Michael Haardt's patch to update Sieve to RFC3028bis.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
5ea81592 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.162 2005/06/17 10:47:05 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
90e9ce59
PH
146PH/18 If the "use_postmaster" option was set for a recipient callout together
147 with the "random" option, the postmaster address was used as the MAIL
148 FROM address for the random test, but not for the subsequent recipient
149 test. It is now used for both.
150
5ea81592
PH
151PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
152 patch removes a few documentation additions to RFC 3028, because the
153 latest draft now contains them. It adds the new en;ascii-case comparator
154 and a new error check for 8bit text in MIME parts. Comparator and
155 require names are now matched exactly. I enabled the subaddress
156 extension, but it is not well tested yet (read: it works for me)."
157
e5a9dba6 158
7982096b
PH
159Exim version 4.51
160-----------------
161
1a46a8c5
PH
162TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
163 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
164
2f079f46 165TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
166
167TK/03 Fix ACL "regex" condition not being available in MIME ACL.
168
169PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
170 to test Sieve filters that use "vacation".
171
172PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
173 that changes the way the GnuTLS parameters are stored in the cache file.
174 The new format can be generated externally. For backward compatibility,
175 if the data in the cache doesn't make sense, Exim assumes it has read an
176 old-format file, and it generates new data and writes a new file. This
177 means that you can't go back to an older release without removing the
178 file.
179
180PH/03 A redirect router that has both "unseen" and "one_time" set does not
181 work if there are any delivery delays because "one_time" forces the
182 parent to be marked "delivered", so its unseen clone is never tried
183 again. For this reason, Exim now forbids the simultaneous setting of
184 these two options.
185
186PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
187 redirected to themselves ("homonym" addresses). Read the long ChangeLog
188 entry if you want to know the details. The fix, however, neglected to
189 consider the case when local delivery batching is involved. The test for
190 "previously delivered" was not happening when checking to see if an
191 address could be batched with a previous (undelivered) one; under
192 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 193 address.
1a46a8c5
PH
194
195PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
196 in its include files, and this causes problems building Exim.
197
198PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
199 header_syntax) but Exim was just ignoring anything given after a slash.
200 In particular, this caused confusion with an attempt to use "verify =
201 reverse_host_lookup/defer_ok". An error is now given when options are
202 supplied for verify items that do not have them. (Maybe reverse_host_
203 lookup should have a defer_ok option, but that's a different point.)
204
205PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
206 defined by RFC 821) to 2048, because there were problems with some AUTH
207 commands, and RFC 1869 says the size should be increased for extended
208 SMTP commands that take arguments.
209
210PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
211 Finch).
212
213PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
214 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 215
49c2d5ea
PH
216PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
217 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
218 whatever). Otherwise the password may appear in the log. Following change
219 PH/42 below, there is no longer a chance of it appearing in a bounce
220 message.
49c2d5ea 221
bf759a8b
PH
222PH/11 Installed exipick version 20050225.0 from John Jetmore.
223
83364d30
PH
224PH/12 If the last host in a fallback_hosts list was multihomed, only the first
225 of its addresses was ever tried. (Bugzilla bug #2.)
226
7999bbd7
PH
227PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
228 the result incorrectly in the debug output. (It correctly added a newline
229 to what was transported.)
230
7dbf77c9
PH
231TF/01 Added $received_time.
232
74e0617f
PH
233PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
234 commented out examples of how to interface to a virus scanner and to
235 SpamAssassin. Also added commented examples of av_scanner and
236 spamd_address settings.
237
2f079f46
PH
238PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
239 and controls are allowed in which ACLs. There were a couple of minor
240 errors. Some of the entries in the conditions table (which is a table of
241 where they are NOT allowed) were getting very unwieldy; rewrote them as a
242 negation of where the condition IS allowed.
243
8c841523
PH
244PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
245
7766a4f0
PH
246PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
247 header file does not have a version number, so I've had to invent a new
248 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
249 API. The code is untested by me (my Linux distribution still has 0.3.2 of
250 radiusclient), but it was contributed by a Radius user.
251
8b417f2c
PH
252PH/18 Installed Lars Mainka's patch for the support of CRL collections in
253 files or directories, for OpenSSL.
254
901f42cb
PH
255PH/19 When an Exim process that is running as root has to create an Exim log
256 file, it does so in a subprocess that runs as exim:exim so as to get the
257 ownership right at creation (otherwise, other Exim processes might see
258 the file with the wrong ownership). There was no test for failure of this
259 fork() call, which would lead to the process getting stuck as it waited
260 for a non-existent subprocess. Forks do occasionally fail when resources
261 run out. I reviewed all the other calls to fork(); they all seem to check
262 for failure.
263
f9b9210e
PH
264PH/20 When checking for unexpected SMTP input at connect time (before writing
265 the banner), Exim was not dealing correctly with a non-positive return
266 from the read() function. If the client had disconnected by this time,
267 the result was a log entry for a synchronization error with an empty
268 string after "input=" when read() returned zero. If read() returned -1
269 (an event I could not check), uninitialized data bytes were printed.
270 There were reports of junk text (parts of files, etc) appearing after
271 "input=".
272
54cdb463
PH
273PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
274
cf00dad6
PH
275PH/22 Added support for macro redefinition, and (re)definition in between
276 driver and ACL definitions.
277
acb1b346
PH
278PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
279 forgetting to use the resulting value; it was using the unexpanded value.
280
c5ddb310
PH
281PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
282 hadn't been configured. The fix is from Juergen Kreileder, who
283 understands it better than I do:
284
285 "Here's what I see happening with three configured cyrus_sasl
286 authenticators configured (plain, login, cram-md5):
287
288 On startup auth_cyrus_sasl_init() gets called for each of these.
289 This means three calls to sasl_listmech() without a specified mech_list.
290 => SASL tests which mechs of all available mechs actually work
291 => three warnings about OTP not working
292 => the returned list contains: plain, login, cram-md5, digest-md5, ...
293
294 With the patch, sasl_listmech() also gets called three times. But now
295 SASL's mech_list option is set to the server_mech specified in the the
296 authenticator. Or in other words, the answer from sasl_listmech()
297 gets limited to just the mech you're testing for (which is different
298 for each call.)
299 => the return list contains just 'plain' or 'login', 'cram-md5' or
300 nothing depending on the value of ob->server_mech.
301
302 I've just tested the patch: Authentication still works fine,
303 unavailable mechs specified in the exim configuration are still
304 caught, and the auth.log warnings about OTP are gone."
305
31619da6
PH
306PH/25 When debugging is enabled, the contents of the command line are added
307 to the debugging output, even when log_selector=+arguments is not
308 specified.
309
bebaf0fc
PH
310PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
311 answer is "GNU", and only if the return is "GNU/something" is the answer
312 "Linux".
313
475fe28a
PH
314PH/27 $acl_verify_message is now set immediately after the failure of a
315 verification in an ACL, and so is available in subsequent modifiers. In
316 particular, the message can be preserved by coding like this:
317
318 warn !verify = sender
319 set acl_m0 = $acl_verify_message
320
321 Previously, $acl_verify_message was set only while expanding "message"
322 and "log_message" when a very denied access.
323
7e8bec7a
PH
324PH/28 Modified OS/os.c-Linux with
325
326 -#ifndef OS_LOAD_AVERAGE
327 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
328
329 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
330 nomenclature these days.)
331
e4a89c47
PH
332PH/29 Installed patch from the Sieve maintainer that adds the options
333 sieve_useraddress and sieve_subaddress to the redirect router.
334
5ca2a9a1
PH
335PH/30 In these circumstances:
336 . Two addresses routed to the same list of hosts;
337 . First host does not offer TLS;
338 . First host accepts first address;
339 . First host gives temporary error to second address;
340 . Second host offers TLS and a TLS session is established;
341 . Second host accepts second address.
342 Exim incorrectly logged both deliveries with the TLS parameters (cipher
343 and peerdn, if requested) that were in fact used only for the second
344 address.
7e8bec7a 345
c688b954
PH
346PH/31 When doing a callout as part of verifying an address, Exim was not paying
347 attention to any local part prefix or suffix that was matched by the
348 router that accepted the address. It now behaves in the same way as it
349 does for delivery: the affixes are removed from the local part unless
350 rcpt_include_affixes is set on the transport.
351
fed77020
PH
352PH/32 Add the sender address, as F=<...>, to the log line when logging a
353 timeout during the DATA phase of an incoming message.
354
7fe1560f
PH
355PH/33 Sieve envelope tests were broken for match types other than :is. I have
356 applied a patch sanctioned by the Sieve maintainer.
c688b954 357
ebb6e6d5
PH
358PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
359 the uid or gid is negative. A case of a negative gid caused this to be
360 noticed. The fix allows for either to be negative.
361
9c4e8f60
PH
362PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
363 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
364 overlooked.
365
366PH/36 The change PH/12 above was broken. Fixed it.
367
d7174846
PH
368PH/37 Exim used to check for duplicate addresses in the middle of routing, on
369 the grounds that routing the same address twice would always produce the
370 same answer. This might have been true once, but it is certainly no
371 longer true now. Routing a child address may depend on the previous
372 routing that produced that child. Some complicated redirection strategies
373 went wrong when messages had multiple recipients, and made Exim's
374 behaviour dependent on the order in which the addresses were given.
375
376 I have moved the duplicate checking until after the routing is complete.
377 Exim scans the addresses that are assigned to local and remote
378 transports, and removes any duplicates. This means that more work will be
379 done, as duplicates will always all be routed, but duplicates are
380 presumably rare, so I don't expect this is of any significance.
381
382 For deliveries to pipes, files, and autoreplies, the duplicate checking
383 still happens during the routing process, since they are not going to be
384 routed further.
385
cfe75fc3
PH
386PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
387 It corrects a timeout issue with spamd. This is Ian's comment: "The
388 background is that sometimes spamd either never reads data from a
389 connection it has accepted, or it never writes response data. The exiscan
390 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
391 blindly assumes that writes won't block so it may never time out."
392
be22d70e
PH
393PH/39 Allow G after quota size as well as K and M.
394
0612b098
PH
395PH/40 The value set for $authenticated_id in an authenticator may not contain
396 binary zeroes or newlines because the value is written to log lines and
397 to spool files. There was no check on this. Now the value is run through
398 the string_printing() function so that such characters are converted to
399 printable escape sequences.
400
2e0c1448
PH
401PH/41 $message_linecount is a new variable that contains the total number of
402 lines in the message. Compare $body_linecount, which is the count for the
403 body only.
404
447d236c
PH
405PH/42 Exim no longer gives details of delivery errors for specific addresses in
406 bounce and delay warning messages, except in certain special cases, which
407 are as follows:
408
409 (a) An SMTP error message from a remote host;
410 (b) A message specified in a :fail: redirection;
411 (c) A message specified in a "fail" command in a system filter;
412 (d) A message specified in a FAIL return from the queryprogram router;
413 (e) A message specified by the cannot_route_message router option.
414
415 In these cases only, Exim does include the error details in bounce and
416 warning messages. There are also a few cases where bland messages such
417 as "unrouteable address" or "local delivery error" are given.
418
d20976dc
PH
419PH/43 $value is now also set for the "else" part of a ${run expansion.
420
f656d135
PH
421PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
422 being worked on, but at least Exim now implements the latest version to
423 play with."
424
2e2a30b4
PH
425PH/45 In a pipe transport, although a timeout while waiting for the pipe
426 process to complete was treated as a delivery failure, a timeout while
427 writing the message to the pipe was logged, but erroneously treated as a
428 successful delivery. Such timeouts include transport filter timeouts. For
429 consistency with the overall process timeout, these timeouts are now
430 treated as errors, giving rise to delivery failures by default. However,
431 there is now a new Boolean option for the pipe transport called
432 timeout_defer, which, if set TRUE, converts the failures into defers for
433 both kinds of timeout. A transport filter timeout is now identified in
434 the log output.
435
9176e9f0
PH
436PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
437 systems where "make" and "gmake" are different, calling "gmake" at top
438 level broke things. I've arranged for the value of $(MAKE) to be passed
439 from the Makefile to this script so that it can call the same version of
440 "make".
441
7982096b 442
bbe902f0
PH
443A note about Exim versions 4.44 and 4.50
444----------------------------------------
445
446Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
447changes of various kinds. As a consequence, a big documentation update was
448needed. This delayed the release for rather longer than seemed good, especially
449in the light of a couple of (minor) security issues. Therefore, the changes
450that fixed bugs were backported into 4.43, to create a 4.44 maintenance
451release. So 4.44 and 4.50 are in effect two different branches that both start
452from 4.43.
453
454I have left the 4.50 change log unchanged; it contains all the changes since
4554.43. The change log for 4.44 is below; many of its items are identical to
456those for 4.50. This seems to be the most sensible way to preserve the
457historical information.
458
459
f7b63901 460Exim version 4.50
495ae4b0
PH
461-----------------
462
5fe762f6
PH
463 1. Minor wording change to the doc/README.SIEVE file.
464
139059f6 465 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 466 computation of the current number of files was incorrect.
495ae4b0 467
7086e875
PH
468 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
469 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
470 place.
471
35af9f61
PH
472 4. Give more explanation in the error message when the command for a transport
473 filter fails to execute.
474
b668c215
PH
475 5. There are several places where Exim runs a non-Exim command in a
476 subprocess. The SIGUSR1 signal should be disabled for these processes. This
477 was being done only for the command run by the queryprogram router. It is
478 now done for all such subprocesses. The other cases are: ${run, transport
479 filters, and the commands run by the lmtp and pipe transports.
480
a494b1e1
PH
481 6. Added CONFIGURE_GROUP build-time option.
482
483 7. Some older OS have a limit of 256 on the maximum number of file
484 descriptors. Exim was using setrlimit() to set 1000 as a large value
485 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
486 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 487
c5fcb476
PH
488 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
489 was an oversight, and furthermore, ever since the addition of extra
490 controls (e.g. 4.43/32), the checks on when to allow different forms of
491 "control" were broken. There should now be diagnostics for all cases when a
492 control that does not make sense is encountered.
493
69358f02
PH
494 9. Added the /retain_sender option to "control=submission".
495
5be20824
PH
49610. $recipients is now available in the predata ACL (oversight).
497
eb2c0248
PH
49811. Tidy the search cache before the fork to do a delivery from a message
499 received from the command line. Otherwise the child will trigger a lookup
500 failure and thereby defer the delivery if it tries to use (for example) a
501 cached ldap connection that the parent has called unbind on.
502
2a3eea10
PH
50312. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
504 of $address_data from the recipient verification was clobbered by the
505 sender verification.
506
50713. The value of address_data from a sender verification is now available in
508 $sender_address_data in subsequent conditions in the ACL statement.
509
23c7ff99
PH
51014. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
511
4deaf07d
PH
51215. Added a new option "connect=<time>" to callout options, to set a different
513 connection timeout.
514
926e1192
PH
51516. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
516 was its contents. (It was OK if the option was not defined at all.)
517
650edc6f
PH
51817. A "Completed" log line is now written for messages that are removed from
519 the spool by the -Mrm option.
520
2c7db3f5
PH
52118. New variables $sender_verify_failure and $recipient_verify_failure contain
522 information about exactly what failed.
523
3d235903
PH
52419. Added -dd to debug only the daemon process.
525
7c7ad977
PH
52620. Incorporated Michael Haardt's patch to ldap.c for improving the way it
527 handles timeouts, both on the server side and network timeouts. Renamed the
528 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
529
981756db
PH
53021. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
531 It is now set to "smtps".
532
d4eb88df
PH
53322. $host_address is now set to the target address during the checking of
534 ignore_target_hosts.
535
53623. When checking ignore_target_hosts for an ipliteral router, no host name was
537 being passed; this would have caused $sender_host_name to have been used if
538 matching the list had actually called for a host name (not very likely,
539 since this list is usually IP addresses). A host name is now passed as
540 "[x.x.x.x]".
541
7d468ab8
PH
54224. Changed the calls that set up the SIGCHLD handler in the daemon to use the
543 code that specifies a non-restarting handler (typically sigaction() in
544 modern systems) in an attempt to fix a rare and obscure crash bug.
545
54625. Narrowed the window for a race in the daemon that could cause it to ignore
547 SIGCHLD signals. This is not a major problem, because they are used only to
548 wake it up if nothing else does.
549
62c0818f
PH
55026. A malformed maildirsize file could cause Exim to calculate negative values
551 for the mailbox size or file count. Odd effects could occur as a result.
552 The maildirsize information is now recalculated if the size or filecount
553 end up negative.
554
26034054
PH
55527. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
556 support for a long time. Removed HAVE_SYS_VFS_H.
557
af66f652
PH
55828. Installed the latest version of exipick from John Jetmore.
559
90af77f4
PH
56029. In an address list, if the pattern was not a regular expression, an empty
561 subject address (from a bounce message) matched only if the pattern was an
562 empty string. Non-empty patterns were not even tested. This was the wrong
563 because it is perfectly reasonable to use an empty address as part of a
564 database query. An empty address is now tested by patterns that are
565 lookups. However, all the other forms of pattern expect the subject to
566 contain a local part and a domain, and therefore, for them, an empty
567 address still always fails if the pattern is not itself empty.
568
d8ef3577
PH
56930. Exim went into a mad DNS loop when attempting to do a callout where the
570 host was specified on an smtp transport, and looking it up yielded more
571 than one IP address.
572
5cb8cbc6
PH
57331. Re-factored the code for checking spool and log partition space into a
574 function that finds that data and another that does the check. The former
575 is then used to implement four new variables: $spool_space, $log_space,
576 $spool_inodes, and $log_inodes.
577
14702f5b
PH
57832. The RFC2047 encoding function was originally intended for short strings
579 such as real names; it was not keeping to the 75-character limit for
580 encoded words that the RFC imposes. It now respects the limit, and
581 generates multiple encoded words if necessary. To be on the safe side, I
582 have increased the buffer size for the ${rfc2047: expansion operator from
583 1024 to 2048 bytes.
584
063b1e99
PH
58533. It is now permitted to omit both strings after an "if" condition; if the
586 condition is true, the result is "true". As before, when the second string
587 is omitted, a false condition yields an empty string. This makes it less
588 cumbersome to write custom ACL and router conditions.
589
652e1b65
PH
59034. Failure to deliver a bounce message always caused it to be frozen, even if
591 there was an errors_to setting on the router. The errors_to setting is now
592 respected.
593
6f0c9a4f
PH
59435. If an IPv6 address is given for -bh or -bhc, it is now converted to the
595 canonical form (fully expanded) before being placed in
596 $sender_host_address.
597
33397d19
PH
59836. The table in the code that translates DNS record types into text (T_A to
599 "A" for instance) was missing entries for NS and CNAME. It is just possible
600 that this could have caused confusion if both these types were looked up
601 for the same domain, because the text type is used as part of Exim's
602 per-process caching. But the chance of anyone hitting this buglet seems
603 very small.
604
7bb56e1f
PH
60537. The dnsdb lookup has been extended in a number of ways.
606
607 (1) There is a new type, "zns", which walks up the domain tree until it
608 finds some nameserver records. It should be used with care.
609
ea3bc19b
PH
610 (2) There is a new type, "mxh", which is like "mx" except that it returns
611 just the host names, not the priorities.
612
613 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
614 looked up. The behaviour when one of the lookups defers can be
615 controlled by a keyword.
7bb56e1f 616
ea3bc19b 617 (4) It is now possible to specify the separator character for use when
7bb56e1f 618 multiple records are returned.
33397d19 619
0bcb2a0e
PH
62038. The dnslists ACL condition has been extended: it is now possible to supply
621 a list of IP addresses and/or domains to be looked up in a particular DNS
622 domain.
623
2ac0e484
PH
62439. Added log_selector=+queue_time_overall.
625
4e1fde53
PH
62640. When running the queue in the test harness, wait just a tad after forking a
627 delivery process, to get repeatability of debugging output.
628
de365ded
PH
62941. Include certificate and key file names in error message when GnuTLS fails
630 to set them up, because the GnuTLS error message doesn't include the name
631 of the failing file when there is a problem reading it.
632
f05da2e8
PH
63342. Allow both -bf and -bF in the same test run.
634
d6453af2
PH
63543. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
636
f7b63901
PH
63744. The "Exiscan patch" is now merged into the mainline Exim source.
638
63945. Sometimes the final signoff response after QUIT could fail to get
640 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
641 < 0 before doing a fflush(). This bug looks as though it goes back to the
642 introduction of TLS in release 3.20, but "sometimes" must have been rare
643 because the tests only now provoked it.
644
a444213a
PH
64546. Reset the locale to "C" after calling embedded Perl, in case it was changed
646 (this can affect the format of dates).
647
0ec020ea
PH
64847. exim_tidydb, when checking for the continued existence of a message for
649 which it has found a message-specific retry record, was not finding
650 messages that were in split spool directories. Consequently, it was
651 deleting retry records that should have stayed in existence.
652
b1206957
PH
65348. Steve fixed some bugs in eximstats.
654
65549. The SPA authentication driver was not abandoning authentication and moving
656 on to the next authenticator when an expansion was forced to fail,
657 contradicting the general specification for all authenticators. Instead it
658 was generating a temporary error. It now behaves as specified.
659
26dd5a95
PH
66050. The default ordering of permitted cipher suites for GnuTLS was pessimal
661 (the order specifies the preference for clients). The order is now AES256,
662 AES128, 3DES, ARCFOUR128.
663
343b2385
PH
66451. Small patch to Sieve code - explicitly set From: when generating an
665 autoreply.
666
1c5466b9
PH
66752. Exim crashed if a remote delivery caused a very long error message to be
668 recorded - for instance if somebody sent an entire SpamAssassin report back
669 as a large number of 550 error lines. This bug was coincidentally fixed by
670 increasing the size of one of Exim's internal buffers (big_buffer) that
671 happened as part of the Exiscan merge. However, to be on the safe side, I
672 have made the code more robust (and fixed the comments that describe what
673 is going on).
674
55ee9ee3
PH
67553. Now that there can be additional text after "Completed" in log lines (if
676 the queue_time_overall log selector is set), a one-byte patch to exigrep
677 was needed to allow it to recognize "Completed" as not the last thing in
678 the line.
679
d38f8232
PH
68054. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
681 patch that reportedly fixes this has been added. I am not expert enough to
682 create a test for it. This is what the patch creator wrote:
683
684 "I found a little strange behaviour of ldap code when working with
685 Windows 2003 AD Domain, where users was placed in more than one
686 Organization Units. When I tried to give exim partial DN, the exit code
687 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
688 But simultaneously result of request was absolutely normal ldap result,
689 so I produce this patch..."
690
3295e65b
PH
691 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
692 REFERENCE, so I have modified the code to exclude the patch when that macro
693 is not defined.
694
7102e136
PH
69555. Some experimental protocols are using DNS PTR records for new purposes. The
696 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
697 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
698 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 699 CAN-2005-0021
7102e136 700
3ca0ba97
PH
70156. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
702
c2bcbe20
PH
70357. Double the size of the debug message buffer (to 2048) so that more of very
704 long debug lines gets shown.
705
18ce445d
PH
70658. The exicyclog utility now does better if the number of log files to keep
707 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
708
1f5b4c3d
PH
70959. Two changes related to the smtp_active_hostname option:
710
711 (1) $smtp_active_hostname is now available as a variable.
712 (2) The default for smtp_banner uses $smtp_active_hostname instead
713 of $primary_hostname.
714
b975ba52
PH
71560. The host_aton() function is supposed to be passed a string that is known
716 to be a valid IP address. However, in the case of IPv6 addresses, it was
717 not checking this. This is a hostage to fortune. Exim now panics and dies
718 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
719 from a dnsdb PTR lookup with an IPv6 address that had more than 8
720 components; fortuitously, this particular loophole had already been fixed
721 by change 4.50/55 above.
722
723 If there are any other similar loopholes, the new check in host_aton()
724 itself should stop them being exploited. The report I received stated that
725 data on the command line could provoke the exploit when Exim was running as
726 exim, but did not say which command line option was involved. All I could
727 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
728 running as the user.
ea3a6f44 729 CAN-2005-0021
85b87bc2
PH
730
73161. There was a buffer overflow vulnerability in the SPA authentication code
732 (which came originally from the Samba project). I have added a test to the
733 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 734 CAN-2005-0022
b975ba52 735
17ffcae7
PH
73662. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
737 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
738
d95f9fdb
PH
73963. The daemon start-up calls getloadavg() while still root for those OS that
740 need the first call to be done as root, but it missed one case: when
741 deliver_queue_load_max is set with deliver_drop_privilege. This is
742 necessary for the benefit of the queue runner, because there is no re-exec
743 when deliver_drop_privilege is set.
744
86b8287f
PH
74564. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
746 This has been fixed.
747
60dc5e56
PH
74865. Caching of lookup data for "hosts =" ACL conditions, when a named host list
749 was in use, was not putting the data itself into the right store pool;
750 consequently, it could be overwritten for a subsequent message in the same
751 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
752 the caching.)
753
533244af
PH
75466. Added hosts_max_try_hardlimit to the smtp transport, default 50.
755
a5a28604
PH
75667. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
757 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
758 the function were treating the return as a boolean value, which happened to
759 work because 0=false and not-0=true, but is not correct code.
760
7e634d24
PH
76168. The host_aton() function was not handling scoped IPv6 addresses (those
762 with, for example, "%eth0" on the end) correctly.
763
3e11c26b
PH
76469. Fixed some compiler warnings in acl.c for the bitmaps specified with
765 negated items (that is, ~something) in unsigned ints. Some compilers
766 apparently mutter when there is no cast.
767
6729cf78
PH
76870. If an address verification called from an ACL failed, and did not produce a
769 user-specific message (i.e. there was only a "system" message), nothing was
770 put in $acl_verify_message. In this situation, it now puts the system
771 message there.
772
00f00ca5
PH
77371. Change 4.23/11 added synchronization checking at the start of an SMTP
774 session; change 4.31/43 added the unwanted input to the log line - except
775 that it did not do this in the start of session case. It now does.
776
c9bdd01c
PH
77772. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
778 This is wrong and can cause the other end to generate a synchronization
779 error if it is another Exim or anything else that does the synchronization
780 check. A QUIT command is no longer sent after a timeout.
781
d43194df
PH
78273. $host_lookup_deferred has been added, to make it easier to detect DEFERs
783 during host lookups.
784
fe5b5d0b
PH
78574. The defer_ok option of callout verification was not working if it was used
786 when verifying addresses in header lines, that is, for this case:
787
788 verify = header_sender/callout=defer_ok
789
76a2d7ba
PH
79075. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
791 those file descriptors could be used for SMTP connections. If anything
792 wrote to stderr (the example that came up was "warn" in embedded Perl), it
793 could be sent to the SMTP client, causing chaos. The daemon now opens
794 stdin, stdout, and stderr to /dev/null when it puts itself into the
795 background.
796
79776. Arrange for output from Perl's "warn" command to be written to Exim's main
798 log by default. The user can override this with suitable Perl magic.
799
04f7d5b9
PH
80077. The use of log_message on a "discard" ACL verb, which is supposed to add to
801 the log message when discard triggers, was not working for the DATA ACL or
802 for the non-SMTP ACL.
803
bc60667e
PH
80478. Error message wording change in sieve.c.
805
bb6e88ff
PH
80679. If smtp_accept_max_per_host was set, the number of connections could be
807 restricted to fewer than expected, because the daemon was trying to set up
808 a new connection before checking whether the processes handling previous
809 connections had finished. The check for completed processes is now done
810 earlier. On busy systems, this bug wouldn't be noticed because something
811 else would have woken the daemon, and it would have reaped the completed
812 process earlier.
813
1e70f85b
PH
81480. If a message was submitted locally by a user whose login name contained one
815 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
816 It caused a spool format error. I have fixed the spool reading code. A
817 related problem was that the "from" clause in the Received: line became
818 illegal because of the space(s). It is now covered by ${quote_local_part.
819
82081. Included the latest eximstats from Steve (adds average sizes to HTML Top
821 tables).
822
4e01f9d6
PH
82382. Updated OS/Makefile-AIX as per message from Mike Meredith.
824
1ee1cef2
PH
82583. Patch from Sieve maintainer to fix unterminated string problem in
826 "vacation" handling.
827
6e2b4ccc
PH
82884. Some minor changes to the Linux configuration files to help with other
829 OS variants using glibc.
830
8e669ac1
PH
83185. One more patch for Sieve to update vacation handling to latest spec.
832
495ae4b0 833
bbe902f0
PH
834----------------------------------------------------
835See the note above about the 4.44 and 4.50 releases.
836----------------------------------------------------
837
838
839Exim version 4.44
840-----------------
841
842 1. Change 4.43/35 introduced a bug that caused file counts to be
843 incorrectly computed when quota_filecount was set in an appendfile
844 transport
845
846 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
847 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
848 place.
849
850 3. Give more explanation in the error message when the command for a transport
851 filter fails to execute.
852
853 4. There are several places where Exim runs a non-Exim command in a
854 subprocess. The SIGUSR1 signal should be disabled for these processes. This
855 was being done only for the command run by the queryprogram router. It is
856 now done for all such subprocesses. The other cases are: ${run, transport
857 filters, and the commands run by the lmtp and pipe transports.
858
859 5. Some older OS have a limit of 256 on the maximum number of file
860 descriptors. Exim was using setrlimit() to set 1000 as a large value
861 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
862 systems. I've change it so that if it can't get 1000, it tries for 256.
863
864 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
865 was an oversight, and furthermore, ever since the addition of extra
866 controls (e.g. 4.43/32), the checks on when to allow different forms of
867 "control" were broken. There should now be diagnostics for all cases when a
868 control that does not make sense is encountered.
869
870 7. $recipients is now available in the predata ACL (oversight).
871
872 8. Tidy the search cache before the fork to do a delivery from a message
873 received from the command line. Otherwise the child will trigger a lookup
874 failure and thereby defer the delivery if it tries to use (for example) a
875 cached ldap connection that the parent has called unbind on.
876
877 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
878 of $address_data from the recipient verification was clobbered by the
879 sender verification.
880
88110. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
882 was its contents. (It was OK if the option was not defined at all.)
883
88411. A "Completed" log line is now written for messages that are removed from
885 the spool by the -Mrm option.
886
88712. $host_address is now set to the target address during the checking of
888 ignore_target_hosts.
889
89013. When checking ignore_target_hosts for an ipliteral router, no host name was
891 being passed; this would have caused $sender_host_name to have been used if
892 matching the list had actually called for a host name (not very likely,
893 since this list is usually IP addresses). A host name is now passed as
894 "[x.x.x.x]".
895
89614. Changed the calls that set up the SIGCHLD handler in the daemon to use the
897 code that specifies a non-restarting handler (typically sigaction() in
898 modern systems) in an attempt to fix a rare and obscure crash bug.
899
90015. Narrowed the window for a race in the daemon that could cause it to ignore
901 SIGCHLD signals. This is not a major problem, because they are used only to
902 wake it up if nothing else does.
903
90416. A malformed maildirsize file could cause Exim to calculate negative values
905 for the mailbox size or file count. Odd effects could occur as a result.
906 The maildirsize information is now recalculated if the size or filecount
907 end up negative.
908
90917. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
910 support for a long time. Removed HAVE_SYS_VFS_H.
911
ea3a6f44 91218. Updated exipick to current release from John Jetmore.
bbe902f0
PH
913
91419. Allow an empty sender to be matched against a lookup in an address list.
915 Previously the only cases considered were a regular expression, or an
916 empty pattern.
917
91820. Exim went into a mad DNS lookup loop when doing a callout where the
919 host was specified on the transport, if the DNS lookup yielded more than
920 one IP address.
921
ea3a6f44
NM
92221. The RFC2047 encoding function was originally intended for short strings
923 such as real names; it was not keeping to the 75-character limit for
924 encoded words that the RFC imposes. It now respects the limit, and
925 generates multiple encoded words if necessary. To be on the safe side, I
926 have increased the buffer size for the ${rfc2047: expansion operator from
927 1024 to 2048 bytes.
bbe902f0 928
ea3a6f44
NM
92922. Failure to deliver a bounce message always caused it to be frozen, even if
930 there was an errors_to setting on the router. The errors_to setting is now
931 respected.
bbe902f0
PH
932
93323. If an IPv6 address is given for -bh or -bhc, it is now converted to the
934 canonical form (fully expanded) before being placed in
935 $sender_host_address.
936
93724. Updated eximstats to version 1.33
938
ea3a6f44
NM
93925. Include certificate and key file names in error message when GnuTLS fails
940 to set them up, because the GnuTLS error message doesn't include the name
941 of the failing file when there is a problem reading it.
bbe902f0
PH
942
94326. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 944 As per change 25.
bbe902f0 945
ea3a6f44
NM
94627. Reset the locale to "C" after calling embedded Perl, in case it was changed
947 (this can affect the format of dates).
bbe902f0 948
ea3a6f44
NM
94928. exim_tidydb, when checking for the continued existence of a message for
950 which it has found a message-specific retry record, was not finding
951 messages that were in split spool directories. Consequently, it was
952 deleting retry records that should have stayed in existence.
bbe902f0
PH
953
95429. eximstats updated to version 1.35
955 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
956 1.35 - bugfix such that pie charts by volume are generated correctly
957
ea3a6f44
NM
95830. The SPA authentication driver was not abandoning authentication and moving
959 on to the next authenticator when an expansion was forced to fail,
960 contradicting the general specification for all authenticators. Instead it
961 was generating a temporary error. It now behaves as specified.
bbe902f0 962
ea3a6f44
NM
96331. The default ordering of permitted cipher suites for GnuTLS was pessimal
964 (the order specifies the preference for clients). The order is now AES256,
965 AES128, 3DES, ARCFOUR128.
bbe902f0 966
ea3a6f44
NM
96731. Small patch to Sieve code - explicitly set From: when generating an
968 autoreply.
bbe902f0 969
ea3a6f44
NM
97032. Exim crashed if a remote delivery caused a very long error message to be
971 recorded - for instance if somebody sent an entire SpamAssassin report back
972 as a large number of 550 error lines. This bug was coincidentally fixed by
973 increasing the size of one of Exim's internal buffers (big_buffer) that
974 happened as part of the Exiscan merge. However, to be on the safe side, I
975 have made the code more robust (and fixed the comments that describe what
976 is going on).
bbe902f0 977
ea3a6f44
NM
97833. Some experimental protocols are using DNS PTR records for new purposes. The
979 keys for these records are domain names, not reversed IP addresses. The
980 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
981 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
982 CAN-2005-0021
983
ea3a6f44
NM
98434. The host_aton() function is supposed to be passed a string that is known
985 to be a valid IP address. However, in the case of IPv6 addresses, it was
986 not checking this. This is a hostage to fortune. Exim now panics and dies
987 if the condition is not met. A case was found where this could be provoked
988 from a dnsdb PTR lookup with an IPv6 address that had more than 8
989 components; fortuitously, this particular loophole had already been fixed
990 by change 4.50/55 or 4.44/33 above.
991
992 If there are any other similar loopholes, the new check in host_aton()
993 itself should stop them being exploited. The report I received stated that
994 data on the command line could provoke the exploit when Exim was running as
995 exim, but did not say which command line option was involved. All I could
996 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
997 running as the user.
bbe902f0
PH
998 CAN-2005-0021
999
ea3a6f44
NM
100035. There was a buffer overflow vulnerability in the SPA authentication code
1001 (which came originally from the Samba project). I have added a test to the
1002 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
1003 CAN-2005-0022
1004
ea3a6f44
NM
100536. The daemon start-up calls getloadavg() while still root for those OS that
1006 need the first call to be done as root, but it missed one case: when
1007 deliver_queue_load_max is set with deliver_drop_privilege. This is
1008 necessary for the benefit of the queue runner, because there is no re-exec
1009 when deliver_drop_privilege is set.
bbe902f0 1010
ea3a6f44
NM
101137. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1012 was in use, was not putting the data itself into the right store pool;
1013 consequently, it could be overwritten for a subsequent message in the same
1014 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1015 the caching.)
bbe902f0 1016
ea3a6f44
NM
101738. Sometimes the final signoff response after QUIT could fail to get
1018 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1019 < 0 before doing a fflush(). This bug looks as though it goes back to the
1020 introduction of TLS in release 3.20, but "sometimes" must have been rare
1021 because the tests only now provoked it.
bbe902f0
PH
1022
1023
495ae4b0
PH
1024Exim version 4.43
1025-----------------
1026
1027 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1028 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1029 Now it returns an integer. A number of calls were still expecting a T/F
1030 return. Fortuitously, in all cases, the tests worked in OK situations,
1031 which is the norm. However, things would have gone wrong on any write
1032 failures on the smtp file descriptor. This function is used when sending
1033 messages over SMTP and also when doing verify callouts.
1034
1035 2. When Exim is called to do synchronous delivery of a locally submitted
1036 message (the -odf or -odi options), it no longer closes stderr before doing
1037 the delivery.
1038
1039 3. Implemented the mua_wrapper option.
1040
1041 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1042
1043 5. Implemented the functions header_remove(), header_testname(),
1044 header_add_at_position(), and receive_remove_recipient(), and exported them
1045 to local_scan().
1046
1047 6. If an ACL "warn" statement specified the addition of headers, Exim already
1048 inserted X-ACL-Warn: at the start if there was no header name. However, it
1049 was not making this test for the second and subsequent header lines if
1050 there were newlines in the string. This meant that an invalid header could
1051 be inserted if Exim was badly configured.
1052
1053 7. Allow an ACL "warn" statement to add header lines at the start or after all
1054 the Received: headers, as well as at the end.
1055
1056 8. Added the rcpt_4xx retry error code.
1057
1058 9. Added postmaster_mailfrom=xxx to callout verification option.
1059
106010. Added mailfrom=xxxx to the callout verification option, for verify=
1061 header_sender only.
1062
106311. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1064 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1065
106612. Inserted some casts to stop certain compilers warning when using pointer
1067 differences as field lengths or precisions in printf-type calls (mostly
1068 affecting debugging statements).
1069
107013. Added optional readline() support for -be (dynamically loaded).
1071
107214. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1073 same clock tick as a message's arrival, so that its received time was the
1074 same as the "first fail" time on the retry record, and that message
1075 remained on the queue past the ultimate address timeout, every queue runner
1076 would try a delivery (because it was past the ultimate address timeout) but
1077 after another failure, the ultimate address timeout, which should have then
1078 bounced the address, did not kick in. This was a "< instead of <=" error;
1079 in most cases the first failure would have been in the next clock tick
1080 after the received time, and all would be well.
1081
108215. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1083 being recognized when the domain list was tested by the match_domain
1084 condition in an expansion string.
1085
108616. Added the ${str2b64: operator.
1087
108817. Exim was always calling setrlimit() to set a large limit for the number of
1089 processes, without checking whether the existing limit was already
1090 adequate. (It did check for the limit on file descriptors.) Furthermore,
1091 errors from getrlimit() and setrlimit() were being ignored. Now they are
1092 logged to the main and panic logs, but Exim does carry on, to try to do its
1093 job under whatever limits there are.
1094
109518. Imported PCRE 5.0.
1096
109719. Trivial typo in log message " temporarily refused connection" (the leading
1098 space).
1099
110020. If the log selector return_path_on_delivery was set and an address was
1101 redirected to /dev/null, the delivery process crashed because it assumed
1102 that a return path would always be set for a "successful" delivery. In this
1103 case, the whole delivery is bypassed as an optimization, and therefore no
1104 return path is set.
1105
110621. Internal re-arrangement: the function for sending a challenge and reading
1107 a response while authentication was assuming a zero-terminated challenge
1108 string. It's now changed to take a pointer and a length, to allow for
1109 binary data in such strings.
1110
111122. Added the cyrus_sasl authenticator (code supplied by MBM).
1112
111323. Exim was not respecting finduser_retries when seeking the login of the
1114 uid under which it was called; it was always trying 10 times. (The default
1115 setting of finduser_retries is zero.) Also, it was sleeping after the final
1116 failure, which is pointless.
1117
111824. Implemented tls_on_connect_ports.
1119
112025. Implemented acl_smtp_predata.
1121
112226. If the domain in control=submission is set empty, Exim assumes that the
1123 authenticated id is a complete email address when it generates From: or
1124 Sender: header lines.
1125
112627. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1127 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1128 chown and chgrp in /bin and hostname in /usr/bin.
1129
113028. Exim was keeping the "process log" file open after each use, just as it
1131 does for the main log. This opens the possibility of it remaining open for
1132 long periods when the USR1 signal hits a daemon. Occasional processlog
1133 errors were reported, that could have been caused by this. Anyway, it seems
1134 much more sensible not to leave this file open at all, so that is what now
1135 happens.
1136
113729. The long-running daemon process does not normally write to the log once it
1138 has entered its main loop, and it closes the log before doing so. This is
1139 so that log files can straightforwardly be renamed and moved. However,
1140 there are a couple of unusual error situations where the daemon does write
1141 log entries, and I had neglected to close the log afterwards.
1142
114330. The text of an SMTP error response that was received during a remote
1144 delivery was being truncated at 512 bytes. This is too short for some of
1145 the long messages that one sometimes sees. I've increased the limit to
1146 1024.
1147
114831. It is now possible to make retry rules that apply only when a message has a
1149 specific sender, in particular, an empty sender.
1150
115132. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1152 it possible to be selective about when SMTP synchronization is enforced.
1153
115433. Added "control = caseful_local_part" and "control = "caselower_local_part".
1155
115632. Implemented hosts_connection_nolog.
1157
115833. Added an ACL for QUIT.
1159
116034. Setting "delay_warning=" to disable warnings was not working; it gave a
1161 syntax error.
1162
116335. Added mailbox_size and mailbox_filecount to appendfile.
1164
116536. Added control = no_multiline_responses to ACLs.
1166
116737. There was a bug in the logic of the code that waits for the clock to tick
1168 in the case where the clock went backwards by a substantial amount such
1169 that the microsecond fraction of "now" was more than the microsecond
1170 fraction of "then" (but the whole seconds number was less).
1171
117238. Added support for the libradius Radius client library this is found on
1173 FreeBSD (previously only the radiusclient library was supported).
1174
1175
1176Exim version 4.42
1177-----------------
1178
1179 1. When certain lookups returned multiple values in the form name=value, the
1180 quoting of the values was not always being done properly. Specifically:
1181 (a) If the value started with a double quote, but contained no whitespace,
1182 it was not quoted.
1183 (b) If the value contained whitespace other than a space character (i.e.
1184 tabs or newlines or carriage returns) it was not quoted.
1185 This fix has been applied to the mysql and pgsql lookups by writing a
1186 separate quoting function and calling it from the lookup code. The fix
1187 should probably also be applied to nisplus, ibase and oracle lookups, but
1188 since I cannot test any of those, I have not disturbed their existing code.
1189
1190 2. A hit in the callout cache for a specific address caused a log line with no
1191 reason for rejecting RCPT. Now it says "Previous (cached) callout
1192 verification failure".
1193
1194 3. There was an off-by-one bug in the queryprogram router. An over-long
1195 return line was truncated at 256 instead of 255 characters, thereby
1196 overflowing its buffer with the terminating zero. As well as fixing this, I
1197 have increased the buffer size to 1024 (and made a note to document this).
1198
1199 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1200 when Exim is waiting for an SMTP response from a remote server, Exim
1201 restarts its select() call on the socket, thereby resetting its timeout.
1202 This is not a problem when such interrupts are rare. Somebody set up a cron
1203 job to run exiwhat every 2 minutes, which is less than the normal select()
1204 timeout (5 or 10 minutes). This meant that the select() timeout never
1205 kicked in because it was always reset. I have fixed this by comparing the
1206 time when an interrupt arrives with the time at the start of the first call
1207 to select(). If more time than the timeout has elapsed, the interrupt is
1208 treated as a timeout.
1209
1210 5. Some internal re-factoring in preparation for the addition of Sieve
1211 extensions (by MH). In particular, the "personal" test is moved to a
1212 separate function, and given an option for scanning Cc: and Bcc: (which is
1213 not set for Exim filters).
1214
1215 6. When Exim created an email address using the login of the caller as the
1216 local part (e.g. when creating a From: or Sender: header line), it was not
1217 quoting the local part when it contained special characters such as @.
1218
1219 7. Installed new OpenBSD configuration files.
1220
1221 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1222 try to make them clearer.
1223
1224 9. Callout options, other than the timeout value, were being ignored when
1225 verifying sender addresses in header lines. For example, when using
1226
1227 verify = header_sender/callout=no_cache
1228
1229 the cache was (incorrectly) being used.
1230
123110. Added a missing instance of ${EXE} to the exim_install script; this affects
1232 only the Cygwin environment.
1233
123411. When return_path_on_delivery was set as a log selector, if different remote
1235 addresses in the same message used different return paths and parallel
1236 remote delivery occurred, the wrong values would sometimes be logged.
1237 (Whenever a remote delivery process finished, the return path value from
1238 the most recently started remote delivery process was logged.)
1239
124012. RFC 3848 specifies standard names for the "with" phrase in Received: header
1241 lines when AUTH and/or TLS are in use. This is the "received protocol"
1242 field. Exim used to use "asmtp" for authenticated SMTP, without any
1243 indication (in the protocol name) for TLS use. Now it follows the RFC and
1244 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1245 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1246 names appear in log lines as well as in Received: header lines.
1247
124813. Installed MH's patches for Sieve to add the "copy" and "vacation"
1249 extensions, and comparison tests, and to fix some bugs.
1250
125114. Changes to the "personal" filter test:
1252
1253 (1) The test was buggy in that it was just doing the equivalent of
1254 "contains" tests on header lines. For example, if a user's address was
1255 anne@some.where, the "personal" test would incorrectly be true for
1256
1257 To: susanne@some.where
1258
1259 This test is now done by extracting each address from the header in turn,
1260 and checking the entire address. Other tests that are part of "personal"
1261 are now done using regular expressions (for example, to check local parts
1262 of addresses in From: header lines).
1263
1264 (2) The list of non-personal local parts in From: addresses has been
1265 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1266 taken from the Sieve specification recommendations.
1267
1268 (3) If the message contains any header line starting with "List-" it is
1269 treated as non-personal.
1270
1271 (4) The test for "circular" in the Subject: header line has been removed
1272 because it now seems ill-conceived.
1273
127415. Minor typos in src/EDITME comments corrected.
1275
127616. Installed latest exipick from John Jetmore.
1277
127817. If headers_add on a router specified a text string that was too long for
1279 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1280 of string_sprintf() is now avoided.
1281
128218. $message_body_size was not set (it was always zero) when running the DATA
1283 ACL and the local_scan() function.
1284
128519. For the "mail" command in an Exim filter, no default was being set for
1286 the once_repeat time, causing a random time value to be used if "once" was
1287 specified. (If the value happened to be <= 0, no repeat happened.) The
1288 default is now 0s, meaning "never repeat". The "vacation" command was OK
1289 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1290 (I found it when inspecting the code).
1291
129220. There is now an overall timeout for performing a callout verification. It
1293 defaults to 4 times the callout timeout, which applies to individual SMTP
1294 commands during the callout. The overall timeout applies when there is more
1295 than one host that can be tried. The timeout is checked before trying the
1296 next host. This prevents very long delays if there are a large number of
1297 hosts and all are timing out (e.g. when the network connections are timing
1298 out). The value of the overall timeout can be changed by specifying an
1299 additional sub-option for "callout", called "maxwait". For example:
1300
1301 verify = sender/callout=5s,maxwait=20s
1302
130321. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1304 to the end before writing, but this should make it even safer).
1305
130622. Exim was forgetting that it had advertised PIPELINING for the second and
1307 subsequent messages on an SMTP connection. It was also not resetting its
1308 memory on STARTTLS and an internal HELO.
1309
131023. When Exim logs an SMTP synchronization error within a session, it now
1311 records whether PIPELINING has been advertised or not.
1312
131324. Added 3 instances of "(long int)" casts to time_t variables that were being
1314 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1315 rather than long int.
1316
131725. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1318
131926. Added the never_mail option to autoreply.
1320
1321
1322Exim version 4.41
1323-----------------
1324
1325 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1326 crash if the getsockname() call failed; this can happen if a connection is
1327 closed very soon after it is established. The problem was simply in the
1328 order in which certain operations were done, causing Exim to try to write
1329 to the SMTP stream before it had set up the file descriptor. The bug has
1330 been fixed by making things happen in the correct order.
1331
1332
1333Exim version 4.40
1334-----------------
1335
1336 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1337 before the connection was closed, thus losing the rejection response.
1338
1339 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1340 some early Solaris releases, but causes trouble in current releases where
1341 socklen_t is defined.
1342
1343 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1344 always exist.
1345
1346 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1347 configured.
1348
1349 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1350 that releasing the top of it at the end releases what was used for sub-
1351 expansions (unless the block got too big). However, discard this block if
1352 the first thing is a variable or header, so that we can use its block when
1353 it is dynamic (useful for very large $message_headers, for example).
1354
1355 6. Lookups now cache *every* query, not just the most recent. A new, separate
1356 store pool is used for this. It can be recovered when all lookup caches are
1357 flushed. Lookups now release memory at the end of their result strings.
1358 This has involved some general refactoring of the lookup sources.
1359
1360 7. Some code has been added to the store_xxx() functions to reduce the amount
1361 of flapping under certain conditions.
1362
1363 8. log_incoming_interface used to affect only the <= reception log lines. Now
1364 it causes the local interface and port to be added to several more SMTP log
1365 lines, for example "SMTP connection from", and rejection lines.
1366
1367 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1368
136910. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1370
137111. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1372 could be overwritten at the end of the current message (or the start of a
1373 new message if it was set in a HELO ACL). The value is now preserved for
1374 the duration of the SMTP connection.
1375
137612. If a transport had a headers_rewrite setting, and a matching header line
1377 contained an unqualified address, that address was qualified, even if it
1378 did not match any rewriting rules. The underlying bug was that the values
1379 of the flags that permit the existence of unqualified sender and recipient
1380 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1381 non-local messages, and by -bnq for local messages) were not being
1382 preserved with the message after it was received.
1383
138413. When Exim was logging an SMTP synchronization error, it could sometimes log
1385 "next input=" as part of the text comprising the host identity instead of
1386 the correct text. The code was using the same buffer for two different
1387 strings. However, depending on which order the printing function evaluated
1388 its arguments, the bug did not always show up. Under Linux, for example, my
1389 test suite worked just fine.
1390
139114. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1392 doesn't work with some older versions of Perl. It has been changed to "my",
1393 which in any case is probably the better facility to use.
1394
139515. A really picky compiler found some instances of statements for creating
1396 error messages that either had too many or two few arguments for the format
1397 string.
1398
139916. The size of the buffer for calls to the DNS resolver has been increased
1400 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1401 for addresses that have a lot of PTR records. This alleviates a problem; it
1402 does not fully solve it.
1403
140417. A dnsdb lookup for PTR records that receives more data than will fit in the
1405 buffer now truncates the list and logs the incident, which is the same
1406 action as happens when Exim is looking up a host name and its aliases.
1407 Previously in this situation something unpredictable would happen;
1408 sometimes it was "internal error: store_reset failed".
1409
141018. If a server dropped the connection unexpectedly when an Exim client was
1411 using GnuTLS and trying to read a response, the client delivery process
1412 crashed while trying to generate an error log message.
1413
141419. If a "warn" verb in an ACL added multiple headers to a message in a single
1415 string, for example:
1416
1417 warn message = H1: something\nH2: something
1418
1419 the text was added as a single header line from Exim's point of view
1420 though it ended up OK in the delivered message. However, searching for the
1421 second and subsequent header lines using $h_h2: did not work. This has been
1422 fixed. Similarly, if a system filter added multiple headers in this way,
1423 the routers could not see them.
1424
142520. Expanded the error message when iplsearch is called with an invalid key to
1426 suggest using net-iplsearch in a host list.
1427
142821. When running tests using -bh, any delays imposed by "delay" modifiers in
1429 ACLs are no longer actually imposed (and a message to that effect is
1430 output).
1431
143222. If a "gecos" field in a passwd entry contained escaped characters, in
1433 particular, if it contained a \" sequence, Exim got it wrong when building
1434 a From: or a Sender: header from that name. A second bug also caused
1435 incorrect handling when an unquoted " was present following a character
1436 that needed quoting.
1437
143823. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1439 was not being matched caselessly.
1440
144124. Arranged for all hyphens in the exim.8 source to be escaped with
1442 backslashes.
1443
144425. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1445 properly. Recipient callout cache records were still being keyed to include
1446 the sender, even when use_sender was set false. This led to far more
1447 callouts that were necessary. The sender is no longer included in the key
1448 when use_sender is false.
1449
145026. Added "control = submission" modifier to ACLs.
1451
145227. Added the ${base62d: operator to decode base 62 numbers.
1453
145428. dnsdb lookups can now access SRV records.
1455
145629. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1457 the configuration file.
1458
145930. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1460 (-v) mode. This makes the output for a verbose queue run more intelligible.
1461
146231. Added a use_postmaster feature to recipient callouts.
1463
146432. Added the $body_zerocount variable, containing the number of binary zero
1465 bytes in the message body.
1466
146733. The time of last modification of the "new" subdirectory is now used as the
1468 "mailbox time last read" when there is a quota error for a maildir
1469 delivery.
1470
147134. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1472
147335. Added +ignore_unknown as a special item in host lists.
1474
147536. Code for decoding IPv6 addresses in host lists is now included, even if
1476 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1477 address was recognized as an IP address, but was then not correctly decoded
1478 into binary, causing unexpected and incorrect effects when compared with
1479 another IP address.
1480
1481
1482Exim version 4.34
1483-----------------
1484
1485 1. Very minor rewording of debugging text in manualroute to say "list of
1486 hosts" instead of "hostlist".
1487
1488 2. If verify=header_syntax was set, and a header line with an unqualified
1489 address (no domain) and a large number of spaces between the end of the
1490 name and the colon was received, the reception process suffered a buffer
1491 overflow, and (when I tested it) crashed. This was caused by some obsolete
1492 code that should have been removed. The fix is to remove it!
1493
1494 3. When running in the test harness, delay a bit after writing a bounce
1495 message to get a bit more predictability in the log output.
1496
1497 4. Added a call to search_tidyup() just before forking a reception process. In
1498 theory, someone could use a lookup in the expansion of smtp_accept_max_
1499 per_host which, without the tidyup, could leave open a database connection.
1500
1501 5. Added the variables $recipient_data and $sender_data which get set from a
1502 lookup success in an ACL "recipients" or "senders" condition, or a router
1503 "senders" option, similar to $domain_data and $local_part_data.
1504
1505 6. Moved the writing of debug_print from before to after the "senders" test
1506 for routers.
1507
1508 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1509 problems for message scanning, either using a data ACL, or using
1510 local_scan() because the Received: header was not generated till after they
1511 were called (in order to set the time as the time of reception completion).
1512 I have revised the way this works. The header is now generated after the
1513 body is received, but before the ACL or local_scan() are called. After they
1514 are run, the timestamp in the header is updated.
1515
1516
1517Exim version 4.33
1518-----------------
1519
1520 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1521 before starting a queue runner without re-exec. This happened only when
1522 deliver_drop_privilege was set or when the Exim user was set to root. The
1523 effect of the bug was that timeouts during subsequent deliveries caused
1524 crashes instead of being properly handled. The handler is now left at its
1525 default (and expected) setting.
1526
1527 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1528 message, again when deliver_drop_privilege is set or Exim is run as root.
1529 The bug described in (1) was not present in this case, but the tidying up
1530 of the other signals was missing. I have made the two cases consistent.
1531
1532 3. The ignore_target_hosts setting on a manualroute router was being ignored
1533 for hosts that were looked up using the /MX notation.
1534
1535 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1536 in domain lists.
1537
1538 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1539 operated on the sender address. After changing the $sender_address to <>
1540 for the sender address verify, Exim was re-instated it as the original
1541 (before rewriting) address, but remembering that it had rewritten it, so it
1542 wasn't rewriting it again. This bug also had the effect of breaking the
1543 sender address verification caching when the sender address was rewritten.
1544
1545 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1546 This has been changed so that if the ip literal address matches
1547 ignore_target_hosts, the router declines.
1548
1549 7. Added expansion conditions match_domain, match_address, and match_local_
1550 part (NOT match_host).
1551
1552 8. The placeholder for the Received: header didn't have a length field set.
1553
1554 9. Added code to Exim itself and to exim_lock to test for a specific race
1555 condition that could lead to file corruption when using MBX delivery. The
1556 issue is with the lockfile that is created in /tmp. If this file is removed
1557 after a process has opened it but before that process has acquired a lock,
1558 there is the potential for a second process to recreate the file and also
1559 acquire a lock. This could lead to two Exim processes writing to the file
1560 at the same time. The added code performs the same test as UW imapd; it
1561 checks after acquiring the lock that its file descriptor still refers to
1562 the same named file.
1563
156410. The buffer for building added header lines was of fixed size, 8192 bytes.
1565 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1566 when Exim is built.
1567
156811. Added the smtp_active_hostname option. If used, this will typically be made
1569 to depend on the incoming interface address. Because $interface_address is
1570 not set up until the daemon has forked a reception process, error responses
1571 that can happen earlier (such as "too many connections") no longer contain
1572 a host name.
1573
157412. If an expansion in a condition on a "warn" statement fails because a lookup
1575 defers, the "warn" statement is abandoned, and the next ACL statement is
1576 processed. Previously this caused the whole ACL to be aborted.
1577
157813. Added the iplsearch lookup type.
1579
158014. Added ident_timeout as a log selector.
1581
158215. Added tls_certificate_verified as a log selector.
1583
158416. Added a global option tls_require_ciphers (compare the smtp transport
1585 option of the same name). This controls incoming TLS connections.
1586
158717. I finally figured out how to make tls_require_ciphers do a similar thing
1588 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1589 before starting the TLS session.
1590
159118. Tabs are now shown as \t in -bP output.
1592
159319. If the log selector return_path_on_delivery was set, Exim crashed when
1594 bouncing a message because it had too many Received: header lines.
1595
159620. If two routers both had headers_remove settings, and the first one included
1597 a superfluous trailing colon, the final name in the first list and the
1598 first name in the second list were incorrectly joined into one item (with a
1599 colon in the middle).
1600
1601
1602Exim version 4.32
1603-----------------
1604
1605 1. Added -C and -D options to the exinext utility, mainly to make it easier
1606 to include in the automated testing, but these could be helpful when
1607 multiple configurations are in use.
1608
1609 2. The exinext utility was not formatting the output nicely when there was
1610 an alternate port involved in the retry record key, nor when there was a
1611 message id as well (for retries that were specific to a specific message
1612 and a specific host). It was also confused by IPv6 addresses, because of
1613 the additional colons they contain. I have fixed the IPv4 problem, and
1614 patched it up to do a reasonable job for IPv6.
1615
1616 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1617 delivery, the log line now contains "pipelined" if PIPELINING was used.
1618
1619 4. An SMTP transport process used to panic and die if the bind() call to set
1620 an explicit outgoing interface failed. This has been changed; it is now
1621 treated in the same way as a connect() failure.
1622
1623 5. A reference to $sender_host_name in the part of a conditional expansion
1624 that was being skipped was still causing a DNS lookup. This no longer
1625 occurs.
1626
1627 6. The def: expansion condition was not recognizing references to header lines
1628 that used bh_ and bheader_.
1629
1630 7. Added the _cache feature to named lists.
1631
1632 8. The code for checking quota_filecount in the appendfile transport was
1633 allowing one more file than it should have been.
1634
1635 9. For compatibility with Sendmail, the command line option
1636
1637 -prval:sval
1638
1639 is equivalent to
1640
1641 -oMr rval -oMs sval
1642
1643 and sets the incoming protocol and host name (for trusted callers). The
1644 host name and its colon can be omitted when only the protocol is to be set.
1645 Note the Exim already has two private options, -pd and -ps, that refer to
1646 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1647 "s", but I don't think that's a major issue.
1648
164910. A number of refactoring changes to the code, none of which should affect
1650 Exim's behaviour:
1651
1652 (a) The number of logging options was getting close to filling up the
1653 32-bit word that was used as a bit map. I have split them into two classes:
1654 those that are passed in the argument to log_write(), and those that are
1655 only ever tested independently outside of that function. These are now in
1656 separate 32-bit words, so there is plenty of room for expansion again.
1657 There is no change in the user interface or the logging behaviour.
1658
1659 (b) When building, for example, log lines, the code previously used a
1660 macro that called string_cat() twice, in order to add two strings. This is
1661 not really sufficiently general. Furthermore, there was one instance where
1662 it was actually wrong because one of the argument was used twice, and in
1663 one call a function was used. (As it happened, calling the function twice
1664 did not affect the overall behaviour.) The macro has been replaced by a
1665 function that can join an arbitrary number of extra strings onto a growing
1666 string.
1667
1668 (c) The code for expansion conditions now uses a table and a binary chop
1669 instead of a serial search (which was left over from when there were very
1670 few conditions). Also, it now recognizes conditions like "pam" even when
1671 the relevant support is not compiled in: a suitably worded error message is
1672 given if an attempt is made to use such a condition.
1673
167411. Added ${time_interval:xxxxx}.
1675
167612. A bug was causing one of the ddress fields not to be passed back correctly
1677 from remote delivery subprocesses. The field in question was not being
1678 subsequently used, so this caused to problems in practice.
1679
168013. Added new log selectors queue_time and deliver_time.
1681
168214. Might have fixed a bug in maildirsizefile handling that threw up
1683 "unexpected character" debug warnings, and recalculated the data
1684 unnecessarily. In any case, I expanded the warning message to give more
1685 information.
1686
168715. Added the message "Restricted characters in address" to the statements in
1688 the default ACL that block characters like @ and % in local parts.
1689
169016. Change 71 for release 4.31 proved to be much less benign that I imagined.
1691 Three changes have been made:
1692
1693 (a) There was a serious bug; a negative response to MAIL caused the whole
1694 recipient domain to be cached as invalid, thereby blocking all messages
1695 to all local parts at the same domain, from all senders. This bug has
1696 been fixed. The domain is no longer cached after a negative response to
1697 MAIL if the sender used is not empty.
1698
1699 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1700 been restored.
1701
1702 (c) A new callout option, "use_sender" has been added for people who want
1703 the modified behaviour.
1704
1705
1706Exim version 4.31
1707-----------------
1708
1709 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1710 Larry Rosenman.
1711
1712 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1713 indeed breaks things for older releases.
1714
1715 3. Added additional logging to the case where there is a problem reading data
1716 from a filter that is running in a subprocess using a pipe, in order to
1717 try to track down a specific problem.
1718
1719 4. Testing facility fudge: when running in the test harness and attempting
1720 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1721 getting "No route to host". Convert this to a timeout.
1722
1723 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1724 warning.
1725
1726 6. Some OS don't have socklen_t but use size_t instead. This affects the
1727 fifth argument of getsockopt() amongst other things. This is now
1728 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1729 can be set for individual OS. I have set it for SunOS5, OSF1, and
1730 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1731 some earlier ones do not.
1732
1733 7. Change 4.30/15 was not doing the test caselessly.
1734
1735 8. The standard form for an IPv6 address literal was being rejected by address
1736 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1737 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1738 this, as well as the form without the "IPv6" on the front (but only when
1739 address literals are enabled, of course).
1740
1741 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1742
174310. Exim crashed if a message with an empty sender address specified by -f
1744 encountered a router with an errors_to setting. This could be provoked only
1745 by a command such as
1746
1747 exim -f "" ...
1748
1749 where an empty string was supplied; "<>" did not hit this bug.
1750
175111. Installed PCRE release 4.5.
1752
175312. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1754 remained set. It is now erased.
1755
175613. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1757 times from message ids (which are base 36 rather than the normal 62).
1758
175914. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1760 were being counted as actual protocol errors, and logged if the log
1761 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1762 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1763 and DATA following a set of rejected RCPTs do not count as protocol errors.
1764 In other words, Exim assumes they were pipelined, though this may not
1765 actually be the case. Of course, in all cases the client gets an
1766 appropriate error code.
1767
176815. If a lookup fails in an ACL condition, a message about the failure may
1769 be available; it is used if testing the ACL cannot continue, because most
1770 such messages specify what the cause of the deferral is. However, some
1771 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1772 that caused an old message to be retained and used if a later statement
1773 caused a defer, replacing the real cause of the deferral.
1774
177516. If an IP address had so many PTR records that the DNS lookup buffer
1776 was not large enough to hold them, Exim could crash while trying to process
1777 the truncated data. It now detects and logs this case.
1778
177917. Further to 4.21/58, another change has been made: if (and only if) the
1780 first line of a message (the first header line) ends with CRLF, a bare LF
1781 in a subsequent header line has a space inserted after it, so as not to
1782 terminate the header.
1783
178418. Refactoring: tidied an ugly bit of code in appendfile that copied data
1785 unnecessarily, used atoi() instead of strtol(), and didn't check the
1786 termination when getting file sizes from file names by regex.
1787
178819. Completely re-implemented the support for maildirsize files, in the light
1789 of a number of problems with the previous contributed implementation
1790 (4.30/29). In particular:
1791
1792 . If the quota is zero, the maildirsize file is maintained, but no quota is
1793 imposed.
1794
1795 . If the maildir directory does not exist, it is created before any attempt
1796 to write a maildirsize file.
1797
1798 . The quota value in the file is just a cache; if the quota is changed in
1799 the transport, the new value overrides.
1800
1801 . A regular expression is available for excluding directories from the
1802 count.
1803
180420. The autoreply transport checks the characters in options that define the
1805 message's headers; it allows continued headers, but it was checking with
1806 isspace() after an embedded newline instead of explicitly looking for a
1807 space or a tab.
1808
180921. If all the "regular" hosts to which an address was routed had passed their
1810 expiry times, and had not reached their retry times, the address was
1811 bounced, even if fallback hosts were defined. Now Exim should go on to try
1812 the fallback hosts.
1813
181422. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1815 equivalent code in the SMTP transport. Some hosts send humungous responses
1816 to HELO/EHLO, more than 1024 it seems.
1817
181823. Refactoring: code in filter.c used (void *) for "any old type" but this
1819 gives compiler warnings in some environments. I've now done it "properly",
1820 using a union.
1821
182224. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1823 (because of problems with the built-in one) was declared to return uschar *
1824 instead of char *, causing compiler failure.
1825
182625. Fixed a file descriptor leak when processing alias/forward files.
1827
182826. Fixed a minor format string issue in dbfn.c.
1829
183027. Typo in exim.c: ("dmbnz" for "dbmnz").
1831
183228. If a filter file refered to $h_xxx or $message_headers, and the headers
1833 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1834 become corrupted.
1835
183629. When a sender address is verified, it is cached, to save repeating the test
1837 when there is more than one recipient in a message. However, when the
1838 verification involves a callout, it is possible for different callout
1839 options to be set for different recipients. It is too complicated to keep
1840 track of this in the cache, so now Exim always runs a verification when a
1841 callout is required, relying on the callout cache for the optimization.
1842 The overhead is duplication of the address routing, but this should not be
1843 too great.
1844
184530. Fixed a bug in callout caching. If a RCPT command caused the sender address
1846 to be verified with callout=postmaster, and the main callout worked but the
1847 postmaster check failed, the verification correctly failed. However, if a
1848 subsequent RCPT command asked for sender verification *without* the
1849 postmaster check, incorrect caching caused this verification also to fail,
1850 incorrectly.
1851
185231. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1853 it was not caching the DNS options (qualify_single, search_parents) that
1854 were used when the lookup failed. A subsequent lookup with different
1855 options therefore always gave the same answer, though there were cases
1856 where it should not have. (Example: a "domains = !$mx_any" option on a
1857 dnslookup router: the "domains" option is always processed without any
1858 widening, but the router might have qualify_single set.) Now Exim uses the
1859 cached value only when the same options are set.
1860
186132. Added John Jetmore's "exipick" utility to the distribution.
1862
186333. GnuTLS: When an attempt to start a TLS session fails for any reason other
1864 than a timeout (e.g. a certificate is required, and is not provided), an
1865 Exim server now closes the connection immediately. Previously it waited for
1866 the client to close - but if the client is SSL, it seems that they each
1867 wait for each other, leading to a delay before one of them times out.
1868
186934: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1870 maintained 0.8.x compatibility because I don't think many are using it, and
1871 it is clearly obsolete.
1872
187335. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1874 transport.
1875
187636. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1877 client certificate was expired. A simple patch fixes this, though I don't
1878 understand the full logic of why the verify callback is called multiple
1879 times.
1880
188137. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1882 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1883 which causes problems with some clients (such as the Certicom SSL Plus
1884 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1885 disables the coutermeasure allowing Eudora to connect."
1886
188738. Exim was not checking that a write() to a log file succeeded. This could
1888 lead to Bad Things if a log got too big, in particular if it hit a file
1889 size limit. Exim now panics and dies if it cannot write to a log file, just
1890 as it does if it cannot open a log file.
1891
189239. Modified OS/Makefile-Linux so that it now contains
1893
1894 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1895
1896 The two -D definitions ensure that Exim is compiled with large file
1897 support, which makes it possible to handle log files that are bigger than
1898 2^31.
1899
190040. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1901 instance) a domain was checked against a named list that involved a lookup,
1902 causing $domain_data to be set, then another domain was checked against the
1903 same list, then the first domain was re-checked, the value of $domain_data
1904 after the final check could be wrong. In particular, if the second check
1905 failed, it could be set empty. This bug probably also applied to
1906 $localpart_data.
1907
190841. The strip_trailing_dot option was not being applied to the address given
1909 with the -f command-line option.
1910
191142. The code for reading a message's header from the spool was incrementing
1912 $received_count, but never initializing it. This meant that the value was
1913 incorrect (doubled) while delivering a message in the same process in which
1914 it was received. In the most common configuration of Exim, this never
1915 happens - a fresh exec is done - but it can happen when
1916 deliver_drop_privilege is set.
1917
191843. When Exim logs an SMTP synchronization error - client data sent too soon -
1919 it now includes up to 150 characters of the unexpected data in the log
1920 line.
1921
192244. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1923 and building data strings. The size of both of these buffers was 10 000
1924 bytes - far larger than anybody would *ever* want, thought I. Needless to
1925 say, somebody hit the limit. I have increased the maximum line length to
1926 20 000 and the maximum data length of concatenated lines to 100 000. I have
1927 also fixed two bugs, because there was no checking on these buffers. Tsk,
1928 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1929 buffer is too small.
1930
193145. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1932 lsearch lookups. Now it does.
1933
193446. When parsing a route_list item in a manualroute router, a fixed-length
1935 buffer was used for the list of hosts. I made this 1024 bytes long,
1936 thinking that nobody would ever have a list of hosts that long. Wrong.
1937 Somebody had a whole pile of complicated expansion conditions, and the
1938 string was silently truncated, leading to an expansion error. It turns out
1939 that it is easier to change to an unlimited length (owing to other changes
1940 that have happened since this code was originally written) than to build
1941 structure for giving a limitation error. The length of the item that
1942 expands into the list of hosts is now unlimited.
1943
194447. The lsearch lookup could not handle data where the length of text line was
1945 more than 4095 characters. Such lines were truncated, leading to shortened
1946 data being returned. It should now handle lines of any length.
1947
194848. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1949 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1950 ACL").
1951
195249. Cosmetic tidy to scripts like exicyclog that are generated by globally
1953 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1954 no longer happens in comment lines. A list of replacements is now placed
1955 at the head of all of the source files, except those whose only change is
1956 to replace PERL_COMMAND in the very first #! line.
1957
195850. Replaced the slow insertion sort in queue.c, for sorting the list of
1959 messages on the queue, with a bottom-up merge sort, using code contributed
1960 by Michael Haardt. This should make operations like -bp somewhat faster on
1961 large queues. It won't affect queue runners, except when queue_run_in_order
1962 is set.
1963
196451. Installed eximstats 1.31 in the distribution.
1965
196652. Added support for SRV lookups to the dnslookup router.
1967
196853. If an ACL referred to $message_body or $message_body_end, the value was not
1969 reset for any messages that followed in the same SMTP session.
1970
197154. The store-handling optimization for building very long strings was not
1972 differentiating between the different store pools. I don't think this
1973 actually made any difference in practice, but I've tidied it.
1974
197555. While running the routers to verify a sender address, $sender_address
1976 was still set to the sender address. This is wrong, because when routing to
1977 send a bounce to the sender, it would be empty. Therefore, I have changed
1978 it so that, while verifying a sender address, $sender_address is set to <>.
1979 (There is no change to what happens when verifying a recipient address.)
1980
198156. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1982 target A or AAAA records (if not already returned) without resetting the
1983 qualify_single or search_parents options of the DNS resolver. These are
1984 inappropriate in this case because the targets of MX and SRV records must
1985 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1986 target that, when qualified, matched something in the local domain. These
1987 two options are now turned off when doing these lookups.
1988
198957. It seems that at least some releases of Reiserfs (which does not have the
1990 concept of a fixed number of inodes) returns zero and not -1 for the
1991 number of available inodes. This interacted badly with check_spool_inodes,
1992 which assumed that -1 was the "no such thing" setting. What I have done is
1993 to check that the total number of inodes is greater than zero before doing
1994 the test of how many are available.
1995
199658. When a "warn" ACL statement has a log_message modifier, the message is
1997 remembered, and not repeated. This is to avoid a lot of repetition when a
1998 message has many recipients that cause the same warning to be written.
1999 Howewer, Exim was preserving the list of already written lines for an
2000 entire SMTP session, which doesn't seem right. The memory is now reset if a
2001 new message is started.
2002
200359. The "rewrite" debugging flag was not showing the result of rewriting in the
2004 debugging output unless log_rewrite was also set.
2005
200660. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2007 of (int)(handle) when we know that handle contains (void *)(-1).
2008
200961. The Exim daemon panic-logs an error return when it closes the incoming
2010 connection. However "connection reset by peer" seems to be common, and
2011 isn't really an error worthy of noting specially, so that particular error
2012 is no long logged.
2013
201462. When Exim is trying to find all the local interfaces, it used to panic and
2015 die if the ioctl to get the interface flags failed. However, it seems that
2016 on at least one OS (Solaris 9) it is possible to have an interface that is
2017 included in the list of interfaces, but for which you get a failure error
2018 for this call. This happens when the interface is not "plumbed" into a
2019 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2020 failure of the "get flags" call assumes that the interface is down.
2021
202263. Added a ${eval10: operator, which assumes all numbers are decimal. This
2023 makes life easier for people who are doing arithmetic on fields extracted
2024 from dates, where you often get leading zeros that should not be
2025 interpreted as octal.
2026
202764. Added qualify_domain to the redirect router, to override the global
2028 setting.
2029
203065. If a pathologically long header line contained very many addresses (the
2031 report of this problem mentioned 10 000) and each of them was rewritten,
2032 Exim could use up a very large amount of memory. (It kept on making new
2033 copies of the header line as it rewrote, and never released the old ones.)
2034 At the expense of a bit more processing, the header rewriting function has
2035 been changed so that it no longer eats memory in this way.
2036
203766. The generation of the Received: header has been moved from the time that a
2038 message starts to be received, to the time that it finishes. The timestamp
2039 in the Received: header should now be very close to that of the <= log
2040 line. There are two side-effects of this change:
2041
2042 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2043 logged header lines no longer include the local Received: line, because
2044 it has not yet been created. The same applies to a copy of the message
2045 that is returned to a non-SMTP sender when a message is rejected.
2046
2047 (b) When a filter file is tested using -bf, no additional Received: header
2048 is added to the test message. After some thought, I decided that this
2049 is a bug fix.
2050
2051 This change does not affect the value of $received_for. It is still set
2052 after address rewriting, but before local_scan() is called.
2053
205467. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2055
205668. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2057 gave an unhelpful panic error message, and a defer error. I have managed to
2058 change this behaviour so that it now rejects any supplied certificate,
2059 which seems right, as the list of acceptable certificates is empty.
2060
206169. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2062 gave an unhelpful defer error. I have not managed to make this reject any
2063 supplied certificates, but the error message it gives is "no certificate
2064 supplied", which is not helpful.
2065
206670. exigrep's output now also includes lines that are not associated with any
2067 message, but which match the given pattern. Implemented by a patch from
2068 Martin Sluka, which also tidied up the Perl a bit.
2069
207071. Recipient callout verification, like sender verification, was using <> in
2071 the MAIL FROM command. This isn't really the right thing, since the actual
2072 sender may affect whether the remote host accepts the recipient or not. I
2073 have changed it to use the actual sender in the callout; this means that
2074 the cache record is now keyed on a recipient/sender pair, not just the
2075 recipient address. There doesn't seem to be a real danger of callout loops,
2076 since a callout by the remote host to check the sender would use <>.
2077 [SEE ABOVE: changed after hitting problems.]
2078
207972. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2080 temporary errors. However, in the case of such a code being given after
2081 the end of a data transmission (i.e. after ".") Exim was failing to write
2082 a retry record for the message. (Yes, there was some broken host that was
2083 actually sending 8xx at this point.)
2084
208573. An unknown lookup type in a host list could cause Exim to panic-die when
2086 the list was checked. (An example that provoked this was putting <; in the
2087 middle of a list instead of at the start.) If this happened during a DATA
2088 ACL check, a -D file could be left lying around. This kind of configuration
2089 error no longer causes Exim to die; instead it causes a defer errror. The
2090 incident is still logged to the main and panic logs.
2091
209274. Buglet left over from Exim 3 conversion. The message "too many messages
2093 in one connection" was written to the rejectlog but not the mainlog, except
2094 when address rewriting (yes!) was being logged.
2095
209675. Added write_rejectlog option.
2097
209876. When a system filter was run not as root (that is, when system_filter_user
2099 was set), the values of the $n variables were not being returned to the
2100 main process; thus, they were not subsequently available in the $sn
2101 variables.
2102
210377. Added +return_path_on_delivery log selector.
2104
210578. A connection timeout was being treated differently from recipients deferred
2106 when testing hosts_max_try with a message that was older than the host's
2107 retry timeout. (The host should not be counted, thus allowing all hosts to
2108 be tried at least once before bouncing.) This may have been the cause of an
2109 occasionally reported bug whereby a message would remain on the queue
2110 longer than the retry timeout, but would be bounced if a delivery was
2111 forced. I say "may" because I never totally pinned down the problem;
2112 setting up timeout/retry tests is difficult. See also the next item.
2113
211479. The ultimate address timeout was not being applied to errors that involved
2115 a combination of host plus message (for example, a timeout on a MAIL
2116 command). When an address resolved to a number of possible hosts, and they
2117 were not all tried for each delivery (e.g. because of hosts_max_try), a
2118 message could remain on the queue longer than the retry timeout.
2119
212080. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2121 Haardt.
2122
212381. Fixed an obscure SMTP outgoing bug which required at least the following
2124 conditions: (a) there was another message waiting for the same server;
2125 (b) the server returned 5xx to all RCPT commands in the first message so
2126 that the message was not completed; (c) the server dropped the connection
2127 or gave a negative response to the RSET that Exim sends to abort the
2128 transaction. The observed case was a dropped connection after DATA that had
2129 been sent in pipelining mode. That is, the server had advertised PIPELINING
2130 but was not implementing it correctly. The effect of the bug was incorrect
2131 behaviour, such as trying another host, and this could lead to a crash.
2132
2133
2134Exim version 4.30
2135-----------------
2136
2137 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2138 and daemon.c were passed as pointers to ints; they should have been
2139 pointers to socklen_t variables (which are typically unsigned ints).
2140
2141 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2142 fixed.
2143
2144 3. Fixed a really odd bug that affected only the testing scheme; patching a
2145 certain fixed string in the binary changed the value of another string that
2146 happened to be identical to the end of the original first string.
2147
2148 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2149 name", it returns that address as the IP address. On some operating
2150 systems (e.g. Solaris), it also passes back the IP address string as the
2151 "host name". However, on others (e.g. Linux), it passes back an empty
2152 string. Exim wasn't checking for this, and was changing the host name to an
2153 empty string, assuming it had been canonicized.
2154
2155 5. Although rare, it is permitted to have more than one PTR record for a given
2156 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2157 all the names associated with an address, because they do in Solaris.
2158 However, it seems that they do not in Linux for data that comes from the
2159 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2160 I found this out when I moved to a new Linux workstation and tried to run
2161 the Exim test suite.
2162
2163 To get round this problem I have changed the code so that it now does its
2164 own call to the DNS to look up PTR records when searching for a host name.
2165 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2166 addresses that are only in /etc/hosts are still found.
2167
2168 This behaviour is, however, controlled by an option called host_lookup_
2169 order, which defaults to "bydns:byaddr". If people want to use the other
2170 order, or indeed, just use one or the other means of lookup, they can
2171 specify it in this variable.
2172
2173 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2174 some operating systems, this comes back from gethostbyaddr() as an empty
2175 string, and this is what Exim used to test for. However, it seems that in
2176 other systems, "." is yielded. Exim now tests for this case too.
2177
2178 7. The values of check_spool_space and check_log_space are now held internally
2179 as a number of kilobytes instead of an absolute number of bytes. If a
2180 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2181 kilobyte. This means that much larger values can be stored.
2182
2183 8. Exim monitor: an attempt to get the action menu when not actually pointing
2184 at a message produces an empty menu entitled "No message selected". This
2185 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2186 no entries in it ("Shell widget menu has zero width and/or height"). So I
2187 have added a single, blank menu entry in this case.
2188
2189 9. Added ${quote_local_part.
2190
219110. MIME decoding is now applied to the contents of Subject: header lines when
2192 they are logged.
2193
219411. Now that a reference to $sender_host_address automatically causes a reverse
2195 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2196 host lookup before query-style lookups in lists that might use this
2197 variable. This has therefore been abolished, and the "net-" prefix is no
2198 longer necessary for query-style lookups.
2199
220012. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2201 have been a typo for LFLAGS, so it has been changed.
2202
220313. The install script calls Exim with "-C /dev/null" in order to find the
2204 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2205 to be output. Howeve, since Exim outputs its version number before the
2206 error, it didn't break the script. It just looked ugly. I fixed this by
2207 always allowing "-C /dev/null" if the caller is root.
2208
220914. Ignore overlarge ACL variable number when reading spool file - insurance
2210 against a later release with more variables having written the file.
2211
221215. The standard form for an IPv6 address literal was being rejected by EHLO.
2213 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2214 this, as well as the form without the "IPv6" on the front.
2215
221616. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2217 OS/Makefile-Darwin file.
2218
221917. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2220 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2221
222218. After change 4.21/52, "%ld" was used to format the contents of the $inode
2223 variable. However, some OS use ints for inodes. I've added cast to long int
2224 to get rid of the compiler warning.
2225
222619. I had forgotten to lock out "/../" in configuration file names when
2227 ALT_CONFIG_PREFIX was set.
2228
222920. Routers used for verification do not need to specify transports. However,
2230 if such a router generated a host list, and callout was configured, Exim
2231 crashed, because it could not find a port number from the (non-existent)
2232 transport. It now assumes port 25 in this circumstance.
2233
223421. Added the -t option to exigrep.
2235
223622. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2237 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2238 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2239 them is set, LOOKUP_LSEARCH is forced.
2240
224123. "exim -bV" now outputs a list of lookups that are included in the binary.
2242
224324. Added sender and host information to the "rejected by local_scan()" log
2244 line; previously there was no indication of these.
2245
224625. Added .include_if_exists.
2247
224826. Change 3.952/11 added an explicit directory sync on top of a file sync for
2249 Linux. It turns out that not all file systems support this. Apparently some
2250 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2251 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2252 not supported on the file descriptor, is now ignored when Exim is trying to
2253 sync a directory. This applies only to Linux.
2254
225527. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2256
225728. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2258 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2259 this.
2260
226129. Added support for maildirsize files from supplied patch (modified a bit).
2262
226330. The use of :fail: followed by an empty string could lead Exim to respond to
2264 sender verification failures with (e.g.):
2265
2266 550 Verification failed for <xxx>
2267 550 Sender verify failed
2268
2269 where the first response line was missing the '-' that indicates it is not
2270 the final line of the response.
2271
227231. The loop for finding the name of the user that called Exim had a hardwired
2273 limit of 10; it now uses the value of finduser_retries, which is used for
2274 all other user lookups.
2275
227632. Added $received_count variable, available in data and not_smtp ACLs, and at
2277 delivery time.
2278
227933. Exim was neglecting to zero errno before one call of strtol() when
2280 expanding a string and expecting an integer value. On some systems this
2281 resulted in spurious "integer overflow" errors. Also, it was casting the
2282 result into an int without checking.
2283
228434. Testing for a connection timeout using "timeout_connect" in the retry rules
2285 did not work. The code looks as if it has *never* worked, though it appears
2286 to have been documented since at least releast 1.62. I have made it work.
2287
228835. The "timeout_DNS" error in retry rules, also documented since at least
2289 1.62, also never worked. As it isn't clear exactly what this means, and
2290 clearly it isn't a major issue, I have abolished the feature by treating it
2291 as "timeout", and writing a warning to the main and panic logs.
2292
229336. The display of retry rules for -brt wasn't always showing the error code
2294 correctly.
2295
229637. Added new error conditions to retry rules: timeout_A, timeout_MX,
2297 timeout_connect_A, timeout_connect_MX.
2298
229938. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2300 to the empty sender.
2301
230239. The daemon was not analysing the content of -oX till after it had closed
2303 stderr and disconnected from the controlling terminal. This meant that any
2304 syntax errors were only noted on the panic log, and the return code from
2305 the command was 0. By re-arranging the code a little, I've made the
2306 decoding happen first, so such errors now appear on stderr, and the return
2307 code is 1. However, the actual setting up of the sockets still happens in
2308 the disconnected process, so errors there are still only recorded on the
2309 panic log.
2310
231140. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2312 connections (as happens on some IP stacks) was logged at start up time as
2313 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2314 it from "IPv6 and IPv4", which means that two separate sockets are being
2315 used.
2316
231741. The debug output for gethostbyname2() or getipnodebyname() failures now
2318 says whether AF_INET or AF_INET6 was passed as an argument.
2319
232042. Exiwhat output was messed up when time zones were included in log
2321 timestamps.
2322
232343. Exiwhat now gives more information about the daemon's listening ports,
2324 and whether -tls-on-connect was used.
2325
232644. The "port" option of the smtp transport is now expanded.
2327
232845. A "message" modifier in a "warn" statement in a non-message ACL was being
2329 silently ignored. Now an error message is written to the main and panic
2330 logs.
2331
233246. There's a new ACL modifier called "logwrite" which writes to a log file
2333 as soon as it is encountered.
2334
233547. Added $local_user_uid and $local_user_gid at routing time.
2336
233748. Exim crashed when trying to verify a sender address that was being
2338 rewritten to "<>".
2339
234049. Exim was recognizing only a space character after ".include". It now also
2341 recognizes a tab character.
2342
234350. Fixed several bugs in the Perl script that creates the exim.8 man page by
2344 extracting the relevant information from the specification. The man page no
2345 longer contains scrambled data for the -d option, and I've added a section
2346 at the front about calling Exim under different names.
2347
234851. Added "extra_headers" argument to the "mail" command in filter files.
2349
235052. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2351 crash.
2352
235353. Installed eximstats 1.29.
2354
235554. Added transport_filter_timeout as a generic transport option.
2356
235755. Exim no longer adds an empty Bcc: header to messages that have no To: or
2358 Cc: header lines. This was required by RFC 822, but it not required by RFC
2359 2822.
2360
236156. Exim used to add From:, Date:, and Message-Id: header lines to any
2362 incoming messages that did not have them. Now it does so only if the
2363 message originates locally, that is, if there is no associated remote host
2364 address. When Resent- header lines are present, this applies to the Resent-
2365 lines rather than the non-Resent- lines.
2366
236757. Drop incoming SMTP connection after too many syntax or protocol errors. The
2368 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2369
237058. Messages for configuration errors now include the name of the main
2371 configuration file - useful now that there may be more than one file in a
2372 list (.included file names were always shown).
2373
237459. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2375 for those rare installations that do not start the daemon as root or run it
2376 setuid root. I've cut out the call to initgroups() if the daemon is not
2377 root at that time.
2378
237960. The Exim user and group can now be bound into the binary as text strings
2380 that are looked up at the start of Exim's processing.
2381
238261. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2383
238462. Added $mailstore_basename variable.
2385
238663. Installed patch to sieve.c from Michael Haardt.
2387
238864. When Exim failed to open the panic log after failing to open the main log,
2389 the original message it was trying to log was written to stderr and debug
2390 output, but if they were not available (the usual case in production), it
2391 was lost. Now it is written to syslog before the two lines that record the
2392 failures to open the logs.
2393
239465. Users' Exim filters run in subprocesses under the user's uid. It is
2395 possible for a "deliver" command or an alias in a "personal" command to
2396 provoke an address rewrite. If logging of address rewriting is configured,
2397 this fails because the process is not running as root or exim. There may be
2398 a better way of dealing with this, but for the moment (because 4.30 needs
2399 to be released), I have disabled address rewrite logging when running a
2400 filter in a non-root, non-exim process.
2401
2402
2403Exim version 4.24
2404-----------------
2405
2406 1. The buildconfig auxiliary program wasn't quoting the value set for
2407 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2408 not defined. This bug was masked in 4.22 by the effect that was fixed in
2409 change 4.23/1.
2410
2411 2. Some messages that were rejected after a message id was allocated were
2412 shown as "incomplete" by exigrep. It no longer does this for messages that
2413 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2414
2415 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2416 have allow_domain_literals set, the ID did not get logged in the <= line.
2417 Domain literals are now always recognized in Message-ID: header lines.
2418
2419 4. The first argument for a ${extract expansion item is the key name or field
2420 number. Leading and trailing spaces in this item were not being ignored,
2421 causing some misleading effects.
2422
2423 5. When deliver_drop_privilege was set, single queue runner processes started
2424 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2425 same command in the process it spins off) were not dropping privilege.
2426
2427 6. When the daemon running as "exim" started a queue runner, it always
2428 re-executed Exim in the spun-off process. This is a waste of effort when
2429 deliver_drop_privilege is set. The new process now just calls the
2430 queue-runner function directly.
2431
2432
2433Exim version 4.23
2434-----------------
2435
2436 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2437 HEADERS_CHARSET.
2438
2439 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2440 ignored. Though the use of -oP was forcing the writing of a pid file, it
2441 was always written to the default place.
2442
2443 3. If the message "no IP address found for host xxxx" is generated during
2444 incoming verification, it is now followed by identification of the incoming
2445 connection (so you can more easily find what provoked it).
2446
2447 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2448
2449 5. Added some features to "harden" Exim a bit more against certain attacks:
2450
2451 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2452 be put in Local/Makefile. This is like the never_users runtime option,
2453 but it cannot be overridden. The default setting is "root".
2454
2455 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2456 prefix string with which any file named in a -C command line option
2457 must start.
2458
2459 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2460 is retained for -C and -D only if the caller of Exim is root. Without
2461 it, the exim user may also use -C and -D and retain privilege.
2462
2463 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2464 command line option is disabled.
2465
2466 6. Macro names set by the -D option must start with an upper case letter, just
2467 like macro names defined in the configuration file.
2468
2469 7. Added "dereference=" facility to LDAP.
2470
2471 8. Two instances of the typo "uknown" in the source files are fixed.
2472
2473 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2474 the Configure-Makefile script screwed up while processing it.
2475
247610. Incorporated PCRE 4.4.
2477
247811. The SMTP synchronization check was not operating right at the start of an
2479 SMTP session. For example, it could not catch a HELO sent before the client
2480 waited for the greeting. There is now a check for outstanding input at the
2481 point when the greeting is written. Because of the duplex, asynchronous
2482 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2483 way, but not yet received, when the check is performed.
2484
248512. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2486 on TCP/IP sockets, because this apparently causes some broken clients to
2487 timeout.
2488
248913. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2490 unchanged) from the Cygwin maintainer.
2491
249214. The code for -bV that shows what is in the binary showed "mbx" when maildir
2493 was supported instead of testing for mbx. Effectively a typo.
2494
249515. The spa authenticator server code was not checking that the input it
2496 received was valid base64.
2497
249816. The debug output line for the "set" modifier in ACLs was not showing the
2499 name of the variable that was being set.
2500
250117. Code tidy: the variable type "vtype_string" was never used. Removed it.
2502
250318. Previously, a reference to $sender_host_name did not cause a DNS reverse
2504 lookup on its own. Something else was needed to trigger the lookup. For
2505 example, a match in host_lookup or the need for a host name in a host list.
2506 Now, if $sender_host_name is referenced and the host name has not yet been
2507 looked up, a lookup is performed. If the lookup fails, the variable remains
2508 empty, and $host_lookup_failed is set to "1".
2509
251019. Added "eqi" as a case-independent comparison operator.
2511
251220. The saslauthd authentication condition could segfault if neither service
2513 nor realm was specified.
2514
251521. If an overflowing value such as "2048M" was set for message_size_limit, the
2516 error message that was logged was misleading, and incoming SMTP
2517 connections were dropped. The message is now more accurate, and temporary
2518 errors are given to SMTP connections.
2519
252022. In some error situations (such as 21 above) Exim rejects all SMTP commands
2521 (except RSET) with a 421 error, until QUIT is received. However, it was
2522 failing to send a response to QUIT.
2523
252423. The HELO ACL was being run before the code for helo_try_verify_hosts,
2525 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2526 ACL is now run after the helo_try_verify_hosts code.
2527
252824. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2529 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2530 case-independent, so other case variants are also recognized). Apparently
2531 some systems use these upper case variants.
2532
253325. If more than two messages were waiting for the same host, and a transport
2534 filter was specified for the transport, Exim sent two messages over the
2535 same TCP/IP connection, and then failed with "socket operation on non-
2536 socket" when it tried to send the third.
2537
253826. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2539
254027. The extern definition of crypt16() in expand.c was not being excluded when
2541 the OS had its own crypt16() function.
2542
254328. Added bounce_return_body as a new option, and bounce_return_size_limit
2544 as a preferred synonym for return_size_limit, both as an option and as an
2545 expansion variable.
2546
254729. Added LIBS=-liconv to OS/Makefile-OSF1.
2548
254930. Changed the default configuration ACL to relax the local part checking rule
2550 for addresses that are not in any local domains. For these addresses,
2551 slashes and pipe symbols are allowed within local parts, but the sequence
2552 /../ is explicitly forbidden.
2553
255431. SPA server authentication was not clearing the challenge buffer before
2555 using it.
2556
255732. log_message in a "warn" ACL statement was writing to the reject log as
2558 well as to the main log, which contradicts the documentation and doesn't
2559 seem right (because no rejection is happening). So I have stopped it.
2560
256133. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2562 However, I am unable to do any testing of this.
2563
256434. Fixed an infelicity in the appendfile transport. When checking directories
2565 for a mailbox, to see if any needed to be created, it was accidentally
2566 using path names with one or more superfluous leading slashes; tracing
2567 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2568
256935. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2570 contents are added to the log line that is written for every discarded
2571 recipient. (Previously a log_message setting was ignored.)
2572
257336. The ${quote: operator now quotes the string if it is empty.
2574
257537. The install script runs exim in order to find its version number. If for
2576 some reason other than non-existence or emptiness, which it checks, it
2577 could not run './exim', it was installing it with an empty version number,
2578 i.e. as "exim-". This error state is now caught, and the installation is
2579 aborted.
2580
258138. An argument was missing from the function that creates an error message
2582 when Exim fails to connect to the socket for saslauthd authentication.
2583 This could cause Exim to crash, or give a corrupted message.
2584
258539. Added isip, isip4, and isip6 to ${if conditions.
2586
258740. The ACL variables $acl_xx are now saved with the message, and can be
2588 accessed later in routers, transports, and filters.
2589
259041. The new lookup type nwildlsearch is like wildlsearch, except that the key
2591 strings in the file are not string-expanded.
2592
259342. If a MAIL command specified a SIZE value that was too large to fit into an
2594 int variable, the check against message_size_limit failed. Such values are
2595 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2596 day this will have to be increased, but I don't think I want to be around
2597 when emails are that large.
2598
2599
2600
2601Exim version 4.22
2602-----------------
2603
2604 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2605 iconv() is not standard in FreeBSD.
2606
2607 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2608 IPv6 enabled. The observed symptom was a segmentation fault on return from
2609 the function os_common_find_running_interfaces() in src/os.c.
2610
2611 3. In the check_special_case() function in daemon.c I had used "errno" as an
2612 argument name, which causes warnings on some systems. This was basically a
2613 typo, since it was named "eno" in the comments!
2614
2615 4. The code that waits for the clock to tick (at a resolution of some fraction
2616 of a second) so as to ensure message-id uniqueness was always waiting for
2617 at least one whole tick, when it could have waited for less. [This is
2618 almost certainly not relevant at current processor speeds, where it is
2619 unlikely to ever wait at all. But we try to future-proof.]
2620
2621 5. The function that sleeps for a time interval that includes fractions of a
2622 second contained a race. It did not block SIGALRM between setting the
2623 timer, and suspending (a couple of lines later). If the interval was short
2624 and the sigsuspend() was delayed until after it had expired, the suspension
2625 never ended. On busy systems this could lead to processes getting stuck for
2626 ever.
2627
2628 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2629 process, before it forks any delivery processes. The open lookup caching
2630 mechanism meant that the open file or database connection was passed into
2631 the delivery process. The problem was that delivery processes always tidy
2632 up cached lookup data. This could cause a problem for the next delivery
2633 process started by the queue runner, because the external queue runner
2634 process does not know about the closure. So the next delivery process
2635 still has data in the lookup cache. In the case of a file lookup, there was
2636 no problem because closing a file descriptor in a subprocess doesn't affect
2637 the parent. However, if the lookup was caching a connection to a database,
2638 the connection was closed, and the second delivery process was likely to
2639 see errors such as "PGSQL: query failed: server closed the connection
2640 unexpectedly". The problem has been fixed by closing all cached lookups
2641 in a queue runner before running a delivery process.
2642
2643 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2644 seem to have the "const" qualifier which it has on other OS. I've
2645 parameterised it.
2646
2647 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2648 *of the same type* (client or server) with the same public name that an
2649 error should be diagnosed.
2650
2651 9. When Exim looked up a host name for an IP address, but failed to find the
2652 original IP address when looking up the host name (a safety check), it
2653 output the message "<ip address> does not match any IP for NULL", which was
2654 confusing, to say the least. The bug was that the host name should have
2655 appeared instead of "NULL".
2656
265710. Since release 3.03, if Exim is called by a uid other than root or the Exim
2658 user that is built into the binary, and the -C or -D options is used, root
2659 privilege is dropped before the configuration file is read. In addition,
2660 logging is switched to stderr instead of the normal log files. If the
2661 configuration then re-defines the Exim user, the unprivileged environment
2662 is probably not what is expected, so Exim logs a panic warning message (but
2663 proceeds).
2664
2665 However, if deliver_drop_privilege is set, the unprivileged state may well
2666 be exactly what is intended, so the warning has been cut out in that case,
2667 and Exim is allowed to try to write to its normal log files.
2668
2669
2670Exim version 4.21
2671-----------------
2672
2673 1. smtp_return_error_details was not giving details for temporary sender
2674 or receiver verification errors.
2675
2676 2. Diagnose a configuration error if two authenticators have the same public
2677 name.
2678
2679 3. Exim used not to create the message log file for a message until the first
2680 delivery attempt. This could be confusing when incoming messages were held
2681 for policy or load reasons. The message log file is now created at the time
2682 the message is received, and an initial "Received" line is written to it.
2683
2684 4. The automatically generated man page for command line options had a minor
2685 bug that caused no ill effects; however, a more serious problem was that
2686 the procedure for building the man page automatically didn't always
2687 operate. Consequently, release 4.20 contains an out-of-date version. This
2688 shouldn't happen again.
2689
2690 5. When building Exim with embedded Perl support, the script that builds the
2691 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2692 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2693
2694 6. The freeze_tell option was not being used for messages that were frozen on
2695 arrival, either by an ACL or by local_scan().
2696
2697 7. Added the smtp_incomplete_transaction log selector.
2698
2699 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2700 was accepting AUTH without a new EHLO.
2701
2702 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2703 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2704 RFC.
2705
270610. Logging of TCP/IP connections (when configured) now happens in the main
2707 daemon process instead of the child process, so that the TCP/IP connection
2708 count is more accurate (but it can never be perfect).
2709
271011. The use of "drop" in a nested ACL was not being handled correctly in the
2711 outer ACL. Now, if condition failure induced by the nested "drop" causes
2712 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2713 or "require"), the connection is dropped.
2714
271512. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2716 that yield "discard" can now be used with an "accept" or a "discard" verb,
2717 but an error is generated for any others (because I can't see a useful way
2718 to define what should happen).
2719
272013. When an ACL is read dynamically from a file (or anywhere else), the lines
2721 are now processed in the same way as lines in the Exim configuration file.
2722 In particular, continuation lines are supported.
2723
272414. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2725
272615. Added -ti meaning -t -i.
2727
272816. Check for letters, digits, hyphens, and dots in the names of dnslist
2729 domains, and warn by logging if others are found.
2730
273117. At least on BSD, alignment is not guarenteed for the array of ifreq's
2732 returned from GIFCONF when Exim is trying to find the list of interfaces on
2733 a host. The code in os.c has been modified to copy each ifreq to an aligned
2734 structure in all cases.
2735
2736 Also, in some cases, the returned ifreq's were being copied to a 'struct
2737 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2738 means the last couple of bytes of an IPv6 address could be chopped if the
2739 ifreq contained only a normal sockaddr (14 bytes storage).
2740
274118. Named domain lists were not supported in the hosts_treat_as_local option.
2742 An entry such as +xxxx was not recognized, and was treated as a literal
2743 domain name.
2744
274519. Ensure that header lines added by a DATA ACL are included in the reject log
2746 if the ACL subsequently rejects the message.
2747
274820. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2749 MD5 (which is deprecated).
2750
275121. When testing a filter file using -bf, Exim was writing a message when it
2752 took the sender from a "From " line in the message, but it was not doing so
2753 when it took $return_path from a Return-Path: header line. It now does.
2754
275522. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2756 with a valid header line field name (a series of printing characters
2757 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2758
275923. Changed "disc" in the source to "disk" to conform to the documentation and
2760 the book and for uniformity.
2761
276224. Ignore Sendmail's -Ooption=value command line item.
2763
276425. When execve() failed while trying to run a command in a pipe transport,
2765 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2766 could be confused with a return value of 69 from the command itself. This
2767 has been changed to 127, the value the shell returns if it is asked to run
2768 a non-existent command. The wording for the related log line suggests a
2769 non-existent command as the problem.
2770
277126. If received_header_text expands to an empty string, do not add a Received:
2772 header line to the message. (Well, it adds a token one on the spool, but
2773 marks it "old" so that it doesn't get used or transmitted.)
2774
277527. Installed eximstats 1.28 (addition of -nt option).
2776
277728. There was no check for failure on the call to getsockname() in the daemon
2778 code. This can fail if there is a shortage of resources on the system, with
2779 ENOMEM, for example. A temporary error is now given on failure.
2780
278129. Contrary to the C standard, it seems that in some environments, the
2782 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2783 program. Exim now does this explicitly; it affects the formatting of
2784 timestamps using strftime().
2785
278630. If exiqsumm was given junk data, it threw up some uninitialized variable
2787 complaints. I've now initialized all the variables, to avoid this.
2788
278932. Header lines added by a system filter were not being "seen" during
2790 transport-time rewrites.
2791
279233. The info_callback() function passed to OpenSSL is set up with type void
2793 (*)(SSL *, int, int), as described somewhere. However, when calling the
2794 function (actually a macro) that sets it up, the type void(*)() is
2795 expected. I've put in a cast to prevent warnings from picky compilers.
2796
279734. If a DNS black list lookup found a CNAME record, but there were no A
2798 records associated with the domain it pointed at, Exim crashed.
2799
280035. If a DNS black list lookup returned more than one A record, Exim ignored
2801 all but the first. It now scans all returned addresses if a particular IP
2802 value is being sought. In this situation, the contents of the
2803 $dnslist_value variable are a list of all the addresses, separated by a
2804 comma and a space.
2805
280636. Tightened up the rules for host name lookups using reverse DNS. Exim used
2807 to accept a host name and all its aliases if the forward lookup for any of
2808 them yielded the IP address of the incoming connection. Now it accepts only
2809 those names whose forward lookup yields the correct IP address. Any other
2810 names are discarded. This closes a loophole whereby a rogue DNS
2811 administrator could create reverse DNS records to break through a
2812 wildcarded host restriction in an ACL.
2813
281437. If a user filter or a system filter that ran in a subprocess used any of
2815 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2816 the wrong values were passed to the pipe command ($thisaddress had the
2817 value of $0, $0 had the value of $1, etc). This bug was introduced by
2818 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2819
282038. Improved the line breaking for long SMTP error messages from ACLs.
2821 Previously, if there was no break point between 40 and 75 characters, Exim
2822 left the rest of the message alone. Two changes have been made: (a) I've
2823 reduced the minimum length to 35 characters; (b) if it can't find a break
2824 point between 35 and 75 characters, it looks ahead and uses the first one
2825 that it finds. This may give the occasional overlong line, but at least the
2826 remaining text gets split now.
2827
282839. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2829 file descriptors might be low, and that setting 1000 would always raise it.
2830 It turns out that in some environments, the limit is already over 1000 and
2831 that lowering it causes trouble. So now Exim takes care not to decrease it.
2832
283340. When delivering a message, the value of $return_path is set to $sender_
2834 address at the start of routing (routers may change the value). By an
2835 oversight, this default was not being set up when an address was tested by
2836 -bt or -bv, which affected the outcome if any router or filter referred to
2837 $return_path.
2838
283941. The idea of the "warn" ACL verb is that it adds a header or writes to the
2840 log only when "message" or "log_message" are set. However, if one of the
2841 conditions was an address verification, or a call to a nested ACL, the
2842 messages generated by the underlying test were being passed through. This
2843 no longer happens. The underlying message is available in $acl_verify_
2844 message for both "message" and "log_message" expansions, so it can be
2845 passed through if needed.
2846
284742. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2848 new expansion $bh_ to give the encoded byte string without charset
2849 translation. Translation happens only if iconv() is available; HAVE_ICONV
2850 indicates this at build time. HEADERS_CHARSET gives the charset to
2851 translate to; headers_charset can change it in the configuration, and
2852 "headers charset" can change it in an individual filter file.
2853
285443. Now that we have a default RFC 2047 charset (see above), the code in Exim
2855 that creates RFC 2047 encoded "words" labels them as that charset instead
2856 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2857 expansion operator; (ii) when Exim creates a From: line for a local
2858 message; (iii) when a header line is rewritten to include a "phrase" part.
2859
286044. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2861 buggy, causing it to skip the first lines of messages whose message ID
2862 ended in 'D'. This would not have bitten before Exim release 4.14, because
2863 message IDs were unlikely to end in 'D' before then. The effect was to have
2864 incorrect size information for certain domains.
2865
286645. #include "config.h" was missing at the start of the crypt16.c module. This
2867 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2868 noticed.
2869
287046. If there was a timeout during a "random" callout check, Exim treated it as
2871 a failure of the random address, and carried on sending RSET and the real
2872 address. If the delay was just some slowness somewhere, the response to the
2873 original RCPT would be taken as a response to RSET and so on, causing
2874 mayhem of various kinds.
2875
287647. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2877 when I implemented it. It didn't allow for the fact that some option values
2878 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2879 the right test for positive values.
2880
288148. Domain names in DNS records are case-independent. Exim always looks them up
2882 in lower case. Some resolvers return domain names in exactly the case they
2883 appear in the zone file, that is, they may contain uppercase letters. Not
2884 all resolvers do this - some return always lower case. Exim was treating a
2885 change of case by a resolver as a change of domain, similar to a widening
2886 of a domain abbreviation. This triggered its re-routing code and so it was
2887 trying to route what was effectively the same domain again. This normally
2888 caused routing to fail (because the router wouldn't handle the domain
2889 twice). Now Exim checks for this case specially, and just changes the
2890 casing of the domain that it ultimately uses when it transmits the message
2891 envelope.
2892
289349. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2894 module.
2895
289650. If a filter generated a file delivery with a non-absolute name (possible if
2897 no home directory exists for the router), the forbid_file option was not
2898 forbidding it.
2899
290051. Added '&' feature to dnslists, to provide bit mask matching in addition to
2901 the existing equality matching.
2902
290352. Exim was using ints instead of ino_t variables in some places where it was
2904 dealing with inode numbers.
2905
290653. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2907 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2908 called TMPDIR, and if it finds it is different, it changes its value.
2909
291054. The smtp_printf() function is now made available to local_scan() so
2911 additional output lines can be written before returning. There is also an
2912 smtp_fflush() function to enable the detection of a dropped connection.
2913 The variables smtp_input and smtp_batched_input are exported to
2914 local_scan().
2915
291655. Changed the default runtime configuration: the message "Unknown user"
2917 has been removed from the ACL, and instead placed on the localuser router,
2918 using the cannot_route_message feature. This means that any verification
2919 failures that generate their own messages won't get overridden. Similarly,
2920 the "Unrouteable address" message that was in the ACL for unverifiable
2921 relay addresses has also been removed.
2922
292356. Added hosts_avoid_esmtp to the smtp transport.
2924
292557. The exicyclog script was not checking for the esoteric option
2926 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2927 will work only if exicyclog is run under the appropriate euid.
2928
292958. Following a discussion on the list, the rules by which Exim recognises line
2930 endings on incoming messages have been changed. The -dropcr and drop_cr
2931 options are now no-ops, retained only for backwards compatibility. The
2932 following line terminators are recognized: LF CRLF CR. However, special
2933 processing applies to CR:
2934
2935 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2936 nor a local message in the state where . is a terminator.
2937
2938 (ii) If a bare CR is encountered in a header line, an extra space is added
2939 after the line terminator so as not to end the header. The reasoning
2940 behind this is that bare CRs in header lines are most likely either
2941 to be mistakes, or people trying to play silly games.
2942
294359. The size of a message, as listed by "-bp" or in the Exim monitor window,
2944 was being incorrectly given as 18 bytes larger than it should have been.
2945 This is a VOB (very old bug).
2946
294760. This may never have affected anything current, but just in case it has:
2948 When the local host is found other than at the start of a list of hosts,
2949 the local host, those with the same MX, and any that follow, are discarded.
2950 When the list in question was part of a longer list of hosts, the following
2951 hosts (not currently being processed) were also being discarded. This no
2952 longer happens. I'm not sure if this situation could ever has previously
2953 arisen.
2954
295561. Added the "/MX" feature to lists of hosts in the manualroute and query
2956 program routers.
2957
295862. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2959 header. This is something that is recommended in a new Internet Draft, and
2960 is something that is documented as being done by Sendmail. There are two
2961 possible values. For messages generated by the autoreply transport, Exim
2962 adds:
2963
2964 Auto-Submitted: auto-replied
2965
2966 whereas for all other generated messages (e.g. bounces) it adds
2967
2968 Auto-Submitted: auto-generated
2969
297063. The "personal" condition in filters now includes a test for the
2971 Auto-Submitted: header. If it contains the string "auto-" the message it
2972 not considered personal.
2973
297464. Added rcpt_include_affixes as a generic transport option.
2975
297665. Added queue_only_override (default true).
2977
297866. Added the syslog_duplication option.
2979
298067. If what should have been the first header line of a message consisted of
2981 a space followed by a colon, Exim was mis-interpreting it as a header line.
2982 It isn't of course - it is syntactically invalid and should therefore be
2983 treated as the start of the message body. The misbehaviour could have
2984 caused a number of strange effects, including loss of data in subsequent
2985 header lines, and spool format errors.
2986
298768. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2988 client host had authenticated. This control can now be exercised by an ACL
2989 for more flexibility.
2990
299169. By default, callouts do not happen when testing with -bh. There is now a
2992 variant, -bhc, which does actually run the callout code, including
2993 consulting and updating the callout cache.
2994
299570. Added support for saslauthd authentication, courtesy of Alexander
2996 Sabourenkov.
2997
299871. If statvfs() failed on the spool or log directories while checking their
2999 size for availability, Exim confusingly gave the error "space shortage".
3000 Furthermore, in debugging mode it crashed with a floating point exception.
3001 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3002 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3003 serious problem, Exim now writes to the main and panic logs when this
3004 happens, with details of the failure. It then refuses to accept the
3005 incoming message, giving the message "spool directory problem" or "log
3006 directory problem" with a 421 code for SMTP messages.
3007
300872. When Exim is about to re-exec itself, it ensures that the file descriptors
3009 0, 1, and 2 exist, because some OS complain for execs without them (see
3010 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3011 descriptors. However, the code omitted to check that the open succeeded,
3012 causing mysterious errors if for some reason the permissions on /dev/null
3013 got screwed. Now Exim writes a message to the main and panic logs, and
3014 bombs out if it can't open /dev/null.
3015
301673. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3017 interact so that it is all more flexible. It is supposed to remain
3018 backwards compatible. Also added extra_local_interfaces.
3019
302074. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3021 to bomb out with an assertion failure - to the client this appears as a
3022 connection drop. This problem occurs in the part of the code that was taken
3023 from the Samba project. Fortunately, the assertion is in a very simple
3024 function, so I have fixed this by reproducing the function inline in the
3025 one place where it is called, and arranging for authentication to fail
3026 instead of killing the process with assert().
3027
302875. The SPA client code was not working when the server requested OEM rather
3029 than Unicode encoding.
3030
303176. Added code to make require_files with a specific uid setting more usable in
3032 the case where statting the file as root fails - usually a non-root-mounted
3033 NFS file system. When this happens and the failure is EACCES, Exim now
3034 forks a subprocess and does the per-uid checking as the relevant uid.
3035
303677. Added process_log_path.
3037
303878. If log_file_path was not explicitly set, a setting of check_log_space or
3039 check_log_inodes was ignored.
3040
304179. If a space check for the spool or log partitions fails, the incident is now
3042 logged. Of course, in the latter case the data may get lost...
3043
304480. Added the %p formatting code to string_format() so that it can be used to
3045 print addresses in debug_print(). Adjusted all the address printing in the
3046 debugging in store.c to use %p rather than %d.
3047
304881. There was a concern that a line of code in smtp_in.c could overflow a
3049 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3050 initially expressed, the concern was not well-founded, because trailing
3051 spaces are removed early. However, if the trailing spaces were followed by
3052 a NULL, they did not get removed, so the overflow was possible. Two fixes
3053 were applied:
3054
3055 (a) I re-wrote the offending code in a cleaner fashion.
3056 (b) If an incoming SMTP command contains a NULL character, it is rejected
3057 as invalid.
3058
305982. When Exim changes uid/gid to the Exim user at daemon start time, it now
3060 runs initgroups(), so that if the Exim user is in any additional groups,
3061 they will be used during message reception.
3062
3063
3064Exim version 4.20
3065-----------------
3066
3067The change log for 4.20 and earlier releases has been archived.
3068
3069****