Applied Michael Haardt's HP-UX patch.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
5bd022fe 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.160 2005/06/17 09:02:40 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
4aac9b49
PH
130PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
131 the configuration file to be ":syslog", then the script "guesses" where
132 the logs files are, rather than using the compiled in default. In our
133 case the guess is not the same as the compiled default, so the script
134 suddenly stopped working when I started to use syslog. The patch checks
135 to see if log_file_path is "". If so, it attempts to read it from exim
136 with no configuration file to get the compiled in version, before it
137 falls back to the previous guessing code."
138
294520c8
TK
139TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
140 implementing BATV in an Exim configuration. See NewStuff for the gory
141 details.
142
5bd022fe
PH
143PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
144 Makefile that are specific to HP-UX.
145
e5a9dba6 146
7982096b
PH
147Exim version 4.51
148-----------------
149
1a46a8c5
PH
150TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
151 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
152
2f079f46 153TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
154
155TK/03 Fix ACL "regex" condition not being available in MIME ACL.
156
157PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
158 to test Sieve filters that use "vacation".
159
160PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
161 that changes the way the GnuTLS parameters are stored in the cache file.
162 The new format can be generated externally. For backward compatibility,
163 if the data in the cache doesn't make sense, Exim assumes it has read an
164 old-format file, and it generates new data and writes a new file. This
165 means that you can't go back to an older release without removing the
166 file.
167
168PH/03 A redirect router that has both "unseen" and "one_time" set does not
169 work if there are any delivery delays because "one_time" forces the
170 parent to be marked "delivered", so its unseen clone is never tried
171 again. For this reason, Exim now forbids the simultaneous setting of
172 these two options.
173
174PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
175 redirected to themselves ("homonym" addresses). Read the long ChangeLog
176 entry if you want to know the details. The fix, however, neglected to
177 consider the case when local delivery batching is involved. The test for
178 "previously delivered" was not happening when checking to see if an
179 address could be batched with a previous (undelivered) one; under
180 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 181 address.
1a46a8c5
PH
182
183PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
184 in its include files, and this causes problems building Exim.
185
186PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
187 header_syntax) but Exim was just ignoring anything given after a slash.
188 In particular, this caused confusion with an attempt to use "verify =
189 reverse_host_lookup/defer_ok". An error is now given when options are
190 supplied for verify items that do not have them. (Maybe reverse_host_
191 lookup should have a defer_ok option, but that's a different point.)
192
193PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
194 defined by RFC 821) to 2048, because there were problems with some AUTH
195 commands, and RFC 1869 says the size should be increased for extended
196 SMTP commands that take arguments.
197
198PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
199 Finch).
200
201PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
202 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 203
49c2d5ea
PH
204PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
205 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
206 whatever). Otherwise the password may appear in the log. Following change
207 PH/42 below, there is no longer a chance of it appearing in a bounce
208 message.
49c2d5ea 209
bf759a8b
PH
210PH/11 Installed exipick version 20050225.0 from John Jetmore.
211
83364d30
PH
212PH/12 If the last host in a fallback_hosts list was multihomed, only the first
213 of its addresses was ever tried. (Bugzilla bug #2.)
214
7999bbd7
PH
215PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
216 the result incorrectly in the debug output. (It correctly added a newline
217 to what was transported.)
218
7dbf77c9
PH
219TF/01 Added $received_time.
220
74e0617f
PH
221PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
222 commented out examples of how to interface to a virus scanner and to
223 SpamAssassin. Also added commented examples of av_scanner and
224 spamd_address settings.
225
2f079f46
PH
226PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
227 and controls are allowed in which ACLs. There were a couple of minor
228 errors. Some of the entries in the conditions table (which is a table of
229 where they are NOT allowed) were getting very unwieldy; rewrote them as a
230 negation of where the condition IS allowed.
231
8c841523
PH
232PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
233
7766a4f0
PH
234PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
235 header file does not have a version number, so I've had to invent a new
236 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
237 API. The code is untested by me (my Linux distribution still has 0.3.2 of
238 radiusclient), but it was contributed by a Radius user.
239
8b417f2c
PH
240PH/18 Installed Lars Mainka's patch for the support of CRL collections in
241 files or directories, for OpenSSL.
242
901f42cb
PH
243PH/19 When an Exim process that is running as root has to create an Exim log
244 file, it does so in a subprocess that runs as exim:exim so as to get the
245 ownership right at creation (otherwise, other Exim processes might see
246 the file with the wrong ownership). There was no test for failure of this
247 fork() call, which would lead to the process getting stuck as it waited
248 for a non-existent subprocess. Forks do occasionally fail when resources
249 run out. I reviewed all the other calls to fork(); they all seem to check
250 for failure.
251
f9b9210e
PH
252PH/20 When checking for unexpected SMTP input at connect time (before writing
253 the banner), Exim was not dealing correctly with a non-positive return
254 from the read() function. If the client had disconnected by this time,
255 the result was a log entry for a synchronization error with an empty
256 string after "input=" when read() returned zero. If read() returned -1
257 (an event I could not check), uninitialized data bytes were printed.
258 There were reports of junk text (parts of files, etc) appearing after
259 "input=".
260
54cdb463
PH
261PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
262
cf00dad6
PH
263PH/22 Added support for macro redefinition, and (re)definition in between
264 driver and ACL definitions.
265
acb1b346
PH
266PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
267 forgetting to use the resulting value; it was using the unexpanded value.
268
c5ddb310
PH
269PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
270 hadn't been configured. The fix is from Juergen Kreileder, who
271 understands it better than I do:
272
273 "Here's what I see happening with three configured cyrus_sasl
274 authenticators configured (plain, login, cram-md5):
275
276 On startup auth_cyrus_sasl_init() gets called for each of these.
277 This means three calls to sasl_listmech() without a specified mech_list.
278 => SASL tests which mechs of all available mechs actually work
279 => three warnings about OTP not working
280 => the returned list contains: plain, login, cram-md5, digest-md5, ...
281
282 With the patch, sasl_listmech() also gets called three times. But now
283 SASL's mech_list option is set to the server_mech specified in the the
284 authenticator. Or in other words, the answer from sasl_listmech()
285 gets limited to just the mech you're testing for (which is different
286 for each call.)
287 => the return list contains just 'plain' or 'login', 'cram-md5' or
288 nothing depending on the value of ob->server_mech.
289
290 I've just tested the patch: Authentication still works fine,
291 unavailable mechs specified in the exim configuration are still
292 caught, and the auth.log warnings about OTP are gone."
293
31619da6
PH
294PH/25 When debugging is enabled, the contents of the command line are added
295 to the debugging output, even when log_selector=+arguments is not
296 specified.
297
bebaf0fc
PH
298PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
299 answer is "GNU", and only if the return is "GNU/something" is the answer
300 "Linux".
301
475fe28a
PH
302PH/27 $acl_verify_message is now set immediately after the failure of a
303 verification in an ACL, and so is available in subsequent modifiers. In
304 particular, the message can be preserved by coding like this:
305
306 warn !verify = sender
307 set acl_m0 = $acl_verify_message
308
309 Previously, $acl_verify_message was set only while expanding "message"
310 and "log_message" when a very denied access.
311
7e8bec7a
PH
312PH/28 Modified OS/os.c-Linux with
313
314 -#ifndef OS_LOAD_AVERAGE
315 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
316
317 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
318 nomenclature these days.)
319
e4a89c47
PH
320PH/29 Installed patch from the Sieve maintainer that adds the options
321 sieve_useraddress and sieve_subaddress to the redirect router.
322
5ca2a9a1
PH
323PH/30 In these circumstances:
324 . Two addresses routed to the same list of hosts;
325 . First host does not offer TLS;
326 . First host accepts first address;
327 . First host gives temporary error to second address;
328 . Second host offers TLS and a TLS session is established;
329 . Second host accepts second address.
330 Exim incorrectly logged both deliveries with the TLS parameters (cipher
331 and peerdn, if requested) that were in fact used only for the second
332 address.
7e8bec7a 333
c688b954
PH
334PH/31 When doing a callout as part of verifying an address, Exim was not paying
335 attention to any local part prefix or suffix that was matched by the
336 router that accepted the address. It now behaves in the same way as it
337 does for delivery: the affixes are removed from the local part unless
338 rcpt_include_affixes is set on the transport.
339
fed77020
PH
340PH/32 Add the sender address, as F=<...>, to the log line when logging a
341 timeout during the DATA phase of an incoming message.
342
7fe1560f
PH
343PH/33 Sieve envelope tests were broken for match types other than :is. I have
344 applied a patch sanctioned by the Sieve maintainer.
c688b954 345
ebb6e6d5
PH
346PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
347 the uid or gid is negative. A case of a negative gid caused this to be
348 noticed. The fix allows for either to be negative.
349
9c4e8f60
PH
350PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
351 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
352 overlooked.
353
354PH/36 The change PH/12 above was broken. Fixed it.
355
d7174846
PH
356PH/37 Exim used to check for duplicate addresses in the middle of routing, on
357 the grounds that routing the same address twice would always produce the
358 same answer. This might have been true once, but it is certainly no
359 longer true now. Routing a child address may depend on the previous
360 routing that produced that child. Some complicated redirection strategies
361 went wrong when messages had multiple recipients, and made Exim's
362 behaviour dependent on the order in which the addresses were given.
363
364 I have moved the duplicate checking until after the routing is complete.
365 Exim scans the addresses that are assigned to local and remote
366 transports, and removes any duplicates. This means that more work will be
367 done, as duplicates will always all be routed, but duplicates are
368 presumably rare, so I don't expect this is of any significance.
369
370 For deliveries to pipes, files, and autoreplies, the duplicate checking
371 still happens during the routing process, since they are not going to be
372 routed further.
373
cfe75fc3
PH
374PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
375 It corrects a timeout issue with spamd. This is Ian's comment: "The
376 background is that sometimes spamd either never reads data from a
377 connection it has accepted, or it never writes response data. The exiscan
378 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
379 blindly assumes that writes won't block so it may never time out."
380
be22d70e
PH
381PH/39 Allow G after quota size as well as K and M.
382
0612b098
PH
383PH/40 The value set for $authenticated_id in an authenticator may not contain
384 binary zeroes or newlines because the value is written to log lines and
385 to spool files. There was no check on this. Now the value is run through
386 the string_printing() function so that such characters are converted to
387 printable escape sequences.
388
2e0c1448
PH
389PH/41 $message_linecount is a new variable that contains the total number of
390 lines in the message. Compare $body_linecount, which is the count for the
391 body only.
392
447d236c
PH
393PH/42 Exim no longer gives details of delivery errors for specific addresses in
394 bounce and delay warning messages, except in certain special cases, which
395 are as follows:
396
397 (a) An SMTP error message from a remote host;
398 (b) A message specified in a :fail: redirection;
399 (c) A message specified in a "fail" command in a system filter;
400 (d) A message specified in a FAIL return from the queryprogram router;
401 (e) A message specified by the cannot_route_message router option.
402
403 In these cases only, Exim does include the error details in bounce and
404 warning messages. There are also a few cases where bland messages such
405 as "unrouteable address" or "local delivery error" are given.
406
d20976dc
PH
407PH/43 $value is now also set for the "else" part of a ${run expansion.
408
f656d135
PH
409PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
410 being worked on, but at least Exim now implements the latest version to
411 play with."
412
2e2a30b4
PH
413PH/45 In a pipe transport, although a timeout while waiting for the pipe
414 process to complete was treated as a delivery failure, a timeout while
415 writing the message to the pipe was logged, but erroneously treated as a
416 successful delivery. Such timeouts include transport filter timeouts. For
417 consistency with the overall process timeout, these timeouts are now
418 treated as errors, giving rise to delivery failures by default. However,
419 there is now a new Boolean option for the pipe transport called
420 timeout_defer, which, if set TRUE, converts the failures into defers for
421 both kinds of timeout. A transport filter timeout is now identified in
422 the log output.
423
9176e9f0
PH
424PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
425 systems where "make" and "gmake" are different, calling "gmake" at top
426 level broke things. I've arranged for the value of $(MAKE) to be passed
427 from the Makefile to this script so that it can call the same version of
428 "make".
429
7982096b 430
bbe902f0
PH
431A note about Exim versions 4.44 and 4.50
432----------------------------------------
433
434Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
435changes of various kinds. As a consequence, a big documentation update was
436needed. This delayed the release for rather longer than seemed good, especially
437in the light of a couple of (minor) security issues. Therefore, the changes
438that fixed bugs were backported into 4.43, to create a 4.44 maintenance
439release. So 4.44 and 4.50 are in effect two different branches that both start
440from 4.43.
441
442I have left the 4.50 change log unchanged; it contains all the changes since
4434.43. The change log for 4.44 is below; many of its items are identical to
444those for 4.50. This seems to be the most sensible way to preserve the
445historical information.
446
447
f7b63901 448Exim version 4.50
495ae4b0
PH
449-----------------
450
5fe762f6
PH
451 1. Minor wording change to the doc/README.SIEVE file.
452
139059f6 453 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 454 computation of the current number of files was incorrect.
495ae4b0 455
7086e875
PH
456 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
457 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
458 place.
459
35af9f61
PH
460 4. Give more explanation in the error message when the command for a transport
461 filter fails to execute.
462
b668c215
PH
463 5. There are several places where Exim runs a non-Exim command in a
464 subprocess. The SIGUSR1 signal should be disabled for these processes. This
465 was being done only for the command run by the queryprogram router. It is
466 now done for all such subprocesses. The other cases are: ${run, transport
467 filters, and the commands run by the lmtp and pipe transports.
468
a494b1e1
PH
469 6. Added CONFIGURE_GROUP build-time option.
470
471 7. Some older OS have a limit of 256 on the maximum number of file
472 descriptors. Exim was using setrlimit() to set 1000 as a large value
473 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
474 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 475
c5fcb476
PH
476 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
477 was an oversight, and furthermore, ever since the addition of extra
478 controls (e.g. 4.43/32), the checks on when to allow different forms of
479 "control" were broken. There should now be diagnostics for all cases when a
480 control that does not make sense is encountered.
481
69358f02
PH
482 9. Added the /retain_sender option to "control=submission".
483
5be20824
PH
48410. $recipients is now available in the predata ACL (oversight).
485
eb2c0248
PH
48611. Tidy the search cache before the fork to do a delivery from a message
487 received from the command line. Otherwise the child will trigger a lookup
488 failure and thereby defer the delivery if it tries to use (for example) a
489 cached ldap connection that the parent has called unbind on.
490
2a3eea10
PH
49112. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
492 of $address_data from the recipient verification was clobbered by the
493 sender verification.
494
49513. The value of address_data from a sender verification is now available in
496 $sender_address_data in subsequent conditions in the ACL statement.
497
23c7ff99
PH
49814. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
499
4deaf07d
PH
50015. Added a new option "connect=<time>" to callout options, to set a different
501 connection timeout.
502
926e1192
PH
50316. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
504 was its contents. (It was OK if the option was not defined at all.)
505
650edc6f
PH
50617. A "Completed" log line is now written for messages that are removed from
507 the spool by the -Mrm option.
508
2c7db3f5
PH
50918. New variables $sender_verify_failure and $recipient_verify_failure contain
510 information about exactly what failed.
511
3d235903
PH
51219. Added -dd to debug only the daemon process.
513
7c7ad977
PH
51420. Incorporated Michael Haardt's patch to ldap.c for improving the way it
515 handles timeouts, both on the server side and network timeouts. Renamed the
516 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
517
981756db
PH
51821. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
519 It is now set to "smtps".
520
d4eb88df
PH
52122. $host_address is now set to the target address during the checking of
522 ignore_target_hosts.
523
52423. When checking ignore_target_hosts for an ipliteral router, no host name was
525 being passed; this would have caused $sender_host_name to have been used if
526 matching the list had actually called for a host name (not very likely,
527 since this list is usually IP addresses). A host name is now passed as
528 "[x.x.x.x]".
529
7d468ab8
PH
53024. Changed the calls that set up the SIGCHLD handler in the daemon to use the
531 code that specifies a non-restarting handler (typically sigaction() in
532 modern systems) in an attempt to fix a rare and obscure crash bug.
533
53425. Narrowed the window for a race in the daemon that could cause it to ignore
535 SIGCHLD signals. This is not a major problem, because they are used only to
536 wake it up if nothing else does.
537
62c0818f
PH
53826. A malformed maildirsize file could cause Exim to calculate negative values
539 for the mailbox size or file count. Odd effects could occur as a result.
540 The maildirsize information is now recalculated if the size or filecount
541 end up negative.
542
26034054
PH
54327. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
544 support for a long time. Removed HAVE_SYS_VFS_H.
545
af66f652
PH
54628. Installed the latest version of exipick from John Jetmore.
547
90af77f4
PH
54829. In an address list, if the pattern was not a regular expression, an empty
549 subject address (from a bounce message) matched only if the pattern was an
550 empty string. Non-empty patterns were not even tested. This was the wrong
551 because it is perfectly reasonable to use an empty address as part of a
552 database query. An empty address is now tested by patterns that are
553 lookups. However, all the other forms of pattern expect the subject to
554 contain a local part and a domain, and therefore, for them, an empty
555 address still always fails if the pattern is not itself empty.
556
d8ef3577
PH
55730. Exim went into a mad DNS loop when attempting to do a callout where the
558 host was specified on an smtp transport, and looking it up yielded more
559 than one IP address.
560
5cb8cbc6
PH
56131. Re-factored the code for checking spool and log partition space into a
562 function that finds that data and another that does the check. The former
563 is then used to implement four new variables: $spool_space, $log_space,
564 $spool_inodes, and $log_inodes.
565
14702f5b
PH
56632. The RFC2047 encoding function was originally intended for short strings
567 such as real names; it was not keeping to the 75-character limit for
568 encoded words that the RFC imposes. It now respects the limit, and
569 generates multiple encoded words if necessary. To be on the safe side, I
570 have increased the buffer size for the ${rfc2047: expansion operator from
571 1024 to 2048 bytes.
572
063b1e99
PH
57333. It is now permitted to omit both strings after an "if" condition; if the
574 condition is true, the result is "true". As before, when the second string
575 is omitted, a false condition yields an empty string. This makes it less
576 cumbersome to write custom ACL and router conditions.
577
652e1b65
PH
57834. Failure to deliver a bounce message always caused it to be frozen, even if
579 there was an errors_to setting on the router. The errors_to setting is now
580 respected.
581
6f0c9a4f
PH
58235. If an IPv6 address is given for -bh or -bhc, it is now converted to the
583 canonical form (fully expanded) before being placed in
584 $sender_host_address.
585
33397d19
PH
58636. The table in the code that translates DNS record types into text (T_A to
587 "A" for instance) was missing entries for NS and CNAME. It is just possible
588 that this could have caused confusion if both these types were looked up
589 for the same domain, because the text type is used as part of Exim's
590 per-process caching. But the chance of anyone hitting this buglet seems
591 very small.
592
7bb56e1f
PH
59337. The dnsdb lookup has been extended in a number of ways.
594
595 (1) There is a new type, "zns", which walks up the domain tree until it
596 finds some nameserver records. It should be used with care.
597
ea3bc19b
PH
598 (2) There is a new type, "mxh", which is like "mx" except that it returns
599 just the host names, not the priorities.
600
601 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
602 looked up. The behaviour when one of the lookups defers can be
603 controlled by a keyword.
7bb56e1f 604
ea3bc19b 605 (4) It is now possible to specify the separator character for use when
7bb56e1f 606 multiple records are returned.
33397d19 607
0bcb2a0e
PH
60838. The dnslists ACL condition has been extended: it is now possible to supply
609 a list of IP addresses and/or domains to be looked up in a particular DNS
610 domain.
611
2ac0e484
PH
61239. Added log_selector=+queue_time_overall.
613
4e1fde53
PH
61440. When running the queue in the test harness, wait just a tad after forking a
615 delivery process, to get repeatability of debugging output.
616
de365ded
PH
61741. Include certificate and key file names in error message when GnuTLS fails
618 to set them up, because the GnuTLS error message doesn't include the name
619 of the failing file when there is a problem reading it.
620
f05da2e8
PH
62142. Allow both -bf and -bF in the same test run.
622
d6453af2
PH
62343. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
624
f7b63901
PH
62544. The "Exiscan patch" is now merged into the mainline Exim source.
626
62745. Sometimes the final signoff response after QUIT could fail to get
628 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
629 < 0 before doing a fflush(). This bug looks as though it goes back to the
630 introduction of TLS in release 3.20, but "sometimes" must have been rare
631 because the tests only now provoked it.
632
a444213a
PH
63346. Reset the locale to "C" after calling embedded Perl, in case it was changed
634 (this can affect the format of dates).
635
0ec020ea
PH
63647. exim_tidydb, when checking for the continued existence of a message for
637 which it has found a message-specific retry record, was not finding
638 messages that were in split spool directories. Consequently, it was
639 deleting retry records that should have stayed in existence.
640
b1206957
PH
64148. Steve fixed some bugs in eximstats.
642
64349. The SPA authentication driver was not abandoning authentication and moving
644 on to the next authenticator when an expansion was forced to fail,
645 contradicting the general specification for all authenticators. Instead it
646 was generating a temporary error. It now behaves as specified.
647
26dd5a95
PH
64850. The default ordering of permitted cipher suites for GnuTLS was pessimal
649 (the order specifies the preference for clients). The order is now AES256,
650 AES128, 3DES, ARCFOUR128.
651
343b2385
PH
65251. Small patch to Sieve code - explicitly set From: when generating an
653 autoreply.
654
1c5466b9
PH
65552. Exim crashed if a remote delivery caused a very long error message to be
656 recorded - for instance if somebody sent an entire SpamAssassin report back
657 as a large number of 550 error lines. This bug was coincidentally fixed by
658 increasing the size of one of Exim's internal buffers (big_buffer) that
659 happened as part of the Exiscan merge. However, to be on the safe side, I
660 have made the code more robust (and fixed the comments that describe what
661 is going on).
662
55ee9ee3
PH
66353. Now that there can be additional text after "Completed" in log lines (if
664 the queue_time_overall log selector is set), a one-byte patch to exigrep
665 was needed to allow it to recognize "Completed" as not the last thing in
666 the line.
667
d38f8232
PH
66854. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
669 patch that reportedly fixes this has been added. I am not expert enough to
670 create a test for it. This is what the patch creator wrote:
671
672 "I found a little strange behaviour of ldap code when working with
673 Windows 2003 AD Domain, where users was placed in more than one
674 Organization Units. When I tried to give exim partial DN, the exit code
675 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
676 But simultaneously result of request was absolutely normal ldap result,
677 so I produce this patch..."
678
3295e65b
PH
679 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
680 REFERENCE, so I have modified the code to exclude the patch when that macro
681 is not defined.
682
7102e136
PH
68355. Some experimental protocols are using DNS PTR records for new purposes. The
684 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
685 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
686 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 687 CAN-2005-0021
7102e136 688
3ca0ba97
PH
68956. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
690
c2bcbe20
PH
69157. Double the size of the debug message buffer (to 2048) so that more of very
692 long debug lines gets shown.
693
18ce445d
PH
69458. The exicyclog utility now does better if the number of log files to keep
695 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
696
1f5b4c3d
PH
69759. Two changes related to the smtp_active_hostname option:
698
699 (1) $smtp_active_hostname is now available as a variable.
700 (2) The default for smtp_banner uses $smtp_active_hostname instead
701 of $primary_hostname.
702
b975ba52
PH
70360. The host_aton() function is supposed to be passed a string that is known
704 to be a valid IP address. However, in the case of IPv6 addresses, it was
705 not checking this. This is a hostage to fortune. Exim now panics and dies
706 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
707 from a dnsdb PTR lookup with an IPv6 address that had more than 8
708 components; fortuitously, this particular loophole had already been fixed
709 by change 4.50/55 above.
710
711 If there are any other similar loopholes, the new check in host_aton()
712 itself should stop them being exploited. The report I received stated that
713 data on the command line could provoke the exploit when Exim was running as
714 exim, but did not say which command line option was involved. All I could
715 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
716 running as the user.
ea3a6f44 717 CAN-2005-0021
85b87bc2
PH
718
71961. There was a buffer overflow vulnerability in the SPA authentication code
720 (which came originally from the Samba project). I have added a test to the
721 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 722 CAN-2005-0022
b975ba52 723
17ffcae7
PH
72462. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
725 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
726
d95f9fdb
PH
72763. The daemon start-up calls getloadavg() while still root for those OS that
728 need the first call to be done as root, but it missed one case: when
729 deliver_queue_load_max is set with deliver_drop_privilege. This is
730 necessary for the benefit of the queue runner, because there is no re-exec
731 when deliver_drop_privilege is set.
732
86b8287f
PH
73364. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
734 This has been fixed.
735
60dc5e56
PH
73665. Caching of lookup data for "hosts =" ACL conditions, when a named host list
737 was in use, was not putting the data itself into the right store pool;
738 consequently, it could be overwritten for a subsequent message in the same
739 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
740 the caching.)
741
533244af
PH
74266. Added hosts_max_try_hardlimit to the smtp transport, default 50.
743
a5a28604
PH
74467. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
745 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
746 the function were treating the return as a boolean value, which happened to
747 work because 0=false and not-0=true, but is not correct code.
748
7e634d24
PH
74968. The host_aton() function was not handling scoped IPv6 addresses (those
750 with, for example, "%eth0" on the end) correctly.
751
3e11c26b
PH
75269. Fixed some compiler warnings in acl.c for the bitmaps specified with
753 negated items (that is, ~something) in unsigned ints. Some compilers
754 apparently mutter when there is no cast.
755
6729cf78
PH
75670. If an address verification called from an ACL failed, and did not produce a
757 user-specific message (i.e. there was only a "system" message), nothing was
758 put in $acl_verify_message. In this situation, it now puts the system
759 message there.
760
00f00ca5
PH
76171. Change 4.23/11 added synchronization checking at the start of an SMTP
762 session; change 4.31/43 added the unwanted input to the log line - except
763 that it did not do this in the start of session case. It now does.
764
c9bdd01c
PH
76572. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
766 This is wrong and can cause the other end to generate a synchronization
767 error if it is another Exim or anything else that does the synchronization
768 check. A QUIT command is no longer sent after a timeout.
769
d43194df
PH
77073. $host_lookup_deferred has been added, to make it easier to detect DEFERs
771 during host lookups.
772
fe5b5d0b
PH
77374. The defer_ok option of callout verification was not working if it was used
774 when verifying addresses in header lines, that is, for this case:
775
776 verify = header_sender/callout=defer_ok
777
76a2d7ba
PH
77875. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
779 those file descriptors could be used for SMTP connections. If anything
780 wrote to stderr (the example that came up was "warn" in embedded Perl), it
781 could be sent to the SMTP client, causing chaos. The daemon now opens
782 stdin, stdout, and stderr to /dev/null when it puts itself into the
783 background.
784
78576. Arrange for output from Perl's "warn" command to be written to Exim's main
786 log by default. The user can override this with suitable Perl magic.
787
04f7d5b9
PH
78877. The use of log_message on a "discard" ACL verb, which is supposed to add to
789 the log message when discard triggers, was not working for the DATA ACL or
790 for the non-SMTP ACL.
791
bc60667e
PH
79278. Error message wording change in sieve.c.
793
bb6e88ff
PH
79479. If smtp_accept_max_per_host was set, the number of connections could be
795 restricted to fewer than expected, because the daemon was trying to set up
796 a new connection before checking whether the processes handling previous
797 connections had finished. The check for completed processes is now done
798 earlier. On busy systems, this bug wouldn't be noticed because something
799 else would have woken the daemon, and it would have reaped the completed
800 process earlier.
801
1e70f85b
PH
80280. If a message was submitted locally by a user whose login name contained one
803 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
804 It caused a spool format error. I have fixed the spool reading code. A
805 related problem was that the "from" clause in the Received: line became
806 illegal because of the space(s). It is now covered by ${quote_local_part.
807
80881. Included the latest eximstats from Steve (adds average sizes to HTML Top
809 tables).
810
4e01f9d6
PH
81182. Updated OS/Makefile-AIX as per message from Mike Meredith.
812
1ee1cef2
PH
81383. Patch from Sieve maintainer to fix unterminated string problem in
814 "vacation" handling.
815
6e2b4ccc
PH
81684. Some minor changes to the Linux configuration files to help with other
817 OS variants using glibc.
818
8e669ac1
PH
81985. One more patch for Sieve to update vacation handling to latest spec.
820
495ae4b0 821
bbe902f0
PH
822----------------------------------------------------
823See the note above about the 4.44 and 4.50 releases.
824----------------------------------------------------
825
826
827Exim version 4.44
828-----------------
829
830 1. Change 4.43/35 introduced a bug that caused file counts to be
831 incorrectly computed when quota_filecount was set in an appendfile
832 transport
833
834 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
835 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
836 place.
837
838 3. Give more explanation in the error message when the command for a transport
839 filter fails to execute.
840
841 4. There are several places where Exim runs a non-Exim command in a
842 subprocess. The SIGUSR1 signal should be disabled for these processes. This
843 was being done only for the command run by the queryprogram router. It is
844 now done for all such subprocesses. The other cases are: ${run, transport
845 filters, and the commands run by the lmtp and pipe transports.
846
847 5. Some older OS have a limit of 256 on the maximum number of file
848 descriptors. Exim was using setrlimit() to set 1000 as a large value
849 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
850 systems. I've change it so that if it can't get 1000, it tries for 256.
851
852 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
853 was an oversight, and furthermore, ever since the addition of extra
854 controls (e.g. 4.43/32), the checks on when to allow different forms of
855 "control" were broken. There should now be diagnostics for all cases when a
856 control that does not make sense is encountered.
857
858 7. $recipients is now available in the predata ACL (oversight).
859
860 8. Tidy the search cache before the fork to do a delivery from a message
861 received from the command line. Otherwise the child will trigger a lookup
862 failure and thereby defer the delivery if it tries to use (for example) a
863 cached ldap connection that the parent has called unbind on.
864
865 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
866 of $address_data from the recipient verification was clobbered by the
867 sender verification.
868
86910. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
870 was its contents. (It was OK if the option was not defined at all.)
871
87211. A "Completed" log line is now written for messages that are removed from
873 the spool by the -Mrm option.
874
87512. $host_address is now set to the target address during the checking of
876 ignore_target_hosts.
877
87813. When checking ignore_target_hosts for an ipliteral router, no host name was
879 being passed; this would have caused $sender_host_name to have been used if
880 matching the list had actually called for a host name (not very likely,
881 since this list is usually IP addresses). A host name is now passed as
882 "[x.x.x.x]".
883
88414. Changed the calls that set up the SIGCHLD handler in the daemon to use the
885 code that specifies a non-restarting handler (typically sigaction() in
886 modern systems) in an attempt to fix a rare and obscure crash bug.
887
88815. Narrowed the window for a race in the daemon that could cause it to ignore
889 SIGCHLD signals. This is not a major problem, because they are used only to
890 wake it up if nothing else does.
891
89216. A malformed maildirsize file could cause Exim to calculate negative values
893 for the mailbox size or file count. Odd effects could occur as a result.
894 The maildirsize information is now recalculated if the size or filecount
895 end up negative.
896
89717. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
898 support for a long time. Removed HAVE_SYS_VFS_H.
899
ea3a6f44 90018. Updated exipick to current release from John Jetmore.
bbe902f0
PH
901
90219. Allow an empty sender to be matched against a lookup in an address list.
903 Previously the only cases considered were a regular expression, or an
904 empty pattern.
905
90620. Exim went into a mad DNS lookup loop when doing a callout where the
907 host was specified on the transport, if the DNS lookup yielded more than
908 one IP address.
909
ea3a6f44
NM
91021. The RFC2047 encoding function was originally intended for short strings
911 such as real names; it was not keeping to the 75-character limit for
912 encoded words that the RFC imposes. It now respects the limit, and
913 generates multiple encoded words if necessary. To be on the safe side, I
914 have increased the buffer size for the ${rfc2047: expansion operator from
915 1024 to 2048 bytes.
bbe902f0 916
ea3a6f44
NM
91722. Failure to deliver a bounce message always caused it to be frozen, even if
918 there was an errors_to setting on the router. The errors_to setting is now
919 respected.
bbe902f0
PH
920
92123. If an IPv6 address is given for -bh or -bhc, it is now converted to the
922 canonical form (fully expanded) before being placed in
923 $sender_host_address.
924
92524. Updated eximstats to version 1.33
926
ea3a6f44
NM
92725. Include certificate and key file names in error message when GnuTLS fails
928 to set them up, because the GnuTLS error message doesn't include the name
929 of the failing file when there is a problem reading it.
bbe902f0
PH
930
93126. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 932 As per change 25.
bbe902f0 933
ea3a6f44
NM
93427. Reset the locale to "C" after calling embedded Perl, in case it was changed
935 (this can affect the format of dates).
bbe902f0 936
ea3a6f44
NM
93728. exim_tidydb, when checking for the continued existence of a message for
938 which it has found a message-specific retry record, was not finding
939 messages that were in split spool directories. Consequently, it was
940 deleting retry records that should have stayed in existence.
bbe902f0
PH
941
94229. eximstats updated to version 1.35
943 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
944 1.35 - bugfix such that pie charts by volume are generated correctly
945
ea3a6f44
NM
94630. The SPA authentication driver was not abandoning authentication and moving
947 on to the next authenticator when an expansion was forced to fail,
948 contradicting the general specification for all authenticators. Instead it
949 was generating a temporary error. It now behaves as specified.
bbe902f0 950
ea3a6f44
NM
95131. The default ordering of permitted cipher suites for GnuTLS was pessimal
952 (the order specifies the preference for clients). The order is now AES256,
953 AES128, 3DES, ARCFOUR128.
bbe902f0 954
ea3a6f44
NM
95531. Small patch to Sieve code - explicitly set From: when generating an
956 autoreply.
bbe902f0 957
ea3a6f44
NM
95832. Exim crashed if a remote delivery caused a very long error message to be
959 recorded - for instance if somebody sent an entire SpamAssassin report back
960 as a large number of 550 error lines. This bug was coincidentally fixed by
961 increasing the size of one of Exim's internal buffers (big_buffer) that
962 happened as part of the Exiscan merge. However, to be on the safe side, I
963 have made the code more robust (and fixed the comments that describe what
964 is going on).
bbe902f0 965
ea3a6f44
NM
96633. Some experimental protocols are using DNS PTR records for new purposes. The
967 keys for these records are domain names, not reversed IP addresses. The
968 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
969 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
970 CAN-2005-0021
971
ea3a6f44
NM
97234. The host_aton() function is supposed to be passed a string that is known
973 to be a valid IP address. However, in the case of IPv6 addresses, it was
974 not checking this. This is a hostage to fortune. Exim now panics and dies
975 if the condition is not met. A case was found where this could be provoked
976 from a dnsdb PTR lookup with an IPv6 address that had more than 8
977 components; fortuitously, this particular loophole had already been fixed
978 by change 4.50/55 or 4.44/33 above.
979
980 If there are any other similar loopholes, the new check in host_aton()
981 itself should stop them being exploited. The report I received stated that
982 data on the command line could provoke the exploit when Exim was running as
983 exim, but did not say which command line option was involved. All I could
984 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
985 running as the user.
bbe902f0
PH
986 CAN-2005-0021
987
ea3a6f44
NM
98835. There was a buffer overflow vulnerability in the SPA authentication code
989 (which came originally from the Samba project). I have added a test to the
990 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
991 CAN-2005-0022
992
ea3a6f44
NM
99336. The daemon start-up calls getloadavg() while still root for those OS that
994 need the first call to be done as root, but it missed one case: when
995 deliver_queue_load_max is set with deliver_drop_privilege. This is
996 necessary for the benefit of the queue runner, because there is no re-exec
997 when deliver_drop_privilege is set.
bbe902f0 998
ea3a6f44
NM
99937. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1000 was in use, was not putting the data itself into the right store pool;
1001 consequently, it could be overwritten for a subsequent message in the same
1002 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1003 the caching.)
bbe902f0 1004
ea3a6f44
NM
100538. Sometimes the final signoff response after QUIT could fail to get
1006 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1007 < 0 before doing a fflush(). This bug looks as though it goes back to the
1008 introduction of TLS in release 3.20, but "sometimes" must have been rare
1009 because the tests only now provoked it.
bbe902f0
PH
1010
1011
495ae4b0
PH
1012Exim version 4.43
1013-----------------
1014
1015 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1016 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1017 Now it returns an integer. A number of calls were still expecting a T/F
1018 return. Fortuitously, in all cases, the tests worked in OK situations,
1019 which is the norm. However, things would have gone wrong on any write
1020 failures on the smtp file descriptor. This function is used when sending
1021 messages over SMTP and also when doing verify callouts.
1022
1023 2. When Exim is called to do synchronous delivery of a locally submitted
1024 message (the -odf or -odi options), it no longer closes stderr before doing
1025 the delivery.
1026
1027 3. Implemented the mua_wrapper option.
1028
1029 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1030
1031 5. Implemented the functions header_remove(), header_testname(),
1032 header_add_at_position(), and receive_remove_recipient(), and exported them
1033 to local_scan().
1034
1035 6. If an ACL "warn" statement specified the addition of headers, Exim already
1036 inserted X-ACL-Warn: at the start if there was no header name. However, it
1037 was not making this test for the second and subsequent header lines if
1038 there were newlines in the string. This meant that an invalid header could
1039 be inserted if Exim was badly configured.
1040
1041 7. Allow an ACL "warn" statement to add header lines at the start or after all
1042 the Received: headers, as well as at the end.
1043
1044 8. Added the rcpt_4xx retry error code.
1045
1046 9. Added postmaster_mailfrom=xxx to callout verification option.
1047
104810. Added mailfrom=xxxx to the callout verification option, for verify=
1049 header_sender only.
1050
105111. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1052 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1053
105412. Inserted some casts to stop certain compilers warning when using pointer
1055 differences as field lengths or precisions in printf-type calls (mostly
1056 affecting debugging statements).
1057
105813. Added optional readline() support for -be (dynamically loaded).
1059
106014. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1061 same clock tick as a message's arrival, so that its received time was the
1062 same as the "first fail" time on the retry record, and that message
1063 remained on the queue past the ultimate address timeout, every queue runner
1064 would try a delivery (because it was past the ultimate address timeout) but
1065 after another failure, the ultimate address timeout, which should have then
1066 bounced the address, did not kick in. This was a "< instead of <=" error;
1067 in most cases the first failure would have been in the next clock tick
1068 after the received time, and all would be well.
1069
107015. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1071 being recognized when the domain list was tested by the match_domain
1072 condition in an expansion string.
1073
107416. Added the ${str2b64: operator.
1075
107617. Exim was always calling setrlimit() to set a large limit for the number of
1077 processes, without checking whether the existing limit was already
1078 adequate. (It did check for the limit on file descriptors.) Furthermore,
1079 errors from getrlimit() and setrlimit() were being ignored. Now they are
1080 logged to the main and panic logs, but Exim does carry on, to try to do its
1081 job under whatever limits there are.
1082
108318. Imported PCRE 5.0.
1084
108519. Trivial typo in log message " temporarily refused connection" (the leading
1086 space).
1087
108820. If the log selector return_path_on_delivery was set and an address was
1089 redirected to /dev/null, the delivery process crashed because it assumed
1090 that a return path would always be set for a "successful" delivery. In this
1091 case, the whole delivery is bypassed as an optimization, and therefore no
1092 return path is set.
1093
109421. Internal re-arrangement: the function for sending a challenge and reading
1095 a response while authentication was assuming a zero-terminated challenge
1096 string. It's now changed to take a pointer and a length, to allow for
1097 binary data in such strings.
1098
109922. Added the cyrus_sasl authenticator (code supplied by MBM).
1100
110123. Exim was not respecting finduser_retries when seeking the login of the
1102 uid under which it was called; it was always trying 10 times. (The default
1103 setting of finduser_retries is zero.) Also, it was sleeping after the final
1104 failure, which is pointless.
1105
110624. Implemented tls_on_connect_ports.
1107
110825. Implemented acl_smtp_predata.
1109
111026. If the domain in control=submission is set empty, Exim assumes that the
1111 authenticated id is a complete email address when it generates From: or
1112 Sender: header lines.
1113
111427. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1115 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1116 chown and chgrp in /bin and hostname in /usr/bin.
1117
111828. Exim was keeping the "process log" file open after each use, just as it
1119 does for the main log. This opens the possibility of it remaining open for
1120 long periods when the USR1 signal hits a daemon. Occasional processlog
1121 errors were reported, that could have been caused by this. Anyway, it seems
1122 much more sensible not to leave this file open at all, so that is what now
1123 happens.
1124
112529. The long-running daemon process does not normally write to the log once it
1126 has entered its main loop, and it closes the log before doing so. This is
1127 so that log files can straightforwardly be renamed and moved. However,
1128 there are a couple of unusual error situations where the daemon does write
1129 log entries, and I had neglected to close the log afterwards.
1130
113130. The text of an SMTP error response that was received during a remote
1132 delivery was being truncated at 512 bytes. This is too short for some of
1133 the long messages that one sometimes sees. I've increased the limit to
1134 1024.
1135
113631. It is now possible to make retry rules that apply only when a message has a
1137 specific sender, in particular, an empty sender.
1138
113932. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1140 it possible to be selective about when SMTP synchronization is enforced.
1141
114233. Added "control = caseful_local_part" and "control = "caselower_local_part".
1143
114432. Implemented hosts_connection_nolog.
1145
114633. Added an ACL for QUIT.
1147
114834. Setting "delay_warning=" to disable warnings was not working; it gave a
1149 syntax error.
1150
115135. Added mailbox_size and mailbox_filecount to appendfile.
1152
115336. Added control = no_multiline_responses to ACLs.
1154
115537. There was a bug in the logic of the code that waits for the clock to tick
1156 in the case where the clock went backwards by a substantial amount such
1157 that the microsecond fraction of "now" was more than the microsecond
1158 fraction of "then" (but the whole seconds number was less).
1159
116038. Added support for the libradius Radius client library this is found on
1161 FreeBSD (previously only the radiusclient library was supported).
1162
1163
1164Exim version 4.42
1165-----------------
1166
1167 1. When certain lookups returned multiple values in the form name=value, the
1168 quoting of the values was not always being done properly. Specifically:
1169 (a) If the value started with a double quote, but contained no whitespace,
1170 it was not quoted.
1171 (b) If the value contained whitespace other than a space character (i.e.
1172 tabs or newlines or carriage returns) it was not quoted.
1173 This fix has been applied to the mysql and pgsql lookups by writing a
1174 separate quoting function and calling it from the lookup code. The fix
1175 should probably also be applied to nisplus, ibase and oracle lookups, but
1176 since I cannot test any of those, I have not disturbed their existing code.
1177
1178 2. A hit in the callout cache for a specific address caused a log line with no
1179 reason for rejecting RCPT. Now it says "Previous (cached) callout
1180 verification failure".
1181
1182 3. There was an off-by-one bug in the queryprogram router. An over-long
1183 return line was truncated at 256 instead of 255 characters, thereby
1184 overflowing its buffer with the terminating zero. As well as fixing this, I
1185 have increased the buffer size to 1024 (and made a note to document this).
1186
1187 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1188 when Exim is waiting for an SMTP response from a remote server, Exim
1189 restarts its select() call on the socket, thereby resetting its timeout.
1190 This is not a problem when such interrupts are rare. Somebody set up a cron
1191 job to run exiwhat every 2 minutes, which is less than the normal select()
1192 timeout (5 or 10 minutes). This meant that the select() timeout never
1193 kicked in because it was always reset. I have fixed this by comparing the
1194 time when an interrupt arrives with the time at the start of the first call
1195 to select(). If more time than the timeout has elapsed, the interrupt is
1196 treated as a timeout.
1197
1198 5. Some internal re-factoring in preparation for the addition of Sieve
1199 extensions (by MH). In particular, the "personal" test is moved to a
1200 separate function, and given an option for scanning Cc: and Bcc: (which is
1201 not set for Exim filters).
1202
1203 6. When Exim created an email address using the login of the caller as the
1204 local part (e.g. when creating a From: or Sender: header line), it was not
1205 quoting the local part when it contained special characters such as @.
1206
1207 7. Installed new OpenBSD configuration files.
1208
1209 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1210 try to make them clearer.
1211
1212 9. Callout options, other than the timeout value, were being ignored when
1213 verifying sender addresses in header lines. For example, when using
1214
1215 verify = header_sender/callout=no_cache
1216
1217 the cache was (incorrectly) being used.
1218
121910. Added a missing instance of ${EXE} to the exim_install script; this affects
1220 only the Cygwin environment.
1221
122211. When return_path_on_delivery was set as a log selector, if different remote
1223 addresses in the same message used different return paths and parallel
1224 remote delivery occurred, the wrong values would sometimes be logged.
1225 (Whenever a remote delivery process finished, the return path value from
1226 the most recently started remote delivery process was logged.)
1227
122812. RFC 3848 specifies standard names for the "with" phrase in Received: header
1229 lines when AUTH and/or TLS are in use. This is the "received protocol"
1230 field. Exim used to use "asmtp" for authenticated SMTP, without any
1231 indication (in the protocol name) for TLS use. Now it follows the RFC and
1232 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1233 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1234 names appear in log lines as well as in Received: header lines.
1235
123613. Installed MH's patches for Sieve to add the "copy" and "vacation"
1237 extensions, and comparison tests, and to fix some bugs.
1238
123914. Changes to the "personal" filter test:
1240
1241 (1) The test was buggy in that it was just doing the equivalent of
1242 "contains" tests on header lines. For example, if a user's address was
1243 anne@some.where, the "personal" test would incorrectly be true for
1244
1245 To: susanne@some.where
1246
1247 This test is now done by extracting each address from the header in turn,
1248 and checking the entire address. Other tests that are part of "personal"
1249 are now done using regular expressions (for example, to check local parts
1250 of addresses in From: header lines).
1251
1252 (2) The list of non-personal local parts in From: addresses has been
1253 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1254 taken from the Sieve specification recommendations.
1255
1256 (3) If the message contains any header line starting with "List-" it is
1257 treated as non-personal.
1258
1259 (4) The test for "circular" in the Subject: header line has been removed
1260 because it now seems ill-conceived.
1261
126215. Minor typos in src/EDITME comments corrected.
1263
126416. Installed latest exipick from John Jetmore.
1265
126617. If headers_add on a router specified a text string that was too long for
1267 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1268 of string_sprintf() is now avoided.
1269
127018. $message_body_size was not set (it was always zero) when running the DATA
1271 ACL and the local_scan() function.
1272
127319. For the "mail" command in an Exim filter, no default was being set for
1274 the once_repeat time, causing a random time value to be used if "once" was
1275 specified. (If the value happened to be <= 0, no repeat happened.) The
1276 default is now 0s, meaning "never repeat". The "vacation" command was OK
1277 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1278 (I found it when inspecting the code).
1279
128020. There is now an overall timeout for performing a callout verification. It
1281 defaults to 4 times the callout timeout, which applies to individual SMTP
1282 commands during the callout. The overall timeout applies when there is more
1283 than one host that can be tried. The timeout is checked before trying the
1284 next host. This prevents very long delays if there are a large number of
1285 hosts and all are timing out (e.g. when the network connections are timing
1286 out). The value of the overall timeout can be changed by specifying an
1287 additional sub-option for "callout", called "maxwait". For example:
1288
1289 verify = sender/callout=5s,maxwait=20s
1290
129121. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1292 to the end before writing, but this should make it even safer).
1293
129422. Exim was forgetting that it had advertised PIPELINING for the second and
1295 subsequent messages on an SMTP connection. It was also not resetting its
1296 memory on STARTTLS and an internal HELO.
1297
129823. When Exim logs an SMTP synchronization error within a session, it now
1299 records whether PIPELINING has been advertised or not.
1300
130124. Added 3 instances of "(long int)" casts to time_t variables that were being
1302 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1303 rather than long int.
1304
130525. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1306
130726. Added the never_mail option to autoreply.
1308
1309
1310Exim version 4.41
1311-----------------
1312
1313 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1314 crash if the getsockname() call failed; this can happen if a connection is
1315 closed very soon after it is established. The problem was simply in the
1316 order in which certain operations were done, causing Exim to try to write
1317 to the SMTP stream before it had set up the file descriptor. The bug has
1318 been fixed by making things happen in the correct order.
1319
1320
1321Exim version 4.40
1322-----------------
1323
1324 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1325 before the connection was closed, thus losing the rejection response.
1326
1327 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1328 some early Solaris releases, but causes trouble in current releases where
1329 socklen_t is defined.
1330
1331 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1332 always exist.
1333
1334 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1335 configured.
1336
1337 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1338 that releasing the top of it at the end releases what was used for sub-
1339 expansions (unless the block got too big). However, discard this block if
1340 the first thing is a variable or header, so that we can use its block when
1341 it is dynamic (useful for very large $message_headers, for example).
1342
1343 6. Lookups now cache *every* query, not just the most recent. A new, separate
1344 store pool is used for this. It can be recovered when all lookup caches are
1345 flushed. Lookups now release memory at the end of their result strings.
1346 This has involved some general refactoring of the lookup sources.
1347
1348 7. Some code has been added to the store_xxx() functions to reduce the amount
1349 of flapping under certain conditions.
1350
1351 8. log_incoming_interface used to affect only the <= reception log lines. Now
1352 it causes the local interface and port to be added to several more SMTP log
1353 lines, for example "SMTP connection from", and rejection lines.
1354
1355 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1356
135710. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1358
135911. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1360 could be overwritten at the end of the current message (or the start of a
1361 new message if it was set in a HELO ACL). The value is now preserved for
1362 the duration of the SMTP connection.
1363
136412. If a transport had a headers_rewrite setting, and a matching header line
1365 contained an unqualified address, that address was qualified, even if it
1366 did not match any rewriting rules. The underlying bug was that the values
1367 of the flags that permit the existence of unqualified sender and recipient
1368 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1369 non-local messages, and by -bnq for local messages) were not being
1370 preserved with the message after it was received.
1371
137213. When Exim was logging an SMTP synchronization error, it could sometimes log
1373 "next input=" as part of the text comprising the host identity instead of
1374 the correct text. The code was using the same buffer for two different
1375 strings. However, depending on which order the printing function evaluated
1376 its arguments, the bug did not always show up. Under Linux, for example, my
1377 test suite worked just fine.
1378
137914. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1380 doesn't work with some older versions of Perl. It has been changed to "my",
1381 which in any case is probably the better facility to use.
1382
138315. A really picky compiler found some instances of statements for creating
1384 error messages that either had too many or two few arguments for the format
1385 string.
1386
138716. The size of the buffer for calls to the DNS resolver has been increased
1388 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1389 for addresses that have a lot of PTR records. This alleviates a problem; it
1390 does not fully solve it.
1391
139217. A dnsdb lookup for PTR records that receives more data than will fit in the
1393 buffer now truncates the list and logs the incident, which is the same
1394 action as happens when Exim is looking up a host name and its aliases.
1395 Previously in this situation something unpredictable would happen;
1396 sometimes it was "internal error: store_reset failed".
1397
139818. If a server dropped the connection unexpectedly when an Exim client was
1399 using GnuTLS and trying to read a response, the client delivery process
1400 crashed while trying to generate an error log message.
1401
140219. If a "warn" verb in an ACL added multiple headers to a message in a single
1403 string, for example:
1404
1405 warn message = H1: something\nH2: something
1406
1407 the text was added as a single header line from Exim's point of view
1408 though it ended up OK in the delivered message. However, searching for the
1409 second and subsequent header lines using $h_h2: did not work. This has been
1410 fixed. Similarly, if a system filter added multiple headers in this way,
1411 the routers could not see them.
1412
141320. Expanded the error message when iplsearch is called with an invalid key to
1414 suggest using net-iplsearch in a host list.
1415
141621. When running tests using -bh, any delays imposed by "delay" modifiers in
1417 ACLs are no longer actually imposed (and a message to that effect is
1418 output).
1419
142022. If a "gecos" field in a passwd entry contained escaped characters, in
1421 particular, if it contained a \" sequence, Exim got it wrong when building
1422 a From: or a Sender: header from that name. A second bug also caused
1423 incorrect handling when an unquoted " was present following a character
1424 that needed quoting.
1425
142623. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1427 was not being matched caselessly.
1428
142924. Arranged for all hyphens in the exim.8 source to be escaped with
1430 backslashes.
1431
143225. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1433 properly. Recipient callout cache records were still being keyed to include
1434 the sender, even when use_sender was set false. This led to far more
1435 callouts that were necessary. The sender is no longer included in the key
1436 when use_sender is false.
1437
143826. Added "control = submission" modifier to ACLs.
1439
144027. Added the ${base62d: operator to decode base 62 numbers.
1441
144228. dnsdb lookups can now access SRV records.
1443
144429. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1445 the configuration file.
1446
144730. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1448 (-v) mode. This makes the output for a verbose queue run more intelligible.
1449
145031. Added a use_postmaster feature to recipient callouts.
1451
145232. Added the $body_zerocount variable, containing the number of binary zero
1453 bytes in the message body.
1454
145533. The time of last modification of the "new" subdirectory is now used as the
1456 "mailbox time last read" when there is a quota error for a maildir
1457 delivery.
1458
145934. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1460
146135. Added +ignore_unknown as a special item in host lists.
1462
146336. Code for decoding IPv6 addresses in host lists is now included, even if
1464 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1465 address was recognized as an IP address, but was then not correctly decoded
1466 into binary, causing unexpected and incorrect effects when compared with
1467 another IP address.
1468
1469
1470Exim version 4.34
1471-----------------
1472
1473 1. Very minor rewording of debugging text in manualroute to say "list of
1474 hosts" instead of "hostlist".
1475
1476 2. If verify=header_syntax was set, and a header line with an unqualified
1477 address (no domain) and a large number of spaces between the end of the
1478 name and the colon was received, the reception process suffered a buffer
1479 overflow, and (when I tested it) crashed. This was caused by some obsolete
1480 code that should have been removed. The fix is to remove it!
1481
1482 3. When running in the test harness, delay a bit after writing a bounce
1483 message to get a bit more predictability in the log output.
1484
1485 4. Added a call to search_tidyup() just before forking a reception process. In
1486 theory, someone could use a lookup in the expansion of smtp_accept_max_
1487 per_host which, without the tidyup, could leave open a database connection.
1488
1489 5. Added the variables $recipient_data and $sender_data which get set from a
1490 lookup success in an ACL "recipients" or "senders" condition, or a router
1491 "senders" option, similar to $domain_data and $local_part_data.
1492
1493 6. Moved the writing of debug_print from before to after the "senders" test
1494 for routers.
1495
1496 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1497 problems for message scanning, either using a data ACL, or using
1498 local_scan() because the Received: header was not generated till after they
1499 were called (in order to set the time as the time of reception completion).
1500 I have revised the way this works. The header is now generated after the
1501 body is received, but before the ACL or local_scan() are called. After they
1502 are run, the timestamp in the header is updated.
1503
1504
1505Exim version 4.33
1506-----------------
1507
1508 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1509 before starting a queue runner without re-exec. This happened only when
1510 deliver_drop_privilege was set or when the Exim user was set to root. The
1511 effect of the bug was that timeouts during subsequent deliveries caused
1512 crashes instead of being properly handled. The handler is now left at its
1513 default (and expected) setting.
1514
1515 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1516 message, again when deliver_drop_privilege is set or Exim is run as root.
1517 The bug described in (1) was not present in this case, but the tidying up
1518 of the other signals was missing. I have made the two cases consistent.
1519
1520 3. The ignore_target_hosts setting on a manualroute router was being ignored
1521 for hosts that were looked up using the /MX notation.
1522
1523 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1524 in domain lists.
1525
1526 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1527 operated on the sender address. After changing the $sender_address to <>
1528 for the sender address verify, Exim was re-instated it as the original
1529 (before rewriting) address, but remembering that it had rewritten it, so it
1530 wasn't rewriting it again. This bug also had the effect of breaking the
1531 sender address verification caching when the sender address was rewritten.
1532
1533 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1534 This has been changed so that if the ip literal address matches
1535 ignore_target_hosts, the router declines.
1536
1537 7. Added expansion conditions match_domain, match_address, and match_local_
1538 part (NOT match_host).
1539
1540 8. The placeholder for the Received: header didn't have a length field set.
1541
1542 9. Added code to Exim itself and to exim_lock to test for a specific race
1543 condition that could lead to file corruption when using MBX delivery. The
1544 issue is with the lockfile that is created in /tmp. If this file is removed
1545 after a process has opened it but before that process has acquired a lock,
1546 there is the potential for a second process to recreate the file and also
1547 acquire a lock. This could lead to two Exim processes writing to the file
1548 at the same time. The added code performs the same test as UW imapd; it
1549 checks after acquiring the lock that its file descriptor still refers to
1550 the same named file.
1551
155210. The buffer for building added header lines was of fixed size, 8192 bytes.
1553 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1554 when Exim is built.
1555
155611. Added the smtp_active_hostname option. If used, this will typically be made
1557 to depend on the incoming interface address. Because $interface_address is
1558 not set up until the daemon has forked a reception process, error responses
1559 that can happen earlier (such as "too many connections") no longer contain
1560 a host name.
1561
156212. If an expansion in a condition on a "warn" statement fails because a lookup
1563 defers, the "warn" statement is abandoned, and the next ACL statement is
1564 processed. Previously this caused the whole ACL to be aborted.
1565
156613. Added the iplsearch lookup type.
1567
156814. Added ident_timeout as a log selector.
1569
157015. Added tls_certificate_verified as a log selector.
1571
157216. Added a global option tls_require_ciphers (compare the smtp transport
1573 option of the same name). This controls incoming TLS connections.
1574
157517. I finally figured out how to make tls_require_ciphers do a similar thing
1576 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1577 before starting the TLS session.
1578
157918. Tabs are now shown as \t in -bP output.
1580
158119. If the log selector return_path_on_delivery was set, Exim crashed when
1582 bouncing a message because it had too many Received: header lines.
1583
158420. If two routers both had headers_remove settings, and the first one included
1585 a superfluous trailing colon, the final name in the first list and the
1586 first name in the second list were incorrectly joined into one item (with a
1587 colon in the middle).
1588
1589
1590Exim version 4.32
1591-----------------
1592
1593 1. Added -C and -D options to the exinext utility, mainly to make it easier
1594 to include in the automated testing, but these could be helpful when
1595 multiple configurations are in use.
1596
1597 2. The exinext utility was not formatting the output nicely when there was
1598 an alternate port involved in the retry record key, nor when there was a
1599 message id as well (for retries that were specific to a specific message
1600 and a specific host). It was also confused by IPv6 addresses, because of
1601 the additional colons they contain. I have fixed the IPv4 problem, and
1602 patched it up to do a reasonable job for IPv6.
1603
1604 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1605 delivery, the log line now contains "pipelined" if PIPELINING was used.
1606
1607 4. An SMTP transport process used to panic and die if the bind() call to set
1608 an explicit outgoing interface failed. This has been changed; it is now
1609 treated in the same way as a connect() failure.
1610
1611 5. A reference to $sender_host_name in the part of a conditional expansion
1612 that was being skipped was still causing a DNS lookup. This no longer
1613 occurs.
1614
1615 6. The def: expansion condition was not recognizing references to header lines
1616 that used bh_ and bheader_.
1617
1618 7. Added the _cache feature to named lists.
1619
1620 8. The code for checking quota_filecount in the appendfile transport was
1621 allowing one more file than it should have been.
1622
1623 9. For compatibility with Sendmail, the command line option
1624
1625 -prval:sval
1626
1627 is equivalent to
1628
1629 -oMr rval -oMs sval
1630
1631 and sets the incoming protocol and host name (for trusted callers). The
1632 host name and its colon can be omitted when only the protocol is to be set.
1633 Note the Exim already has two private options, -pd and -ps, that refer to
1634 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1635 "s", but I don't think that's a major issue.
1636
163710. A number of refactoring changes to the code, none of which should affect
1638 Exim's behaviour:
1639
1640 (a) The number of logging options was getting close to filling up the
1641 32-bit word that was used as a bit map. I have split them into two classes:
1642 those that are passed in the argument to log_write(), and those that are
1643 only ever tested independently outside of that function. These are now in
1644 separate 32-bit words, so there is plenty of room for expansion again.
1645 There is no change in the user interface or the logging behaviour.
1646
1647 (b) When building, for example, log lines, the code previously used a
1648 macro that called string_cat() twice, in order to add two strings. This is
1649 not really sufficiently general. Furthermore, there was one instance where
1650 it was actually wrong because one of the argument was used twice, and in
1651 one call a function was used. (As it happened, calling the function twice
1652 did not affect the overall behaviour.) The macro has been replaced by a
1653 function that can join an arbitrary number of extra strings onto a growing
1654 string.
1655
1656 (c) The code for expansion conditions now uses a table and a binary chop
1657 instead of a serial search (which was left over from when there were very
1658 few conditions). Also, it now recognizes conditions like "pam" even when
1659 the relevant support is not compiled in: a suitably worded error message is
1660 given if an attempt is made to use such a condition.
1661
166211. Added ${time_interval:xxxxx}.
1663
166412. A bug was causing one of the ddress fields not to be passed back correctly
1665 from remote delivery subprocesses. The field in question was not being
1666 subsequently used, so this caused to problems in practice.
1667
166813. Added new log selectors queue_time and deliver_time.
1669
167014. Might have fixed a bug in maildirsizefile handling that threw up
1671 "unexpected character" debug warnings, and recalculated the data
1672 unnecessarily. In any case, I expanded the warning message to give more
1673 information.
1674
167515. Added the message "Restricted characters in address" to the statements in
1676 the default ACL that block characters like @ and % in local parts.
1677
167816. Change 71 for release 4.31 proved to be much less benign that I imagined.
1679 Three changes have been made:
1680
1681 (a) There was a serious bug; a negative response to MAIL caused the whole
1682 recipient domain to be cached as invalid, thereby blocking all messages
1683 to all local parts at the same domain, from all senders. This bug has
1684 been fixed. The domain is no longer cached after a negative response to
1685 MAIL if the sender used is not empty.
1686
1687 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1688 been restored.
1689
1690 (c) A new callout option, "use_sender" has been added for people who want
1691 the modified behaviour.
1692
1693
1694Exim version 4.31
1695-----------------
1696
1697 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1698 Larry Rosenman.
1699
1700 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1701 indeed breaks things for older releases.
1702
1703 3. Added additional logging to the case where there is a problem reading data
1704 from a filter that is running in a subprocess using a pipe, in order to
1705 try to track down a specific problem.
1706
1707 4. Testing facility fudge: when running in the test harness and attempting
1708 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1709 getting "No route to host". Convert this to a timeout.
1710
1711 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1712 warning.
1713
1714 6. Some OS don't have socklen_t but use size_t instead. This affects the
1715 fifth argument of getsockopt() amongst other things. This is now
1716 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1717 can be set for individual OS. I have set it for SunOS5, OSF1, and
1718 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1719 some earlier ones do not.
1720
1721 7. Change 4.30/15 was not doing the test caselessly.
1722
1723 8. The standard form for an IPv6 address literal was being rejected by address
1724 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1725 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1726 this, as well as the form without the "IPv6" on the front (but only when
1727 address literals are enabled, of course).
1728
1729 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1730
173110. Exim crashed if a message with an empty sender address specified by -f
1732 encountered a router with an errors_to setting. This could be provoked only
1733 by a command such as
1734
1735 exim -f "" ...
1736
1737 where an empty string was supplied; "<>" did not hit this bug.
1738
173911. Installed PCRE release 4.5.
1740
174112. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1742 remained set. It is now erased.
1743
174413. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1745 times from message ids (which are base 36 rather than the normal 62).
1746
174714. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1748 were being counted as actual protocol errors, and logged if the log
1749 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1750 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1751 and DATA following a set of rejected RCPTs do not count as protocol errors.
1752 In other words, Exim assumes they were pipelined, though this may not
1753 actually be the case. Of course, in all cases the client gets an
1754 appropriate error code.
1755
175615. If a lookup fails in an ACL condition, a message about the failure may
1757 be available; it is used if testing the ACL cannot continue, because most
1758 such messages specify what the cause of the deferral is. However, some
1759 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1760 that caused an old message to be retained and used if a later statement
1761 caused a defer, replacing the real cause of the deferral.
1762
176316. If an IP address had so many PTR records that the DNS lookup buffer
1764 was not large enough to hold them, Exim could crash while trying to process
1765 the truncated data. It now detects and logs this case.
1766
176717. Further to 4.21/58, another change has been made: if (and only if) the
1768 first line of a message (the first header line) ends with CRLF, a bare LF
1769 in a subsequent header line has a space inserted after it, so as not to
1770 terminate the header.
1771
177218. Refactoring: tidied an ugly bit of code in appendfile that copied data
1773 unnecessarily, used atoi() instead of strtol(), and didn't check the
1774 termination when getting file sizes from file names by regex.
1775
177619. Completely re-implemented the support for maildirsize files, in the light
1777 of a number of problems with the previous contributed implementation
1778 (4.30/29). In particular:
1779
1780 . If the quota is zero, the maildirsize file is maintained, but no quota is
1781 imposed.
1782
1783 . If the maildir directory does not exist, it is created before any attempt
1784 to write a maildirsize file.
1785
1786 . The quota value in the file is just a cache; if the quota is changed in
1787 the transport, the new value overrides.
1788
1789 . A regular expression is available for excluding directories from the
1790 count.
1791
179220. The autoreply transport checks the characters in options that define the
1793 message's headers; it allows continued headers, but it was checking with
1794 isspace() after an embedded newline instead of explicitly looking for a
1795 space or a tab.
1796
179721. If all the "regular" hosts to which an address was routed had passed their
1798 expiry times, and had not reached their retry times, the address was
1799 bounced, even if fallback hosts were defined. Now Exim should go on to try
1800 the fallback hosts.
1801
180222. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1803 equivalent code in the SMTP transport. Some hosts send humungous responses
1804 to HELO/EHLO, more than 1024 it seems.
1805
180623. Refactoring: code in filter.c used (void *) for "any old type" but this
1807 gives compiler warnings in some environments. I've now done it "properly",
1808 using a union.
1809
181024. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1811 (because of problems with the built-in one) was declared to return uschar *
1812 instead of char *, causing compiler failure.
1813
181425. Fixed a file descriptor leak when processing alias/forward files.
1815
181626. Fixed a minor format string issue in dbfn.c.
1817
181827. Typo in exim.c: ("dmbnz" for "dbmnz").
1819
182028. If a filter file refered to $h_xxx or $message_headers, and the headers
1821 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1822 become corrupted.
1823
182429. When a sender address is verified, it is cached, to save repeating the test
1825 when there is more than one recipient in a message. However, when the
1826 verification involves a callout, it is possible for different callout
1827 options to be set for different recipients. It is too complicated to keep
1828 track of this in the cache, so now Exim always runs a verification when a
1829 callout is required, relying on the callout cache for the optimization.
1830 The overhead is duplication of the address routing, but this should not be
1831 too great.
1832
183330. Fixed a bug in callout caching. If a RCPT command caused the sender address
1834 to be verified with callout=postmaster, and the main callout worked but the
1835 postmaster check failed, the verification correctly failed. However, if a
1836 subsequent RCPT command asked for sender verification *without* the
1837 postmaster check, incorrect caching caused this verification also to fail,
1838 incorrectly.
1839
184031. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1841 it was not caching the DNS options (qualify_single, search_parents) that
1842 were used when the lookup failed. A subsequent lookup with different
1843 options therefore always gave the same answer, though there were cases
1844 where it should not have. (Example: a "domains = !$mx_any" option on a
1845 dnslookup router: the "domains" option is always processed without any
1846 widening, but the router might have qualify_single set.) Now Exim uses the
1847 cached value only when the same options are set.
1848
184932. Added John Jetmore's "exipick" utility to the distribution.
1850
185133. GnuTLS: When an attempt to start a TLS session fails for any reason other
1852 than a timeout (e.g. a certificate is required, and is not provided), an
1853 Exim server now closes the connection immediately. Previously it waited for
1854 the client to close - but if the client is SSL, it seems that they each
1855 wait for each other, leading to a delay before one of them times out.
1856
185734: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1858 maintained 0.8.x compatibility because I don't think many are using it, and
1859 it is clearly obsolete.
1860
186135. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1862 transport.
1863
186436. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1865 client certificate was expired. A simple patch fixes this, though I don't
1866 understand the full logic of why the verify callback is called multiple
1867 times.
1868
186937. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1870 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1871 which causes problems with some clients (such as the Certicom SSL Plus
1872 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1873 disables the coutermeasure allowing Eudora to connect."
1874
187538. Exim was not checking that a write() to a log file succeeded. This could
1876 lead to Bad Things if a log got too big, in particular if it hit a file
1877 size limit. Exim now panics and dies if it cannot write to a log file, just
1878 as it does if it cannot open a log file.
1879
188039. Modified OS/Makefile-Linux so that it now contains
1881
1882 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1883
1884 The two -D definitions ensure that Exim is compiled with large file
1885 support, which makes it possible to handle log files that are bigger than
1886 2^31.
1887
188840. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1889 instance) a domain was checked against a named list that involved a lookup,
1890 causing $domain_data to be set, then another domain was checked against the
1891 same list, then the first domain was re-checked, the value of $domain_data
1892 after the final check could be wrong. In particular, if the second check
1893 failed, it could be set empty. This bug probably also applied to
1894 $localpart_data.
1895
189641. The strip_trailing_dot option was not being applied to the address given
1897 with the -f command-line option.
1898
189942. The code for reading a message's header from the spool was incrementing
1900 $received_count, but never initializing it. This meant that the value was
1901 incorrect (doubled) while delivering a message in the same process in which
1902 it was received. In the most common configuration of Exim, this never
1903 happens - a fresh exec is done - but it can happen when
1904 deliver_drop_privilege is set.
1905
190643. When Exim logs an SMTP synchronization error - client data sent too soon -
1907 it now includes up to 150 characters of the unexpected data in the log
1908 line.
1909
191044. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1911 and building data strings. The size of both of these buffers was 10 000
1912 bytes - far larger than anybody would *ever* want, thought I. Needless to
1913 say, somebody hit the limit. I have increased the maximum line length to
1914 20 000 and the maximum data length of concatenated lines to 100 000. I have
1915 also fixed two bugs, because there was no checking on these buffers. Tsk,
1916 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1917 buffer is too small.
1918
191945. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1920 lsearch lookups. Now it does.
1921
192246. When parsing a route_list item in a manualroute router, a fixed-length
1923 buffer was used for the list of hosts. I made this 1024 bytes long,
1924 thinking that nobody would ever have a list of hosts that long. Wrong.
1925 Somebody had a whole pile of complicated expansion conditions, and the
1926 string was silently truncated, leading to an expansion error. It turns out
1927 that it is easier to change to an unlimited length (owing to other changes
1928 that have happened since this code was originally written) than to build
1929 structure for giving a limitation error. The length of the item that
1930 expands into the list of hosts is now unlimited.
1931
193247. The lsearch lookup could not handle data where the length of text line was
1933 more than 4095 characters. Such lines were truncated, leading to shortened
1934 data being returned. It should now handle lines of any length.
1935
193648. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1937 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1938 ACL").
1939
194049. Cosmetic tidy to scripts like exicyclog that are generated by globally
1941 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1942 no longer happens in comment lines. A list of replacements is now placed
1943 at the head of all of the source files, except those whose only change is
1944 to replace PERL_COMMAND in the very first #! line.
1945
194650. Replaced the slow insertion sort in queue.c, for sorting the list of
1947 messages on the queue, with a bottom-up merge sort, using code contributed
1948 by Michael Haardt. This should make operations like -bp somewhat faster on
1949 large queues. It won't affect queue runners, except when queue_run_in_order
1950 is set.
1951
195251. Installed eximstats 1.31 in the distribution.
1953
195452. Added support for SRV lookups to the dnslookup router.
1955
195653. If an ACL referred to $message_body or $message_body_end, the value was not
1957 reset for any messages that followed in the same SMTP session.
1958
195954. The store-handling optimization for building very long strings was not
1960 differentiating between the different store pools. I don't think this
1961 actually made any difference in practice, but I've tidied it.
1962
196355. While running the routers to verify a sender address, $sender_address
1964 was still set to the sender address. This is wrong, because when routing to
1965 send a bounce to the sender, it would be empty. Therefore, I have changed
1966 it so that, while verifying a sender address, $sender_address is set to <>.
1967 (There is no change to what happens when verifying a recipient address.)
1968
196956. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1970 target A or AAAA records (if not already returned) without resetting the
1971 qualify_single or search_parents options of the DNS resolver. These are
1972 inappropriate in this case because the targets of MX and SRV records must
1973 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1974 target that, when qualified, matched something in the local domain. These
1975 two options are now turned off when doing these lookups.
1976
197757. It seems that at least some releases of Reiserfs (which does not have the
1978 concept of a fixed number of inodes) returns zero and not -1 for the
1979 number of available inodes. This interacted badly with check_spool_inodes,
1980 which assumed that -1 was the "no such thing" setting. What I have done is
1981 to check that the total number of inodes is greater than zero before doing
1982 the test of how many are available.
1983
198458. When a "warn" ACL statement has a log_message modifier, the message is
1985 remembered, and not repeated. This is to avoid a lot of repetition when a
1986 message has many recipients that cause the same warning to be written.
1987 Howewer, Exim was preserving the list of already written lines for an
1988 entire SMTP session, which doesn't seem right. The memory is now reset if a
1989 new message is started.
1990
199159. The "rewrite" debugging flag was not showing the result of rewriting in the
1992 debugging output unless log_rewrite was also set.
1993
199460. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1995 of (int)(handle) when we know that handle contains (void *)(-1).
1996
199761. The Exim daemon panic-logs an error return when it closes the incoming
1998 connection. However "connection reset by peer" seems to be common, and
1999 isn't really an error worthy of noting specially, so that particular error
2000 is no long logged.
2001
200262. When Exim is trying to find all the local interfaces, it used to panic and
2003 die if the ioctl to get the interface flags failed. However, it seems that
2004 on at least one OS (Solaris 9) it is possible to have an interface that is
2005 included in the list of interfaces, but for which you get a failure error
2006 for this call. This happens when the interface is not "plumbed" into a
2007 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2008 failure of the "get flags" call assumes that the interface is down.
2009
201063. Added a ${eval10: operator, which assumes all numbers are decimal. This
2011 makes life easier for people who are doing arithmetic on fields extracted
2012 from dates, where you often get leading zeros that should not be
2013 interpreted as octal.
2014
201564. Added qualify_domain to the redirect router, to override the global
2016 setting.
2017
201865. If a pathologically long header line contained very many addresses (the
2019 report of this problem mentioned 10 000) and each of them was rewritten,
2020 Exim could use up a very large amount of memory. (It kept on making new
2021 copies of the header line as it rewrote, and never released the old ones.)
2022 At the expense of a bit more processing, the header rewriting function has
2023 been changed so that it no longer eats memory in this way.
2024
202566. The generation of the Received: header has been moved from the time that a
2026 message starts to be received, to the time that it finishes. The timestamp
2027 in the Received: header should now be very close to that of the <= log
2028 line. There are two side-effects of this change:
2029
2030 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2031 logged header lines no longer include the local Received: line, because
2032 it has not yet been created. The same applies to a copy of the message
2033 that is returned to a non-SMTP sender when a message is rejected.
2034
2035 (b) When a filter file is tested using -bf, no additional Received: header
2036 is added to the test message. After some thought, I decided that this
2037 is a bug fix.
2038
2039 This change does not affect the value of $received_for. It is still set
2040 after address rewriting, but before local_scan() is called.
2041
204267. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2043
204468. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2045 gave an unhelpful panic error message, and a defer error. I have managed to
2046 change this behaviour so that it now rejects any supplied certificate,
2047 which seems right, as the list of acceptable certificates is empty.
2048
204969. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2050 gave an unhelpful defer error. I have not managed to make this reject any
2051 supplied certificates, but the error message it gives is "no certificate
2052 supplied", which is not helpful.
2053
205470. exigrep's output now also includes lines that are not associated with any
2055 message, but which match the given pattern. Implemented by a patch from
2056 Martin Sluka, which also tidied up the Perl a bit.
2057
205871. Recipient callout verification, like sender verification, was using <> in
2059 the MAIL FROM command. This isn't really the right thing, since the actual
2060 sender may affect whether the remote host accepts the recipient or not. I
2061 have changed it to use the actual sender in the callout; this means that
2062 the cache record is now keyed on a recipient/sender pair, not just the
2063 recipient address. There doesn't seem to be a real danger of callout loops,
2064 since a callout by the remote host to check the sender would use <>.
2065 [SEE ABOVE: changed after hitting problems.]
2066
206772. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2068 temporary errors. However, in the case of such a code being given after
2069 the end of a data transmission (i.e. after ".") Exim was failing to write
2070 a retry record for the message. (Yes, there was some broken host that was
2071 actually sending 8xx at this point.)
2072
207373. An unknown lookup type in a host list could cause Exim to panic-die when
2074 the list was checked. (An example that provoked this was putting <; in the
2075 middle of a list instead of at the start.) If this happened during a DATA
2076 ACL check, a -D file could be left lying around. This kind of configuration
2077 error no longer causes Exim to die; instead it causes a defer errror. The
2078 incident is still logged to the main and panic logs.
2079
208074. Buglet left over from Exim 3 conversion. The message "too many messages
2081 in one connection" was written to the rejectlog but not the mainlog, except
2082 when address rewriting (yes!) was being logged.
2083
208475. Added write_rejectlog option.
2085
208676. When a system filter was run not as root (that is, when system_filter_user
2087 was set), the values of the $n variables were not being returned to the
2088 main process; thus, they were not subsequently available in the $sn
2089 variables.
2090
209177. Added +return_path_on_delivery log selector.
2092
209378. A connection timeout was being treated differently from recipients deferred
2094 when testing hosts_max_try with a message that was older than the host's
2095 retry timeout. (The host should not be counted, thus allowing all hosts to
2096 be tried at least once before bouncing.) This may have been the cause of an
2097 occasionally reported bug whereby a message would remain on the queue
2098 longer than the retry timeout, but would be bounced if a delivery was
2099 forced. I say "may" because I never totally pinned down the problem;
2100 setting up timeout/retry tests is difficult. See also the next item.
2101
210279. The ultimate address timeout was not being applied to errors that involved
2103 a combination of host plus message (for example, a timeout on a MAIL
2104 command). When an address resolved to a number of possible hosts, and they
2105 were not all tried for each delivery (e.g. because of hosts_max_try), a
2106 message could remain on the queue longer than the retry timeout.
2107
210880. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2109 Haardt.
2110
211181. Fixed an obscure SMTP outgoing bug which required at least the following
2112 conditions: (a) there was another message waiting for the same server;
2113 (b) the server returned 5xx to all RCPT commands in the first message so
2114 that the message was not completed; (c) the server dropped the connection
2115 or gave a negative response to the RSET that Exim sends to abort the
2116 transaction. The observed case was a dropped connection after DATA that had
2117 been sent in pipelining mode. That is, the server had advertised PIPELINING
2118 but was not implementing it correctly. The effect of the bug was incorrect
2119 behaviour, such as trying another host, and this could lead to a crash.
2120
2121
2122Exim version 4.30
2123-----------------
2124
2125 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2126 and daemon.c were passed as pointers to ints; they should have been
2127 pointers to socklen_t variables (which are typically unsigned ints).
2128
2129 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2130 fixed.
2131
2132 3. Fixed a really odd bug that affected only the testing scheme; patching a
2133 certain fixed string in the binary changed the value of another string that
2134 happened to be identical to the end of the original first string.
2135
2136 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2137 name", it returns that address as the IP address. On some operating
2138 systems (e.g. Solaris), it also passes back the IP address string as the
2139 "host name". However, on others (e.g. Linux), it passes back an empty
2140 string. Exim wasn't checking for this, and was changing the host name to an
2141 empty string, assuming it had been canonicized.
2142
2143 5. Although rare, it is permitted to have more than one PTR record for a given
2144 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2145 all the names associated with an address, because they do in Solaris.
2146 However, it seems that they do not in Linux for data that comes from the
2147 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2148 I found this out when I moved to a new Linux workstation and tried to run
2149 the Exim test suite.
2150
2151 To get round this problem I have changed the code so that it now does its
2152 own call to the DNS to look up PTR records when searching for a host name.
2153 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2154 addresses that are only in /etc/hosts are still found.
2155
2156 This behaviour is, however, controlled by an option called host_lookup_
2157 order, which defaults to "bydns:byaddr". If people want to use the other
2158 order, or indeed, just use one or the other means of lookup, they can
2159 specify it in this variable.
2160
2161 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2162 some operating systems, this comes back from gethostbyaddr() as an empty
2163 string, and this is what Exim used to test for. However, it seems that in
2164 other systems, "." is yielded. Exim now tests for this case too.
2165
2166 7. The values of check_spool_space and check_log_space are now held internally
2167 as a number of kilobytes instead of an absolute number of bytes. If a
2168 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2169 kilobyte. This means that much larger values can be stored.
2170
2171 8. Exim monitor: an attempt to get the action menu when not actually pointing
2172 at a message produces an empty menu entitled "No message selected". This
2173 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2174 no entries in it ("Shell widget menu has zero width and/or height"). So I
2175 have added a single, blank menu entry in this case.
2176
2177 9. Added ${quote_local_part.
2178
217910. MIME decoding is now applied to the contents of Subject: header lines when
2180 they are logged.
2181
218211. Now that a reference to $sender_host_address automatically causes a reverse
2183 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2184 host lookup before query-style lookups in lists that might use this
2185 variable. This has therefore been abolished, and the "net-" prefix is no
2186 longer necessary for query-style lookups.
2187
218812. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2189 have been a typo for LFLAGS, so it has been changed.
2190
219113. The install script calls Exim with "-C /dev/null" in order to find the
2192 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2193 to be output. Howeve, since Exim outputs its version number before the
2194 error, it didn't break the script. It just looked ugly. I fixed this by
2195 always allowing "-C /dev/null" if the caller is root.
2196
219714. Ignore overlarge ACL variable number when reading spool file - insurance
2198 against a later release with more variables having written the file.
2199
220015. The standard form for an IPv6 address literal was being rejected by EHLO.
2201 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2202 this, as well as the form without the "IPv6" on the front.
2203
220416. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2205 OS/Makefile-Darwin file.
2206
220717. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2208 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2209
221018. After change 4.21/52, "%ld" was used to format the contents of the $inode
2211 variable. However, some OS use ints for inodes. I've added cast to long int
2212 to get rid of the compiler warning.
2213
221419. I had forgotten to lock out "/../" in configuration file names when
2215 ALT_CONFIG_PREFIX was set.
2216
221720. Routers used for verification do not need to specify transports. However,
2218 if such a router generated a host list, and callout was configured, Exim
2219 crashed, because it could not find a port number from the (non-existent)
2220 transport. It now assumes port 25 in this circumstance.
2221
222221. Added the -t option to exigrep.
2223
222422. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2225 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2226 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2227 them is set, LOOKUP_LSEARCH is forced.
2228
222923. "exim -bV" now outputs a list of lookups that are included in the binary.
2230
223124. Added sender and host information to the "rejected by local_scan()" log
2232 line; previously there was no indication of these.
2233
223425. Added .include_if_exists.
2235
223626. Change 3.952/11 added an explicit directory sync on top of a file sync for
2237 Linux. It turns out that not all file systems support this. Apparently some
2238 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2239 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2240 not supported on the file descriptor, is now ignored when Exim is trying to
2241 sync a directory. This applies only to Linux.
2242
224327. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2244
224528. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2246 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2247 this.
2248
224929. Added support for maildirsize files from supplied patch (modified a bit).
2250
225130. The use of :fail: followed by an empty string could lead Exim to respond to
2252 sender verification failures with (e.g.):
2253
2254 550 Verification failed for <xxx>
2255 550 Sender verify failed
2256
2257 where the first response line was missing the '-' that indicates it is not
2258 the final line of the response.
2259
226031. The loop for finding the name of the user that called Exim had a hardwired
2261 limit of 10; it now uses the value of finduser_retries, which is used for
2262 all other user lookups.
2263
226432. Added $received_count variable, available in data and not_smtp ACLs, and at
2265 delivery time.
2266
226733. Exim was neglecting to zero errno before one call of strtol() when
2268 expanding a string and expecting an integer value. On some systems this
2269 resulted in spurious "integer overflow" errors. Also, it was casting the
2270 result into an int without checking.
2271
227234. Testing for a connection timeout using "timeout_connect" in the retry rules
2273 did not work. The code looks as if it has *never* worked, though it appears
2274 to have been documented since at least releast 1.62. I have made it work.
2275
227635. The "timeout_DNS" error in retry rules, also documented since at least
2277 1.62, also never worked. As it isn't clear exactly what this means, and
2278 clearly it isn't a major issue, I have abolished the feature by treating it
2279 as "timeout", and writing a warning to the main and panic logs.
2280
228136. The display of retry rules for -brt wasn't always showing the error code
2282 correctly.
2283
228437. Added new error conditions to retry rules: timeout_A, timeout_MX,
2285 timeout_connect_A, timeout_connect_MX.
2286
228738. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2288 to the empty sender.
2289
229039. The daemon was not analysing the content of -oX till after it had closed
2291 stderr and disconnected from the controlling terminal. This meant that any
2292 syntax errors were only noted on the panic log, and the return code from
2293 the command was 0. By re-arranging the code a little, I've made the
2294 decoding happen first, so such errors now appear on stderr, and the return
2295 code is 1. However, the actual setting up of the sockets still happens in
2296 the disconnected process, so errors there are still only recorded on the
2297 panic log.
2298
229940. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2300 connections (as happens on some IP stacks) was logged at start up time as
2301 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2302 it from "IPv6 and IPv4", which means that two separate sockets are being
2303 used.
2304
230541. The debug output for gethostbyname2() or getipnodebyname() failures now
2306 says whether AF_INET or AF_INET6 was passed as an argument.
2307
230842. Exiwhat output was messed up when time zones were included in log
2309 timestamps.
2310
231143. Exiwhat now gives more information about the daemon's listening ports,
2312 and whether -tls-on-connect was used.
2313
231444. The "port" option of the smtp transport is now expanded.
2315
231645. A "message" modifier in a "warn" statement in a non-message ACL was being
2317 silently ignored. Now an error message is written to the main and panic
2318 logs.
2319
232046. There's a new ACL modifier called "logwrite" which writes to a log file
2321 as soon as it is encountered.
2322
232347. Added $local_user_uid and $local_user_gid at routing time.
2324
232548. Exim crashed when trying to verify a sender address that was being
2326 rewritten to "<>".
2327
232849. Exim was recognizing only a space character after ".include". It now also
2329 recognizes a tab character.
2330
233150. Fixed several bugs in the Perl script that creates the exim.8 man page by
2332 extracting the relevant information from the specification. The man page no
2333 longer contains scrambled data for the -d option, and I've added a section
2334 at the front about calling Exim under different names.
2335
233651. Added "extra_headers" argument to the "mail" command in filter files.
2337
233852. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2339 crash.
2340
234153. Installed eximstats 1.29.
2342
234354. Added transport_filter_timeout as a generic transport option.
2344
234555. Exim no longer adds an empty Bcc: header to messages that have no To: or
2346 Cc: header lines. This was required by RFC 822, but it not required by RFC
2347 2822.
2348
234956. Exim used to add From:, Date:, and Message-Id: header lines to any
2350 incoming messages that did not have them. Now it does so only if the
2351 message originates locally, that is, if there is no associated remote host
2352 address. When Resent- header lines are present, this applies to the Resent-
2353 lines rather than the non-Resent- lines.
2354
235557. Drop incoming SMTP connection after too many syntax or protocol errors. The
2356 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2357
235858. Messages for configuration errors now include the name of the main
2359 configuration file - useful now that there may be more than one file in a
2360 list (.included file names were always shown).
2361
236259. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2363 for those rare installations that do not start the daemon as root or run it
2364 setuid root. I've cut out the call to initgroups() if the daemon is not
2365 root at that time.
2366
236760. The Exim user and group can now be bound into the binary as text strings
2368 that are looked up at the start of Exim's processing.
2369
237061. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2371
237262. Added $mailstore_basename variable.
2373
237463. Installed patch to sieve.c from Michael Haardt.
2375
237664. When Exim failed to open the panic log after failing to open the main log,
2377 the original message it was trying to log was written to stderr and debug
2378 output, but if they were not available (the usual case in production), it
2379 was lost. Now it is written to syslog before the two lines that record the
2380 failures to open the logs.
2381
238265. Users' Exim filters run in subprocesses under the user's uid. It is
2383 possible for a "deliver" command or an alias in a "personal" command to
2384 provoke an address rewrite. If logging of address rewriting is configured,
2385 this fails because the process is not running as root or exim. There may be
2386 a better way of dealing with this, but for the moment (because 4.30 needs
2387 to be released), I have disabled address rewrite logging when running a
2388 filter in a non-root, non-exim process.
2389
2390
2391Exim version 4.24
2392-----------------
2393
2394 1. The buildconfig auxiliary program wasn't quoting the value set for
2395 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2396 not defined. This bug was masked in 4.22 by the effect that was fixed in
2397 change 4.23/1.
2398
2399 2. Some messages that were rejected after a message id was allocated were
2400 shown as "incomplete" by exigrep. It no longer does this for messages that
2401 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2402
2403 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2404 have allow_domain_literals set, the ID did not get logged in the <= line.
2405 Domain literals are now always recognized in Message-ID: header lines.
2406
2407 4. The first argument for a ${extract expansion item is the key name or field
2408 number. Leading and trailing spaces in this item were not being ignored,
2409 causing some misleading effects.
2410
2411 5. When deliver_drop_privilege was set, single queue runner processes started
2412 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2413 same command in the process it spins off) were not dropping privilege.
2414
2415 6. When the daemon running as "exim" started a queue runner, it always
2416 re-executed Exim in the spun-off process. This is a waste of effort when
2417 deliver_drop_privilege is set. The new process now just calls the
2418 queue-runner function directly.
2419
2420
2421Exim version 4.23
2422-----------------
2423
2424 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2425 HEADERS_CHARSET.
2426
2427 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2428 ignored. Though the use of -oP was forcing the writing of a pid file, it
2429 was always written to the default place.
2430
2431 3. If the message "no IP address found for host xxxx" is generated during
2432 incoming verification, it is now followed by identification of the incoming
2433 connection (so you can more easily find what provoked it).
2434
2435 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2436
2437 5. Added some features to "harden" Exim a bit more against certain attacks:
2438
2439 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2440 be put in Local/Makefile. This is like the never_users runtime option,
2441 but it cannot be overridden. The default setting is "root".
2442
2443 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2444 prefix string with which any file named in a -C command line option
2445 must start.
2446
2447 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2448 is retained for -C and -D only if the caller of Exim is root. Without
2449 it, the exim user may also use -C and -D and retain privilege.
2450
2451 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2452 command line option is disabled.
2453
2454 6. Macro names set by the -D option must start with an upper case letter, just
2455 like macro names defined in the configuration file.
2456
2457 7. Added "dereference=" facility to LDAP.
2458
2459 8. Two instances of the typo "uknown" in the source files are fixed.
2460
2461 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2462 the Configure-Makefile script screwed up while processing it.
2463
246410. Incorporated PCRE 4.4.
2465
246611. The SMTP synchronization check was not operating right at the start of an
2467 SMTP session. For example, it could not catch a HELO sent before the client
2468 waited for the greeting. There is now a check for outstanding input at the
2469 point when the greeting is written. Because of the duplex, asynchronous
2470 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2471 way, but not yet received, when the check is performed.
2472
247312. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2474 on TCP/IP sockets, because this apparently causes some broken clients to
2475 timeout.
2476
247713. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2478 unchanged) from the Cygwin maintainer.
2479
248014. The code for -bV that shows what is in the binary showed "mbx" when maildir
2481 was supported instead of testing for mbx. Effectively a typo.
2482
248315. The spa authenticator server code was not checking that the input it
2484 received was valid base64.
2485
248616. The debug output line for the "set" modifier in ACLs was not showing the
2487 name of the variable that was being set.
2488
248917. Code tidy: the variable type "vtype_string" was never used. Removed it.
2490
249118. Previously, a reference to $sender_host_name did not cause a DNS reverse
2492 lookup on its own. Something else was needed to trigger the lookup. For
2493 example, a match in host_lookup or the need for a host name in a host list.
2494 Now, if $sender_host_name is referenced and the host name has not yet been
2495 looked up, a lookup is performed. If the lookup fails, the variable remains
2496 empty, and $host_lookup_failed is set to "1".
2497
249819. Added "eqi" as a case-independent comparison operator.
2499
250020. The saslauthd authentication condition could segfault if neither service
2501 nor realm was specified.
2502
250321. If an overflowing value such as "2048M" was set for message_size_limit, the
2504 error message that was logged was misleading, and incoming SMTP
2505 connections were dropped. The message is now more accurate, and temporary
2506 errors are given to SMTP connections.
2507
250822. In some error situations (such as 21 above) Exim rejects all SMTP commands
2509 (except RSET) with a 421 error, until QUIT is received. However, it was
2510 failing to send a response to QUIT.
2511
251223. The HELO ACL was being run before the code for helo_try_verify_hosts,
2513 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2514 ACL is now run after the helo_try_verify_hosts code.
2515
251624. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2517 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2518 case-independent, so other case variants are also recognized). Apparently
2519 some systems use these upper case variants.
2520
252125. If more than two messages were waiting for the same host, and a transport
2522 filter was specified for the transport, Exim sent two messages over the
2523 same TCP/IP connection, and then failed with "socket operation on non-
2524 socket" when it tried to send the third.
2525
252626. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2527
252827. The extern definition of crypt16() in expand.c was not being excluded when
2529 the OS had its own crypt16() function.
2530
253128. Added bounce_return_body as a new option, and bounce_return_size_limit
2532 as a preferred synonym for return_size_limit, both as an option and as an
2533 expansion variable.
2534
253529. Added LIBS=-liconv to OS/Makefile-OSF1.
2536
253730. Changed the default configuration ACL to relax the local part checking rule
2538 for addresses that are not in any local domains. For these addresses,
2539 slashes and pipe symbols are allowed within local parts, but the sequence
2540 /../ is explicitly forbidden.
2541
254231. SPA server authentication was not clearing the challenge buffer before
2543 using it.
2544
254532. log_message in a "warn" ACL statement was writing to the reject log as
2546 well as to the main log, which contradicts the documentation and doesn't
2547 seem right (because no rejection is happening). So I have stopped it.
2548
254933. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2550 However, I am unable to do any testing of this.
2551
255234. Fixed an infelicity in the appendfile transport. When checking directories
2553 for a mailbox, to see if any needed to be created, it was accidentally
2554 using path names with one or more superfluous leading slashes; tracing
2555 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2556
255735. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2558 contents are added to the log line that is written for every discarded
2559 recipient. (Previously a log_message setting was ignored.)
2560
256136. The ${quote: operator now quotes the string if it is empty.
2562
256337. The install script runs exim in order to find its version number. If for
2564 some reason other than non-existence or emptiness, which it checks, it
2565 could not run './exim', it was installing it with an empty version number,
2566 i.e. as "exim-". This error state is now caught, and the installation is
2567 aborted.
2568
256938. An argument was missing from the function that creates an error message
2570 when Exim fails to connect to the socket for saslauthd authentication.
2571 This could cause Exim to crash, or give a corrupted message.
2572
257339. Added isip, isip4, and isip6 to ${if conditions.
2574
257540. The ACL variables $acl_xx are now saved with the message, and can be
2576 accessed later in routers, transports, and filters.
2577
257841. The new lookup type nwildlsearch is like wildlsearch, except that the key
2579 strings in the file are not string-expanded.
2580
258142. If a MAIL command specified a SIZE value that was too large to fit into an
2582 int variable, the check against message_size_limit failed. Such values are
2583 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2584 day this will have to be increased, but I don't think I want to be around
2585 when emails are that large.
2586
2587
2588
2589Exim version 4.22
2590-----------------
2591
2592 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2593 iconv() is not standard in FreeBSD.
2594
2595 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2596 IPv6 enabled. The observed symptom was a segmentation fault on return from
2597 the function os_common_find_running_interfaces() in src/os.c.
2598
2599 3. In the check_special_case() function in daemon.c I had used "errno" as an
2600 argument name, which causes warnings on some systems. This was basically a
2601 typo, since it was named "eno" in the comments!
2602
2603 4. The code that waits for the clock to tick (at a resolution of some fraction
2604 of a second) so as to ensure message-id uniqueness was always waiting for
2605 at least one whole tick, when it could have waited for less. [This is
2606 almost certainly not relevant at current processor speeds, where it is
2607 unlikely to ever wait at all. But we try to future-proof.]
2608
2609 5. The function that sleeps for a time interval that includes fractions of a
2610 second contained a race. It did not block SIGALRM between setting the
2611 timer, and suspending (a couple of lines later). If the interval was short
2612 and the sigsuspend() was delayed until after it had expired, the suspension
2613 never ended. On busy systems this could lead to processes getting stuck for
2614 ever.
2615
2616 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2617 process, before it forks any delivery processes. The open lookup caching
2618 mechanism meant that the open file or database connection was passed into
2619 the delivery process. The problem was that delivery processes always tidy
2620 up cached lookup data. This could cause a problem for the next delivery
2621 process started by the queue runner, because the external queue runner
2622 process does not know about the closure. So the next delivery process
2623 still has data in the lookup cache. In the case of a file lookup, there was
2624 no problem because closing a file descriptor in a subprocess doesn't affect
2625 the parent. However, if the lookup was caching a connection to a database,
2626 the connection was closed, and the second delivery process was likely to
2627 see errors such as "PGSQL: query failed: server closed the connection
2628 unexpectedly". The problem has been fixed by closing all cached lookups
2629 in a queue runner before running a delivery process.
2630
2631 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2632 seem to have the "const" qualifier which it has on other OS. I've
2633 parameterised it.
2634
2635 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2636 *of the same type* (client or server) with the same public name that an
2637 error should be diagnosed.
2638
2639 9. When Exim looked up a host name for an IP address, but failed to find the
2640 original IP address when looking up the host name (a safety check), it
2641 output the message "<ip address> does not match any IP for NULL", which was
2642 confusing, to say the least. The bug was that the host name should have
2643 appeared instead of "NULL".
2644
264510. Since release 3.03, if Exim is called by a uid other than root or the Exim
2646 user that is built into the binary, and the -C or -D options is used, root
2647 privilege is dropped before the configuration file is read. In addition,
2648 logging is switched to stderr instead of the normal log files. If the
2649 configuration then re-defines the Exim user, the unprivileged environment
2650 is probably not what is expected, so Exim logs a panic warning message (but
2651 proceeds).
2652
2653 However, if deliver_drop_privilege is set, the unprivileged state may well
2654 be exactly what is intended, so the warning has been cut out in that case,
2655 and Exim is allowed to try to write to its normal log files.
2656
2657
2658Exim version 4.21
2659-----------------
2660
2661 1. smtp_return_error_details was not giving details for temporary sender
2662 or receiver verification errors.
2663
2664 2. Diagnose a configuration error if two authenticators have the same public
2665 name.
2666
2667 3. Exim used not to create the message log file for a message until the first
2668 delivery attempt. This could be confusing when incoming messages were held
2669 for policy or load reasons. The message log file is now created at the time
2670 the message is received, and an initial "Received" line is written to it.
2671
2672 4. The automatically generated man page for command line options had a minor
2673 bug that caused no ill effects; however, a more serious problem was that
2674 the procedure for building the man page automatically didn't always
2675 operate. Consequently, release 4.20 contains an out-of-date version. This
2676 shouldn't happen again.
2677
2678 5. When building Exim with embedded Perl support, the script that builds the
2679 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2680 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2681
2682 6. The freeze_tell option was not being used for messages that were frozen on
2683 arrival, either by an ACL or by local_scan().
2684
2685 7. Added the smtp_incomplete_transaction log selector.
2686
2687 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2688 was accepting AUTH without a new EHLO.
2689
2690 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2691 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2692 RFC.
2693
269410. Logging of TCP/IP connections (when configured) now happens in the main
2695 daemon process instead of the child process, so that the TCP/IP connection
2696 count is more accurate (but it can never be perfect).
2697
269811. The use of "drop" in a nested ACL was not being handled correctly in the
2699 outer ACL. Now, if condition failure induced by the nested "drop" causes
2700 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2701 or "require"), the connection is dropped.
2702
270312. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2704 that yield "discard" can now be used with an "accept" or a "discard" verb,
2705 but an error is generated for any others (because I can't see a useful way
2706 to define what should happen).
2707
270813. When an ACL is read dynamically from a file (or anywhere else), the lines
2709 are now processed in the same way as lines in the Exim configuration file.
2710 In particular, continuation lines are supported.
2711
271214. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2713
271415. Added -ti meaning -t -i.
2715
271616. Check for letters, digits, hyphens, and dots in the names of dnslist
2717 domains, and warn by logging if others are found.
2718
271917. At least on BSD, alignment is not guarenteed for the array of ifreq's
2720 returned from GIFCONF when Exim is trying to find the list of interfaces on
2721 a host. The code in os.c has been modified to copy each ifreq to an aligned
2722 structure in all cases.
2723
2724 Also, in some cases, the returned ifreq's were being copied to a 'struct
2725 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2726 means the last couple of bytes of an IPv6 address could be chopped if the
2727 ifreq contained only a normal sockaddr (14 bytes storage).
2728
272918. Named domain lists were not supported in the hosts_treat_as_local option.
2730 An entry such as +xxxx was not recognized, and was treated as a literal
2731 domain name.
2732
273319. Ensure that header lines added by a DATA ACL are included in the reject log
2734 if the ACL subsequently rejects the message.
2735
273620. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2737 MD5 (which is deprecated).
2738
273921. When testing a filter file using -bf, Exim was writing a message when it
2740 took the sender from a "From " line in the message, but it was not doing so
2741 when it took $return_path from a Return-Path: header line. It now does.
2742
274322. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2744 with a valid header line field name (a series of printing characters
2745 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2746
274723. Changed "disc" in the source to "disk" to conform to the documentation and
2748 the book and for uniformity.
2749
275024. Ignore Sendmail's -Ooption=value command line item.
2751
275225. When execve() failed while trying to run a command in a pipe transport,
2753 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2754 could be confused with a return value of 69 from the command itself. This
2755 has been changed to 127, the value the shell returns if it is asked to run
2756 a non-existent command. The wording for the related log line suggests a
2757 non-existent command as the problem.
2758
275926. If received_header_text expands to an empty string, do not add a Received:
2760 header line to the message. (Well, it adds a token one on the spool, but
2761 marks it "old" so that it doesn't get used or transmitted.)
2762
276327. Installed eximstats 1.28 (addition of -nt option).
2764
276528. There was no check for failure on the call to getsockname() in the daemon
2766 code. This can fail if there is a shortage of resources on the system, with
2767 ENOMEM, for example. A temporary error is now given on failure.
2768
276929. Contrary to the C standard, it seems that in some environments, the
2770 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2771 program. Exim now does this explicitly; it affects the formatting of
2772 timestamps using strftime().
2773
277430. If exiqsumm was given junk data, it threw up some uninitialized variable
2775 complaints. I've now initialized all the variables, to avoid this.
2776
277732. Header lines added by a system filter were not being "seen" during
2778 transport-time rewrites.
2779
278033. The info_callback() function passed to OpenSSL is set up with type void
2781 (*)(SSL *, int, int), as described somewhere. However, when calling the
2782 function (actually a macro) that sets it up, the type void(*)() is
2783 expected. I've put in a cast to prevent warnings from picky compilers.
2784
278534. If a DNS black list lookup found a CNAME record, but there were no A
2786 records associated with the domain it pointed at, Exim crashed.
2787
278835. If a DNS black list lookup returned more than one A record, Exim ignored
2789 all but the first. It now scans all returned addresses if a particular IP
2790 value is being sought. In this situation, the contents of the
2791 $dnslist_value variable are a list of all the addresses, separated by a
2792 comma and a space.
2793
279436. Tightened up the rules for host name lookups using reverse DNS. Exim used
2795 to accept a host name and all its aliases if the forward lookup for any of
2796 them yielded the IP address of the incoming connection. Now it accepts only
2797 those names whose forward lookup yields the correct IP address. Any other
2798 names are discarded. This closes a loophole whereby a rogue DNS
2799 administrator could create reverse DNS records to break through a
2800 wildcarded host restriction in an ACL.
2801
280237. If a user filter or a system filter that ran in a subprocess used any of
2803 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2804 the wrong values were passed to the pipe command ($thisaddress had the
2805 value of $0, $0 had the value of $1, etc). This bug was introduced by
2806 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2807
280838. Improved the line breaking for long SMTP error messages from ACLs.
2809 Previously, if there was no break point between 40 and 75 characters, Exim
2810 left the rest of the message alone. Two changes have been made: (a) I've
2811 reduced the minimum length to 35 characters; (b) if it can't find a break
2812 point between 35 and 75 characters, it looks ahead and uses the first one
2813 that it finds. This may give the occasional overlong line, but at least the
2814 remaining text gets split now.
2815
281639. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2817 file descriptors might be low, and that setting 1000 would always raise it.
2818 It turns out that in some environments, the limit is already over 1000 and
2819 that lowering it causes trouble. So now Exim takes care not to decrease it.
2820
282140. When delivering a message, the value of $return_path is set to $sender_
2822 address at the start of routing (routers may change the value). By an
2823 oversight, this default was not being set up when an address was tested by
2824 -bt or -bv, which affected the outcome if any router or filter referred to
2825 $return_path.
2826
282741. The idea of the "warn" ACL verb is that it adds a header or writes to the
2828 log only when "message" or "log_message" are set. However, if one of the
2829 conditions was an address verification, or a call to a nested ACL, the
2830 messages generated by the underlying test were being passed through. This
2831 no longer happens. The underlying message is available in $acl_verify_
2832 message for both "message" and "log_message" expansions, so it can be
2833 passed through if needed.
2834
283542. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2836 new expansion $bh_ to give the encoded byte string without charset
2837 translation. Translation happens only if iconv() is available; HAVE_ICONV
2838 indicates this at build time. HEADERS_CHARSET gives the charset to
2839 translate to; headers_charset can change it in the configuration, and
2840 "headers charset" can change it in an individual filter file.
2841
284243. Now that we have a default RFC 2047 charset (see above), the code in Exim
2843 that creates RFC 2047 encoded "words" labels them as that charset instead
2844 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2845 expansion operator; (ii) when Exim creates a From: line for a local
2846 message; (iii) when a header line is rewritten to include a "phrase" part.
2847
284844. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2849 buggy, causing it to skip the first lines of messages whose message ID
2850 ended in 'D'. This would not have bitten before Exim release 4.14, because
2851 message IDs were unlikely to end in 'D' before then. The effect was to have
2852 incorrect size information for certain domains.
2853
285445. #include "config.h" was missing at the start of the crypt16.c module. This
2855 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2856 noticed.
2857
285846. If there was a timeout during a "random" callout check, Exim treated it as
2859 a failure of the random address, and carried on sending RSET and the real
2860 address. If the delay was just some slowness somewhere, the response to the
2861 original RCPT would be taken as a response to RSET and so on, causing
2862 mayhem of various kinds.
2863
286447. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2865 when I implemented it. It didn't allow for the fact that some option values
2866 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2867 the right test for positive values.
2868
286948. Domain names in DNS records are case-independent. Exim always looks them up
2870 in lower case. Some resolvers return domain names in exactly the case they
2871 appear in the zone file, that is, they may contain uppercase letters. Not
2872 all resolvers do this - some return always lower case. Exim was treating a
2873 change of case by a resolver as a change of domain, similar to a widening
2874 of a domain abbreviation. This triggered its re-routing code and so it was
2875 trying to route what was effectively the same domain again. This normally
2876 caused routing to fail (because the router wouldn't handle the domain
2877 twice). Now Exim checks for this case specially, and just changes the
2878 casing of the domain that it ultimately uses when it transmits the message
2879 envelope.
2880
288149. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2882 module.
2883
288450. If a filter generated a file delivery with a non-absolute name (possible if
2885 no home directory exists for the router), the forbid_file option was not
2886 forbidding it.
2887
288851. Added '&' feature to dnslists, to provide bit mask matching in addition to
2889 the existing equality matching.
2890
289152. Exim was using ints instead of ino_t variables in some places where it was
2892 dealing with inode numbers.
2893
289453. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2895 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2896 called TMPDIR, and if it finds it is different, it changes its value.
2897
289854. The smtp_printf() function is now made available to local_scan() so
2899 additional output lines can be written before returning. There is also an
2900 smtp_fflush() function to enable the detection of a dropped connection.
2901 The variables smtp_input and smtp_batched_input are exported to
2902 local_scan().
2903
290455. Changed the default runtime configuration: the message "Unknown user"
2905 has been removed from the ACL, and instead placed on the localuser router,
2906 using the cannot_route_message feature. This means that any verification
2907 failures that generate their own messages won't get overridden. Similarly,
2908 the "Unrouteable address" message that was in the ACL for unverifiable
2909 relay addresses has also been removed.
2910
291156. Added hosts_avoid_esmtp to the smtp transport.
2912
291357. The exicyclog script was not checking for the esoteric option
2914 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2915 will work only if exicyclog is run under the appropriate euid.
2916
291758. Following a discussion on the list, the rules by which Exim recognises line
2918 endings on incoming messages have been changed. The -dropcr and drop_cr
2919 options are now no-ops, retained only for backwards compatibility. The
2920 following line terminators are recognized: LF CRLF CR. However, special
2921 processing applies to CR:
2922
2923 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2924 nor a local message in the state where . is a terminator.
2925
2926 (ii) If a bare CR is encountered in a header line, an extra space is added
2927 after the line terminator so as not to end the header. The reasoning
2928 behind this is that bare CRs in header lines are most likely either
2929 to be mistakes, or people trying to play silly games.
2930
293159. The size of a message, as listed by "-bp" or in the Exim monitor window,
2932 was being incorrectly given as 18 bytes larger than it should have been.
2933 This is a VOB (very old bug).
2934
293560. This may never have affected anything current, but just in case it has:
2936 When the local host is found other than at the start of a list of hosts,
2937 the local host, those with the same MX, and any that follow, are discarded.
2938 When the list in question was part of a longer list of hosts, the following
2939 hosts (not currently being processed) were also being discarded. This no
2940 longer happens. I'm not sure if this situation could ever has previously
2941 arisen.
2942
294361. Added the "/MX" feature to lists of hosts in the manualroute and query
2944 program routers.
2945
294662. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2947 header. This is something that is recommended in a new Internet Draft, and
2948 is something that is documented as being done by Sendmail. There are two
2949 possible values. For messages generated by the autoreply transport, Exim
2950 adds:
2951
2952 Auto-Submitted: auto-replied
2953
2954 whereas for all other generated messages (e.g. bounces) it adds
2955
2956 Auto-Submitted: auto-generated
2957
295863. The "personal" condition in filters now includes a test for the
2959 Auto-Submitted: header. If it contains the string "auto-" the message it
2960 not considered personal.
2961
296264. Added rcpt_include_affixes as a generic transport option.
2963
296465. Added queue_only_override (default true).
2965
296666. Added the syslog_duplication option.
2967
296867. If what should have been the first header line of a message consisted of
2969 a space followed by a colon, Exim was mis-interpreting it as a header line.
2970 It isn't of course - it is syntactically invalid and should therefore be
2971 treated as the start of the message body. The misbehaviour could have
2972 caused a number of strange effects, including loss of data in subsequent
2973 header lines, and spool format errors.
2974
297568. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2976 client host had authenticated. This control can now be exercised by an ACL
2977 for more flexibility.
2978
297969. By default, callouts do not happen when testing with -bh. There is now a
2980 variant, -bhc, which does actually run the callout code, including
2981 consulting and updating the callout cache.
2982
298370. Added support for saslauthd authentication, courtesy of Alexander
2984 Sabourenkov.
2985
298671. If statvfs() failed on the spool or log directories while checking their
2987 size for availability, Exim confusingly gave the error "space shortage".
2988 Furthermore, in debugging mode it crashed with a floating point exception.
2989 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2990 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2991 serious problem, Exim now writes to the main and panic logs when this
2992 happens, with details of the failure. It then refuses to accept the
2993 incoming message, giving the message "spool directory problem" or "log
2994 directory problem" with a 421 code for SMTP messages.
2995
299672. When Exim is about to re-exec itself, it ensures that the file descriptors
2997 0, 1, and 2 exist, because some OS complain for execs without them (see
2998 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2999 descriptors. However, the code omitted to check that the open succeeded,
3000 causing mysterious errors if for some reason the permissions on /dev/null
3001 got screwed. Now Exim writes a message to the main and panic logs, and
3002 bombs out if it can't open /dev/null.
3003
300473. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3005 interact so that it is all more flexible. It is supposed to remain
3006 backwards compatible. Also added extra_local_interfaces.
3007
300874. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3009 to bomb out with an assertion failure - to the client this appears as a
3010 connection drop. This problem occurs in the part of the code that was taken
3011 from the Samba project. Fortunately, the assertion is in a very simple
3012 function, so I have fixed this by reproducing the function inline in the
3013 one place where it is called, and arranging for authentication to fail
3014 instead of killing the process with assert().
3015
301675. The SPA client code was not working when the server requested OEM rather
3017 than Unicode encoding.
3018
301976. Added code to make require_files with a specific uid setting more usable in
3020 the case where statting the file as root fails - usually a non-root-mounted
3021 NFS file system. When this happens and the failure is EACCES, Exim now
3022 forks a subprocess and does the per-uid checking as the relevant uid.
3023
302477. Added process_log_path.
3025
302678. If log_file_path was not explicitly set, a setting of check_log_space or
3027 check_log_inodes was ignored.
3028
302979. If a space check for the spool or log partitions fails, the incident is now
3030 logged. Of course, in the latter case the data may get lost...
3031
303280. Added the %p formatting code to string_format() so that it can be used to
3033 print addresses in debug_print(). Adjusted all the address printing in the
3034 debugging in store.c to use %p rather than %d.
3035
303681. There was a concern that a line of code in smtp_in.c could overflow a
3037 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3038 initially expressed, the concern was not well-founded, because trailing
3039 spaces are removed early. However, if the trailing spaces were followed by
3040 a NULL, they did not get removed, so the overflow was possible. Two fixes
3041 were applied:
3042
3043 (a) I re-wrote the offending code in a cleaner fashion.
3044 (b) If an incoming SMTP command contains a NULL character, it is rejected
3045 as invalid.
3046
304782. When Exim changes uid/gid to the Exim user at daemon start time, it now
3048 runs initgroups(), so that if the Exim user is in any additional groups,
3049 they will be used during message reception.
3050
3051
3052Exim version 4.20
3053-----------------
3054
3055The change log for 4.20 and earlier releases has been archived.
3056
3057****