6e61fa9b6786adeb4b6e5ea1ae83e8f9a3ebdda7
[exim.git] / doc / doc-txt / ChangeLog
1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.484 2007/02/20 15:58:02 ph10 Exp $
2
3 Change log file for Exim from version 4.21
4 -------------------------------------------
5
6 Exim version 4.67
7 -----------------
8
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
11 Jan Srzednicki.
12
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
14 issue a MAIL command.
15
16 PH/02 In an ACL statement such as
17
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
19
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
25
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
28
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
31
32 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
33 FSYNC, which compiles an option called disable_fsync that allows for
34 bypassing fsync(). The documentation is heavily laced with warnings.
35
36 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
37
38 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
39 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
40 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
41 including adding "make clean"; (3) Added -fPIC when compiling the test
42 dynamically loaded module, to get rid of a warning.
43
44 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
45 message fails, move_frozen_messages = true and ignore_bounce_errors_after
46 = 0s. The bug is otherwise harmless.
47
48 PH/07 There was a bug in the dovecot authenticator such that the value of
49 $auth1 could be overwritten, and so not correctly preserved, after a
50 successful authentication. This usually meant that the value preserved by
51 the server_setid option was incorrect.
52
53 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
54
55 PH/09 Installed PCRE release 7.0.
56
57 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
58 run for batched SMTP input. It is now run at the start of every message
59 in the batch. While fixing this I discovered that the process information
60 (output by running exiwhat) was not always getting set for -bs and -bS
61 input. This is fixed, and it now also says "batched" for BSMTP.
62
63 PH/11 Added control=no_pipelining.
64
65 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
66 patch, slightly modified), and move the expansion of helo_data till after
67 the connection is made in the smtp transport (so it can use these
68 values).
69
70 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
71
72 PH/14 Added log_selector = +pid.
73
74 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
75
76 PH/16 Add ${if forany and ${if forall.
77
78 PH/17 Added dsn_from option to vary the From: line in DSNs.
79
80 PH/18 Flush SMTP output before performing a callout, unless control =
81 no_callout_flush is set.
82
83 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
84 was true (the default) a successful delivery failed to delete the retry
85 item, thus causing premature timeout of the address. The bug is now
86 fixed.
87
88 PH/20 Added hosts_avoid_pipelining to the smtp transport.
89
90 PH/21 Long custom messages for fakedefer and fakereject are now split up
91 into multiline reponses in the same way that messages for "deny" and
92 other ACL rejections are.
93
94 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
95 with slight modification.
96
97 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
98 draft, changing the syntax and factoring some duplicate code".
99
100 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
101 for deliveries of the second and subsequent messages over the same SMTP
102 connection.
103
104 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
105 ${reduce, with only minor "tidies".
106
107 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
108
109 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
110 expansion side effects.
111
112 PH/27 When a message times out after an over-quota error from an Exim-imposed
113 quota, the bounce message says "mailbox is full". This message was not
114 being given when it was a system quota that was exceeded. It now should
115 be the same.
116
117 MH/03 Made $recipients available in local_scan(). local_scan() already has
118 better access to the recipient list through recipients_list[], but
119 $recipients can be useful in postmaster-provided expansion strings.
120
121 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
122 in the case of a MAIL command with additional options following the
123 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
124 were accidentally chopped off.
125
126 PH/29 SMTP synchronization checks are implemented when a command is read -
127 there is a check that no more input is waiting when there shouldn't be
128 any. However, for some commands, a delay in an ACL can mean that it is
129 some time before the response is written. In this time, more input might
130 arrive, invalidly. So now there are extra checks after an ACL has run for
131 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
132 pipelining has not been advertised.
133
134
135 Exim version 4.66
136 -----------------
137
138 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
139 fixed by 4.65/MH/01 (is this a record?) are fixed:
140
141 (i) An empty string was always treated as zero by the numeric comparison
142 operators. This behaviour has been restored.
143
144 (ii) It is documented that the numeric comparison operators always treat
145 their arguments as decimal numbers. This was broken in that numbers
146 starting with 0 were being interpreted as octal.
147
148 While fixing these problems I realized that there was another issue that
149 hadn't been noticed. Values of message_size_limit (both the global option
150 and the transport option) were treated as octal if they started with 0.
151 The documentation was vague. These values are now always treated as
152 decimal, and I will make that clear in the documentation.
153
154
155 Exim version 4.65
156 -----------------
157
158 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
159 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
160 versions. (#438)
161
162 MH/01 Don't check that the operands of numeric comparison operators are
163 integers when their expansion is in "skipping" mode (fixes bug
164 introduced by 4.64-PH/07).
165
166 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
167 child addresses, Exim now panics and dies. Previously, because the count
168 is held in a short int, deliveries were likely to be lost. As such a
169 large number of recipients for a single message is ridiculous
170 (performance will be very, very poor), I have chosen to impose a limit
171 rather than extend the field.
172
173
174 Exim version 4.64
175 -----------------
176
177 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
178 leftover -K file (the existence of which was triggered by #402).
179 While we were at it, introduced process PID as part of the -K
180 filename. This should rule out race conditions when creating
181 these files.
182
183 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
184 processing considerably. Previous code took too long for large mails,
185 triggering a timeout which in turn triggers #401.
186
187 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
188 in the DK code in transports.c. sendfile() is not really portable,
189 hence the _LINUX specificness.
190
191 TF/01 In the add_headers option to the mail command in an Exim filter,
192 there was a bug that Exim would claim a syntax error in any
193 header after the first one which had an odd number of characters
194 in the field name.
195
196 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
197 callout verification, Exim cached a "reject" for the entire domain. This
198 is correct for most verifications, but it is not correct for a recipient
199 verification with use_sender or use_postmaster set, because in that case
200 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
201 case of MAIL FROM:<> rejection from other early rejections (e.g.
202 rejection of HELO). When verifying a recipient using a non-null MAIL
203 address, the cache is ignored if it shows MAIL FROM:<> rejection.
204 Whatever the result of the callout, the value of the domain cache is
205 left unchanged (for any other kind of callout, getting as far as trying
206 RCPT means that the domain itself is ok).
207
208 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
209 gcc 4.1.1 threw up.
210
211 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
212 manifest itself as EPIPE rather than ECONNECT. When tidying away a
213 session, the daemon ignores ECONNECT errors and logs others; it now
214 ignores EPIPE as well.
215
216 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
217 (quoted-printable decoding).
218
219 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
220 later the small subsequent patch to fix an introduced bug.
221
222 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
223
224 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
225
226 PH/08 An error is now given if message_size_limit is specified negative.
227
228 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
229 to be given (somewhat) arbitrary names.
230
231 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
232 in 4.64-PH/09.
233
234 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
235 miscellaneous code fixes
236
237 PH/10 Added the log_reject_target ACL modifier to specify where to log
238 rejections.
239
240 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
241 hostname. This is wrong, because it relates to the incoming message (and
242 probably the interface on which it is arriving) and not to the outgoing
243 callout (which could be using a different interface). This has been
244 changed to use the value of the helo_data option from the smtp transport
245 instead - this is what is used when a message is actually being sent. If
246 there is no remote transport (possible with a router that sets up host
247 addresses), $smtp_active_hostname is used.
248
249 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
250 tweaks were necessary in order to get it to work (see also 21 below):
251 (a) The code assumed that strncpy() returns a negative number on buffer
252 overflow, which isn't the case. Replaced with Exim's string_format()
253 function.
254 (b) There were several signed/unsigned issues. I just did the minimum
255 hacking in of casts. There is scope for a larger refactoring.
256 (c) The code used strcasecmp() which is not a standard C function.
257 Replaced with Exim's strcmpic() function.
258 (d) The code set only $1; it now sets $auth1 as well.
259 (e) A simple test gave the error "authentication client didn't specify
260 service in request". It would seem that Dovecot has changed its
261 interface. Fortunately there's a specification; I followed it and
262 changed what the client sends and it appears to be working now.
263
264 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
265 decoding.
266
267 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
268 address A is aliased to B and C, where B exists and C does not. Without
269 -v the output is "A verified" because verification stops after a
270 successful redirection if more than one address is generated. However,
271 with -v the child addresses are also verified. Exim was outputting "A
272 failed to verify" and then showing the successful verification for C,
273 with its parentage. It now outputs "B failed to verify", showing B's
274 parentage before showing the successful verification of C.
275
276 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
277 look up a TXT record in a specific list after matching in a combined
278 list.
279
280 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
281 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
282 they consult the DNS. I had assumed they would set it the way they
283 wanted; and indeed my experiments on Linux seem to show that in some
284 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
285 To be on the safe side, however, I have now made the interface to
286 host_find_byname() similar to host_find_bydns(), with an argument
287 containing the DNS resolver options. The host_find_byname() function now
288 sets these options at its start, just as host_find_bydns() does. The smtp
289 transport options dns_qualify_single and dns_search_parents are passed to
290 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
291 of host_find_byname() use the default settings of RES_DEFNAMES
292 (qualify_single) but not RES_DNSRCH (search_parents).
293
294 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
295 spool_read_header() do less string testing, by means of a preliminary
296 switch on the second character of optional "-foo" lines. (This is
297 overdue, caused by the large number of possibilities that now exist.
298 Originally there were few.) While I was there, I also converted the
299 str(n)cmp tests so they don't re-test the leading "-" and the first
300 character, in the hope this might squeeze out yet more improvement.
301
302 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
303 flag allowing group syntax was set by the header_syntax check but not
304 turned off, possible causing trouble later; (2) The flag was not being
305 set at all for the header_verify test, causing "group"-style headers to
306 be rejected. I have now set it in this case, and also caused header_
307 verify to ignore an empty address taken from a group. While doing this, I
308 came across some other cases where the code for allowing group syntax
309 while scanning a header line wasn't quite right (mostly, not resetting
310 the flag correctly in the right place). These bugs could have caused
311 trouble for malformed header lines. I hope it is now all correct.
312
313 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
314 with the "reply" argument non-NULL. The code, however (which originally
315 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
316 but it didn't always do it. This confused somebody who was copying the
317 code for some other use. I have removed all the tests.
318
319 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
320 feature that was used to support insecure browsers during the U.S. crypto
321 embargo. It requires special client support, and Exim is probably the
322 only MTA that supported it -- and would never use it because real RSA is
323 always available. This code has been removed, because it had the bad
324 effect of slowing Exim down by computing (never used) parameters for the
325 RSA_EXPORT functionality.
326
327 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
328 authenticator to fail if there's a tab character in the incoming data
329 (there should never be unless someone is messing about, as it's supposed
330 to be base64-encoded). Also added, on Timo's advice, the "secured" option
331 if the connection is using TLS or if the remote IP is the same as the
332 local IP, and the "valid-client-cert option" if a client certificate has
333 been verified.
334
335 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
336 authenticators. This can be used for authorization after authentication
337 succeeds. (In the case of plaintext, it servers for both authentication
338 and authorization.)
339
340 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
341 if any retry times were supplied.
342
343 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
344 connection, where there is no client IP address to check. In this
345 situation, the verify now always succeeds.
346
347 PH/25 Applied John Jetmore's -Mset patch.
348
349 PH/26 Added -bem to be like -Mset, but loading a message from a file.
350
351 PH/27 In a string expansion for a processed (not raw) header when multiple
352 headers of the same name were present, leading whitespace was being
353 removed from all of them, but trailing whitespace was being removed only
354 from the last one. Now trailing whitespace is removed from each header
355 before concatenation. Completely empty headers in a concatenation (as
356 before) are ignored.
357
358 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
359 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
360
361 PH/29 [Removed. This was a change that I later backed out, and forgot to
362 correct the ChangeLog entry (that I had efficiently created) before
363 committing the later change.]
364
365 PH/30 Exim was sometimes attempting to deliver messages that had suffered
366 address errors (4xx response to RCPT) over the same connection as other
367 messages routed to the same hosts. Such deliveries are always "forced",
368 so retry times are not inspected. This resulted in far too many retries
369 for the affected addresses. The effect occurred only when there were more
370 hosts than the hosts_max_try setting in the smtp transport when it had
371 the 4xx errors. Those hosts that it had tried were not added to the list
372 of hosts for which the message was waiting, so if all were tried, there
373 was no problem. Two fixes have been applied:
374
375 (i) If there are any address or message errors in an SMTP delivery, none
376 of the hosts (tried or untried) are now added to the list of hosts
377 for which the message is waiting, so the message should not be a
378 candidate for sending over the same connection that was used for a
379 successful delivery of some other message. This seems entirely
380 reasonable: after all the message is NOT "waiting for some host".
381 This is so "obvious" that I'm not sure why it wasn't done
382 previously. Hope I haven't missed anything, but it can't do any
383 harm, as the worst effect is to miss an optimization.
384
385 (ii) If, despite (i), such a delivery is accidentally attempted, the
386 routing retry time is respected, so at least it doesn't keep
387 hammering the server.
388
389 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
390 in ${readsocket because some servers need this prod.
391
392 PH/32 Added some extra debug output when updating a wait-xxx database.
393
394 PH/33 The hint "could be header name not terminated by colon", which has been
395 given for certain expansion errors for a long time, was not being given
396 for the ${if def:h_colon_omitted{... case.
397
398 PH/34 The spec says: "With one important exception, whenever a domain list is
399 being scanned, $domain contains the subject domain." There was at least
400 one case where this was not true.
401
402 PH/35 The error "getsockname() failed: connection reset by peer" was being
403 written to the panic log as well as the main log, but it isn't really
404 panic-worthy as it just means the connection died rather early on. I have
405 removed the panic log writing for the ECONNRESET error when getsockname()
406 fails.
407
408 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
409 runs only) independently of the message's sender address. This meant
410 that, if the 4xx error was in fact related to the sender, a different
411 message to the same recipient with a different sender could confuse
412 things. In particualar, this can happen when sending to a greylisting
413 server, but other circumstances could also provoke similar problems.
414 I have changed the default so that the retry time for these errors is now
415 based a combination of the sender and recipient addresses. This change
416 can be overridden by setting address_retry_include_sender=false in the
417 smtp transport.
418
419 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
420 remote server are returned as part of bounce messages. This was not
421 happening for LMTP over a pipe (the lmtp transport), but now it is the
422 same for both kinds of LMTP.
423
424 PH/38 Despite being documented as not happening, Exim was rewriting addresses
425 in header lines that were in fact CNAMEs. This is no longer the case.
426
427 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
428 and queue runs started by the daemon processed all messages. This has
429 been fixed so that -R and -S can now usefully be given with -q<time>.
430
431 PH/40 Import PCRE release 6.7 (fixes some bugs).
432
433 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
434
435 PH/42 Give an error if -q is specified more than once.
436
437 PH/43 Renamed the variables $interface_address and $interface_port as
438 $received_ip_address and $received_port, to make it clear that these
439 values apply to message reception, and not to the outgoing interface when
440 a message is delivered. (The old names remain recognized, of course.)
441
442 PH/44 There was no timeout on the connect() call when using a Unix domain
443 socket in the ${readsocket expansion. There now is.
444
445 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
446 be meaningful with "accept".
447
448 SC/01 Eximstats V1.43
449 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
450
451 SC/02 Eximstats V1.44
452 Use a glob alias rather than an array ref in the generated
453 parser. This improves both readability and performance.
454
455 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
456 Collect SpamAssassin and rejection statistics.
457 Don't display local sender or destination tables unless
458 there is data to show.
459 Added average volumes into the top table text output.
460
461 SC/04 Eximstats V1.46
462 Collect data on the number of addresses (recipients)
463 as well as the number of messages.
464
465 SC/05 Eximstats V1.47
466 Added 'Message too big' to the list of mail rejection
467 reasons (thanks to Marco Gaiarin).
468
469 SC/06 Eximstats V1.48
470 Mainlog lines which have GMT offsets and are too short to
471 have a flag are now skipped.
472
473 SC/07 Eximstats V1.49 (Alain Williams)
474 Added the -emptyok flag.
475
476 SC/08 Eximstats V1.50
477 Fixes for obtaining the IP address from reject messages.
478
479 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
480 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
481 whitesspace changes from 4.64-PH/27
482
483 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
484 match 4.64-PH/13
485
486 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
487 are found, allow negative numbers in numeric criteria)
488
489 JJ/06 exipick.20061117.2, added new $message_body_missing variable
490
491 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
492 to match changes made in 4.64-PH/43
493
494 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
495
496 PH/47 Put in an explicit test for a DNS lookup of an address record where the
497 "domain" is actually an IP address, and force a failure. This locks out
498 those revolvers/nameservers that support "A-for-A" lookups, in
499 contravention of the specifications.
500
501 PH/48 When a host name was looked up from an IP address, and the subsequent
502 forward lookup of the name timed out, the host name was left in
503 $sender_host_name, contrary to the specification.
504
505 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
506 restricted to single-key lookups, Exim was not diagnosing an error if
507 * or *@ was used with a query-style lookup.
508
509 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
510
511 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
512 long ago, but noone interested enough thought of it. Let's just say that
513 the "1.1" means that there are some new functions that weren't there at
514 some point in the past.
515
516 PH/51 Error processing for expansion failure of helo_data from an smtp
517 transport during callout processing was broken.
518
519 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
520 tested/used via the -bh/-bhc/-bs options.
521
522 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
523 bug, fixed in subsequent PCRE releases).
524
525 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
526 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
527
528 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
529
530
531 Exim version 4.63
532 -----------------
533
534 SC/01 Use a glob alias rather than an array ref in eximstats generated
535 parser. This improves both readability and performance.
536
537 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
538 Don't display local sender or destination tables in eximstats unless
539 there is data to show.
540 Added average volumes into the eximstats top table text output.
541
542 SC/03 Collect data on the number of addresses (recipients) as well
543 as the number of messages in eximstats.
544
545 TF/01 Correct an error in the documentation for the redirect router. Exim
546 does (usually) call initgroups() when daemonizing.
547
548 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
549 with consistent privilege compared to when running as a daemon.
550
551 TF/03 Note in the spec that $authenticated_id is not set for local
552 submissions from trusted users.
553
554 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
555 Thanks to Dean Brooks <dean@iglou.com> for the patch.
556
557 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
558 by adding some example configuration directives to the default
559 configuration file. A little bit of work is required to uncomment the
560 directives and define how usernames and passwords are checked, but
561 there is now a framework to start from.
562
563 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
564 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
565 without this. I don't know how relevant this is to other LDAP libraries.
566
567 PH/02 Add the verb name to the "unknown ACL verb" error.
568
569 PH/03 Magnus Holmgren's patch for filter_prepend_home.
570
571 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
572
573 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
574 directory not expanded when it should be if an expanded home directory
575 was set for the address (which is overridden by the transport).
576
577 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
578 libradius.
579
580 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
581 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
582 because it is too late at that time, and has no effect.
583
584 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
585 security issue with \' (bugzilla #107). I could not use the
586 PQescapeStringConn() function, because it needs a PGconn value as one of
587 its arguments.
588
589 PH/08 When testing addresses using -bt, indicate those final addresses that
590 are duplicates that would not cause an additional delivery. At least one
591 person was confused, thinking that -bt output corresponded to deliveries.
592 (Suppressing duplicates isn't a good idea as you lose the information
593 about possibly different redirections that led to the duplicates.)
594
595 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
596 systems where poll() doesn't work, in particular OS X.
597
598 PH/10 Added more information to debugging output for retry time not reached.
599
600 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
601 operations in malware.c.
602
603 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
604 signatures.
605
606 PH/13 If write_rejectlog was set false when logging was sent to syslog with
607 syslog_duplication set false, log lines that would normally be written
608 both the the main log and to the reject log were not written to syslog at
609 all.
610
611 PH/14 In the default configuration, change the use of "message" in ACL warn
612 statements to "add_header".
613
614 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
615 not followed by a command (e.g. "seen endif").
616
617 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
618 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
619 latter.
620
621 PH/17 Added extra conditions to the default value of delay_warning_condition
622 so that it is now:
623
624 ${if or { \
625 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
626 { match{$h_precedence:}{(?i)bulk|list|junk} } \
627 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
628 }{no}{yes}}
629
630 The Auto-Submitted: and various List- headers are standardised, whereas I
631 don't think Precedence: ever was.
632
633 PH/18 Refactored debugging code in route_finduser() to show more information,
634 in particular, the error code if getpwnam() issues one.
635
636 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
637 This is apparently needed in addition to the PH/07 change above to avoid
638 any possible encoding problems.
639
640 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
641 but not after initializing Perl.
642
643 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
644 output them only if debugging. By default they are written stderr,
645 apparently, which is not desirable.
646
647 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
648 queries.
649
650 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
651 --not options
652
653 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
654
655 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
656 authenticated or an ident call has been made. Suppress the default
657 values for $authenticated_id and $authenticated_sender (but permit -oMai
658 and -oMas) when testing with -bh.
659
660 PH/24 Re-jigged the order of the tests in the default configuration so that the
661 tests for valid domains and recipients precede the DNS black list and CSA
662 tests, on the grounds that those ones are more expensive.
663
664 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
665 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
666 This bug exists in every version of Exim that I still have, right back to
667 0.12.
668
669 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
670 However, an attempt to turn on case-sensitivity in a regex key by
671 including (?-i) didn't work because the subject string was already
672 lowercased, and the effects were non-intuitive. It turns out that a
673 one-line patch can be used to allow (?-i) to work as expected.
674
675
676 Exim version 4.62
677 -----------------
678
679 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
680 other effects) broke the use of negated acl sub-conditions.
681
682 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
683 patch).
684
685 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
686 "Deny" causes Exim to reject the incoming connection with a 554 error.
687 Unfortunately, if there is a major crisis, such as a disk failure,
688 tcp-wrappers gives "deny", whereas what one would like would be some
689 kind of temporary error. A kludge has been added to help with this.
690 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
691 554 error is used if errno is still zero or contains ENOENT (which occurs
692 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
693 451 error is used.
694
695 PH/03 Add -lutil to the default FreeBSD LIBS setting.
696
697 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
698 errors. Otherwise a message that provokes a temporary error (when other
699 messages do not) can cause a whole host to time out.
700
701 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
702 addresses were routed directly to files or pipes from a redirect router.
703 File deliveries just didn't batch; pipe deliveries might have suffered
704 odd errors.
705
706 PH/06 A failure to get a lock for a hints database would erroneously always say
707 "Failed to get write lock", even when it was really a read lock.
708
709 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
710 of 0600. This has been changed to use the value of the lockfile_mode
711 option (which defaults to 0600).
712
713 PH/08 Applied small patch from the Sieve maintainer.
714
715 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
716 folder from quota calculations, a direct delivery into this folder messed
717 up the contents of the maildirsize file. This was because the regex was
718 used only to exclude .Trash (or whatever) when the size of the mailbox
719 was calculated. There was no check that a delivery was happening into an
720 excluded directory. This bug has been fixed by ignoring all quota
721 processing for deliveries into excluded directories.
722
723 PH/10 Added the maildirfolder_create_regex option to appendfile.
724
725
726 Exim version 4.61
727 -----------------
728
729 PH/01 The code for finding all the local interface addresses on a FreeBSD
730 system running IPv6 was broken. This may well have applied to all BSD
731 systems, as well as to others that have similar system calls. The broken
732 code found IPv4 interfaces correctly, but gave incorrect values for the
733 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
734 that it would not match correctly against @[] and not recognize the IPv6
735 addresses as local.
736
737 PH/02 The ipliteral router was not recognizing addresses of the form user@
738 [ipv6:....] because it didn't know about the "ipv6:" prefix.
739
740 PH/03 Added disable_ipv6.
741
742 PH/04 Changed $reply_address to use the raw form of the headers instead of the
743 decoded form, because it is most often used to construct To: headers
744 lines in autoreplies, and the decoded form may well be syntactically
745 invalid. However, $reply_address has leading white space removed, and all
746 newlines turned into spaces so that the autoreply transport does not
747 grumble.
748
749 PH/05 If group was specified without a user on a router, and no group or user
750 was specified on a transport, the group from the router was ignored.
751
752 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
753 for visible compile-time settings that can be used to change these
754 numbers, for those that want even more. Backwards compatibility with old
755 spool files has been maintained. However, going back to a previous Exim
756 release will lost any variables that are in spool files.
757
758 PH/07 Two small changes when running in the test harness: increase delay when
759 passing a TCP/IP connection to a new process, in case the original
760 process has to generate a bounce, and remove special handling of
761 127.0.0.2 (sic), which is no longer necessary.
762
763 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
764 be the same on different OS.
765
766 PH/09 Moved a debug statement in filter processing to avoid a race problem when
767 testing.
768
769 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
770 whether --show-vars was specified or not
771
772 JJ/02 exipick: Added support for new ACL variable spool format introduced
773 in 4.61-PH/06
774
775 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
776 syntactically invalid From: or Reply-to: line, and a filter used this to
777 generate an autoreply, and therefore failed to obtain an address for the
778 autoreply, Exim could try to deliver to a non-existent relative file
779 name, causing unrelated and misleading errors. What now happens is that
780 it logs this as a hard delivery error, but does not attempt to create a
781 bounce message.
782
783 PH/11 The exinext utility has a -C option for testing purposes, but although
784 the given file was scanned by exinext itself; it wasn't being passed on
785 when Exim was called.
786
787 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
788 an end-of-file indication when reading a command response.
789
790 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
791 compiled. In many other places in Exim, IPv6 addresses are always
792 recognized, so I have changed this. It also means that IPv4 domain
793 literals of the form [IPV4:n.n.n.n] are now always recognized.
794
795 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
796 used if the router is not running as root, for example, when verifying at
797 ACL time, or when using -bh. The debugging output from this situation was
798 non-existent - all you got was a failure to exec. I have made two
799 changes:
800
801 (a) Failures to set uid/gid, the current directory, or a process leader
802 in a subprocess such as that created by queryprogram now generate
803 suitable debugging ouput when -d is set.
804
805 (b) The queryprogram router detects when it is not running as root,
806 outputs suitable debugging information if -d is set, and then runs
807 the subprocess without attempting to change uid/gid.
808
809 PH/15 Minor change to Makefile for building test_host (undocumented testing
810 feature).
811
812 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
813 additional section of a DNS packet that returns MX or SRV records.
814 Instead, it always explicitly searches for A/AAAA records. This avoids
815 major problems that occur when a DNS server includes only records of one
816 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
817 fixed another bug: if SRV records were looked up and the corresponding
818 address records were *not* found in the additional section, the port
819 values from the SRV records were lost.
820
821 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
822 using the correct key (the original address) when searching the retry
823 rules in order to find which one to use for generating the retry hint.
824
825 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
826 adding the default one. Similarly, if it contains a Reply-To: header, the
827 errors_reply_to option, if set, is not used.
828
829 PH/19 When calculating a retry time, Exim used to measure the "time since
830 failure" by looking at the "first failed" field in the retry record. Now
831 it does not use this if it is later than than the arrival time of the
832 message. Instead it uses the arrival time. This makes for better
833 behaviour in cases where some deliveries succeed, thus re-setting the
834 "first failed" field. An example is a quota failure for a huge message
835 when small messages continue to be delivered. Without this change, the
836 "time since failure" will always be short, possible causing more frequent
837 delivery attempts for the huge message than are intended.
838 [Note: This change was subsequently modified - see PH/04 for 4.62.]
839
840 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
841 $1, $2, $3) because the numerical variables can be reset during some
842 expansion items (e.g. "match"), thereby losing the authentication data.
843
844 PH/21 Make -bV show the size of off_t variables so that the test suite can
845 decide whether to run tests for quotas > 2G.
846
847 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
848 mailbox_size, and mailbox_filecount in the appendfile transport. If a
849 filecount value is greater than 2G or if a quota value is greater than 2G
850 on a system where the size of off_t is not greater than 4, a panic error
851 is given.
852
853 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
854 never match. The debug and -bh output now contains an explicit error
855 message indicating a malformed IPv4 address or mask.
856
857 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
858 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
859 PH/23 above applies.
860
861 PH/25 Do not write to syslog when running in the test harness. The only
862 occasion when this arises is a failure to open the main or panic logs
863 (for which there is an explicit test).
864
865 PH/26 Added the /no_tell option to "control=freeze".
866
867 PH/27 If a host name lookup failed very early in a connection, for example, if
868 the IP address matched host_lookup and the reverse lookup yielded a name
869 that did not have a forward lookup, an error message of the form "no IP
870 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
871 could be logged. Now it outputs the IP address instead of "NULL".
872
873 PH/28 An enabling patch from MH: add new function child_open_exim2() which
874 allows the sender and the authenticated sender to be set when
875 submitting a message from within Exim. Since child_open_exim() is
876 documented for local_scan(), the new function should be too.
877
878 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
879 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
880 results in an empty string is now treated as unset.
881
882 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
883
884 PH/31 Added sender_verify_fail logging option.
885
886 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
887 needed by RFC 822 but not by RFC 2822 was commented out. I have now
888 tidied the source and removed it altogether.
889
890 PH/33 When a queue run was abandoned because the load average was too high, a
891 log line was always written; now it is written only if the queue_run log
892 selector is set. In addition, the log line for abandonment now contains
893 information about the queue run such as the pid. This is always present
894 in "start" and "stop" lines but was omitted from the "abandon" line.
895
896 PH/34 Omit spaces between a header name and the colon in the error message that
897 is given when verify = headers_syntax fails (if there are lots of them,
898 the message gets confusing).
899
900 PH/35 Change the default for dns_check_names_pattern to allow slashes within
901 names, as there are now some PTR records that contain slashes. This check
902 is only to protect against broken name servers that fall over on strange
903 characters, so the fact that it applies to all lookups doesn't matter.
904
905 PH/36 Now that the new test suite is complete, we can remove some of the
906 special code in Exim that was needed for the old test suite. For example,
907 sorting DNS records because real resolvers return them in an arbitrary
908 order. The new test suite's fake resolver always returns records in the
909 same order.
910
911 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
912 bounces) except when queue_only is set, to avoid logging races between
913 the different processes.
914
915 PH/38 Panic-die if .include specifies a non-absolute path.
916
917 PH/39 A tweak to the "H" retry rule from its user.
918
919 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
920 a label. They prevented compilation on older perls.
921
922 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
923 a warning to be raised on newish perls.
924
925 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
926 on queue. Changes to match documented behaviour of showing count of
927 messages matching specified criteria.
928
929 PH/40 Changed the default ident timeout from 30s to 5s.
930
931 PH/41 Added support for the use of login_cap features, on those BSD systems
932 that have them, for controlling the resources used by pipe deliveries.
933
934 PH/42 The content-scanning code uses fopen() to create files in which to put
935 message data. Previously it was not paying any attention to the mode of
936 the files. Exim runs with umask(0) because the rest of the code creates
937 files with open(), and sets the required mode explicitly. Thus, these
938 files were ending up world-writeable. This was not a big issue, because,
939 being within the spool directory, they were not world-accessible. I have
940 created a function called modefopen, which takes an additional mode
941 argument. It sets umask(777), creates the file, chmods it to the required
942 mode, then resets the umask. All the relevant calls to fopen() in the
943 content scanning code have been changed to use this function.
944
945 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
946 to 24 hours. This avoids potential overflow problems when processing G
947 and H retry rules. I suspect nobody ever tinkers with this value.
948
949 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
950
951 PH/45 When the plaintext authenticator is running as a client, the server's
952 challenges are checked to ensure they are valid base64 strings. By
953 default, the authentication attempt is cancelled if an invalid string is
954 received. Setting client_ignore_invalid_base64 true ignores these errors.
955 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
956 they are received. Thus, the responses can be made to depend on the
957 challenges. If an invalid string is ignored, an empty string is placed in
958 the variable.
959
960 PH/46 Messages that are created by the autoreply transport now contains a
961 References: header, in accordance with RFCs 2822 and 3834.
962
963 PH/47 Added authenticated_sender_force to the smtp transport.
964
965 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
966
967 PH/49 Installed latest patch from the Sieve maintainer.
968
969 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
970 was also set, the appendfile transport was unnecessarily scanning a
971 directory of message files (e.g. for maildir delivery) to find the count
972 of files (along with the size), even though it did not need this
973 information. It now does the scan only if it needs to find either the
974 size of the count of files.
975
976 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
977
978 PH/52 Two bugs concerned with error handling when the smtp transport is
979 used in LMTP mode:
980
981 (i) Exim was not creating retry information for temporary errors given
982 for individual recipients after the DATA command when the smtp transport
983 was used in LMTP mode. This meant that they could be retried too
984 frequently, and not timed out correctly.
985
986 (ii) Exim was setting the flag that allows error details to be returned
987 for LMTP errors on RCPT commands, but not for LMTP errors for individual
988 recipients that were returned after the DATA command.
989
990 PH/53 This is related to PH/52, but is more general: for any failing address,
991 when detailed error information was permitted to be returned to the
992 sender, but the error was temporary, then after the final timeout, only
993 "retry timeout exceeded" was returned. Now it returns the full error as
994 well as "retry timeout exceeded".
995
996 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
997 do this, and (what is worse) MTAs that accept it.
998
999 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1000 will now be deprecated.
1001
1002 PH/56 New os.c-cygwin from the Cygwin maintainer.
1003
1004 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1005 formats (previously only available in exim formats via -bpr, -bpru,
1006 and -bpra. Now also available in native and exiqgrep formats)
1007
1008 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1009 with very large, slow to parse queues
1010
1011 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1012
1013 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1014
1015 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1016 responses to authentication challenges, though it was showing the
1017 challenges; (ii) I've removed the CR characters from the debug output for
1018 SMTP output lines.
1019
1020 PH/58 Allow for the insertion of a newline as well as a space when a string
1021 is turned into more than one encoded-word during RFC 2047 encoding. The
1022 Sieve code now uses this.
1023
1024 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1025 data_4xx, lost_connection, tls_required.
1026
1027 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1028 message was being sent as an SMTP response.
1029
1030 PH/61 Add -l and -k options to exicyclog.
1031
1032 PH/62 When verifying, if an address was redirected to one new address, so that
1033 verification continued, and the new address failed or deferred after
1034 having set something in $address_data, the value of $address_data was not
1035 passed back to the ACL. This was different to the case when no
1036 redirection occurred. The value is now passed back in both cases.
1037
1038 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1039 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1040 use login_cap.h, so on its own it isn't the distinguishing feature. The
1041 new name refers directly to the setclassresources() function.
1042
1043 PH/65 Added configuration files for NetBSD3.
1044
1045 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1046
1047 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1048 is preferred over IPv4.
1049
1050 PH/68 The bounce_return_message and bounce_return_body options were not being
1051 honoured for bounces generated during the reception of non-SMTP messages.
1052 In particular, this applied to messages rejected by the ACL. This bug has
1053 been fixed. However, if bounce_return_message is true and bounce_return_
1054 body is false, the headers that are returned for a non-SMTP message
1055 include only those that have been read before the error was detected.
1056 (In the case of an ACL rejection, they have all been read.)
1057
1058 PH/69 The HTML version of the specification is now built in a directory called
1059 spec_html instead of spec.html, because the latter looks like a path with
1060 a MIME-type, and this confuses some software.
1061
1062 PH/70 Catch two compiler warnings in sieve.c.
1063
1064 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1065 function verify_get_ident() calls ip_connect() to connect a socket, but
1066 if the "connect()" function timed out, ip_connect() used to close the
1067 socket. However, verify_get_ident() also closes the socket later, and in
1068 between Exim writes to the log, which may get opened at this point. When
1069 the socket was closed in ip_connect(), the log could get the same file
1070 descriptor number as the socket. This naturally causes chaos. The fix is
1071 not to close the socket in ip_connect(); the socket should be closed by
1072 the function that creates it. There was only one place in the code where
1073 this was missing, in the iplookup router, which I don't think anybody now
1074 uses, but I've fixed it anyway.
1075
1076 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1077 well as to direct DNS lookups. Otherwise the handling of names in host
1078 lists is inconsistent and therefore confusing.
1079
1080
1081 Exim version 4.60
1082 -----------------
1083
1084 PH/01 Two changes to the default runtime configuration:
1085
1086 (1) Move the checks for relay_from_hosts and authenticated clients from
1087 after to before the (commented out) DNS black list checks.
1088
1089 (2) Add control=submission to the relay_from_hosts and authenticated
1090 clients checks, on the grounds that messages accepted by these
1091 statements are most likely to be submissions.
1092
1093 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1094
1095 (1) Generate an error if the third argument for the ${prvs expansion is
1096 not a single digit.
1097
1098 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1099 string.
1100
1101 (3) Reset the variables that are obtained from the first argument of
1102 ${prvscheck and used in the second argument before leaving the code,
1103 because their memory is reclaimed, so using them afterwards may do
1104 silly things.
1105
1106 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1107 one (it's much easier than Tom thought :-).
1108
1109 (5) Because of (4), we can now allow for the use of $prvscheck_result
1110 inside the third argument.
1111
1112 PH/03 For some reason, the default setting of PATH when running a command from
1113 a pipe transport was just "/usr/bin". I have changed it to
1114 "/bin:/usr/bin".
1115
1116 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1117 anything to be listed in the output from -bV.
1118
1119 PH/05 When a filter generated an autoreply, the entire To: header line was
1120 quoted in the delivery log line, like this:
1121
1122 => >A.N.Other <ano@some.domain> <original@ddress> ...
1123
1124 This has been changed so that it extracts the operative address. There
1125 may be more than one such address. If so, they are comma-separated, like
1126 this:
1127
1128 => >ano@some.domain,ona@other.domain <original@ddress> ...
1129
1130 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1131 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1132 not being looked up in the rDNS to get a host name, Exim was showing the
1133 IP address twice in Received: lines, even though the IP addresses were
1134 identical. For example:
1135
1136 Received: from [1.2.3.4] (helo=[1.2.3.4])
1137
1138 However, if the real host name was known, it was omitting the HELO data
1139 if it matched the actual IP address. This has been tidied up so that it
1140 doesn't show the same IP address twice.
1141
1142 PH/07 When both +timestamp and +memory debugging was on, the value given by
1143 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1144 called by the debug printing, thereby overwriting the timestamp buffer.
1145 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1146
1147 PH/08 When the original message was included in an autoreply transport, it
1148 always said "this is a copy of the message, including all the headers",
1149 even if body_only or headers_only was set. It now gives an appropriate
1150 message.
1151
1152 PH/09 Applied a patch from the Sieve maintainer which:
1153
1154 o fixes some comments
1155 o adds the (disabled) notify extension core
1156 o adds some debug output for the result of if/elsif tests
1157 o points to the current vacation draft in the documentation
1158 and documents the missing references header update
1159
1160 and most important:
1161
1162 o fixes a bug in processing the envelope test (when testing
1163 multiple envelope elements, the last element determinted the
1164 result)
1165
1166 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1167 Electronic Mail") by including:
1168
1169 Auto-submitted: auto-generated
1170
1171 in the messages that it generates (bounce messages and others, such as
1172 warnings). In the case of bounce messages for non-SMTP mesages, there was
1173 also a typo: it was using "Auto_submitted" (underscore instead of
1174 hyphen). Since every message generated by Exim is necessarily in response
1175 to another message, thes have all been changed to:
1176
1177 Auto-Submitted: auto-replied
1178
1179 in accordance with these statements in the RFC:
1180
1181 The auto-replied keyword:
1182
1183 - SHOULD be used on messages sent in direct response to another
1184 message by an automatic process,
1185
1186 - MUST NOT be used on manually-generated messages,
1187
1188 - MAY be used on Delivery Status Notifications (DSNs) and Message
1189 Disposition Notifications (MDNs),
1190
1191 - MUST NOT be used on messages generated by automatic or periodic
1192 processes, except for messages which are automatic responses to
1193 other messages.
1194
1195 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1196 to the default Received: header definition.
1197
1198 PH/12 Added log selector acl_warn_skipped (default on).
1199
1200 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1201 variables because (a) they are in the wrong storage pool and (b) even if
1202 they were copied, it wouldn't work properly because of the caching.
1203
1204 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1205 checking when decoding. Apparently there are clients that generate
1206 overlong encoded strings. Why am I not surprised?
1207
1208 PH/15 If the first argument of "${if match_address" was not empty, but did not
1209 contain an "@" character, Exim crashed. Now it writes a panic log message
1210 and treats the condition as false.
1211
1212 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1213
1214 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1215 extension "envelope-auth". The code is finished and in agreement with
1216 other implementations, but there is no documentation so far and in fact,
1217 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1218 not changing the active code.
1219
1220 Print executed "if" and "elsif" statements when debugging is used. This
1221 helps a great deal to understand what a filter does.
1222
1223 Document more things not specified clearly in RFC3028. I had all this
1224 sorted out, when out of a sudden new issues came to my mind. Oops."
1225
1226 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1227 (Bugzilla #53).
1228
1229 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1230 canonical form (as documented). However, after a host name lookup from
1231 the IP address, check_host() was doing a simple string comparison with
1232 addresses acquired from the DNS when checking that the found name did
1233 have the original IP as one of its addresses. Since any found IPv6
1234 addresses are likely to be in abbreviated form, the comparison could
1235 fail. Luckily, there already exists a function for doing the comparison
1236 by converting both addresses to binary, so now that is used instead of
1237 the text comparison.
1238
1239 PH/20 There was another similar case to PH/19, when a complete host name was
1240 given in a host list; looking up its IP address could give an abbreviated
1241 form, whereas the current host's name might or might not be abbreviated.
1242 The same fix has been applied.
1243
1244
1245 Exim version 4.54
1246 -----------------
1247
1248 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1249 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1250 It now does.
1251
1252 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1253 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1254
1255 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1256
1257 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1258 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1259 Also, for "Auto-Submitted", treat anything other than "no" as
1260 non-personal, in accordance with RFC 3834. (Previously it treated
1261 anything starting "auto-" as non-personal.)
1262
1263 TF/01 The control=submission/name=... option had a problem with syntax
1264 errors if the name included a slash character. The /name= option
1265 now slurps the rest of the string, so it can include any characters
1266 but it must come last in the list of options (after /sender_retain
1267 or /domain=).
1268
1269 PH/05 Some modifications to the interface to the fake nameserver for the new
1270 testing suite.
1271
1272
1273
1274 Exim version 4.53
1275 -----------------
1276
1277 TK/01 Added the "success_on_redirect" address verification option. See
1278 NewStuff for rationale and an example.
1279
1280 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1281
1282 PH/02 Patch to exigrep to allow it to work on syslog lines.
1283
1284 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1285 fread() to skip over the body file's header line, because in Cygwin the
1286 header line is locked and is inaccessible.
1287
1288 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1289 co-exist for some time) to make it clear that it is the Exim ID that is
1290 referenced, not the Message-ID: header line.
1291
1292 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1293 string_format() function, because snprintf() does not exist on all
1294 operating systems.
1295
1296 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1297 ${stat: expansion item.
1298
1299 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1300 protocol synchronization error", to keep the pedants happy.
1301
1302 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1303 well as for IRIX systems, when gcc is being used. See the host.c source
1304 file for comments.
1305
1306 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1307
1308 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1309 setting.
1310
1311 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1312 transport and to the smtp transport in LMTP mode.
1313
1314 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1315
1316 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1317 run a filter in a subprocess. This could lead to confusion in subsequent
1318 lookups in the parent process. There should also be a search_tidyup() at
1319 the end of the subprocess.
1320
1321 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1322 only if the host matched helo_try_verify_hosts, which caused the
1323 verification to occur when the EHLO/HELO command was issued. The ACL just
1324 tested the remembered result. Now, if a previous verification attempt has
1325 not happened, "verify = helo" does it there and then.
1326
1327 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1328
1329 TK/03 Fix log output including CR from clamd.
1330
1331 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1332 exist provoked a memory error which could cause a segfault.
1333
1334 PH/15 Installed PCRE 6.2
1335
1336 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1337
1338 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1339 of the problem. Specifically, suggested +O2 rather than +O1 for the
1340 HP-UX compiler.
1341
1342 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1343
1344 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1345 record, the port was not correctly logged when the outgoing_port log
1346 selector was set (it logged the transort's default port).
1347
1348 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1349 fallback_hosts, and "hosts" in the smtp transport.
1350
1351 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1352 host errors such as "Connection refused".
1353
1354 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1355 authentication with radiusclient 0.4.9:
1356
1357 - Error returned from rc_read_config was caught wrongly
1358 - Username/password not passed on to radius server due to wrong length.
1359
1360 The presumption is that some radiusclient API changes for 4.51/PH/17
1361 were not taken care of correctly. The code is still untested by me (my
1362 Linux distribution still has 0.3.2 of radiusclient), but it was
1363 contributed by a Radius user.
1364
1365 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1366 expanding the "port" option of the smtp transport.
1367
1368 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1369 while reading a MIME header. Thanks to Tom Hughes for a patch.
1370
1371 PH/24 Include config.h inside local_scan.h so that configuration settings are
1372 available.
1373
1374 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1375 that in an ACL for RCPT (for example), you can examine exactly what was
1376 received.
1377
1378 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1379 commands, but it was not correctly comparing the address with the actual
1380 client host address. Thus, it would show the EHLO address in Received:
1381 header lines when this was not necessary.
1382
1383 PH/27 Added the % operator to ${eval:}.
1384
1385 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1386 it should be ignoring failures (because with -C, for example, it has lost
1387 privilege). It wasn't ignoring creation failures other than "already
1388 exists".
1389
1390 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1391 -bV or -d is used.
1392
1393 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1394 because an input line was too long, either on its own, or by virtue of
1395 too many continuations, the temporary file was not being removed, and the
1396 return code was incorrect.
1397
1398 PH/31 Missing "BOOL" in function definition in filtertest.c.
1399
1400 PH/32 Applied Sieve patches from the maintainer.
1401
1402 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1403
1404 PH/33 Added "verify = not_blind".
1405
1406 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1407 Local/Makefile (with some defaults set). These are used in built scripts
1408 such as exicyclog, but they have never been used in the exim_install
1409 script (though there are many overriding facilities there). I have
1410 arranged that the exim_install script now takes note of these two
1411 settings.
1412
1413 PH/35 Installed configuration files for Dragonfly.
1414
1415 PH/36 When a locally submitted message by a trusted user did not contain a
1416 From: header, and the sender address was obtained from -f or from an SMTP
1417 MAIL command, and the trusted user did not use -F to supply a sender
1418 name, $originator_name was incorrectly used when constructing a From:
1419 header. Furthermore, $originator_name was used for submission mode
1420 messages from external hosts without From: headers in a similar way,
1421 which is clearly wrong.
1422
1423 PH/37 Added control=suppress_local_fixups.
1424
1425 PH/38 When log_selector = +received_sender was set, and the addition of the
1426 sender made the log line's construction buffer exactly full, or one byte
1427 less than full, an overflow happened when the terminating "\n" was
1428 subsequently added.
1429
1430 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1431 when the result of a list match is failure because a DNS lookup failed.
1432
1433 PH/40 RM_COMMAND is now used in the building process.
1434
1435 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1436 the "build-* directories that it finds.
1437
1438 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1439 address in a domain literal was a prefix of an interface address.
1440
1441 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1442 when verifying a sender address, unless rewrite_headers is false.
1443
1444 PH/44 Wrote a long comment about why errors_to addresses are verified as
1445 recipients, not senders.
1446
1447 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1448 the ratelimit ACL was added.
1449
1450 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1451
1452 PH/46 Added extra information about PostgreSQL errors to the error string.
1453
1454 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1455 avoiding the need to install special zones in a real server. This is
1456 backwards compatible; if it can't find the fake resolver, it drops back.
1457 Thus, both old and new test suites can be run.
1458
1459 TF/02 Added util/ratelimit.pl
1460
1461 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1462 clock is set back in time.
1463
1464 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1465 Candler <B.Candler@pobox.com>.
1466
1467 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1468 OK for addresses that are the result of redirections.
1469
1470 PH/48 A number of further additions for the benefit of the new test suite,
1471 including a fake gethostbyname() that interfaces to the fake DNS resolver
1472 (see PH/47 above).
1473
1474 TF/06 The fix for widen_domains has also been applied to qualify_single and
1475 search_parents which are the other dnslookup options that can cause
1476 header rewrites.
1477
1478 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1479 type ("H").
1480
1481 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1482
1483 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1484 service" if it encountered an MX record with an empty target hostname.
1485 The message is now "an MX or SRV record indicated no SMTP service".
1486
1487 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1488 if the DNS of the sending site is misconfigured. This is quite a
1489 common situation. This change restores the behaviour of treating a
1490 helo verification defer as a failure.
1491
1492 PH/51 If self=fail was set on a router, the bounce message did not include the
1493 actual error message.
1494
1495
1496 Exim version 4.52
1497 -----------------
1498
1499 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1500
1501 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1502 command itself ended in error, the underlying message about the transport
1503 filter timeout was being overwritten with the pipe command error. Now the
1504 underlying error message should be appended to the second error message.
1505
1506 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1507
1508 PH/02 Reduce the amount of output that "make" produces by default. Full output
1509 can still be requested.
1510
1511 PH/03 The warning log line about a condition test deferring for a "warn" verb
1512 was being output only once per connection, rather than after each
1513 occurrence (because it was using the same function as for successful
1514 "warn" verbs). This seems wrong, so I have changed it.
1515
1516 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1517 it should not have, which might have caused a crash in the right
1518 circumstances, but probably never did.
1519
1520 PH/04 Installed a modified version of Tony Finch's patch to make submission
1521 mode fix the return path as well as the Sender: header line, and to
1522 add a /name= option so that you can make the user's friendly name appear
1523 in the header line.
1524
1525 TF/03 Added the control = fakedefer ACL modifier.
1526
1527 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1528 Mark Lowes for thorough testing.
1529
1530 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1531
1532 TK/03 Merged latest SRS patch from Miles Wilton.
1533
1534 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1535 with the definition in sysexits.h (which is #included earlier).
1536 Fortunately, Exim does not actually use EX_OK. The code used to try to
1537 preserve the sysexits.h value, by assumimg that macro definitions were
1538 scanned for macro replacements. I have been disabused of this notion,
1539 so now the code just undefines EX_OK before #including unistd.h.
1540
1541 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1542 in the smtp transport. When a block could not be written in a single
1543 write() function, the timeout was being re-applied to each part-write.
1544 This seems wrong - if the receiver was accepting one byte at a time it
1545 would take for ever. The timeout is now adjusted when this happens. It
1546 doesn't have to be particularly precise.
1547
1548 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1549 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1550
1551 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1552 without a domain if the check to <postmaster@domain> fails.
1553
1554 SC/01 Eximstats: added -xls and the ability to specify output files
1555 (patch written by Frank Heydlauf).
1556
1557 SC/02 Eximstats: use FileHandles for outputing results.
1558
1559 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1560
1561 SC/04 Eximstats: fixed display of large numbers with -nvr option
1562
1563 SC/05 Eximstats: fixed merging of reports with empty tables.
1564
1565 SC/06 Eximstats: added the -include_original_destination flag
1566
1567 SC/07 Eximstats: removed tabs and trailing whitespace.
1568
1569 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1570
1571 TK/06 MBOX spool code: Add real "From " MBOX separator line
1572 so the .eml file is really in mbox format (even though
1573 most programs do not really care). Patch from Alex Miller.
1574
1575 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1576 The latter is generated from $received_to and is only set if the
1577 message has one envelope recipient. SA can use these headers,
1578 obviously out-of-the-box. Patch from Alex Miller.
1579
1580 PH/08 The ${def test on a variable was returning false if the variable's
1581 value was "0", contrary to what the specification has always said!
1582 The result should be true unless the variable is empty.
1583
1584 PH/09 The syntax error of a character other than { following "${if
1585 def:variable_name" (after optional whitespace) was not being diagnosed.
1586 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1587 accidental colon was present, for example, could give incorrect results.
1588
1589 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1590 result is used (not including appendfile, where other changes are about
1591 to be made).
1592
1593 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1594 This involved changing a lot of size variables from int to off_t. It
1595 should work with maildirs and everything.
1596
1597 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1598 spamd dying while we are connected to it.
1599
1600 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1601 <jgh@wizmail.org>
1602
1603 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1604 function for BDB 4.3.
1605
1606 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1607
1608 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1609 the sources for PCRE have been re-arranged and more files are now
1610 involved.
1611
1612 PH/15 The code I had for printing potentially long long variables in PH/11
1613 above was not the best (it lost precision). The length of off_t variables
1614 is now inspected at build time, and an appropriate printing format (%ld
1615 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1616 to be "long long int" or "long int". This is needed for the internal
1617 formatting function string_vformat().
1618
1619 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1620 the configuration file to be ":syslog", then the script "guesses" where
1621 the logs files are, rather than using the compiled in default. In our
1622 case the guess is not the same as the compiled default, so the script
1623 suddenly stopped working when I started to use syslog. The patch checks
1624 to see if log_file_path is "". If so, it attempts to read it from exim
1625 with no configuration file to get the compiled in version, before it
1626 falls back to the previous guessing code."
1627
1628 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1629 implementing BATV in an Exim configuration. See NewStuff for the gory
1630 details.
1631
1632 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1633 Makefile that are specific to HP-UX.
1634
1635 PH/18 If the "use_postmaster" option was set for a recipient callout together
1636 with the "random" option, the postmaster address was used as the MAIL
1637 FROM address for the random test, but not for the subsequent recipient
1638 test. It is now used for both.
1639
1640 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1641 patch removes a few documentation additions to RFC 3028, because the
1642 latest draft now contains them. It adds the new en;ascii-case comparator
1643 and a new error check for 8bit text in MIME parts. Comparator and
1644 require names are now matched exactly. I enabled the subaddress
1645 extension, but it is not well tested yet (read: it works for me)."
1646
1647 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1648 rework some of the code of TK/09 above to avoid the hardwired use of
1649 "%lld" and "long long". Replaced the call to snprintf() with a call to
1650 string_vformat().
1651
1652 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1653 records point to non-existent hosts", "retry timeout exceeded", and
1654 "retry time not reached for any host after a long failure period".
1655
1656 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1657 experimental DomainKeys support:
1658
1659 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1660 (2) On an error such as an illegally used "control", the wrong name for
1661 the control was given.
1662
1663 These problems did NOT occur unless DomainKeys support was compiled.
1664
1665 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1666
1667 PH/24 Added ${if match_ip condition.
1668
1669 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1670 databases so that it will be absolutely obvious if a crash occurs in the
1671 DB library. This is a regular occurrence (often caused by mis-matched
1672 db.h files).
1673
1674 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1675 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1676 picked up on a user's system that detects such things. There doesn't seem
1677 to be a gcc warning option for this - only an attribute that has to be
1678 put on the function's prototype. It seems that in Fedora Core 4 they have
1679 set this on a number of new functions. No doubt there will be more in due
1680 course.
1681
1682 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1683 specify a transport. However, if an address that was verified by such a
1684 router was the subject of a callout, Exim crashed because it tried to
1685 read the rcpt_include_affixes from the non-existent transport. Now it
1686 just assumes that the setting of that option is false. This bug was
1687 introduced by 4.51/PH/31.
1688
1689 PH/28 Changed -d+all to exclude +memory, because that information is very
1690 rarely of interest, but it makes the output a lot bigger. People tend to
1691 do -d+all out of habit.
1692
1693 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1694 code in os-type was giving problems when libc.so lives in lib64, like on
1695 x86_64 Fedora Core.
1696
1697 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1698 aren't the modern standard, and it seems that some systems' include files
1699 don't always have them. Exim was already checking for some of the newer
1700 ones like T_AAAA, and defining it itself. I've added checks for all the
1701 record types that Exim uses.
1702
1703 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1704 not automatically generating a new one, as it is supposed to. This
1705 prevented TLS from working. If the file did exist, but contained invalid
1706 data, a new version was generated, as expected. It was only the case of a
1707 non-existent file that was broken.
1708
1709 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1710 with a change in libdomainkeys > 0.64.
1711
1712 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1713 from DNS. If the selector record carries the flag, it now has
1714 precedence over the domain-wide flag.
1715
1716 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1717
1718 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1719 the use of a port name that isn't defined in /etc/services) occurred, the
1720 message was deferred as in a normal delivery, and thus remained on the
1721 spool, instead of being failed because of the mua_wrapper setting. This
1722 is now fixed, and I tidied up some of the mua_wrapper messages at the
1723 same time.
1724
1725 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1726 the messages in a hash of arrays rather than using individual hashes.
1727 This is a bit cleaner and results in dramatic memory savings, albeit
1728 at a slight CPU cost.
1729
1730 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1731 as requested by Marc Sherman.
1732
1733 SC/10 Eximstats: added histograms for user specified patterns as requested
1734 by Marc Sherman.
1735
1736 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1737
1738 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1739 fopen() in the content-scanning modules that did not already have it.
1740
1741
1742 Exim version 4.51
1743 -----------------
1744
1745 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1746 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1747
1748 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1749
1750 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1751
1752 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1753 to test Sieve filters that use "vacation".
1754
1755 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1756 that changes the way the GnuTLS parameters are stored in the cache file.
1757 The new format can be generated externally. For backward compatibility,
1758 if the data in the cache doesn't make sense, Exim assumes it has read an
1759 old-format file, and it generates new data and writes a new file. This
1760 means that you can't go back to an older release without removing the
1761 file.
1762
1763 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1764 work if there are any delivery delays because "one_time" forces the
1765 parent to be marked "delivered", so its unseen clone is never tried
1766 again. For this reason, Exim now forbids the simultaneous setting of
1767 these two options.
1768
1769 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1770 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1771 entry if you want to know the details. The fix, however, neglected to
1772 consider the case when local delivery batching is involved. The test for
1773 "previously delivered" was not happening when checking to see if an
1774 address could be batched with a previous (undelivered) one; under
1775 certain circumstances this could lead to multiple deliveries to the same
1776 address.
1777
1778 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1779 in its include files, and this causes problems building Exim.
1780
1781 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1782 header_syntax) but Exim was just ignoring anything given after a slash.
1783 In particular, this caused confusion with an attempt to use "verify =
1784 reverse_host_lookup/defer_ok". An error is now given when options are
1785 supplied for verify items that do not have them. (Maybe reverse_host_
1786 lookup should have a defer_ok option, but that's a different point.)
1787
1788 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1789 defined by RFC 821) to 2048, because there were problems with some AUTH
1790 commands, and RFC 1869 says the size should be increased for extended
1791 SMTP commands that take arguments.
1792
1793 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1794 Finch).
1795
1796 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1797 "unknown" error; now it says that the functionality isn't in the binary.
1798
1799 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1800 an address' error message when a string expansion fails (syntax or
1801 whatever). Otherwise the password may appear in the log. Following change
1802 PH/42 below, there is no longer a chance of it appearing in a bounce
1803 message.
1804
1805 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1806
1807 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1808 of its addresses was ever tried. (Bugzilla bug #2.)
1809
1810 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1811 the result incorrectly in the debug output. (It correctly added a newline
1812 to what was transported.)
1813
1814 TF/01 Added $received_time.
1815
1816 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1817 commented out examples of how to interface to a virus scanner and to
1818 SpamAssassin. Also added commented examples of av_scanner and
1819 spamd_address settings.
1820
1821 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1822 and controls are allowed in which ACLs. There were a couple of minor
1823 errors. Some of the entries in the conditions table (which is a table of
1824 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1825 negation of where the condition IS allowed.
1826
1827 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1828
1829 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1830 header file does not have a version number, so I've had to invent a new
1831 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1832 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1833 radiusclient), but it was contributed by a Radius user.
1834
1835 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1836 files or directories, for OpenSSL.
1837
1838 PH/19 When an Exim process that is running as root has to create an Exim log
1839 file, it does so in a subprocess that runs as exim:exim so as to get the
1840 ownership right at creation (otherwise, other Exim processes might see
1841 the file with the wrong ownership). There was no test for failure of this
1842 fork() call, which would lead to the process getting stuck as it waited
1843 for a non-existent subprocess. Forks do occasionally fail when resources
1844 run out. I reviewed all the other calls to fork(); they all seem to check
1845 for failure.
1846
1847 PH/20 When checking for unexpected SMTP input at connect time (before writing
1848 the banner), Exim was not dealing correctly with a non-positive return
1849 from the read() function. If the client had disconnected by this time,
1850 the result was a log entry for a synchronization error with an empty
1851 string after "input=" when read() returned zero. If read() returned -1
1852 (an event I could not check), uninitialized data bytes were printed.
1853 There were reports of junk text (parts of files, etc) appearing after
1854 "input=".
1855
1856 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1857
1858 PH/22 Added support for macro redefinition, and (re)definition in between
1859 driver and ACL definitions.
1860
1861 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1862 forgetting to use the resulting value; it was using the unexpanded value.
1863
1864 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1865 hadn't been configured. The fix is from Juergen Kreileder, who
1866 understands it better than I do:
1867
1868 "Here's what I see happening with three configured cyrus_sasl
1869 authenticators configured (plain, login, cram-md5):
1870
1871 On startup auth_cyrus_sasl_init() gets called for each of these.
1872 This means three calls to sasl_listmech() without a specified mech_list.
1873 => SASL tests which mechs of all available mechs actually work
1874 => three warnings about OTP not working
1875 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1876
1877 With the patch, sasl_listmech() also gets called three times. But now
1878 SASL's mech_list option is set to the server_mech specified in the the
1879 authenticator. Or in other words, the answer from sasl_listmech()
1880 gets limited to just the mech you're testing for (which is different
1881 for each call.)
1882 => the return list contains just 'plain' or 'login', 'cram-md5' or
1883 nothing depending on the value of ob->server_mech.
1884
1885 I've just tested the patch: Authentication still works fine,
1886 unavailable mechs specified in the exim configuration are still
1887 caught, and the auth.log warnings about OTP are gone."
1888
1889 PH/25 When debugging is enabled, the contents of the command line are added
1890 to the debugging output, even when log_selector=+arguments is not
1891 specified.
1892
1893 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1894 answer is "GNU", and only if the return is "GNU/something" is the answer
1895 "Linux".
1896
1897 PH/27 $acl_verify_message is now set immediately after the failure of a
1898 verification in an ACL, and so is available in subsequent modifiers. In
1899 particular, the message can be preserved by coding like this:
1900
1901 warn !verify = sender
1902 set acl_m0 = $acl_verify_message
1903
1904 Previously, $acl_verify_message was set only while expanding "message"
1905 and "log_message" when a very denied access.
1906
1907 PH/28 Modified OS/os.c-Linux with
1908
1909 -#ifndef OS_LOAD_AVERAGE
1910 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1911
1912 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1913 nomenclature these days.)
1914
1915 PH/29 Installed patch from the Sieve maintainer that adds the options
1916 sieve_useraddress and sieve_subaddress to the redirect router.
1917
1918 PH/30 In these circumstances:
1919 . Two addresses routed to the same list of hosts;
1920 . First host does not offer TLS;
1921 . First host accepts first address;
1922 . First host gives temporary error to second address;
1923 . Second host offers TLS and a TLS session is established;
1924 . Second host accepts second address.
1925 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1926 and peerdn, if requested) that were in fact used only for the second
1927 address.
1928
1929 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1930 attention to any local part prefix or suffix that was matched by the
1931 router that accepted the address. It now behaves in the same way as it
1932 does for delivery: the affixes are removed from the local part unless
1933 rcpt_include_affixes is set on the transport.
1934
1935 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1936 timeout during the DATA phase of an incoming message.
1937
1938 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1939 applied a patch sanctioned by the Sieve maintainer.
1940
1941 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1942 the uid or gid is negative. A case of a negative gid caused this to be
1943 noticed. The fix allows for either to be negative.
1944
1945 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1946 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1947 overlooked.
1948
1949 PH/36 The change PH/12 above was broken. Fixed it.
1950
1951 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1952 the grounds that routing the same address twice would always produce the
1953 same answer. This might have been true once, but it is certainly no
1954 longer true now. Routing a child address may depend on the previous
1955 routing that produced that child. Some complicated redirection strategies
1956 went wrong when messages had multiple recipients, and made Exim's
1957 behaviour dependent on the order in which the addresses were given.
1958
1959 I have moved the duplicate checking until after the routing is complete.
1960 Exim scans the addresses that are assigned to local and remote
1961 transports, and removes any duplicates. This means that more work will be
1962 done, as duplicates will always all be routed, but duplicates are
1963 presumably rare, so I don't expect this is of any significance.
1964
1965 For deliveries to pipes, files, and autoreplies, the duplicate checking
1966 still happens during the routing process, since they are not going to be
1967 routed further.
1968
1969 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1970 It corrects a timeout issue with spamd. This is Ian's comment: "The
1971 background is that sometimes spamd either never reads data from a
1972 connection it has accepted, or it never writes response data. The exiscan
1973 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1974 blindly assumes that writes won't block so it may never time out."
1975
1976 PH/39 Allow G after quota size as well as K and M.
1977
1978 PH/40 The value set for $authenticated_id in an authenticator may not contain
1979 binary zeroes or newlines because the value is written to log lines and
1980 to spool files. There was no check on this. Now the value is run through
1981 the string_printing() function so that such characters are converted to
1982 printable escape sequences.
1983
1984 PH/41 $message_linecount is a new variable that contains the total number of
1985 lines in the message. Compare $body_linecount, which is the count for the
1986 body only.
1987
1988 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1989 bounce and delay warning messages, except in certain special cases, which
1990 are as follows:
1991
1992 (a) An SMTP error message from a remote host;
1993 (b) A message specified in a :fail: redirection;
1994 (c) A message specified in a "fail" command in a system filter;
1995 (d) A message specified in a FAIL return from the queryprogram router;
1996 (e) A message specified by the cannot_route_message router option.
1997
1998 In these cases only, Exim does include the error details in bounce and
1999 warning messages. There are also a few cases where bland messages such
2000 as "unrouteable address" or "local delivery error" are given.
2001
2002 PH/43 $value is now also set for the "else" part of a ${run expansion.
2003
2004 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2005 being worked on, but at least Exim now implements the latest version to
2006 play with."
2007
2008 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2009 process to complete was treated as a delivery failure, a timeout while
2010 writing the message to the pipe was logged, but erroneously treated as a
2011 successful delivery. Such timeouts include transport filter timeouts. For
2012 consistency with the overall process timeout, these timeouts are now
2013 treated as errors, giving rise to delivery failures by default. However,
2014 there is now a new Boolean option for the pipe transport called
2015 timeout_defer, which, if set TRUE, converts the failures into defers for
2016 both kinds of timeout. A transport filter timeout is now identified in
2017 the log output.
2018
2019 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2020 systems where "make" and "gmake" are different, calling "gmake" at top
2021 level broke things. I've arranged for the value of $(MAKE) to be passed
2022 from the Makefile to this script so that it can call the same version of
2023 "make".
2024
2025
2026 A note about Exim versions 4.44 and 4.50
2027 ----------------------------------------
2028
2029 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2030 changes of various kinds. As a consequence, a big documentation update was
2031 needed. This delayed the release for rather longer than seemed good, especially
2032 in the light of a couple of (minor) security issues. Therefore, the changes
2033 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2034 release. So 4.44 and 4.50 are in effect two different branches that both start
2035 from 4.43.
2036
2037 I have left the 4.50 change log unchanged; it contains all the changes since
2038 4.43. The change log for 4.44 is below; many of its items are identical to
2039 those for 4.50. This seems to be the most sensible way to preserve the
2040 historical information.
2041
2042
2043 Exim version 4.50
2044 -----------------
2045
2046 1. Minor wording change to the doc/README.SIEVE file.
2047
2048 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2049 computation of the current number of files was incorrect.
2050
2051 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2052 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2053 place.
2054
2055 4. Give more explanation in the error message when the command for a transport
2056 filter fails to execute.
2057
2058 5. There are several places where Exim runs a non-Exim command in a
2059 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2060 was being done only for the command run by the queryprogram router. It is
2061 now done for all such subprocesses. The other cases are: ${run, transport
2062 filters, and the commands run by the lmtp and pipe transports.
2063
2064 6. Added CONFIGURE_GROUP build-time option.
2065
2066 7. Some older OS have a limit of 256 on the maximum number of file
2067 descriptors. Exim was using setrlimit() to set 1000 as a large value
2068 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2069 systems. I've change it so that if it can't get 1000, it tries for 256.
2070
2071 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2072 was an oversight, and furthermore, ever since the addition of extra
2073 controls (e.g. 4.43/32), the checks on when to allow different forms of
2074 "control" were broken. There should now be diagnostics for all cases when a
2075 control that does not make sense is encountered.
2076
2077 9. Added the /retain_sender option to "control=submission".
2078
2079 10. $recipients is now available in the predata ACL (oversight).
2080
2081 11. Tidy the search cache before the fork to do a delivery from a message
2082 received from the command line. Otherwise the child will trigger a lookup
2083 failure and thereby defer the delivery if it tries to use (for example) a
2084 cached ldap connection that the parent has called unbind on.
2085
2086 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2087 of $address_data from the recipient verification was clobbered by the
2088 sender verification.
2089
2090 13. The value of address_data from a sender verification is now available in
2091 $sender_address_data in subsequent conditions in the ACL statement.
2092
2093 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2094
2095 15. Added a new option "connect=<time>" to callout options, to set a different
2096 connection timeout.
2097
2098 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2099 was its contents. (It was OK if the option was not defined at all.)
2100
2101 17. A "Completed" log line is now written for messages that are removed from
2102 the spool by the -Mrm option.
2103
2104 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2105 information about exactly what failed.
2106
2107 19. Added -dd to debug only the daemon process.
2108
2109 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2110 handles timeouts, both on the server side and network timeouts. Renamed the
2111 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2112
2113 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2114 It is now set to "smtps".
2115
2116 22. $host_address is now set to the target address during the checking of
2117 ignore_target_hosts.
2118
2119 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2120 being passed; this would have caused $sender_host_name to have been used if
2121 matching the list had actually called for a host name (not very likely,
2122 since this list is usually IP addresses). A host name is now passed as
2123 "[x.x.x.x]".
2124
2125 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2126 code that specifies a non-restarting handler (typically sigaction() in
2127 modern systems) in an attempt to fix a rare and obscure crash bug.
2128
2129 25. Narrowed the window for a race in the daemon that could cause it to ignore
2130 SIGCHLD signals. This is not a major problem, because they are used only to
2131 wake it up if nothing else does.
2132
2133 26. A malformed maildirsize file could cause Exim to calculate negative values
2134 for the mailbox size or file count. Odd effects could occur as a result.
2135 The maildirsize information is now recalculated if the size or filecount
2136 end up negative.
2137
2138 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2139 support for a long time. Removed HAVE_SYS_VFS_H.
2140
2141 28. Installed the latest version of exipick from John Jetmore.
2142
2143 29. In an address list, if the pattern was not a regular expression, an empty
2144 subject address (from a bounce message) matched only if the pattern was an
2145 empty string. Non-empty patterns were not even tested. This was the wrong
2146 because it is perfectly reasonable to use an empty address as part of a
2147 database query. An empty address is now tested by patterns that are
2148 lookups. However, all the other forms of pattern expect the subject to
2149 contain a local part and a domain, and therefore, for them, an empty
2150 address still always fails if the pattern is not itself empty.
2151
2152 30. Exim went into a mad DNS loop when attempting to do a callout where the
2153 host was specified on an smtp transport, and looking it up yielded more
2154 than one IP address.
2155
2156 31. Re-factored the code for checking spool and log partition space into a
2157 function that finds that data and another that does the check. The former
2158 is then used to implement four new variables: $spool_space, $log_space,
2159 $spool_inodes, and $log_inodes.
2160
2161 32. The RFC2047 encoding function was originally intended for short strings
2162 such as real names; it was not keeping to the 75-character limit for
2163 encoded words that the RFC imposes. It now respects the limit, and
2164 generates multiple encoded words if necessary. To be on the safe side, I
2165 have increased the buffer size for the ${rfc2047: expansion operator from
2166 1024 to 2048 bytes.
2167
2168 33. It is now permitted to omit both strings after an "if" condition; if the
2169 condition is true, the result is "true". As before, when the second string
2170 is omitted, a false condition yields an empty string. This makes it less
2171 cumbersome to write custom ACL and router conditions.
2172
2173 34. Failure to deliver a bounce message always caused it to be frozen, even if
2174 there was an errors_to setting on the router. The errors_to setting is now
2175 respected.
2176
2177 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2178 canonical form (fully expanded) before being placed in
2179 $sender_host_address.
2180
2181 36. The table in the code that translates DNS record types into text (T_A to
2182 "A" for instance) was missing entries for NS and CNAME. It is just possible
2183 that this could have caused confusion if both these types were looked up
2184 for the same domain, because the text type is used as part of Exim's
2185 per-process caching. But the chance of anyone hitting this buglet seems
2186 very small.
2187
2188 37. The dnsdb lookup has been extended in a number of ways.
2189
2190 (1) There is a new type, "zns", which walks up the domain tree until it
2191 finds some nameserver records. It should be used with care.
2192
2193 (2) There is a new type, "mxh", which is like "mx" except that it returns
2194 just the host names, not the priorities.
2195
2196 (3) It is now possible to give a list of domains (or IP addresses) to be
2197 looked up. The behaviour when one of the lookups defers can be
2198 controlled by a keyword.
2199
2200 (4) It is now possible to specify the separator character for use when
2201 multiple records are returned.
2202
2203 38. The dnslists ACL condition has been extended: it is now possible to supply
2204 a list of IP addresses and/or domains to be looked up in a particular DNS
2205 domain.
2206
2207 39. Added log_selector=+queue_time_overall.
2208
2209 40. When running the queue in the test harness, wait just a tad after forking a
2210 delivery process, to get repeatability of debugging output.
2211
2212 41. Include certificate and key file names in error message when GnuTLS fails
2213 to set them up, because the GnuTLS error message doesn't include the name
2214 of the failing file when there is a problem reading it.
2215
2216 42. Allow both -bf and -bF in the same test run.
2217
2218 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2219
2220 44. The "Exiscan patch" is now merged into the mainline Exim source.
2221
2222 45. Sometimes the final signoff response after QUIT could fail to get
2223 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2224 < 0 before doing a fflush(). This bug looks as though it goes back to the
2225 introduction of TLS in release 3.20, but "sometimes" must have been rare
2226 because the tests only now provoked it.
2227
2228 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2229 (this can affect the format of dates).
2230
2231 47. exim_tidydb, when checking for the continued existence of a message for
2232 which it has found a message-specific retry record, was not finding
2233 messages that were in split spool directories. Consequently, it was
2234 deleting retry records that should have stayed in existence.
2235
2236 48. Steve fixed some bugs in eximstats.
2237
2238 49. The SPA authentication driver was not abandoning authentication and moving
2239 on to the next authenticator when an expansion was forced to fail,
2240 contradicting the general specification for all authenticators. Instead it
2241 was generating a temporary error. It now behaves as specified.
2242
2243 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2244 (the order specifies the preference for clients). The order is now AES256,
2245 AES128, 3DES, ARCFOUR128.
2246
2247 51. Small patch to Sieve code - explicitly set From: when generating an
2248 autoreply.
2249
2250 52. Exim crashed if a remote delivery caused a very long error message to be
2251 recorded - for instance if somebody sent an entire SpamAssassin report back
2252 as a large number of 550 error lines. This bug was coincidentally fixed by
2253 increasing the size of one of Exim's internal buffers (big_buffer) that
2254 happened as part of the Exiscan merge. However, to be on the safe side, I
2255 have made the code more robust (and fixed the comments that describe what
2256 is going on).
2257
2258 53. Now that there can be additional text after "Completed" in log lines (if
2259 the queue_time_overall log selector is set), a one-byte patch to exigrep
2260 was needed to allow it to recognize "Completed" as not the last thing in
2261 the line.
2262
2263 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2264 patch that reportedly fixes this has been added. I am not expert enough to
2265 create a test for it. This is what the patch creator wrote:
2266
2267 "I found a little strange behaviour of ldap code when working with
2268 Windows 2003 AD Domain, where users was placed in more than one
2269 Organization Units. When I tried to give exim partial DN, the exit code
2270 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2271 But simultaneously result of request was absolutely normal ldap result,
2272 so I produce this patch..."
2273
2274 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2275 REFERENCE, so I have modified the code to exclude the patch when that macro
2276 is not defined.
2277
2278 55. Some experimental protocols are using DNS PTR records for new purposes. The
2279 keys for these records are domain names, not reversed IP addresses. The
2280 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2281 leaves it alone. Component reversal etc. now happens only for IP addresses.
2282 CAN-2005-0021
2283
2284 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2285
2286 57. Double the size of the debug message buffer (to 2048) so that more of very
2287 long debug lines gets shown.
2288
2289 58. The exicyclog utility now does better if the number of log files to keep
2290 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2291
2292 59. Two changes related to the smtp_active_hostname option:
2293
2294 (1) $smtp_active_hostname is now available as a variable.
2295 (2) The default for smtp_banner uses $smtp_active_hostname instead
2296 of $primary_hostname.
2297
2298 60. The host_aton() function is supposed to be passed a string that is known
2299 to be a valid IP address. However, in the case of IPv6 addresses, it was
2300 not checking this. This is a hostage to fortune. Exim now panics and dies
2301 if the condition is not met. A case was found where this could be provoked
2302 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2303 components; fortuitously, this particular loophole had already been fixed
2304 by change 4.50/55 above.
2305
2306 If there are any other similar loopholes, the new check in host_aton()
2307 itself should stop them being exploited. The report I received stated that
2308 data on the command line could provoke the exploit when Exim was running as
2309 exim, but did not say which command line option was involved. All I could
2310 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2311 running as the user.
2312 CAN-2005-0021
2313
2314 61. There was a buffer overflow vulnerability in the SPA authentication code
2315 (which came originally from the Samba project). I have added a test to the
2316 spa_base64_to_bits() function which I hope fixes it.
2317 CAN-2005-0022
2318
2319 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2320 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2321
2322 63. The daemon start-up calls getloadavg() while still root for those OS that
2323 need the first call to be done as root, but it missed one case: when
2324 deliver_queue_load_max is set with deliver_drop_privilege. This is
2325 necessary for the benefit of the queue runner, because there is no re-exec
2326 when deliver_drop_privilege is set.
2327
2328 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2329 This has been fixed.
2330
2331 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2332 was in use, was not putting the data itself into the right store pool;
2333 consequently, it could be overwritten for a subsequent message in the same
2334 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2335 the caching.)
2336
2337 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2338
2339 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2340 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2341 the function were treating the return as a boolean value, which happened to
2342 work because 0=false and not-0=true, but is not correct code.
2343
2344 68. The host_aton() function was not handling scoped IPv6 addresses (those
2345 with, for example, "%eth0" on the end) correctly.
2346
2347 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2348 negated items (that is, ~something) in unsigned ints. Some compilers
2349 apparently mutter when there is no cast.
2350
2351 70. If an address verification called from an ACL failed, and did not produce a
2352 user-specific message (i.e. there was only a "system" message), nothing was
2353 put in $acl_verify_message. In this situation, it now puts the system
2354 message there.
2355
2356 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2357 session; change 4.31/43 added the unwanted input to the log line - except
2358 that it did not do this in the start of session case. It now does.
2359
2360 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2361 This is wrong and can cause the other end to generate a synchronization
2362 error if it is another Exim or anything else that does the synchronization
2363 check. A QUIT command is no longer sent after a timeout.
2364
2365 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2366 during host lookups.
2367
2368 74. The defer_ok option of callout verification was not working if it was used
2369 when verifying addresses in header lines, that is, for this case:
2370
2371 verify = header_sender/callout=defer_ok
2372
2373 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2374 those file descriptors could be used for SMTP connections. If anything
2375 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2376 could be sent to the SMTP client, causing chaos. The daemon now opens
2377 stdin, stdout, and stderr to /dev/null when it puts itself into the
2378 background.
2379
2380 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2381 log by default. The user can override this with suitable Perl magic.
2382
2383 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2384 the log message when discard triggers, was not working for the DATA ACL or
2385 for the non-SMTP ACL.
2386
2387 78. Error message wording change in sieve.c.
2388
2389 79. If smtp_accept_max_per_host was set, the number of connections could be
2390 restricted to fewer than expected, because the daemon was trying to set up
2391 a new connection before checking whether the processes handling previous
2392 connections had finished. The check for completed processes is now done
2393 earlier. On busy systems, this bug wouldn't be noticed because something
2394 else would have woken the daemon, and it would have reaped the completed
2395 process earlier.
2396
2397 80. If a message was submitted locally by a user whose login name contained one
2398 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2399 It caused a spool format error. I have fixed the spool reading code. A
2400 related problem was that the "from" clause in the Received: line became
2401 illegal because of the space(s). It is now covered by ${quote_local_part.
2402
2403 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2404 tables).
2405
2406 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2407
2408 83. Patch from Sieve maintainer to fix unterminated string problem in
2409 "vacation" handling.
2410
2411 84. Some minor changes to the Linux configuration files to help with other
2412 OS variants using glibc.
2413
2414 85. One more patch for Sieve to update vacation handling to latest spec.
2415
2416
2417 ----------------------------------------------------
2418 See the note above about the 4.44 and 4.50 releases.
2419 ----------------------------------------------------
2420
2421
2422 Exim version 4.44
2423 -----------------
2424
2425 1. Change 4.43/35 introduced a bug that caused file counts to be
2426 incorrectly computed when quota_filecount was set in an appendfile
2427 transport
2428
2429 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2430 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2431 place.
2432
2433 3. Give more explanation in the error message when the command for a transport
2434 filter fails to execute.
2435
2436 4. There are several places where Exim runs a non-Exim command in a
2437 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2438 was being done only for the command run by the queryprogram router. It is
2439 now done for all such subprocesses. The other cases are: ${run, transport
2440 filters, and the commands run by the lmtp and pipe transports.
2441
2442 5. Some older OS have a limit of 256 on the maximum number of file
2443 descriptors. Exim was using setrlimit() to set 1000 as a large value
2444 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2445 systems. I've change it so that if it can't get 1000, it tries for 256.
2446
2447 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2448 was an oversight, and furthermore, ever since the addition of extra
2449 controls (e.g. 4.43/32), the checks on when to allow different forms of
2450 "control" were broken. There should now be diagnostics for all cases when a
2451 control that does not make sense is encountered.
2452
2453 7. $recipients is now available in the predata ACL (oversight).
2454
2455 8. Tidy the search cache before the fork to do a delivery from a message
2456 received from the command line. Otherwise the child will trigger a lookup
2457 failure and thereby defer the delivery if it tries to use (for example) a
2458 cached ldap connection that the parent has called unbind on.
2459
2460 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2461 of $address_data from the recipient verification was clobbered by the
2462 sender verification.
2463
2464 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2465 was its contents. (It was OK if the option was not defined at all.)
2466
2467 11. A "Completed" log line is now written for messages that are removed from
2468 the spool by the -Mrm option.
2469
2470 12. $host_address is now set to the target address during the checking of
2471 ignore_target_hosts.
2472
2473 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2474 being passed; this would have caused $sender_host_name to have been used if
2475 matching the list had actually called for a host name (not very likely,
2476 since this list is usually IP addresses). A host name is now passed as
2477 "[x.x.x.x]".
2478
2479 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2480 code that specifies a non-restarting handler (typically sigaction() in
2481 modern systems) in an attempt to fix a rare and obscure crash bug.
2482
2483 15. Narrowed the window for a race in the daemon that could cause it to ignore
2484 SIGCHLD signals. This is not a major problem, because they are used only to
2485 wake it up if nothing else does.
2486
2487 16. A malformed maildirsize file could cause Exim to calculate negative values
2488 for the mailbox size or file count. Odd effects could occur as a result.
2489 The maildirsize information is now recalculated if the size or filecount
2490 end up negative.
2491
2492 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2493 support for a long time. Removed HAVE_SYS_VFS_H.
2494
2495 18. Updated exipick to current release from John Jetmore.
2496
2497 19. Allow an empty sender to be matched against a lookup in an address list.
2498 Previously the only cases considered were a regular expression, or an
2499 empty pattern.
2500
2501 20. Exim went into a mad DNS lookup loop when doing a callout where the
2502 host was specified on the transport, if the DNS lookup yielded more than
2503 one IP address.
2504
2505 21. The RFC2047 encoding function was originally intended for short strings
2506 such as real names; it was not keeping to the 75-character limit for
2507 encoded words that the RFC imposes. It now respects the limit, and
2508 generates multiple encoded words if necessary. To be on the safe side, I
2509 have increased the buffer size for the ${rfc2047: expansion operator from
2510 1024 to 2048 bytes.
2511
2512 22. Failure to deliver a bounce message always caused it to be frozen, even if
2513 there was an errors_to setting on the router. The errors_to setting is now
2514 respected.
2515
2516 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2517 canonical form (fully expanded) before being placed in
2518 $sender_host_address.
2519
2520 24. Updated eximstats to version 1.33
2521
2522 25. Include certificate and key file names in error message when GnuTLS fails
2523 to set them up, because the GnuTLS error message doesn't include the name
2524 of the failing file when there is a problem reading it.
2525
2526 26. Expand error message when OpenSSL has problems setting up cert/key files.
2527 As per change 25.
2528
2529 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2530 (this can affect the format of dates).
2531
2532 28. exim_tidydb, when checking for the continued existence of a message for
2533 which it has found a message-specific retry record, was not finding
2534 messages that were in split spool directories. Consequently, it was
2535 deleting retry records that should have stayed in existence.
2536
2537 29. eximstats updated to version 1.35
2538 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2539 1.35 - bugfix such that pie charts by volume are generated correctly
2540
2541 30. The SPA authentication driver was not abandoning authentication and moving
2542 on to the next authenticator when an expansion was forced to fail,
2543 contradicting the general specification for all authenticators. Instead it
2544 was generating a temporary error. It now behaves as specified.
2545
2546 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2547 (the order specifies the preference for clients). The order is now AES256,
2548 AES128, 3DES, ARCFOUR128.
2549
2550 31. Small patch to Sieve code - explicitly set From: when generating an
2551 autoreply.
2552
2553 32. Exim crashed if a remote delivery caused a very long error message to be
2554 recorded - for instance if somebody sent an entire SpamAssassin report back
2555 as a large number of 550 error lines. This bug was coincidentally fixed by
2556 increasing the size of one of Exim's internal buffers (big_buffer) that
2557 happened as part of the Exiscan merge. However, to be on the safe side, I
2558 have made the code more robust (and fixed the comments that describe what
2559 is going on).
2560
2561 33. Some experimental protocols are using DNS PTR records for new purposes. The
2562 keys for these records are domain names, not reversed IP addresses. The
2563 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2564 leaves it alone. Component reversal etc. now happens only for IP addresses.
2565 CAN-2005-0021
2566
2567 34. The host_aton() function is supposed to be passed a string that is known
2568 to be a valid IP address. However, in the case of IPv6 addresses, it was
2569 not checking this. This is a hostage to fortune. Exim now panics and dies
2570 if the condition is not met. A case was found where this could be provoked
2571 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2572 components; fortuitously, this particular loophole had already been fixed
2573 by change 4.50/55 or 4.44/33 above.
2574
2575 If there are any other similar loopholes, the new check in host_aton()
2576 itself should stop them being exploited. The report I received stated that
2577 data on the command line could provoke the exploit when Exim was running as
2578 exim, but did not say which command line option was involved. All I could
2579 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2580 running as the user.
2581 CAN-2005-0021
2582
2583 35. There was a buffer overflow vulnerability in the SPA authentication code
2584 (which came originally from the Samba project). I have added a test to the
2585 spa_base64_to_bits() function which I hope fixes it.
2586 CAN-2005-0022
2587
2588 36. The daemon start-up calls getloadavg() while still root for those OS that
2589 need the first call to be done as root, but it missed one case: when
2590 deliver_queue_load_max is set with deliver_drop_privilege. This is
2591 necessary for the benefit of the queue runner, because there is no re-exec
2592 when deliver_drop_privilege is set.
2593
2594 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2595 was in use, was not putting the data itself into the right store pool;
2596 consequently, it could be overwritten for a subsequent message in the same
2597 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2598 the caching.)
2599
2600 38. Sometimes the final signoff response after QUIT could fail to get
2601 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2602 < 0 before doing a fflush(). This bug looks as though it goes back to the
2603 introduction of TLS in release 3.20, but "sometimes" must have been rare
2604 because the tests only now provoked it.
2605
2606
2607 Exim version 4.43
2608 -----------------
2609
2610 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2611 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2612 Now it returns an integer. A number of calls were still expecting a T/F
2613 return. Fortuitously, in all cases, the tests worked in OK situations,
2614 which is the norm. However, things would have gone wrong on any write
2615 failures on the smtp file descriptor. This function is used when sending
2616 messages over SMTP and also when doing verify callouts.
2617
2618 2. When Exim is called to do synchronous delivery of a locally submitted
2619 message (the -odf or -odi options), it no longer closes stderr before doing
2620 the delivery.
2621
2622 3. Implemented the mua_wrapper option.
2623
2624 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2625
2626 5. Implemented the functions header_remove(), header_testname(),
2627 header_add_at_position(), and receive_remove_recipient(), and exported them
2628 to local_scan().
2629
2630 6. If an ACL "warn" statement specified the addition of headers, Exim already
2631 inserted X-ACL-Warn: at the start if there was no header name. However, it
2632 was not making this test for the second and subsequent header lines if
2633 there were newlines in the string. This meant that an invalid header could
2634 be inserted if Exim was badly configured.
2635
2636 7. Allow an ACL "warn" statement to add header lines at the start or after all
2637 the Received: headers, as well as at the end.
2638
2639 8. Added the rcpt_4xx retry error code.
2640
2641 9. Added postmaster_mailfrom=xxx to callout verification option.
2642
2643 10. Added mailfrom=xxxx to the callout verification option, for verify=
2644 header_sender only.
2645
2646 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2647 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2648
2649 12. Inserted some casts to stop certain compilers warning when using pointer
2650 differences as field lengths or precisions in printf-type calls (mostly
2651 affecting debugging statements).
2652
2653 13. Added optional readline() support for -be (dynamically loaded).
2654
2655 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2656 same clock tick as a message's arrival, so that its received time was the
2657 same as the "first fail" time on the retry record, and that message
2658 remained on the queue past the ultimate address timeout, every queue runner
2659 would try a delivery (because it was past the ultimate address timeout) but
2660 after another failure, the ultimate address timeout, which should have then
2661 bounced the address, did not kick in. This was a "< instead of <=" error;
2662 in most cases the first failure would have been in the next clock tick
2663 after the received time, and all would be well.
2664
2665 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2666 being recognized when the domain list was tested by the match_domain
2667 condition in an expansion string.
2668
2669 16. Added the ${str2b64: operator.
2670
2671 17. Exim was always calling setrlimit() to set a large limit for the number of
2672 processes, without checking whether the existing limit was already
2673 adequate. (It did check for the limit on file descriptors.) Furthermore,
2674 errors from getrlimit() and setrlimit() were being ignored. Now they are
2675 logged to the main and panic logs, but Exim does carry on, to try to do its
2676 job under whatever limits there are.
2677
2678 18. Imported PCRE 5.0.
2679
2680 19. Trivial typo in log message " temporarily refused connection" (the leading
2681 space).
2682
2683 20. If the log selector return_path_on_delivery was set and an address was
2684 redirected to /dev/null, the delivery process crashed because it assumed
2685 that a return path would always be set for a "successful" delivery. In this
2686 case, the whole delivery is bypassed as an optimization, and therefore no
2687 return path is set.
2688
2689 21. Internal re-arrangement: the function for sending a challenge and reading
2690 a response while authentication was assuming a zero-terminated challenge
2691 string. It's now changed to take a pointer and a length, to allow for
2692 binary data in such strings.
2693
2694 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2695
2696 23. Exim was not respecting finduser_retries when seeking the login of the
2697 uid under which it was called; it was always trying 10 times. (The default
2698 setting of finduser_retries is zero.) Also, it was sleeping after the final
2699 failure, which is pointless.
2700
2701 24. Implemented tls_on_connect_ports.
2702
2703 25. Implemented acl_smtp_predata.
2704
2705 26. If the domain in control=submission is set empty, Exim assumes that the
2706 authenticated id is a complete email address when it generates From: or
2707 Sender: header lines.
2708
2709 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2710 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2711 chown and chgrp in /bin and hostname in /usr/bin.
2712
2713 28. Exim was keeping the "process log" file open after each use, just as it
2714 does for the main log. This opens the possibility of it remaining open for
2715 long periods when the USR1 signal hits a daemon. Occasional processlog
2716 errors were reported, that could have been caused by this. Anyway, it seems
2717 much more sensible not to leave this file open at all, so that is what now
2718 happens.
2719
2720 29. The long-running daemon process does not normally write to the log once it
2721 has entered its main loop, and it closes the log before doing so. This is
2722 so that log files can straightforwardly be renamed and moved. However,
2723 there are a couple of unusual error situations where the daemon does write
2724 log entries, and I had neglected to close the log afterwards.
2725
2726 30. The text of an SMTP error response that was received during a remote
2727 delivery was being truncated at 512 bytes. This is too short for some of
2728 the long messages that one sometimes sees. I've increased the limit to
2729 1024.
2730
2731 31. It is now possible to make retry rules that apply only when a message has a
2732 specific sender, in particular, an empty sender.
2733
2734 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2735 it possible to be selective about when SMTP synchronization is enforced.
2736
2737 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2738
2739 32. Implemented hosts_connection_nolog.
2740
2741 33. Added an ACL for QUIT.
2742
2743 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2744 syntax error.
2745
2746 35. Added mailbox_size and mailbox_filecount to appendfile.
2747
2748 36. Added control = no_multiline_responses to ACLs.
2749
2750 37. There was a bug in the logic of the code that waits for the clock to tick
2751 in the case where the clock went backwards by a substantial amount such
2752 that the microsecond fraction of "now" was more than the microsecond
2753 fraction of "then" (but the whole seconds number was less).
2754
2755 38. Added support for the libradius Radius client library this is found on
2756 FreeBSD (previously only the radiusclient library was supported).
2757
2758
2759 Exim version 4.42
2760 -----------------
2761
2762 1. When certain lookups returned multiple values in the form name=value, the
2763 quoting of the values was not always being done properly. Specifically:
2764 (a) If the value started with a double quote, but contained no whitespace,
2765 it was not quoted.
2766 (b) If the value contained whitespace other than a space character (i.e.
2767 tabs or newlines or carriage returns) it was not quoted.
2768 This fix has been applied to the mysql and pgsql lookups by writing a
2769 separate quoting function and calling it from the lookup code. The fix
2770 should probably also be applied to nisplus, ibase and oracle lookups, but
2771 since I cannot test any of those, I have not disturbed their existing code.
2772
2773 2. A hit in the callout cache for a specific address caused a log line with no
2774 reason for rejecting RCPT. Now it says "Previous (cached) callout
2775 verification failure".
2776
2777 3. There was an off-by-one bug in the queryprogram router. An over-long
2778 return line was truncated at 256 instead of 255 characters, thereby
2779 overflowing its buffer with the terminating zero. As well as fixing this, I
2780 have increased the buffer size to 1024 (and made a note to document this).
2781
2782 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2783 when Exim is waiting for an SMTP response from a remote server, Exim
2784 restarts its select() call on the socket, thereby resetting its timeout.
2785 This is not a problem when such interrupts are rare. Somebody set up a cron
2786 job to run exiwhat every 2 minutes, which is less than the normal select()
2787 timeout (5 or 10 minutes). This meant that the select() timeout never
2788 kicked in because it was always reset. I have fixed this by comparing the
2789 time when an interrupt arrives with the time at the start of the first call
2790 to select(). If more time than the timeout has elapsed, the interrupt is
2791 treated as a timeout.
2792
2793 5. Some internal re-factoring in preparation for the addition of Sieve
2794 extensions (by MH). In particular, the "personal" test is moved to a
2795 separate function, and given an option for scanning Cc: and Bcc: (which is
2796 not set for Exim filters).
2797
2798 6. When Exim created an email address using the login of the caller as the
2799 local part (e.g. when creating a From: or Sender: header line), it was not
2800 quoting the local part when it contained special characters such as @.
2801
2802 7. Installed new OpenBSD configuration files.
2803
2804 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2805 try to make them clearer.
2806
2807 9. Callout options, other than the timeout value, were being ignored when
2808 verifying sender addresses in header lines. For example, when using
2809
2810 verify = header_sender/callout=no_cache
2811
2812 the cache was (incorrectly) being used.
2813
2814 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2815 only the Cygwin environment.
2816
2817 11. When return_path_on_delivery was set as a log selector, if different remote
2818 addresses in the same message used different return paths and parallel
2819 remote delivery occurred, the wrong values would sometimes be logged.
2820 (Whenever a remote delivery process finished, the return path value from
2821 the most recently started remote delivery process was logged.)
2822
2823 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2824 lines when AUTH and/or TLS are in use. This is the "received protocol"
2825 field. Exim used to use "asmtp" for authenticated SMTP, without any
2826 indication (in the protocol name) for TLS use. Now it follows the RFC and
2827 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2828 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2829 names appear in log lines as well as in Received: header lines.
2830
2831 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2832 extensions, and comparison tests, and to fix some bugs.
2833
2834 14. Changes to the "personal" filter test:
2835
2836 (1) The test was buggy in that it was just doing the equivalent of
2837 "contains" tests on header lines. For example, if a user's address was
2838 anne@some.where, the "personal" test would incorrectly be true for
2839
2840 To: susanne@some.where
2841
2842 This test is now done by extracting each address from the header in turn,
2843 and checking the entire address. Other tests that are part of "personal"
2844 are now done using regular expressions (for example, to check local parts
2845 of addresses in From: header lines).
2846
2847 (2) The list of non-personal local parts in From: addresses has been
2848 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2849 taken from the Sieve specification recommendations.
2850
2851 (3) If the message contains any header line starting with "List-" it is
2852 treated as non-personal.
2853
2854 (4) The test for "circular" in the Subject: header line has been removed
2855 because it now seems ill-conceived.
2856
2857 15. Minor typos in src/EDITME comments corrected.
2858
2859 16. Installed latest exipick from John Jetmore.
2860
2861 17. If headers_add on a router specified a text string that was too long for
2862 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2863 of string_sprintf() is now avoided.
2864
2865 18. $message_body_size was not set (it was always zero) when running the DATA
2866 ACL and the local_scan() function.
2867
2868 19. For the "mail" command in an Exim filter, no default was being set for
2869 the once_repeat time, causing a random time value to be used if "once" was
2870 specified. (If the value happened to be <= 0, no repeat happened.) The
2871 default is now 0s, meaning "never repeat". The "vacation" command was OK
2872 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2873 (I found it when inspecting the code).
2874
2875 20. There is now an overall timeout for performing a callout verification. It
2876 defaults to 4 times the callout timeout, which applies to individual SMTP
2877 commands during the callout. The overall timeout applies when there is more
2878 than one host that can be tried. The timeout is checked before trying the
2879 next host. This prevents very long delays if there are a large number of
2880 hosts and all are timing out (e.g. when the network connections are timing
2881 out). The value of the overall timeout can be changed by specifying an
2882 additional sub-option for "callout", called "maxwait". For example:
2883
2884 verify = sender/callout=5s,maxwait=20s
2885
2886 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2887 to the end before writing, but this should make it even safer).
2888
2889 22. Exim was forgetting that it had advertised PIPELINING for the second and
2890 subsequent messages on an SMTP connection. It was also not resetting its
2891 memory on STARTTLS and an internal HELO.
2892
2893 23. When Exim logs an SMTP synchronization error within a session, it now
2894 records whether PIPELINING has been advertised or not.
2895
2896 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2897 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2898 rather than long int.
2899
2900 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2901
2902 26. Added the never_mail option to autoreply.
2903
2904
2905 Exim version 4.41
2906 -----------------
2907
2908 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2909 crash if the getsockname() call failed; this can happen if a connection is
2910 closed very soon after it is established. The problem was simply in the
2911 order in which certain operations were done, causing Exim to try to write
2912 to the SMTP stream before it had set up the file descriptor. The bug has
2913 been fixed by making things happen in the correct order.
2914
2915
2916 Exim version 4.40
2917 -----------------
2918
2919 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2920 before the connection was closed, thus losing the rejection response.
2921
2922 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2923 some early Solaris releases, but causes trouble in current releases where
2924 socklen_t is defined.
2925
2926 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2927 always exist.
2928
2929 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2930 configured.
2931
2932 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2933 that releasing the top of it at the end releases what was used for sub-
2934 expansions (unless the block got too big). However, discard this block if
2935 the first thing is a variable or header, so that we can use its block when
2936 it is dynamic (useful for very large $message_headers, for example).
2937
2938 6. Lookups now cache *every* query, not just the most recent. A new, separate
2939 store pool is used for this. It can be recovered when all lookup caches are
2940 flushed. Lookups now release memory at the end of their result strings.
2941 This has involved some general refactoring of the lookup sources.
2942
2943 7. Some code has been added to the store_xxx() functions to reduce the amount
2944 of flapping under certain conditions.
2945
2946 8. log_incoming_interface used to affect only the <= reception log lines. Now
2947 it causes the local interface and port to be added to several more SMTP log
2948 lines, for example "SMTP connection from", and rejection lines.
2949
2950 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2951
2952 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2953
2954 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2955 could be overwritten at the end of the current message (or the start of a
2956 new message if it was set in a HELO ACL). The value is now preserved for
2957 the duration of the SMTP connection.
2958
2959 12. If a transport had a headers_rewrite setting, and a matching header line
2960 contained an unqualified address, that address was qualified, even if it
2961 did not match any rewriting rules. The underlying bug was that the values
2962 of the flags that permit the existence of unqualified sender and recipient
2963 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2964 non-local messages, and by -bnq for local messages) were not being
2965 preserved with the message after it was received.
2966
2967 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2968 "next input=" as part of the text comprising the host identity instead of
2969 the correct text. The code was using the same buffer for two different
2970 strings. However, depending on which order the printing function evaluated
2971 its arguments, the bug did not always show up. Under Linux, for example, my
2972 test suite worked just fine.
2973
2974 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2975 doesn't work with some older versions of Perl. It has been changed to "my",
2976 which in any case is probably the better facility to use.
2977
2978 15. A really picky compiler found some instances of statements for creating
2979 error messages that either had too many or two few arguments for the format
2980 string.
2981
2982 16. The size of the buffer for calls to the DNS resolver has been increased
2983 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2984 for addresses that have a lot of PTR records. This alleviates a problem; it
2985 does not fully solve it.
2986
2987 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2988 buffer now truncates the list and logs the incident, which is the same
2989 action as happens when Exim is looking up a host name and its aliases.
2990 Previously in this situation something unpredictable would happen;
2991 sometimes it was "internal error: store_reset failed".
2992
2993 18. If a server dropped the connection unexpectedly when an Exim client was
2994 using GnuTLS and trying to read a response, the client delivery process
2995 crashed while trying to generate an error log message.
2996
2997 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2998 string, for example:
2999
3000 warn message = H1: something\nH2: something
3001
3002 the text was added as a single header line from Exim's point of view
3003 though it ended up OK in the delivered message. However, searching for the
3004 second and subsequent header lines using $h_h2: did not work. This has been
3005 fixed. Similarly, if a system filter added multiple headers in this way,
3006 the routers could not see them.
3007
3008 20. Expanded the error message when iplsearch is called with an invalid key to
3009 suggest using net-iplsearch in a host list.
3010
3011 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3012 ACLs are no longer actually imposed (and a message to that effect is
3013 output).
3014
3015 22. If a "gecos" field in a passwd entry contained escaped characters, in
3016 particular, if it contained a \" sequence, Exim got it wrong when building
3017 a From: or a Sender: header from that name. A second bug also caused
3018 incorrect handling when an unquoted " was present following a character
3019 that needed quoting.
3020
3021 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3022 was not being matched caselessly.
3023
3024 24. Arranged for all hyphens in the exim.8 source to be escaped with
3025 backslashes.
3026
3027 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3028 properly. Recipient callout cache records were still being keyed to include
3029 the sender, even when use_sender was set false. This led to far more
3030 callouts that were necessary. The sender is no longer included in the key
3031 when use_sender is false.
3032
3033 26. Added "control = submission" modifier to ACLs.
3034
3035 27. Added the ${base62d: operator to decode base 62 numbers.
3036
3037 28. dnsdb lookups can now access SRV records.
3038
3039 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3040 the configuration file.
3041
3042 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3043 (-v) mode. This makes the output for a verbose queue run more intelligible.
3044
3045 31. Added a use_postmaster feature to recipient callouts.
3046
3047 32. Added the $body_zerocount variable, containing the number of binary zero
3048 bytes in the message body.
3049
3050 33. The time of last modification of the "new" subdirectory is now used as the
3051 "mailbox time last read" when there is a quota error for a maildir
3052 delivery.
3053
3054 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3055
3056 35. Added +ignore_unknown as a special item in host lists.
3057
3058 36. Code for decoding IPv6 addresses in host lists is now included, even if
3059 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3060 address was recognized as an IP address, but was then not correctly decoded
3061 into binary, causing unexpected and incorrect effects when compared with
3062 another IP address.
3063
3064
3065 Exim version 4.34
3066 -----------------
3067
3068 1. Very minor rewording of debugging text in manualroute to say "list of
3069 hosts" instead of "hostlist".
3070
3071 2. If verify=header_syntax was set, and a header line with an unqualified
3072 address (no domain) and a large number of spaces between the end of the
3073 name and the colon was received, the reception process suffered a buffer
3074 overflow, and (when I tested it) crashed. This was caused by some obsolete
3075 code that should have been removed. The fix is to remove it!
3076
3077 3. When running in the test harness, delay a bit after writing a bounce
3078 message to get a bit more predictability in the log output.
3079
3080 4. Added a call to search_tidyup() just before forking a reception process. In
3081 theory, someone could use a lookup in the expansion of smtp_accept_max_
3082 per_host which, without the tidyup, could leave open a database connection.
3083
3084 5. Added the variables $recipient_data and $sender_data which get set from a
3085 lookup success in an ACL "recipients" or "senders" condition, or a router
3086 "senders" option, similar to $domain_data and $local_part_data.
3087
3088 6. Moved the writing of debug_print from before to after the "senders" test
3089 for routers.
3090
3091 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3092 problems for message scanning, either using a data ACL, or using
3093 local_scan() because the Received: header was not generated till after they
3094 were called (in order to set the time as the time of reception completion).
3095 I have revised the way this works. The header is now generated after the
3096 body is received, but before the ACL or local_scan() are called. After they
3097 are run, the timestamp in the header is updated.
3098
3099
3100 Exim version 4.33
3101 -----------------
3102
3103 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3104 before starting a queue runner without re-exec. This happened only when
3105 deliver_drop_privilege was set or when the Exim user was set to root. The
3106 effect of the bug was that timeouts during subsequent deliveries caused
3107 crashes instead of being properly handled. The handler is now left at its
3108 default (and expected) setting.
3109
3110 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3111 message, again when deliver_drop_privilege is set or Exim is run as root.
3112 The bug described in (1) was not present in this case, but the tidying up
3113 of the other signals was missing. I have made the two cases consistent.
3114
3115 3. The ignore_target_hosts setting on a manualroute router was being ignored
3116 for hosts that were looked up using the /MX notation.
3117
3118 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3119 in domain lists.
3120
3121 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3122 operated on the sender address. After changing the $sender_address to <>
3123 for the sender address verify, Exim was re-instated it as the original
3124 (before rewriting) address, but remembering that it had rewritten it, so it
3125 wasn't rewriting it again. This bug also had the effect of breaking the
3126 sender address verification caching when the sender address was rewritten.
3127
3128 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3129 This has been changed so that if the ip literal address matches
3130 ignore_target_hosts, the router declines.
3131
3132 7. Added expansion conditions match_domain, match_address, and match_local_
3133 part (NOT match_host).
3134
3135 8. The placeholder for the Received: header didn't have a length field set.
3136
3137 9. Added code to Exim itself and to exim_lock to test for a specific race
3138 condition that could lead to file corruption when using MBX delivery. The
3139 issue is with the lockfile that is created in /tmp. If this file is removed
3140 after a process has opened it but before that process has acquired a lock,
3141 there is the potential for a second process to recreate the file and also
3142 acquire a lock. This could lead to two Exim processes writing to the file
3143 at the same time. The added code performs the same test as UW imapd; it
3144 checks after acquiring the lock that its file descriptor still refers to
3145 the same named file.
3146
3147 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3148 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3149 when Exim is built.
3150
3151 11. Added the smtp_active_hostname option. If used, this will typically be made
3152 to depend on the incoming interface address. Because $interface_address is
3153 not set up until the daemon has forked a reception process, error responses
3154 that can happen earlier (such as "too many connections") no longer contain
3155 a host name.
3156
3157 12. If an expansion in a condition on a "warn" statement fails because a lookup
3158 defers, the "warn" statement is abandoned, and the next ACL statement is
3159 processed. Previously this caused the whole ACL to be aborted.
3160
3161 13. Added the iplsearch lookup type.
3162
3163 14. Added ident_timeout as a log selector.
3164
3165 15. Added tls_certificate_verified as a log selector.
3166
3167 16. Added a global option tls_require_ciphers (compare the smtp transport
3168 option of the same name). This controls incoming TLS connections.
3169
3170 17. I finally figured out how to make tls_require_ciphers do a similar thing
3171 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3172 before starting the TLS session.
3173
3174 18. Tabs are now shown as \t in -bP output.
3175
3176 19. If the log selector return_path_on_delivery was set, Exim crashed when
3177 bouncing a message because it had too many Received: header lines.
3178
3179 20. If two routers both had headers_remove settings, and the first one included
3180 a superfluous trailing colon, the final name in the first list and the
3181 first name in the second list were incorrectly joined into one item (with a
3182 colon in the middle).
3183
3184
3185 Exim version 4.32
3186 -----------------
3187
3188 1. Added -C and -D options to the exinext utility, mainly to make it easier
3189 to include in the automated testing, but these could be helpful when
3190 multiple configurations are in use.
3191
3192 2. The exinext utility was not formatting the output nicely when there was
3193 an alternate port involved in the retry record key, nor when there was a
3194 message id as well (for retries that were specific to a specific message
3195 and a specific host). It was also confused by IPv6 addresses, because of
3196 the additional colons they contain. I have fixed the IPv4 problem, and
3197 patched it up to do a reasonable job for IPv6.
3198
3199 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3200 delivery, the log line now contains "pipelined" if PIPELINING was used.
3201
3202 4. An SMTP transport process used to panic and die if the bind() call to set
3203 an explicit outgoing interface failed. This has been changed; it is now
3204 treated in the same way as a connect() failure.
3205
3206 5. A reference to $sender_host_name in the part of a conditional expansion
3207 that was being skipped was still causing a DNS lookup. This no longer
3208 occurs.
3209
3210 6. The def: expansion condition was not recognizing references to header lines
3211 that used bh_ and bheader_.
3212
3213 7. Added the _cache feature to named lists.
3214
3215 8. The code for checking quota_filecount in the appendfile transport was
3216 allowing one more file than it should have been.
3217
3218 9. For compatibility with Sendmail, the command line option
3219
3220 -prval:sval
3221
3222 is equivalent to
3223
3224 -oMr rval -oMs sval
3225
3226 and sets the incoming protocol and host name (for trusted callers). The
3227 host name and its colon can be omitted when only the protocol is to be set.
3228 Note the Exim already has two private options, -pd and -ps, that refer to
3229 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3230 "s", but I don't think that's a major issue.
3231
3232 10. A number of refactoring changes to the code, none of which should affect
3233 Exim's behaviour:
3234
3235 (a) The number of logging options was getting close to filling up the
3236 32-bit word that was used as a bit map. I have split them into two classes:
3237 those that are passed in the argument to log_write(), and those that are
3238 only ever tested independently outside of that function. These are now in
3239 separate 32-bit words, so there is plenty of room for expansion again.
3240 There is no change in the user interface or the logging behaviour.
3241
3242 (b) When building, for example, log lines, the code previously used a
3243 macro that called string_cat() twice, in order to add two strings. This is
3244 not really sufficiently general. Furthermore, there was one instance where
3245 it was actually wrong because one of the argument was used twice, and in
3246 one call a function was used. (As it happened, calling the function twice
3247 did not affect the overall behaviour.) The macro has been replaced by a
3248 function that can join an arbitrary number of extra strings onto a growing
3249 string.
3250
3251 (c) The code for expansion conditions now uses a table and a binary chop
3252 instead of a serial search (which was left over from when there were very
3253 few conditions). Also, it now recognizes conditions like "pam" even when
3254 the relevant support is not compiled in: a suitably worded error message is
3255 given if an attempt is made to use such a condition.
3256
3257 11. Added ${time_interval:xxxxx}.
3258
3259 12. A bug was causing one of the ddress fields not to be passed back correctly
3260 from remote delivery subprocesses. The field in question was not being
3261 subsequently used, so this caused to problems in practice.
3262
3263 13. Added new log selectors queue_time and deliver_time.
3264
3265 14. Might have fixed a bug in maildirsizefile handling that threw up
3266 "unexpected character" debug warnings, and recalculated the data
3267 unnecessarily. In any case, I expanded the warning message to give more
3268 information.
3269
3270 15. Added the message "Restricted characters in address" to the statements in
3271 the default ACL that block characters like @ and % in local parts.
3272
3273 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3274 Three changes have been made:
3275
3276 (a) There was a serious bug; a negative response to MAIL caused the whole
3277 recipient domain to be cached as invalid, thereby blocking all messages
3278 to all local parts at the same domain, from all senders. This bug has
3279 been fixed. The domain is no longer cached after a negative response to
3280 MAIL if the sender used is not empty.
3281
3282 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3283 been restored.
3284
3285 (c) A new callout option, "use_sender" has been added for people who want
3286 the modified behaviour.
3287
3288
3289 Exim version 4.31
3290 -----------------
3291
3292 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3293 Larry Rosenman.
3294
3295 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3296 indeed breaks things for older releases.
3297
3298 3. Added additional logging to the case where there is a problem reading data
3299 from a filter that is running in a subprocess using a pipe, in order to
3300 try to track down a specific problem.
3301
3302 4. Testing facility fudge: when running in the test harness and attempting
3303 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3304 getting "No route to host". Convert this to a timeout.
3305
3306 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3307 warning.
3308
3309 6. Some OS don't have socklen_t but use size_t instead. This affects the
3310 fifth argument of getsockopt() amongst other things. This is now
3311 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3312 can be set for individual OS. I have set it for SunOS5, OSF1, and
3313 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3314 some earlier ones do not.
3315
3316 7. Change 4.30/15 was not doing the test caselessly.
3317
3318 8. The standard form for an IPv6 address literal was being rejected by address
3319 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3320 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3321 this, as well as the form without the "IPv6" on the front (but only when
3322 address literals are enabled, of course).
3323
3324 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3325
3326 10. Exim crashed if a message with an empty sender address specified by -f
3327 encountered a router with an errors_to setting. This could be provoked only
3328 by a command such as
3329
3330 exim -f "" ...
3331
3332 where an empty string was supplied; "<>" did not hit this bug.
3333
3334 11. Installed PCRE release 4.5.
3335
3336 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3337 remained set. It is now erased.
3338
3339 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3340 times from message ids (which are base 36 rather than the normal 62).
3341
3342 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3343 were being counted as actual protocol errors, and logged if the log
3344 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3345 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3346 and DATA following a set of rejected RCPTs do not count as protocol errors.
3347 In other words, Exim assumes they were pipelined, though this may not
3348 actually be the case. Of course, in all cases the client gets an
3349 appropriate error code.
3350
3351 15. If a lookup fails in an ACL condition, a message about the failure may
3352 be available; it is used if testing the ACL cannot continue, because most
3353 such messages specify what the cause of the deferral is. However, some
3354 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3355 that caused an old message to be retained and used if a later statement
3356 caused a defer, replacing the real cause of the deferral.
3357
3358 16. If an IP address had so many PTR records that the DNS lookup buffer
3359 was not large enough to hold them, Exim could crash while trying to process
3360 the truncated data. It now detects and logs this case.
3361
3362 17. Further to 4.21/58, another change has been made: if (and only if) the
3363 first line of a message (the first header line) ends with CRLF, a bare LF
3364 in a subsequent header line has a space inserted after it, so as not to
3365 terminate the header.
3366
3367 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3368 unnecessarily, used atoi() instead of strtol(), and didn't check the
3369 termination when getting file sizes from file names by regex.
3370
3371 19. Completely re-implemented the support for maildirsize files, in the light
3372 of a number of problems with the previous contributed implementation
3373 (4.30/29). In particular:
3374
3375 . If the quota is zero, the maildirsize file is maintained, but no quota is
3376 imposed.
3377
3378 . If the maildir directory does not exist, it is created before any attempt
3379 to write a maildirsize file.
3380
3381 . The quota value in the file is just a cache; if the quota is changed in
3382 the transport, the new value overrides.
3383
3384 . A regular expression is available for excluding directories from the
3385 count.
3386
3387 20. The autoreply transport checks the characters in options that define the
3388 message's headers; it allows continued headers, but it was checking with
3389 isspace() after an embedded newline instead of explicitly looking for a
3390 space or a tab.
3391
3392 21. If all the "regular" hosts to which an address was routed had passed their
3393 expiry times, and had not reached their retry times, the address was
3394 bounced, even if fallback hosts were defined. Now Exim should go on to try
3395 the fallback hosts.
3396
3397 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3398 equivalent code in the SMTP transport. Some hosts send humungous responses
3399 to HELO/EHLO, more than 1024 it seems.
3400
3401 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3402 gives compiler warnings in some environments. I've now done it "properly",
3403 using a union.
3404
3405 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3406 (because of problems with the built-in one) was declared to return uschar *
3407 instead of char *, causing compiler failure.
3408
3409 25. Fixed a file descriptor leak when processing alias/forward files.
3410
3411 26. Fixed a minor format string issue in dbfn.c.
3412
3413 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3414
3415 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3416 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3417 become corrupted.
3418
3419 29. When a sender address is verified, it is cached, to save repeating the test
3420 when there is more than one recipient in a message. However, when the
3421 verification involves a callout, it is possible for different callout
3422 options to be set for different recipients. It is too complicated to keep
3423 track of this in the cache, so now Exim always runs a verification when a
3424 callout is required, relying on the callout cache for the optimization.
3425 The overhead is duplication of the address routing, but this should not be
3426 too great.
3427
3428 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3429 to be verified with callout=postmaster, and the main callout worked but the
3430 postmaster check failed, the verification correctly failed. However, if a
3431 subsequent RCPT command asked for sender verification *without* the
3432 postmaster check, incorrect caching caused this verification also to fail,
3433 incorrectly.
3434
3435 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3436 it was not caching the DNS options (qualify_single, search_parents) that
3437 were used when the lookup failed. A subsequent lookup with different
3438 options therefore always gave the same answer, though there were cases
3439 where it should not have. (Example: a "domains = !$mx_any" option on a
3440 dnslookup router: the "domains" option is always processed without any
3441 widening, but the router might have qualify_single set.) Now Exim uses the
3442 cached value only when the same options are set.
3443
3444 32. Added John Jetmore's "exipick" utility to the distribution.
3445
3446 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3447 than a timeout (e.g. a certificate is required, and is not provided), an
3448 Exim server now closes the connection immediately. Previously it waited for
3449 the client to close - but if the client is SSL, it seems that they each
3450 wait for each other, leading to a delay before one of them times out.
3451
3452 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3453 maintained 0.8.x compatibility because I don't think many are using it, and
3454 it is clearly obsolete.
3455
3456 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3457 transport.
3458
3459 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3460 client certificate was expired. A simple patch fixes this, though I don't
3461 understand the full logic of why the verify callback is called multiple
3462 times.
3463
3464 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3465 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3466 which causes problems with some clients (such as the Certicom SSL Plus
3467 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3468 disables the coutermeasure allowing Eudora to connect."
3469
3470 38. Exim was not checking that a write() to a log file succeeded. This could
3471 lead to Bad Things if a log got too big, in particular if it hit a file
3472 size limit. Exim now panics and dies if it cannot write to a log file, just
3473 as it does if it cannot open a log file.
3474
3475 39. Modified OS/Makefile-Linux so that it now contains
3476
3477 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3478
3479 The two -D definitions ensure that Exim is compiled with large file
3480 support, which makes it possible to handle log files that are bigger than
3481 2^31.
3482
3483 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3484 instance) a domain was checked against a named list that involved a lookup,
3485 causing $domain_data to be set, then another domain was checked against the
3486 same list, then the first domain was re-checked, the value of $domain_data
3487 after the final check could be wrong. In particular, if the second check
3488 failed, it could be set empty. This bug probably also applied to
3489 $localpart_data.
3490
3491 41. The strip_trailing_dot option was not being applied to the address given
3492 with the -f command-line option.
3493
3494 42. The code for reading a message's header from the spool was incrementing
3495 $received_count, but never initializing it. This meant that the value was
3496 incorrect (doubled) while delivering a message in the same process in which
3497 it was received. In the most common configuration of Exim, this never
3498 happens - a fresh exec is done - but it can happen when
3499 deliver_drop_privilege is set.
3500
3501 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3502 it now includes up to 150 characters of the unexpected data in the log
3503 line.
3504
3505 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3506 and building data strings. The size of both of these buffers was 10 000
3507 bytes - far larger than anybody would *ever* want, thought I. Needless to
3508 say, somebody hit the limit. I have increased the maximum line length to
3509 20 000 and the maximum data length of concatenated lines to 100 000. I have
3510 also fixed two bugs, because there was no checking on these buffers. Tsk,
3511 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3512 buffer is too small.
3513
3514 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3515 lsearch lookups. Now it does.
3516
3517 46. When parsing a route_list item in a manualroute router, a fixed-length
3518 buffer was used for the list of hosts. I made this 1024 bytes long,
3519 thinking that nobody would ever have a list of hosts that long. Wrong.
3520 Somebody had a whole pile of complicated expansion conditions, and the
3521 string was silently truncated, leading to an expansion error. It turns out
3522 that it is easier to change to an unlimited length (owing to other changes
3523 that have happened since this code was originally written) than to build
3524 structure for giving a limitation error. The length of the item that
3525 expands into the list of hosts is now unlimited.
3526
3527 47. The lsearch lookup could not handle data where the length of text line was
3528 more than 4095 characters. Such lines were truncated, leading to shortened
3529 data being returned. It should now handle lines of any length.
3530
3531 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3532 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3533 ACL").
3534
3535 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3536 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3537 no longer happens in comment lines. A list of replacements is now placed
3538 at the head of all of the source files, except those whose only change is
3539 to replace PERL_COMMAND in the very first #! line.
3540
3541 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3542 messages on the queue, with a bottom-up merge sort, using code contributed
3543 by Michael Haardt. This should make operations like -bp somewhat faster on
3544 large queues. It won't affect queue runners, except when queue_run_in_order
3545 is set.
3546
3547 51. Installed eximstats 1.31 in the distribution.
3548
3549 52. Added support for SRV lookups to the dnslookup router.
3550
3551 53. If an ACL referred to $message_body or $message_body_end, the value was not
3552 reset for any messages that followed in the same SMTP session.
3553
3554 54. The store-handling optimization for building very long strings was not
3555 differentiating between the different store pools. I don't think this
3556 actually made any difference in practice, but I've tidied it.
3557
3558 55. While running the routers to verify a sender address, $sender_address
3559 was still set to the sender address. This is wrong, because when routing to
3560 send a bounce to the sender, it would be empty. Therefore, I have changed
3561 it so that, while verifying a sender address, $sender_address is set to <>.
3562 (There is no change to what happens when verifying a recipient address.)
3563
3564 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3565 target A or AAAA records (if not already returned) without resetting the
3566 qualify_single or search_parents options of the DNS resolver. These are
3567 inappropriate in this case because the targets of MX and SRV records must
3568 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3569 target that, when qualified, matched something in the local domain. These
3570 two options are now turned off when doing these lookups.
3571
3572 57. It seems that at least some releases of Reiserfs (which does not have the
3573 concept of a fixed number of inodes) returns zero and not -1 for the
3574 number of available inodes. This interacted badly with check_spool_inodes,
3575 which assumed that -1 was the "no such thing" setting. What I have done is
3576 to check that the total number of inodes is greater than zero before doing
3577 the test of how many are available.
3578
3579 58. When a "warn" ACL statement has a log_message modifier, the message is
3580 remembered, and not repeated. This is to avoid a lot of repetition when a
3581 message has many recipients that cause the same warning to be written.
3582 Howewer, Exim was preserving the list of already written lines for an
3583 entire SMTP session, which doesn't seem right. The memory is now reset if a
3584 new message is started.
3585
3586 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3587 debugging output unless log_rewrite was also set.
3588
3589 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3590 of (int)(handle) when we know that handle contains (void *)(-1).
3591
3592 61. The Exim daemon panic-logs an error return when it closes the incoming
3593 connection. However "connection reset by peer" seems to be common, and
3594 isn't really an error worthy of noting specially, so that particular error
3595 is no long logged.
3596
3597 62. When Exim is trying to find all the local interfaces, it used to panic and
3598 die if the ioctl to get the interface flags failed. However, it seems that
3599 on at least one OS (Solaris 9) it is possible to have an interface that is
3600 included in the list of interfaces, but for which you get a failure error
3601 for this call. This happens when the interface is not "plumbed" into a
3602 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3603 failure of the "get flags" call assumes that the interface is down.
3604
3605 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3606 makes life easier for people who are doing arithmetic on fields extracted
3607 from dates, where you often get leading zeros that should not be
3608 interpreted as octal.
3609
3610 64. Added qualify_domain to the redirect router, to override the global
3611 setting.
3612
3613 65. If a pathologically long header line contained very many addresses (the
3614 report of this problem mentioned 10 000) and each of them was rewritten,
3615 Exim could use up a very large amount of memory. (It kept on making new
3616 copies of the header line as it rewrote, and never released the old ones.)
3617 At the expense of a bit more processing, the header rewriting function has
3618 been changed so that it no longer eats memory in this way.
3619
3620 66. The generation of the Received: header has been moved from the time that a
3621 message starts to be received, to the time that it finishes. The timestamp
3622 in the Received: header should now be very close to that of the <= log
3623 line. There are two side-effects of this change:
3624
3625 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3626 logged header lines no longer include the local Received: line, because
3627 it has not yet been created. The same applies to a copy of the message
3628 that is returned to a non-SMTP sender when a message is rejected.
3629
3630 (b) When a filter file is tested using -bf, no additional Received: header
3631 is added to the test message. After some thought, I decided that this
3632 is a bug fix.
3633
3634 This change does not affect the value of $received_for. It is still set
3635 after address rewriting, but before local_scan() is called.
3636
3637 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3638
3639 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3640 gave an unhelpful panic error message, and a defer error. I have managed to
3641 change this behaviour so that it now rejects any supplied certificate,
3642 which seems right, as the list of acceptable certificates is empty.
3643
3644 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3645 gave an unhelpful defer error. I have not managed to make this reject any
3646 supplied certificates, but the error message it gives is "no certificate
3647 supplied", which is not helpful.
3648
3649 70. exigrep's output now also includes lines that are not associated with any
3650 message, but which match the given pattern. Implemented by a patch from
3651 Martin Sluka, which also tidied up the Perl a bit.
3652
3653 71. Recipient callout verification, like sender verification, was using <> in
3654 the MAIL FROM command. This isn't really the right thing, since the actual
3655 sender may affect whether the remote host accepts the recipient or not. I
3656 have changed it to use the actual sender in the callout; this means that
3657 the cache record is now keyed on a recipient/sender pair, not just the
3658 recipient address. There doesn't seem to be a real danger of callout loops,
3659 since a callout by the remote host to check the sender would use <>.
3660 [SEE ABOVE: changed after hitting problems.]
3661
3662 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3663 temporary errors. However, in the case of such a code being given after
3664 the end of a data transmission (i.e. after ".") Exim was failing to write
3665 a retry record for the message. (Yes, there was some broken host that was
3666 actually sending 8xx at this point.)
3667
3668 73. An unknown lookup type in a host list could cause Exim to panic-die when
3669 the list was checked. (An example that provoked this was putting <; in the
3670 middle of a list instead of at the start.) If this happened during a DATA
3671 ACL check, a -D file could be left lying around. This kind of configuration
3672 error no longer causes Exim to die; instead it causes a defer errror. The
3673 incident is still logged to the main and panic logs.
3674
3675 74. Buglet left over from Exim 3 conversion. The message "too many messages
3676 in one connection" was written to the rejectlog but not the mainlog, except
3677 when address rewriting (yes!) was being logged.
3678
3679 75. Added write_rejectlog option.
3680
3681 76. When a system filter was run not as root (that is, when system_filter_user
3682 was set), the values of the $n variables were not being returned to the
3683 main process; thus, they were not subsequently available in the $sn
3684 variables.
3685
3686 77. Added +return_path_on_delivery log selector.
3687
3688 78. A connection timeout was being treated differently from recipients deferred
3689 when testing hosts_max_try with a message that was older than the host's
3690 retry timeout. (The host should not be counted, thus allowing all hosts to
3691 be tried at least once before bouncing.) This may have been the cause of an
3692 occasionally reported bug whereby a message would remain on the queue
3693 longer than the retry timeout, but would be bounced if a delivery was
3694 forced. I say "may" because I never totally pinned down the problem;
3695 setting up timeout/retry tests is difficult. See also the next item.
3696
3697 79. The ultimate address timeout was not being applied to errors that involved
3698 a combination of host plus message (for example, a timeout on a MAIL
3699 command). When an address resolved to a number of possible hosts, and they
3700 were not all tried for each delivery (e.g. because of hosts_max_try), a
3701 message could remain on the queue longer than the retry timeout.
3702
3703 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3704 Haardt.
3705
3706 81. Fixed an obscure SMTP outgoing bug which required at least the following
3707 conditions: (a) there was another message waiting for the same server;
3708 (b) the server returned 5xx to all RCPT commands in the first message so
3709 that the message was not completed; (c) the server dropped the connection
3710 or gave a negative response to the RSET that Exim sends to abort the
3711 transaction. The observed case was a dropped connection after DATA that had
3712 been sent in pipelining mode. That is, the server had advertised PIPELINING
3713 but was not implementing it correctly. The effect of the bug was incorrect
3714 behaviour, such as trying another host, and this could lead to a crash.
3715
3716
3717 Exim version 4.30
3718 -----------------
3719
3720 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3721 and daemon.c were passed as pointers to ints; they should have been
3722 pointers to socklen_t variables (which are typically unsigned ints).
3723
3724 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3725 fixed.
3726
3727 3. Fixed a really odd bug that affected only the testing scheme; patching a
3728 certain fixed string in the binary changed the value of another string that
3729 happened to be identical to the end of the original first string.
3730
3731 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3732 name", it returns that address as the IP address. On some operating
3733 systems (e.g. Solaris), it also passes back the IP address string as the
3734 "host name". However, on others (e.g. Linux), it passes back an empty
3735 string. Exim wasn't checking for this, and was changing the host name to an
3736 empty string, assuming it had been canonicized.
3737
3738 5. Although rare, it is permitted to have more than one PTR record for a given
3739 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3740 all the names associated with an address, because they do in Solaris.
3741 However, it seems that they do not in Linux for data that comes from the
3742 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3743 I found this out when I moved to a new Linux workstation and tried to run
3744 the Exim test suite.
3745
3746 To get round this problem I have changed the code so that it now does its
3747 own call to the DNS to look up PTR records when searching for a host name.
3748 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3749 addresses that are only in /etc/hosts are still found.
3750
3751 This behaviour is, however, controlled by an option called host_lookup_
3752 order, which defaults to "bydns:byaddr". If people want to use the other
3753 order, or indeed, just use one or the other means of lookup, they can
3754 specify it in this variable.
3755
3756 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3757 some operating systems, this comes back from gethostbyaddr() as an empty
3758 string, and this is what Exim used to test for. However, it seems that in
3759 other systems, "." is yielded. Exim now tests for this case too.
3760
3761 7. The values of check_spool_space and check_log_space are now held internally
3762 as a number of kilobytes instead of an absolute number of bytes. If a
3763 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3764 kilobyte. This means that much larger values can be stored.
3765
3766 8. Exim monitor: an attempt to get the action menu when not actually pointing
3767 at a message produces an empty menu entitled "No message selected". This
3768 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3769 no entries in it ("Shell widget menu has zero width and/or height"). So I
3770 have added a single, blank menu entry in this case.
3771
3772 9. Added ${quote_local_part.
3773
3774 10. MIME decoding is now applied to the contents of Subject: header lines when
3775 they are logged.
3776
3777 11. Now that a reference to $sender_host_address automatically causes a reverse
3778 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3779 host lookup before query-style lookups in lists that might use this
3780 variable. This has therefore been abolished, and the "net-" prefix is no
3781 longer necessary for query-style lookups.
3782
3783 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3784 have been a typo for LFLAGS, so it has been changed.
3785
3786 13. The install script calls Exim with "-C /dev/null" in order to find the
3787 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3788 to be output. Howeve, since Exim outputs its version number before the
3789 error, it didn't break the script. It just looked ugly. I fixed this by
3790 always allowing "-C /dev/null" if the caller is root.
3791
3792 14. Ignore overlarge ACL variable number when reading spool file - insurance
3793 against a later release with more variables having written the file.
3794
3795 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3796 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3797 this, as well as the form without the "IPv6" on the front.
3798
3799 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3800 OS/Makefile-Darwin file.
3801
3802 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3803 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3804
3805 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3806 variable. However, some OS use ints for inodes. I've added cast to long int
3807 to get rid of the compiler warning.
3808
3809 19. I had forgotten to lock out "/../" in configuration file names when
3810 ALT_CONFIG_PREFIX was set.
3811
3812 20. Routers used for verification do not need to specify transports. However,
3813 if such a router generated a host list, and callout was configured, Exim
3814 crashed, because it could not find a port number from the (non-existent)
3815 transport. It now assumes port 25 in this circumstance.
3816
3817 21. Added the -t option to exigrep.
3818
3819 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3820 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3821 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3822 them is set, LOOKUP_LSEARCH is forced.
3823
3824 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3825
3826 24. Added sender and host information to the "rejected by local_scan()" log
3827 line; previously there was no indication of these.
3828
3829 25. Added .include_if_exists.
3830
3831 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3832 Linux. It turns out that not all file systems support this. Apparently some
3833 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3834 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3835 not supported on the file descriptor, is now ignored when Exim is trying to
3836 sync a directory. This applies only to Linux.
3837
3838 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3839
3840 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3841 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3842 this.
3843
3844 29. Added support for maildirsize files from supplied patch (modified a bit).
3845
3846 30. The use of :fail: followed by an empty string could lead Exim to respond to
3847 sender verification failures with (e.g.):
3848
3849 550 Verification failed for <xxx>
3850 550 Sender verify failed
3851
3852 where the first response line was missing the '-' that indicates it is not
3853 the final line of the response.
3854
3855 31. The loop for finding the name of the user that called Exim had a hardwired
3856 limit of 10; it now uses the value of finduser_retries, which is used for
3857 all other user lookups.
3858
3859 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3860 delivery time.
3861
3862 33. Exim was neglecting to zero errno before one call of strtol() when
3863 expanding a string and expecting an integer value. On some systems this
3864 resulted in spurious "integer overflow" errors. Also, it was casting the
3865 result into an int without checking.
3866
3867 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3868 did not work. The code looks as if it has *never* worked, though it appears
3869 to have been documented since at least releast 1.62. I have made it work.
3870
3871 35. The "timeout_DNS" error in retry rules, also documented since at least
3872 1.62, also never worked. As it isn't clear exactly what this means, and
3873 clearly it isn't a major issue, I have abolished the feature by treating it
3874 as "timeout", and writing a warning to the main and panic logs.
3875
3876 36. The display of retry rules for -brt wasn't always showing the error code
3877 correctly.
3878
3879 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3880 timeout_connect_A, timeout_connect_MX.
3881
3882 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3883 to the empty sender.
3884
3885 39. The daemon was not analysing the content of -oX till after it had closed
3886 stderr and disconnected from the controlling terminal. This meant that any
3887 syntax errors were only noted on the panic log, and the return code from
3888 the command was 0. By re-arranging the code a little, I've made the
3889 decoding happen first, so such errors now appear on stderr, and the return
3890 code is 1. However, the actual setting up of the sockets still happens in
3891 the disconnected process, so errors there are still only recorded on the
3892 panic log.
3893
3894 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3895 connections (as happens on some IP stacks) was logged at start up time as
3896 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3897 it from "IPv6 and IPv4", which means that two separate sockets are being
3898 used.
3899
3900 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3901 says whether AF_INET or AF_INET6 was passed as an argument.
3902
3903 42. Exiwhat output was messed up when time zones were included in log
3904 timestamps.
3905
3906 43. Exiwhat now gives more information about the daemon's listening ports,
3907 and whether -tls-on-connect was used.
3908
3909 44. The "port" option of the smtp transport is now expanded.
3910
3911 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3912 silently ignored. Now an error message is written to the main and panic
3913 logs.
3914
3915 46. There's a new ACL modifier called "logwrite" which writes to a log file
3916 as soon as it is encountered.
3917
3918 47. Added $local_user_uid and $local_user_gid at routing time.
3919
3920 48. Exim crashed when trying to verify a sender address that was being
3921 rewritten to "<>".
3922
3923 49. Exim was recognizing only a space character after ".include". It now also
3924 recognizes a tab character.
3925
3926 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3927 extracting the relevant information from the specification. The man page no
3928 longer contains scrambled data for the -d option, and I've added a section
3929 at the front about calling Exim under different names.
3930
3931 51. Added "extra_headers" argument to the "mail" command in filter files.
3932
3933 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3934 crash.
3935
3936 53. Installed eximstats 1.29.
3937
3938 54. Added transport_filter_timeout as a generic transport option.
3939
3940 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3941 Cc: header lines. This was required by RFC 822, but it not required by RFC
3942 2822.
3943
3944 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3945 incoming messages that did not have them. Now it does so only if the
3946 message originates locally, that is, if there is no associated remote host
3947 address. When Resent- header lines are present, this applies to the Resent-
3948 lines rather than the non-Resent- lines.
3949
3950 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3951 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3952
3953 58. Messages for configuration errors now include the name of the main
3954 configuration file - useful now that there may be more than one file in a
3955 list (.included file names were always shown).
3956
3957 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3958 for those rare installations that do not start the daemon as root or run it
3959 setuid root. I've cut out the call to initgroups() if the daemon is not
3960 root at that time.
3961
3962 60. The Exim user and group can now be bound into the binary as text strings
3963 that are looked up at the start of Exim's processing.
3964
3965 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3966
3967 62. Added $mailstore_basename variable.
3968
3969 63. Installed patch to sieve.c from Michael Haardt.
3970
3971 64. When Exim failed to open the panic log after failing to open the main log,
3972 the original message it was trying to log was written to stderr and debug
3973 output, but if they were not available (the usual case in production), it
3974 was lost. Now it is written to syslog before the two lines that record the
3975 failures to open the logs.
3976
3977 65. Users' Exim filters run in subprocesses under the user's uid. It is
3978 possible for a "deliver" command or an alias in a "personal" command to
3979 provoke an address rewrite. If logging of address rewriting is configured,
3980 this fails because the process is not running as root or exim. There may be
3981 a better way of dealing with this, but for the moment (because 4.30 needs
3982 to be released), I have disabled address rewrite logging when running a
3983 filter in a non-root, non-exim process.
3984
3985
3986 Exim version 4.24
3987 -----------------
3988
3989 1. The buildconfig auxiliary program wasn't quoting the value set for
3990 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3991 not defined. This bug was masked in 4.22 by the effect that was fixed in
3992 change 4.23/1.
3993
3994 2. Some messages that were rejected after a message id was allocated were
3995 shown as "incomplete" by exigrep. It no longer does this for messages that
3996 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3997
3998 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3999 have allow_domain_literals set, the ID did not get logged in the <= line.
4000 Domain literals are now always recognized in Message-ID: header lines.
4001
4002 4. The first argument for a ${extract expansion item is the key name or field
4003 number. Leading and trailing spaces in this item were not being ignored,
4004 causing some misleading effects.
4005
4006 5. When deliver_drop_privilege was set, single queue runner processes started
4007 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4008 same command in the process it spins off) were not dropping privilege.
4009
4010 6. When the daemon running as "exim" started a queue runner, it always
4011 re-executed Exim in the spun-off process. This is a waste of effort when
4012 deliver_drop_privilege is set. The new process now just calls the
4013 queue-runner function directly.
4014
4015
4016 Exim version 4.23
4017 -----------------
4018
4019 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4020 HEADERS_CHARSET.
4021
4022 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4023 ignored. Though the use of -oP was forcing the writing of a pid file, it
4024 was always written to the default place.
4025
4026 3. If the message "no IP address found for host xxxx" is generated during
4027 incoming verification, it is now followed by identification of the incoming
4028 connection (so you can more easily find what provoked it).
4029
4030 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4031
4032 5. Added some features to "harden" Exim a bit more against certain attacks:
4033
4034 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4035 be put in Local/Makefile. This is like the never_users runtime option,
4036 but it cannot be overridden. The default setting is "root".
4037
4038 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4039 prefix string with which any file named in a -C command line option
4040 must start.
4041
4042 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4043 is retained for -C and -D only if the caller of Exim is root. Without
4044 it, the exim user may also use -C and -D and retain privilege.
4045
4046 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4047 command line option is disabled.
4048
4049 6. Macro names set by the -D option must start with an upper case letter, just
4050 like macro names defined in the configuration file.
4051
4052 7. Added "dereference=" facility to LDAP.
4053
4054 8. Two instances of the typo "uknown" in the source files are fixed.
4055
4056 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4057 the Configure-Makefile script screwed up while processing it.
4058
4059 10. Incorporated PCRE 4.4.
4060
4061 11. The SMTP synchronization check was not operating right at the start of an
4062 SMTP session. For example, it could not catch a HELO sent before the client
4063 waited for the greeting. There is now a check for outstanding input at the
4064 point when the greeting is written. Because of the duplex, asynchronous
4065 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4066 way, but not yet received, when the check is performed.
4067
4068 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4069 on TCP/IP sockets, because this apparently causes some broken clients to
4070 timeout.
4071
4072 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4073 unchanged) from the Cygwin maintainer.
4074
4075 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4076 was supported instead of testing for mbx. Effectively a typo.
4077
4078 15. The spa authenticator server code was not checking that the input it
4079 received was valid base64.
4080
4081 16. The debug output line for the "set" modifier in ACLs was not showing the
4082 name of the variable that was being set.
4083
4084 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4085
4086 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4087 lookup on its own. Something else was needed to trigger the lookup. For
4088 example, a match in host_lookup or the need for a host name in a host list.
4089 Now, if $sender_host_name is referenced and the host name has not yet been
4090 looked up, a lookup is performed. If the lookup fails, the variable remains
4091 empty, and $host_lookup_failed is set to "1".
4092
4093 19. Added "eqi" as a case-independent comparison operator.
4094
4095 20. The saslauthd authentication condition could segfault if neither service
4096 nor realm was specified.
4097
4098 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4099 error message that was logged was misleading, and incoming SMTP
4100 connections were dropped. The message is now more accurate, and temporary
4101 errors are given to SMTP connections.
4102
4103 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4104 (except RSET) with a 421 error, until QUIT is received. However, it was
4105 failing to send a response to QUIT.
4106
4107 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4108 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4109 ACL is now run after the helo_try_verify_hosts code.
4110
4111 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4112 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4113 case-independent, so other case variants are also recognized). Apparently
4114 some systems use these upper case variants.
4115
4116 25. If more than two messages were waiting for the same host, and a transport
4117 filter was specified for the transport, Exim sent two messages over the
4118 same TCP/IP connection, and then failed with "socket operation on non-
4119 socket" when it tried to send the third.
4120
4121 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4122
4123 27. The extern definition of crypt16() in expand.c was not being excluded when
4124 the OS had its own crypt16() function.
4125
4126 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4127 as a preferred synonym for return_size_limit, both as an option and as an
4128 expansion variable.
4129
4130 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4131
4132 30. Changed the default configuration ACL to relax the local part checking rule
4133 for addresses that are not in any local domains. For these addresses,
4134 slashes and pipe symbols are allowed within local parts, but the sequence
4135 /../ is explicitly forbidden.
4136
4137 31. SPA server authentication was not clearing the challenge buffer before
4138 using it.
4139
4140 32. log_message in a "warn" ACL statement was writing to the reject log as
4141 well as to the main log, which contradicts the documentation and doesn't
4142 seem right (because no rejection is happening). So I have stopped it.
4143
4144 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4145 However, I am unable to do any testing of this.
4146
4147 34. Fixed an infelicity in the appendfile transport. When checking directories
4148 for a mailbox, to see if any needed to be created, it was accidentally
4149 using path names with one or more superfluous leading slashes; tracing
4150 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4151
4152 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4153 contents are added to the log line that is written for every discarded
4154 recipient. (Previously a log_message setting was ignored.)
4155
4156 36. The ${quote: operator now quotes the string if it is empty.
4157
4158 37. The install script runs exim in order to find its version number. If for
4159 some reason other than non-existence or emptiness, which it checks, it
4160 could not run './exim', it was installing it with an empty version number,
4161 i.e. as "exim-". This error state is now caught, and the installation is
4162 aborted.
4163
4164 38. An argument was missing from the function that creates an error message
4165 when Exim fails to connect to the socket for saslauthd authentication.
4166 This could cause Exim to crash, or give a corrupted message.
4167
4168 39. Added isip, isip4, and isip6 to ${if conditions.
4169
4170 40. The ACL variables $acl_xx are now saved with the message, and can be
4171 accessed later in routers, transports, and filters.
4172
4173 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4174 strings in the file are not string-expanded.
4175
4176 42. If a MAIL command specified a SIZE value that was too large to fit into an
4177 int variable, the check against message_size_limit failed. Such values are
4178 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4179 day this will have to be increased, but I don't think I want to be around
4180 when emails are that large.
4181
4182
4183
4184 Exim version 4.22
4185 -----------------
4186
4187 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4188 iconv() is not standard in FreeBSD.
4189
4190 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4191 IPv6 enabled. The observed symptom was a segmentation fault on return from
4192 the function os_common_find_running_interfaces() in src/os.c.
4193
4194 3. In the check_special_case() function in daemon.c I had used "errno" as an
4195 argument name, which causes warnings on some systems. This was basically a
4196 typo, since it was named "eno" in the comments!
4197
4198 4. The code that waits for the clock to tick (at a resolution of some fraction
4199 of a second) so as to ensure message-id uniqueness was always waiting for
4200 at least one whole tick, when it could have waited for less. [This is
4201 almost certainly not relevant at current processor speeds, where it is
4202 unlikely to ever wait at all. But we try to future-proof.]
4203
4204 5. The function that sleeps for a time interval that includes fractions of a
4205 second contained a race. It did not block SIGALRM between setting the
4206 timer, and suspending (a couple of lines later). If the interval was short
4207 and the sigsuspend() was delayed until after it had expired, the suspension
4208 never ended. On busy systems this could lead to processes getting stuck for
4209 ever.
4210
4211 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4212 process, before it forks any delivery processes. The open lookup caching
4213 mechanism meant that the open file or database connection was passed into
4214 the delivery process. The problem was that delivery processes always tidy
4215 up cached lookup data. This could cause a problem for the next delivery
4216 process started by the queue runner, because the external queue runner
4217 process does not know about the closure. So the next delivery process
4218 still has data in the lookup cache. In the case of a file lookup, there was
4219 no problem because closing a file descriptor in a subprocess doesn't affect
4220 the parent. However, if the lookup was caching a connection to a database,
4221 the connection was closed, and the second delivery process was likely to
4222 see errors such as "PGSQL: query failed: server closed the connection
4223 unexpectedly". The problem has been fixed by closing all cached lookups
4224 in a queue runner before running a delivery process.
4225
4226 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4227 seem to have the "const" qualifier which it has on other OS. I've
4228 parameterised it.
4229
4230 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4231 *of the same type* (client or server) with the same public name that an
4232 error should be diagnosed.
4233
4234 9. When Exim looked up a host name for an IP address, but failed to find the
4235 original IP address when looking up the host name (a safety check), it
4236 output the message "<ip address> does not match any IP for NULL", which was
4237 confusing, to say the least. The bug was that the host name should have
4238 appeared instead of "NULL".
4239
4240 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4241 user that is built into the binary, and the -C or -D options is used, root
4242 privilege is dropped before the configuration file is read. In addition,
4243 logging is switched to stderr instead of the normal log files. If the
4244 configuration then re-defines the Exim user, the unprivileged environment
4245 is probably not what is expected, so Exim logs a panic warning message (but
4246 proceeds).
4247
4248 However, if deliver_drop_privilege is set, the unprivileged state may well
4249 be exactly what is intended, so the warning has been cut out in that case,
4250 and Exim is allowed to try to write to its normal log files.
4251
4252
4253 Exim version 4.21
4254 -----------------
4255
4256 1. smtp_return_error_details was not giving details for temporary sender
4257 or receiver verification errors.
4258
4259 2. Diagnose a configuration error if two authenticators have the same public
4260 name.
4261
4262 3. Exim used not to create the message log file for a message until the first
4263 delivery attempt. This could be confusing when incoming messages were held
4264 for policy or load reasons. The message log file is now created at the time
4265 the message is received, and an initial "Received" line is written to it.
4266
4267 4. The automatically generated man page for command line options had a minor
4268 bug that caused no ill effects; however, a more serious problem was that
4269 the procedure for building the man page automatically didn't always
4270 operate. Consequently, release 4.20 contains an out-of-date version. This
4271 shouldn't happen again.
4272
4273 5. When building Exim with embedded Perl support, the script that builds the
4274 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4275 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4276
4277 6. The freeze_tell option was not being used for messages that were frozen on
4278 arrival, either by an ACL or by local_scan().
4279
4280 7. Added the smtp_incomplete_transaction log selector.
4281
4282 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4283 was accepting AUTH without a new EHLO.
4284
4285 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4286 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4287 RFC.
4288
4289 10. Logging of TCP/IP connections (when configured) now happens in the main
4290 daemon process instead of the child process, so that the TCP/IP connection
4291 count is more accurate (but it can never be perfect).
4292
4293 11. The use of "drop" in a nested ACL was not being handled correctly in the
4294 outer ACL. Now, if condition failure induced by the nested "drop" causes
4295 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4296 or "require"), the connection is dropped.
4297
4298 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4299 that yield "discard" can now be used with an "accept" or a "discard" verb,
4300 but an error is generated for any others (because I can't see a useful way
4301 to define what should happen).
4302
4303 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4304 are now processed in the same way as lines in the Exim configuration file.
4305 In particular, continuation lines are supported.
4306
4307 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4308
4309 15. Added -ti meaning -t -i.
4310
4311 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4312 domains, and warn by logging if others are found.
4313
4314 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4315 returned from GIFCONF when Exim is trying to find the list of interfaces on
4316 a host. The code in os.c has been modified to copy each ifreq to an aligned
4317 structure in all cases.
4318
4319 Also, in some cases, the returned ifreq's were being copied to a 'struct
4320 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4321 means the last couple of bytes of an IPv6 address could be chopped if the
4322 ifreq contained only a normal sockaddr (14 bytes storage).
4323
4324 18. Named domain lists were not supported in the hosts_treat_as_local option.
4325 An entry such as +xxxx was not recognized, and was treated as a literal
4326 domain name.
4327
4328 19. Ensure that header lines added by a DATA ACL are included in the reject log
4329 if the ACL subsequently rejects the message.
4330
4331 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4332 MD5 (which is deprecated).
4333
4334 21. When testing a filter file using -bf, Exim was writing a message when it
4335 took the sender from a "From " line in the message, but it was not doing so
4336 when it took $return_path from a Return-Path: header line. It now does.
4337
4338 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4339 with a valid header line field name (a series of printing characters
4340 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4341
4342 23. Changed "disc" in the source to "disk" to conform to the documentation and
4343 the book and for uniformity.
4344
4345 24. Ignore Sendmail's -Ooption=value command line item.
4346
4347 25. When execve() failed while trying to run a command in a pipe transport,
4348 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4349 could be confused with a return value of 69 from the command itself. This
4350 has been changed to 127, the value the shell returns if it is asked to run
4351 a non-existent command. The wording for the related log line suggests a
4352 non-existent command as the problem.
4353
4354 26. If received_header_text expands to an empty string, do not add a Received:
4355 header line to the message. (Well, it adds a token one on the spool, but
4356 marks it "old" so that it doesn't get used or transmitted.)
4357
4358 27. Installed eximstats 1.28 (addition of -nt option).
4359
4360 28. There was no check for failure on the call to getsockname() in the daemon
4361 code. This can fail if there is a shortage of resources on the system, with
4362 ENOMEM, for example. A temporary error is now given on failure.
4363
4364 29. Contrary to the C standard, it seems that in some environments, the
4365 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4366 program. Exim now does this explicitly; it affects the formatting of
4367 timestamps using strftime().
4368
4369 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4370 complaints. I've now initialized all the variables, to avoid this.
4371
4372 32. Header lines added by a system filter were not being "seen" during
4373 transport-time rewrites.
4374
4375 33. The info_callback() function passed to OpenSSL is set up with type void
4376 (*)(SSL *, int, int), as described somewhere. However, when calling the
4377 function (actually a macro) that sets it up, the type void(*)() is
4378 expected. I've put in a cast to prevent warnings from picky compilers.
4379
4380 34. If a DNS black list lookup found a CNAME record, but there were no A
4381 records associated with the domain it pointed at, Exim crashed.
4382
4383 35. If a DNS black list lookup returned more than one A record, Exim ignored
4384 all but the first. It now scans all returned addresses if a particular IP
4385 value is being sought. In this situation, the contents of the
4386 $dnslist_value variable are a list of all the addresses, separated by a
4387 comma and a space.
4388
4389 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4390 to accept a host name and all its aliases if the forward lookup for any of
4391 them yielded the IP address of the incoming connection. Now it accepts only
4392 those names whose forward lookup yields the correct IP address. Any other
4393 names are discarded. This closes a loophole whereby a rogue DNS
4394 administrator could create reverse DNS records to break through a
4395 wildcarded host restriction in an ACL.
4396
4397 37. If a user filter or a system filter that ran in a subprocess used any of
4398 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4399 the wrong values were passed to the pipe command ($thisaddress had the
4400 value of $0, $0 had the value of $1, etc). This bug was introduced by
4401 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4402
4403 38. Improved the line breaking for long SMTP error messages from ACLs.
4404 Previously, if there was no break point between 40 and 75 characters, Exim
4405 left the rest of the message alone. Two changes have been made: (a) I've
4406 reduced the minimum length to 35 characters; (b) if it can't find a break
4407 point between 35 and 75 characters, it looks ahead and uses the first one
4408 that it finds. This may give the occasional overlong line, but at least the
4409 remaining text gets split now.
4410
4411 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4412 file descriptors might be low, and that setting 1000 would always raise it.
4413 It turns out that in some environments, the limit is already over 1000 and
4414 that lowering it causes trouble. So now Exim takes care not to decrease it.
4415
4416 40. When delivering a message, the value of $return_path is set to $sender_
4417 address at the start of routing (routers may change the value). By an
4418 oversight, this default was not being set up when an address was tested by
4419 -bt or -bv, which affected the outcome if any router or filter referred to
4420 $return_path.
4421
4422 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4423 log only when "message" or "log_message" are set. However, if one of the
4424 conditions was an address verification, or a call to a nested ACL, the
4425 messages generated by the underlying test were being passed through. This
4426 no longer happens. The underlying message is available in $acl_verify_
4427 message for both "message" and "log_message" expansions, so it can be
4428 passed through if needed.
4429
4430 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4431 new expansion $bh_ to give the encoded byte string without charset
4432 translation. Translation happens only if iconv() is available; HAVE_ICONV
4433 indicates this at build time. HEADERS_CHARSET gives the charset to
4434 translate to; headers_charset can change it in the configuration, and
4435 "headers charset" can change it in an individual filter file.
4436
4437 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4438 that creates RFC 2047 encoded "words" labels them as that charset instead
4439 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4440 expansion operator; (ii) when Exim creates a From: line for a local
4441 message; (iii) when a header line is rewritten to include a "phrase" part.
4442
4443 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4444 buggy, causing it to skip the first lines of messages whose message ID
4445 ended in 'D'. This would not have bitten before Exim release 4.14, because
4446 message IDs were unlikely to end in 'D' before then. The effect was to have
4447 incorrect size information for certain domains.
4448
4449 45. #include "config.h" was missing at the start of the crypt16.c module. This
4450 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4451 noticed.
4452
4453 46. If there was a timeout during a "random" callout check, Exim treated it as
4454 a failure of the random address, and carried on sending RSET and the real
4455 address. If the delay was just some slowness somewhere, the response to the
4456 original RCPT would be taken as a response to RSET and so on, causing
4457 mayhem of various kinds.
4458
4459 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4460 when I implemented it. It didn't allow for the fact that some option values
4461 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4462 the right test for positive values.
4463
4464 48. Domain names in DNS records are case-independent. Exim always looks them up
4465 in lower case. Some resolvers return domain names in exactly the case they
4466 appear in the zone file, that is, they may contain uppercase letters. Not
4467 all resolvers do this - some return always lower case. Exim was treating a
4468 change of case by a resolver as a change of domain, similar to a widening
4469 of a domain abbreviation. This triggered its re-routing code and so it was
4470 trying to route what was effectively the same domain again. This normally
4471 caused routing to fail (because the router wouldn't handle the domain
4472 twice). Now Exim checks for this case specially, and just changes the
4473 casing of the domain that it ultimately uses when it transmits the message
4474 envelope.
4475
4476 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4477 module.
4478
4479 50. If a filter generated a file delivery with a non-absolute name (possible if
4480 no home directory exists for the router), the forbid_file option was not
4481 forbidding it.
4482
4483 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4484 the existing equality matching.
4485
4486 52. Exim was using ints instead of ino_t variables in some places where it was
4487 dealing with inode numbers.
4488
4489 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4490 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4491 called TMPDIR, and if it finds it is different, it changes its value.
4492
4493 54. The smtp_printf() function is now made available to local_scan() so
4494 additional output lines can be written before returning. There is also an
4495 smtp_fflush() function to enable the detection of a dropped connection.
4496 The variables smtp_input and smtp_batched_input are exported to
4497 local_scan().
4498
4499 55. Changed the default runtime configuration: the message "Unknown user"
4500 has been removed from the ACL, and instead placed on the localuser router,
4501 using the cannot_route_message feature. This means that any verification
4502 failures that generate their own messages won't get overridden. Similarly,
4503 the "Unrouteable address" message that was in the ACL for unverifiable
4504 relay addresses has also been removed.
4505
4506 56. Added hosts_avoid_esmtp to the smtp transport.
4507
4508 57. The exicyclog script was not checking for the esoteric option
4509 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4510 will work only if exicyclog is run under the appropriate euid.
4511
4512 58. Following a discussion on the list, the rules by which Exim recognises line
4513 endings on incoming messages have been changed. The -dropcr and drop_cr
4514 options are now no-ops, retained only for backwards compatibility. The
4515 following line terminators are recognized: LF CRLF CR. However, special
4516 processing applies to CR:
4517
4518 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4519 nor a local message in the state where . is a terminator.
4520
4521 (ii) If a bare CR is encountered in a header line, an extra space is added
4522 after the line terminator so as not to end the header. The reasoning
4523 behind this is that bare CRs in header lines are most likely either
4524 to be mistakes, or people trying to play silly games.
4525
4526 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4527 was being incorrectly given as 18 bytes larger than it should have been.
4528 This is a VOB (very old bug).
4529
4530 60. This may never have affected anything current, but just in case it has:
4531 When the local host is found other than at the start of a list of hosts,
4532 the local host, those with the same MX, and any that follow, are discarded.
4533 When the list in question was part of a longer list of hosts, the following
4534 hosts (not currently being processed) were also being discarded. This no
4535 longer happens. I'm not sure if this situation could ever has previously
4536 arisen.
4537
4538 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4539 program routers.
4540
4541 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4542 header. This is something that is recommended in a new Internet Draft, and
4543 is something that is documented as being done by Sendmail. There are two
4544 possible values. For messages generated by the autoreply transport, Exim
4545 adds:
4546
4547 Auto-Submitted: auto-replied
4548
4549 whereas for all other generated messages (e.g. bounces) it adds
4550
4551 Auto-Submitted: auto-generated
4552
4553 63. The "personal" condition in filters now includes a test for the
4554 Auto-Submitted: header. If it contains the string "auto-" the message it
4555 not considered personal.
4556
4557 64. Added rcpt_include_affixes as a generic transport option.
4558
4559 65. Added queue_only_override (default true).
4560
4561 66. Added the syslog_duplication option.
4562
4563 67. If what should have been the first header line of a message consisted of
4564 a space followed by a colon, Exim was mis-interpreting it as a header line.
4565 It isn't of course - it is syntactically invalid and should therefore be
4566 treated as the start of the message body. The misbehaviour could have
4567 caused a number of strange effects, including loss of data in subsequent
4568 header lines, and spool format errors.
4569
4570 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4571 client host had authenticated. This control can now be exercised by an ACL
4572 for more flexibility.
4573
4574 69. By default, callouts do not happen when testing with -bh. There is now a
4575 variant, -bhc, which does actually run the callout code, including
4576 consulting and updating the callout cache.
4577
4578 70. Added support for saslauthd authentication, courtesy of Alexander
4579 Sabourenkov.
4580
4581 71. If statvfs() failed on the spool or log directories while checking their
4582 size for availability, Exim confusingly gave the error "space shortage".
4583 Furthermore, in debugging mode it crashed with a floating point exception.
4584 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4585 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4586 serious problem, Exim now writes to the main and panic logs when this
4587 happens, with details of the failure. It then refuses to accept the
4588 incoming message, giving the message "spool directory problem" or "log
4589 directory problem" with a 421 code for SMTP messages.
4590
4591 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4592 0, 1, and 2 exist, because some OS complain for execs without them (see
4593 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4594 descriptors. However, the code omitted to check that the open succeeded,
4595 causing mysterious errors if for some reason the permissions on /dev/null
4596 got screwed. Now Exim writes a message to the main and panic logs, and
4597 bombs out if it can't open /dev/null.
4598
4599 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4600 interact so that it is all more flexible. It is supposed to remain
4601 backwards compatible. Also added extra_local_interfaces.
4602
4603 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4604 to bomb out with an assertion failure - to the client this appears as a
4605 connection drop. This problem occurs in the part of the code that was taken
4606 from the Samba project. Fortunately, the assertion is in a very simple
4607 function, so I have fixed this by reproducing the function inline in the
4608 one place where it is called, and arranging for authentication to fail
4609 instead of killing the process with assert().
4610
4611 75. The SPA client code was not working when the server requested OEM rather
4612 than Unicode encoding.
4613
4614 76. Added code to make require_files with a specific uid setting more usable in
4615 the case where statting the file as root fails - usually a non-root-mounted
4616 NFS file system. When this happens and the failure is EACCES, Exim now
4617 forks a subprocess and does the per-uid checking as the relevant uid.
4618
4619 77. Added process_log_path.
4620
4621 78. If log_file_path was not explicitly set, a setting of check_log_space or
4622 check_log_inodes was ignored.
4623
4624 79. If a space check for the spool or log partitions fails, the incident is now
4625 logged. Of course, in the latter case the data may get lost...
4626
4627 80. Added the %p formatting code to string_format() so that it can be used to
4628 print addresses in debug_print(). Adjusted all the address printing in the
4629 debugging in store.c to use %p rather than %d.
4630
4631 81. There was a concern that a line of code in smtp_in.c could overflow a
4632 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4633 initially expressed, the concern was not well-founded, because trailing
4634 spaces are removed early. However, if the trailing spaces were followed by
4635 a NULL, they did not get removed, so the overflow was possible. Two fixes
4636 were applied:
4637
4638 (a) I re-wrote the offending code in a cleaner fashion.
4639 (b) If an incoming SMTP command contains a NULL character, it is rejected
4640 as invalid.
4641
4642 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4643 runs initgroups(), so that if the Exim user is in any additional groups,
4644 they will be used during message reception.
4645
4646
4647 Exim version 4.20
4648 -----------------
4649
4650 The change log for 4.20 and earlier releases has been archived.
4651
4652 ****