Check the length of off_t at build time and use %ld or %lld to print
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
b1c749bb 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.157 2005/06/16 14:10:13 ph10 Exp $
495ae4b0
PH
2
3Change log file for Exim from version 4.21
4-------------------------------------------
5
bbe902f0 6
e5a9dba6
PH
7Exim version 4.52
8-----------------
9
10TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
11
22c3b60b
PH
12PH/01 When a transport filter timed out in a pipe delivery, and the pipe
13 command itself ended in error, the underlying message about the transport
14 filter timeout was being overwritten with the pipe command error. Now the
15 underlying error message should be appended to the second error message.
16
06a9b4b5
PH
17TK/01 Fix poll() being unavailable on Mac OSX 10.2.
18
c1ac6996
PH
19PH/02 Reduce the amount of output that "make" produces by default. Full output
20 can still be requested.
21
9c7a242c
PH
22PH/03 The warning log line about a condition test deferring for a "warn" verb
23 was being output only once per connection, rather than after each
24 occurrence (because it was using the same function as for successful
25 "warn" verbs). This seems wrong, so I have changed it.
26
87ba3f5f
PH
27TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
28 it should not have, which might have caused a crash in the right
29 circumstances, but probably never did.
30
31PH/04 Installed a modified version of Tony Finch's patch to make submission
32 mode fix the return path as well as the Sender: header line, and to
33 add a /name= option so that you can make the user's friendly name appear
34 in the header line.
35
29aba418
TF
36TF/03 Added the control = fakedefer ACL modifier.
37
fe0dab11
TF
38TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
39 Mark Lowes for thorough testing.
870f6ba8 40
11d337a4
TK
41TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
42
43TK/03 Merged latest SRS patch from Miles Wilton.
44
415c8f3b
PH
45PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
46 with the definition in sysexits.h (which is #included earlier).
47 Fortunately, Exim does not actually use EX_OK. The code used to try to
48 preserve the sysexits.h value, by assumimg that macro definitions were
49 scanned for macro replacements. I have been disabused of this notion,
50 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 51
958541e9
PH
52PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
53 in the smtp transport. When a block could not be written in a single
54 write() function, the timeout was being re-applied to each part-write.
55 This seems wrong - if the receiver was accepting one byte at a time it
56 would take for ever. The timeout is now adjusted when this happens. It
57 doesn't have to be particularly precise.
58
c206415f
TK
59TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
60 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
61
2a4be8f9
PH
62PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
63 without a domain if the check to <postmaster@domain> fails.
64
1cba11c5
SC
65SC/01 Eximstats: added -xls and the ability to specify output files
66 (patch written by Frank Heydlauf).
67
68SC/02 Eximstats: use FileHandles for outputing results.
69
70SC/03 Eximstats: allow any combination of xls, txt, and html output.
71
72SC/04 Eximstats: fixed display of large numbers with -nvr option
73
74SC/05 Eximstats: fixed merging of reports with empty tables.
75
76SC/06 Eximstats: added the -include_original_destination flag
77
78SC/07 Eximstats: removed tabs and trailing whitespace.
79
1005d00e
TK
80TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
81
82TK/06 MBOX spool code: Add real "From " MBOX separator line
83 so the .eml file is really in mbox format (even though
84 most programs do not really care). Patch from Alex Miller.
85
86TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
87 The latter is generated from $received_to and is only set if the
88 message has one envelope recipient. SA can use these headers,
89 obviously out-of-the-box. Patch from Alex Miller.
90
9b4768fa
PH
91PH/08 The ${def test on a variable was returning false if the variable's
92 value was "0", contrary to what the specification has always said!
93 The result should be true unless the variable is empty.
94
95PH/09 The syntax error of a character other than { following "${if
96 def:variable_name" (after optional whitespace) was not being diagnosed.
97 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
98 accidental colon was present, for example, could give incorrect results.
99
0d7eb84a
PH
100PH/10 Tidied the code in a number of places where the st_size field of a stat()
101 result is used (not including appendfile, where other changes are about
102 to be made).
103
104PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
105 This involved changing a lot of size variables from int to off_t. It
106 should work with maildirs and everything.
107
40727bee
TK
108TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
109 spamd dying while we are connected to it.
110
554d2369
TF
111TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
112 <jgh@wizmail.org>
113
1f922db1
PH
114PH/12 Applied Alex Kiernan's patch for the API change for the error callback
115 function for BDB 4.3.
116
ef213c3b
PH
117PH/13 Changed auto_thaw such that it does not apply to bounce messages.
118
8ac170f3
PH
119PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
120 the sources for PCRE have been re-arranged and more files are now
121 involved.
122
b1c749bb
PH
123PH/15 The code I had for printing potentially long long variables in PH/11
124 above was not the best (it lost precision). The length of off_t variables
125 is now inspected at build time, and an appropriate printing format (%ld
126 or %lld) is chosen and #defined by OFF_T_FMT. We also define ASSUME_
127 LONG_LONG_SUPPORT if the length is greater than 4. This is needed for the
128 internal formatting function string_vformat().
129
e5a9dba6 130
7982096b
PH
131Exim version 4.51
132-----------------
133
1a46a8c5
PH
134TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
135 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
136
2f079f46 137TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
138
139TK/03 Fix ACL "regex" condition not being available in MIME ACL.
140
141PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
142 to test Sieve filters that use "vacation".
143
144PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
145 that changes the way the GnuTLS parameters are stored in the cache file.
146 The new format can be generated externally. For backward compatibility,
147 if the data in the cache doesn't make sense, Exim assumes it has read an
148 old-format file, and it generates new data and writes a new file. This
149 means that you can't go back to an older release without removing the
150 file.
151
152PH/03 A redirect router that has both "unseen" and "one_time" set does not
153 work if there are any delivery delays because "one_time" forces the
154 parent to be marked "delivered", so its unseen clone is never tried
155 again. For this reason, Exim now forbids the simultaneous setting of
156 these two options.
157
158PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
159 redirected to themselves ("homonym" addresses). Read the long ChangeLog
160 entry if you want to know the details. The fix, however, neglected to
161 consider the case when local delivery batching is involved. The test for
162 "previously delivered" was not happening when checking to see if an
163 address could be batched with a previous (undelivered) one; under
164 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 165 address.
1a46a8c5
PH
166
167PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
168 in its include files, and this causes problems building Exim.
169
170PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
171 header_syntax) but Exim was just ignoring anything given after a slash.
172 In particular, this caused confusion with an attempt to use "verify =
173 reverse_host_lookup/defer_ok". An error is now given when options are
174 supplied for verify items that do not have them. (Maybe reverse_host_
175 lookup should have a defer_ok option, but that's a different point.)
176
177PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
178 defined by RFC 821) to 2048, because there were problems with some AUTH
179 commands, and RFC 1869 says the size should be increased for extended
180 SMTP commands that take arguments.
181
182PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
183 Finch).
184
185PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
186 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 187
49c2d5ea
PH
188PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
189 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
190 whatever). Otherwise the password may appear in the log. Following change
191 PH/42 below, there is no longer a chance of it appearing in a bounce
192 message.
49c2d5ea 193
bf759a8b
PH
194PH/11 Installed exipick version 20050225.0 from John Jetmore.
195
83364d30
PH
196PH/12 If the last host in a fallback_hosts list was multihomed, only the first
197 of its addresses was ever tried. (Bugzilla bug #2.)
198
7999bbd7
PH
199PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
200 the result incorrectly in the debug output. (It correctly added a newline
201 to what was transported.)
202
7dbf77c9
PH
203TF/01 Added $received_time.
204
74e0617f
PH
205PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
206 commented out examples of how to interface to a virus scanner and to
207 SpamAssassin. Also added commented examples of av_scanner and
208 spamd_address settings.
209
2f079f46
PH
210PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
211 and controls are allowed in which ACLs. There were a couple of minor
212 errors. Some of the entries in the conditions table (which is a table of
213 where they are NOT allowed) were getting very unwieldy; rewrote them as a
214 negation of where the condition IS allowed.
215
8c841523
PH
216PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
217
7766a4f0
PH
218PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
219 header file does not have a version number, so I've had to invent a new
220 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
221 API. The code is untested by me (my Linux distribution still has 0.3.2 of
222 radiusclient), but it was contributed by a Radius user.
223
8b417f2c
PH
224PH/18 Installed Lars Mainka's patch for the support of CRL collections in
225 files or directories, for OpenSSL.
226
901f42cb
PH
227PH/19 When an Exim process that is running as root has to create an Exim log
228 file, it does so in a subprocess that runs as exim:exim so as to get the
229 ownership right at creation (otherwise, other Exim processes might see
230 the file with the wrong ownership). There was no test for failure of this
231 fork() call, which would lead to the process getting stuck as it waited
232 for a non-existent subprocess. Forks do occasionally fail when resources
233 run out. I reviewed all the other calls to fork(); they all seem to check
234 for failure.
235
f9b9210e
PH
236PH/20 When checking for unexpected SMTP input at connect time (before writing
237 the banner), Exim was not dealing correctly with a non-positive return
238 from the read() function. If the client had disconnected by this time,
239 the result was a log entry for a synchronization error with an empty
240 string after "input=" when read() returned zero. If read() returned -1
241 (an event I could not check), uninitialized data bytes were printed.
242 There were reports of junk text (parts of files, etc) appearing after
243 "input=".
244
54cdb463
PH
245PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
246
cf00dad6
PH
247PH/22 Added support for macro redefinition, and (re)definition in between
248 driver and ACL definitions.
249
acb1b346
PH
250PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
251 forgetting to use the resulting value; it was using the unexpanded value.
252
c5ddb310
PH
253PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
254 hadn't been configured. The fix is from Juergen Kreileder, who
255 understands it better than I do:
256
257 "Here's what I see happening with three configured cyrus_sasl
258 authenticators configured (plain, login, cram-md5):
259
260 On startup auth_cyrus_sasl_init() gets called for each of these.
261 This means three calls to sasl_listmech() without a specified mech_list.
262 => SASL tests which mechs of all available mechs actually work
263 => three warnings about OTP not working
264 => the returned list contains: plain, login, cram-md5, digest-md5, ...
265
266 With the patch, sasl_listmech() also gets called three times. But now
267 SASL's mech_list option is set to the server_mech specified in the the
268 authenticator. Or in other words, the answer from sasl_listmech()
269 gets limited to just the mech you're testing for (which is different
270 for each call.)
271 => the return list contains just 'plain' or 'login', 'cram-md5' or
272 nothing depending on the value of ob->server_mech.
273
274 I've just tested the patch: Authentication still works fine,
275 unavailable mechs specified in the exim configuration are still
276 caught, and the auth.log warnings about OTP are gone."
277
31619da6
PH
278PH/25 When debugging is enabled, the contents of the command line are added
279 to the debugging output, even when log_selector=+arguments is not
280 specified.
281
bebaf0fc
PH
282PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
283 answer is "GNU", and only if the return is "GNU/something" is the answer
284 "Linux".
285
475fe28a
PH
286PH/27 $acl_verify_message is now set immediately after the failure of a
287 verification in an ACL, and so is available in subsequent modifiers. In
288 particular, the message can be preserved by coding like this:
289
290 warn !verify = sender
291 set acl_m0 = $acl_verify_message
292
293 Previously, $acl_verify_message was set only while expanding "message"
294 and "log_message" when a very denied access.
295
7e8bec7a
PH
296PH/28 Modified OS/os.c-Linux with
297
298 -#ifndef OS_LOAD_AVERAGE
299 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
300
301 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
302 nomenclature these days.)
303
e4a89c47
PH
304PH/29 Installed patch from the Sieve maintainer that adds the options
305 sieve_useraddress and sieve_subaddress to the redirect router.
306
5ca2a9a1
PH
307PH/30 In these circumstances:
308 . Two addresses routed to the same list of hosts;
309 . First host does not offer TLS;
310 . First host accepts first address;
311 . First host gives temporary error to second address;
312 . Second host offers TLS and a TLS session is established;
313 . Second host accepts second address.
314 Exim incorrectly logged both deliveries with the TLS parameters (cipher
315 and peerdn, if requested) that were in fact used only for the second
316 address.
7e8bec7a 317
c688b954
PH
318PH/31 When doing a callout as part of verifying an address, Exim was not paying
319 attention to any local part prefix or suffix that was matched by the
320 router that accepted the address. It now behaves in the same way as it
321 does for delivery: the affixes are removed from the local part unless
322 rcpt_include_affixes is set on the transport.
323
fed77020
PH
324PH/32 Add the sender address, as F=<...>, to the log line when logging a
325 timeout during the DATA phase of an incoming message.
326
7fe1560f
PH
327PH/33 Sieve envelope tests were broken for match types other than :is. I have
328 applied a patch sanctioned by the Sieve maintainer.
c688b954 329
ebb6e6d5
PH
330PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
331 the uid or gid is negative. A case of a negative gid caused this to be
332 noticed. The fix allows for either to be negative.
333
9c4e8f60
PH
334PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
335 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
336 overlooked.
337
338PH/36 The change PH/12 above was broken. Fixed it.
339
d7174846
PH
340PH/37 Exim used to check for duplicate addresses in the middle of routing, on
341 the grounds that routing the same address twice would always produce the
342 same answer. This might have been true once, but it is certainly no
343 longer true now. Routing a child address may depend on the previous
344 routing that produced that child. Some complicated redirection strategies
345 went wrong when messages had multiple recipients, and made Exim's
346 behaviour dependent on the order in which the addresses were given.
347
348 I have moved the duplicate checking until after the routing is complete.
349 Exim scans the addresses that are assigned to local and remote
350 transports, and removes any duplicates. This means that more work will be
351 done, as duplicates will always all be routed, but duplicates are
352 presumably rare, so I don't expect this is of any significance.
353
354 For deliveries to pipes, files, and autoreplies, the duplicate checking
355 still happens during the routing process, since they are not going to be
356 routed further.
357
cfe75fc3
PH
358PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
359 It corrects a timeout issue with spamd. This is Ian's comment: "The
360 background is that sometimes spamd either never reads data from a
361 connection it has accepted, or it never writes response data. The exiscan
362 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
363 blindly assumes that writes won't block so it may never time out."
364
be22d70e
PH
365PH/39 Allow G after quota size as well as K and M.
366
0612b098
PH
367PH/40 The value set for $authenticated_id in an authenticator may not contain
368 binary zeroes or newlines because the value is written to log lines and
369 to spool files. There was no check on this. Now the value is run through
370 the string_printing() function so that such characters are converted to
371 printable escape sequences.
372
2e0c1448
PH
373PH/41 $message_linecount is a new variable that contains the total number of
374 lines in the message. Compare $body_linecount, which is the count for the
375 body only.
376
447d236c
PH
377PH/42 Exim no longer gives details of delivery errors for specific addresses in
378 bounce and delay warning messages, except in certain special cases, which
379 are as follows:
380
381 (a) An SMTP error message from a remote host;
382 (b) A message specified in a :fail: redirection;
383 (c) A message specified in a "fail" command in a system filter;
384 (d) A message specified in a FAIL return from the queryprogram router;
385 (e) A message specified by the cannot_route_message router option.
386
387 In these cases only, Exim does include the error details in bounce and
388 warning messages. There are also a few cases where bland messages such
389 as "unrouteable address" or "local delivery error" are given.
390
d20976dc
PH
391PH/43 $value is now also set for the "else" part of a ${run expansion.
392
f656d135
PH
393PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
394 being worked on, but at least Exim now implements the latest version to
395 play with."
396
2e2a30b4
PH
397PH/45 In a pipe transport, although a timeout while waiting for the pipe
398 process to complete was treated as a delivery failure, a timeout while
399 writing the message to the pipe was logged, but erroneously treated as a
400 successful delivery. Such timeouts include transport filter timeouts. For
401 consistency with the overall process timeout, these timeouts are now
402 treated as errors, giving rise to delivery failures by default. However,
403 there is now a new Boolean option for the pipe transport called
404 timeout_defer, which, if set TRUE, converts the failures into defers for
405 both kinds of timeout. A transport filter timeout is now identified in
406 the log output.
407
9176e9f0
PH
408PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
409 systems where "make" and "gmake" are different, calling "gmake" at top
410 level broke things. I've arranged for the value of $(MAKE) to be passed
411 from the Makefile to this script so that it can call the same version of
412 "make".
413
7982096b 414
bbe902f0
PH
415A note about Exim versions 4.44 and 4.50
416----------------------------------------
417
418Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
419changes of various kinds. As a consequence, a big documentation update was
420needed. This delayed the release for rather longer than seemed good, especially
421in the light of a couple of (minor) security issues. Therefore, the changes
422that fixed bugs were backported into 4.43, to create a 4.44 maintenance
423release. So 4.44 and 4.50 are in effect two different branches that both start
424from 4.43.
425
426I have left the 4.50 change log unchanged; it contains all the changes since
4274.43. The change log for 4.44 is below; many of its items are identical to
428those for 4.50. This seems to be the most sensible way to preserve the
429historical information.
430
431
f7b63901 432Exim version 4.50
495ae4b0
PH
433-----------------
434
5fe762f6
PH
435 1. Minor wording change to the doc/README.SIEVE file.
436
139059f6 437 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 438 computation of the current number of files was incorrect.
495ae4b0 439
7086e875
PH
440 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
441 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
442 place.
443
35af9f61
PH
444 4. Give more explanation in the error message when the command for a transport
445 filter fails to execute.
446
b668c215
PH
447 5. There are several places where Exim runs a non-Exim command in a
448 subprocess. The SIGUSR1 signal should be disabled for these processes. This
449 was being done only for the command run by the queryprogram router. It is
450 now done for all such subprocesses. The other cases are: ${run, transport
451 filters, and the commands run by the lmtp and pipe transports.
452
a494b1e1
PH
453 6. Added CONFIGURE_GROUP build-time option.
454
455 7. Some older OS have a limit of 256 on the maximum number of file
456 descriptors. Exim was using setrlimit() to set 1000 as a large value
457 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
458 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 459
c5fcb476
PH
460 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
461 was an oversight, and furthermore, ever since the addition of extra
462 controls (e.g. 4.43/32), the checks on when to allow different forms of
463 "control" were broken. There should now be diagnostics for all cases when a
464 control that does not make sense is encountered.
465
69358f02
PH
466 9. Added the /retain_sender option to "control=submission".
467
5be20824
PH
46810. $recipients is now available in the predata ACL (oversight).
469
eb2c0248
PH
47011. Tidy the search cache before the fork to do a delivery from a message
471 received from the command line. Otherwise the child will trigger a lookup
472 failure and thereby defer the delivery if it tries to use (for example) a
473 cached ldap connection that the parent has called unbind on.
474
2a3eea10
PH
47512. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
476 of $address_data from the recipient verification was clobbered by the
477 sender verification.
478
47913. The value of address_data from a sender verification is now available in
480 $sender_address_data in subsequent conditions in the ACL statement.
481
23c7ff99
PH
48214. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
483
4deaf07d
PH
48415. Added a new option "connect=<time>" to callout options, to set a different
485 connection timeout.
486
926e1192
PH
48716. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
488 was its contents. (It was OK if the option was not defined at all.)
489
650edc6f
PH
49017. A "Completed" log line is now written for messages that are removed from
491 the spool by the -Mrm option.
492
2c7db3f5
PH
49318. New variables $sender_verify_failure and $recipient_verify_failure contain
494 information about exactly what failed.
495
3d235903
PH
49619. Added -dd to debug only the daemon process.
497
7c7ad977
PH
49820. Incorporated Michael Haardt's patch to ldap.c for improving the way it
499 handles timeouts, both on the server side and network timeouts. Renamed the
500 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
501
981756db
PH
50221. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
503 It is now set to "smtps".
504
d4eb88df
PH
50522. $host_address is now set to the target address during the checking of
506 ignore_target_hosts.
507
50823. When checking ignore_target_hosts for an ipliteral router, no host name was
509 being passed; this would have caused $sender_host_name to have been used if
510 matching the list had actually called for a host name (not very likely,
511 since this list is usually IP addresses). A host name is now passed as
512 "[x.x.x.x]".
513
7d468ab8
PH
51424. Changed the calls that set up the SIGCHLD handler in the daemon to use the
515 code that specifies a non-restarting handler (typically sigaction() in
516 modern systems) in an attempt to fix a rare and obscure crash bug.
517
51825. Narrowed the window for a race in the daemon that could cause it to ignore
519 SIGCHLD signals. This is not a major problem, because they are used only to
520 wake it up if nothing else does.
521
62c0818f
PH
52226. A malformed maildirsize file could cause Exim to calculate negative values
523 for the mailbox size or file count. Odd effects could occur as a result.
524 The maildirsize information is now recalculated if the size or filecount
525 end up negative.
526
26034054
PH
52727. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
528 support for a long time. Removed HAVE_SYS_VFS_H.
529
af66f652
PH
53028. Installed the latest version of exipick from John Jetmore.
531
90af77f4
PH
53229. In an address list, if the pattern was not a regular expression, an empty
533 subject address (from a bounce message) matched only if the pattern was an
534 empty string. Non-empty patterns were not even tested. This was the wrong
535 because it is perfectly reasonable to use an empty address as part of a
536 database query. An empty address is now tested by patterns that are
537 lookups. However, all the other forms of pattern expect the subject to
538 contain a local part and a domain, and therefore, for them, an empty
539 address still always fails if the pattern is not itself empty.
540
d8ef3577
PH
54130. Exim went into a mad DNS loop when attempting to do a callout where the
542 host was specified on an smtp transport, and looking it up yielded more
543 than one IP address.
544
5cb8cbc6
PH
54531. Re-factored the code for checking spool and log partition space into a
546 function that finds that data and another that does the check. The former
547 is then used to implement four new variables: $spool_space, $log_space,
548 $spool_inodes, and $log_inodes.
549
14702f5b
PH
55032. The RFC2047 encoding function was originally intended for short strings
551 such as real names; it was not keeping to the 75-character limit for
552 encoded words that the RFC imposes. It now respects the limit, and
553 generates multiple encoded words if necessary. To be on the safe side, I
554 have increased the buffer size for the ${rfc2047: expansion operator from
555 1024 to 2048 bytes.
556
063b1e99
PH
55733. It is now permitted to omit both strings after an "if" condition; if the
558 condition is true, the result is "true". As before, when the second string
559 is omitted, a false condition yields an empty string. This makes it less
560 cumbersome to write custom ACL and router conditions.
561
652e1b65
PH
56234. Failure to deliver a bounce message always caused it to be frozen, even if
563 there was an errors_to setting on the router. The errors_to setting is now
564 respected.
565
6f0c9a4f
PH
56635. If an IPv6 address is given for -bh or -bhc, it is now converted to the
567 canonical form (fully expanded) before being placed in
568 $sender_host_address.
569
33397d19
PH
57036. The table in the code that translates DNS record types into text (T_A to
571 "A" for instance) was missing entries for NS and CNAME. It is just possible
572 that this could have caused confusion if both these types were looked up
573 for the same domain, because the text type is used as part of Exim's
574 per-process caching. But the chance of anyone hitting this buglet seems
575 very small.
576
7bb56e1f
PH
57737. The dnsdb lookup has been extended in a number of ways.
578
579 (1) There is a new type, "zns", which walks up the domain tree until it
580 finds some nameserver records. It should be used with care.
581
ea3bc19b
PH
582 (2) There is a new type, "mxh", which is like "mx" except that it returns
583 just the host names, not the priorities.
584
585 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
586 looked up. The behaviour when one of the lookups defers can be
587 controlled by a keyword.
7bb56e1f 588
ea3bc19b 589 (4) It is now possible to specify the separator character for use when
7bb56e1f 590 multiple records are returned.
33397d19 591
0bcb2a0e
PH
59238. The dnslists ACL condition has been extended: it is now possible to supply
593 a list of IP addresses and/or domains to be looked up in a particular DNS
594 domain.
595
2ac0e484
PH
59639. Added log_selector=+queue_time_overall.
597
4e1fde53
PH
59840. When running the queue in the test harness, wait just a tad after forking a
599 delivery process, to get repeatability of debugging output.
600
de365ded
PH
60141. Include certificate and key file names in error message when GnuTLS fails
602 to set them up, because the GnuTLS error message doesn't include the name
603 of the failing file when there is a problem reading it.
604
f05da2e8
PH
60542. Allow both -bf and -bF in the same test run.
606
d6453af2
PH
60743. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
608
f7b63901
PH
60944. The "Exiscan patch" is now merged into the mainline Exim source.
610
61145. Sometimes the final signoff response after QUIT could fail to get
612 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
613 < 0 before doing a fflush(). This bug looks as though it goes back to the
614 introduction of TLS in release 3.20, but "sometimes" must have been rare
615 because the tests only now provoked it.
616
a444213a
PH
61746. Reset the locale to "C" after calling embedded Perl, in case it was changed
618 (this can affect the format of dates).
619
0ec020ea
PH
62047. exim_tidydb, when checking for the continued existence of a message for
621 which it has found a message-specific retry record, was not finding
622 messages that were in split spool directories. Consequently, it was
623 deleting retry records that should have stayed in existence.
624
b1206957
PH
62548. Steve fixed some bugs in eximstats.
626
62749. The SPA authentication driver was not abandoning authentication and moving
628 on to the next authenticator when an expansion was forced to fail,
629 contradicting the general specification for all authenticators. Instead it
630 was generating a temporary error. It now behaves as specified.
631
26dd5a95
PH
63250. The default ordering of permitted cipher suites for GnuTLS was pessimal
633 (the order specifies the preference for clients). The order is now AES256,
634 AES128, 3DES, ARCFOUR128.
635
343b2385
PH
63651. Small patch to Sieve code - explicitly set From: when generating an
637 autoreply.
638
1c5466b9
PH
63952. Exim crashed if a remote delivery caused a very long error message to be
640 recorded - for instance if somebody sent an entire SpamAssassin report back
641 as a large number of 550 error lines. This bug was coincidentally fixed by
642 increasing the size of one of Exim's internal buffers (big_buffer) that
643 happened as part of the Exiscan merge. However, to be on the safe side, I
644 have made the code more robust (and fixed the comments that describe what
645 is going on).
646
55ee9ee3
PH
64753. Now that there can be additional text after "Completed" in log lines (if
648 the queue_time_overall log selector is set), a one-byte patch to exigrep
649 was needed to allow it to recognize "Completed" as not the last thing in
650 the line.
651
d38f8232
PH
65254. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
653 patch that reportedly fixes this has been added. I am not expert enough to
654 create a test for it. This is what the patch creator wrote:
655
656 "I found a little strange behaviour of ldap code when working with
657 Windows 2003 AD Domain, where users was placed in more than one
658 Organization Units. When I tried to give exim partial DN, the exit code
659 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
660 But simultaneously result of request was absolutely normal ldap result,
661 so I produce this patch..."
662
3295e65b
PH
663 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
664 REFERENCE, so I have modified the code to exclude the patch when that macro
665 is not defined.
666
7102e136
PH
66755. Some experimental protocols are using DNS PTR records for new purposes. The
668 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
669 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
670 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 671 CAN-2005-0021
7102e136 672
3ca0ba97
PH
67356. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
674
c2bcbe20
PH
67557. Double the size of the debug message buffer (to 2048) so that more of very
676 long debug lines gets shown.
677
18ce445d
PH
67858. The exicyclog utility now does better if the number of log files to keep
679 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
680
1f5b4c3d
PH
68159. Two changes related to the smtp_active_hostname option:
682
683 (1) $smtp_active_hostname is now available as a variable.
684 (2) The default for smtp_banner uses $smtp_active_hostname instead
685 of $primary_hostname.
686
b975ba52
PH
68760. The host_aton() function is supposed to be passed a string that is known
688 to be a valid IP address. However, in the case of IPv6 addresses, it was
689 not checking this. This is a hostage to fortune. Exim now panics and dies
690 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
691 from a dnsdb PTR lookup with an IPv6 address that had more than 8
692 components; fortuitously, this particular loophole had already been fixed
693 by change 4.50/55 above.
694
695 If there are any other similar loopholes, the new check in host_aton()
696 itself should stop them being exploited. The report I received stated that
697 data on the command line could provoke the exploit when Exim was running as
698 exim, but did not say which command line option was involved. All I could
699 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
700 running as the user.
ea3a6f44 701 CAN-2005-0021
85b87bc2
PH
702
70361. There was a buffer overflow vulnerability in the SPA authentication code
704 (which came originally from the Samba project). I have added a test to the
705 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 706 CAN-2005-0022
b975ba52 707
17ffcae7
PH
70862. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
709 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
710
d95f9fdb
PH
71163. The daemon start-up calls getloadavg() while still root for those OS that
712 need the first call to be done as root, but it missed one case: when
713 deliver_queue_load_max is set with deliver_drop_privilege. This is
714 necessary for the benefit of the queue runner, because there is no re-exec
715 when deliver_drop_privilege is set.
716
86b8287f
PH
71764. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
718 This has been fixed.
719
60dc5e56
PH
72065. Caching of lookup data for "hosts =" ACL conditions, when a named host list
721 was in use, was not putting the data itself into the right store pool;
722 consequently, it could be overwritten for a subsequent message in the same
723 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
724 the caching.)
725
533244af
PH
72666. Added hosts_max_try_hardlimit to the smtp transport, default 50.
727
a5a28604
PH
72867. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
729 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
730 the function were treating the return as a boolean value, which happened to
731 work because 0=false and not-0=true, but is not correct code.
732
7e634d24
PH
73368. The host_aton() function was not handling scoped IPv6 addresses (those
734 with, for example, "%eth0" on the end) correctly.
735
3e11c26b
PH
73669. Fixed some compiler warnings in acl.c for the bitmaps specified with
737 negated items (that is, ~something) in unsigned ints. Some compilers
738 apparently mutter when there is no cast.
739
6729cf78
PH
74070. If an address verification called from an ACL failed, and did not produce a
741 user-specific message (i.e. there was only a "system" message), nothing was
742 put in $acl_verify_message. In this situation, it now puts the system
743 message there.
744
00f00ca5
PH
74571. Change 4.23/11 added synchronization checking at the start of an SMTP
746 session; change 4.31/43 added the unwanted input to the log line - except
747 that it did not do this in the start of session case. It now does.
748
c9bdd01c
PH
74972. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
750 This is wrong and can cause the other end to generate a synchronization
751 error if it is another Exim or anything else that does the synchronization
752 check. A QUIT command is no longer sent after a timeout.
753
d43194df
PH
75473. $host_lookup_deferred has been added, to make it easier to detect DEFERs
755 during host lookups.
756
fe5b5d0b
PH
75774. The defer_ok option of callout verification was not working if it was used
758 when verifying addresses in header lines, that is, for this case:
759
760 verify = header_sender/callout=defer_ok
761
76a2d7ba
PH
76275. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
763 those file descriptors could be used for SMTP connections. If anything
764 wrote to stderr (the example that came up was "warn" in embedded Perl), it
765 could be sent to the SMTP client, causing chaos. The daemon now opens
766 stdin, stdout, and stderr to /dev/null when it puts itself into the
767 background.
768
76976. Arrange for output from Perl's "warn" command to be written to Exim's main
770 log by default. The user can override this with suitable Perl magic.
771
04f7d5b9
PH
77277. The use of log_message on a "discard" ACL verb, which is supposed to add to
773 the log message when discard triggers, was not working for the DATA ACL or
774 for the non-SMTP ACL.
775
bc60667e
PH
77678. Error message wording change in sieve.c.
777
bb6e88ff
PH
77879. If smtp_accept_max_per_host was set, the number of connections could be
779 restricted to fewer than expected, because the daemon was trying to set up
780 a new connection before checking whether the processes handling previous
781 connections had finished. The check for completed processes is now done
782 earlier. On busy systems, this bug wouldn't be noticed because something
783 else would have woken the daemon, and it would have reaped the completed
784 process earlier.
785
1e70f85b
PH
78680. If a message was submitted locally by a user whose login name contained one
787 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
788 It caused a spool format error. I have fixed the spool reading code. A
789 related problem was that the "from" clause in the Received: line became
790 illegal because of the space(s). It is now covered by ${quote_local_part.
791
79281. Included the latest eximstats from Steve (adds average sizes to HTML Top
793 tables).
794
4e01f9d6
PH
79582. Updated OS/Makefile-AIX as per message from Mike Meredith.
796
1ee1cef2
PH
79783. Patch from Sieve maintainer to fix unterminated string problem in
798 "vacation" handling.
799
6e2b4ccc
PH
80084. Some minor changes to the Linux configuration files to help with other
801 OS variants using glibc.
802
8e669ac1
PH
80385. One more patch for Sieve to update vacation handling to latest spec.
804
495ae4b0 805
bbe902f0
PH
806----------------------------------------------------
807See the note above about the 4.44 and 4.50 releases.
808----------------------------------------------------
809
810
811Exim version 4.44
812-----------------
813
814 1. Change 4.43/35 introduced a bug that caused file counts to be
815 incorrectly computed when quota_filecount was set in an appendfile
816 transport
817
818 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
819 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
820 place.
821
822 3. Give more explanation in the error message when the command for a transport
823 filter fails to execute.
824
825 4. There are several places where Exim runs a non-Exim command in a
826 subprocess. The SIGUSR1 signal should be disabled for these processes. This
827 was being done only for the command run by the queryprogram router. It is
828 now done for all such subprocesses. The other cases are: ${run, transport
829 filters, and the commands run by the lmtp and pipe transports.
830
831 5. Some older OS have a limit of 256 on the maximum number of file
832 descriptors. Exim was using setrlimit() to set 1000 as a large value
833 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
834 systems. I've change it so that if it can't get 1000, it tries for 256.
835
836 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
837 was an oversight, and furthermore, ever since the addition of extra
838 controls (e.g. 4.43/32), the checks on when to allow different forms of
839 "control" were broken. There should now be diagnostics for all cases when a
840 control that does not make sense is encountered.
841
842 7. $recipients is now available in the predata ACL (oversight).
843
844 8. Tidy the search cache before the fork to do a delivery from a message
845 received from the command line. Otherwise the child will trigger a lookup
846 failure and thereby defer the delivery if it tries to use (for example) a
847 cached ldap connection that the parent has called unbind on.
848
849 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
850 of $address_data from the recipient verification was clobbered by the
851 sender verification.
852
85310. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
854 was its contents. (It was OK if the option was not defined at all.)
855
85611. A "Completed" log line is now written for messages that are removed from
857 the spool by the -Mrm option.
858
85912. $host_address is now set to the target address during the checking of
860 ignore_target_hosts.
861
86213. When checking ignore_target_hosts for an ipliteral router, no host name was
863 being passed; this would have caused $sender_host_name to have been used if
864 matching the list had actually called for a host name (not very likely,
865 since this list is usually IP addresses). A host name is now passed as
866 "[x.x.x.x]".
867
86814. Changed the calls that set up the SIGCHLD handler in the daemon to use the
869 code that specifies a non-restarting handler (typically sigaction() in
870 modern systems) in an attempt to fix a rare and obscure crash bug.
871
87215. Narrowed the window for a race in the daemon that could cause it to ignore
873 SIGCHLD signals. This is not a major problem, because they are used only to
874 wake it up if nothing else does.
875
87616. A malformed maildirsize file could cause Exim to calculate negative values
877 for the mailbox size or file count. Odd effects could occur as a result.
878 The maildirsize information is now recalculated if the size or filecount
879 end up negative.
880
88117. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
882 support for a long time. Removed HAVE_SYS_VFS_H.
883
ea3a6f44 88418. Updated exipick to current release from John Jetmore.
bbe902f0
PH
885
88619. Allow an empty sender to be matched against a lookup in an address list.
887 Previously the only cases considered were a regular expression, or an
888 empty pattern.
889
89020. Exim went into a mad DNS lookup loop when doing a callout where the
891 host was specified on the transport, if the DNS lookup yielded more than
892 one IP address.
893
ea3a6f44
NM
89421. The RFC2047 encoding function was originally intended for short strings
895 such as real names; it was not keeping to the 75-character limit for
896 encoded words that the RFC imposes. It now respects the limit, and
897 generates multiple encoded words if necessary. To be on the safe side, I
898 have increased the buffer size for the ${rfc2047: expansion operator from
899 1024 to 2048 bytes.
bbe902f0 900
ea3a6f44
NM
90122. Failure to deliver a bounce message always caused it to be frozen, even if
902 there was an errors_to setting on the router. The errors_to setting is now
903 respected.
bbe902f0
PH
904
90523. If an IPv6 address is given for -bh or -bhc, it is now converted to the
906 canonical form (fully expanded) before being placed in
907 $sender_host_address.
908
90924. Updated eximstats to version 1.33
910
ea3a6f44
NM
91125. Include certificate and key file names in error message when GnuTLS fails
912 to set them up, because the GnuTLS error message doesn't include the name
913 of the failing file when there is a problem reading it.
bbe902f0
PH
914
91526. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 916 As per change 25.
bbe902f0 917
ea3a6f44
NM
91827. Reset the locale to "C" after calling embedded Perl, in case it was changed
919 (this can affect the format of dates).
bbe902f0 920
ea3a6f44
NM
92128. exim_tidydb, when checking for the continued existence of a message for
922 which it has found a message-specific retry record, was not finding
923 messages that were in split spool directories. Consequently, it was
924 deleting retry records that should have stayed in existence.
bbe902f0
PH
925
92629. eximstats updated to version 1.35
927 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
928 1.35 - bugfix such that pie charts by volume are generated correctly
929
ea3a6f44
NM
93030. The SPA authentication driver was not abandoning authentication and moving
931 on to the next authenticator when an expansion was forced to fail,
932 contradicting the general specification for all authenticators. Instead it
933 was generating a temporary error. It now behaves as specified.
bbe902f0 934
ea3a6f44
NM
93531. The default ordering of permitted cipher suites for GnuTLS was pessimal
936 (the order specifies the preference for clients). The order is now AES256,
937 AES128, 3DES, ARCFOUR128.
bbe902f0 938
ea3a6f44
NM
93931. Small patch to Sieve code - explicitly set From: when generating an
940 autoreply.
bbe902f0 941
ea3a6f44
NM
94232. Exim crashed if a remote delivery caused a very long error message to be
943 recorded - for instance if somebody sent an entire SpamAssassin report back
944 as a large number of 550 error lines. This bug was coincidentally fixed by
945 increasing the size of one of Exim's internal buffers (big_buffer) that
946 happened as part of the Exiscan merge. However, to be on the safe side, I
947 have made the code more robust (and fixed the comments that describe what
948 is going on).
bbe902f0 949
ea3a6f44
NM
95033. Some experimental protocols are using DNS PTR records for new purposes. The
951 keys for these records are domain names, not reversed IP addresses. The
952 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
953 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
954 CAN-2005-0021
955
ea3a6f44
NM
95634. The host_aton() function is supposed to be passed a string that is known
957 to be a valid IP address. However, in the case of IPv6 addresses, it was
958 not checking this. This is a hostage to fortune. Exim now panics and dies
959 if the condition is not met. A case was found where this could be provoked
960 from a dnsdb PTR lookup with an IPv6 address that had more than 8
961 components; fortuitously, this particular loophole had already been fixed
962 by change 4.50/55 or 4.44/33 above.
963
964 If there are any other similar loopholes, the new check in host_aton()
965 itself should stop them being exploited. The report I received stated that
966 data on the command line could provoke the exploit when Exim was running as
967 exim, but did not say which command line option was involved. All I could
968 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
969 running as the user.
bbe902f0
PH
970 CAN-2005-0021
971
ea3a6f44
NM
97235. There was a buffer overflow vulnerability in the SPA authentication code
973 (which came originally from the Samba project). I have added a test to the
974 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
975 CAN-2005-0022
976
ea3a6f44
NM
97736. The daemon start-up calls getloadavg() while still root for those OS that
978 need the first call to be done as root, but it missed one case: when
979 deliver_queue_load_max is set with deliver_drop_privilege. This is
980 necessary for the benefit of the queue runner, because there is no re-exec
981 when deliver_drop_privilege is set.
bbe902f0 982
ea3a6f44
NM
98337. Caching of lookup data for "hosts =" ACL conditions, when a named host list
984 was in use, was not putting the data itself into the right store pool;
985 consequently, it could be overwritten for a subsequent message in the same
986 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
987 the caching.)
bbe902f0 988
ea3a6f44
NM
98938. Sometimes the final signoff response after QUIT could fail to get
990 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
991 < 0 before doing a fflush(). This bug looks as though it goes back to the
992 introduction of TLS in release 3.20, but "sometimes" must have been rare
993 because the tests only now provoked it.
bbe902f0
PH
994
995
495ae4b0
PH
996Exim version 4.43
997-----------------
998
999 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1000 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1001 Now it returns an integer. A number of calls were still expecting a T/F
1002 return. Fortuitously, in all cases, the tests worked in OK situations,
1003 which is the norm. However, things would have gone wrong on any write
1004 failures on the smtp file descriptor. This function is used when sending
1005 messages over SMTP and also when doing verify callouts.
1006
1007 2. When Exim is called to do synchronous delivery of a locally submitted
1008 message (the -odf or -odi options), it no longer closes stderr before doing
1009 the delivery.
1010
1011 3. Implemented the mua_wrapper option.
1012
1013 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1014
1015 5. Implemented the functions header_remove(), header_testname(),
1016 header_add_at_position(), and receive_remove_recipient(), and exported them
1017 to local_scan().
1018
1019 6. If an ACL "warn" statement specified the addition of headers, Exim already
1020 inserted X-ACL-Warn: at the start if there was no header name. However, it
1021 was not making this test for the second and subsequent header lines if
1022 there were newlines in the string. This meant that an invalid header could
1023 be inserted if Exim was badly configured.
1024
1025 7. Allow an ACL "warn" statement to add header lines at the start or after all
1026 the Received: headers, as well as at the end.
1027
1028 8. Added the rcpt_4xx retry error code.
1029
1030 9. Added postmaster_mailfrom=xxx to callout verification option.
1031
103210. Added mailfrom=xxxx to the callout verification option, for verify=
1033 header_sender only.
1034
103511. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1036 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1037
103812. Inserted some casts to stop certain compilers warning when using pointer
1039 differences as field lengths or precisions in printf-type calls (mostly
1040 affecting debugging statements).
1041
104213. Added optional readline() support for -be (dynamically loaded).
1043
104414. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1045 same clock tick as a message's arrival, so that its received time was the
1046 same as the "first fail" time on the retry record, and that message
1047 remained on the queue past the ultimate address timeout, every queue runner
1048 would try a delivery (because it was past the ultimate address timeout) but
1049 after another failure, the ultimate address timeout, which should have then
1050 bounced the address, did not kick in. This was a "< instead of <=" error;
1051 in most cases the first failure would have been in the next clock tick
1052 after the received time, and all would be well.
1053
105415. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1055 being recognized when the domain list was tested by the match_domain
1056 condition in an expansion string.
1057
105816. Added the ${str2b64: operator.
1059
106017. Exim was always calling setrlimit() to set a large limit for the number of
1061 processes, without checking whether the existing limit was already
1062 adequate. (It did check for the limit on file descriptors.) Furthermore,
1063 errors from getrlimit() and setrlimit() were being ignored. Now they are
1064 logged to the main and panic logs, but Exim does carry on, to try to do its
1065 job under whatever limits there are.
1066
106718. Imported PCRE 5.0.
1068
106919. Trivial typo in log message " temporarily refused connection" (the leading
1070 space).
1071
107220. If the log selector return_path_on_delivery was set and an address was
1073 redirected to /dev/null, the delivery process crashed because it assumed
1074 that a return path would always be set for a "successful" delivery. In this
1075 case, the whole delivery is bypassed as an optimization, and therefore no
1076 return path is set.
1077
107821. Internal re-arrangement: the function for sending a challenge and reading
1079 a response while authentication was assuming a zero-terminated challenge
1080 string. It's now changed to take a pointer and a length, to allow for
1081 binary data in such strings.
1082
108322. Added the cyrus_sasl authenticator (code supplied by MBM).
1084
108523. Exim was not respecting finduser_retries when seeking the login of the
1086 uid under which it was called; it was always trying 10 times. (The default
1087 setting of finduser_retries is zero.) Also, it was sleeping after the final
1088 failure, which is pointless.
1089
109024. Implemented tls_on_connect_ports.
1091
109225. Implemented acl_smtp_predata.
1093
109426. If the domain in control=submission is set empty, Exim assumes that the
1095 authenticated id is a complete email address when it generates From: or
1096 Sender: header lines.
1097
109827. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1099 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1100 chown and chgrp in /bin and hostname in /usr/bin.
1101
110228. Exim was keeping the "process log" file open after each use, just as it
1103 does for the main log. This opens the possibility of it remaining open for
1104 long periods when the USR1 signal hits a daemon. Occasional processlog
1105 errors were reported, that could have been caused by this. Anyway, it seems
1106 much more sensible not to leave this file open at all, so that is what now
1107 happens.
1108
110929. The long-running daemon process does not normally write to the log once it
1110 has entered its main loop, and it closes the log before doing so. This is
1111 so that log files can straightforwardly be renamed and moved. However,
1112 there are a couple of unusual error situations where the daemon does write
1113 log entries, and I had neglected to close the log afterwards.
1114
111530. The text of an SMTP error response that was received during a remote
1116 delivery was being truncated at 512 bytes. This is too short for some of
1117 the long messages that one sometimes sees. I've increased the limit to
1118 1024.
1119
112031. It is now possible to make retry rules that apply only when a message has a
1121 specific sender, in particular, an empty sender.
1122
112332. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1124 it possible to be selective about when SMTP synchronization is enforced.
1125
112633. Added "control = caseful_local_part" and "control = "caselower_local_part".
1127
112832. Implemented hosts_connection_nolog.
1129
113033. Added an ACL for QUIT.
1131
113234. Setting "delay_warning=" to disable warnings was not working; it gave a
1133 syntax error.
1134
113535. Added mailbox_size and mailbox_filecount to appendfile.
1136
113736. Added control = no_multiline_responses to ACLs.
1138
113937. There was a bug in the logic of the code that waits for the clock to tick
1140 in the case where the clock went backwards by a substantial amount such
1141 that the microsecond fraction of "now" was more than the microsecond
1142 fraction of "then" (but the whole seconds number was less).
1143
114438. Added support for the libradius Radius client library this is found on
1145 FreeBSD (previously only the radiusclient library was supported).
1146
1147
1148Exim version 4.42
1149-----------------
1150
1151 1. When certain lookups returned multiple values in the form name=value, the
1152 quoting of the values was not always being done properly. Specifically:
1153 (a) If the value started with a double quote, but contained no whitespace,
1154 it was not quoted.
1155 (b) If the value contained whitespace other than a space character (i.e.
1156 tabs or newlines or carriage returns) it was not quoted.
1157 This fix has been applied to the mysql and pgsql lookups by writing a
1158 separate quoting function and calling it from the lookup code. The fix
1159 should probably also be applied to nisplus, ibase and oracle lookups, but
1160 since I cannot test any of those, I have not disturbed their existing code.
1161
1162 2. A hit in the callout cache for a specific address caused a log line with no
1163 reason for rejecting RCPT. Now it says "Previous (cached) callout
1164 verification failure".
1165
1166 3. There was an off-by-one bug in the queryprogram router. An over-long
1167 return line was truncated at 256 instead of 255 characters, thereby
1168 overflowing its buffer with the terminating zero. As well as fixing this, I
1169 have increased the buffer size to 1024 (and made a note to document this).
1170
1171 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1172 when Exim is waiting for an SMTP response from a remote server, Exim
1173 restarts its select() call on the socket, thereby resetting its timeout.
1174 This is not a problem when such interrupts are rare. Somebody set up a cron
1175 job to run exiwhat every 2 minutes, which is less than the normal select()
1176 timeout (5 or 10 minutes). This meant that the select() timeout never
1177 kicked in because it was always reset. I have fixed this by comparing the
1178 time when an interrupt arrives with the time at the start of the first call
1179 to select(). If more time than the timeout has elapsed, the interrupt is
1180 treated as a timeout.
1181
1182 5. Some internal re-factoring in preparation for the addition of Sieve
1183 extensions (by MH). In particular, the "personal" test is moved to a
1184 separate function, and given an option for scanning Cc: and Bcc: (which is
1185 not set for Exim filters).
1186
1187 6. When Exim created an email address using the login of the caller as the
1188 local part (e.g. when creating a From: or Sender: header line), it was not
1189 quoting the local part when it contained special characters such as @.
1190
1191 7. Installed new OpenBSD configuration files.
1192
1193 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1194 try to make them clearer.
1195
1196 9. Callout options, other than the timeout value, were being ignored when
1197 verifying sender addresses in header lines. For example, when using
1198
1199 verify = header_sender/callout=no_cache
1200
1201 the cache was (incorrectly) being used.
1202
120310. Added a missing instance of ${EXE} to the exim_install script; this affects
1204 only the Cygwin environment.
1205
120611. When return_path_on_delivery was set as a log selector, if different remote
1207 addresses in the same message used different return paths and parallel
1208 remote delivery occurred, the wrong values would sometimes be logged.
1209 (Whenever a remote delivery process finished, the return path value from
1210 the most recently started remote delivery process was logged.)
1211
121212. RFC 3848 specifies standard names for the "with" phrase in Received: header
1213 lines when AUTH and/or TLS are in use. This is the "received protocol"
1214 field. Exim used to use "asmtp" for authenticated SMTP, without any
1215 indication (in the protocol name) for TLS use. Now it follows the RFC and
1216 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1217 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1218 names appear in log lines as well as in Received: header lines.
1219
122013. Installed MH's patches for Sieve to add the "copy" and "vacation"
1221 extensions, and comparison tests, and to fix some bugs.
1222
122314. Changes to the "personal" filter test:
1224
1225 (1) The test was buggy in that it was just doing the equivalent of
1226 "contains" tests on header lines. For example, if a user's address was
1227 anne@some.where, the "personal" test would incorrectly be true for
1228
1229 To: susanne@some.where
1230
1231 This test is now done by extracting each address from the header in turn,
1232 and checking the entire address. Other tests that are part of "personal"
1233 are now done using regular expressions (for example, to check local parts
1234 of addresses in From: header lines).
1235
1236 (2) The list of non-personal local parts in From: addresses has been
1237 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1238 taken from the Sieve specification recommendations.
1239
1240 (3) If the message contains any header line starting with "List-" it is
1241 treated as non-personal.
1242
1243 (4) The test for "circular" in the Subject: header line has been removed
1244 because it now seems ill-conceived.
1245
124615. Minor typos in src/EDITME comments corrected.
1247
124816. Installed latest exipick from John Jetmore.
1249
125017. If headers_add on a router specified a text string that was too long for
1251 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1252 of string_sprintf() is now avoided.
1253
125418. $message_body_size was not set (it was always zero) when running the DATA
1255 ACL and the local_scan() function.
1256
125719. For the "mail" command in an Exim filter, no default was being set for
1258 the once_repeat time, causing a random time value to be used if "once" was
1259 specified. (If the value happened to be <= 0, no repeat happened.) The
1260 default is now 0s, meaning "never repeat". The "vacation" command was OK
1261 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1262 (I found it when inspecting the code).
1263
126420. There is now an overall timeout for performing a callout verification. It
1265 defaults to 4 times the callout timeout, which applies to individual SMTP
1266 commands during the callout. The overall timeout applies when there is more
1267 than one host that can be tried. The timeout is checked before trying the
1268 next host. This prevents very long delays if there are a large number of
1269 hosts and all are timing out (e.g. when the network connections are timing
1270 out). The value of the overall timeout can be changed by specifying an
1271 additional sub-option for "callout", called "maxwait". For example:
1272
1273 verify = sender/callout=5s,maxwait=20s
1274
127521. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1276 to the end before writing, but this should make it even safer).
1277
127822. Exim was forgetting that it had advertised PIPELINING for the second and
1279 subsequent messages on an SMTP connection. It was also not resetting its
1280 memory on STARTTLS and an internal HELO.
1281
128223. When Exim logs an SMTP synchronization error within a session, it now
1283 records whether PIPELINING has been advertised or not.
1284
128524. Added 3 instances of "(long int)" casts to time_t variables that were being
1286 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1287 rather than long int.
1288
128925. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1290
129126. Added the never_mail option to autoreply.
1292
1293
1294Exim version 4.41
1295-----------------
1296
1297 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1298 crash if the getsockname() call failed; this can happen if a connection is
1299 closed very soon after it is established. The problem was simply in the
1300 order in which certain operations were done, causing Exim to try to write
1301 to the SMTP stream before it had set up the file descriptor. The bug has
1302 been fixed by making things happen in the correct order.
1303
1304
1305Exim version 4.40
1306-----------------
1307
1308 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1309 before the connection was closed, thus losing the rejection response.
1310
1311 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1312 some early Solaris releases, but causes trouble in current releases where
1313 socklen_t is defined.
1314
1315 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1316 always exist.
1317
1318 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1319 configured.
1320
1321 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1322 that releasing the top of it at the end releases what was used for sub-
1323 expansions (unless the block got too big). However, discard this block if
1324 the first thing is a variable or header, so that we can use its block when
1325 it is dynamic (useful for very large $message_headers, for example).
1326
1327 6. Lookups now cache *every* query, not just the most recent. A new, separate
1328 store pool is used for this. It can be recovered when all lookup caches are
1329 flushed. Lookups now release memory at the end of their result strings.
1330 This has involved some general refactoring of the lookup sources.
1331
1332 7. Some code has been added to the store_xxx() functions to reduce the amount
1333 of flapping under certain conditions.
1334
1335 8. log_incoming_interface used to affect only the <= reception log lines. Now
1336 it causes the local interface and port to be added to several more SMTP log
1337 lines, for example "SMTP connection from", and rejection lines.
1338
1339 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1340
134110. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1342
134311. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1344 could be overwritten at the end of the current message (or the start of a
1345 new message if it was set in a HELO ACL). The value is now preserved for
1346 the duration of the SMTP connection.
1347
134812. If a transport had a headers_rewrite setting, and a matching header line
1349 contained an unqualified address, that address was qualified, even if it
1350 did not match any rewriting rules. The underlying bug was that the values
1351 of the flags that permit the existence of unqualified sender and recipient
1352 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1353 non-local messages, and by -bnq for local messages) were not being
1354 preserved with the message after it was received.
1355
135613. When Exim was logging an SMTP synchronization error, it could sometimes log
1357 "next input=" as part of the text comprising the host identity instead of
1358 the correct text. The code was using the same buffer for two different
1359 strings. However, depending on which order the printing function evaluated
1360 its arguments, the bug did not always show up. Under Linux, for example, my
1361 test suite worked just fine.
1362
136314. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1364 doesn't work with some older versions of Perl. It has been changed to "my",
1365 which in any case is probably the better facility to use.
1366
136715. A really picky compiler found some instances of statements for creating
1368 error messages that either had too many or two few arguments for the format
1369 string.
1370
137116. The size of the buffer for calls to the DNS resolver has been increased
1372 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1373 for addresses that have a lot of PTR records. This alleviates a problem; it
1374 does not fully solve it.
1375
137617. A dnsdb lookup for PTR records that receives more data than will fit in the
1377 buffer now truncates the list and logs the incident, which is the same
1378 action as happens when Exim is looking up a host name and its aliases.
1379 Previously in this situation something unpredictable would happen;
1380 sometimes it was "internal error: store_reset failed".
1381
138218. If a server dropped the connection unexpectedly when an Exim client was
1383 using GnuTLS and trying to read a response, the client delivery process
1384 crashed while trying to generate an error log message.
1385
138619. If a "warn" verb in an ACL added multiple headers to a message in a single
1387 string, for example:
1388
1389 warn message = H1: something\nH2: something
1390
1391 the text was added as a single header line from Exim's point of view
1392 though it ended up OK in the delivered message. However, searching for the
1393 second and subsequent header lines using $h_h2: did not work. This has been
1394 fixed. Similarly, if a system filter added multiple headers in this way,
1395 the routers could not see them.
1396
139720. Expanded the error message when iplsearch is called with an invalid key to
1398 suggest using net-iplsearch in a host list.
1399
140021. When running tests using -bh, any delays imposed by "delay" modifiers in
1401 ACLs are no longer actually imposed (and a message to that effect is
1402 output).
1403
140422. If a "gecos" field in a passwd entry contained escaped characters, in
1405 particular, if it contained a \" sequence, Exim got it wrong when building
1406 a From: or a Sender: header from that name. A second bug also caused
1407 incorrect handling when an unquoted " was present following a character
1408 that needed quoting.
1409
141023. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1411 was not being matched caselessly.
1412
141324. Arranged for all hyphens in the exim.8 source to be escaped with
1414 backslashes.
1415
141625. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1417 properly. Recipient callout cache records were still being keyed to include
1418 the sender, even when use_sender was set false. This led to far more
1419 callouts that were necessary. The sender is no longer included in the key
1420 when use_sender is false.
1421
142226. Added "control = submission" modifier to ACLs.
1423
142427. Added the ${base62d: operator to decode base 62 numbers.
1425
142628. dnsdb lookups can now access SRV records.
1427
142829. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1429 the configuration file.
1430
143130. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1432 (-v) mode. This makes the output for a verbose queue run more intelligible.
1433
143431. Added a use_postmaster feature to recipient callouts.
1435
143632. Added the $body_zerocount variable, containing the number of binary zero
1437 bytes in the message body.
1438
143933. The time of last modification of the "new" subdirectory is now used as the
1440 "mailbox time last read" when there is a quota error for a maildir
1441 delivery.
1442
144334. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1444
144535. Added +ignore_unknown as a special item in host lists.
1446
144736. Code for decoding IPv6 addresses in host lists is now included, even if
1448 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1449 address was recognized as an IP address, but was then not correctly decoded
1450 into binary, causing unexpected and incorrect effects when compared with
1451 another IP address.
1452
1453
1454Exim version 4.34
1455-----------------
1456
1457 1. Very minor rewording of debugging text in manualroute to say "list of
1458 hosts" instead of "hostlist".
1459
1460 2. If verify=header_syntax was set, and a header line with an unqualified
1461 address (no domain) and a large number of spaces between the end of the
1462 name and the colon was received, the reception process suffered a buffer
1463 overflow, and (when I tested it) crashed. This was caused by some obsolete
1464 code that should have been removed. The fix is to remove it!
1465
1466 3. When running in the test harness, delay a bit after writing a bounce
1467 message to get a bit more predictability in the log output.
1468
1469 4. Added a call to search_tidyup() just before forking a reception process. In
1470 theory, someone could use a lookup in the expansion of smtp_accept_max_
1471 per_host which, without the tidyup, could leave open a database connection.
1472
1473 5. Added the variables $recipient_data and $sender_data which get set from a
1474 lookup success in an ACL "recipients" or "senders" condition, or a router
1475 "senders" option, similar to $domain_data and $local_part_data.
1476
1477 6. Moved the writing of debug_print from before to after the "senders" test
1478 for routers.
1479
1480 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1481 problems for message scanning, either using a data ACL, or using
1482 local_scan() because the Received: header was not generated till after they
1483 were called (in order to set the time as the time of reception completion).
1484 I have revised the way this works. The header is now generated after the
1485 body is received, but before the ACL or local_scan() are called. After they
1486 are run, the timestamp in the header is updated.
1487
1488
1489Exim version 4.33
1490-----------------
1491
1492 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1493 before starting a queue runner without re-exec. This happened only when
1494 deliver_drop_privilege was set or when the Exim user was set to root. The
1495 effect of the bug was that timeouts during subsequent deliveries caused
1496 crashes instead of being properly handled. The handler is now left at its
1497 default (and expected) setting.
1498
1499 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1500 message, again when deliver_drop_privilege is set or Exim is run as root.
1501 The bug described in (1) was not present in this case, but the tidying up
1502 of the other signals was missing. I have made the two cases consistent.
1503
1504 3. The ignore_target_hosts setting on a manualroute router was being ignored
1505 for hosts that were looked up using the /MX notation.
1506
1507 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1508 in domain lists.
1509
1510 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1511 operated on the sender address. After changing the $sender_address to <>
1512 for the sender address verify, Exim was re-instated it as the original
1513 (before rewriting) address, but remembering that it had rewritten it, so it
1514 wasn't rewriting it again. This bug also had the effect of breaking the
1515 sender address verification caching when the sender address was rewritten.
1516
1517 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1518 This has been changed so that if the ip literal address matches
1519 ignore_target_hosts, the router declines.
1520
1521 7. Added expansion conditions match_domain, match_address, and match_local_
1522 part (NOT match_host).
1523
1524 8. The placeholder for the Received: header didn't have a length field set.
1525
1526 9. Added code to Exim itself and to exim_lock to test for a specific race
1527 condition that could lead to file corruption when using MBX delivery. The
1528 issue is with the lockfile that is created in /tmp. If this file is removed
1529 after a process has opened it but before that process has acquired a lock,
1530 there is the potential for a second process to recreate the file and also
1531 acquire a lock. This could lead to two Exim processes writing to the file
1532 at the same time. The added code performs the same test as UW imapd; it
1533 checks after acquiring the lock that its file descriptor still refers to
1534 the same named file.
1535
153610. The buffer for building added header lines was of fixed size, 8192 bytes.
1537 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1538 when Exim is built.
1539
154011. Added the smtp_active_hostname option. If used, this will typically be made
1541 to depend on the incoming interface address. Because $interface_address is
1542 not set up until the daemon has forked a reception process, error responses
1543 that can happen earlier (such as "too many connections") no longer contain
1544 a host name.
1545
154612. If an expansion in a condition on a "warn" statement fails because a lookup
1547 defers, the "warn" statement is abandoned, and the next ACL statement is
1548 processed. Previously this caused the whole ACL to be aborted.
1549
155013. Added the iplsearch lookup type.
1551
155214. Added ident_timeout as a log selector.
1553
155415. Added tls_certificate_verified as a log selector.
1555
155616. Added a global option tls_require_ciphers (compare the smtp transport
1557 option of the same name). This controls incoming TLS connections.
1558
155917. I finally figured out how to make tls_require_ciphers do a similar thing
1560 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1561 before starting the TLS session.
1562
156318. Tabs are now shown as \t in -bP output.
1564
156519. If the log selector return_path_on_delivery was set, Exim crashed when
1566 bouncing a message because it had too many Received: header lines.
1567
156820. If two routers both had headers_remove settings, and the first one included
1569 a superfluous trailing colon, the final name in the first list and the
1570 first name in the second list were incorrectly joined into one item (with a
1571 colon in the middle).
1572
1573
1574Exim version 4.32
1575-----------------
1576
1577 1. Added -C and -D options to the exinext utility, mainly to make it easier
1578 to include in the automated testing, but these could be helpful when
1579 multiple configurations are in use.
1580
1581 2. The exinext utility was not formatting the output nicely when there was
1582 an alternate port involved in the retry record key, nor when there was a
1583 message id as well (for retries that were specific to a specific message
1584 and a specific host). It was also confused by IPv6 addresses, because of
1585 the additional colons they contain. I have fixed the IPv4 problem, and
1586 patched it up to do a reasonable job for IPv6.
1587
1588 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1589 delivery, the log line now contains "pipelined" if PIPELINING was used.
1590
1591 4. An SMTP transport process used to panic and die if the bind() call to set
1592 an explicit outgoing interface failed. This has been changed; it is now
1593 treated in the same way as a connect() failure.
1594
1595 5. A reference to $sender_host_name in the part of a conditional expansion
1596 that was being skipped was still causing a DNS lookup. This no longer
1597 occurs.
1598
1599 6. The def: expansion condition was not recognizing references to header lines
1600 that used bh_ and bheader_.
1601
1602 7. Added the _cache feature to named lists.
1603
1604 8. The code for checking quota_filecount in the appendfile transport was
1605 allowing one more file than it should have been.
1606
1607 9. For compatibility with Sendmail, the command line option
1608
1609 -prval:sval
1610
1611 is equivalent to
1612
1613 -oMr rval -oMs sval
1614
1615 and sets the incoming protocol and host name (for trusted callers). The
1616 host name and its colon can be omitted when only the protocol is to be set.
1617 Note the Exim already has two private options, -pd and -ps, that refer to
1618 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1619 "s", but I don't think that's a major issue.
1620
162110. A number of refactoring changes to the code, none of which should affect
1622 Exim's behaviour:
1623
1624 (a) The number of logging options was getting close to filling up the
1625 32-bit word that was used as a bit map. I have split them into two classes:
1626 those that are passed in the argument to log_write(), and those that are
1627 only ever tested independently outside of that function. These are now in
1628 separate 32-bit words, so there is plenty of room for expansion again.
1629 There is no change in the user interface or the logging behaviour.
1630
1631 (b) When building, for example, log lines, the code previously used a
1632 macro that called string_cat() twice, in order to add two strings. This is
1633 not really sufficiently general. Furthermore, there was one instance where
1634 it was actually wrong because one of the argument was used twice, and in
1635 one call a function was used. (As it happened, calling the function twice
1636 did not affect the overall behaviour.) The macro has been replaced by a
1637 function that can join an arbitrary number of extra strings onto a growing
1638 string.
1639
1640 (c) The code for expansion conditions now uses a table and a binary chop
1641 instead of a serial search (which was left over from when there were very
1642 few conditions). Also, it now recognizes conditions like "pam" even when
1643 the relevant support is not compiled in: a suitably worded error message is
1644 given if an attempt is made to use such a condition.
1645
164611. Added ${time_interval:xxxxx}.
1647
164812. A bug was causing one of the ddress fields not to be passed back correctly
1649 from remote delivery subprocesses. The field in question was not being
1650 subsequently used, so this caused to problems in practice.
1651
165213. Added new log selectors queue_time and deliver_time.
1653
165414. Might have fixed a bug in maildirsizefile handling that threw up
1655 "unexpected character" debug warnings, and recalculated the data
1656 unnecessarily. In any case, I expanded the warning message to give more
1657 information.
1658
165915. Added the message "Restricted characters in address" to the statements in
1660 the default ACL that block characters like @ and % in local parts.
1661
166216. Change 71 for release 4.31 proved to be much less benign that I imagined.
1663 Three changes have been made:
1664
1665 (a) There was a serious bug; a negative response to MAIL caused the whole
1666 recipient domain to be cached as invalid, thereby blocking all messages
1667 to all local parts at the same domain, from all senders. This bug has
1668 been fixed. The domain is no longer cached after a negative response to
1669 MAIL if the sender used is not empty.
1670
1671 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1672 been restored.
1673
1674 (c) A new callout option, "use_sender" has been added for people who want
1675 the modified behaviour.
1676
1677
1678Exim version 4.31
1679-----------------
1680
1681 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1682 Larry Rosenman.
1683
1684 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1685 indeed breaks things for older releases.
1686
1687 3. Added additional logging to the case where there is a problem reading data
1688 from a filter that is running in a subprocess using a pipe, in order to
1689 try to track down a specific problem.
1690
1691 4. Testing facility fudge: when running in the test harness and attempting
1692 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1693 getting "No route to host". Convert this to a timeout.
1694
1695 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1696 warning.
1697
1698 6. Some OS don't have socklen_t but use size_t instead. This affects the
1699 fifth argument of getsockopt() amongst other things. This is now
1700 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1701 can be set for individual OS. I have set it for SunOS5, OSF1, and
1702 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1703 some earlier ones do not.
1704
1705 7. Change 4.30/15 was not doing the test caselessly.
1706
1707 8. The standard form for an IPv6 address literal was being rejected by address
1708 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1709 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1710 this, as well as the form without the "IPv6" on the front (but only when
1711 address literals are enabled, of course).
1712
1713 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1714
171510. Exim crashed if a message with an empty sender address specified by -f
1716 encountered a router with an errors_to setting. This could be provoked only
1717 by a command such as
1718
1719 exim -f "" ...
1720
1721 where an empty string was supplied; "<>" did not hit this bug.
1722
172311. Installed PCRE release 4.5.
1724
172512. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1726 remained set. It is now erased.
1727
172813. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1729 times from message ids (which are base 36 rather than the normal 62).
1730
173114. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1732 were being counted as actual protocol errors, and logged if the log
1733 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1734 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1735 and DATA following a set of rejected RCPTs do not count as protocol errors.
1736 In other words, Exim assumes they were pipelined, though this may not
1737 actually be the case. Of course, in all cases the client gets an
1738 appropriate error code.
1739
174015. If a lookup fails in an ACL condition, a message about the failure may
1741 be available; it is used if testing the ACL cannot continue, because most
1742 such messages specify what the cause of the deferral is. However, some
1743 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1744 that caused an old message to be retained and used if a later statement
1745 caused a defer, replacing the real cause of the deferral.
1746
174716. If an IP address had so many PTR records that the DNS lookup buffer
1748 was not large enough to hold them, Exim could crash while trying to process
1749 the truncated data. It now detects and logs this case.
1750
175117. Further to 4.21/58, another change has been made: if (and only if) the
1752 first line of a message (the first header line) ends with CRLF, a bare LF
1753 in a subsequent header line has a space inserted after it, so as not to
1754 terminate the header.
1755
175618. Refactoring: tidied an ugly bit of code in appendfile that copied data
1757 unnecessarily, used atoi() instead of strtol(), and didn't check the
1758 termination when getting file sizes from file names by regex.
1759
176019. Completely re-implemented the support for maildirsize files, in the light
1761 of a number of problems with the previous contributed implementation
1762 (4.30/29). In particular:
1763
1764 . If the quota is zero, the maildirsize file is maintained, but no quota is
1765 imposed.
1766
1767 . If the maildir directory does not exist, it is created before any attempt
1768 to write a maildirsize file.
1769
1770 . The quota value in the file is just a cache; if the quota is changed in
1771 the transport, the new value overrides.
1772
1773 . A regular expression is available for excluding directories from the
1774 count.
1775
177620. The autoreply transport checks the characters in options that define the
1777 message's headers; it allows continued headers, but it was checking with
1778 isspace() after an embedded newline instead of explicitly looking for a
1779 space or a tab.
1780
178121. If all the "regular" hosts to which an address was routed had passed their
1782 expiry times, and had not reached their retry times, the address was
1783 bounced, even if fallback hosts were defined. Now Exim should go on to try
1784 the fallback hosts.
1785
178622. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1787 equivalent code in the SMTP transport. Some hosts send humungous responses
1788 to HELO/EHLO, more than 1024 it seems.
1789
179023. Refactoring: code in filter.c used (void *) for "any old type" but this
1791 gives compiler warnings in some environments. I've now done it "properly",
1792 using a union.
1793
179424. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1795 (because of problems with the built-in one) was declared to return uschar *
1796 instead of char *, causing compiler failure.
1797
179825. Fixed a file descriptor leak when processing alias/forward files.
1799
180026. Fixed a minor format string issue in dbfn.c.
1801
180227. Typo in exim.c: ("dmbnz" for "dbmnz").
1803
180428. If a filter file refered to $h_xxx or $message_headers, and the headers
1805 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1806 become corrupted.
1807
180829. When a sender address is verified, it is cached, to save repeating the test
1809 when there is more than one recipient in a message. However, when the
1810 verification involves a callout, it is possible for different callout
1811 options to be set for different recipients. It is too complicated to keep
1812 track of this in the cache, so now Exim always runs a verification when a
1813 callout is required, relying on the callout cache for the optimization.
1814 The overhead is duplication of the address routing, but this should not be
1815 too great.
1816
181730. Fixed a bug in callout caching. If a RCPT command caused the sender address
1818 to be verified with callout=postmaster, and the main callout worked but the
1819 postmaster check failed, the verification correctly failed. However, if a
1820 subsequent RCPT command asked for sender verification *without* the
1821 postmaster check, incorrect caching caused this verification also to fail,
1822 incorrectly.
1823
182431. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1825 it was not caching the DNS options (qualify_single, search_parents) that
1826 were used when the lookup failed. A subsequent lookup with different
1827 options therefore always gave the same answer, though there were cases
1828 where it should not have. (Example: a "domains = !$mx_any" option on a
1829 dnslookup router: the "domains" option is always processed without any
1830 widening, but the router might have qualify_single set.) Now Exim uses the
1831 cached value only when the same options are set.
1832
183332. Added John Jetmore's "exipick" utility to the distribution.
1834
183533. GnuTLS: When an attempt to start a TLS session fails for any reason other
1836 than a timeout (e.g. a certificate is required, and is not provided), an
1837 Exim server now closes the connection immediately. Previously it waited for
1838 the client to close - but if the client is SSL, it seems that they each
1839 wait for each other, leading to a delay before one of them times out.
1840
184134: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1842 maintained 0.8.x compatibility because I don't think many are using it, and
1843 it is clearly obsolete.
1844
184535. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1846 transport.
1847
184836. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1849 client certificate was expired. A simple patch fixes this, though I don't
1850 understand the full logic of why the verify callback is called multiple
1851 times.
1852
185337. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1854 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1855 which causes problems with some clients (such as the Certicom SSL Plus
1856 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1857 disables the coutermeasure allowing Eudora to connect."
1858
185938. Exim was not checking that a write() to a log file succeeded. This could
1860 lead to Bad Things if a log got too big, in particular if it hit a file
1861 size limit. Exim now panics and dies if it cannot write to a log file, just
1862 as it does if it cannot open a log file.
1863
186439. Modified OS/Makefile-Linux so that it now contains
1865
1866 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1867
1868 The two -D definitions ensure that Exim is compiled with large file
1869 support, which makes it possible to handle log files that are bigger than
1870 2^31.
1871
187240. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1873 instance) a domain was checked against a named list that involved a lookup,
1874 causing $domain_data to be set, then another domain was checked against the
1875 same list, then the first domain was re-checked, the value of $domain_data
1876 after the final check could be wrong. In particular, if the second check
1877 failed, it could be set empty. This bug probably also applied to
1878 $localpart_data.
1879
188041. The strip_trailing_dot option was not being applied to the address given
1881 with the -f command-line option.
1882
188342. The code for reading a message's header from the spool was incrementing
1884 $received_count, but never initializing it. This meant that the value was
1885 incorrect (doubled) while delivering a message in the same process in which
1886 it was received. In the most common configuration of Exim, this never
1887 happens - a fresh exec is done - but it can happen when
1888 deliver_drop_privilege is set.
1889
189043. When Exim logs an SMTP synchronization error - client data sent too soon -
1891 it now includes up to 150 characters of the unexpected data in the log
1892 line.
1893
189444. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1895 and building data strings. The size of both of these buffers was 10 000
1896 bytes - far larger than anybody would *ever* want, thought I. Needless to
1897 say, somebody hit the limit. I have increased the maximum line length to
1898 20 000 and the maximum data length of concatenated lines to 100 000. I have
1899 also fixed two bugs, because there was no checking on these buffers. Tsk,
1900 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1901 buffer is too small.
1902
190345. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1904 lsearch lookups. Now it does.
1905
190646. When parsing a route_list item in a manualroute router, a fixed-length
1907 buffer was used for the list of hosts. I made this 1024 bytes long,
1908 thinking that nobody would ever have a list of hosts that long. Wrong.
1909 Somebody had a whole pile of complicated expansion conditions, and the
1910 string was silently truncated, leading to an expansion error. It turns out
1911 that it is easier to change to an unlimited length (owing to other changes
1912 that have happened since this code was originally written) than to build
1913 structure for giving a limitation error. The length of the item that
1914 expands into the list of hosts is now unlimited.
1915
191647. The lsearch lookup could not handle data where the length of text line was
1917 more than 4095 characters. Such lines were truncated, leading to shortened
1918 data being returned. It should now handle lines of any length.
1919
192048. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1921 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1922 ACL").
1923
192449. Cosmetic tidy to scripts like exicyclog that are generated by globally
1925 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1926 no longer happens in comment lines. A list of replacements is now placed
1927 at the head of all of the source files, except those whose only change is
1928 to replace PERL_COMMAND in the very first #! line.
1929
193050. Replaced the slow insertion sort in queue.c, for sorting the list of
1931 messages on the queue, with a bottom-up merge sort, using code contributed
1932 by Michael Haardt. This should make operations like -bp somewhat faster on
1933 large queues. It won't affect queue runners, except when queue_run_in_order
1934 is set.
1935
193651. Installed eximstats 1.31 in the distribution.
1937
193852. Added support for SRV lookups to the dnslookup router.
1939
194053. If an ACL referred to $message_body or $message_body_end, the value was not
1941 reset for any messages that followed in the same SMTP session.
1942
194354. The store-handling optimization for building very long strings was not
1944 differentiating between the different store pools. I don't think this
1945 actually made any difference in practice, but I've tidied it.
1946
194755. While running the routers to verify a sender address, $sender_address
1948 was still set to the sender address. This is wrong, because when routing to
1949 send a bounce to the sender, it would be empty. Therefore, I have changed
1950 it so that, while verifying a sender address, $sender_address is set to <>.
1951 (There is no change to what happens when verifying a recipient address.)
1952
195356. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1954 target A or AAAA records (if not already returned) without resetting the
1955 qualify_single or search_parents options of the DNS resolver. These are
1956 inappropriate in this case because the targets of MX and SRV records must
1957 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1958 target that, when qualified, matched something in the local domain. These
1959 two options are now turned off when doing these lookups.
1960
196157. It seems that at least some releases of Reiserfs (which does not have the
1962 concept of a fixed number of inodes) returns zero and not -1 for the
1963 number of available inodes. This interacted badly with check_spool_inodes,
1964 which assumed that -1 was the "no such thing" setting. What I have done is
1965 to check that the total number of inodes is greater than zero before doing
1966 the test of how many are available.
1967
196858. When a "warn" ACL statement has a log_message modifier, the message is
1969 remembered, and not repeated. This is to avoid a lot of repetition when a
1970 message has many recipients that cause the same warning to be written.
1971 Howewer, Exim was preserving the list of already written lines for an
1972 entire SMTP session, which doesn't seem right. The memory is now reset if a
1973 new message is started.
1974
197559. The "rewrite" debugging flag was not showing the result of rewriting in the
1976 debugging output unless log_rewrite was also set.
1977
197860. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1979 of (int)(handle) when we know that handle contains (void *)(-1).
1980
198161. The Exim daemon panic-logs an error return when it closes the incoming
1982 connection. However "connection reset by peer" seems to be common, and
1983 isn't really an error worthy of noting specially, so that particular error
1984 is no long logged.
1985
198662. When Exim is trying to find all the local interfaces, it used to panic and
1987 die if the ioctl to get the interface flags failed. However, it seems that
1988 on at least one OS (Solaris 9) it is possible to have an interface that is
1989 included in the list of interfaces, but for which you get a failure error
1990 for this call. This happens when the interface is not "plumbed" into a
1991 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1992 failure of the "get flags" call assumes that the interface is down.
1993
199463. Added a ${eval10: operator, which assumes all numbers are decimal. This
1995 makes life easier for people who are doing arithmetic on fields extracted
1996 from dates, where you often get leading zeros that should not be
1997 interpreted as octal.
1998
199964. Added qualify_domain to the redirect router, to override the global
2000 setting.
2001
200265. If a pathologically long header line contained very many addresses (the
2003 report of this problem mentioned 10 000) and each of them was rewritten,
2004 Exim could use up a very large amount of memory. (It kept on making new
2005 copies of the header line as it rewrote, and never released the old ones.)
2006 At the expense of a bit more processing, the header rewriting function has
2007 been changed so that it no longer eats memory in this way.
2008
200966. The generation of the Received: header has been moved from the time that a
2010 message starts to be received, to the time that it finishes. The timestamp
2011 in the Received: header should now be very close to that of the <= log
2012 line. There are two side-effects of this change:
2013
2014 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2015 logged header lines no longer include the local Received: line, because
2016 it has not yet been created. The same applies to a copy of the message
2017 that is returned to a non-SMTP sender when a message is rejected.
2018
2019 (b) When a filter file is tested using -bf, no additional Received: header
2020 is added to the test message. After some thought, I decided that this
2021 is a bug fix.
2022
2023 This change does not affect the value of $received_for. It is still set
2024 after address rewriting, but before local_scan() is called.
2025
202667. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2027
202868. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2029 gave an unhelpful panic error message, and a defer error. I have managed to
2030 change this behaviour so that it now rejects any supplied certificate,
2031 which seems right, as the list of acceptable certificates is empty.
2032
203369. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2034 gave an unhelpful defer error. I have not managed to make this reject any
2035 supplied certificates, but the error message it gives is "no certificate
2036 supplied", which is not helpful.
2037
203870. exigrep's output now also includes lines that are not associated with any
2039 message, but which match the given pattern. Implemented by a patch from
2040 Martin Sluka, which also tidied up the Perl a bit.
2041
204271. Recipient callout verification, like sender verification, was using <> in
2043 the MAIL FROM command. This isn't really the right thing, since the actual
2044 sender may affect whether the remote host accepts the recipient or not. I
2045 have changed it to use the actual sender in the callout; this means that
2046 the cache record is now keyed on a recipient/sender pair, not just the
2047 recipient address. There doesn't seem to be a real danger of callout loops,
2048 since a callout by the remote host to check the sender would use <>.
2049 [SEE ABOVE: changed after hitting problems.]
2050
205172. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2052 temporary errors. However, in the case of such a code being given after
2053 the end of a data transmission (i.e. after ".") Exim was failing to write
2054 a retry record for the message. (Yes, there was some broken host that was
2055 actually sending 8xx at this point.)
2056
205773. An unknown lookup type in a host list could cause Exim to panic-die when
2058 the list was checked. (An example that provoked this was putting <; in the
2059 middle of a list instead of at the start.) If this happened during a DATA
2060 ACL check, a -D file could be left lying around. This kind of configuration
2061 error no longer causes Exim to die; instead it causes a defer errror. The
2062 incident is still logged to the main and panic logs.
2063
206474. Buglet left over from Exim 3 conversion. The message "too many messages
2065 in one connection" was written to the rejectlog but not the mainlog, except
2066 when address rewriting (yes!) was being logged.
2067
206875. Added write_rejectlog option.
2069
207076. When a system filter was run not as root (that is, when system_filter_user
2071 was set), the values of the $n variables were not being returned to the
2072 main process; thus, they were not subsequently available in the $sn
2073 variables.
2074
207577. Added +return_path_on_delivery log selector.
2076
207778. A connection timeout was being treated differently from recipients deferred
2078 when testing hosts_max_try with a message that was older than the host's
2079 retry timeout. (The host should not be counted, thus allowing all hosts to
2080 be tried at least once before bouncing.) This may have been the cause of an
2081 occasionally reported bug whereby a message would remain on the queue
2082 longer than the retry timeout, but would be bounced if a delivery was
2083 forced. I say "may" because I never totally pinned down the problem;
2084 setting up timeout/retry tests is difficult. See also the next item.
2085
208679. The ultimate address timeout was not being applied to errors that involved
2087 a combination of host plus message (for example, a timeout on a MAIL
2088 command). When an address resolved to a number of possible hosts, and they
2089 were not all tried for each delivery (e.g. because of hosts_max_try), a
2090 message could remain on the queue longer than the retry timeout.
2091
209280. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2093 Haardt.
2094
209581. Fixed an obscure SMTP outgoing bug which required at least the following
2096 conditions: (a) there was another message waiting for the same server;
2097 (b) the server returned 5xx to all RCPT commands in the first message so
2098 that the message was not completed; (c) the server dropped the connection
2099 or gave a negative response to the RSET that Exim sends to abort the
2100 transaction. The observed case was a dropped connection after DATA that had
2101 been sent in pipelining mode. That is, the server had advertised PIPELINING
2102 but was not implementing it correctly. The effect of the bug was incorrect
2103 behaviour, such as trying another host, and this could lead to a crash.
2104
2105
2106Exim version 4.30
2107-----------------
2108
2109 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2110 and daemon.c were passed as pointers to ints; they should have been
2111 pointers to socklen_t variables (which are typically unsigned ints).
2112
2113 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2114 fixed.
2115
2116 3. Fixed a really odd bug that affected only the testing scheme; patching a
2117 certain fixed string in the binary changed the value of another string that
2118 happened to be identical to the end of the original first string.
2119
2120 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2121 name", it returns that address as the IP address. On some operating
2122 systems (e.g. Solaris), it also passes back the IP address string as the
2123 "host name". However, on others (e.g. Linux), it passes back an empty
2124 string. Exim wasn't checking for this, and was changing the host name to an
2125 empty string, assuming it had been canonicized.
2126
2127 5. Although rare, it is permitted to have more than one PTR record for a given
2128 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2129 all the names associated with an address, because they do in Solaris.
2130 However, it seems that they do not in Linux for data that comes from the
2131 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2132 I found this out when I moved to a new Linux workstation and tried to run
2133 the Exim test suite.
2134
2135 To get round this problem I have changed the code so that it now does its
2136 own call to the DNS to look up PTR records when searching for a host name.
2137 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2138 addresses that are only in /etc/hosts are still found.
2139
2140 This behaviour is, however, controlled by an option called host_lookup_
2141 order, which defaults to "bydns:byaddr". If people want to use the other
2142 order, or indeed, just use one or the other means of lookup, they can
2143 specify it in this variable.
2144
2145 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2146 some operating systems, this comes back from gethostbyaddr() as an empty
2147 string, and this is what Exim used to test for. However, it seems that in
2148 other systems, "." is yielded. Exim now tests for this case too.
2149
2150 7. The values of check_spool_space and check_log_space are now held internally
2151 as a number of kilobytes instead of an absolute number of bytes. If a
2152 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2153 kilobyte. This means that much larger values can be stored.
2154
2155 8. Exim monitor: an attempt to get the action menu when not actually pointing
2156 at a message produces an empty menu entitled "No message selected". This
2157 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2158 no entries in it ("Shell widget menu has zero width and/or height"). So I
2159 have added a single, blank menu entry in this case.
2160
2161 9. Added ${quote_local_part.
2162
216310. MIME decoding is now applied to the contents of Subject: header lines when
2164 they are logged.
2165
216611. Now that a reference to $sender_host_address automatically causes a reverse
2167 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2168 host lookup before query-style lookups in lists that might use this
2169 variable. This has therefore been abolished, and the "net-" prefix is no
2170 longer necessary for query-style lookups.
2171
217212. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2173 have been a typo for LFLAGS, so it has been changed.
2174
217513. The install script calls Exim with "-C /dev/null" in order to find the
2176 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2177 to be output. Howeve, since Exim outputs its version number before the
2178 error, it didn't break the script. It just looked ugly. I fixed this by
2179 always allowing "-C /dev/null" if the caller is root.
2180
218114. Ignore overlarge ACL variable number when reading spool file - insurance
2182 against a later release with more variables having written the file.
2183
218415. The standard form for an IPv6 address literal was being rejected by EHLO.
2185 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2186 this, as well as the form without the "IPv6" on the front.
2187
218816. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2189 OS/Makefile-Darwin file.
2190
219117. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2192 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2193
219418. After change 4.21/52, "%ld" was used to format the contents of the $inode
2195 variable. However, some OS use ints for inodes. I've added cast to long int
2196 to get rid of the compiler warning.
2197
219819. I had forgotten to lock out "/../" in configuration file names when
2199 ALT_CONFIG_PREFIX was set.
2200
220120. Routers used for verification do not need to specify transports. However,
2202 if such a router generated a host list, and callout was configured, Exim
2203 crashed, because it could not find a port number from the (non-existent)
2204 transport. It now assumes port 25 in this circumstance.
2205
220621. Added the -t option to exigrep.
2207
220822. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2209 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2210 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2211 them is set, LOOKUP_LSEARCH is forced.
2212
221323. "exim -bV" now outputs a list of lookups that are included in the binary.
2214
221524. Added sender and host information to the "rejected by local_scan()" log
2216 line; previously there was no indication of these.
2217
221825. Added .include_if_exists.
2219
222026. Change 3.952/11 added an explicit directory sync on top of a file sync for
2221 Linux. It turns out that not all file systems support this. Apparently some
2222 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2223 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2224 not supported on the file descriptor, is now ignored when Exim is trying to
2225 sync a directory. This applies only to Linux.
2226
222727. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2228
222928. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2230 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2231 this.
2232
223329. Added support for maildirsize files from supplied patch (modified a bit).
2234
223530. The use of :fail: followed by an empty string could lead Exim to respond to
2236 sender verification failures with (e.g.):
2237
2238 550 Verification failed for <xxx>
2239 550 Sender verify failed
2240
2241 where the first response line was missing the '-' that indicates it is not
2242 the final line of the response.
2243
224431. The loop for finding the name of the user that called Exim had a hardwired
2245 limit of 10; it now uses the value of finduser_retries, which is used for
2246 all other user lookups.
2247
224832. Added $received_count variable, available in data and not_smtp ACLs, and at
2249 delivery time.
2250
225133. Exim was neglecting to zero errno before one call of strtol() when
2252 expanding a string and expecting an integer value. On some systems this
2253 resulted in spurious "integer overflow" errors. Also, it was casting the
2254 result into an int without checking.
2255
225634. Testing for a connection timeout using "timeout_connect" in the retry rules
2257 did not work. The code looks as if it has *never* worked, though it appears
2258 to have been documented since at least releast 1.62. I have made it work.
2259
226035. The "timeout_DNS" error in retry rules, also documented since at least
2261 1.62, also never worked. As it isn't clear exactly what this means, and
2262 clearly it isn't a major issue, I have abolished the feature by treating it
2263 as "timeout", and writing a warning to the main and panic logs.
2264
226536. The display of retry rules for -brt wasn't always showing the error code
2266 correctly.
2267
226837. Added new error conditions to retry rules: timeout_A, timeout_MX,
2269 timeout_connect_A, timeout_connect_MX.
2270
227138. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2272 to the empty sender.
2273
227439. The daemon was not analysing the content of -oX till after it had closed
2275 stderr and disconnected from the controlling terminal. This meant that any
2276 syntax errors were only noted on the panic log, and the return code from
2277 the command was 0. By re-arranging the code a little, I've made the
2278 decoding happen first, so such errors now appear on stderr, and the return
2279 code is 1. However, the actual setting up of the sockets still happens in
2280 the disconnected process, so errors there are still only recorded on the
2281 panic log.
2282
228340. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2284 connections (as happens on some IP stacks) was logged at start up time as
2285 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2286 it from "IPv6 and IPv4", which means that two separate sockets are being
2287 used.
2288
228941. The debug output for gethostbyname2() or getipnodebyname() failures now
2290 says whether AF_INET or AF_INET6 was passed as an argument.
2291
229242. Exiwhat output was messed up when time zones were included in log
2293 timestamps.
2294
229543. Exiwhat now gives more information about the daemon's listening ports,
2296 and whether -tls-on-connect was used.
2297
229844. The "port" option of the smtp transport is now expanded.
2299
230045. A "message" modifier in a "warn" statement in a non-message ACL was being
2301 silently ignored. Now an error message is written to the main and panic
2302 logs.
2303
230446. There's a new ACL modifier called "logwrite" which writes to a log file
2305 as soon as it is encountered.
2306
230747. Added $local_user_uid and $local_user_gid at routing time.
2308
230948. Exim crashed when trying to verify a sender address that was being
2310 rewritten to "<>".
2311
231249. Exim was recognizing only a space character after ".include". It now also
2313 recognizes a tab character.
2314
231550. Fixed several bugs in the Perl script that creates the exim.8 man page by
2316 extracting the relevant information from the specification. The man page no
2317 longer contains scrambled data for the -d option, and I've added a section
2318 at the front about calling Exim under different names.
2319
232051. Added "extra_headers" argument to the "mail" command in filter files.
2321
232252. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2323 crash.
2324
232553. Installed eximstats 1.29.
2326
232754. Added transport_filter_timeout as a generic transport option.
2328
232955. Exim no longer adds an empty Bcc: header to messages that have no To: or
2330 Cc: header lines. This was required by RFC 822, but it not required by RFC
2331 2822.
2332
233356. Exim used to add From:, Date:, and Message-Id: header lines to any
2334 incoming messages that did not have them. Now it does so only if the
2335 message originates locally, that is, if there is no associated remote host
2336 address. When Resent- header lines are present, this applies to the Resent-
2337 lines rather than the non-Resent- lines.
2338
233957. Drop incoming SMTP connection after too many syntax or protocol errors. The
2340 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2341
234258. Messages for configuration errors now include the name of the main
2343 configuration file - useful now that there may be more than one file in a
2344 list (.included file names were always shown).
2345
234659. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2347 for those rare installations that do not start the daemon as root or run it
2348 setuid root. I've cut out the call to initgroups() if the daemon is not
2349 root at that time.
2350
235160. The Exim user and group can now be bound into the binary as text strings
2352 that are looked up at the start of Exim's processing.
2353
235461. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2355
235662. Added $mailstore_basename variable.
2357
235863. Installed patch to sieve.c from Michael Haardt.
2359
236064. When Exim failed to open the panic log after failing to open the main log,
2361 the original message it was trying to log was written to stderr and debug
2362 output, but if they were not available (the usual case in production), it
2363 was lost. Now it is written to syslog before the two lines that record the
2364 failures to open the logs.
2365
236665. Users' Exim filters run in subprocesses under the user's uid. It is
2367 possible for a "deliver" command or an alias in a "personal" command to
2368 provoke an address rewrite. If logging of address rewriting is configured,
2369 this fails because the process is not running as root or exim. There may be
2370 a better way of dealing with this, but for the moment (because 4.30 needs
2371 to be released), I have disabled address rewrite logging when running a
2372 filter in a non-root, non-exim process.
2373
2374
2375Exim version 4.24
2376-----------------
2377
2378 1. The buildconfig auxiliary program wasn't quoting the value set for
2379 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2380 not defined. This bug was masked in 4.22 by the effect that was fixed in
2381 change 4.23/1.
2382
2383 2. Some messages that were rejected after a message id was allocated were
2384 shown as "incomplete" by exigrep. It no longer does this for messages that
2385 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2386
2387 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2388 have allow_domain_literals set, the ID did not get logged in the <= line.
2389 Domain literals are now always recognized in Message-ID: header lines.
2390
2391 4. The first argument for a ${extract expansion item is the key name or field
2392 number. Leading and trailing spaces in this item were not being ignored,
2393 causing some misleading effects.
2394
2395 5. When deliver_drop_privilege was set, single queue runner processes started
2396 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2397 same command in the process it spins off) were not dropping privilege.
2398
2399 6. When the daemon running as "exim" started a queue runner, it always
2400 re-executed Exim in the spun-off process. This is a waste of effort when
2401 deliver_drop_privilege is set. The new process now just calls the
2402 queue-runner function directly.
2403
2404
2405Exim version 4.23
2406-----------------
2407
2408 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2409 HEADERS_CHARSET.
2410
2411 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2412 ignored. Though the use of -oP was forcing the writing of a pid file, it
2413 was always written to the default place.
2414
2415 3. If the message "no IP address found for host xxxx" is generated during
2416 incoming verification, it is now followed by identification of the incoming
2417 connection (so you can more easily find what provoked it).
2418
2419 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2420
2421 5. Added some features to "harden" Exim a bit more against certain attacks:
2422
2423 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2424 be put in Local/Makefile. This is like the never_users runtime option,
2425 but it cannot be overridden. The default setting is "root".
2426
2427 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2428 prefix string with which any file named in a -C command line option
2429 must start.
2430
2431 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2432 is retained for -C and -D only if the caller of Exim is root. Without
2433 it, the exim user may also use -C and -D and retain privilege.
2434
2435 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2436 command line option is disabled.
2437
2438 6. Macro names set by the -D option must start with an upper case letter, just
2439 like macro names defined in the configuration file.
2440
2441 7. Added "dereference=" facility to LDAP.
2442
2443 8. Two instances of the typo "uknown" in the source files are fixed.
2444
2445 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2446 the Configure-Makefile script screwed up while processing it.
2447
244810. Incorporated PCRE 4.4.
2449
245011. The SMTP synchronization check was not operating right at the start of an
2451 SMTP session. For example, it could not catch a HELO sent before the client
2452 waited for the greeting. There is now a check for outstanding input at the
2453 point when the greeting is written. Because of the duplex, asynchronous
2454 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2455 way, but not yet received, when the check is performed.
2456
245712. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2458 on TCP/IP sockets, because this apparently causes some broken clients to
2459 timeout.
2460
246113. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2462 unchanged) from the Cygwin maintainer.
2463
246414. The code for -bV that shows what is in the binary showed "mbx" when maildir
2465 was supported instead of testing for mbx. Effectively a typo.
2466
246715. The spa authenticator server code was not checking that the input it
2468 received was valid base64.
2469
247016. The debug output line for the "set" modifier in ACLs was not showing the
2471 name of the variable that was being set.
2472
247317. Code tidy: the variable type "vtype_string" was never used. Removed it.
2474
247518. Previously, a reference to $sender_host_name did not cause a DNS reverse
2476 lookup on its own. Something else was needed to trigger the lookup. For
2477 example, a match in host_lookup or the need for a host name in a host list.
2478 Now, if $sender_host_name is referenced and the host name has not yet been
2479 looked up, a lookup is performed. If the lookup fails, the variable remains
2480 empty, and $host_lookup_failed is set to "1".
2481
248219. Added "eqi" as a case-independent comparison operator.
2483
248420. The saslauthd authentication condition could segfault if neither service
2485 nor realm was specified.
2486
248721. If an overflowing value such as "2048M" was set for message_size_limit, the
2488 error message that was logged was misleading, and incoming SMTP
2489 connections were dropped. The message is now more accurate, and temporary
2490 errors are given to SMTP connections.
2491
249222. In some error situations (such as 21 above) Exim rejects all SMTP commands
2493 (except RSET) with a 421 error, until QUIT is received. However, it was
2494 failing to send a response to QUIT.
2495
249623. The HELO ACL was being run before the code for helo_try_verify_hosts,
2497 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2498 ACL is now run after the helo_try_verify_hosts code.
2499
250024. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2501 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2502 case-independent, so other case variants are also recognized). Apparently
2503 some systems use these upper case variants.
2504
250525. If more than two messages were waiting for the same host, and a transport
2506 filter was specified for the transport, Exim sent two messages over the
2507 same TCP/IP connection, and then failed with "socket operation on non-
2508 socket" when it tried to send the third.
2509
251026. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2511
251227. The extern definition of crypt16() in expand.c was not being excluded when
2513 the OS had its own crypt16() function.
2514
251528. Added bounce_return_body as a new option, and bounce_return_size_limit
2516 as a preferred synonym for return_size_limit, both as an option and as an
2517 expansion variable.
2518
251929. Added LIBS=-liconv to OS/Makefile-OSF1.
2520
252130. Changed the default configuration ACL to relax the local part checking rule
2522 for addresses that are not in any local domains. For these addresses,
2523 slashes and pipe symbols are allowed within local parts, but the sequence
2524 /../ is explicitly forbidden.
2525
252631. SPA server authentication was not clearing the challenge buffer before
2527 using it.
2528
252932. log_message in a "warn" ACL statement was writing to the reject log as
2530 well as to the main log, which contradicts the documentation and doesn't
2531 seem right (because no rejection is happening). So I have stopped it.
2532
253333. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2534 However, I am unable to do any testing of this.
2535
253634. Fixed an infelicity in the appendfile transport. When checking directories
2537 for a mailbox, to see if any needed to be created, it was accidentally
2538 using path names with one or more superfluous leading slashes; tracing
2539 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2540
254135. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2542 contents are added to the log line that is written for every discarded
2543 recipient. (Previously a log_message setting was ignored.)
2544
254536. The ${quote: operator now quotes the string if it is empty.
2546
254737. The install script runs exim in order to find its version number. If for
2548 some reason other than non-existence or emptiness, which it checks, it
2549 could not run './exim', it was installing it with an empty version number,
2550 i.e. as "exim-". This error state is now caught, and the installation is
2551 aborted.
2552
255338. An argument was missing from the function that creates an error message
2554 when Exim fails to connect to the socket for saslauthd authentication.
2555 This could cause Exim to crash, or give a corrupted message.
2556
255739. Added isip, isip4, and isip6 to ${if conditions.
2558
255940. The ACL variables $acl_xx are now saved with the message, and can be
2560 accessed later in routers, transports, and filters.
2561
256241. The new lookup type nwildlsearch is like wildlsearch, except that the key
2563 strings in the file are not string-expanded.
2564
256542. If a MAIL command specified a SIZE value that was too large to fit into an
2566 int variable, the check against message_size_limit failed. Such values are
2567 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2568 day this will have to be increased, but I don't think I want to be around
2569 when emails are that large.
2570
2571
2572
2573Exim version 4.22
2574-----------------
2575
2576 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2577 iconv() is not standard in FreeBSD.
2578
2579 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2580 IPv6 enabled. The observed symptom was a segmentation fault on return from
2581 the function os_common_find_running_interfaces() in src/os.c.
2582
2583 3. In the check_special_case() function in daemon.c I had used "errno" as an
2584 argument name, which causes warnings on some systems. This was basically a
2585 typo, since it was named "eno" in the comments!
2586
2587 4. The code that waits for the clock to tick (at a resolution of some fraction
2588 of a second) so as to ensure message-id uniqueness was always waiting for
2589 at least one whole tick, when it could have waited for less. [This is
2590 almost certainly not relevant at current processor speeds, where it is
2591 unlikely to ever wait at all. But we try to future-proof.]
2592
2593 5. The function that sleeps for a time interval that includes fractions of a
2594 second contained a race. It did not block SIGALRM between setting the
2595 timer, and suspending (a couple of lines later). If the interval was short
2596 and the sigsuspend() was delayed until after it had expired, the suspension
2597 never ended. On busy systems this could lead to processes getting stuck for
2598 ever.
2599
2600 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2601 process, before it forks any delivery processes. The open lookup caching
2602 mechanism meant that the open file or database connection was passed into
2603 the delivery process. The problem was that delivery processes always tidy
2604 up cached lookup data. This could cause a problem for the next delivery
2605 process started by the queue runner, because the external queue runner
2606 process does not know about the closure. So the next delivery process
2607 still has data in the lookup cache. In the case of a file lookup, there was
2608 no problem because closing a file descriptor in a subprocess doesn't affect
2609 the parent. However, if the lookup was caching a connection to a database,
2610 the connection was closed, and the second delivery process was likely to
2611 see errors such as "PGSQL: query failed: server closed the connection
2612 unexpectedly". The problem has been fixed by closing all cached lookups
2613 in a queue runner before running a delivery process.
2614
2615 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2616 seem to have the "const" qualifier which it has on other OS. I've
2617 parameterised it.
2618
2619 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2620 *of the same type* (client or server) with the same public name that an
2621 error should be diagnosed.
2622
2623 9. When Exim looked up a host name for an IP address, but failed to find the
2624 original IP address when looking up the host name (a safety check), it
2625 output the message "<ip address> does not match any IP for NULL", which was
2626 confusing, to say the least. The bug was that the host name should have
2627 appeared instead of "NULL".
2628
262910. Since release 3.03, if Exim is called by a uid other than root or the Exim
2630 user that is built into the binary, and the -C or -D options is used, root
2631 privilege is dropped before the configuration file is read. In addition,
2632 logging is switched to stderr instead of the normal log files. If the
2633 configuration then re-defines the Exim user, the unprivileged environment
2634 is probably not what is expected, so Exim logs a panic warning message (but
2635 proceeds).
2636
2637 However, if deliver_drop_privilege is set, the unprivileged state may well
2638 be exactly what is intended, so the warning has been cut out in that case,
2639 and Exim is allowed to try to write to its normal log files.
2640
2641
2642Exim version 4.21
2643-----------------
2644
2645 1. smtp_return_error_details was not giving details for temporary sender
2646 or receiver verification errors.
2647
2648 2. Diagnose a configuration error if two authenticators have the same public
2649 name.
2650
2651 3. Exim used not to create the message log file for a message until the first
2652 delivery attempt. This could be confusing when incoming messages were held
2653 for policy or load reasons. The message log file is now created at the time
2654 the message is received, and an initial "Received" line is written to it.
2655
2656 4. The automatically generated man page for command line options had a minor
2657 bug that caused no ill effects; however, a more serious problem was that
2658 the procedure for building the man page automatically didn't always
2659 operate. Consequently, release 4.20 contains an out-of-date version. This
2660 shouldn't happen again.
2661
2662 5. When building Exim with embedded Perl support, the script that builds the
2663 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2664 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2665
2666 6. The freeze_tell option was not being used for messages that were frozen on
2667 arrival, either by an ACL or by local_scan().
2668
2669 7. Added the smtp_incomplete_transaction log selector.
2670
2671 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2672 was accepting AUTH without a new EHLO.
2673
2674 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2675 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2676 RFC.
2677
267810. Logging of TCP/IP connections (when configured) now happens in the main
2679 daemon process instead of the child process, so that the TCP/IP connection
2680 count is more accurate (but it can never be perfect).
2681
268211. The use of "drop" in a nested ACL was not being handled correctly in the
2683 outer ACL. Now, if condition failure induced by the nested "drop" causes
2684 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2685 or "require"), the connection is dropped.
2686
268712. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2688 that yield "discard" can now be used with an "accept" or a "discard" verb,
2689 but an error is generated for any others (because I can't see a useful way
2690 to define what should happen).
2691
269213. When an ACL is read dynamically from a file (or anywhere else), the lines
2693 are now processed in the same way as lines in the Exim configuration file.
2694 In particular, continuation lines are supported.
2695
269614. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2697
269815. Added -ti meaning -t -i.
2699
270016. Check for letters, digits, hyphens, and dots in the names of dnslist
2701 domains, and warn by logging if others are found.
2702
270317. At least on BSD, alignment is not guarenteed for the array of ifreq's
2704 returned from GIFCONF when Exim is trying to find the list of interfaces on
2705 a host. The code in os.c has been modified to copy each ifreq to an aligned
2706 structure in all cases.
2707
2708 Also, in some cases, the returned ifreq's were being copied to a 'struct
2709 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2710 means the last couple of bytes of an IPv6 address could be chopped if the
2711 ifreq contained only a normal sockaddr (14 bytes storage).
2712
271318. Named domain lists were not supported in the hosts_treat_as_local option.
2714 An entry such as +xxxx was not recognized, and was treated as a literal
2715 domain name.
2716
271719. Ensure that header lines added by a DATA ACL are included in the reject log
2718 if the ACL subsequently rejects the message.
2719
272020. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2721 MD5 (which is deprecated).
2722
272321. When testing a filter file using -bf, Exim was writing a message when it
2724 took the sender from a "From " line in the message, but it was not doing so
2725 when it took $return_path from a Return-Path: header line. It now does.
2726
272722. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2728 with a valid header line field name (a series of printing characters
2729 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2730
273123. Changed "disc" in the source to "disk" to conform to the documentation and
2732 the book and for uniformity.
2733
273424. Ignore Sendmail's -Ooption=value command line item.
2735
273625. When execve() failed while trying to run a command in a pipe transport,
2737 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2738 could be confused with a return value of 69 from the command itself. This
2739 has been changed to 127, the value the shell returns if it is asked to run
2740 a non-existent command. The wording for the related log line suggests a
2741 non-existent command as the problem.
2742
274326. If received_header_text expands to an empty string, do not add a Received:
2744 header line to the message. (Well, it adds a token one on the spool, but
2745 marks it "old" so that it doesn't get used or transmitted.)
2746
274727. Installed eximstats 1.28 (addition of -nt option).
2748
274928. There was no check for failure on the call to getsockname() in the daemon
2750 code. This can fail if there is a shortage of resources on the system, with
2751 ENOMEM, for example. A temporary error is now given on failure.
2752
275329. Contrary to the C standard, it seems that in some environments, the
2754 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2755 program. Exim now does this explicitly; it affects the formatting of
2756 timestamps using strftime().
2757
275830. If exiqsumm was given junk data, it threw up some uninitialized variable
2759 complaints. I've now initialized all the variables, to avoid this.
2760
276132. Header lines added by a system filter were not being "seen" during
2762 transport-time rewrites.
2763
276433. The info_callback() function passed to OpenSSL is set up with type void
2765 (*)(SSL *, int, int), as described somewhere. However, when calling the
2766 function (actually a macro) that sets it up, the type void(*)() is
2767 expected. I've put in a cast to prevent warnings from picky compilers.
2768
276934. If a DNS black list lookup found a CNAME record, but there were no A
2770 records associated with the domain it pointed at, Exim crashed.
2771
277235. If a DNS black list lookup returned more than one A record, Exim ignored
2773 all but the first. It now scans all returned addresses if a particular IP
2774 value is being sought. In this situation, the contents of the
2775 $dnslist_value variable are a list of all the addresses, separated by a
2776 comma and a space.
2777
277836. Tightened up the rules for host name lookups using reverse DNS. Exim used
2779 to accept a host name and all its aliases if the forward lookup for any of
2780 them yielded the IP address of the incoming connection. Now it accepts only
2781 those names whose forward lookup yields the correct IP address. Any other
2782 names are discarded. This closes a loophole whereby a rogue DNS
2783 administrator could create reverse DNS records to break through a
2784 wildcarded host restriction in an ACL.
2785
278637. If a user filter or a system filter that ran in a subprocess used any of
2787 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2788 the wrong values were passed to the pipe command ($thisaddress had the
2789 value of $0, $0 had the value of $1, etc). This bug was introduced by
2790 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2791
279238. Improved the line breaking for long SMTP error messages from ACLs.
2793 Previously, if there was no break point between 40 and 75 characters, Exim
2794 left the rest of the message alone. Two changes have been made: (a) I've
2795 reduced the minimum length to 35 characters; (b) if it can't find a break
2796 point between 35 and 75 characters, it looks ahead and uses the first one
2797 that it finds. This may give the occasional overlong line, but at least the
2798 remaining text gets split now.
2799
280039. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2801 file descriptors might be low, and that setting 1000 would always raise it.
2802 It turns out that in some environments, the limit is already over 1000 and
2803 that lowering it causes trouble. So now Exim takes care not to decrease it.
2804
280540. When delivering a message, the value of $return_path is set to $sender_
2806 address at the start of routing (routers may change the value). By an
2807 oversight, this default was not being set up when an address was tested by
2808 -bt or -bv, which affected the outcome if any router or filter referred to
2809 $return_path.
2810
281141. The idea of the "warn" ACL verb is that it adds a header or writes to the
2812 log only when "message" or "log_message" are set. However, if one of the
2813 conditions was an address verification, or a call to a nested ACL, the
2814 messages generated by the underlying test were being passed through. This
2815 no longer happens. The underlying message is available in $acl_verify_
2816 message for both "message" and "log_message" expansions, so it can be
2817 passed through if needed.
2818
281942. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2820 new expansion $bh_ to give the encoded byte string without charset
2821 translation. Translation happens only if iconv() is available; HAVE_ICONV
2822 indicates this at build time. HEADERS_CHARSET gives the charset to
2823 translate to; headers_charset can change it in the configuration, and
2824 "headers charset" can change it in an individual filter file.
2825
282643. Now that we have a default RFC 2047 charset (see above), the code in Exim
2827 that creates RFC 2047 encoded "words" labels them as that charset instead
2828 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2829 expansion operator; (ii) when Exim creates a From: line for a local
2830 message; (iii) when a header line is rewritten to include a "phrase" part.
2831
283244. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2833 buggy, causing it to skip the first lines of messages whose message ID
2834 ended in 'D'. This would not have bitten before Exim release 4.14, because
2835 message IDs were unlikely to end in 'D' before then. The effect was to have
2836 incorrect size information for certain domains.
2837
283845. #include "config.h" was missing at the start of the crypt16.c module. This
2839 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2840 noticed.
2841
284246. If there was a timeout during a "random" callout check, Exim treated it as
2843 a failure of the random address, and carried on sending RSET and the real
2844 address. If the delay was just some slowness somewhere, the response to the
2845 original RCPT would be taken as a response to RSET and so on, causing
2846 mayhem of various kinds.
2847
284847. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2849 when I implemented it. It didn't allow for the fact that some option values
2850 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2851 the right test for positive values.
2852
285348. Domain names in DNS records are case-independent. Exim always looks them up
2854 in lower case. Some resolvers return domain names in exactly the case they
2855 appear in the zone file, that is, they may contain uppercase letters. Not
2856 all resolvers do this - some return always lower case. Exim was treating a
2857 change of case by a resolver as a change of domain, similar to a widening
2858 of a domain abbreviation. This triggered its re-routing code and so it was
2859 trying to route what was effectively the same domain again. This normally
2860 caused routing to fail (because the router wouldn't handle the domain
2861 twice). Now Exim checks for this case specially, and just changes the
2862 casing of the domain that it ultimately uses when it transmits the message
2863 envelope.
2864
286549. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2866 module.
2867
286850. If a filter generated a file delivery with a non-absolute name (possible if
2869 no home directory exists for the router), the forbid_file option was not
2870 forbidding it.
2871
287251. Added '&' feature to dnslists, to provide bit mask matching in addition to
2873 the existing equality matching.
2874
287552. Exim was using ints instead of ino_t variables in some places where it was
2876 dealing with inode numbers.
2877
287853. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2879 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2880 called TMPDIR, and if it finds it is different, it changes its value.
2881
288254. The smtp_printf() function is now made available to local_scan() so
2883 additional output lines can be written before returning. There is also an
2884 smtp_fflush() function to enable the detection of a dropped connection.
2885 The variables smtp_input and smtp_batched_input are exported to
2886 local_scan().
2887
288855. Changed the default runtime configuration: the message "Unknown user"
2889 has been removed from the ACL, and instead placed on the localuser router,
2890 using the cannot_route_message feature. This means that any verification
2891 failures that generate their own messages won't get overridden. Similarly,
2892 the "Unrouteable address" message that was in the ACL for unverifiable
2893 relay addresses has also been removed.
2894
289556. Added hosts_avoid_esmtp to the smtp transport.
2896
289757. The exicyclog script was not checking for the esoteric option
2898 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2899 will work only if exicyclog is run under the appropriate euid.
2900
290158. Following a discussion on the list, the rules by which Exim recognises line
2902 endings on incoming messages have been changed. The -dropcr and drop_cr
2903 options are now no-ops, retained only for backwards compatibility. The
2904 following line terminators are recognized: LF CRLF CR. However, special
2905 processing applies to CR:
2906
2907 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2908 nor a local message in the state where . is a terminator.
2909
2910 (ii) If a bare CR is encountered in a header line, an extra space is added
2911 after the line terminator so as not to end the header. The reasoning
2912 behind this is that bare CRs in header lines are most likely either
2913 to be mistakes, or people trying to play silly games.
2914
291559. The size of a message, as listed by "-bp" or in the Exim monitor window,
2916 was being incorrectly given as 18 bytes larger than it should have been.
2917 This is a VOB (very old bug).
2918
291960. This may never have affected anything current, but just in case it has:
2920 When the local host is found other than at the start of a list of hosts,
2921 the local host, those with the same MX, and any that follow, are discarded.
2922 When the list in question was part of a longer list of hosts, the following
2923 hosts (not currently being processed) were also being discarded. This no
2924 longer happens. I'm not sure if this situation could ever has previously
2925 arisen.
2926
292761. Added the "/MX" feature to lists of hosts in the manualroute and query
2928 program routers.
2929
293062. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2931 header. This is something that is recommended in a new Internet Draft, and
2932 is something that is documented as being done by Sendmail. There are two
2933 possible values. For messages generated by the autoreply transport, Exim
2934 adds:
2935
2936 Auto-Submitted: auto-replied
2937
2938 whereas for all other generated messages (e.g. bounces) it adds
2939
2940 Auto-Submitted: auto-generated
2941
294263. The "personal" condition in filters now includes a test for the
2943 Auto-Submitted: header. If it contains the string "auto-" the message it
2944 not considered personal.
2945
294664. Added rcpt_include_affixes as a generic transport option.
2947
294865. Added queue_only_override (default true).
2949
295066. Added the syslog_duplication option.
2951
295267. If what should have been the first header line of a message consisted of
2953 a space followed by a colon, Exim was mis-interpreting it as a header line.
2954 It isn't of course - it is syntactically invalid and should therefore be
2955 treated as the start of the message body. The misbehaviour could have
2956 caused a number of strange effects, including loss of data in subsequent
2957 header lines, and spool format errors.
2958
295968. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2960 client host had authenticated. This control can now be exercised by an ACL
2961 for more flexibility.
2962
296369. By default, callouts do not happen when testing with -bh. There is now a
2964 variant, -bhc, which does actually run the callout code, including
2965 consulting and updating the callout cache.
2966
296770. Added support for saslauthd authentication, courtesy of Alexander
2968 Sabourenkov.
2969
297071. If statvfs() failed on the spool or log directories while checking their
2971 size for availability, Exim confusingly gave the error "space shortage".
2972 Furthermore, in debugging mode it crashed with a floating point exception.
2973 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2974 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2975 serious problem, Exim now writes to the main and panic logs when this
2976 happens, with details of the failure. It then refuses to accept the
2977 incoming message, giving the message "spool directory problem" or "log
2978 directory problem" with a 421 code for SMTP messages.
2979
298072. When Exim is about to re-exec itself, it ensures that the file descriptors
2981 0, 1, and 2 exist, because some OS complain for execs without them (see
2982 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2983 descriptors. However, the code omitted to check that the open succeeded,
2984 causing mysterious errors if for some reason the permissions on /dev/null
2985 got screwed. Now Exim writes a message to the main and panic logs, and
2986 bombs out if it can't open /dev/null.
2987
298873. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2989 interact so that it is all more flexible. It is supposed to remain
2990 backwards compatible. Also added extra_local_interfaces.
2991
299274. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2993 to bomb out with an assertion failure - to the client this appears as a
2994 connection drop. This problem occurs in the part of the code that was taken
2995 from the Samba project. Fortunately, the assertion is in a very simple
2996 function, so I have fixed this by reproducing the function inline in the
2997 one place where it is called, and arranging for authentication to fail
2998 instead of killing the process with assert().
2999
300075. The SPA client code was not working when the server requested OEM rather
3001 than Unicode encoding.
3002
300376. Added code to make require_files with a specific uid setting more usable in
3004 the case where statting the file as root fails - usually a non-root-mounted
3005 NFS file system. When this happens and the failure is EACCES, Exim now
3006 forks a subprocess and does the per-uid checking as the relevant uid.
3007
300877. Added process_log_path.
3009
301078. If log_file_path was not explicitly set, a setting of check_log_space or
3011 check_log_inodes was ignored.
3012
301379. If a space check for the spool or log partitions fails, the incident is now
3014 logged. Of course, in the latter case the data may get lost...
3015
301680. Added the %p formatting code to string_format() so that it can be used to
3017 print addresses in debug_print(). Adjusted all the address printing in the
3018 debugging in store.c to use %p rather than %d.
3019
302081. There was a concern that a line of code in smtp_in.c could overflow a
3021 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3022 initially expressed, the concern was not well-founded, because trailing
3023 spaces are removed early. However, if the trailing spaces were followed by
3024 a NULL, they did not get removed, so the overflow was possible. Two fixes
3025 were applied:
3026
3027 (a) I re-wrote the offending code in a cleaner fashion.
3028 (b) If an incoming SMTP command contains a NULL character, it is rejected
3029 as invalid.
3030
303182. When Exim changes uid/gid to the Exim user at daemon start time, it now
3032 runs initgroups(), so that if the Exim user is in any additional groups,
3033 they will be used during message reception.
3034
3035
3036Exim version 4.20
3037-----------------
3038
3039The change log for 4.20 and earlier releases has been archived.
3040
3041****