Avoid segfault on ref:name specified as uid.
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
5dc43717 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
495ae4b0
PH
2
3Change log file for Exim from version 4.21
4-------------------------------------------
5
10906672
PP
6Exim version 4.76
7-----------------
8
9PP/01 The new ldap_require_cert option would segfault if used. Fixed.
10
754a0503
PP
11PP/02 Harmonised TLS library version reporting; only show if debugging.
12 Layout now matches that introduced for other libraries in 4.74 PP/03.
13
c0c7b2da
PP
14PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
15
e97d1f08
PP
16PP/04 New "dns_use_edns0" global option.
17
084c1d8c
PP
18PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
19 Bugzilla 1098.
20
10906672 21
aa097c4c
NM
22Exim version 4.75
23-----------------
24
4c9ef03a 25NM/01 Workround for PCRE version dependency in version reporting
aa097c4c
NM
26 Bugzilla 1073
27
7f3d9eff
TF
28TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
29 This fixes portability to compilers other than gcc, notably
30 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
31
159f52d2
TF
32TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
33 makefiles for portability to HP-UX and POSIX correctness.
34
0cc9542a
PP
35PP/01 Permit LOOKUP_foo enabling on the make command-line.
36 Also via indented variable definition in the Makefile.
37 (Debugging by Oliver Heesakkers).
38
f7274286
PP
39PP/02 Restore caching of spamd results with expanded spamd_address.
40 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
41
7b797365
PP
42PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
43 Improves build reliability. Fix from: Frank Elsner
44
caacae52
NM
45NM/02 Fix wide character breakage in the rfc2047 coding
46 Fixes bug 1064. Patch from Andrey N. Oktyabrski
47
09dcaba9
NM
48NM/03 Allow underscore in dnslist lookups
49 Fixes bug 1026. Patch from Graeme Fowler
50
bc19a55b
PP
51PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
52 Code patches from Adam Ciarcinski of NetBSD.
caacae52 53
bd4c9759
NM
54NM/04 Fixed exiqgrep to cope with mailq missing size issue
55 Fixes bug 943.
56
b72aab72
PP
57PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
58 is logged, to avoid truncation. Patch from John Horne.
59
2fe76745
PP
60PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
61 Patch from Jakob Hirsch.
62
76aa570c
PP
63PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
64 SQL string expansion failure details.
65 Patch from Andrey Oktyabrski.
66
f1e5fef5
PP
67PP/08 Bugzilla 486: implement %M datestamping in log filenames.
68 Patch from Simon Arlott.
69
4d805ee9
PP
70PP/09 New lookups functionality failed to compile on old gcc which rejects
71 extern declarations in function scope.
72 Patch from Oliver Fleischmann
73
cd59ab18
PP
74PP/10 Use sig_atomic_t for flags set from signal handlers.
75 Check getgroups() return and improve debugging.
76 Fixed developed for diagnosis in bug 927 (which turned out to be
77 a kernel bug).
78
332f5cf3
PP
79PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
80 Patch from Mark Zealey.
81
29cfeb94
PP
82PP/12 Bugzilla 1056: Improved spamd server selection.
83 Patch from Mark Zealey.
84
660242ad
PP
85PP/13 Bugzilla 1086: Deal with maildir quota file races.
86 Based on patch from Heiko Schlittermann.
87
bc4bc4c5
PP
88PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
89 Patch from Uwe Doering, sign-off by Michael Haardt.
90
2e64baa9
NM
91NM/05 Fix to spam.c to accommodate older gcc versions which dislike
92 variable declaration deep within a block. Bug and patch from
93 Dennis Davis.
94
bddd7526
PP
95PP/15 lookups-Makefile IRIX compatibilty coercion.
96
6bac1a9a
PP
97PP/16 Make DISABLE_DKIM build knob functional.
98
552193f0
NM
99NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
100 Patch by Simon Arlott
baeee2c1 101
1b587e48
TF
102TF/03 Fix valgrind.h portability to C89 compilers that do not support
103 variable argument macros. Our copy now differs from upstream.
104
aa097c4c 105
8c07b69f
TF
106Exim version 4.74
107-----------------
108
109TF/01 Failure to get a lock on a hints database can have serious
110 consequences so log it to the panic log.
111
c0ea85ab
TF
112TF/02 Log LMTP confirmation messages in the same way as SMTP,
113 controlled using the smtp_confirmation log selector.
114
0761d44e
TF
115TF/03 Include the error message when we fail to unlink a spool file.
116
0a349494
PP
117DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
118 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
119 for maintaining out-of-tree patches for some time.
120
121PP/01 Bugzilla 139: Documentation and portability issues.
122 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
123 Handle per-OS dynamic-module compilation flags.
124
fea24b2e
PP
125PP/02 Let /dev/null have normal permissions.
126 The 4.73 fixes were a little too stringent and complained about the
127 permissions on /dev/null. Exempt it from some checks.
128 Reported by Andreas M. Kirchwitz.
129
6545de78
PP
130PP/03 Report version information for many libraries, including
131 Exim version information for dynamically loaded libraries. Created
132 version.h, now support a version extension string for distributors
133 who patch heavily. Dynamic module ABI change.
134
1670ef10
PP
135PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
136 privilege escalation vulnerability whereby the Exim run-time user
137 can cause root to append content of the attacker's choosing to
138 arbitrary files.
139
c0886197
PP
140PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
141 (Wolfgang Breyha)
142
b7487bce
PP
143PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
144 If dropping privileges for untrusted macros, we disabled normal logging
145 on the basis that it would fail; for the Exim run-time user, this is not
146 the case, and it resulted in successful deliveries going unlogged.
147 Fixed. Reported by Andreas Metzler.
148
8c07b69f 149
97fd1e48 150Exim version 4.73
ed7f7860 151-----------------
97fd1e48
PP
152
153PP/01 Date: & Message-Id: revert to normally being appended to a message,
154 only prepend for the Resent-* case. Fixes regression introduced in
155 Exim 4.70 by NM/22 for Bugzilla 607.
156
6901c596
PP
157PP/02 Include check_rfc2047_length in configure.default because we're seeing
158 increasing numbers of administrators be bitten by this.
159
a8c8d6b5
JJ
160JJ/01 Added DISABLE_DKIM and comment to src/EDITME
161
77bb000f
PP
162PP/03 Bugzilla 994: added openssl_options main configuration option.
163
a29e5231
PP
164PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
165
ec5a0394 166PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
a29e5231 167
55c75993
PP
168PP/06 Adjust NTLM authentication to handle SASL Initial Response.
169
453a6645 170PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
ec5a0394
PP
171 without a peer certificate, leading to a segfault because of an
172 assumption that peers always have certificates. Be a little more
453a6645
PP
173 paranoid. Problem reported by Martin Tscholak.
174
8544e77a
PP
175PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
176 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
177 NB: ClamAV planning to remove STREAM in "middle of 2010".
3346ab01
PP
178 CL also introduces -bmalware, various -d+acl logging additions and
179 more caution in buffer sizes.
8544e77a 180
83e029d5
PP
181PP/09 Implemented reverse_ip expansion operator.
182
ed7f7860
PP
183PP/10 Bugzilla 937: provide a "debug" ACL control.
184
7d9f747b
PP
185PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
186
4b2241d2
PP
187PP/12 Bugzilla 973: Implement --version.
188
10385c15
PP
189PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
190
dbc4b90d
PP
191PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
192
532be449
PP
193PP/15 Bugzilla 816: support multiple condition rules on Routers.
194
6a8de854 195PP/16 Add bool_lax{} expansion operator and use that for combining multiple
71265ae9
PP
196 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
197 ignore trailing whitespace.
6a8de854 198
5dc43717
JJ
199JJ/02 prevent non-panic DKIM error from being sent to paniclog
200
201JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
202 "exim" to be used
55c75993 203
3346ab01
PP
204PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
205 Notification from Dr Andrew Aitchison.
206
491fab4c
PP
207PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
208 ExtendedDetectionInfo response format.
209 Notification from John Horne.
210
13eb9497
PP
211PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
212 compatible.
213
214PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
215 XSL and documented dependency on system catalogs, with examples of how
216 it normally works.
217
7f36d675
DW
218DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
219 access.
220
c1d94452
DW
221DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
222 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
223 configuration file which is writeable by the Exim user or group.
224
e2f5dc15
DW
225DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
226 of configuration files to cover files specified with the -C option if
227 they are going to be used with root privileges, not just the default
228 configuration file.
229
cd25e41d
DW
230DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
231 option (effectively making it always true).
232
261dc43e
DW
233DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
234 files to be used while preserving root privileges.
235
fa32850b
DW
236DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
237 that rogue child processes cannot use them.
238
79d4bc3d
PP
239PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
240 run-time user, instead of root.
241
2cfd3221
PP
242PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
243 Exim run-time user without dropping privileges.
244
fb08281f
DW
245DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
246 result string, instead of calling string_vformat() twice with the same
247 arguments.
3346ab01 248
74935b98
DW
249DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
250 for other users. Others should always drop root privileges if they use
251 -C on the command line, even for a whitelisted configure file.
252
90b6341f
DW
253DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
254
57730b52
ML
255NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
256
66581d1e 257
465e92cf
JJ
258Exim version 4.72
259-----------------
260
453a6645
PP
261JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
262 $data_path, and $header_path variables; fixed documentation bugs and
263 typos
465e92cf 264
453a6645
PP
265JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
266 exipick to access non-standard spools, including the "frozen" queue
267 (Finput)
edae0343 268
9bd3e22c
NM
269NM/01 Bugzilla 965: Support mysql stored procedures.
270 Patch from Alain Williams
271
bb576ff7
NM
272NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
273
5a1a5845
NM
274NM/03 Bugzilla 955: Documentation fix for max_rcpts.
275 Patch from Andreas Metzler
276
981a9fad
NM
277NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
278 Patch from Kirill Miazine
279
7fc497ee
NM
280NM/05 Bugzilla 671: Added umask to procmail example.
281
1a41defa
JJ
282JJ/03 installed exipick 20100323.0, fixing doc bug
283
a466095c 284NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
b26eacf1 285 directory. Notification and patch from Dan Rosenberg.
a466095c 286
94a6bd0b
NM
287TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
288
289TK/02 Improve log output when DKIM signing operation fails.
290
291MH/01 Treat the transport option dkim_domain as a colon separated
292 list, not as a single string, and sign the message with each element,
293 omitting multiple occurences of the same signer.
294
c1b141a8
NM
295NM/07 Null terminate DKIM strings, Null initialise DKIM variable
296 Bugzilla 985, 986. Patch by Simon Arlott
94a6bd0b 297
b26eacf1 298NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
0d0c6357
NM
299 Patch by Simon Arlott
300
179c5980 301PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
b26eacf1 302 MBX locking. Notification from Dan Rosenberg.
179c5980 303
9bd3e22c 304
7c6d71af
NM
305Exim version 4.71
306-----------------
307
7d9f747b 308TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
7c6d71af 309
f013fb92
NM
310NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
311
0eb8eedd
NM
312NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
313
663ee6d9
NM
314NM/03 Bugzilla 847: Enable DNSDB lookup by default.
315
177ebd9b
NM
316NM/04 Bugzilla 915: Flag broken perl installation during build.
317
7c6d71af 318
210f147e
NM
319Exim version 4.70
320-----------------
321
cdd3bb85 322TK/01 Added patch by Johannes Berg that expands the main option
e739e3d9 323 "spamd_address" if it starts with a dollar sign.
cdd3bb85
TK
324
325TK/02 Write list of recipients to X-Envelope-Sender header when building
326 the mbox-format spool file for content scanning (suggested by Jakob
7d9f747b 327 Hirsch).
cdd3bb85
TK
328
329TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
330 (http://www.dcc-servers.net/) support via dccifd. Activated by
e739e3d9 331 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
cdd3bb85
TK
332
333TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
334 by Mark Daniel Reidel <mr@df.eu>.
335
210f147e
NM
336NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
337 When building exim an external PCRE library is now needed -
338 PCRE is a system library on the majority of modern systems.
339 See entry on PCRE_LIBS in EDITME file.
340
deafd5b3
NM
341NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
342 conversation. Added nologin parameter to request.
7d9f747b 343 Patch contributed by Kirill Miazine.
deafd5b3 344
089793a4
TF
345TF/01 Do not log submission mode rewrites if they do not change the address.
346
5f16ca82
TF
347TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
348
dae9d94e 349NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
7d9f747b 350 log files in place. Contributed by Roberto Lima.
dae9d94e 351
7d9f747b 352NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3f0da4d0 353
06864c44
TF
354TF/03 Bugzilla 615: When checking the local_parts router precondition
355 after a local_part_suffix or local_part_prefix option, Exim now
356 does not use the address's named list lookup cache, since this
357 contains cached lookups for the whole local part.
358
65a7d8c3 359NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
7d9f747b 360 Robert Millan. Documentation is in experimental-spec.txt.
65a7d8c3 361
23510047 362TF/04 Bugzilla 668: Fix parallel build (make -j).
65a7d8c3 363
7d9f747b 364NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
5f28a6e8 365
7d8eec3a 366NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
7d9f747b 367 Patch provided by Jan Srzednicki.
6c588e74 368
89dec7b6
TF
369TF/05 Leading white space used to be stripped from $spam_report which
370 wrecked the formatting. Now it is preserved.
5f28a6e8 371
a99de90c
TF
372TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
373 that they are available at delivery time.
374
e2803e40
TF
375TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
376
7199e1ee
TF
377TF/08 TLS error reporting now respects the incoming_interface and
378 incoming_port log selectors.
379
e276e04b
TF
380TF/09 Produce a more useful error message if an SMTP transport's hosts
381 setting expands to an empty string.
382
ce552449 383NM/06 Bugzilla 744: EXPN did not work under TLS.
7d9f747b 384 Patch provided by Phil Pennock.
ce552449 385
e765a0f1 386NM/07 Bugzilla 769: Extraneous comma in usage fprintf
7d9f747b 387 Patch provided by Richard Godbee.
e765a0f1 388
4f054c63 389NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
447de4b0 390 acl_smtp_notquit, added index entry.
4f054c63 391
7d9f747b
PP
392NM/09 Bugzilla 787: Potential buffer overflow in string_format.
393 Patch provided by Eugene Bujak.
24c929a2 394
7d9f747b
PP
395NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
396 accept(). Patch provided by Maxim Dounin.
cf73943b 397
b52bc06e 398NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
7d9f747b 399 Patch provided by Phil Pennock.
b52bc06e 400
447de4b0
NM
401NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
402
4c69d561 403NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
7d9f747b 404 Patch provided by Brad "anomie" Jorsch.
4c69d561 405
d5c39246 406NM/14 Bugzilla 622: Added timeout setting to transport filter.
7d9f747b 407 Patch provided by Dean Brooks.
9b989985 408
0b23848a
TK
409TK/05 Add native DKIM support (does not depend on external libraries).
410
8f3414a1 411NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
7d9f747b 412 Patch provided by Graeme Fowler.
e2aacdfd 413
fb6f955d
NM
414NM/16 Bugzilla 851: Documentation example syntax fix.
415
416NM/17 Changed NOTICE file to remove references to embedded PCRE.
8f3414a1 417
7d9f747b
PP
418NM/18 Bugzilla 894: Fix issue with very long lines including comments in
419 lsearch.
dbb0bf41 420
7d9f747b
PP
421NM/19 Bugzilla 745: TLS version reporting.
422 Patch provided by Phil Pennock.
f3766eb5 423
7d9f747b
PP
424NM/20 Bugzilla 167: bool: condition support.
425 Patch provided by Phil Pennock.
36f12725 426
7d9f747b
PP
427NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
428 clients. Patch provided by Phil Pennock.
e6060e2c 429
7d9f747b
PP
430NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
431 Patch provided by Brad "anomie" Jorsch.
5eb690a1 432
7d9f747b
PP
433NM/23 Bugzilla 687: Fix misparses in eximstats.
434 Patch provided by Heiko Schlittermann.
d5c13d66 435
7d9f747b
PP
436NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
437 Patch provided by Heiko Schlittermann.
b2335c0b 438
7d9f747b 439NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1da77999 440 plus update to original patch.
f4cd9433 441
7d9f747b 442NM/26 Bugzilla 799: Documentation correction for ratelimit.
dc988b7e 443
7d9f747b
PP
444NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
445 Patch provided by David Brownlee.
8dc71ab3 446
7d9f747b 447NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
400eda43 448
7d9f747b 449NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
ec5a421b 450
7d9f747b 451NM/30 Bugzilla 888: TLS documentation bugfixes.
07af267e 452
7d9f747b 453NM/31 Bugzilla 896: Dovecot buffer overrun fix.
51473862 454
17792b53 455NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
7d9f747b 456 Unlike the original bugzilla I have changed all shell scripts in src tree.
17792b53 457
7d9f747b
PP
458NM/33 Bugzilla 898: Transport filter timeout fix.
459 Patch by Todd Rinaldo.
52383f8f 460
7d9f747b
PP
461NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
462 Patch by Serge Demonchaux.
5ca6d115 463
7d9f747b
PP
464NM/35 Bugzilla 39: Base64 decode bug fixes.
465 Patch by Jakob Hirsch.
baee9eee 466
7d9f747b 467NM/36 Bugzilla 909: Correct connect() call in dcc code.
e93a964c 468
7d9f747b 469NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
9bf3d68f 470
7d9f747b 471NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
96535b98 472
7d9f747b 473NM/39 Bugzilla 911: Fixed MakeLinks build script.
30339e0f 474
deafd5b3 475
47db1125
NM
476Exim version 4.69
477-----------------
478
4b3504d0
TK
479TK/01 Add preliminary DKIM support. Currently requires a forked version of
480 ALT-N's libdkim that I have put here:
481 http://duncanthrax.net/exim-experimental/
482
483 Note to Michael Haardt: I had to rename some vars in sieve.c. They
484 were called 'true' and it seems that C99 defines that as a reserved
485 keyword to be used with 'bool' variable types. That means you could
486 not include C99-style headers which use bools without triggering
487 build errors in sieve.c.
488
81ea09ca
NM
489NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
490 as mailq or other aliases. Changed the --help handling significantly
491 to do whats expected. exim_usage() emits usage/help information.
492
f13cddcb
SC
493SC/01 Added the -bylocaldomain option to eximstats.
494
7d9f747b 495NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
8ad076b2 496
7d9f747b 497NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
a843aaa6 498
7d9f747b 499NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
47db1125
NM
500
501
eb4c0de6
PH
502Exim version 4.68
503-----------------
504
505PH/01 Another patch from the Sieve maintainer.
506
6a3bceb1
PH
507PH/02 When an IPv6 address is converted to a string for single-key lookup
508 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
509 dots are used instead of colons so that keys in lsearch files need not
510 contain colons. This was done some time before quoting was made available
511 in lsearch files. However, iplsearch files do require colons in IPv6 keys
512 (notated using the quote facility) so as to distinguish them from IPv4
513 keys. This meant that lookups for IP addresses in host lists did not work
514 for iplsearch lookups.
515
516 This has been fixed by arranging for IPv6 addresses to be expressed with
517 colons if the lookup type is iplsearch. This is not incompatible, because
518 previously such lookups could never work.
519
520 The situation is now rather anomolous, since one *can* have colons in
521 ordinary lsearch keys. However, making the change in all cases is
522 incompatible and would probably break a number of configurations.
523
2e30fa9d
TK
524TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
525 version.
526
0806a9c5
MH
527MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
528 conversion specification without a maximum field width, thereby enabling
529 a rogue spamd server to cause a buffer overflow. While nobody in their
530 right mind would setup Exim to query an untrusted spamd server, an
531 attacker that gains access to a server running spamd could potentially
532 exploit this vulnerability to run arbitrary code as the Exim user.
533
ae276964
TK
534TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
535 $primary_hostname instead of what libspf2 thinks the hosts name is.
536
0f2cbd1b
MH
537MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
538 a directory entry by the name of the lookup key. Previously, if a
539 symlink pointed to a non-existing file or a file in a directory that
540 Exim lacked permissions to read, a lookup for a key matching that
541 symlink would fail. Now it is enough that a matching directory entry
542 exists, symlink or not. (Bugzilla 503.)
543
2b85bce7
PH
544PH/03 The body_linecount and body_zerocount variables are now exported in the
545 local_scan API.
546
93655c46
PH
547PH/04 Added the $dnslist_matched variable.
548
6c512171
PH
549PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
550 This means they are set thereafter only if the connection becomes
551 encrypted.
552
553PH/06 Added the client_condition to authenticators so that some can be skipped
554 by clients under certain conditions.
555
aa6dc513
PH
556PH/07 The error message for a badly-placed control=no_multiline_responses left
557 "_responses" off the end of the name.
558
a96603a0
PH
559PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
560
8f240103
PH
561PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
562 (without spaces) instead of just copying the configuration text.
563
564PH/10 Added the /noupdate option to the ratelimit ACL condition.
565
d677b2f2
PH
566PH/11 Added $max_received_linelength.
567
d52120f2
PH
568PH/12 Added +ignore_defer and +include_defer to host lists.
569
64f2600a
PH
570PH/13 Installed PCRE version 7.2. This needed some changes because of the new
571 way in which PCRE > 7.0 is built.
572
8669f003
PH
573PH/14 Implemented queue_only_load_latch.
574
a4dc33a8
PH
575PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
576 MAIL command. The effect was to mangle the value on 64-bit systems.
577
d6a60c0f
PH
578PH/16 Another patch from the Sieve maintainer.
579
8f128379
PH
580PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
581
8932dffe
PH
582PH/18 If a system quota error occurred while trying to create the file for
583 a maildir delivery, the message "Mailbox is full" was not appended to the
584 bounce if the delivery eventually timed out. Change 4.67/27 below applied
585 only to a quota excession during the actual writing of the file.
d6a60c0f 586
ddea74fa 587PH/19 It seems that peer DN values may contain newlines (and other non-printing
48ed62d9
PH
588 characters?) which causes problems in log lines. The DN values are now
589 passed through string_printing() before being added to log lines.
590
ddea74fa 591PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
b7670459
PH
592 and InterBase are left for another time.)
593
ddea74fa
PH
594PH/21 Added message_body_newlines option.
595
ce9f225c
PH
596PH/22 Guard against possible overflow in moan_check_errorcopy().
597
19897d52
PH
598PH/23 POSIX allows open() to be a macro; guard against that.
599
bc64a74d
PH
600PH/24 If the recipient of an error message contained an @ in the local part
601 (suitably quoted, of course), incorrect values were put in $domain and
602 $local_part during the evaluation of errors_copy.
603
eb4c0de6 604
b4ed4da0
PH
605Exim version 4.67
606-----------------
607
22ad45c9
MH
608MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
609 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
610 Jan Srzednicki.
611
b4ed4da0
PH
612PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
613 issue a MAIL command.
614
431b7361
PH
615PH/02 In an ACL statement such as
616
617 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
618
619 if a client was not listed at all, or was listed with a value other than
620 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
621 the condition was not true (as it should be), so access was not denied.
622 The bug was that the ! inversion was incorrectly passed on to the second
623 item. This has been fixed.
624
625PH/03 Added additional dnslists conditions == and =& which are different from
626 = and & when the dns lookup returns more than one IP address.
627
83da1223
PH
628PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
629 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
630
54fc8428
PH
631PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
632 FSYNC, which compiles an option called disable_fsync that allows for
633 bypassing fsync(). The documentation is heavily laced with warnings.
634
34c5e8dd
SC
635SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
636
bbe15da8
PH
637PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
638 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
639 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
640 including adding "make clean"; (3) Added -fPIC when compiling the test
641 dynamically loaded module, to get rid of a warning.
642
0e8a9471
MH
643MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
644 message fails, move_frozen_messages = true and ignore_bounce_errors_after
645 = 0s. The bug is otherwise harmless.
646
f0872424
PH
647PH/07 There was a bug in the dovecot authenticator such that the value of
648 $auth1 could be overwritten, and so not correctly preserved, after a
649 successful authentication. This usually meant that the value preserved by
650 the server_setid option was incorrect.
651
b01dd148
PH
652PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
653
6bf342e1
PH
654PH/09 Installed PCRE release 7.0.
655
273f34d0
PH
656PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
657 run for batched SMTP input. It is now run at the start of every message
658 in the batch. While fixing this I discovered that the process information
659 (output by running exiwhat) was not always getting set for -bs and -bS
660 input. This is fixed, and it now also says "batched" for BSMTP.
661
cf8b11a5
PH
662PH/11 Added control=no_pipelining.
663
41c7c167
PH
664PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
665 patch, slightly modified), and move the expansion of helo_data till after
666 the connection is made in the smtp transport (so it can use these
667 values).
668
9c57cbc0
PH
669PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
670
f3f065bb
PH
671PH/14 Added log_selector = +pid.
672
047bdd8c
PH
673PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
674
0ce9abe6
PH
675PH/16 Add ${if forany and ${if forall.
676
0e22dfd1
PH
677PH/17 Added dsn_from option to vary the From: line in DSNs.
678
4c590bd1
PH
679PH/18 Flush SMTP output before performing a callout, unless control =
680 no_callout_flush is set.
681
09945f1e
PH
682PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
683 was true (the default) a successful delivery failed to delete the retry
684 item, thus causing premature timeout of the address. The bug is now
685 fixed.
686
c51b8e75
PH
687PH/20 Added hosts_avoid_pipelining to the smtp transport.
688
e28326d8
PH
689PH/21 Long custom messages for fakedefer and fakereject are now split up
690 into multiline reponses in the same way that messages for "deny" and
691 other ACL rejections are.
692
75b1493f
PH
693PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
694 with slight modification.
695
7c5214ec
PH
696PH/23 Applied sieve patches from the maintainer "tracking the latest notify
697 draft, changing the syntax and factoring some duplicate code".
698
4311097e
PH
699PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
700 for deliveries of the second and subsequent messages over the same SMTP
701 connection.
702
29f89cad
PH
703PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
704 ${reduce, with only minor "tidies".
705
5e687460
SC
706SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
707
c3611384
PH
708PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
709 expansion side effects.
710
5a11a7b4
PH
711PH/27 When a message times out after an over-quota error from an Exim-imposed
712 quota, the bounce message says "mailbox is full". This message was not
713 being given when it was a system quota that was exceeded. It now should
714 be the same.
715
0e20aff9
MH
716MH/03 Made $recipients available in local_scan(). local_scan() already has
717 better access to the recipient list through recipients_list[], but
718 $recipients can be useful in postmaster-provided expansion strings.
719
ca86f471
PH
720PH/28 The $smtp_command and $smtp_command_argument variables were not correct
721 in the case of a MAIL command with additional options following the
722 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
723 were accidentally chopped off.
724
a14e5636
PH
725PH/29 SMTP synchronization checks are implemented when a command is read -
726 there is a check that no more input is waiting when there shouldn't be
727 any. However, for some commands, a delay in an ACL can mean that it is
728 some time before the response is written. In this time, more input might
729 arrive, invalidly. So now there are extra checks after an ACL has run for
730 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
731 pipelining has not been advertised.
732
ec95d1a6
PH
733PH/30 MH's patch to allow iscntrl() characters to be list separators.
734
42855d71
PH
735PH/31 Unlike :fail:, a custom message specified with :defer: was not being
736 returned in the SMTP response when smtp_return_error_details was false.
737 This has been fixed.
738
57c2c631
PH
739PH/32 Change the Dovecot authenticator to use read() and write() on the socket
740 instead of the C I/O that was originally supplied, because problems were
741 reported on Solaris.
742
58c01c94
PH
743PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
744 Exim which did not show up earlier: it was assuming that a call to
745 SSL_CTX_set_info_callback() might give an error value. In fact, there is
746 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
747 was a macro that became an assignment, so it seemed to work. This has
748 changed to a proper function call with a void return, hence the compile
749 error. Exim's code has been fixed.
750
dee5a20a
PH
751PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
752 cpus.
753
d2ee6114
PH
754PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
755
b2d5182b
PH
756PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
757
79749a79
PH
758PH/37 If a message is not accepted after it has had an id assigned (e.g.
759 because it turns out to be too big or there is a timeout) there is no
3ce62588
PH
760 "Completed" line in the log. When some messages of this type were
761 selected by exigrep, they were listed as "not completed". Others were
762 picked up by some special patterns. I have improved the selection
763 criteria to be more general.
79749a79 764
c456d9bb
PH
765PH/38 The host_find_failed option in the manualroute router can now be set
766 to "ignore", to completely ignore a host whose IP address cannot be
767 found. If all hosts are ignored, the behaviour is controlled by the new
768 host_all_ignored option.
769
cd9868ec
PH
770PH/39 In a list of hosts for manualroute, if one item (either because of multi-
771 homing or because of multiple MX records with /mx) generated more than
772 one IP address, and the following item turned out to be the local host,
773 all the secondary addresses of the first item were incorrectly removed
774 from the list, along with the local host and any following hosts (which
775 is what is supposed to happen).
776
ebeaf996
PH
777PH/40 When Exim receives a message, it writes the login name, uid, and gid of
778 whoever called Exim into the -H file. In the case of the daemon it was
779 behaving confusingly. When first started, it used values for whoever
780 started the daemon, but after a SIGHUP it used the Exim user (because it
781 calls itself on a restart). I have changed the code so that it now always
782 uses the Exim user.
783
2679d413
PH
784PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
785 message are rejected with the same error (e.g. no authentication or bad
786 sender address), and a DATA command is nevertheless sent (as can happen
787 with PIPELINING or a stupid MUA), the error message that was given to the
788 RCPT commands is included in the rejection of the DATA command. This is
789 intended to be helpful for MUAs that show only the final error to their
790 users.
791
84024b72
PH
792PH/42 Another patch from the Sieve maintainer.
793
8005d38e
SC
794SC/02 Eximstats - Differentiate between permanent and temporary rejects.
795 Eximstats - Fixed some broken HTML links and added missing column headers
796 (Jez Hancock).
797 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
798 columns for Rejects, Temp Rejects, Ham, and Spam rows.
799
3298c6c6
SC
800SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
801
a43a27c5
PH
802PH/43 Yet another patch from the Sieve maintainer.
803
58eb016e 804PH/44 I found a way to check for a TCP/IP connection going away before sending
563b63fa
PH
805 the response to the final '.' that terminates a message, but only in the
806 case where the client has not sent further data following the '.'
807 (unfortunately, this is allowed). However, in many cases there won't be
808 any further data because there won't be any more messages to send. A call
809 to select() can be used: if it shows that the input is "ready", there is
810 either input waiting, or the socket has been closed. An attempt to read
811 the next input character can distinguish the two cases. Previously, Exim
58eb016e 812 would have sent an OK response which the client would never have see.
563b63fa
PH
813 This could lead to message repetition. This fix should cure that, at
814 least in a lot of common cases.
58eb016e 815
b43a74ea
PH
816PH/45 Do not advertise STARTTLS in response to HELP unless it would be
817 advertised in response to EHLO.
818
b4ed4da0 819
5dd1517f
PH
820Exim version 4.66
821-----------------
822
823PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
824 fixed by 4.65/MH/01 (is this a record?) are fixed:
825
826 (i) An empty string was always treated as zero by the numeric comparison
827 operators. This behaviour has been restored.
828
829 (ii) It is documented that the numeric comparison operators always treat
830 their arguments as decimal numbers. This was broken in that numbers
831 starting with 0 were being interpreted as octal.
832
833 While fixing these problems I realized that there was another issue that
834 hadn't been noticed. Values of message_size_limit (both the global option
835 and the transport option) were treated as octal if they started with 0.
836 The documentation was vague. These values are now always treated as
837 decimal, and I will make that clear in the documentation.
838
839
93cfa765
TK
840Exim version 4.65
841-----------------
842
843TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
844 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
845 versions. (#438)
846
d6066548
MH
847MH/01 Don't check that the operands of numeric comparison operators are
848 integers when their expansion is in "skipping" mode (fixes bug
849 introduced by 4.64-PH/07).
850
4362ff0d
PH
851PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
852 child addresses, Exim now panics and dies. Previously, because the count
853 is held in a short int, deliveries were likely to be lost. As such a
854 large number of recipients for a single message is ridiculous
855 (performance will be very, very poor), I have chosen to impose a limit
856 rather than extend the field.
857
93cfa765 858
944e9e9c
TF
859Exim version 4.64
860-----------------
aa41d2de 861
21d74bd9
TK
862TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
863 leftover -K file (the existence of which was triggered by #402).
864 While we were at it, introduced process PID as part of the -K
865 filename. This should rule out race conditions when creating
866 these files.
867
868TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
869 processing considerably. Previous code took too long for large mails,
870 triggering a timeout which in turn triggers #401.
871
872TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
873 in the DK code in transports.c. sendfile() is not really portable,
874 hence the _LINUX specificness.
944e9e9c
TF
875
876TF/01 In the add_headers option to the mail command in an Exim filter,
877 there was a bug that Exim would claim a syntax error in any
878 header after the first one which had an odd number of characters
879 in the field name.
880
2b1c6e3a
PH
881PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
882 callout verification, Exim cached a "reject" for the entire domain. This
883 is correct for most verifications, but it is not correct for a recipient
884 verification with use_sender or use_postmaster set, because in that case
885 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
886 case of MAIL FROM:<> rejection from other early rejections (e.g.
887 rejection of HELO). When verifying a recipient using a non-null MAIL
888 address, the cache is ignored if it shows MAIL FROM:<> rejection.
889 Whatever the result of the callout, the value of the domain cache is
890 left unchanged (for any other kind of callout, getting as far as trying
891 RCPT means that the domain itself is ok).
892
1f872c80
PH
893PH/02 Tidied a number of unused variable and signed/unsigned warnings that
894 gcc 4.1.1 threw up.
895
896PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
897 manifest itself as EPIPE rather than ECONNECT. When tidying away a
898 session, the daemon ignores ECONNECT errors and logs others; it now
899 ignores EPIPE as well.
900
d203e649
PH
901PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
902 (quoted-printable decoding).
903
cc2ed8f7 904PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
21a04aa3 905 later the small subsequent patch to fix an introduced bug.
f951fd57 906
ddfcd446
PH
907PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
908
d45b1de8
PH
909PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
910
911PH/08 An error is now given if message_size_limit is specified negative.
912
38a0a95f 913PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
641cb756 914 to be given (somewhat) arbitrary names.
38a0a95f 915
a2405d83
JJ
916JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
917 in 4.64-PH/09.
918
919JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
920 miscellaneous code fixes
921
6ea85e9a
PH
922PH/10 Added the log_reject_target ACL modifier to specify where to log
923 rejections.
924
26da7e20
PH
925PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
926 hostname. This is wrong, because it relates to the incoming message (and
927 probably the interface on which it is arriving) and not to the outgoing
928 callout (which could be using a different interface). This has been
929 changed to use the value of the helo_data option from the smtp transport
930 instead - this is what is used when a message is actually being sent. If
931 there is no remote transport (possible with a router that sets up host
932 addresses), $smtp_active_hostname is used.
6ea85e9a 933
14aa5a05 934PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
7befa435 935 tweaks were necessary in order to get it to work (see also 21 below):
14aa5a05
PH
936 (a) The code assumed that strncpy() returns a negative number on buffer
937 overflow, which isn't the case. Replaced with Exim's string_format()
938 function.
939 (b) There were several signed/unsigned issues. I just did the minimum
940 hacking in of casts. There is scope for a larger refactoring.
941 (c) The code used strcasecmp() which is not a standard C function.
942 Replaced with Exim's strcmpic() function.
943 (d) The code set only $1; it now sets $auth1 as well.
944 (e) A simple test gave the error "authentication client didn't specify
945 service in request". It would seem that Dovecot has changed its
946 interface. Fortunately there's a specification; I followed it and
947 changed what the client sends and it appears to be working now.
948
ff75a1f7
PH
949PH/13 Added $message_headers_raw to provide the headers without RFC 2047
950 decoding.
951
e6f6568e
PH
952PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
953 address A is aliased to B and C, where B exists and C does not. Without
954 -v the output is "A verified" because verification stops after a
955 successful redirection if more than one address is generated. However,
956 with -v the child addresses are also verified. Exim was outputting "A
957 failed to verify" and then showing the successful verification for C,
958 with its parentage. It now outputs "B failed to verify", showing B's
959 parentage before showing the successful verification of C.
960
d6f6e0dc
PH
961PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
962 look up a TXT record in a specific list after matching in a combined
963 list.
964
322050c2
PH
965PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
966 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
967 they consult the DNS. I had assumed they would set it the way they
968 wanted; and indeed my experiments on Linux seem to show that in some
969 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
970 To be on the safe side, however, I have now made the interface to
971 host_find_byname() similar to host_find_bydns(), with an argument
972 containing the DNS resolver options. The host_find_byname() function now
973 sets these options at its start, just as host_find_bydns() does. The smtp
974 transport options dns_qualify_single and dns_search_parents are passed to
975 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
976 of host_find_byname() use the default settings of RES_DEFNAMES
977 (qualify_single) but not RES_DNSRCH (search_parents).
978
08955dd3
PH
979PH/17 Applied (a modified version of) Nico Erfurth's patch to make
980 spool_read_header() do less string testing, by means of a preliminary
981 switch on the second character of optional "-foo" lines. (This is
982 overdue, caused by the large number of possibilities that now exist.
983 Originally there were few.) While I was there, I also converted the
984 str(n)cmp tests so they don't re-test the leading "-" and the first
985 character, in the hope this might squeeze out yet more improvement.
986
1eccaa59
PH
987PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
988 flag allowing group syntax was set by the header_syntax check but not
989 turned off, possible causing trouble later; (2) The flag was not being
990 set at all for the header_verify test, causing "group"-style headers to
991 be rejected. I have now set it in this case, and also caused header_
992 verify to ignore an empty address taken from a group. While doing this, I
993 came across some other cases where the code for allowing group syntax
994 while scanning a header line wasn't quite right (mostly, not resetting
995 the flag correctly in the right place). These bugs could have caused
996 trouble for malformed header lines. I hope it is now all correct.
997
602e59e5
PH
998PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
999 with the "reply" argument non-NULL. The code, however (which originally
1000 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1001 but it didn't always do it. This confused somebody who was copying the
1002 code for some other use. I have removed all the tests.
1003
411ef850
PH
1004PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1005 feature that was used to support insecure browsers during the U.S. crypto
1006 embargo. It requires special client support, and Exim is probably the
1007 only MTA that supported it -- and would never use it because real RSA is
1008 always available. This code has been removed, because it had the bad
1009 effect of slowing Exim down by computing (never used) parameters for the
1010 RSA_EXPORT functionality.
1011
7befa435
PH
1012PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1013 authenticator to fail if there's a tab character in the incoming data
1014 (there should never be unless someone is messing about, as it's supposed
1015 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1016 if the connection is using TLS or if the remote IP is the same as the
1017 local IP, and the "valid-client-cert option" if a client certificate has
1018 been verified.
1019
48da4259 1020PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
16ff981e
PH
1021 authenticators. This can be used for authorization after authentication
1022 succeeds. (In the case of plaintext, it servers for both authentication
1023 and authorization.)
1024
48da4259
PH
1025PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1026 if any retry times were supplied.
1027
d1d5595c
PH
1028PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1029 connection, where there is no client IP address to check. In this
1030 situation, the verify now always succeeds.
1031
0ef732d9
PH
1032PH/25 Applied John Jetmore's -Mset patch.
1033
328895cc
PH
1034PH/26 Added -bem to be like -Mset, but loading a message from a file.
1035
fd700877
PH
1036PH/27 In a string expansion for a processed (not raw) header when multiple
1037 headers of the same name were present, leading whitespace was being
1038 removed from all of them, but trailing whitespace was being removed only
1039 from the last one. Now trailing whitespace is removed from each header
f6c332bd
PH
1040 before concatenation. Completely empty headers in a concatenation (as
1041 before) are ignored.
fd700877 1042
8dce1a6f
PH
1043PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1044 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1045
17af4a17
PH
1046PH/29 [Removed. This was a change that I later backed out, and forgot to
1047 correct the ChangeLog entry (that I had efficiently created) before
1048 committing the later change.]
f6c332bd
PH
1049
1050PH/30 Exim was sometimes attempting to deliver messages that had suffered
1051 address errors (4xx response to RCPT) over the same connection as other
1052 messages routed to the same hosts. Such deliveries are always "forced",
1053 so retry times are not inspected. This resulted in far too many retries
1054 for the affected addresses. The effect occurred only when there were more
1055 hosts than the hosts_max_try setting in the smtp transport when it had
1056 the 4xx errors. Those hosts that it had tried were not added to the list
1057 of hosts for which the message was waiting, so if all were tried, there
1058 was no problem. Two fixes have been applied:
1059
1060 (i) If there are any address or message errors in an SMTP delivery, none
1061 of the hosts (tried or untried) are now added to the list of hosts
1062 for which the message is waiting, so the message should not be a
1063 candidate for sending over the same connection that was used for a
1064 successful delivery of some other message. This seems entirely
1065 reasonable: after all the message is NOT "waiting for some host".
1066 This is so "obvious" that I'm not sure why it wasn't done
1067 previously. Hope I haven't missed anything, but it can't do any
1068 harm, as the worst effect is to miss an optimization.
1069
1070 (ii) If, despite (i), such a delivery is accidentally attempted, the
1071 routing retry time is respected, so at least it doesn't keep
1072 hammering the server.
1073
c1114884
PH
1074PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1075 in ${readsocket because some servers need this prod.
1076
7a0743eb
PH
1077PH/32 Added some extra debug output when updating a wait-xxx database.
1078
0d85fa3f
PH
1079PH/33 The hint "could be header name not terminated by colon", which has been
1080 given for certain expansion errors for a long time, was not being given
1081 for the ${if def:h_colon_omitted{... case.
1082
1bf43b78
PH
1083PH/34 The spec says: "With one important exception, whenever a domain list is
1084 being scanned, $domain contains the subject domain." There was at least
1085 one case where this was not true.
1086
520de300
PH
1087PH/35 The error "getsockname() failed: connection reset by peer" was being
1088 written to the panic log as well as the main log, but it isn't really
1089 panic-worthy as it just means the connection died rather early on. I have
1090 removed the panic log writing for the ECONNRESET error when getsockname()
1091 fails.
1092
48c7f9e2
PH
1093PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1094 runs only) independently of the message's sender address. This meant
1095 that, if the 4xx error was in fact related to the sender, a different
1096 message to the same recipient with a different sender could confuse
1097 things. In particualar, this can happen when sending to a greylisting
1098 server, but other circumstances could also provoke similar problems.
1099 I have changed the default so that the retry time for these errors is now
1100 based a combination of the sender and recipient addresses. This change
1101 can be overridden by setting address_retry_include_sender=false in the
1102 smtp transport.
1103
99ea1c86
PH
1104PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1105 remote server are returned as part of bounce messages. This was not
1106 happening for LMTP over a pipe (the lmtp transport), but now it is the
1107 same for both kinds of LMTP.
1108
a2042e78
PH
1109PH/38 Despite being documented as not happening, Exim was rewriting addresses
1110 in header lines that were in fact CNAMEs. This is no longer the case.
1111
4fbcfc2e
PH
1112PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1113 and queue runs started by the daemon processed all messages. This has
1114 been fixed so that -R and -S can now usefully be given with -q<time>.
1115
aa41d2de
PH
1116PH/40 Import PCRE release 6.7 (fixes some bugs).
1117
af561417
PH
1118PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1119
3cc66b45
PH
1120PH/42 Give an error if -q is specified more than once.
1121
194cc0e4
PH
1122PH/43 Renamed the variables $interface_address and $interface_port as
1123 $received_ip_address and $received_port, to make it clear that these
1124 values apply to message reception, and not to the outgoing interface when
1125 a message is delivered. (The old names remain recognized, of course.)
1126
a401ddaa
PH
1127PH/44 There was no timeout on the connect() call when using a Unix domain
1128 socket in the ${readsocket expansion. There now is.
1129
4e88a19f
PH
1130PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1131 be meaningful with "accept".
1132
d7d7b289
SC
1133SC/01 Eximstats V1.43
1134 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1135
1136SC/02 Eximstats V1.44
1137 Use a glob alias rather than an array ref in the generated
1138 parser. This improves both readability and performance.
1139
1140SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1141 Collect SpamAssassin and rejection statistics.
1142 Don't display local sender or destination tables unless
1143 there is data to show.
1144 Added average volumes into the top table text output.
1145
1146SC/04 Eximstats V1.46
1147 Collect data on the number of addresses (recipients)
1148 as well as the number of messages.
1149
1150SC/05 Eximstats V1.47
1151 Added 'Message too big' to the list of mail rejection
1152 reasons (thanks to Marco Gaiarin).
1153
1154SC/06 Eximstats V1.48
1155 Mainlog lines which have GMT offsets and are too short to
1156 have a flag are now skipped.
1157
1158SC/07 Eximstats V1.49 (Alain Williams)
1159 Added the -emptyok flag.
1160
1161SC/08 Eximstats V1.50
1162 Fixes for obtaining the IP address from reject messages.
1163
0ea2a468
JJ
1164JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1165 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1166 whitesspace changes from 4.64-PH/27
1167
1168JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1169 match 4.64-PH/13
1170
1171JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1172 are found, allow negative numbers in numeric criteria)
1173
1174JJ/06 exipick.20061117.2, added new $message_body_missing variable
1175
1176JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1177 to match changes made in 4.64-PH/43
1178
8a10f5a4
PH
1179PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1180
30e18802
PH
1181PH/47 Put in an explicit test for a DNS lookup of an address record where the
1182 "domain" is actually an IP address, and force a failure. This locks out
1183 those revolvers/nameservers that support "A-for-A" lookups, in
1184 contravention of the specifications.
1185
55728a4f
PH
1186PH/48 When a host name was looked up from an IP address, and the subsequent
1187 forward lookup of the name timed out, the host name was left in
1188 $sender_host_name, contrary to the specification.
d7d7b289 1189
d7837193
PH
1190PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1191 restricted to single-key lookups, Exim was not diagnosing an error if
1192 * or *@ was used with a query-style lookup.
1193
87054a31
PH
1194PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1195
ea2c01d2
MH
1196MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1197 long ago, but noone interested enough thought of it. Let's just say that
1198 the "1.1" means that there are some new functions that weren't there at
1199 some point in the past.
1200
e4fa6968
PH
1201PH/51 Error processing for expansion failure of helo_data from an smtp
1202 transport during callout processing was broken.
1203
56f5d9bd
PH
1204PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1205 tested/used via the -bh/-bhc/-bs options.
1206
922e1c28
PH
1207PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1208 bug, fixed in subsequent PCRE releases).
1209
21eb6e72
PH
1210PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1211 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1212
a0540757
PH
1213PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1214
944e9e9c 1215
478be7b0
SC
1216Exim version 4.63
1217-----------------
1218
1219SC/01 Use a glob alias rather than an array ref in eximstats generated
1220 parser. This improves both readability and performance.
1221
1222SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1223 Don't display local sender or destination tables in eximstats unless
1224 there is data to show.
1225 Added average volumes into the eximstats top table text output.
1226
1227SC/03 Collect data on the number of addresses (recipients) as well
1228 as the number of messages in eximstats.
1229
2b965a65
TF
1230TF/01 Correct an error in the documentation for the redirect router. Exim
1231 does (usually) call initgroups() when daemonizing.
478be7b0 1232
45b91596
PH
1233TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1234 with consistent privilege compared to when running as a daemon.
478be7b0 1235
c59f5781
TF
1236TF/03 Note in the spec that $authenticated_id is not set for local
1237 submissions from trusted users.
1238
90fc3069
TF
1239TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1240 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1241
6083aca0
TF
1242TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1243 by adding some example configuration directives to the default
1244 configuration file. A little bit of work is required to uncomment the
1245 directives and define how usernames and passwords are checked, but
1246 there is now a framework to start from.
1247
765b530f
PH
1248PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1249 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1250 without this. I don't know how relevant this is to other LDAP libraries.
1251
4e167a8c
PH
1252PH/02 Add the verb name to the "unknown ACL verb" error.
1253
4608d683
PH
1254PH/03 Magnus Holmgren's patch for filter_prepend_home.
1255
b8dc3e4a
PH
1256PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1257
5418e93b
PH
1258PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1259 directory not expanded when it should be if an expanded home directory
1260 was set for the address (which is overridden by the transport).
1261
b4a9bda2
PH
1262PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1263 libradius.
1264
45b91596
PH
1265PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1266 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1267 because it is too late at that time, and has no effect.
1268
5547e2c5
PH
1269PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1270 security issue with \' (bugzilla #107). I could not use the
1271 PQescapeStringConn() function, because it needs a PGconn value as one of
1272 its arguments.
1273
dbcef0ea
PH
1274PH/08 When testing addresses using -bt, indicate those final addresses that
1275 are duplicates that would not cause an additional delivery. At least one
1276 person was confused, thinking that -bt output corresponded to deliveries.
1277 (Suppressing duplicates isn't a good idea as you lose the information
1278 about possibly different redirections that led to the duplicates.)
1279
25257489
PH
1280PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1281 systems where poll() doesn't work, in particular OS X.
1282
c816d124
PH
1283PH/10 Added more information to debugging output for retry time not reached.
1284
a9ccd69a
PH
1285PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1286 operations in malware.c.
1287
75fa1910
PH
1288PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1289 signatures.
1290
a7d7aa58
PH
1291PH/13 If write_rejectlog was set false when logging was sent to syslog with
1292 syslog_duplication set false, log lines that would normally be written
1293 both the the main log and to the reject log were not written to syslog at
1294 all.
1295
42119b09
PH
1296PH/14 In the default configuration, change the use of "message" in ACL warn
1297 statements to "add_header".
1298
41609df5
PH
1299PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1300 not followed by a command (e.g. "seen endif").
1301
a5bd321b
PH
1302PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1303 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1304 latter.
1305
e85a7ad5 1306PH/17 Added extra conditions to the default value of delay_warning_condition
5dff5817
PH
1307 so that it is now:
1308
e85a7ad5
PH
1309 ${if or { \
1310 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1311 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1312 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
5dff5817
PH
1313 }{no}{yes}}
1314
e85a7ad5
PH
1315 The Auto-Submitted: and various List- headers are standardised, whereas I
1316 don't think Precedence: ever was.
5dff5817 1317
d8fe1c03
PH
1318PH/18 Refactored debugging code in route_finduser() to show more information,
1319 in particular, the error code if getpwnam() issues one.
1320
16282d2b
PH
1321PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1322 This is apparently needed in addition to the PH/07 change above to avoid
1323 any possible encoding problems.
1324
35d40a98
PH
1325PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1326 but not after initializing Perl.
1327
034d99ab
PH
1328PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1329 output them only if debugging. By default they are written stderr,
1330 apparently, which is not desirable.
1331
6ec97b1b
PH
1332PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1333 queries.
1334
e22ca4ac
JJ
1335JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1336 --not options
1337
1338JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1339
33d73e3b
PH
1340PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1341 authenticated or an ident call has been made. Suppress the default
1342 values for $authenticated_id and $authenticated_sender (but permit -oMai
1343 and -oMas) when testing with -bh.
1344
9ecb03f3
PH
1345PH/24 Re-jigged the order of the tests in the default configuration so that the
1346 tests for valid domains and recipients precede the DNS black list and CSA
1347 tests, on the grounds that those ones are more expensive.
1348
084efe8d
PH
1349PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1350 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1351 This bug exists in every version of Exim that I still have, right back to
1352 0.12.
1353
366fc9f0
PH
1354PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1355 However, an attempt to turn on case-sensitivity in a regex key by
1356 including (?-i) didn't work because the subject string was already
1357 lowercased, and the effects were non-intuitive. It turns out that a
1358 one-line patch can be used to allow (?-i) to work as expected.
1359
c59f5781 1360
c887c79e
TF
1361Exim version 4.62
1362-----------------
1363
1364TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1365 other effects) broke the use of negated acl sub-conditions.
1366
1cce3af8
PH
1367PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1368 patch).
1369
afb3eaaf
PH
1370PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1371 "Deny" causes Exim to reject the incoming connection with a 554 error.
1372 Unfortunately, if there is a major crisis, such as a disk failure,
1373 tcp-wrappers gives "deny", whereas what one would like would be some
1374 kind of temporary error. A kludge has been added to help with this.
1375 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1376 554 error is used if errno is still zero or contains ENOENT (which occurs
1377 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1378 451 error is used.
1379
e173618b
PH
1380PH/03 Add -lutil to the default FreeBSD LIBS setting.
1381
dd16e114
PH
1382PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1383 errors. Otherwise a message that provokes a temporary error (when other
1384 messages do not) can cause a whole host to time out.
1385
f7fd3850
PH
1386PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1387 addresses were routed directly to files or pipes from a redirect router.
1388 File deliveries just didn't batch; pipe deliveries might have suffered
1389 odd errors.
1390
d87df92c
PH
1391PH/06 A failure to get a lock for a hints database would erroneously always say
1392 "Failed to get write lock", even when it was really a read lock.
1393
7e9f683d
PH
1394PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1395 of 0600. This has been changed to use the value of the lockfile_mode
1396 option (which defaults to 0600).
1397
bfad5236
PH
1398PH/08 Applied small patch from the Sieve maintainer.
1399
01c490df
PH
1400PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1401 folder from quota calculations, a direct delivery into this folder messed
1402 up the contents of the maildirsize file. This was because the regex was
1403 used only to exclude .Trash (or whatever) when the size of the mailbox
1404 was calculated. There was no check that a delivery was happening into an
1405 excluded directory. This bug has been fixed by ignoring all quota
1406 processing for deliveries into excluded directories.
1407
d6629cdc
PH
1408PH/10 Added the maildirfolder_create_regex option to appendfile.
1409
1cce3af8 1410
214e2000
PH
1411Exim version 4.61
1412-----------------
1413
1414PH/01 The code for finding all the local interface addresses on a FreeBSD
1415 system running IPv6 was broken. This may well have applied to all BSD
1416 systems, as well as to others that have similar system calls. The broken
1417 code found IPv4 interfaces correctly, but gave incorrect values for the
1418 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1419 that it would not match correctly against @[] and not recognize the IPv6
1420 addresses as local.
1421
f9daeae0
PH
1422PH/02 The ipliteral router was not recognizing addresses of the form user@
1423 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1424
7e66e54d
PH
1425PH/03 Added disable_ipv6.
1426
c8ea1597
PH
1427PH/04 Changed $reply_address to use the raw form of the headers instead of the
1428 decoded form, because it is most often used to construct To: headers
1429 lines in autoreplies, and the decoded form may well be syntactically
1430 invalid. However, $reply_address has leading white space removed, and all
1431 newlines turned into spaces so that the autoreply transport does not
1432 grumble.
1433
911f6fde
PH
1434PH/05 If group was specified without a user on a router, and no group or user
1435 was specified on a transport, the group from the router was ignored.
1436
47ca6d6c
PH
1437PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1438 for visible compile-time settings that can be used to change these
1439 numbers, for those that want even more. Backwards compatibility with old
1440 spool files has been maintained. However, going back to a previous Exim
1441 release will lost any variables that are in spool files.
1442
ed0e9820
PH
1443PH/07 Two small changes when running in the test harness: increase delay when
1444 passing a TCP/IP connection to a new process, in case the original
1445 process has to generate a bounce, and remove special handling of
1446 127.0.0.2 (sic), which is no longer necessary.
1447
eff37e47
PH
1448PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1449 be the same on different OS.
1450
1921d2ea
PH
1451PH/09 Moved a debug statement in filter processing to avoid a race problem when
1452 testing.
1453
b3f69ca8
JJ
1454JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1455 whether --show-vars was specified or not
1456
1457JJ/02 exipick: Added support for new ACL variable spool format introduced
1458 in 4.61-PH/06
1459
424a1c63
PH
1460PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1461 syntactically invalid From: or Reply-to: line, and a filter used this to
1462 generate an autoreply, and therefore failed to obtain an address for the
1463 autoreply, Exim could try to deliver to a non-existent relative file
1464 name, causing unrelated and misleading errors. What now happens is that
1465 it logs this as a hard delivery error, but does not attempt to create a
1466 bounce message.
1467
7a100415
PH
1468PH/11 The exinext utility has a -C option for testing purposes, but although
1469 the given file was scanned by exinext itself; it wasn't being passed on
1470 when Exim was called.
1471
19b9dc85
PH
1472PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1473 an end-of-file indication when reading a command response.
1474
309bd837
PH
1475PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1476 compiled. In many other places in Exim, IPv6 addresses are always
1477 recognized, so I have changed this. It also means that IPv4 domain
1478 literals of the form [IPV4:n.n.n.n] are now always recognized.
1479
59e82a2a
PH
1480PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1481 used if the router is not running as root, for example, when verifying at
1482 ACL time, or when using -bh. The debugging output from this situation was
1483 non-existent - all you got was a failure to exec. I have made two
1484 changes:
1485
1486 (a) Failures to set uid/gid, the current directory, or a process leader
1487 in a subprocess such as that created by queryprogram now generate
1488 suitable debugging ouput when -d is set.
1489
1490 (b) The queryprogram router detects when it is not running as root,
1491 outputs suitable debugging information if -d is set, and then runs
1492 the subprocess without attempting to change uid/gid.
1493
9edc04ce
PH
1494PH/15 Minor change to Makefile for building test_host (undocumented testing
1495 feature).
1496
1349e1e5
PH
1497PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1498 additional section of a DNS packet that returns MX or SRV records.
1499 Instead, it always explicitly searches for A/AAAA records. This avoids
1500 major problems that occur when a DNS server includes only records of one
1501 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1502 fixed another bug: if SRV records were looked up and the corresponding
1503 address records were *not* found in the additional section, the port
1504 values from the SRV records were lost.
1505
ea49d0e1
PH
1506PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1507 using the correct key (the original address) when searching the retry
1508 rules in order to find which one to use for generating the retry hint.
1509
064a94c9
PH
1510PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1511 adding the default one. Similarly, if it contains a Reply-To: header, the
1512 errors_reply_to option, if set, is not used.
1513
727071f8
PH
1514PH/19 When calculating a retry time, Exim used to measure the "time since
1515 failure" by looking at the "first failed" field in the retry record. Now
1516 it does not use this if it is later than than the arrival time of the
1517 message. Instead it uses the arrival time. This makes for better
1518 behaviour in cases where some deliveries succeed, thus re-setting the
1519 "first failed" field. An example is a quota failure for a huge message
1520 when small messages continue to be delivered. Without this change, the
1521 "time since failure" will always be short, possible causing more frequent
1522 delivery attempts for the huge message than are intended.
dd16e114 1523 [Note: This change was subsequently modified - see PH/04 for 4.62.]
727071f8 1524
f78eb7c6
PH
1525PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1526 $1, $2, $3) because the numerical variables can be reset during some
1527 expansion items (e.g. "match"), thereby losing the authentication data.
1528
21c28500
PH
1529PH/21 Make -bV show the size of off_t variables so that the test suite can
1530 decide whether to run tests for quotas > 2G.
1531
1532PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1533 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1534 filecount value is greater than 2G or if a quota value is greater than 2G
1535 on a system where the size of off_t is not greater than 4, a panic error
1536 is given.
1537
1688f43b
PH
1538PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1539 never match. The debug and -bh output now contains an explicit error
1540 message indicating a malformed IPv4 address or mask.
1541
1542PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1543 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1544 PH/23 above applies.
1545
9675b384
PH
1546PH/25 Do not write to syslog when running in the test harness. The only
1547 occasion when this arises is a failure to open the main or panic logs
1548 (for which there is an explicit test).
1549
6a3f1455
PH
1550PH/26 Added the /no_tell option to "control=freeze".
1551
dac79d3e
PH
1552PH/27 If a host name lookup failed very early in a connection, for example, if
1553 the IP address matched host_lookup and the reverse lookup yielded a name
1554 that did not have a forward lookup, an error message of the form "no IP
1555 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1556 could be logged. Now it outputs the IP address instead of "NULL".
1349e1e5 1557
5977a0b3
PH
1558PH/28 An enabling patch from MH: add new function child_open_exim2() which
1559 allows the sender and the authenticated sender to be set when
1560 submitting a message from within Exim. Since child_open_exim() is
1561 documented for local_scan(), the new function should be too.
1562
c91535f3
PH
1563PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1564 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1565 results in an empty string is now treated as unset.
1566
0d46a8c8
PH
1567PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1568
278c6e6c
PH
1569PH/31 Added sender_verify_fail logging option.
1570
2cbb4081
PH
1571PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1572 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1573 tidied the source and removed it altogether.
1574
3eef829e
PH
1575PH/33 When a queue run was abandoned because the load average was too high, a
1576 log line was always written; now it is written only if the queue_run log
1577 selector is set. In addition, the log line for abandonment now contains
1578 information about the queue run such as the pid. This is always present
1579 in "start" and "stop" lines but was omitted from the "abandon" line.
1580
1ab95fa6
PH
1581PH/34 Omit spaces between a header name and the colon in the error message that
1582 is given when verify = headers_syntax fails (if there are lots of them,
1583 the message gets confusing).
1584
230205fc
PH
1585PH/35 Change the default for dns_check_names_pattern to allow slashes within
1586 names, as there are now some PTR records that contain slashes. This check
1587 is only to protect against broken name servers that fall over on strange
1588 characters, so the fact that it applies to all lookups doesn't matter.
1589
75e0e026
PH
1590PH/36 Now that the new test suite is complete, we can remove some of the
1591 special code in Exim that was needed for the old test suite. For example,
1592 sorting DNS records because real resolvers return them in an arbitrary
1593 order. The new test suite's fake resolver always returns records in the
1594 same order.
1595
1596PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1597 bounces) except when queue_only is set, to avoid logging races between
1598 the different processes.
1599
145396a6
PH
1600PH/38 Panic-die if .include specifies a non-absolute path.
1601
3cd34f13
PH
1602PH/39 A tweak to the "H" retry rule from its user.
1603
11121d3d
JJ
1604JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1605 a label. They prevented compilation on older perls.
1606
1607JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1608 a warning to be raised on newish perls.
1609
1610JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1611 on queue. Changes to match documented behaviour of showing count of
1612 messages matching specified criteria.
1613
8def5aaf
PH
1614PH/40 Changed the default ident timeout from 30s to 5s.
1615
929ba01c
PH
1616PH/41 Added support for the use of login_cap features, on those BSD systems
1617 that have them, for controlling the resources used by pipe deliveries.
1618
2632889e
PH
1619PH/42 The content-scanning code uses fopen() to create files in which to put
1620 message data. Previously it was not paying any attention to the mode of
1621 the files. Exim runs with umask(0) because the rest of the code creates
1622 files with open(), and sets the required mode explicitly. Thus, these
1623 files were ending up world-writeable. This was not a big issue, because,
1624 being within the spool directory, they were not world-accessible. I have
1625 created a function called modefopen, which takes an additional mode
1626 argument. It sets umask(777), creates the file, chmods it to the required
1627 mode, then resets the umask. All the relevant calls to fopen() in the
1628 content scanning code have been changed to use this function.
1629
944a9c55
PH
1630PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1631 to 24 hours. This avoids potential overflow problems when processing G
1632 and H retry rules. I suspect nobody ever tinkers with this value.
1633
4a23603b
PH
1634PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1635
4730f942
PH
1636PH/45 When the plaintext authenticator is running as a client, the server's
1637 challenges are checked to ensure they are valid base64 strings. By
1638 default, the authentication attempt is cancelled if an invalid string is
1639 received. Setting client_ignore_invalid_base64 true ignores these errors.
1640 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1641 they are received. Thus, the responses can be made to depend on the
1642 challenges. If an invalid string is ignored, an empty string is placed in
1643 the variable.
1644
30dba1e6
PH
1645PH/46 Messages that are created by the autoreply transport now contains a
1646 References: header, in accordance with RFCs 2822 and 3834.
1647
382afc6b
PH
1648PH/47 Added authenticated_sender_force to the smtp transport.
1649
a86229cf
PH
1650PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1651
50c99ba6
PH
1652PH/49 Installed latest patch from the Sieve maintainer.
1653
d35e429d
PH
1654PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1655 was also set, the appendfile transport was unnecessarily scanning a
1656 directory of message files (e.g. for maildir delivery) to find the count
1657 of files (along with the size), even though it did not need this
1658 information. It now does the scan only if it needs to find either the
1659 size of the count of files.
1660
f90d018c
PH
1661PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1662
75def545
PH
1663PH/52 Two bugs concerned with error handling when the smtp transport is
1664 used in LMTP mode:
1665
1666 (i) Exim was not creating retry information for temporary errors given
1667 for individual recipients after the DATA command when the smtp transport
1668 was used in LMTP mode. This meant that they could be retried too
1669 frequently, and not timed out correctly.
1670
1671 (ii) Exim was setting the flag that allows error details to be returned
1672 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1673 recipients that were returned after the DATA command.
1674
1675PH/53 This is related to PH/52, but is more general: for any failing address,
1676 when detailed error information was permitted to be returned to the
1677 sender, but the error was temporary, then after the final timeout, only
1678 "retry timeout exceeded" was returned. Now it returns the full error as
1679 well as "retry timeout exceeded".
1680
c46782ef
PH
1681PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1682 do this, and (what is worse) MTAs that accept it.
1683
71fafd95
PH
1684PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1685 will now be deprecated.
1686
2c5db4fd
PH
1687PH/56 New os.c-cygwin from the Cygwin maintainer.
1688
9cf6b11a
JJ
1689JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1690 formats (previously only available in exim formats via -bpr, -bpru,
1691 and -bpra. Now also available in native and exiqgrep formats)
1692
1693JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1694 with very large, slow to parse queues
1695
1696JJ/08 exipick: added ! as generic prefix to negate any criteria format
1697
1698JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1699
898d150f
PH
1700PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1701 responses to authentication challenges, though it was showing the
1702 challenges; (ii) I've removed the CR characters from the debug output for
1703 SMTP output lines.
1704
46218253
PH
1705PH/58 Allow for the insertion of a newline as well as a space when a string
1706 is turned into more than one encoded-word during RFC 2047 encoding. The
1707 Sieve code now uses this.
1708
e97957bc
PH
1709PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1710 data_4xx, lost_connection, tls_required.
1711
81e509d7
PH
1712PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1713 message was being sent as an SMTP response.
1714
3d240ff7
PH
1715PH/61 Add -l and -k options to exicyclog.
1716
b37c4101
PH
1717PH/62 When verifying, if an address was redirected to one new address, so that
1718 verification continued, and the new address failed or deferred after
1719 having set something in $address_data, the value of $address_data was not
1720 passed back to the ACL. This was different to the case when no
1721 redirection occurred. The value is now passed back in both cases.
1722
79378e0f
PH
1723PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1724 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1725 use login_cap.h, so on its own it isn't the distinguishing feature. The
1726 new name refers directly to the setclassresources() function.
1727
e49c7bb4
PH
1728PH/65 Added configuration files for NetBSD3.
1729
d114ec46
PH
1730PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1731
f3d7df6c
PH
1732PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1733 is preferred over IPv4.
1734
715ab376
PH
1735PH/68 The bounce_return_message and bounce_return_body options were not being
1736 honoured for bounces generated during the reception of non-SMTP messages.
1737 In particular, this applied to messages rejected by the ACL. This bug has
1738 been fixed. However, if bounce_return_message is true and bounce_return_
1739 body is false, the headers that are returned for a non-SMTP message
1740 include only those that have been read before the error was detected.
1741 (In the case of an ACL rejection, they have all been read.)
1742
6b31b150
PH
1743PH/69 The HTML version of the specification is now built in a directory called
1744 spec_html instead of spec.html, because the latter looks like a path with
1745 a MIME-type, and this confuses some software.
1746
1747PH/70 Catch two compiler warnings in sieve.c.
1748
d515a917
PH
1749PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1750 function verify_get_ident() calls ip_connect() to connect a socket, but
1751 if the "connect()" function timed out, ip_connect() used to close the
1752 socket. However, verify_get_ident() also closes the socket later, and in
1753 between Exim writes to the log, which may get opened at this point. When
1754 the socket was closed in ip_connect(), the log could get the same file
1755 descriptor number as the socket. This naturally causes chaos. The fix is
1756 not to close the socket in ip_connect(); the socket should be closed by
1757 the function that creates it. There was only one place in the code where
1758 this was missing, in the iplookup router, which I don't think anybody now
1759 uses, but I've fixed it anyway.
1760
9b8fadde
PH
1761PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1762 well as to direct DNS lookups. Otherwise the handling of names in host
1763 lists is inconsistent and therefore confusing.
1764
214e2000 1765
5de37277
PH
1766Exim version 4.60
1767-----------------
1768
cc38ddbf
PH
1769PH/01 Two changes to the default runtime configuration:
1770
1771 (1) Move the checks for relay_from_hosts and authenticated clients from
1772 after to before the (commented out) DNS black list checks.
1773
1774 (2) Add control=submission to the relay_from_hosts and authenticated
1775 clients checks, on the grounds that messages accepted by these
1776 statements are most likely to be submissions.
5de37277 1777
72fdd6ae
PH
1778PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1779
1780 (1) Generate an error if the third argument for the ${prvs expansion is
1781 not a single digit.
1782
1783 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1784 string.
1785
1786 (3) Reset the variables that are obtained from the first argument of
1787 ${prvscheck and used in the second argument before leaving the code,
1788 because their memory is reclaimed, so using them afterwards may do
1789 silly things.
1790
1791 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1792 one (it's much easier than Tom thought :-).
1793
1794 (5) Because of (4), we can now allow for the use of $prvscheck_result
1795 inside the third argument.
cb9328de 1796
cb741023
PH
1797PH/03 For some reason, the default setting of PATH when running a command from
1798 a pipe transport was just "/usr/bin". I have changed it to
1799 "/bin:/usr/bin".
1800
f174f16e
PH
1801PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1802 anything to be listed in the output from -bV.
b2f5a032 1803
c25242d7
PH
1804PH/05 When a filter generated an autoreply, the entire To: header line was
1805 quoted in the delivery log line, like this:
1806
1807 => >A.N.Other <ano@some.domain> <original@ddress> ...
1808
1809 This has been changed so that it extracts the operative address. There
1810 may be more than one such address. If so, they are comma-separated, like
1811 this:
1812
1813 => >ano@some.domain,ona@other.domain <original@ddress> ...
1814
82c19f95
PH
1815PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1816 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1817 not being looked up in the rDNS to get a host name, Exim was showing the
1818 IP address twice in Received: lines, even though the IP addresses were
1819 identical. For example:
1820
1821 Received: from [1.2.3.4] (helo=[1.2.3.4])
1822
1823 However, if the real host name was known, it was omitting the HELO data
1824 if it matched the actual IP address. This has been tidied up so that it
1825 doesn't show the same IP address twice.
1826
d7ffbc12
PH
1827PH/07 When both +timestamp and +memory debugging was on, the value given by
1828 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1829 called by the debug printing, thereby overwriting the timestamp buffer.
1830 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1831
9f526266
PH
1832PH/08 When the original message was included in an autoreply transport, it
1833 always said "this is a copy of the message, including all the headers",
1834 even if body_only or headers_only was set. It now gives an appropriate
1835 message.
1836
87fcc8b9
PH
1837PH/09 Applied a patch from the Sieve maintainer which:
1838
1839 o fixes some comments
1840 o adds the (disabled) notify extension core
1841 o adds some debug output for the result of if/elsif tests
1842 o points to the current vacation draft in the documentation
1843 and documents the missing references header update
1844
1845 and most important:
1846
1847 o fixes a bug in processing the envelope test (when testing
1848 multiple envelope elements, the last element determinted the
1849 result)
1850
456682f5
PH
1851PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1852 Electronic Mail") by including:
1853
1854 Auto-submitted: auto-generated
1855
1856 in the messages that it generates (bounce messages and others, such as
1857 warnings). In the case of bounce messages for non-SMTP mesages, there was
1858 also a typo: it was using "Auto_submitted" (underscore instead of
1859 hyphen). Since every message generated by Exim is necessarily in response
1860 to another message, thes have all been changed to:
1861
1862 Auto-Submitted: auto-replied
1863
1864 in accordance with these statements in the RFC:
1865
1866 The auto-replied keyword:
1867
1868 - SHOULD be used on messages sent in direct response to another
1869 message by an automatic process,
1870
1871 - MUST NOT be used on manually-generated messages,
1872
1873 - MAY be used on Delivery Status Notifications (DSNs) and Message
1874 Disposition Notifications (MDNs),
1875
1876 - MUST NOT be used on messages generated by automatic or periodic
1877 processes, except for messages which are automatic responses to
1878 other messages.
1879
3e46c1aa
PH
1880PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1881 to the default Received: header definition.
456682f5 1882
49826d12
PH
1883PH/12 Added log selector acl_warn_skipped (default on).
1884
eba0c039
PH
1885PH/13 After a successful wildlsearch lookup, discard the values of numeric
1886 variables because (a) they are in the wrong storage pool and (b) even if
1887 they were copied, it wouldn't work properly because of the caching.
1888
a0d6ba8a
PH
1889PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1890 checking when decoding. Apparently there are clients that generate
1891 overlong encoded strings. Why am I not surprised?
1892
f0917727
PH
1893PH/15 If the first argument of "${if match_address" was not empty, but did not
1894 contain an "@" character, Exim crashed. Now it writes a panic log message
1895 and treats the condition as false.
1896
096fee00
PH
1897PH/16 In autoreply, treat an empty string for "once" the same as unset.
1898
024bd3c2
PH
1899PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1900 extension "envelope-auth". The code is finished and in agreement with
1901 other implementations, but there is no documentation so far and in fact,
1902 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1903 not changing the active code.
1904
1905 Print executed "if" and "elsif" statements when debugging is used. This
1906 helps a great deal to understand what a filter does.
1907
1908 Document more things not specified clearly in RFC3028. I had all this
1909 sorted out, when out of a sudden new issues came to my mind. Oops."
1910
df199fec
PH
1911PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1912 (Bugzilla #53).
1913
d27f1df3
PH
1914PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1915 canonical form (as documented). However, after a host name lookup from
1916 the IP address, check_host() was doing a simple string comparison with
1917 addresses acquired from the DNS when checking that the found name did
1918 have the original IP as one of its addresses. Since any found IPv6
1919 addresses are likely to be in abbreviated form, the comparison could
1920 fail. Luckily, there already exists a function for doing the comparison
1921 by converting both addresses to binary, so now that is used instead of
1922 the text comparison.
1923
96776534
PH
1924PH/20 There was another similar case to PH/19, when a complete host name was
1925 given in a host list; looking up its IP address could give an abbreviated
1926 form, whereas the current host's name might or might not be abbreviated.
1927 The same fix has been applied.
1928
5de37277 1929
9a799bc0
PH
1930Exim version 4.54
1931-----------------
1932
1933PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1934 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1935 It now does.
1936
99a4b039
PH
1937PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1938 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1939
1940PH/03 Typo: missing ".o" in src/pcre/Makefile.
1941
4b233853
PH
1942PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1943 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1944 Also, for "Auto-Submitted", treat anything other than "no" as
1945 non-personal, in accordance with RFC 3834. (Previously it treated
1946 anything starting "auto-" as non-personal.)
1947
8857ccfd
PH
1948TF/01 The control=submission/name=... option had a problem with syntax
1949 errors if the name included a slash character. The /name= option
1950 now slurps the rest of the string, so it can include any characters
1951 but it must come last in the list of options (after /sender_retain
1952 or /domain=).
1953
433a2980
PH
1954PH/05 Some modifications to the interface to the fake nameserver for the new
1955 testing suite.
1956
3e46c1aa 1957
9a799bc0 1958
e3a311ba
TK
1959Exim version 4.53
1960-----------------
1961
1962TK/01 Added the "success_on_redirect" address verification option. See
1963 NewStuff for rationale and an example.
1964
13b685f9
PH
1965PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1966
395ff96d
PH
1967PH/02 Patch to exigrep to allow it to work on syslog lines.
1968
5b68f6e4
PH
1969PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1970 fread() to skip over the body file's header line, because in Cygwin the
1971 header line is locked and is inaccessible.
1972
1ab52c69
PH
1973PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1974 co-exist for some time) to make it clear that it is the Exim ID that is
1975 referenced, not the Message-ID: header line.
1976
b07e6aa3
PH
1977PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1978 string_format() function, because snprintf() does not exist on all
1979 operating systems.
1980
254e032f
PH
1981PH/06 The use of forbid_filter_existstest now also locks out the use of the
1982 ${stat: expansion item.
1983
3af76a81
PH
1984PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1985 protocol synchronization error", to keep the pedants happy.
1986
2548ba04
PH
1987PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1988 well as for IRIX systems, when gcc is being used. See the host.c source
1989 file for comments.
1990
b6c6011d
PH
1991PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1992
cf39cf57
PH
1993PH/10 Named domain lists were not working if used in a queue_smtp_domains
1994 setting.
1995
f1513293
PH
1996PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1997 transport and to the smtp transport in LMTP mode.
1998
727549a4
PH
1999TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2000
af46795e
PH
2001PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2002 run a filter in a subprocess. This could lead to confusion in subsequent
2003 lookups in the parent process. There should also be a search_tidyup() at
2004 the end of the subprocess.
2005
d7b47fd0
PH
2006PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2007 only if the host matched helo_try_verify_hosts, which caused the
2008 verification to occur when the EHLO/HELO command was issued. The ACL just
2009 tested the remembered result. Now, if a previous verification attempt has
2010 not happened, "verify = helo" does it there and then.
2011
ee744174
JJ
2012JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2013
b582ab87
PH
2014TK/03 Fix log output including CR from clamd.
2015
41a13e0a
PH
2016PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2017 exist provoked a memory error which could cause a segfault.
2018
f625cc5a
PH
2019PH/15 Installed PCRE 6.2
2020
2021PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2022
21f7af35
PH
2023PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2024 of the problem. Specifically, suggested +O2 rather than +O1 for the
2025 HP-UX compiler.
2026
31480e42
PH
2027PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2028
2d280592
PH
2029PH/20 If a delivery was routed to a non-standard port by means of an SRV
2030 record, the port was not correctly logged when the outgoing_port log
2031 selector was set (it logged the transort's default port).
2032
7cd1141b
PH
2033PH/21 Added support for host-specific ports to manualroute, queryprogram,
2034 fallback_hosts, and "hosts" in the smtp transport.
2035
2036PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2037 host errors such as "Connection refused".
2038
750af86e
PH
2039PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2040 authentication with radiusclient 0.4.9:
2041
2042 - Error returned from rc_read_config was caught wrongly
2043 - Username/password not passed on to radius server due to wrong length.
2044
2045 The presumption is that some radiusclient API changes for 4.51/PH/17
2046 were not taken care of correctly. The code is still untested by me (my
2047 Linux distribution still has 0.3.2 of radiusclient), but it was
2048 contributed by a Radius user.
2049
2050PH/24 When doing a callout, the value of $domain wasn't set correctly when
2051 expanding the "port" option of the smtp transport.
2052
4304270b
TK
2053TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2054 while reading a MIME header. Thanks to Tom Hughes for a patch.
2055
750af86e
PH
2056PH/24 Include config.h inside local_scan.h so that configuration settings are
2057 available.
2058
64ffc24f
PH
2059PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2060 that in an ACL for RCPT (for example), you can examine exactly what was
2061 received.
2062
5dd9625b
PH
2063PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2064 commands, but it was not correctly comparing the address with the actual
2065 client host address. Thus, it would show the EHLO address in Received:
2066 header lines when this was not necessary.
2067
5591031b
PH
2068PH/27 Added the % operator to ${eval:}.
2069
ba18e66a
PH
2070PH/28 Exim tries to create and chdir to its spool directory when it starts;
2071 it should be ignoring failures (because with -C, for example, it has lost
2072 privilege). It wasn't ignoring creation failures other than "already
2073 exists".
2074
9cec981f
PH
2075PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2076 -bV or -d is used.
2077
aa2b5c79
PH
2078PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2079 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
2080 too many continuations, the temporary file was not being removed, and the
2081 return code was incorrect.
aa2b5c79 2082
48a53b7f
PH
2083PH/31 Missing "BOOL" in function definition in filtertest.c.
2084
1c59d63b
PH
2085PH/32 Applied Sieve patches from the maintainer.
2086
671012da
TK
2087TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2088
1509d3a8
PH
2089PH/33 Added "verify = not_blind".
2090
2091PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2092 Local/Makefile (with some defaults set). These are used in built scripts
2093 such as exicyclog, but they have never been used in the exim_install
2094 script (though there are many overriding facilities there). I have
2095 arranged that the exim_install script now takes note of these two
2096 settings.
2097
2098PH/35 Installed configuration files for Dragonfly.
2099
2fe1a124
PH
2100PH/36 When a locally submitted message by a trusted user did not contain a
2101 From: header, and the sender address was obtained from -f or from an SMTP
2102 MAIL command, and the trusted user did not use -F to supply a sender
2103 name, $originator_name was incorrectly used when constructing a From:
2104 header. Furthermore, $originator_name was used for submission mode
2105 messages from external hosts without From: headers in a similar way,
2106 which is clearly wrong.
2107
8800895a
PH
2108PH/37 Added control=suppress_local_fixups.
2109
ccfdb010
PH
2110PH/38 When log_selector = +received_sender was set, and the addition of the
2111 sender made the log line's construction buffer exactly full, or one byte
2112 less than full, an overflow happened when the terminating "\n" was
2113 subsequently added.
2114
1130bfb0
PH
2115PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2116 when the result of a list match is failure because a DNS lookup failed.
2117
ebcb507f
PH
2118PH/40 RM_COMMAND is now used in the building process.
2119
c35e155c
PH
2120PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2121 the "build-* directories that it finds.
2122
95d1f782
PH
2123PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2124 address in a domain literal was a prefix of an interface address.
2125
fd6de02e
PH
2126PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2127 when verifying a sender address, unless rewrite_headers is false.
2128
58de37c5
PH
2129PH/44 Wrote a long comment about why errors_to addresses are verified as
2130 recipients, not senders.
2131
261cf466
TF
2132TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2133 the ratelimit ACL was added.
2134
3ee512ff
PH
2135PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2136
e08c430f
PH
2137PH/46 Added extra information about PostgreSQL errors to the error string.
2138
bef5a11f
PH
2139PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2140 avoiding the need to install special zones in a real server. This is
2141 backwards compatible; if it can't find the fake resolver, it drops back.
2142 Thus, both old and new test suites can be run.
2143
7546de58
TF
2144TF/02 Added util/ratelimit.pl
2145
e5d5a95f
TF
2146TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2147 clock is set back in time.
2148
2e88a017
TF
2149TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2150 Candler <B.Candler@pobox.com>.
2151
a5f65aa4
TF
2152TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2153 OK for addresses that are the result of redirections.
2154
e7726cbf
PH
2155PH/48 A number of further additions for the benefit of the new test suite,
2156 including a fake gethostbyname() that interfaces to the fake DNS resolver
2157 (see PH/47 above).
2158
a7fdad5b
TF
2159TF/06 The fix for widen_domains has also been applied to qualify_single and
2160 search_parents which are the other dnslookup options that can cause
2161 header rewrites.
2162
6af56900
PH
2163PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2164 type ("H").
2165
0925ede6
PH
2166PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2167
66afa403
TF
2168TF/07 Exim produced the error message "an SRV record indicated no SMTP
2169 service" if it encountered an MX record with an empty target hostname.
2170 The message is now "an MX or SRV record indicated no SMTP service".
2171
0154e85a
TF
2172TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2173 if the DNS of the sending site is misconfigured. This is quite a
2174 common situation. This change restores the behaviour of treating a
2175 helo verification defer as a failure.
2176
16f12c76
PH
2177PH/51 If self=fail was set on a router, the bounce message did not include the
2178 actual error message.
2179
bbe902f0 2180
e5a9dba6
PH
2181Exim version 4.52
2182-----------------
2183
2184TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2185
22c3b60b
PH
2186PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2187 command itself ended in error, the underlying message about the transport
2188 filter timeout was being overwritten with the pipe command error. Now the
2189 underlying error message should be appended to the second error message.
2190
06a9b4b5
PH
2191TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2192
c1ac6996
PH
2193PH/02 Reduce the amount of output that "make" produces by default. Full output
2194 can still be requested.
2195
9c7a242c
PH
2196PH/03 The warning log line about a condition test deferring for a "warn" verb
2197 was being output only once per connection, rather than after each
2198 occurrence (because it was using the same function as for successful
2199 "warn" verbs). This seems wrong, so I have changed it.
2200
87ba3f5f
PH
2201TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2202 it should not have, which might have caused a crash in the right
2203 circumstances, but probably never did.
2204
2205PH/04 Installed a modified version of Tony Finch's patch to make submission
2206 mode fix the return path as well as the Sender: header line, and to
2207 add a /name= option so that you can make the user's friendly name appear
2208 in the header line.
2209
29aba418
TF
2210TF/03 Added the control = fakedefer ACL modifier.
2211
fe0dab11
TF
2212TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2213 Mark Lowes for thorough testing.
870f6ba8 2214
11d337a4
TK
2215TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2216
2217TK/03 Merged latest SRS patch from Miles Wilton.
2218
415c8f3b
PH
2219PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2220 with the definition in sysexits.h (which is #included earlier).
2221 Fortunately, Exim does not actually use EX_OK. The code used to try to
2222 preserve the sysexits.h value, by assumimg that macro definitions were
2223 scanned for macro replacements. I have been disabused of this notion,
2224 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 2225
958541e9
PH
2226PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2227 in the smtp transport. When a block could not be written in a single
2228 write() function, the timeout was being re-applied to each part-write.
2229 This seems wrong - if the receiver was accepting one byte at a time it
2230 would take for ever. The timeout is now adjusted when this happens. It
2231 doesn't have to be particularly precise.
2232
c206415f
TK
2233TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2234 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2235
2a4be8f9
PH
2236PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2237 without a domain if the check to <postmaster@domain> fails.
2238
1cba11c5
SC
2239SC/01 Eximstats: added -xls and the ability to specify output files
2240 (patch written by Frank Heydlauf).
2241
2242SC/02 Eximstats: use FileHandles for outputing results.
2243
2244SC/03 Eximstats: allow any combination of xls, txt, and html output.
2245
2246SC/04 Eximstats: fixed display of large numbers with -nvr option
2247
2248SC/05 Eximstats: fixed merging of reports with empty tables.
2249
2250SC/06 Eximstats: added the -include_original_destination flag
2251
2252SC/07 Eximstats: removed tabs and trailing whitespace.
2253
1005d00e
TK
2254TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2255
2256TK/06 MBOX spool code: Add real "From " MBOX separator line
2257 so the .eml file is really in mbox format (even though
2258 most programs do not really care). Patch from Alex Miller.
2259
2260TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2261 The latter is generated from $received_to and is only set if the
2262 message has one envelope recipient. SA can use these headers,
2263 obviously out-of-the-box. Patch from Alex Miller.
2264
9b4768fa
PH
2265PH/08 The ${def test on a variable was returning false if the variable's
2266 value was "0", contrary to what the specification has always said!
2267 The result should be true unless the variable is empty.
2268
2269PH/09 The syntax error of a character other than { following "${if
2270 def:variable_name" (after optional whitespace) was not being diagnosed.
2271 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2272 accidental colon was present, for example, could give incorrect results.
2273
0d7eb84a
PH
2274PH/10 Tidied the code in a number of places where the st_size field of a stat()
2275 result is used (not including appendfile, where other changes are about
2276 to be made).
2277
2278PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2279 This involved changing a lot of size variables from int to off_t. It
2280 should work with maildirs and everything.
2281
40727bee
TK
2282TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2283 spamd dying while we are connected to it.
2284
554d2369
TF
2285TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2286 <jgh@wizmail.org>
2287
1f922db1
PH
2288PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2289 function for BDB 4.3.
2290
ef213c3b
PH
2291PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2292
8ac170f3
PH
2293PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2294 the sources for PCRE have been re-arranged and more files are now
2295 involved.
2296
b1c749bb
PH
2297PH/15 The code I had for printing potentially long long variables in PH/11
2298 above was not the best (it lost precision). The length of off_t variables
2299 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
2300 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2301 to be "long long int" or "long int". This is needed for the internal
2302 formatting function string_vformat().
b1c749bb 2303
4aac9b49
PH
2304PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2305 the configuration file to be ":syslog", then the script "guesses" where
2306 the logs files are, rather than using the compiled in default. In our
2307 case the guess is not the same as the compiled default, so the script
2308 suddenly stopped working when I started to use syslog. The patch checks
2309 to see if log_file_path is "". If so, it attempts to read it from exim
2310 with no configuration file to get the compiled in version, before it
2311 falls back to the previous guessing code."
2312
294520c8
TK
2313TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2314 implementing BATV in an Exim configuration. See NewStuff for the gory
2315 details.
2316
5bd022fe
PH
2317PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2318 Makefile that are specific to HP-UX.
2319
90e9ce59
PH
2320PH/18 If the "use_postmaster" option was set for a recipient callout together
2321 with the "random" option, the postmaster address was used as the MAIL
2322 FROM address for the random test, but not for the subsequent recipient
2323 test. It is now used for both.
2324
5ea81592
PH
2325PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2326 patch removes a few documentation additions to RFC 3028, because the
2327 latest draft now contains them. It adds the new en;ascii-case comparator
2328 and a new error check for 8bit text in MIME parts. Comparator and
2329 require names are now matched exactly. I enabled the subaddress
2330 extension, but it is not well tested yet (read: it works for me)."
2331
c6c2dc1d
PH
2332PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2333 rework some of the code of TK/09 above to avoid the hardwired use of
2334 "%lld" and "long long". Replaced the call to snprintf() with a call to
2335 string_vformat().
2336
fffffe4c
PH
2337PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2338 records point to non-existent hosts", "retry timeout exceeded", and
2339 "retry time not reached for any host after a long failure period".
ca02eafb 2340
9a26b6b2
PH
2341PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2342 experimental DomainKeys support:
2343
2344 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2345 (2) On an error such as an illegally used "control", the wrong name for
2346 the control was given.
2347
2348 These problems did NOT occur unless DomainKeys support was compiled.
2349
4aee0225
PH
2350PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2351
32d668a5
PH
2352PH/24 Added ${if match_ip condition.
2353
8187c3f3
PH
2354PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2355 databases so that it will be absolutely obvious if a crash occurs in the
2356 DB library. This is a regular occurrence (often caused by mis-matched
2357 db.h files).
2358
ff790e47 2359PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
2360 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2361 picked up on a user's system that detects such things. There doesn't seem
2362 to be a gcc warning option for this - only an attribute that has to be
2363 put on the function's prototype. It seems that in Fedora Core 4 they have
2364 set this on a number of new functions. No doubt there will be more in due
2365 course.
ff790e47 2366
5417f6d1
PH
2367PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2368 specify a transport. However, if an address that was verified by such a
2369 router was the subject of a callout, Exim crashed because it tried to
2370 read the rcpt_include_affixes from the non-existent transport. Now it
2371 just assumes that the setting of that option is false. This bug was
2372 introduced by 4.51/PH/31.
2373
59cf8544
PH
2374PH/28 Changed -d+all to exclude +memory, because that information is very
2375 rarely of interest, but it makes the output a lot bigger. People tend to
2376 do -d+all out of habit.
2377
e7ad8a65
PH
2378PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2379 code in os-type was giving problems when libc.so lives in lib64, like on
2380 x86_64 Fedora Core.
2381
ade42478
PH
2382PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2383 aren't the modern standard, and it seems that some systems' include files
2384 don't always have them. Exim was already checking for some of the newer
2385 ones like T_AAAA, and defining it itself. I've added checks for all the
2386 record types that Exim uses.
2387
182ad5cf
PH
2388PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2389 not automatically generating a new one, as it is supposed to. This
2390 prevented TLS from working. If the file did exist, but contained invalid
2391 data, a new version was generated, as expected. It was only the case of a
2392 non-existent file that was broken.
2393
b0d9fc80
TK
2394TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2395 with a change in libdomainkeys > 0.64.
2396
2397TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2398 from DNS. If the selector record carries the flag, it now has
2399 precedence over the domain-wide flag.
2400
2401TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2402
47c7a64a
PH
2403PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2404 the use of a port name that isn't defined in /etc/services) occurred, the
2405 message was deferred as in a normal delivery, and thus remained on the
2406 spool, instead of being failed because of the mua_wrapper setting. This
2407 is now fixed, and I tidied up some of the mua_wrapper messages at the
2408 same time.
2409
a388bce4
SC
2410SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2411 the messages in a hash of arrays rather than using individual hashes.
2412 This is a bit cleaner and results in dramatic memory savings, albeit
2413 at a slight CPU cost.
2414
2415SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2416 as requested by Marc Sherman.
2417
2418SC/10 Eximstats: added histograms for user specified patterns as requested
2419 by Marc Sherman.
2420
0793e4ed
SC
2421SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2422
c58b88df
PH
2423PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2424 fopen() in the content-scanning modules that did not already have it.
2425
e7ad8a65 2426
7982096b
PH
2427Exim version 4.51
2428-----------------
2429
1a46a8c5
PH
2430TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2431 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2432
2f079f46 2433TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
2434
2435TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2436
2437PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2438 to test Sieve filters that use "vacation".
2439
2440PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2441 that changes the way the GnuTLS parameters are stored in the cache file.
2442 The new format can be generated externally. For backward compatibility,
2443 if the data in the cache doesn't make sense, Exim assumes it has read an
2444 old-format file, and it generates new data and writes a new file. This
2445 means that you can't go back to an older release without removing the
2446 file.
2447
2448PH/03 A redirect router that has both "unseen" and "one_time" set does not
2449 work if there are any delivery delays because "one_time" forces the
2450 parent to be marked "delivered", so its unseen clone is never tried
2451 again. For this reason, Exim now forbids the simultaneous setting of
2452 these two options.
2453
2454PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2455 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2456 entry if you want to know the details. The fix, however, neglected to
2457 consider the case when local delivery batching is involved. The test for
2458 "previously delivered" was not happening when checking to see if an
2459 address could be batched with a previous (undelivered) one; under
2460 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 2461 address.
1a46a8c5
PH
2462
2463PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2464 in its include files, and this causes problems building Exim.
2465
2466PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2467 header_syntax) but Exim was just ignoring anything given after a slash.
2468 In particular, this caused confusion with an attempt to use "verify =
2469 reverse_host_lookup/defer_ok". An error is now given when options are
2470 supplied for verify items that do not have them. (Maybe reverse_host_
2471 lookup should have a defer_ok option, but that's a different point.)
2472
2473PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2474 defined by RFC 821) to 2048, because there were problems with some AUTH
2475 commands, and RFC 1869 says the size should be increased for extended
2476 SMTP commands that take arguments.
2477
2478PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2479 Finch).
2480
2481PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2482 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 2483
49c2d5ea
PH
2484PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2485 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
2486 whatever). Otherwise the password may appear in the log. Following change
2487 PH/42 below, there is no longer a chance of it appearing in a bounce
2488 message.
49c2d5ea 2489
bf759a8b
PH
2490PH/11 Installed exipick version 20050225.0 from John Jetmore.
2491
83364d30
PH
2492PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2493 of its addresses was ever tried. (Bugzilla bug #2.)
2494
7999bbd7
PH
2495PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2496 the result incorrectly in the debug output. (It correctly added a newline
2497 to what was transported.)
2498
7dbf77c9
PH
2499TF/01 Added $received_time.
2500
74e0617f
PH
2501PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2502 commented out examples of how to interface to a virus scanner and to
2503 SpamAssassin. Also added commented examples of av_scanner and
2504 spamd_address settings.
2505
2f079f46
PH
2506PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2507 and controls are allowed in which ACLs. There were a couple of minor
2508 errors. Some of the entries in the conditions table (which is a table of
2509 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2510 negation of where the condition IS allowed.
2511
8c841523
PH
2512PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2513
7766a4f0
PH
2514PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2515 header file does not have a version number, so I've had to invent a new
2516 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2517 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2518 radiusclient), but it was contributed by a Radius user.
2519
8b417f2c
PH
2520PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2521 files or directories, for OpenSSL.
2522
901f42cb
PH
2523PH/19 When an Exim process that is running as root has to create an Exim log
2524 file, it does so in a subprocess that runs as exim:exim so as to get the
2525 ownership right at creation (otherwise, other Exim processes might see
2526 the file with the wrong ownership). There was no test for failure of this
2527 fork() call, which would lead to the process getting stuck as it waited
2528 for a non-existent subprocess. Forks do occasionally fail when resources
2529 run out. I reviewed all the other calls to fork(); they all seem to check
2530 for failure.
2531
f9b9210e
PH
2532PH/20 When checking for unexpected SMTP input at connect time (before writing
2533 the banner), Exim was not dealing correctly with a non-positive return
2534 from the read() function. If the client had disconnected by this time,
2535 the result was a log entry for a synchronization error with an empty
2536 string after "input=" when read() returned zero. If read() returned -1
2537 (an event I could not check), uninitialized data bytes were printed.
2538 There were reports of junk text (parts of files, etc) appearing after
2539 "input=".
2540
54cdb463
PH
2541PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2542
cf00dad6
PH
2543PH/22 Added support for macro redefinition, and (re)definition in between
2544 driver and ACL definitions.
2545
acb1b346
PH
2546PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2547 forgetting to use the resulting value; it was using the unexpanded value.
2548
c5ddb310
PH
2549PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2550 hadn't been configured. The fix is from Juergen Kreileder, who
2551 understands it better than I do:
2552
2553 "Here's what I see happening with three configured cyrus_sasl
2554 authenticators configured (plain, login, cram-md5):
2555
2556 On startup auth_cyrus_sasl_init() gets called for each of these.
2557 This means three calls to sasl_listmech() without a specified mech_list.
2558 => SASL tests which mechs of all available mechs actually work
2559 => three warnings about OTP not working
2560 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2561
2562 With the patch, sasl_listmech() also gets called three times. But now
2563 SASL's mech_list option is set to the server_mech specified in the the
2564 authenticator. Or in other words, the answer from sasl_listmech()
2565 gets limited to just the mech you're testing for (which is different
2566 for each call.)
2567 => the return list contains just 'plain' or 'login', 'cram-md5' or
2568 nothing depending on the value of ob->server_mech.
2569
2570 I've just tested the patch: Authentication still works fine,
2571 unavailable mechs specified in the exim configuration are still
2572 caught, and the auth.log warnings about OTP are gone."
2573
31619da6
PH
2574PH/25 When debugging is enabled, the contents of the command line are added
2575 to the debugging output, even when log_selector=+arguments is not
2576 specified.
2577
bebaf0fc
PH
2578PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2579 answer is "GNU", and only if the return is "GNU/something" is the answer
2580 "Linux".
2581
475fe28a
PH
2582PH/27 $acl_verify_message is now set immediately after the failure of a
2583 verification in an ACL, and so is available in subsequent modifiers. In
2584 particular, the message can be preserved by coding like this:
2585
2586 warn !verify = sender
2587 set acl_m0 = $acl_verify_message
2588
2589 Previously, $acl_verify_message was set only while expanding "message"
2590 and "log_message" when a very denied access.
2591
7e8bec7a
PH
2592PH/28 Modified OS/os.c-Linux with
2593
2594 -#ifndef OS_LOAD_AVERAGE
2595 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2596
2597 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2598 nomenclature these days.)
2599
e4a89c47
PH
2600PH/29 Installed patch from the Sieve maintainer that adds the options
2601 sieve_useraddress and sieve_subaddress to the redirect router.
2602
5ca2a9a1
PH
2603PH/30 In these circumstances:
2604 . Two addresses routed to the same list of hosts;
2605 . First host does not offer TLS;
2606 . First host accepts first address;
2607 . First host gives temporary error to second address;
2608 . Second host offers TLS and a TLS session is established;
2609 . Second host accepts second address.
2610 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2611 and peerdn, if requested) that were in fact used only for the second
2612 address.
7e8bec7a 2613
c688b954
PH
2614PH/31 When doing a callout as part of verifying an address, Exim was not paying
2615 attention to any local part prefix or suffix that was matched by the
2616 router that accepted the address. It now behaves in the same way as it
2617 does for delivery: the affixes are removed from the local part unless
2618 rcpt_include_affixes is set on the transport.
2619
fed77020
PH
2620PH/32 Add the sender address, as F=<...>, to the log line when logging a
2621 timeout during the DATA phase of an incoming message.
2622
7fe1560f
PH
2623PH/33 Sieve envelope tests were broken for match types other than :is. I have
2624 applied a patch sanctioned by the Sieve maintainer.
c688b954 2625
ebb6e6d5
PH
2626PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2627 the uid or gid is negative. A case of a negative gid caused this to be
2628 noticed. The fix allows for either to be negative.
2629
9c4e8f60
PH
2630PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2631 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2632 overlooked.
2633
2634PH/36 The change PH/12 above was broken. Fixed it.
2635
d7174846
PH
2636PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2637 the grounds that routing the same address twice would always produce the
2638 same answer. This might have been true once, but it is certainly no
2639 longer true now. Routing a child address may depend on the previous
2640 routing that produced that child. Some complicated redirection strategies
2641 went wrong when messages had multiple recipients, and made Exim's
2642 behaviour dependent on the order in which the addresses were given.
2643
2644 I have moved the duplicate checking until after the routing is complete.
2645 Exim scans the addresses that are assigned to local and remote
2646 transports, and removes any duplicates. This means that more work will be
2647 done, as duplicates will always all be routed, but duplicates are
2648 presumably rare, so I don't expect this is of any significance.
2649
2650 For deliveries to pipes, files, and autoreplies, the duplicate checking
2651 still happens during the routing process, since they are not going to be
2652 routed further.
2653
cfe75fc3
PH
2654PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2655 It corrects a timeout issue with spamd. This is Ian's comment: "The
2656 background is that sometimes spamd either never reads data from a
2657 connection it has accepted, or it never writes response data. The exiscan
2658 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2659 blindly assumes that writes won't block so it may never time out."
2660
be22d70e
PH
2661PH/39 Allow G after quota size as well as K and M.
2662
0612b098
PH
2663PH/40 The value set for $authenticated_id in an authenticator may not contain
2664 binary zeroes or newlines because the value is written to log lines and
2665 to spool files. There was no check on this. Now the value is run through
2666 the string_printing() function so that such characters are converted to
2667 printable escape sequences.
2668
2e0c1448
PH
2669PH/41 $message_linecount is a new variable that contains the total number of
2670 lines in the message. Compare $body_linecount, which is the count for the
2671 body only.
2672
447d236c
PH
2673PH/42 Exim no longer gives details of delivery errors for specific addresses in
2674 bounce and delay warning messages, except in certain special cases, which
2675 are as follows:
2676
2677 (a) An SMTP error message from a remote host;
2678 (b) A message specified in a :fail: redirection;
2679 (c) A message specified in a "fail" command in a system filter;
2680 (d) A message specified in a FAIL return from the queryprogram router;
2681 (e) A message specified by the cannot_route_message router option.
2682
2683 In these cases only, Exim does include the error details in bounce and
2684 warning messages. There are also a few cases where bland messages such
2685 as "unrouteable address" or "local delivery error" are given.
2686
d20976dc
PH
2687PH/43 $value is now also set for the "else" part of a ${run expansion.
2688
f656d135
PH
2689PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2690 being worked on, but at least Exim now implements the latest version to
2691 play with."
2692
2e2a30b4
PH
2693PH/45 In a pipe transport, although a timeout while waiting for the pipe
2694 process to complete was treated as a delivery failure, a timeout while
2695 writing the message to the pipe was logged, but erroneously treated as a
2696 successful delivery. Such timeouts include transport filter timeouts. For
2697 consistency with the overall process timeout, these timeouts are now
2698 treated as errors, giving rise to delivery failures by default. However,
2699 there is now a new Boolean option for the pipe transport called
2700 timeout_defer, which, if set TRUE, converts the failures into defers for
2701 both kinds of timeout. A transport filter timeout is now identified in
2702 the log output.
2703
9176e9f0
PH
2704PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2705 systems where "make" and "gmake" are different, calling "gmake" at top
2706 level broke things. I've arranged for the value of $(MAKE) to be passed
2707 from the Makefile to this script so that it can call the same version of
2708 "make".
2709
7982096b 2710
bbe902f0
PH
2711A note about Exim versions 4.44 and 4.50
2712----------------------------------------
2713
2714Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2715changes of various kinds. As a consequence, a big documentation update was
2716needed. This delayed the release for rather longer than seemed good, especially
2717in the light of a couple of (minor) security issues. Therefore, the changes
2718that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2719release. So 4.44 and 4.50 are in effect two different branches that both start
2720from 4.43.
2721
2722I have left the 4.50 change log unchanged; it contains all the changes since
27234.43. The change log for 4.44 is below; many of its items are identical to
2724those for 4.50. This seems to be the most sensible way to preserve the
2725historical information.
2726
2727
f7b63901 2728Exim version 4.50
495ae4b0
PH
2729-----------------
2730
5fe762f6
PH
2731 1. Minor wording change to the doc/README.SIEVE file.
2732
139059f6 2733 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 2734 computation of the current number of files was incorrect.
495ae4b0 2735
7086e875
PH
2736 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2737 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2738 place.
2739
35af9f61
PH
2740 4. Give more explanation in the error message when the command for a transport
2741 filter fails to execute.
2742
b668c215
PH
2743 5. There are several places where Exim runs a non-Exim command in a
2744 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2745 was being done only for the command run by the queryprogram router. It is
2746 now done for all such subprocesses. The other cases are: ${run, transport
2747 filters, and the commands run by the lmtp and pipe transports.
2748
a494b1e1
PH
2749 6. Added CONFIGURE_GROUP build-time option.
2750
2751 7. Some older OS have a limit of 256 on the maximum number of file
2752 descriptors. Exim was using setrlimit() to set 1000 as a large value
2753 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2754 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 2755
c5fcb476
PH
2756 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2757 was an oversight, and furthermore, ever since the addition of extra
2758 controls (e.g. 4.43/32), the checks on when to allow different forms of
2759 "control" were broken. There should now be diagnostics for all cases when a
2760 control that does not make sense is encountered.
2761
69358f02
PH
2762 9. Added the /retain_sender option to "control=submission".
2763
5be20824
PH
276410. $recipients is now available in the predata ACL (oversight).
2765
eb2c0248
PH
276611. Tidy the search cache before the fork to do a delivery from a message
2767 received from the command line. Otherwise the child will trigger a lookup
2768 failure and thereby defer the delivery if it tries to use (for example) a
2769 cached ldap connection that the parent has called unbind on.
2770
2a3eea10
PH
277112. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2772 of $address_data from the recipient verification was clobbered by the
2773 sender verification.
2774
277513. The value of address_data from a sender verification is now available in
2776 $sender_address_data in subsequent conditions in the ACL statement.
2777
23c7ff99
PH
277814. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2779
4deaf07d
PH
278015. Added a new option "connect=<time>" to callout options, to set a different
2781 connection timeout.
2782
926e1192
PH
278316. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2784 was its contents. (It was OK if the option was not defined at all.)
2785
650edc6f
PH
278617. A "Completed" log line is now written for messages that are removed from
2787 the spool by the -Mrm option.
2788
2c7db3f5
PH
278918. New variables $sender_verify_failure and $recipient_verify_failure contain
2790 information about exactly what failed.
2791
3d235903
PH
279219. Added -dd to debug only the daemon process.
2793
7c7ad977
PH
279420. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2795 handles timeouts, both on the server side and network timeouts. Renamed the
2796 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2797
981756db
PH
279821. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2799 It is now set to "smtps".
2800
d4eb88df
PH
280122. $host_address is now set to the target address during the checking of
2802 ignore_target_hosts.
2803
280423. When checking ignore_target_hosts for an ipliteral router, no host name was
2805 being passed; this would have caused $sender_host_name to have been used if
2806 matching the list had actually called for a host name (not very likely,
2807 since this list is usually IP addresses). A host name is now passed as
2808 "[x.x.x.x]".
2809
7d468ab8
PH
281024. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2811 code that specifies a non-restarting handler (typically sigaction() in
2812 modern systems) in an attempt to fix a rare and obscure crash bug.
2813
281425. Narrowed the window for a race in the daemon that could cause it to ignore
2815 SIGCHLD signals. This is not a major problem, because they are used only to
2816 wake it up if nothing else does.
2817
62c0818f
PH
281826. A malformed maildirsize file could cause Exim to calculate negative values
2819 for the mailbox size or file count. Odd effects could occur as a result.
2820 The maildirsize information is now recalculated if the size or filecount
2821 end up negative.
2822
26034054
PH
282327. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2824 support for a long time. Removed HAVE_SYS_VFS_H.
2825
af66f652
PH
282628. Installed the latest version of exipick from John Jetmore.
2827
90af77f4
PH
282829. In an address list, if the pattern was not a regular expression, an empty
2829 subject address (from a bounce message) matched only if the pattern was an
2830 empty string. Non-empty patterns were not even tested. This was the wrong
2831 because it is perfectly reasonable to use an empty address as part of a
2832 database query. An empty address is now tested by patterns that are
2833 lookups. However, all the other forms of pattern expect the subject to
2834 contain a local part and a domain, and therefore, for them, an empty
2835 address still always fails if the pattern is not itself empty.
2836
d8ef3577
PH
283730. Exim went into a mad DNS loop when attempting to do a callout where the
2838 host was specified on an smtp transport, and looking it up yielded more
2839 than one IP address.
2840
5cb8cbc6
PH
284131. Re-factored the code for checking spool and log partition space into a
2842 function that finds that data and another that does the check. The former
2843 is then used to implement four new variables: $spool_space, $log_space,
2844 $spool_inodes, and $log_inodes.
2845
14702f5b
PH
284632. The RFC2047 encoding function was originally intended for short strings
2847 such as real names; it was not keeping to the 75-character limit for
2848 encoded words that the RFC imposes. It now respects the limit, and
2849 generates multiple encoded words if necessary. To be on the safe side, I
2850 have increased the buffer size for the ${rfc2047: expansion operator from
2851 1024 to 2048 bytes.
2852
063b1e99
PH
285333. It is now permitted to omit both strings after an "if" condition; if the
2854 condition is true, the result is "true". As before, when the second string
2855 is omitted, a false condition yields an empty string. This makes it less
2856 cumbersome to write custom ACL and router conditions.
2857
652e1b65
PH
285834. Failure to deliver a bounce message always caused it to be frozen, even if
2859 there was an errors_to setting on the router. The errors_to setting is now
2860 respected.
2861
6f0c9a4f
PH
286235. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2863 canonical form (fully expanded) before being placed in
2864 $sender_host_address.
2865
33397d19
PH
286636. The table in the code that translates DNS record types into text (T_A to
2867 "A" for instance) was missing entries for NS and CNAME. It is just possible
2868 that this could have caused confusion if both these types were looked up
2869 for the same domain, because the text type is used as part of Exim's
2870 per-process caching. But the chance of anyone hitting this buglet seems
2871 very small.
2872
7bb56e1f
PH
287337. The dnsdb lookup has been extended in a number of ways.
2874
2875 (1) There is a new type, "zns", which walks up the domain tree until it
2876 finds some nameserver records. It should be used with care.
2877
ea3bc19b
PH
2878 (2) There is a new type, "mxh", which is like "mx" except that it returns
2879 just the host names, not the priorities.
2880
2881 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
2882 looked up. The behaviour when one of the lookups defers can be
2883 controlled by a keyword.
7bb56e1f 2884
ea3bc19b 2885 (4) It is now possible to specify the separator character for use when
7bb56e1f 2886 multiple records are returned.
33397d19 2887
0bcb2a0e
PH
288838. The dnslists ACL condition has been extended: it is now possible to supply
2889 a list of IP addresses and/or domains to be looked up in a particular DNS
2890 domain.
2891
2ac0e484
PH
289239. Added log_selector=+queue_time_overall.
2893
4e1fde53
PH
289440. When running the queue in the test harness, wait just a tad after forking a
2895 delivery process, to get repeatability of debugging output.
2896
de365ded
PH
289741. Include certificate and key file names in error message when GnuTLS fails
2898 to set them up, because the GnuTLS error message doesn't include the name
2899 of the failing file when there is a problem reading it.
2900
f05da2e8
PH
290142. Allow both -bf and -bF in the same test run.
2902
d6453af2
PH
290343. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2904
f7b63901
PH
290544. The "Exiscan patch" is now merged into the mainline Exim source.
2906
290745. Sometimes the final signoff response after QUIT could fail to get
2908 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2909 < 0 before doing a fflush(). This bug looks as though it goes back to the
2910 introduction of TLS in release 3.20, but "sometimes" must have been rare
2911 because the tests only now provoked it.
2912
a444213a
PH
291346. Reset the locale to "C" after calling embedded Perl, in case it was changed
2914 (this can affect the format of dates).
2915
0ec020ea
PH
291647. exim_tidydb, when checking for the continued existence of a message for
2917 which it has found a message-specific retry record, was not finding
2918 messages that were in split spool directories. Consequently, it was
2919 deleting retry records that should have stayed in existence.
2920
b1206957
PH
292148. Steve fixed some bugs in eximstats.
2922
292349. The SPA authentication driver was not abandoning authentication and moving
2924 on to the next authenticator when an expansion was forced to fail,
2925 contradicting the general specification for all authenticators. Instead it
2926 was generating a temporary error. It now behaves as specified.
2927
26dd5a95
PH
292850. The default ordering of permitted cipher suites for GnuTLS was pessimal
2929 (the order specifies the preference for clients). The order is now AES256,
2930 AES128, 3DES, ARCFOUR128.
2931
343b2385
PH
293251. Small patch to Sieve code - explicitly set From: when generating an
2933 autoreply.
2934
1c5466b9
PH
293552. Exim crashed if a remote delivery caused a very long error message to be
2936 recorded - for instance if somebody sent an entire SpamAssassin report back
2937 as a large number of 550 error lines. This bug was coincidentally fixed by
2938 increasing the size of one of Exim's internal buffers (big_buffer) that
2939 happened as part of the Exiscan merge. However, to be on the safe side, I
2940 have made the code more robust (and fixed the comments that describe what
2941 is going on).
2942
55ee9ee3
PH
294353. Now that there can be additional text after "Completed" in log lines (if
2944 the queue_time_overall log selector is set), a one-byte patch to exigrep
2945 was needed to allow it to recognize "Completed" as not the last thing in
2946 the line.
2947
d38f8232
PH
294854. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2949 patch that reportedly fixes this has been added. I am not expert enough to
2950 create a test for it. This is what the patch creator wrote:
2951
2952 "I found a little strange behaviour of ldap code when working with
2953 Windows 2003 AD Domain, where users was placed in more than one
2954 Organization Units. When I tried to give exim partial DN, the exit code
2955 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2956 But simultaneously result of request was absolutely normal ldap result,
2957 so I produce this patch..."
2958
3295e65b
PH
2959 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2960 REFERENCE, so I have modified the code to exclude the patch when that macro
2961 is not defined.
2962
7102e136
PH
296355. Some experimental protocols are using DNS PTR records for new purposes. The
2964 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
2965 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2966 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 2967 CAN-2005-0021
7102e136 2968
3ca0ba97
PH
296956. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2970
c2bcbe20
PH
297157. Double the size of the debug message buffer (to 2048) so that more of very
2972 long debug lines gets shown.
2973
18ce445d
PH
297458. The exicyclog utility now does better if the number of log files to keep
2975 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2976
1f5b4c3d
PH
297759. Two changes related to the smtp_active_hostname option:
2978
2979 (1) $smtp_active_hostname is now available as a variable.
2980 (2) The default for smtp_banner uses $smtp_active_hostname instead
2981 of $primary_hostname.
2982
b975ba52
PH
298360. The host_aton() function is supposed to be passed a string that is known
2984 to be a valid IP address. However, in the case of IPv6 addresses, it was
2985 not checking this. This is a hostage to fortune. Exim now panics and dies
2986 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
2987 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2988 components; fortuitously, this particular loophole had already been fixed
2989 by change 4.50/55 above.
2990
2991 If there are any other similar loopholes, the new check in host_aton()
2992 itself should stop them being exploited. The report I received stated that
2993 data on the command line could provoke the exploit when Exim was running as
2994 exim, but did not say which command line option was involved. All I could
2995 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2996 running as the user.
ea3a6f44 2997 CAN-2005-0021
85b87bc2
PH
2998
299961. There was a buffer overflow vulnerability in the SPA authentication code
3000 (which came originally from the Samba project). I have added a test to the
3001 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 3002 CAN-2005-0022
b975ba52 3003
17ffcae7
PH
300462. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3005 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3006
d95f9fdb
PH
300763. The daemon start-up calls getloadavg() while still root for those OS that
3008 need the first call to be done as root, but it missed one case: when
3009 deliver_queue_load_max is set with deliver_drop_privilege. This is
3010 necessary for the benefit of the queue runner, because there is no re-exec
3011 when deliver_drop_privilege is set.
3012
86b8287f
PH
301364. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3014 This has been fixed.
3015
60dc5e56
PH
301665. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3017 was in use, was not putting the data itself into the right store pool;
3018 consequently, it could be overwritten for a subsequent message in the same
3019 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3020 the caching.)
3021
533244af
PH
302266. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3023
a5a28604
PH
302467. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3025 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3026 the function were treating the return as a boolean value, which happened to
3027 work because 0=false and not-0=true, but is not correct code.
3028
7e634d24
PH
302968. The host_aton() function was not handling scoped IPv6 addresses (those
3030 with, for example, "%eth0" on the end) correctly.
3031
3e11c26b
PH
303269. Fixed some compiler warnings in acl.c for the bitmaps specified with
3033 negated items (that is, ~something) in unsigned ints. Some compilers
3034 apparently mutter when there is no cast.
3035
6729cf78
PH
303670. If an address verification called from an ACL failed, and did not produce a
3037 user-specific message (i.e. there was only a "system" message), nothing was
3038 put in $acl_verify_message. In this situation, it now puts the system
3039 message there.
3040
00f00ca5
PH
304171. Change 4.23/11 added synchronization checking at the start of an SMTP
3042 session; change 4.31/43 added the unwanted input to the log line - except
3043 that it did not do this in the start of session case. It now does.
3044
c9bdd01c
PH
304572. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3046 This is wrong and can cause the other end to generate a synchronization
3047 error if it is another Exim or anything else that does the synchronization
3048 check. A QUIT command is no longer sent after a timeout.
3049
d43194df
PH
305073. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3051 during host lookups.
3052
fe5b5d0b
PH
305374. The defer_ok option of callout verification was not working if it was used
3054 when verifying addresses in header lines, that is, for this case:
3055
3056 verify = header_sender/callout=defer_ok
3057
76a2d7ba
PH
305875. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3059 those file descriptors could be used for SMTP connections. If anything
3060 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3061 could be sent to the SMTP client, causing chaos. The daemon now opens
3062 stdin, stdout, and stderr to /dev/null when it puts itself into the
3063 background.
3064
306576. Arrange for output from Perl's "warn" command to be written to Exim's main
3066 log by default. The user can override this with suitable Perl magic.
3067
04f7d5b9
PH
306877. The use of log_message on a "discard" ACL verb, which is supposed to add to
3069 the log message when discard triggers, was not working for the DATA ACL or
3070 for the non-SMTP ACL.
3071
bc60667e
PH
307278. Error message wording change in sieve.c.
3073
bb6e88ff
PH
307479. If smtp_accept_max_per_host was set, the number of connections could be
3075 restricted to fewer than expected, because the daemon was trying to set up
3076 a new connection before checking whether the processes handling previous
3077 connections had finished. The check for completed processes is now done
3078 earlier. On busy systems, this bug wouldn't be noticed because something
3079 else would have woken the daemon, and it would have reaped the completed
3080 process earlier.
3081
1e70f85b
PH
308280. If a message was submitted locally by a user whose login name contained one
3083 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3084 It caused a spool format error. I have fixed the spool reading code. A
3085 related problem was that the "from" clause in the Received: line became
3086 illegal because of the space(s). It is now covered by ${quote_local_part.
3087
308881. Included the latest eximstats from Steve (adds average sizes to HTML Top
3089 tables).
3090
4e01f9d6
PH
309182. Updated OS/Makefile-AIX as per message from Mike Meredith.
3092
1ee1cef2
PH
309383. Patch from Sieve maintainer to fix unterminated string problem in
3094 "vacation" handling.
3095
6e2b4ccc
PH
309684. Some minor changes to the Linux configuration files to help with other
3097 OS variants using glibc.
3098
8e669ac1
PH
309985. One more patch for Sieve to update vacation handling to latest spec.
3100
495ae4b0 3101
bbe902f0
PH
3102----------------------------------------------------
3103See the note above about the 4.44 and 4.50 releases.
3104----------------------------------------------------
3105
3106
3107Exim version 4.44
3108-----------------
3109
3110 1. Change 4.43/35 introduced a bug that caused file counts to be
3111 incorrectly computed when quota_filecount was set in an appendfile
3112 transport
3113
3114 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3115 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3116 place.
3117
3118 3. Give more explanation in the error message when the command for a transport
3119 filter fails to execute.
3120
3121 4. There are several places where Exim runs a non-Exim command in a
3122 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3123 was being done only for the command run by the queryprogram router. It is
3124 now done for all such subprocesses. The other cases are: ${run, transport
3125 filters, and the commands run by the lmtp and pipe transports.
3126
3127 5. Some older OS have a limit of 256 on the maximum number of file
3128 descriptors. Exim was using setrlimit() to set 1000 as a large value
3129 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3130 systems. I've change it so that if it can't get 1000, it tries for 256.
3131
3132 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3133 was an oversight, and furthermore, ever since the addition of extra
3134 controls (e.g. 4.43/32), the checks on when to allow different forms of
3135 "control" were broken. There should now be diagnostics for all cases when a
3136 control that does not make sense is encountered.
3137
3138 7. $recipients is now available in the predata ACL (oversight).
3139
3140 8. Tidy the search cache before the fork to do a delivery from a message
3141 received from the command line. Otherwise the child will trigger a lookup
3142 failure and thereby defer the delivery if it tries to use (for example) a
3143 cached ldap connection that the parent has called unbind on.
3144
3145 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3146 of $address_data from the recipient verification was clobbered by the
3147 sender verification.
3148
314910. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3150 was its contents. (It was OK if the option was not defined at all.)
3151
315211. A "Completed" log line is now written for messages that are removed from
3153 the spool by the -Mrm option.
3154
315512. $host_address is now set to the target address during the checking of
3156 ignore_target_hosts.
3157
315813. When checking ignore_target_hosts for an ipliteral router, no host name was
3159 being passed; this would have caused $sender_host_name to have been used if
3160 matching the list had actually called for a host name (not very likely,
3161 since this list is usually IP addresses). A host name is now passed as
3162 "[x.x.x.x]".
3163
316414. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3165 code that specifies a non-restarting handler (typically sigaction() in
3166 modern systems) in an attempt to fix a rare and obscure crash bug.
3167
316815. Narrowed the window for a race in the daemon that could cause it to ignore
3169 SIGCHLD signals. This is not a major problem, because they are used only to
3170 wake it up if nothing else does.
3171
317216. A malformed maildirsize file could cause Exim to calculate negative values
3173 for the mailbox size or file count. Odd effects could occur as a result.
3174 The maildirsize information is now recalculated if the size or filecount
3175 end up negative.
3176
317717. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3178 support for a long time. Removed HAVE_SYS_VFS_H.
3179
ea3a6f44 318018. Updated exipick to current release from John Jetmore.
bbe902f0
PH
3181
318219. Allow an empty sender to be matched against a lookup in an address list.
3183 Previously the only cases considered were a regular expression, or an
3184 empty pattern.
3185
318620. Exim went into a mad DNS lookup loop when doing a callout where the
3187 host was specified on the transport, if the DNS lookup yielded more than
3188 one IP address.
3189
ea3a6f44
NM
319021. The RFC2047 encoding function was originally intended for short strings
3191 such as real names; it was not keeping to the 75-character limit for
3192 encoded words that the RFC imposes. It now respects the limit, and
3193 generates multiple encoded words if necessary. To be on the safe side, I
3194 have increased the buffer size for the ${rfc2047: expansion operator from
3195 1024 to 2048 bytes.
bbe902f0 3196
ea3a6f44
NM
319722. Failure to deliver a bounce message always caused it to be frozen, even if
3198 there was an errors_to setting on the router. The errors_to setting is now
3199 respected.
bbe902f0
PH
3200
320123. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3202 canonical form (fully expanded) before being placed in
3203 $sender_host_address.
3204
320524. Updated eximstats to version 1.33
3206
ea3a6f44
NM
320725. Include certificate and key file names in error message when GnuTLS fails
3208 to set them up, because the GnuTLS error message doesn't include the name
3209 of the failing file when there is a problem reading it.
bbe902f0
PH
3210
321126. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 3212 As per change 25.
bbe902f0 3213
ea3a6f44
NM
321427. Reset the locale to "C" after calling embedded Perl, in case it was changed
3215 (this can affect the format of dates).
bbe902f0 3216
ea3a6f44
NM
321728. exim_tidydb, when checking for the continued existence of a message for
3218 which it has found a message-specific retry record, was not finding
3219 messages that were in split spool directories. Consequently, it was
3220 deleting retry records that should have stayed in existence.
bbe902f0
PH
3221
322229. eximstats updated to version 1.35
3223 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3224 1.35 - bugfix such that pie charts by volume are generated correctly
3225
ea3a6f44
NM
322630. The SPA authentication driver was not abandoning authentication and moving
3227 on to the next authenticator when an expansion was forced to fail,
3228 contradicting the general specification for all authenticators. Instead it
3229 was generating a temporary error. It now behaves as specified.
bbe902f0 3230
ea3a6f44
NM
323131. The default ordering of permitted cipher suites for GnuTLS was pessimal
3232 (the order specifies the preference for clients). The order is now AES256,
3233 AES128, 3DES, ARCFOUR128.
bbe902f0 3234
ea3a6f44
NM
323531. Small patch to Sieve code - explicitly set From: when generating an
3236 autoreply.
bbe902f0 3237
ea3a6f44
NM
323832. Exim crashed if a remote delivery caused a very long error message to be
3239 recorded - for instance if somebody sent an entire SpamAssassin report back
3240 as a large number of 550 error lines. This bug was coincidentally fixed by
3241 increasing the size of one of Exim's internal buffers (big_buffer) that
3242 happened as part of the Exiscan merge. However, to be on the safe side, I
3243 have made the code more robust (and fixed the comments that describe what
3244 is going on).
bbe902f0 3245
ea3a6f44
NM
324633. Some experimental protocols are using DNS PTR records for new purposes. The
3247 keys for these records are domain names, not reversed IP addresses. The
3248 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3249 leaves it alone. Component reversal etc. now happens only for IP addresses.
bbe902f0
PH
3250 CAN-2005-0021
3251
ea3a6f44
NM
325234. The host_aton() function is supposed to be passed a string that is known
3253 to be a valid IP address. However, in the case of IPv6 addresses, it was
3254 not checking this. This is a hostage to fortune. Exim now panics and dies
3255 if the condition is not met. A case was found where this could be provoked
3256 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3257 components; fortuitously, this particular loophole had already been fixed
3258 by change 4.50/55 or 4.44/33 above.
3259
3260 If there are any other similar loopholes, the new check in host_aton()
3261 itself should stop them being exploited. The report I received stated that
3262 data on the command line could provoke the exploit when Exim was running as
3263 exim, but did not say which command line option was involved. All I could
3264 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3265 running as the user.
bbe902f0
PH
3266 CAN-2005-0021
3267
ea3a6f44
NM
326835. There was a buffer overflow vulnerability in the SPA authentication code
3269 (which came originally from the Samba project). I have added a test to the
3270 spa_base64_to_bits() function which I hope fixes it.
bbe902f0
PH
3271 CAN-2005-0022
3272
ea3a6f44
NM
327336. The daemon start-up calls getloadavg() while still root for those OS that
3274 need the first call to be done as root, but it missed one case: when
3275 deliver_queue_load_max is set with deliver_drop_privilege. This is
3276 necessary for the benefit of the queue runner, because there is no re-exec
3277 when deliver_drop_privilege is set.
bbe902f0 3278
ea3a6f44
NM
327937. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3280 was in use, was not putting the data itself into the right store pool;
3281 consequently, it could be overwritten for a subsequent message in the same
3282 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3283 the caching.)
bbe902f0 3284
ea3a6f44
NM
328538. Sometimes the final signoff response after QUIT could fail to get
3286 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3287 < 0 before doing a fflush(). This bug looks as though it goes back to the
3288 introduction of TLS in release 3.20, but "sometimes" must have been rare
3289 because the tests only now provoked it.
bbe902f0
PH
3290
3291
495ae4b0
PH
3292Exim version 4.43
3293-----------------
3294
3295 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3296 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3297 Now it returns an integer. A number of calls were still expecting a T/F
3298 return. Fortuitously, in all cases, the tests worked in OK situations,
3299 which is the norm. However, things would have gone wrong on any write
3300 failures on the smtp file descriptor. This function is used when sending
3301 messages over SMTP and also when doing verify callouts.
3302
3303 2. When Exim is called to do synchronous delivery of a locally submitted
3304 message (the -odf or -odi options), it no longer closes stderr before doing
3305 the delivery.
3306
3307 3. Implemented the mua_wrapper option.
3308
3309 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3310
3311 5. Implemented the functions header_remove(), header_testname(),
3312 header_add_at_position(), and receive_remove_recipient(), and exported them
3313 to local_scan().
3314
3315 6. If an ACL "warn" statement specified the addition of headers, Exim already
3316 inserted X-ACL-Warn: at the start if there was no header name. However, it
3317 was not making this test for the second and subsequent header lines if
3318 there were newlines in the string. This meant that an invalid header could
3319 be inserted if Exim was badly configured.
3320
3321 7. Allow an ACL "warn" statement to add header lines at the start or after all
3322 the Received: headers, as well as at the end.
3323
3324 8. Added the rcpt_4xx retry error code.
3325
3326 9. Added postmaster_mailfrom=xxx to callout verification option.
3327
332810. Added mailfrom=xxxx to the callout verification option, for verify=
3329 header_sender only.
3330
333111. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3332 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3333
333412. Inserted some casts to stop certain compilers warning when using pointer
3335 differences as field lengths or precisions in printf-type calls (mostly
3336 affecting debugging statements).
3337
333813. Added optional readline() support for -be (dynamically loaded).
3339
334014. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3341 same clock tick as a message's arrival, so that its received time was the
3342 same as the "first fail" time on the retry record, and that message
3343 remained on the queue past the ultimate address timeout, every queue runner
3344 would try a delivery (because it was past the ultimate address timeout) but
3345 after another failure, the ultimate address timeout, which should have then
3346 bounced the address, did not kick in. This was a "< instead of <=" error;
3347 in most cases the first failure would have been in the next clock tick
3348 after the received time, and all would be well.
3349
335015. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3351 being recognized when the domain list was tested by the match_domain
3352 condition in an expansion string.
3353
335416. Added the ${str2b64: operator.
3355
335617. Exim was always calling setrlimit() to set a large limit for the number of
3357 processes, without checking whether the existing limit was already
3358 adequate. (It did check for the limit on file descriptors.) Furthermore,
3359 errors from getrlimit() and setrlimit() were being ignored. Now they are
3360 logged to the main and panic logs, but Exim does carry on, to try to do its
3361 job under whatever limits there are.
3362
336318. Imported PCRE 5.0.
3364
336519. Trivial typo in log message " temporarily refused connection" (the leading
3366 space).
3367
336820. If the log selector return_path_on_delivery was set and an address was
3369 redirected to /dev/null, the delivery process crashed because it assumed
3370 that a return path would always be set for a "successful" delivery. In this
3371 case, the whole delivery is bypassed as an optimization, and therefore no
3372 return path is set.
3373
337421. Internal re-arrangement: the function for sending a challenge and reading
3375 a response while authentication was assuming a zero-terminated challenge
3376 string. It's now changed to take a pointer and a length, to allow for
3377 binary data in such strings.
3378
337922. Added the cyrus_sasl authenticator (code supplied by MBM).
3380
338123. Exim was not respecting finduser_retries when seeking the login of the
3382 uid under which it was called; it was always trying 10 times. (The default
3383 setting of finduser_retries is zero.) Also, it was sleeping after the final
3384 failure, which is pointless.
3385
338624. Implemented tls_on_connect_ports.
3387
338825. Implemented acl_smtp_predata.
3389
339026. If the domain in control=submission is set empty, Exim assumes that the
3391 authenticated id is a complete email address when it generates From: or
3392 Sender: header lines.
3393
339427. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3395 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3396 chown and chgrp in /bin and hostname in /usr/bin.
3397
339828. Exim was keeping the "process log" file open after each use, just as it
3399 does for the main log. This opens the possibility of it remaining open for
3400 long periods when the USR1 signal hits a daemon. Occasional processlog
3401 errors were reported, that could have been caused by this. Anyway, it seems
3402 much more sensible not to leave this file open at all, so that is what now
3403 happens.
3404
340529. The long-running daemon process does not normally write to the log once it
3406 has entered its main loop, and it closes the log before doing so. This is
3407 so that log files can straightforwardly be renamed and moved. However,
3408 there are a couple of unusual error situations where the daemon does write
3409 log entries, and I had neglected to close the log afterwards.
3410
341130. The text of an SMTP error response that was received during a remote
3412 delivery was being truncated at 512 bytes. This is too short for some of
3413 the long messages that one sometimes sees. I've increased the limit to
3414 1024.
3415
341631. It is now possible to make retry rules that apply only when a message has a
3417 specific sender, in particular, an empty sender.
3418
341932. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3420 it possible to be selective about when SMTP synchronization is enforced.
3421
342233. Added "control = caseful_local_part" and "control = "caselower_local_part".