, match_address, and match_local_
2748 part (NOT match_host).
2749
2750 8. The placeholder for the Received: header didn't have a length field set.
2751
2752 9. Added code to Exim itself and to exim_lock to test for a specific race
2753 condition that could lead to file corruption when using MBX delivery. The
2754 issue is with the lockfile that is created in /tmp. If this file is removed
2755 after a process has opened it but before that process has acquired a lock,
2756 there is the potential for a second process to recreate the file and also
2757 acquire a lock. This could lead to two Exim processes writing to the file
2758 at the same time. The added code performs the same test as UW imapd; it
2759 checks after acquiring the lock that its file descriptor still refers to
2760 the same named file.
2761
276210. The buffer for building added header lines was of fixed size, 8192 bytes.
2763 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2764 when Exim is built.
2765
276611. Added the smtp_active_hostname option. If used, this will typically be made
2767 to depend on the incoming interface address. Because $interface_address is
2768 not set up until the daemon has forked a reception process, error responses
2769 that can happen earlier (such as "too many connections") no longer contain
2770 a host name.
2771
277212. If an expansion in a condition on a "warn" statement fails because a lookup
2773 defers, the "warn" statement is abandoned, and the next ACL statement is
2774 processed. Previously this caused the whole ACL to be aborted.
2775
277613. Added the iplsearch lookup type.
2777
277814. Added ident_timeout as a log selector.
2779
278015. Added tls_certificate_verified as a log selector.
2781
278216. Added a global option tls_require_ciphers (compare the smtp transport
2783 option of the same name). This controls incoming TLS connections.
2784
278517. I finally figured out how to make tls_require_ciphers do a similar thing
2786 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2787 before starting the TLS session.
2788
278918. Tabs are now shown as \t in -bP output.
2790
279119. If the log selector return_path_on_delivery was set, Exim crashed when
2792 bouncing a message because it had too many Received: header lines.
2793
279420. If two routers both had headers_remove settings, and the first one included
2795 a superfluous trailing colon, the final name in the first list and the
2796 first name in the second list were incorrectly joined into one item (with a
2797 colon in the middle).
2798
2799
2800Exim version 4.32
2801-----------------
2802
2803 1. Added -C and -D options to the exinext utility, mainly to make it easier
2804 to include in the automated testing, but these could be helpful when
2805 multiple configurations are in use.
2806
2807 2. The exinext utility was not formatting the output nicely when there was
2808 an alternate port involved in the retry record key, nor when there was a
2809 message id as well (for retries that were specific to a specific message
2810 and a specific host). It was also confused by IPv6 addresses, because of
2811 the additional colons they contain. I have fixed the IPv4 problem, and
2812 patched it up to do a reasonable job for IPv6.
2813
2814 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2815 delivery, the log line now contains "pipelined" if PIPELINING was used.
2816
2817 4. An SMTP transport process used to panic and die if the bind() call to set
2818 an explicit outgoing interface failed. This has been changed; it is now
2819 treated in the same way as a connect() failure.
2820
2821 5. A reference to $sender_host_name in the part of a conditional expansion
2822 that was being skipped was still causing a DNS lookup. This no longer
2823 occurs.
2824
2825 6. The def: expansion condition was not recognizing references to header lines
2826 that used bh_ and bheader_.
2827
2828 7. Added the _cache feature to named lists.
2829
2830 8. The code for checking quota_filecount in the appendfile transport was
2831 allowing one more file than it should have been.
2832
2833 9. For compatibility with Sendmail, the command line option
2834
2835 -prval:sval
2836
2837 is equivalent to
2838
2839 -oMr rval -oMs sval
2840
2841 and sets the incoming protocol and host name (for trusted callers). The
2842 host name and its colon can be omitted when only the protocol is to be set.
2843 Note the Exim already has two private options, -pd and -ps, that refer to
2844 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2845 "s", but I don't think that's a major issue.
2846
284710. A number of refactoring changes to the code, none of which should affect
2848 Exim's behaviour:
2849
2850 (a) The number of logging options was getting close to filling up the
2851 32-bit word that was used as a bit map. I have split them into two classes:
2852 those that are passed in the argument to log_write(), and those that are
2853 only ever tested independently outside of that function. These are now in
2854 separate 32-bit words, so there is plenty of room for expansion again.
2855 There is no change in the user interface or the logging behaviour.
2856
2857 (b) When building, for example, log lines, the code previously used a
2858 macro that called string_cat() twice, in order to add two strings. This is
2859 not really sufficiently general. Furthermore, there was one instance where
2860 it was actually wrong because one of the argument was used twice, and in
2861 one call a function was used. (As it happened, calling the function twice
2862 did not affect the overall behaviour.) The macro has been replaced by a
2863 function that can join an arbitrary number of extra strings onto a growing
2864 string.
2865
2866 (c) The code for expansion conditions now uses a table and a binary chop
2867 instead of a serial search (which was left over from when there were very
2868 few conditions). Also, it now recognizes conditions like "pam" even when
2869 the relevant support is not compiled in: a suitably worded error message is
2870 given if an attempt is made to use such a condition.
2871
287211. Added ${time_interval:xxxxx}.
2873
287412. A bug was causing one of the ddress fields not to be passed back correctly
2875 from remote delivery subprocesses. The field in question was not being
2876 subsequently used, so this caused to problems in practice.
2877
287813. Added new log selectors queue_time and deliver_time.
2879
288014. Might have fixed a bug in maildirsizefile handling that threw up
2881 "unexpected character" debug warnings, and recalculated the data
2882 unnecessarily. In any case, I expanded the warning message to give more
2883 information.
2884
288515. Added the message "Restricted characters in address" to the statements in
2886 the default ACL that block characters like @ and % in local parts.
2887
288816. Change 71 for release 4.31 proved to be much less benign that I imagined.
2889 Three changes have been made:
2890
2891 (a) There was a serious bug; a negative response to MAIL caused the whole
2892 recipient domain to be cached as invalid, thereby blocking all messages
2893 to all local parts at the same domain, from all senders. This bug has
2894 been fixed. The domain is no longer cached after a negative response to
2895 MAIL if the sender used is not empty.
2896
2897 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2898 been restored.
2899
2900 (c) A new callout option, "use_sender" has been added for people who want
2901 the modified behaviour.
2902
2903
2904Exim version 4.31
2905-----------------
2906
2907 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2908 Larry Rosenman.
2909
2910 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2911 indeed breaks things for older releases.
2912
2913 3. Added additional logging to the case where there is a problem reading data
2914 from a filter that is running in a subprocess using a pipe, in order to
2915 try to track down a specific problem.
2916
2917 4. Testing facility fudge: when running in the test harness and attempting
2918 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2919 getting "No route to host". Convert this to a timeout.
2920
2921 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2922 warning.
2923
2924 6. Some OS don't have socklen_t but use size_t instead. This affects the
2925 fifth argument of getsockopt() amongst other things. This is now
2926 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2927 can be set for individual OS. I have set it for SunOS5, OSF1, and
2928 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2929 some earlier ones do not.
2930
2931 7. Change 4.30/15 was not doing the test caselessly.
2932
2933 8. The standard form for an IPv6 address literal was being rejected by address
2934 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2935 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2936 this, as well as the form without the "IPv6" on the front (but only when
2937 address literals are enabled, of course).
2938
2939 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2940
294110. Exim crashed if a message with an empty sender address specified by -f
2942 encountered a router with an errors_to setting. This could be provoked only
2943 by a command such as
2944
2945 exim -f "" ...
2946
2947 where an empty string was supplied; "<>" did not hit this bug.
2948
294911. Installed PCRE release 4.5.
2950
295112. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2952 remained set. It is now erased.
2953
295413. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2955 times from message ids (which are base 36 rather than the normal 62).
2956
295714. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2958 were being counted as actual protocol errors, and logged if the log
2959 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2960 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2961 and DATA following a set of rejected RCPTs do not count as protocol errors.
2962 In other words, Exim assumes they were pipelined, though this may not
2963 actually be the case. Of course, in all cases the client gets an
2964 appropriate error code.
2965
296615. If a lookup fails in an ACL condition, a message about the failure may
2967 be available; it is used if testing the ACL cannot continue, because most
2968 such messages specify what the cause of the deferral is. However, some
2969 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2970 that caused an old message to be retained and used if a later statement
2971 caused a defer, replacing the real cause of the deferral.
2972
297316. If an IP address had so many PTR records that the DNS lookup buffer
2974 was not large enough to hold them, Exim could crash while trying to process
2975 the truncated data. It now detects and logs this case.
2976
297717. Further to 4.21/58, another change has been made: if (and only if) the
2978 first line of a message (the first header line) ends with CRLF, a bare LF
2979 in a subsequent header line has a space inserted after it, so as not to
2980 terminate the header.
2981
298218. Refactoring: tidied an ugly bit of code in appendfile that copied data
2983 unnecessarily, used atoi() instead of strtol(), and didn't check the
2984 termination when getting file sizes from file names by regex.
2985
298619. Completely re-implemented the support for maildirsize files, in the light
2987 of a number of problems with the previous contributed implementation
2988 (4.30/29). In particular:
2989
2990 . If the quota is zero, the maildirsize file is maintained, but no quota is
2991 imposed.
2992
2993 . If the maildir directory does not exist, it is created before any attempt
2994 to write a maildirsize file.
2995
2996 . The quota value in the file is just a cache; if the quota is changed in
2997 the transport, the new value overrides.
2998
2999 . A regular expression is available for excluding directories from the
3000 count.
3001
300220. The autoreply transport checks the characters in options that define the
3003 message's headers; it allows continued headers, but it was checking with
3004 isspace() after an embedded newline instead of explicitly looking for a
3005 space or a tab.
3006
300721. If all the "regular" hosts to which an address was routed had passed their
3008 expiry times, and had not reached their retry times, the address was
3009 bounced, even if fallback hosts were defined. Now Exim should go on to try
3010 the fallback hosts.
3011
301222. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3013 equivalent code in the SMTP transport. Some hosts send humungous responses
3014 to HELO/EHLO, more than 1024 it seems.
3015
301623. Refactoring: code in filter.c used (void *) for "any old type" but this
3017 gives compiler warnings in some environments. I've now done it "properly",
3018 using a union.
3019
302024. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3021 (because of problems with the built-in one) was declared to return uschar *
3022 instead of char *, causing compiler failure.
3023
302425. Fixed a file descriptor leak when processing alias/forward files.
3025
302626. Fixed a minor format string issue in dbfn.c.
3027
302827. Typo in exim.c: ("dmbnz" for "dbmnz").
3029
303028. If a filter file refered to $h_xxx or $message_headers, and the headers
3031 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3032 become corrupted.
3033
303429. When a sender address is verified, it is cached, to save repeating the test
3035 when there is more than one recipient in a message. However, when the
3036 verification involves a callout, it is possible for different callout
3037 options to be set for different recipients. It is too complicated to keep
3038 track of this in the cache, so now Exim always runs a verification when a
3039 callout is required, relying on the callout cache for the optimization.
3040 The overhead is duplication of the address routing, but this should not be
3041 too great.
3042
304330. Fixed a bug in callout caching. If a RCPT command caused the sender address
3044 to be verified with callout=postmaster, and the main callout worked but the
3045 postmaster check failed, the verification correctly failed. However, if a
3046 subsequent RCPT command asked for sender verification *without* the
3047 postmaster check, incorrect caching caused this verification also to fail,
3048 incorrectly.
3049
305031. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3051 it was not caching the DNS options (qualify_single, search_parents) that
3052 were used when the lookup failed. A subsequent lookup with different
3053 options therefore always gave the same answer, though there were cases
3054 where it should not have. (Example: a "domains = !$mx_any" option on a
3055 dnslookup router: the "domains" option is always processed without any
3056 widening, but the router might have qualify_single set.) Now Exim uses the
3057 cached value only when the same options are set.
3058
305932. Added John Jetmore's "exipick" utility to the distribution.
3060
306133. GnuTLS: When an attempt to start a TLS session fails for any reason other
3062 than a timeout (e.g. a certificate is required, and is not provided), an
3063 Exim server now closes the connection immediately. Previously it waited for
3064 the client to close - but if the client is SSL, it seems that they each
3065 wait for each other, leading to a delay before one of them times out.
3066
306734: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3068 maintained 0.8.x compatibility because I don't think many are using it, and
3069 it is clearly obsolete.
3070
307135. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3072 transport.
3073
307436. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3075 client certificate was expired. A simple patch fixes this, though I don't
3076 understand the full logic of why the verify callback is called multiple
3077 times.
3078
307937. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3080 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3081 which causes problems with some clients (such as the Certicom SSL Plus
3082 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3083 disables the coutermeasure allowing Eudora to connect."
3084
308538. Exim was not checking that a write() to a log file succeeded. This could
3086 lead to Bad Things if a log got too big, in particular if it hit a file
3087 size limit. Exim now panics and dies if it cannot write to a log file, just
3088 as it does if it cannot open a log file.
3089
309039. Modified OS/Makefile-Linux so that it now contains
3091
3092 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3093
3094 The two -D definitions ensure that Exim is compiled with large file
3095 support, which makes it possible to handle log files that are bigger than
3096 2^31.
3097
309840. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3099 instance) a domain was checked against a named list that involved a lookup,
3100 causing $domain_data to be set, then another domain was checked against the
3101 same list, then the first domain was re-checked, the value of $domain_data
3102 after the final check could be wrong. In particular, if the second check
3103 failed, it could be set empty. This bug probably also applied to
3104 $localpart_data.
3105
310641. The strip_trailing_dot option was not being applied to the address given
3107 with the -f command-line option.
3108
310942. The code for reading a message's header from the spool was incrementing
3110 $received_count, but never initializing it. This meant that the value was
3111 incorrect (doubled) while delivering a message in the same process in which
3112 it was received. In the most common configuration of Exim, this never
3113 happens - a fresh exec is done - but it can happen when
3114 deliver_drop_privilege is set.
3115
311643. When Exim logs an SMTP synchronization error - client data sent too soon -
3117 it now includes up to 150 characters of the unexpected data in the log
3118 line.
3119
312044. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3121 and building data strings. The size of both of these buffers was 10 000
3122 bytes - far larger than anybody would *ever* want, thought I. Needless to
3123 say, somebody hit the limit. I have increased the maximum line length to
3124 20 000 and the maximum data length of concatenated lines to 100 000. I have
3125 also fixed two bugs, because there was no checking on these buffers. Tsk,
3126 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3127 buffer is too small.
3128
312945. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3130 lsearch lookups. Now it does.
3131
313246. When parsing a route_list item in a manualroute router, a fixed-length
3133 buffer was used for the list of hosts. I made this 1024 bytes long,
3134 thinking that nobody would ever have a list of hosts that long. Wrong.
3135 Somebody had a whole pile of complicated expansion conditions, and the
3136 string was silently truncated, leading to an expansion error. It turns out
3137 that it is easier to change to an unlimited length (owing to other changes
3138 that have happened since this code was originally written) than to build
3139 structure for giving a limitation error. The length of the item that
3140 expands into the list of hosts is now unlimited.
3141
314247. The lsearch lookup could not handle data where the length of text line was
3143 more than 4095 characters. Such lines were truncated, leading to shortened
3144 data being returned. It should now handle lines of any length.
3145
314648. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3147 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3148 ACL").
3149
315049. Cosmetic tidy to scripts like exicyclog that are generated by globally
3151 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3152 no longer happens in comment lines. A list of replacements is now placed
3153 at the head of all of the source files, except those whose only change is
3154 to replace PERL_COMMAND in the very first #! line.
3155
315650. Replaced the slow insertion sort in queue.c, for sorting the list of
3157 messages on the queue, with a bottom-up merge sort, using code contributed
3158 by Michael Haardt. This should make operations like -bp somewhat faster on
3159 large queues. It won't affect queue runners, except when queue_run_in_order
3160 is set.
3161
316251. Installed eximstats 1.31 in the distribution.
3163
316452. Added support for SRV lookups to the dnslookup router.
3165
316653. If an ACL referred to $message_body or $message_body_end, the value was not
3167 reset for any messages that followed in the same SMTP session.
3168
316954. The store-handling optimization for building very long strings was not
3170 differentiating between the different store pools. I don't think this
3171 actually made any difference in practice, but I've tidied it.
3172
317355. While running the routers to verify a sender address, $sender_address
3174 was still set to the sender address. This is wrong, because when routing to
3175 send a bounce to the sender, it would be empty. Therefore, I have changed
3176 it so that, while verifying a sender address, $sender_address is set to <>.
3177 (There is no change to what happens when verifying a recipient address.)
3178
317956. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3180 target A or AAAA records (if not already returned) without resetting the
3181 qualify_single or search_parents options of the DNS resolver. These are
3182 inappropriate in this case because the targets of MX and SRV records must
3183 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3184 target that, when qualified, matched something in the local domain. These
3185 two options are now turned off when doing these lookups.
3186
318757. It seems that at least some releases of Reiserfs (which does not have the
3188 concept of a fixed number of inodes) returns zero and not -1 for the
3189 number of available inodes. This interacted badly with check_spool_inodes,
3190 which assumed that -1 was the "no such thing" setting. What I have done is
3191 to check that the total number of inodes is greater than zero before doing
3192 the test of how many are available.
3193
319458. When a "warn" ACL statement has a log_message modifier, the message is
3195 remembered, and not repeated. This is to avoid a lot of repetition when a
3196 message has many recipients that cause the same warning to be written.
3197 Howewer, Exim was preserving the list of already written lines for an
3198 entire SMTP session, which doesn't seem right. The memory is now reset if a
3199 new message is started.
3200
320159. The "rewrite" debugging flag was not showing the result of rewriting in the
3202 debugging output unless log_rewrite was also set.
3203
320460. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3205 of (int)(handle) when we know that handle contains (void *)(-1).
3206
320761. The Exim daemon panic-logs an error return when it closes the incoming
3208 connection. However "connection reset by peer" seems to be common, and
3209 isn't really an error worthy of noting specially, so that particular error
3210 is no long logged.
3211
321262. When Exim is trying to find all the local interfaces, it used to panic and
3213 die if the ioctl to get the interface flags failed. However, it seems that
3214 on at least one OS (Solaris 9) it is possible to have an interface that is
3215 included in the list of interfaces, but for which you get a failure error
3216 for this call. This happens when the interface is not "plumbed" into a
3217 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3218 failure of the "get flags" call assumes that the interface is down.
3219
322063. Added a ${eval10: operator, which assumes all numbers are decimal. This
3221 makes life easier for people who are doing arithmetic on fields extracted
3222 from dates, where you often get leading zeros that should not be
3223 interpreted as octal.
3224
322564. Added qualify_domain to the redirect router, to override the global
3226 setting.
3227
322865. If a pathologically long header line contained very many addresses (the
3229 report of this problem mentioned 10 000) and each of them was rewritten,
3230 Exim could use up a very large amount of memory. (It kept on making new
3231 copies of the header line as it rewrote, and never released the old ones.)
3232 At the expense of a bit more processing, the header rewriting function has
3233 been changed so that it no longer eats memory in this way.
3234
323566. The generation of the Received: header has been moved from the time that a
3236 message starts to be received, to the time that it finishes. The timestamp
3237 in the Received: header should now be very close to that of the <= log
3238 line. There are two side-effects of this change:
3239
3240 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3241 logged header lines no longer include the local Received: line, because
3242 it has not yet been created. The same applies to a copy of the message
3243 that is returned to a non-SMTP sender when a message is rejected.
3244
3245 (b) When a filter file is tested using -bf, no additional Received: header
3246 is added to the test message. After some thought, I decided that this
3247 is a bug fix.
3248
3249 This change does not affect the value of $received_for. It is still set
3250 after address rewriting, but before local_scan() is called.
3251
325267. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3253
325468. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3255 gave an unhelpful panic error message, and a defer error. I have managed to
3256 change this behaviour so that it now rejects any supplied certificate,
3257 which seems right, as the list of acceptable certificates is empty.
3258
325969. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3260 gave an unhelpful defer error. I have not managed to make this reject any
3261 supplied certificates, but the error message it gives is "no certificate
3262 supplied", which is not helpful.
3263
326470. exigrep's output now also includes lines that are not associated with any
3265 message, but which match the given pattern. Implemented by a patch from
3266 Martin Sluka, which also tidied up the Perl a bit.
3267
326871. Recipient callout verification, like sender verification, was using <> in
3269 the MAIL FROM command. This isn't really the right thing, since the actual
3270 sender may affect whether the remote host accepts the recipient or not. I
3271 have changed it to use the actual sender in the callout; this means that
3272 the cache record is now keyed on a recipient/sender pair, not just the
3273 recipient address. There doesn't seem to be a real danger of callout loops,
3274 since a callout by the remote host to check the sender would use <>.
3275 [SEE ABOVE: changed after hitting problems.]
3276
327772. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3278 temporary errors. However, in the case of such a code being given after
3279 the end of a data transmission (i.e. after ".") Exim was failing to write
3280 a retry record for the message. (Yes, there was some broken host that was
3281 actually sending 8xx at this point.)
3282
328373. An unknown lookup type in a host list could cause Exim to panic-die when
3284 the list was checked. (An example that provoked this was putting <; in the
3285 middle of a list instead of at the start.) If this happened during a DATA
3286 ACL check, a -D file could be left lying around. This kind of configuration
3287 error no longer causes Exim to die; instead it causes a defer errror. The
3288 incident is still logged to the main and panic logs.
3289
329074. Buglet left over from Exim 3 conversion. The message "too many messages
3291 in one connection" was written to the rejectlog but not the mainlog, except
3292 when address rewriting (yes!) was being logged.
3293
329475. Added write_rejectlog option.
3295
329676. When a system filter was run not as root (that is, when system_filter_user
3297 was set), the values of the $n variables were not being returned to the
3298 main process; thus, they were not subsequently available in the $sn
3299 variables.
3300
330177. Added +return_path_on_delivery log selector.
3302
330378. A connection timeout was being treated differently from recipients deferred
3304 when testing hosts_max_try with a message that was older than the host's
3305 retry timeout. (The host should not be counted, thus allowing all hosts to
3306 be tried at least once before bouncing.) This may have been the cause of an
3307 occasionally reported bug whereby a message would remain on the queue
3308 longer than the retry timeout, but would be bounced if a delivery was
3309 forced. I say "may" because I never totally pinned down the problem;
3310 setting up timeout/retry tests is difficult. See also the next item.
3311
331279. The ultimate address timeout was not being applied to errors that involved
3313 a combination of host plus message (for example, a timeout on a MAIL
3314 command). When an address resolved to a number of possible hosts, and they
3315 were not all tried for each delivery (e.g. because of hosts_max_try), a
3316 message could remain on the queue longer than the retry timeout.
3317
331880. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3319 Haardt.
3320
332181. Fixed an obscure SMTP outgoing bug which required at least the following
3322 conditions: (a) there was another message waiting for the same server;
3323 (b) the server returned 5xx to all RCPT commands in the first message so
3324 that the message was not completed; (c) the server dropped the connection
3325 or gave a negative response to the RSET that Exim sends to abort the
3326 transaction. The observed case was a dropped connection after DATA that had
3327 been sent in pipelining mode. That is, the server had advertised PIPELINING
3328 but was not implementing it correctly. The effect of the bug was incorrect
3329 behaviour, such as trying another host, and this could lead to a crash.
3330
3331
3332Exim version 4.30
3333-----------------
3334
3335 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3336 and daemon.c were passed as pointers to ints; they should have been
3337 pointers to socklen_t variables (which are typically unsigned ints).
3338
3339 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3340 fixed.
3341
3342 3. Fixed a really odd bug that affected only the testing scheme; patching a
3343 certain fixed string in the binary changed the value of another string that
3344 happened to be identical to the end of the original first string.
3345
3346 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3347 name", it returns that address as the IP address. On some operating
3348 systems (e.g. Solaris), it also passes back the IP address string as the
3349 "host name". However, on others (e.g. Linux), it passes back an empty
3350 string. Exim wasn't checking for this, and was changing the host name to an
3351 empty string, assuming it had been canonicized.
3352
3353 5. Although rare, it is permitted to have more than one PTR record for a given
3354 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3355 all the names associated with an address, because they do in Solaris.
3356 However, it seems that they do not in Linux for data that comes from the
3357 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3358 I found this out when I moved to a new Linux workstation and tried to run
3359 the Exim test suite.
3360
3361 To get round this problem I have changed the code so that it now does its
3362 own call to the DNS to look up PTR records when searching for a host name.
3363 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3364 addresses that are only in /etc/hosts are still found.
3365
3366 This behaviour is, however, controlled by an option called host_lookup_
3367 order, which defaults to "bydns:byaddr". If people want to use the other
3368 order, or indeed, just use one or the other means of lookup, they can
3369 specify it in this variable.
3370
3371 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3372 some operating systems, this comes back from gethostbyaddr() as an empty
3373 string, and this is what Exim used to test for. However, it seems that in
3374 other systems, "." is yielded. Exim now tests for this case too.
3375
3376 7. The values of check_spool_space and check_log_space are now held internally
3377 as a number of kilobytes instead of an absolute number of bytes. If a
3378 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3379 kilobyte. This means that much larger values can be stored.
3380
3381 8. Exim monitor: an attempt to get the action menu when not actually pointing
3382 at a message produces an empty menu entitled "No message selected". This
3383 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3384 no entries in it ("Shell widget menu has zero width and/or height"). So I
3385 have added a single, blank menu entry in this case.
3386
3387 9. Added ${quote_local_part.
3388
338910. MIME decoding is now applied to the contents of Subject: header lines when
3390 they are logged.
3391
339211. Now that a reference to $sender_host_address automatically causes a reverse
3393 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3394 host lookup before query-style lookups in lists that might use this
3395 variable. This has therefore been abolished, and the "net-" prefix is no
3396 longer necessary for query-style lookups.
3397
339812. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3399 have been a typo for LFLAGS, so it has been changed.
3400
340113. The install script calls Exim with "-C /dev/null" in order to find the
3402 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3403 to be output. Howeve, since Exim outputs its version number before the
3404 error, it didn't break the script. It just looked ugly. I fixed this by
3405 always allowing "-C /dev/null" if the caller is root.
3406
340714. Ignore overlarge ACL variable number when reading spool file - insurance
3408 against a later release with more variables having written the file.
3409
341015. The standard form for an IPv6 address literal was being rejected by EHLO.
3411 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3412 this, as well as the form without the "IPv6" on the front.
3413
341416. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3415 OS/Makefile-Darwin file.
3416
341717. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3418 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3419
342018. After change 4.21/52, "%ld" was used to format the contents of the $inode
3421 variable. However, some OS use ints for inodes. I've added cast to long int
3422 to get rid of the compiler warning.
3423
342419. I had forgotten to lock out "/../" in configuration file names when
3425 ALT_CONFIG_PREFIX was set.
3426
342720. Routers used for verification do not need to specify transports. However,
3428 if such a router generated a host list, and callout was configured, Exim
3429 crashed, because it could not find a port number from the (non-existent)
3430 transport. It now assumes port 25 in this circumstance.
3431
343221. Added the -t option to exigrep.
3433
343422. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3435 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3436 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3437 them is set, LOOKUP_LSEARCH is forced.
3438
343923. "exim -bV" now outputs a list of lookups that are included in the binary.
3440
344124. Added sender and host information to the "rejected by local_scan()" log
3442 line; previously there was no indication of these.
3443
344425. Added .include_if_exists.
3445
344626. Change 3.952/11 added an explicit directory sync on top of a file sync for
3447 Linux. It turns out that not all file systems support this. Apparently some
3448 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3449 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3450 not supported on the file descriptor, is now ignored when Exim is trying to
3451 sync a directory. This applies only to Linux.
3452
345327. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3454
345528. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3456 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3457 this.
3458
345929. Added support for maildirsize files from supplied patch (modified a bit).
3460
346130. The use of :fail: followed by an empty string could lead Exim to respond to
3462 sender verification failures with (e.g.):
3463
3464 550 Verification failed for <xxx>
3465 550 Sender verify failed
3466
3467 where the first response line was missing the '-' that indicates it is not
3468 the final line of the response.
3469
347031. The loop for finding the name of the user that called Exim had a hardwired
3471 limit of 10; it now uses the value of finduser_retries, which is used for
3472 all other user lookups.
3473
347432. Added $received_count variable, available in data and not_smtp ACLs, and at
3475 delivery time.
3476
347733. Exim was neglecting to zero errno before one call of strtol() when
3478 expanding a string and expecting an integer value. On some systems this
3479 resulted in spurious "integer overflow" errors. Also, it was casting the
3480 result into an int without checking.
3481
348234. Testing for a connection timeout using "timeout_connect" in the retry rules
3483 did not work. The code looks as if it has *never* worked, though it appears
3484 to have been documented since at least releast 1.62. I have made it work.
3485
348635. The "timeout_DNS" error in retry rules, also documented since at least
3487 1.62, also never worked. As it isn't clear exactly what this means, and
3488 clearly it isn't a major issue, I have abolished the feature by treating it
3489 as "timeout", and writing a warning to the main and panic logs.
3490
349136. The display of retry rules for -brt wasn't always showing the error code
3492 correctly.
3493
349437. Added new error conditions to retry rules: timeout_A, timeout_MX,
3495 timeout_connect_A, timeout_connect_MX.
3496
349738. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3498 to the empty sender.
3499
350039. The daemon was not analysing the content of -oX till after it had closed
3501 stderr and disconnected from the controlling terminal. This meant that any
3502 syntax errors were only noted on the panic log, and the return code from
3503 the command was 0. By re-arranging the code a little, I've made the
3504 decoding happen first, so such errors now appear on stderr, and the return
3505 code is 1. However, the actual setting up of the sockets still happens in
3506 the disconnected process, so errors there are still only recorded on the
3507 panic log.
3508
350940. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3510 connections (as happens on some IP stacks) was logged at start up time as
3511 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3512 it from "IPv6 and IPv4", which means that two separate sockets are being
3513 used.
3514
351541. The debug output for gethostbyname2() or getipnodebyname() failures now
3516 says whether AF_INET or AF_INET6 was passed as an argument.
3517
351842. Exiwhat output was messed up when time zones were included in log
3519 timestamps.
3520
352143. Exiwhat now gives more information about the daemon's listening ports,
3522 and whether -tls-on-connect was used.
3523
352444. The "port" option of the smtp transport is now expanded.
3525
352645. A "message" modifier in a "warn" statement in a non-message ACL was being
3527 silently ignored. Now an error message is written to the main and panic
3528 logs.
3529
353046. There's a new ACL modifier called "logwrite" which writes to a log file
3531 as soon as it is encountered.
3532
353347. Added $local_user_uid and $local_user_gid at routing time.
3534
353548. Exim crashed when trying to verify a sender address that was being
3536 rewritten to "<>".
3537
353849. Exim was recognizing only a space character after ".include". It now also
3539 recognizes a tab character.
3540
354150. Fixed several bugs in the Perl script that creates the exim.8 man page by
3542 extracting the relevant information from the specification. The man page no
3543 longer contains scrambled data for the -d option, and I've added a section
3544 at the front about calling Exim under different names.
3545
354651. Added "extra_headers" argument to the "mail" command in filter files.
3547
354852. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3549 crash.
3550
355153. Installed eximstats 1.29.
3552
355354. Added transport_filter_timeout as a generic transport option.
3554
355555. Exim no longer adds an empty Bcc: header to messages that have no To: or
3556 Cc: header lines. This was required by RFC 822, but it not required by RFC
3557 2822.
3558
355956. Exim used to add From:, Date:, and Message-Id: header lines to any
3560 incoming messages that did not have them. Now it does so only if the
3561 message originates locally, that is, if there is no associated remote host
3562 address. When Resent- header lines are present, this applies to the Resent-
3563 lines rather than the non-Resent- lines.
3564
356557. Drop incoming SMTP connection after too many syntax or protocol errors. The
3566 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3567
356858. Messages for configuration errors now include the name of the main
3569 configuration file - useful now that there may be more than one file in a
3570 list (.included file names were always shown).
3571
357259. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3573 for those rare installations that do not start the daemon as root or run it
3574 setuid root. I've cut out the call to initgroups() if the daemon is not
3575 root at that time.
3576
357760. The Exim user and group can now be bound into the binary as text strings
3578 that are looked up at the start of Exim's processing.
3579
358061. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3581
358262. Added $mailstore_basename variable.
3583
358463. Installed patch to sieve.c from Michael Haardt.
3585
358664. When Exim failed to open the panic log after failing to open the main log,
3587 the original message it was trying to log was written to stderr and debug
3588 output, but if they were not available (the usual case in production), it
3589 was lost. Now it is written to syslog before the two lines that record the
3590 failures to open the logs.
3591
359265. Users' Exim filters run in subprocesses under the user's uid. It is
3593 possible for a "deliver" command or an alias in a "personal" command to
3594 provoke an address rewrite. If logging of address rewriting is configured,
3595 this fails because the process is not running as root or exim. There may be
3596 a better way of dealing with this, but for the moment (because 4.30 needs
3597 to be released), I have disabled address rewrite logging when running a
3598 filter in a non-root, non-exim process.
3599
3600
3601Exim version 4.24
3602-----------------
3603
3604 1. The buildconfig auxiliary program wasn't quoting the value set for
3605 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3606 not defined. This bug was masked in 4.22 by the effect that was fixed in
3607 change 4.23/1.
3608
3609 2. Some messages that were rejected after a message id was allocated were
3610 shown as "incomplete" by exigrep. It no longer does this for messages that
3611 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3612
3613 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3614 have allow_domain_literals set, the ID did not get logged in the <= line.
3615 Domain literals are now always recognized in Message-ID: header lines.
3616
3617 4. The first argument for a ${extract expansion item is the key name or field
3618 number. Leading and trailing spaces in this item were not being ignored,
3619 causing some misleading effects.
3620
3621 5. When deliver_drop_privilege was set, single queue runner processes started
3622 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3623 same command in the process it spins off) were not dropping privilege.
3624
3625 6. When the daemon running as "exim" started a queue runner, it always
3626 re-executed Exim in the spun-off process. This is a waste of effort when