Avoid using a waiting db for single-message-only transports. Performance
[exim.git] / doc / doc-txt / ChangeLog
CommitLineData
495ae4b0
PH
1Change log file for Exim from version 4.21
2-------------------------------------------
3
98a90c36
PP
4Exim version 4.81
5-----------------
6
7PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
8
12f69989
PP
9PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
11
54c90be1
PP
12PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
13 by GnuTLS.
14
1f4a55da
PP
15PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
17 routines.
18
13363eba 19PP/05 DSCP support for outbound connections and control modifier for inbound.
36a3ae5f 20
66645890 21PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
e402235f
PP
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
66645890 26
a3fb9793 27PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
f4ee74ac
PP
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
a3fb9793 31 Also parse but ignore: -Ac -Am -X<logfile>
f4ee74ac 32 Bugzilla 1117.
a3fb9793 33
d27f98fe 34TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
98a90c36 35
6822b909
TL
36TL/02 Add +smtp_confirmation as a default logging option.
37
e7568d51
TL
38TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
40
ae0e32ee 41TL/04 Bugzilla 1281 - Spec typo.
ca0ff207 42 Bugzilla 1283 - Spec typo.
97f42f10 43 Bugzilla 1290 - Spec grammar fixes.
ca0ff207
TL
44
45TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
ae0e32ee 46
976b7e9f
JH
47JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
48
be4a1376
JH
49JH/02 Support "G" suffix to numbers in ${if comparisons.
50
ec4b68e5
PP
51PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
52
d7148a07
NM
53NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
ec4b68e5 55
585121e2 56JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
ec4b68e5 57
a5f239e4
PP
58PP/09 Add gnutls_enable_pkcs11 option.
59
13d08c90
PP
60PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
62
bef3ea7f
JH
63JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
65 "acl = name arg..."
a5f239e4 66
846726c5
JH
67JH/05 Permit multiple router/transport headers_add/remove lines.
68
3a796370
JH
69JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
70
ea722490
JH
71JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher.
73
b1770b6e 74Exim version 4.80
0599f9cf
PP
75-----------------
76
77PP/01 Handle short writes when writing local log-files.
78 In practice, only affects FreeBSD (8 onwards).
79 Bugzilla 1053, with thanks to Dmitry Isaikin.
80
23c7e742
NM
81NM/01 Bugzilla 949 - Documentation tweak
82
b322aac8
NM
83NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
84 improved.
85
4a891427
NM
86NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
87
c1e794ba 88PP/02 Implemented gsasl authenticator.
b322aac8 89
97753960
PP
90PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
91
92PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
93 `pkg-config foo` for cflags/libs.
94
df6303fa
PP
95PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
96 with rest of GSASL and with heimdal_gssapi.
97
7e6a8985
PP
98PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
99 `pkg-config foo` for cflags/libs for the TLS implementation.
100
f1e05cc7 101PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
20aa9dbd
PP
102 properties get this fed in as external SSF. A number of robustness
103 and debugging improvements to the cyrus_sasl authenticator.
b322aac8 104
4c287009
PP
105PP/08 cyrus_sasl server now expands the server_realm option.
106
b98bb9ac
PP
107PP/09 Bugzilla 1214 - Log authentication information in reject log.
108 Patch by Jeremy Harris.
109
4a6a987a
PP
110PP/10 Added dbmjz lookup type.
111
c45dd180 112PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
c7955b11 113
7db8d074
PP
114PP/12 MAIL args handles TAB as well as SP, for better interop with
115 non-compliant senders.
116 Analysis and variant patch by Todd Lyons.
117
eae0036b 118NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
e0df1c83
DM
119 Bug report from Lars Müller <lars@samba.org> (via SUSE),
120 Patch from Dirk Mueller <dmueller@suse.com>
121
dec5017e
PP
122PP/13 tls_peerdn now print-escaped for spool files.
123 Observed some $tls_peerdn in wild which contained \n, which resulted
124 in spool file corruption.
125
c80c5570
PP
126PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
127 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
128 or write after TLS renegotiation, which otherwise led to messages
129 "Got SSL error 2".
130
076b11e2
PP
131TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
132 as a tracking header (ie: a signed header comes before the signature).
133 Patch from Wolfgang Breyha.
134
5407bfff
JH
135JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
136 comma-sep list; embedded commas doubled.
137
9e45c72b
PP
138JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
139
e74376d8
PP
140PP/15 LDAP: Check for errors of TLS initialisation, to give correct
141 diagnostics.
142 Report and patch from Dmitry Banschikov.
143
da3ad30d
PP
144PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
145 Removed SSL_clear() after SSL_new() which led to protocol negotiation
146 failures. We appear to now support TLS1.1+ with Exim.
147
7be682ca
PP
148PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
149 lets Exim select keys and certificates based upon TLS SNI from client.
3f0945ff
PP
150 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
151 before an outbound SMTP session. New log_selector, +tls_sni.
7be682ca 152
ef840681
PP
153PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
154 NULL dereference. Report and patch from Alun Jones.
155
5bfb4cdf
PP
156PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
157 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
158 Not seeing resolver debug output on NetBSD, but suspect this is a
159 resolver implementation change.
160
c6e95d22
PP
161PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
162 Left warnings. Added "eximon gdb" invocation mode.
163
9cbad13b
PP
164PP/21 Defaulting "accept_8bitmime" to true, not false.
165
9ee44efb
PP
166PP/22 Added -bw for inetd wait mode support.
167
6a6084f8
PP
168PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
169 locate the relevant includes and libraries. Made this the default.
170
12dd53c7
PP
171PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
172 Bugzilla 1246, report and most of solution from Tomasz Kusy.
173
9e45c72b 174JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
97d17305
JH
175 This may cause build issues on older platforms.
176
17c76198
PP
177PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
178 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
179 gnutls_require_mac & gnutls_require_protocols (no longer supported).
180 Added SNI support via GnuTLS too.
af3498d6 181 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
17c76198 182
53947857 183PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3f7eeb86 184
eae0036b 185PP/27 Applied dnsdb SPF support patch from Janne Snabb.
8ee4b30e
PP
186 Applied second patch from Janne, implementing suggestion to default
187 multiple-strings-in-record handling to match SPF spec.
eae0036b 188
9e45c72b 189JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2605c55b 190
7390e768
PP
191PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
192 read-only, out of scope).
193 Patch from Wolfgang Breyha, report from Stuart Northfield.
194
08488c86
PP
195PP/29 Fix three issues highlighted by clang analyser static analysis.
196 Only crash-plausible issue would require the Cambridge-specific
197 iplookup router and a misconfiguration.
198 Report from Marcin Mirosław.
199
6475bd82
PP
200PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
201
81f91683
PP
202PP/31 %D in printf continues to cause issues (-Wformat=security), so for
203 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
204 As part of this, removing so much warning spew let me fix some minor
205 real issues in debug logging.
206
5779e6aa
PP
207PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
208 assignment on my part. Fixed.
209
3375e053
PP
210PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
211 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
212 Janne Snabb (who went above and beyond: thank you).
213
214PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
215 string otherwise requires a connection and a bunch more work and it's
78e0c7a3
PP
216 relatively easy to get wrong. Should also expose TLS library linkage
217 problems.
3375e053 218
9d26b8c0
PP
219PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
220 64-bit ${eval} (JH/03).
221
57eb9e91 222PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
b87a6e0e
PP
223 GNU libc to support some of the 64-bit stuff, should not lead to
224 conflicts. Defined before os.h is pulled in, so if a given platform
225 needs to override this, it can.
226
16880d1a
PP
227PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
228 protection layer was required, which is not implemented.
229 Bugzilla 1254, patch from Wolfgang Breyha.
230
a799883d
PP
231PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
232 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
233 tls_dhparam take prime identifiers. Also unbreak combination of
234 OpenSSL+DH_params+TLSSNI.
235
3ecab157 236PP/39 Disable SSLv2 by default in OpenSSL support.
f0f5a555 237
0599f9cf 238
867fcbf5
PP
239Exim version 4.77
240-----------------
241
242PP/01 Solaris build fix for Oracle's LDAP libraries.
243 Bugzilla 1109, patch from Stephen Usher.
244
f1a29782
TF
245TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
246
ab42bd23
TK
247TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
248 whitespace trailer
867fcbf5 249
0ca0cf52
TF
250TF/02 Fix a couple more cases where we did not log the error message
251 when unlink() failed. See also change 4.74-TF/03.
252
921b12ca
TF
253TF/03 Make the exiwhat support code safe for signals. Previously Exim might
254 lock up or crash if it happened to be inside a call to libc when it
255 got a SIGUSR1 from exiwhat.
256
257 The SIGUSR1 handler appends the current process status to the process
258 log which is later printed by exiwhat. It used to use the general
259 purpose logging code to do this, but several functions it calls are
260 not safe for signals.
261
262 The new output code in the SIGUSR1 handler is specific to the process
263 log, and simple enough that it's easy to inspect for signal safety.
264 Removing some special cases also simplifies the general logging code.
265 Removing the spurious timestamps from the process log simplifies
266 exiwhat.
267
c99ce5c9
TF
268TF/04 Improved ratelimit ACL condition.
269
270 The /noupdate option has been deprecated in favour of /readonly which
271 has clearer semantics. The /leaky, /strict, and /readonly update modes
272 are mutually exclusive. The update mode is no longer included in the
273 database key; it just determines when the database is updated. (This
274 means that when you upgrde Exim will forget old rate measurements.)
275
276 Exim now checks that the per_* options are used with an update mode that
277 makes sense for the current ACL. For example, when Exim is processing a
278 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
279 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
280 must specify per_mail/readonly. If you omit the update mode it defaults to
281 /leaky where that makes sense (as before) or /readonly where required.
282
283 The /noupdate option is now undocumented but still supported for
284 backwards compatibility. It is equivalent to /readonly except that in
285 ACLs where /readonly is required you may specify /leaky/noupdate or
286 /strict/noupdate which are treated the same as /readonly.
287
288 A useful new feature is the /count= option. This is a generalization
289 of the per_byte option, so that you can measure the throughput of other
290 aggregate values. For example, the per_byte option is now equivalent
291 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
292
293 The per_rcpt option has been generalized using the /count= mechanism
294 (though it's more complicated than the per_byte equivalence). When it is
295 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
296 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
297 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
298 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
299 Note that using per_rcpt with a non-readonly update mode in more than
300 one ACL will cause the recipients to be double-counted. (The per_mail
301 and per_byte options don't have this problem.)
302
303 The handling of very low rates has changed slightly. If the computed rate
304 is less than the event's count (usually one) then this event is the first
305 after a long gap. In this case the rate is set to the same as this event's
306 count, so that the first message of a spam run is counted properly.
307
308 The major new feature is a mechanism for counting the rate of unique
309 events. The new per_addr option counts the number of different
310 recipients that someone has sent messages to in the last time period. It
311 behaves like per_rcpt if all the recipient addresses are different, but
312 duplicate recipient addresses do not increase the measured rate. Like
313 the /count= option this is a general mechanism, so the per_addr option
314 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
315 example, measure the rate that a client uses different sender addresses
316 with the options per_mail/unique=$sender_address. There are further
317 details in the main documentation.
318
3634fc25
TF
319TF/05 Removed obsolete $Cambridge$ CVS revision strings.
320
792e8a19
TF
321TF/06 Removed a few PCRE remnants.
322
5901f0ab
TF
323TF/07 Automatically extract Exim's version number from tags in the git
324 repository when doing development or release builds.
325
7f2a2a43
PP
326PP/02 Raise smtp_cmd_buffer_size to 16kB.
327 Bugzilla 879. Patch from Paul Fisher.
e2ca7082 328
061b7ebd
PP
329PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
330 Heavily based on revision 40f9a89a from Simon Arlott's tree.
331 Bugzilla 97.
332
e12f8c32
PP
333PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
334
9e949f00 335PP/05 Variable $av_failed, true if the AV scanner deferred.
7f2a2a43
PP
336 Bugzilla 1078. Patch from John Horne.
337
338PP/06 Stop make process more reliably on build failure.
339 Bugzilla 1087. Patch from Heiko Schlittermann.
9e949f00 340
555ae6af 341PP/07 Make maildir_use_size_file an _expandable_ boolean.
ac53fcda
PP
342 Bugzilla 1089. Patch from Heiko Schlittermann.
343
344PP/08 Handle ${run} returning more data than OS pipe buffer size.
345 Bugzilla 1131. Patch from Holger Weiß.
555ae6af 346
6f7fe114
PP
347PP/09 Handle IPv6 addresses with SPF.
348 Bugzilla 860. Patch from Wolfgang Breyha.
349
c566dd90
PP
350PP/10 GnuTLS: support TLS 1.2 & 1.1.
351 Bugzilla 1156.
89f897c3
PP
352 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
353 Bugzilla 1095.
c566dd90 354
d6cc7c78 355PP/11 match_* no longer expand right-hand-side by default.
39257585
PP
356 New compile-time build option, EXPAND_LISTMATCH_RHS.
357 New expansion conditions, "inlist", "inlisti".
358
0d0e4455
PP
359PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
360
3399bb60 361PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
d690cbdc
PP
362
363PP/14 fix log_write() format string regression from TF/03.
364 Bugzilla 1152. Patch from Dmitry Isaikin.
365
0ca0cf52 366
10906672
PP
367Exim version 4.76
368-----------------
369
370PP/01 The new ldap_require_cert option would segfault if used. Fixed.
371
754a0503
PP
372PP/02 Harmonised TLS library version reporting; only show if debugging.
373 Layout now matches that introduced for other libraries in 4.74 PP/03.
374
c0c7b2da
PP
375PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
376
e97d1f08
PP
377PP/04 New "dns_use_edns0" global option.
378
084c1d8c
PP
379PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
380 Bugzilla 1098.
381
4e7ee012
PP
382PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
383 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
da80c2a8 384
c8d52a00
PP
385TK/01 Updated PolarSSL code to 0.14.2.
386 Bugzilla 1097. Patch from Andreas Metzler.
387
54e7ce4a
PP
388PP/07 Catch divide-by-zero in ${eval:...}.
389 Fixes bugzilla 1102.
390
5ee6f336
PP
391PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
392 Bugzilla 1104.
393
c8d52a00 394TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
6ea4a851
PP
395 format-string attack -- SECURITY: remote arbitrary code execution.
396
397TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
398 time unintentionally subject to list matching rules, letting the header
399 cause arbitrary Exim lookups (of items which can occur in lists, *not*
400 arbitrary string expansion). This allowed for information disclosure.
401
402PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
403 INT_MIN/-1 -- value coerced to INT_MAX.
c8d52a00 404
10906672 405
aa097c4c
NM
406Exim version 4.75
407-----------------
408
4c9ef03a 409NM/01 Workround for PCRE version dependency in version reporting
aa097c4c
NM
410 Bugzilla 1073
411
7f3d9eff
TF
412TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
413 This fixes portability to compilers other than gcc, notably
414 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
415
159f52d2
TF
416TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
417 makefiles for portability to HP-UX and POSIX correctness.
418
0cc9542a
PP
419PP/01 Permit LOOKUP_foo enabling on the make command-line.
420 Also via indented variable definition in the Makefile.
421 (Debugging by Oliver Heesakkers).
422
f7274286
PP
423PP/02 Restore caching of spamd results with expanded spamd_address.
424 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
425
7b797365
PP
426PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
427 Improves build reliability. Fix from: Frank Elsner
428
caacae52
NM
429NM/02 Fix wide character breakage in the rfc2047 coding
430 Fixes bug 1064. Patch from Andrey N. Oktyabrski
431
09dcaba9
NM
432NM/03 Allow underscore in dnslist lookups
433 Fixes bug 1026. Patch from Graeme Fowler
434
bc19a55b
PP
435PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
436 Code patches from Adam Ciarcinski of NetBSD.
caacae52 437
bd4c9759
NM
438NM/04 Fixed exiqgrep to cope with mailq missing size issue
439 Fixes bug 943.
440
b72aab72
PP
441PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
442 is logged, to avoid truncation. Patch from John Horne.
443
2fe76745
PP
444PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
445 Patch from Jakob Hirsch.
446
76aa570c
PP
447PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
448 SQL string expansion failure details.
449 Patch from Andrey Oktyabrski.
450
f1e5fef5
PP
451PP/08 Bugzilla 486: implement %M datestamping in log filenames.
452 Patch from Simon Arlott.
453
4d805ee9
PP
454PP/09 New lookups functionality failed to compile on old gcc which rejects
455 extern declarations in function scope.
456 Patch from Oliver Fleischmann
457
cd59ab18
PP
458PP/10 Use sig_atomic_t for flags set from signal handlers.
459 Check getgroups() return and improve debugging.
460 Fixed developed for diagnosis in bug 927 (which turned out to be
461 a kernel bug).
462
332f5cf3
PP
463PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
464 Patch from Mark Zealey.
465
29cfeb94
PP
466PP/12 Bugzilla 1056: Improved spamd server selection.
467 Patch from Mark Zealey.
468
660242ad
PP
469PP/13 Bugzilla 1086: Deal with maildir quota file races.
470 Based on patch from Heiko Schlittermann.
471
bc4bc4c5
PP
472PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
473 Patch from Uwe Doering, sign-off by Michael Haardt.
474
2e64baa9
NM
475NM/05 Fix to spam.c to accommodate older gcc versions which dislike
476 variable declaration deep within a block. Bug and patch from
477 Dennis Davis.
478
bddd7526
PP
479PP/15 lookups-Makefile IRIX compatibilty coercion.
480
6bac1a9a
PP
481PP/16 Make DISABLE_DKIM build knob functional.
482
552193f0
NM
483NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
484 Patch by Simon Arlott
baeee2c1 485
1b587e48
TF
486TF/03 Fix valgrind.h portability to C89 compilers that do not support
487 variable argument macros. Our copy now differs from upstream.
488
aa097c4c 489
8c07b69f
TF
490Exim version 4.74
491-----------------
492
493TF/01 Failure to get a lock on a hints database can have serious
494 consequences so log it to the panic log.
495
c0ea85ab
TF
496TF/02 Log LMTP confirmation messages in the same way as SMTP,
497 controlled using the smtp_confirmation log selector.
498
0761d44e
TF
499TF/03 Include the error message when we fail to unlink a spool file.
500
0a349494
PP
501DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
502 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
503 for maintaining out-of-tree patches for some time.
504
505PP/01 Bugzilla 139: Documentation and portability issues.
506 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
507 Handle per-OS dynamic-module compilation flags.
508
fea24b2e
PP
509PP/02 Let /dev/null have normal permissions.
510 The 4.73 fixes were a little too stringent and complained about the
511 permissions on /dev/null. Exempt it from some checks.
512 Reported by Andreas M. Kirchwitz.
513
6545de78
PP
514PP/03 Report version information for many libraries, including
515 Exim version information for dynamically loaded libraries. Created
516 version.h, now support a version extension string for distributors
517 who patch heavily. Dynamic module ABI change.
518
1670ef10
PP
519PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
520 privilege escalation vulnerability whereby the Exim run-time user
521 can cause root to append content of the attacker's choosing to
522 arbitrary files.
523
c0886197
PP
524PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
525 (Wolfgang Breyha)
526
b7487bce
PP
527PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
528 If dropping privileges for untrusted macros, we disabled normal logging
529 on the basis that it would fail; for the Exim run-time user, this is not
530 the case, and it resulted in successful deliveries going unlogged.
531 Fixed. Reported by Andreas Metzler.
532
8c07b69f 533
97fd1e48 534Exim version 4.73
ed7f7860 535-----------------
97fd1e48
PP
536
537PP/01 Date: & Message-Id: revert to normally being appended to a message,
538 only prepend for the Resent-* case. Fixes regression introduced in
539 Exim 4.70 by NM/22 for Bugzilla 607.
540
6901c596
PP
541PP/02 Include check_rfc2047_length in configure.default because we're seeing
542 increasing numbers of administrators be bitten by this.
543
a8c8d6b5
JJ
544JJ/01 Added DISABLE_DKIM and comment to src/EDITME
545
77bb000f
PP
546PP/03 Bugzilla 994: added openssl_options main configuration option.
547
a29e5231
PP
548PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
549
ec5a0394 550PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
a29e5231 551
55c75993
PP
552PP/06 Adjust NTLM authentication to handle SASL Initial Response.
553
453a6645 554PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
ec5a0394
PP
555 without a peer certificate, leading to a segfault because of an
556 assumption that peers always have certificates. Be a little more
453a6645
PP
557 paranoid. Problem reported by Martin Tscholak.
558
8544e77a
PP
559PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
560 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
561 NB: ClamAV planning to remove STREAM in "middle of 2010".
3346ab01
PP
562 CL also introduces -bmalware, various -d+acl logging additions and
563 more caution in buffer sizes.
8544e77a 564
83e029d5
PP
565PP/09 Implemented reverse_ip expansion operator.
566
ed7f7860
PP
567PP/10 Bugzilla 937: provide a "debug" ACL control.
568
7d9f747b
PP
569PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
570
4b2241d2
PP
571PP/12 Bugzilla 973: Implement --version.
572
10385c15
PP
573PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
574
dbc4b90d
PP
575PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
576
532be449
PP
577PP/15 Bugzilla 816: support multiple condition rules on Routers.
578
6a8de854 579PP/16 Add bool_lax{} expansion operator and use that for combining multiple
71265ae9
PP
580 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
581 ignore trailing whitespace.
6a8de854 582
5dc43717
JJ
583JJ/02 prevent non-panic DKIM error from being sent to paniclog
584
585JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
586 "exim" to be used
55c75993 587
3346ab01
PP
588PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
589 Notification from Dr Andrew Aitchison.
590
491fab4c
PP
591PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
592 ExtendedDetectionInfo response format.
593 Notification from John Horne.
594
13eb9497
PP
595PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
596 compatible.
597
598PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
599 XSL and documented dependency on system catalogs, with examples of how
600 it normally works.
601
7f36d675
DW
602DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
603 access.
604
c1d94452
DW
605DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
606 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
607 configuration file which is writeable by the Exim user or group.
608
e2f5dc15
DW
609DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
610 of configuration files to cover files specified with the -C option if
611 they are going to be used with root privileges, not just the default
612 configuration file.
613
cd25e41d
DW
614DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
615 option (effectively making it always true).
616
261dc43e
DW
617DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
618 files to be used while preserving root privileges.
619
fa32850b
DW
620DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
621 that rogue child processes cannot use them.
622
79d4bc3d
PP
623PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
624 run-time user, instead of root.
625
43236f35 626PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2cfd3221
PP
627 Exim run-time user without dropping privileges.
628
fb08281f
DW
629DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
630 result string, instead of calling string_vformat() twice with the same
631 arguments.
3346ab01 632
74935b98
DW
633DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
634 for other users. Others should always drop root privileges if they use
635 -C on the command line, even for a whitelisted configure file.
636
90b6341f
DW
637DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
638
57730b52
ML
639NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
640
66581d1e 641
465e92cf
JJ
642Exim version 4.72
643-----------------
644
453a6645
PP
645JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
646 $data_path, and $header_path variables; fixed documentation bugs and
647 typos
465e92cf 648
453a6645
PP
649JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
650 exipick to access non-standard spools, including the "frozen" queue
651 (Finput)
edae0343 652
9bd3e22c
NM
653NM/01 Bugzilla 965: Support mysql stored procedures.
654 Patch from Alain Williams
655
bb576ff7
NM
656NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
657
5a1a5845
NM
658NM/03 Bugzilla 955: Documentation fix for max_rcpts.
659 Patch from Andreas Metzler
660
981a9fad
NM
661NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
662 Patch from Kirill Miazine
663
7fc497ee
NM
664NM/05 Bugzilla 671: Added umask to procmail example.
665
1a41defa
JJ
666JJ/03 installed exipick 20100323.0, fixing doc bug
667
a466095c 668NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
b26eacf1 669 directory. Notification and patch from Dan Rosenberg.
a466095c 670
94a6bd0b
NM
671TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
672
673TK/02 Improve log output when DKIM signing operation fails.
674
675MH/01 Treat the transport option dkim_domain as a colon separated
676 list, not as a single string, and sign the message with each element,
677 omitting multiple occurences of the same signer.
678
c1b141a8
NM
679NM/07 Null terminate DKIM strings, Null initialise DKIM variable
680 Bugzilla 985, 986. Patch by Simon Arlott
94a6bd0b 681
b26eacf1 682NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
0d0c6357
NM
683 Patch by Simon Arlott
684
179c5980 685PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
b26eacf1 686 MBX locking. Notification from Dan Rosenberg.
179c5980 687
9bd3e22c 688
7c6d71af
NM
689Exim version 4.71
690-----------------
691
7d9f747b 692TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
7c6d71af 693
f013fb92
NM
694NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
695
0eb8eedd
NM
696NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
697
663ee6d9
NM
698NM/03 Bugzilla 847: Enable DNSDB lookup by default.
699
177ebd9b
NM
700NM/04 Bugzilla 915: Flag broken perl installation during build.
701
7c6d71af 702
210f147e
NM
703Exim version 4.70
704-----------------
705
cdd3bb85 706TK/01 Added patch by Johannes Berg that expands the main option
e739e3d9 707 "spamd_address" if it starts with a dollar sign.
cdd3bb85
TK
708
709TK/02 Write list of recipients to X-Envelope-Sender header when building
710 the mbox-format spool file for content scanning (suggested by Jakob
7d9f747b 711 Hirsch).
cdd3bb85
TK
712
713TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
714 (http://www.dcc-servers.net/) support via dccifd. Activated by
e739e3d9 715 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
cdd3bb85
TK
716
717TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
718 by Mark Daniel Reidel <mr@df.eu>.
719
210f147e
NM
720NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
721 When building exim an external PCRE library is now needed -
722 PCRE is a system library on the majority of modern systems.
723 See entry on PCRE_LIBS in EDITME file.
724
deafd5b3
NM
725NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
726 conversation. Added nologin parameter to request.
7d9f747b 727 Patch contributed by Kirill Miazine.
deafd5b3 728
089793a4
TF
729TF/01 Do not log submission mode rewrites if they do not change the address.
730
5f16ca82
TF
731TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
732
dae9d94e 733NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
7d9f747b 734 log files in place. Contributed by Roberto Lima.
dae9d94e 735
7d9f747b 736NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3f0da4d0 737
06864c44
TF
738TF/03 Bugzilla 615: When checking the local_parts router precondition
739 after a local_part_suffix or local_part_prefix option, Exim now
740 does not use the address's named list lookup cache, since this
741 contains cached lookups for the whole local part.
742
65a7d8c3 743NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
7d9f747b 744 Robert Millan. Documentation is in experimental-spec.txt.
65a7d8c3 745
23510047 746TF/04 Bugzilla 668: Fix parallel build (make -j).
65a7d8c3 747
7d9f747b 748NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
5f28a6e8 749
7d8eec3a 750NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
7d9f747b 751 Patch provided by Jan Srzednicki.
6c588e74 752
89dec7b6
TF
753TF/05 Leading white space used to be stripped from $spam_report which
754 wrecked the formatting. Now it is preserved.
5f28a6e8 755
a99de90c
TF
756TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
757 that they are available at delivery time.
758
e2803e40
TF
759TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
760
7199e1ee
TF
761TF/08 TLS error reporting now respects the incoming_interface and
762 incoming_port log selectors.
763
e276e04b
TF
764TF/09 Produce a more useful error message if an SMTP transport's hosts
765 setting expands to an empty string.
766
ce552449 767NM/06 Bugzilla 744: EXPN did not work under TLS.
7d9f747b 768 Patch provided by Phil Pennock.
ce552449 769
e765a0f1 770NM/07 Bugzilla 769: Extraneous comma in usage fprintf
7d9f747b 771 Patch provided by Richard Godbee.
e765a0f1 772
4f054c63 773NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
447de4b0 774 acl_smtp_notquit, added index entry.
4f054c63 775
7d9f747b
PP
776NM/09 Bugzilla 787: Potential buffer overflow in string_format.
777 Patch provided by Eugene Bujak.
24c929a2 778
7d9f747b
PP
779NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
780 accept(). Patch provided by Maxim Dounin.
cf73943b 781
b52bc06e 782NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
7d9f747b 783 Patch provided by Phil Pennock.
b52bc06e 784
447de4b0
NM
785NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
786
4c69d561 787NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
7d9f747b 788 Patch provided by Brad "anomie" Jorsch.
4c69d561 789
d5c39246 790NM/14 Bugzilla 622: Added timeout setting to transport filter.
7d9f747b 791 Patch provided by Dean Brooks.
9b989985 792
0b23848a
TK
793TK/05 Add native DKIM support (does not depend on external libraries).
794
8f3414a1 795NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
7d9f747b 796 Patch provided by Graeme Fowler.
e2aacdfd 797
fb6f955d
NM
798NM/16 Bugzilla 851: Documentation example syntax fix.
799
800NM/17 Changed NOTICE file to remove references to embedded PCRE.
8f3414a1 801
7d9f747b
PP
802NM/18 Bugzilla 894: Fix issue with very long lines including comments in
803 lsearch.
dbb0bf41 804
7d9f747b
PP
805NM/19 Bugzilla 745: TLS version reporting.
806 Patch provided by Phil Pennock.
f3766eb5 807
7d9f747b
PP
808NM/20 Bugzilla 167: bool: condition support.
809 Patch provided by Phil Pennock.
36f12725 810
7d9f747b
PP
811NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
812 clients. Patch provided by Phil Pennock.
e6060e2c 813
7d9f747b
PP
814NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
815 Patch provided by Brad "anomie" Jorsch.
5eb690a1 816
7d9f747b
PP
817NM/23 Bugzilla 687: Fix misparses in eximstats.
818 Patch provided by Heiko Schlittermann.
d5c13d66 819
7d9f747b
PP
820NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
821 Patch provided by Heiko Schlittermann.
b2335c0b 822
7d9f747b 823NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1da77999 824 plus update to original patch.
f4cd9433 825
7d9f747b 826NM/26 Bugzilla 799: Documentation correction for ratelimit.
dc988b7e 827
7d9f747b
PP
828NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
829 Patch provided by David Brownlee.
8dc71ab3 830
7d9f747b 831NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
400eda43 832
7d9f747b 833NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
ec5a421b 834
7d9f747b 835NM/30 Bugzilla 888: TLS documentation bugfixes.
07af267e 836
7d9f747b 837NM/31 Bugzilla 896: Dovecot buffer overrun fix.
51473862 838
17792b53 839NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
7d9f747b 840 Unlike the original bugzilla I have changed all shell scripts in src tree.
17792b53 841
7d9f747b
PP
842NM/33 Bugzilla 898: Transport filter timeout fix.
843 Patch by Todd Rinaldo.
52383f8f 844
91576cec 845NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
7d9f747b 846 Patch by Serge Demonchaux.
5ca6d115 847
7d9f747b
PP
848NM/35 Bugzilla 39: Base64 decode bug fixes.
849 Patch by Jakob Hirsch.
baee9eee 850
7d9f747b 851NM/36 Bugzilla 909: Correct connect() call in dcc code.
e93a964c 852
7d9f747b 853NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
9bf3d68f 854
7d9f747b 855NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
96535b98 856
7d9f747b 857NM/39 Bugzilla 911: Fixed MakeLinks build script.
30339e0f 858
deafd5b3 859
47db1125
NM
860Exim version 4.69
861-----------------
862
4b3504d0
TK
863TK/01 Add preliminary DKIM support. Currently requires a forked version of
864 ALT-N's libdkim that I have put here:
865 http://duncanthrax.net/exim-experimental/
866
867 Note to Michael Haardt: I had to rename some vars in sieve.c. They
868 were called 'true' and it seems that C99 defines that as a reserved
869 keyword to be used with 'bool' variable types. That means you could
870 not include C99-style headers which use bools without triggering
871 build errors in sieve.c.
872
81ea09ca
NM
873NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
874 as mailq or other aliases. Changed the --help handling significantly
875 to do whats expected. exim_usage() emits usage/help information.
876
f13cddcb
SC
877SC/01 Added the -bylocaldomain option to eximstats.
878
7d9f747b 879NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
8ad076b2 880
7d9f747b 881NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
a843aaa6 882
7d9f747b 883NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
47db1125
NM
884
885
eb4c0de6
PH
886Exim version 4.68
887-----------------
888
889PH/01 Another patch from the Sieve maintainer.
890
6a3bceb1
PH
891PH/02 When an IPv6 address is converted to a string for single-key lookup
892 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
893 dots are used instead of colons so that keys in lsearch files need not
894 contain colons. This was done some time before quoting was made available
895 in lsearch files. However, iplsearch files do require colons in IPv6 keys
896 (notated using the quote facility) so as to distinguish them from IPv4
897 keys. This meant that lookups for IP addresses in host lists did not work
898 for iplsearch lookups.
899
900 This has been fixed by arranging for IPv6 addresses to be expressed with
901 colons if the lookup type is iplsearch. This is not incompatible, because
902 previously such lookups could never work.
903
904 The situation is now rather anomolous, since one *can* have colons in
905 ordinary lsearch keys. However, making the change in all cases is
906 incompatible and would probably break a number of configurations.
907
2e30fa9d
TK
908TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
909 version.
910
0806a9c5
MH
911MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
912 conversion specification without a maximum field width, thereby enabling
913 a rogue spamd server to cause a buffer overflow. While nobody in their
914 right mind would setup Exim to query an untrusted spamd server, an
915 attacker that gains access to a server running spamd could potentially
916 exploit this vulnerability to run arbitrary code as the Exim user.
917
ae276964
TK
918TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
919 $primary_hostname instead of what libspf2 thinks the hosts name is.
920
0f2cbd1b
MH
921MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
922 a directory entry by the name of the lookup key. Previously, if a
923 symlink pointed to a non-existing file or a file in a directory that
924 Exim lacked permissions to read, a lookup for a key matching that
925 symlink would fail. Now it is enough that a matching directory entry
926 exists, symlink or not. (Bugzilla 503.)
927
2b85bce7
PH
928PH/03 The body_linecount and body_zerocount variables are now exported in the
929 local_scan API.
930
93655c46
PH
931PH/04 Added the $dnslist_matched variable.
932
6c512171
PH
933PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
934 This means they are set thereafter only if the connection becomes
935 encrypted.
936
937PH/06 Added the client_condition to authenticators so that some can be skipped
938 by clients under certain conditions.
939
aa6dc513
PH
940PH/07 The error message for a badly-placed control=no_multiline_responses left
941 "_responses" off the end of the name.
942
a96603a0
PH
943PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
944
8f240103
PH
945PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
946 (without spaces) instead of just copying the configuration text.
947
948PH/10 Added the /noupdate option to the ratelimit ACL condition.
949
d677b2f2
PH
950PH/11 Added $max_received_linelength.
951
d52120f2
PH
952PH/12 Added +ignore_defer and +include_defer to host lists.
953
64f2600a
PH
954PH/13 Installed PCRE version 7.2. This needed some changes because of the new
955 way in which PCRE > 7.0 is built.
956
8669f003
PH
957PH/14 Implemented queue_only_load_latch.
958
a4dc33a8
PH
959PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
960 MAIL command. The effect was to mangle the value on 64-bit systems.
961
d6a60c0f
PH
962PH/16 Another patch from the Sieve maintainer.
963
8f128379
PH
964PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
965
8932dffe
PH
966PH/18 If a system quota error occurred while trying to create the file for
967 a maildir delivery, the message "Mailbox is full" was not appended to the
968 bounce if the delivery eventually timed out. Change 4.67/27 below applied
969 only to a quota excession during the actual writing of the file.
d6a60c0f 970
ddea74fa 971PH/19 It seems that peer DN values may contain newlines (and other non-printing
48ed62d9
PH
972 characters?) which causes problems in log lines. The DN values are now
973 passed through string_printing() before being added to log lines.
974
ddea74fa 975PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
b7670459
PH
976 and InterBase are left for another time.)
977
ddea74fa
PH
978PH/21 Added message_body_newlines option.
979
ce9f225c
PH
980PH/22 Guard against possible overflow in moan_check_errorcopy().
981
19897d52
PH
982PH/23 POSIX allows open() to be a macro; guard against that.
983
bc64a74d
PH
984PH/24 If the recipient of an error message contained an @ in the local part
985 (suitably quoted, of course), incorrect values were put in $domain and
986 $local_part during the evaluation of errors_copy.
987
eb4c0de6 988
b4ed4da0
PH
989Exim version 4.67
990-----------------
991
22ad45c9
MH
992MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
993 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
994 Jan Srzednicki.
995
b4ed4da0
PH
996PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
997 issue a MAIL command.
998
431b7361
PH
999PH/02 In an ACL statement such as
1000
1001 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1002
1003 if a client was not listed at all, or was listed with a value other than
1004 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1005 the condition was not true (as it should be), so access was not denied.
1006 The bug was that the ! inversion was incorrectly passed on to the second
1007 item. This has been fixed.
1008
1009PH/03 Added additional dnslists conditions == and =& which are different from
1010 = and & when the dns lookup returns more than one IP address.
1011
83da1223
PH
1012PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1013 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1014
54fc8428
PH
1015PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1016 FSYNC, which compiles an option called disable_fsync that allows for
1017 bypassing fsync(). The documentation is heavily laced with warnings.
1018
34c5e8dd
SC
1019SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1020
bbe15da8
PH
1021PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1022 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1023 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1024 including adding "make clean"; (3) Added -fPIC when compiling the test
1025 dynamically loaded module, to get rid of a warning.
1026
0e8a9471
MH
1027MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1028 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1029 = 0s. The bug is otherwise harmless.
1030
f0872424
PH
1031PH/07 There was a bug in the dovecot authenticator such that the value of
1032 $auth1 could be overwritten, and so not correctly preserved, after a
1033 successful authentication. This usually meant that the value preserved by
1034 the server_setid option was incorrect.
1035
b01dd148
PH
1036PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1037
6bf342e1
PH
1038PH/09 Installed PCRE release 7.0.
1039
273f34d0
PH
1040PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1041 run for batched SMTP input. It is now run at the start of every message
1042 in the batch. While fixing this I discovered that the process information
1043 (output by running exiwhat) was not always getting set for -bs and -bS
1044 input. This is fixed, and it now also says "batched" for BSMTP.
1045
cf8b11a5
PH
1046PH/11 Added control=no_pipelining.
1047
41c7c167
PH
1048PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1049 patch, slightly modified), and move the expansion of helo_data till after
1050 the connection is made in the smtp transport (so it can use these
1051 values).
1052
9c57cbc0
PH
1053PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1054
f3f065bb
PH
1055PH/14 Added log_selector = +pid.
1056
047bdd8c
PH
1057PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1058
0ce9abe6
PH
1059PH/16 Add ${if forany and ${if forall.
1060
0e22dfd1
PH
1061PH/17 Added dsn_from option to vary the From: line in DSNs.
1062
4c590bd1
PH
1063PH/18 Flush SMTP output before performing a callout, unless control =
1064 no_callout_flush is set.
1065
09945f1e
PH
1066PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1067 was true (the default) a successful delivery failed to delete the retry
1068 item, thus causing premature timeout of the address. The bug is now
1069 fixed.
1070
c51b8e75
PH
1071PH/20 Added hosts_avoid_pipelining to the smtp transport.
1072
e28326d8
PH
1073PH/21 Long custom messages for fakedefer and fakereject are now split up
1074 into multiline reponses in the same way that messages for "deny" and
1075 other ACL rejections are.
1076
75b1493f
PH
1077PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1078 with slight modification.
1079
7c5214ec
PH
1080PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1081 draft, changing the syntax and factoring some duplicate code".
1082
4311097e
PH
1083PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1084 for deliveries of the second and subsequent messages over the same SMTP
1085 connection.
1086
29f89cad
PH
1087PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1088 ${reduce, with only minor "tidies".
1089
5e687460
SC
1090SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1091
c3611384
PH
1092PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1093 expansion side effects.
1094
5a11a7b4
PH
1095PH/27 When a message times out after an over-quota error from an Exim-imposed
1096 quota, the bounce message says "mailbox is full". This message was not
1097 being given when it was a system quota that was exceeded. It now should
1098 be the same.
1099
0e20aff9
MH
1100MH/03 Made $recipients available in local_scan(). local_scan() already has
1101 better access to the recipient list through recipients_list[], but
1102 $recipients can be useful in postmaster-provided expansion strings.
1103
ca86f471
PH
1104PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1105 in the case of a MAIL command with additional options following the
1106 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1107 were accidentally chopped off.
1108
a14e5636
PH
1109PH/29 SMTP synchronization checks are implemented when a command is read -
1110 there is a check that no more input is waiting when there shouldn't be
1111 any. However, for some commands, a delay in an ACL can mean that it is
1112 some time before the response is written. In this time, more input might
1113 arrive, invalidly. So now there are extra checks after an ACL has run for
1114 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1115 pipelining has not been advertised.
1116
ec95d1a6
PH
1117PH/30 MH's patch to allow iscntrl() characters to be list separators.
1118
42855d71
PH
1119PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1120 returned in the SMTP response when smtp_return_error_details was false.
1121 This has been fixed.
1122
57c2c631
PH
1123PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1124 instead of the C I/O that was originally supplied, because problems were
1125 reported on Solaris.
1126
58c01c94
PH
1127PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1128 Exim which did not show up earlier: it was assuming that a call to
1129 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1130 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1131 was a macro that became an assignment, so it seemed to work. This has
1132 changed to a proper function call with a void return, hence the compile
1133 error. Exim's code has been fixed.
1134
dee5a20a
PH
1135PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1136 cpus.
1137
d2ee6114
PH
1138PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1139
b2d5182b
PH
1140PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1141
79749a79
PH
1142PH/37 If a message is not accepted after it has had an id assigned (e.g.
1143 because it turns out to be too big or there is a timeout) there is no
3ce62588
PH
1144 "Completed" line in the log. When some messages of this type were
1145 selected by exigrep, they were listed as "not completed". Others were
1146 picked up by some special patterns. I have improved the selection
1147 criteria to be more general.
79749a79 1148
c456d9bb
PH
1149PH/38 The host_find_failed option in the manualroute router can now be set
1150 to "ignore", to completely ignore a host whose IP address cannot be
1151 found. If all hosts are ignored, the behaviour is controlled by the new
1152 host_all_ignored option.
1153
cd9868ec
PH
1154PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1155 homing or because of multiple MX records with /mx) generated more than
1156 one IP address, and the following item turned out to be the local host,
1157 all the secondary addresses of the first item were incorrectly removed
1158 from the list, along with the local host and any following hosts (which
1159 is what is supposed to happen).
1160
ebeaf996
PH
1161PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1162 whoever called Exim into the -H file. In the case of the daemon it was
1163 behaving confusingly. When first started, it used values for whoever
1164 started the daemon, but after a SIGHUP it used the Exim user (because it
1165 calls itself on a restart). I have changed the code so that it now always
1166 uses the Exim user.
1167
2679d413
PH
1168PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1169 message are rejected with the same error (e.g. no authentication or bad
1170 sender address), and a DATA command is nevertheless sent (as can happen
1171 with PIPELINING or a stupid MUA), the error message that was given to the
1172 RCPT commands is included in the rejection of the DATA command. This is
1173 intended to be helpful for MUAs that show only the final error to their
1174 users.
1175
84024b72
PH
1176PH/42 Another patch from the Sieve maintainer.
1177
8005d38e
SC
1178SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1179 Eximstats - Fixed some broken HTML links and added missing column headers
1180 (Jez Hancock).
1181 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1182 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1183
3298c6c6
SC
1184SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1185
a43a27c5
PH
1186PH/43 Yet another patch from the Sieve maintainer.
1187
58eb016e 1188PH/44 I found a way to check for a TCP/IP connection going away before sending
563b63fa
PH
1189 the response to the final '.' that terminates a message, but only in the
1190 case where the client has not sent further data following the '.'
1191 (unfortunately, this is allowed). However, in many cases there won't be
1192 any further data because there won't be any more messages to send. A call
1193 to select() can be used: if it shows that the input is "ready", there is
1194 either input waiting, or the socket has been closed. An attempt to read
1195 the next input character can distinguish the two cases. Previously, Exim
58eb016e 1196 would have sent an OK response which the client would never have see.
563b63fa
PH
1197 This could lead to message repetition. This fix should cure that, at
1198 least in a lot of common cases.
58eb016e 1199
b43a74ea
PH
1200PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1201 advertised in response to EHLO.
1202
b4ed4da0 1203
5dd1517f
PH
1204Exim version 4.66
1205-----------------
1206
1207PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1208 fixed by 4.65/MH/01 (is this a record?) are fixed:
1209
1210 (i) An empty string was always treated as zero by the numeric comparison
1211 operators. This behaviour has been restored.
1212
1213 (ii) It is documented that the numeric comparison operators always treat
1214 their arguments as decimal numbers. This was broken in that numbers
1215 starting with 0 were being interpreted as octal.
1216
1217 While fixing these problems I realized that there was another issue that
1218 hadn't been noticed. Values of message_size_limit (both the global option
1219 and the transport option) were treated as octal if they started with 0.
1220 The documentation was vague. These values are now always treated as
1221 decimal, and I will make that clear in the documentation.
1222
1223
93cfa765
TK
1224Exim version 4.65
1225-----------------
1226
1227TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1228 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1229 versions. (#438)
1230
d6066548
MH
1231MH/01 Don't check that the operands of numeric comparison operators are
1232 integers when their expansion is in "skipping" mode (fixes bug
1233 introduced by 4.64-PH/07).
1234
4362ff0d
PH
1235PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1236 child addresses, Exim now panics and dies. Previously, because the count
1237 is held in a short int, deliveries were likely to be lost. As such a
1238 large number of recipients for a single message is ridiculous
1239 (performance will be very, very poor), I have chosen to impose a limit
1240 rather than extend the field.
1241
93cfa765 1242
944e9e9c
TF
1243Exim version 4.64
1244-----------------
aa41d2de 1245
21d74bd9
TK
1246TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1247 leftover -K file (the existence of which was triggered by #402).
1248 While we were at it, introduced process PID as part of the -K
1249 filename. This should rule out race conditions when creating
1250 these files.
1251
1252TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1253 processing considerably. Previous code took too long for large mails,
1254 triggering a timeout which in turn triggers #401.
1255
1256TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1257 in the DK code in transports.c. sendfile() is not really portable,
1258 hence the _LINUX specificness.
944e9e9c
TF
1259
1260TF/01 In the add_headers option to the mail command in an Exim filter,
1261 there was a bug that Exim would claim a syntax error in any
1262 header after the first one which had an odd number of characters
1263 in the field name.
1264
2b1c6e3a
PH
1265PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1266 callout verification, Exim cached a "reject" for the entire domain. This
1267 is correct for most verifications, but it is not correct for a recipient
1268 verification with use_sender or use_postmaster set, because in that case
1269 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1270 case of MAIL FROM:<> rejection from other early rejections (e.g.
1271 rejection of HELO). When verifying a recipient using a non-null MAIL
1272 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1273 Whatever the result of the callout, the value of the domain cache is
1274 left unchanged (for any other kind of callout, getting as far as trying
1275 RCPT means that the domain itself is ok).
1276
1f872c80
PH
1277PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1278 gcc 4.1.1 threw up.
1279
1280PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1281 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1282 session, the daemon ignores ECONNECT errors and logs others; it now
1283 ignores EPIPE as well.
1284
d203e649
PH
1285PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1286 (quoted-printable decoding).
1287
cc2ed8f7 1288PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
21a04aa3 1289 later the small subsequent patch to fix an introduced bug.
f951fd57 1290
ddfcd446
PH
1291PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1292
d45b1de8
PH
1293PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1294
1295PH/08 An error is now given if message_size_limit is specified negative.
1296
38a0a95f 1297PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
641cb756 1298 to be given (somewhat) arbitrary names.
38a0a95f 1299
a2405d83
JJ
1300JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1301 in 4.64-PH/09.
1302
1303JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1304 miscellaneous code fixes
1305
6ea85e9a
PH
1306PH/10 Added the log_reject_target ACL modifier to specify where to log
1307 rejections.
1308
26da7e20
PH
1309PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1310 hostname. This is wrong, because it relates to the incoming message (and
1311 probably the interface on which it is arriving) and not to the outgoing
1312 callout (which could be using a different interface). This has been
1313 changed to use the value of the helo_data option from the smtp transport
1314 instead - this is what is used when a message is actually being sent. If
1315 there is no remote transport (possible with a router that sets up host
1316 addresses), $smtp_active_hostname is used.
6ea85e9a 1317
14aa5a05 1318PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
7befa435 1319 tweaks were necessary in order to get it to work (see also 21 below):
14aa5a05
PH
1320 (a) The code assumed that strncpy() returns a negative number on buffer
1321 overflow, which isn't the case. Replaced with Exim's string_format()
1322 function.
1323 (b) There were several signed/unsigned issues. I just did the minimum
1324 hacking in of casts. There is scope for a larger refactoring.
1325 (c) The code used strcasecmp() which is not a standard C function.
1326 Replaced with Exim's strcmpic() function.
1327 (d) The code set only $1; it now sets $auth1 as well.
1328 (e) A simple test gave the error "authentication client didn't specify
1329 service in request". It would seem that Dovecot has changed its
1330 interface. Fortunately there's a specification; I followed it and
1331 changed what the client sends and it appears to be working now.
1332
ff75a1f7
PH
1333PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1334 decoding.
1335
e6f6568e
PH
1336PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1337 address A is aliased to B and C, where B exists and C does not. Without
1338 -v the output is "A verified" because verification stops after a
1339 successful redirection if more than one address is generated. However,
1340 with -v the child addresses are also verified. Exim was outputting "A
1341 failed to verify" and then showing the successful verification for C,
1342 with its parentage. It now outputs "B failed to verify", showing B's
1343 parentage before showing the successful verification of C.
1344
d6f6e0dc
PH
1345PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1346 look up a TXT record in a specific list after matching in a combined
1347 list.
1348
322050c2
PH
1349PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1350 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1351 they consult the DNS. I had assumed they would set it the way they
1352 wanted; and indeed my experiments on Linux seem to show that in some
1353 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1354 To be on the safe side, however, I have now made the interface to
1355 host_find_byname() similar to host_find_bydns(), with an argument
1356 containing the DNS resolver options. The host_find_byname() function now
1357 sets these options at its start, just as host_find_bydns() does. The smtp
1358 transport options dns_qualify_single and dns_search_parents are passed to
1359 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1360 of host_find_byname() use the default settings of RES_DEFNAMES
1361 (qualify_single) but not RES_DNSRCH (search_parents).
1362
08955dd3
PH
1363PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1364 spool_read_header() do less string testing, by means of a preliminary
1365 switch on the second character of optional "-foo" lines. (This is
1366 overdue, caused by the large number of possibilities that now exist.
1367 Originally there were few.) While I was there, I also converted the
1368 str(n)cmp tests so they don't re-test the leading "-" and the first
1369 character, in the hope this might squeeze out yet more improvement.
1370
1eccaa59
PH
1371PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1372 flag allowing group syntax was set by the header_syntax check but not
1373 turned off, possible causing trouble later; (2) The flag was not being
1374 set at all for the header_verify test, causing "group"-style headers to
1375 be rejected. I have now set it in this case, and also caused header_
1376 verify to ignore an empty address taken from a group. While doing this, I
1377 came across some other cases where the code for allowing group syntax
1378 while scanning a header line wasn't quite right (mostly, not resetting
1379 the flag correctly in the right place). These bugs could have caused
1380 trouble for malformed header lines. I hope it is now all correct.
1381
602e59e5
PH
1382PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1383 with the "reply" argument non-NULL. The code, however (which originally
1384 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1385 but it didn't always do it. This confused somebody who was copying the
1386 code for some other use. I have removed all the tests.
1387
411ef850
PH
1388PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1389 feature that was used to support insecure browsers during the U.S. crypto
1390 embargo. It requires special client support, and Exim is probably the
1391 only MTA that supported it -- and would never use it because real RSA is
1392 always available. This code has been removed, because it had the bad
1393 effect of slowing Exim down by computing (never used) parameters for the
1394 RSA_EXPORT functionality.
1395
7befa435
PH
1396PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1397 authenticator to fail if there's a tab character in the incoming data
1398 (there should never be unless someone is messing about, as it's supposed
1399 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1400 if the connection is using TLS or if the remote IP is the same as the
1401 local IP, and the "valid-client-cert option" if a client certificate has
1402 been verified.
1403
48da4259 1404PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
16ff981e
PH
1405 authenticators. This can be used for authorization after authentication
1406 succeeds. (In the case of plaintext, it servers for both authentication
1407 and authorization.)
1408
48da4259
PH
1409PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1410 if any retry times were supplied.
1411
d1d5595c
PH
1412PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1413 connection, where there is no client IP address to check. In this
1414 situation, the verify now always succeeds.
1415
0ef732d9
PH
1416PH/25 Applied John Jetmore's -Mset patch.
1417
328895cc
PH
1418PH/26 Added -bem to be like -Mset, but loading a message from a file.
1419
fd700877
PH
1420PH/27 In a string expansion for a processed (not raw) header when multiple
1421 headers of the same name were present, leading whitespace was being
1422 removed from all of them, but trailing whitespace was being removed only
1423 from the last one. Now trailing whitespace is removed from each header
f6c332bd
PH
1424 before concatenation. Completely empty headers in a concatenation (as
1425 before) are ignored.
fd700877 1426
8dce1a6f
PH
1427PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1428 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1429
17af4a17
PH
1430PH/29 [Removed. This was a change that I later backed out, and forgot to
1431 correct the ChangeLog entry (that I had efficiently created) before
1432 committing the later change.]
f6c332bd
PH
1433
1434PH/30 Exim was sometimes attempting to deliver messages that had suffered
1435 address errors (4xx response to RCPT) over the same connection as other
1436 messages routed to the same hosts. Such deliveries are always "forced",
1437 so retry times are not inspected. This resulted in far too many retries
1438 for the affected addresses. The effect occurred only when there were more
1439 hosts than the hosts_max_try setting in the smtp transport when it had
1440 the 4xx errors. Those hosts that it had tried were not added to the list
1441 of hosts for which the message was waiting, so if all were tried, there
1442 was no problem. Two fixes have been applied:
1443
1444 (i) If there are any address or message errors in an SMTP delivery, none
1445 of the hosts (tried or untried) are now added to the list of hosts
1446 for which the message is waiting, so the message should not be a
1447 candidate for sending over the same connection that was used for a
1448 successful delivery of some other message. This seems entirely
1449 reasonable: after all the message is NOT "waiting for some host".
1450 This is so "obvious" that I'm not sure why it wasn't done
1451 previously. Hope I haven't missed anything, but it can't do any
1452 harm, as the worst effect is to miss an optimization.
1453
1454 (ii) If, despite (i), such a delivery is accidentally attempted, the
1455 routing retry time is respected, so at least it doesn't keep
1456 hammering the server.
1457
c1114884
PH
1458PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1459 in ${readsocket because some servers need this prod.
1460
7a0743eb
PH
1461PH/32 Added some extra debug output when updating a wait-xxx database.
1462
0d85fa3f
PH
1463PH/33 The hint "could be header name not terminated by colon", which has been
1464 given for certain expansion errors for a long time, was not being given
1465 for the ${if def:h_colon_omitted{... case.
1466
1bf43b78
PH
1467PH/34 The spec says: "With one important exception, whenever a domain list is
1468 being scanned, $domain contains the subject domain." There was at least
1469 one case where this was not true.
1470
520de300
PH
1471PH/35 The error "getsockname() failed: connection reset by peer" was being
1472 written to the panic log as well as the main log, but it isn't really
1473 panic-worthy as it just means the connection died rather early on. I have
1474 removed the panic log writing for the ECONNRESET error when getsockname()
1475 fails.
1476
48c7f9e2
PH
1477PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1478 runs only) independently of the message's sender address. This meant
1479 that, if the 4xx error was in fact related to the sender, a different
1480 message to the same recipient with a different sender could confuse
1481 things. In particualar, this can happen when sending to a greylisting
1482 server, but other circumstances could also provoke similar problems.
1483 I have changed the default so that the retry time for these errors is now
1484 based a combination of the sender and recipient addresses. This change
1485 can be overridden by setting address_retry_include_sender=false in the
1486 smtp transport.
1487
99ea1c86
PH
1488PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1489 remote server are returned as part of bounce messages. This was not
1490 happening for LMTP over a pipe (the lmtp transport), but now it is the
1491 same for both kinds of LMTP.
1492
a2042e78
PH
1493PH/38 Despite being documented as not happening, Exim was rewriting addresses
1494 in header lines that were in fact CNAMEs. This is no longer the case.
1495
4fbcfc2e
PH
1496PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1497 and queue runs started by the daemon processed all messages. This has
1498 been fixed so that -R and -S can now usefully be given with -q<time>.
1499
aa41d2de
PH
1500PH/40 Import PCRE release 6.7 (fixes some bugs).
1501
af561417
PH
1502PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1503
3cc66b45
PH
1504PH/42 Give an error if -q is specified more than once.
1505
194cc0e4
PH
1506PH/43 Renamed the variables $interface_address and $interface_port as
1507 $received_ip_address and $received_port, to make it clear that these
1508 values apply to message reception, and not to the outgoing interface when
1509 a message is delivered. (The old names remain recognized, of course.)
1510
a401ddaa
PH
1511PH/44 There was no timeout on the connect() call when using a Unix domain
1512 socket in the ${readsocket expansion. There now is.
1513
4e88a19f
PH
1514PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1515 be meaningful with "accept".
1516
d7d7b289
SC
1517SC/01 Eximstats V1.43
1518 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1519
1520SC/02 Eximstats V1.44
1521 Use a glob alias rather than an array ref in the generated
1522 parser. This improves both readability and performance.
1523
1524SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1525 Collect SpamAssassin and rejection statistics.
1526 Don't display local sender or destination tables unless
1527 there is data to show.
1528 Added average volumes into the top table text output.
1529
1530SC/04 Eximstats V1.46
1531 Collect data on the number of addresses (recipients)
1532 as well as the number of messages.
1533
1534SC/05 Eximstats V1.47
1535 Added 'Message too big' to the list of mail rejection
1536 reasons (thanks to Marco Gaiarin).
1537
1538SC/06 Eximstats V1.48
1539 Mainlog lines which have GMT offsets and are too short to
1540 have a flag are now skipped.
1541
1542SC/07 Eximstats V1.49 (Alain Williams)
1543 Added the -emptyok flag.
1544
1545SC/08 Eximstats V1.50
1546 Fixes for obtaining the IP address from reject messages.
1547
0ea2a468
JJ
1548JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1549 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1550 whitesspace changes from 4.64-PH/27
1551
1552JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1553 match 4.64-PH/13
1554
1555JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1556 are found, allow negative numbers in numeric criteria)
1557
1558JJ/06 exipick.20061117.2, added new $message_body_missing variable
1559
1560JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1561 to match changes made in 4.64-PH/43
1562
8a10f5a4
PH
1563PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1564
30e18802
PH
1565PH/47 Put in an explicit test for a DNS lookup of an address record where the
1566 "domain" is actually an IP address, and force a failure. This locks out
1567 those revolvers/nameservers that support "A-for-A" lookups, in
1568 contravention of the specifications.
1569
55728a4f
PH
1570PH/48 When a host name was looked up from an IP address, and the subsequent
1571 forward lookup of the name timed out, the host name was left in
1572 $sender_host_name, contrary to the specification.
d7d7b289 1573
d7837193
PH
1574PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1575 restricted to single-key lookups, Exim was not diagnosing an error if
1576 * or *@ was used with a query-style lookup.
1577
87054a31
PH
1578PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1579
ea2c01d2
MH
1580MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1581 long ago, but noone interested enough thought of it. Let's just say that
1582 the "1.1" means that there are some new functions that weren't there at
1583 some point in the past.
1584
e4fa6968
PH
1585PH/51 Error processing for expansion failure of helo_data from an smtp
1586 transport during callout processing was broken.
1587
56f5d9bd
PH
1588PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1589 tested/used via the -bh/-bhc/-bs options.
1590
922e1c28
PH
1591PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1592 bug, fixed in subsequent PCRE releases).
1593
21eb6e72
PH
1594PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1595 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1596
a0540757
PH
1597PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1598
944e9e9c 1599
478be7b0
SC
1600Exim version 4.63
1601-----------------
1602
1603SC/01 Use a glob alias rather than an array ref in eximstats generated
1604 parser. This improves both readability and performance.
1605
1606SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1607 Don't display local sender or destination tables in eximstats unless
1608 there is data to show.
1609 Added average volumes into the eximstats top table text output.
1610
1611SC/03 Collect data on the number of addresses (recipients) as well
1612 as the number of messages in eximstats.
1613
2b965a65
TF
1614TF/01 Correct an error in the documentation for the redirect router. Exim
1615 does (usually) call initgroups() when daemonizing.
478be7b0 1616
45b91596
PH
1617TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1618 with consistent privilege compared to when running as a daemon.
478be7b0 1619
c59f5781
TF
1620TF/03 Note in the spec that $authenticated_id is not set for local
1621 submissions from trusted users.
1622
90fc3069
TF
1623TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1624 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1625
6083aca0
TF
1626TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1627 by adding some example configuration directives to the default
1628 configuration file. A little bit of work is required to uncomment the
1629 directives and define how usernames and passwords are checked, but
1630 there is now a framework to start from.
1631
765b530f
PH
1632PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1633 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1634 without this. I don't know how relevant this is to other LDAP libraries.
1635
4e167a8c
PH
1636PH/02 Add the verb name to the "unknown ACL verb" error.
1637
4608d683
PH
1638PH/03 Magnus Holmgren's patch for filter_prepend_home.
1639
b8dc3e4a
PH
1640PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1641
5418e93b
PH
1642PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1643 directory not expanded when it should be if an expanded home directory
1644 was set for the address (which is overridden by the transport).
1645
b4a9bda2
PH
1646PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1647 libradius.
1648
45b91596
PH
1649PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1650 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1651 because it is too late at that time, and has no effect.
1652
5547e2c5
PH
1653PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1654 security issue with \' (bugzilla #107). I could not use the
1655 PQescapeStringConn() function, because it needs a PGconn value as one of
1656 its arguments.
1657
dbcef0ea
PH
1658PH/08 When testing addresses using -bt, indicate those final addresses that
1659 are duplicates that would not cause an additional delivery. At least one
1660 person was confused, thinking that -bt output corresponded to deliveries.
1661 (Suppressing duplicates isn't a good idea as you lose the information
1662 about possibly different redirections that led to the duplicates.)
1663
25257489
PH
1664PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1665 systems where poll() doesn't work, in particular OS X.
1666
c816d124
PH
1667PH/10 Added more information to debugging output for retry time not reached.
1668
a9ccd69a
PH
1669PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1670 operations in malware.c.
1671
75fa1910
PH
1672PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1673 signatures.
1674
a7d7aa58
PH
1675PH/13 If write_rejectlog was set false when logging was sent to syslog with
1676 syslog_duplication set false, log lines that would normally be written
1677 both the the main log and to the reject log were not written to syslog at
1678 all.
1679
42119b09
PH
1680PH/14 In the default configuration, change the use of "message" in ACL warn
1681 statements to "add_header".
1682
41609df5
PH
1683PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1684 not followed by a command (e.g. "seen endif").
1685
a5bd321b
PH
1686PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1687 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1688 latter.
1689
e85a7ad5 1690PH/17 Added extra conditions to the default value of delay_warning_condition
5dff5817
PH
1691 so that it is now:
1692
e85a7ad5
PH
1693 ${if or { \
1694 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1695 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1696 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
5dff5817
PH
1697 }{no}{yes}}
1698
e85a7ad5
PH
1699 The Auto-Submitted: and various List- headers are standardised, whereas I
1700 don't think Precedence: ever was.
5dff5817 1701
d8fe1c03
PH
1702PH/18 Refactored debugging code in route_finduser() to show more information,
1703 in particular, the error code if getpwnam() issues one.
1704
16282d2b
PH
1705PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1706 This is apparently needed in addition to the PH/07 change above to avoid
1707 any possible encoding problems.
1708
35d40a98
PH
1709PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1710 but not after initializing Perl.
1711
034d99ab
PH
1712PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1713 output them only if debugging. By default they are written stderr,
1714 apparently, which is not desirable.
1715
6ec97b1b
PH
1716PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1717 queries.
1718
e22ca4ac
JJ
1719JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1720 --not options
1721
1722JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1723
33d73e3b
PH
1724PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1725 authenticated or an ident call has been made. Suppress the default
1726 values for $authenticated_id and $authenticated_sender (but permit -oMai
1727 and -oMas) when testing with -bh.
1728
9ecb03f3
PH
1729PH/24 Re-jigged the order of the tests in the default configuration so that the
1730 tests for valid domains and recipients precede the DNS black list and CSA
1731 tests, on the grounds that those ones are more expensive.
1732
084efe8d
PH
1733PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1734 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1735 This bug exists in every version of Exim that I still have, right back to
1736 0.12.
1737
366fc9f0
PH
1738PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1739 However, an attempt to turn on case-sensitivity in a regex key by
1740 including (?-i) didn't work because the subject string was already
1741 lowercased, and the effects were non-intuitive. It turns out that a
1742 one-line patch can be used to allow (?-i) to work as expected.
1743
c59f5781 1744
c887c79e
TF
1745Exim version 4.62
1746-----------------
1747
1748TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1749 other effects) broke the use of negated acl sub-conditions.
1750
1cce3af8
PH
1751PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1752 patch).
1753
afb3eaaf
PH
1754PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1755 "Deny" causes Exim to reject the incoming connection with a 554 error.
1756 Unfortunately, if there is a major crisis, such as a disk failure,
1757 tcp-wrappers gives "deny", whereas what one would like would be some
1758 kind of temporary error. A kludge has been added to help with this.
1759 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1760 554 error is used if errno is still zero or contains ENOENT (which occurs
1761 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1762 451 error is used.
1763
e173618b
PH
1764PH/03 Add -lutil to the default FreeBSD LIBS setting.
1765
dd16e114
PH
1766PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1767 errors. Otherwise a message that provokes a temporary error (when other
1768 messages do not) can cause a whole host to time out.
1769
f7fd3850
PH
1770PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1771 addresses were routed directly to files or pipes from a redirect router.
1772 File deliveries just didn't batch; pipe deliveries might have suffered
1773 odd errors.
1774
d87df92c
PH
1775PH/06 A failure to get a lock for a hints database would erroneously always say
1776 "Failed to get write lock", even when it was really a read lock.
1777
7e9f683d
PH
1778PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1779 of 0600. This has been changed to use the value of the lockfile_mode
1780 option (which defaults to 0600).
1781
bfad5236
PH
1782PH/08 Applied small patch from the Sieve maintainer.
1783
01c490df
PH
1784PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1785 folder from quota calculations, a direct delivery into this folder messed
1786 up the contents of the maildirsize file. This was because the regex was
1787 used only to exclude .Trash (or whatever) when the size of the mailbox
1788 was calculated. There was no check that a delivery was happening into an
1789 excluded directory. This bug has been fixed by ignoring all quota
1790 processing for deliveries into excluded directories.
1791
d6629cdc
PH
1792PH/10 Added the maildirfolder_create_regex option to appendfile.
1793
1cce3af8 1794
214e2000
PH
1795Exim version 4.61
1796-----------------
1797
1798PH/01 The code for finding all the local interface addresses on a FreeBSD
1799 system running IPv6 was broken. This may well have applied to all BSD
1800 systems, as well as to others that have similar system calls. The broken
1801 code found IPv4 interfaces correctly, but gave incorrect values for the
1802 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1803 that it would not match correctly against @[] and not recognize the IPv6
1804 addresses as local.
1805
f9daeae0
PH
1806PH/02 The ipliteral router was not recognizing addresses of the form user@
1807 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1808
7e66e54d
PH
1809PH/03 Added disable_ipv6.
1810
c8ea1597
PH
1811PH/04 Changed $reply_address to use the raw form of the headers instead of the
1812 decoded form, because it is most often used to construct To: headers
1813 lines in autoreplies, and the decoded form may well be syntactically
1814 invalid. However, $reply_address has leading white space removed, and all
1815 newlines turned into spaces so that the autoreply transport does not
1816 grumble.
1817
911f6fde
PH
1818PH/05 If group was specified without a user on a router, and no group or user
1819 was specified on a transport, the group from the router was ignored.
1820
47ca6d6c
PH
1821PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1822 for visible compile-time settings that can be used to change these
1823 numbers, for those that want even more. Backwards compatibility with old
1824 spool files has been maintained. However, going back to a previous Exim
1825 release will lost any variables that are in spool files.
1826
ed0e9820
PH
1827PH/07 Two small changes when running in the test harness: increase delay when
1828 passing a TCP/IP connection to a new process, in case the original
1829 process has to generate a bounce, and remove special handling of
1830 127.0.0.2 (sic), which is no longer necessary.
1831
eff37e47
PH
1832PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1833 be the same on different OS.
1834
1921d2ea
PH
1835PH/09 Moved a debug statement in filter processing to avoid a race problem when
1836 testing.
1837
b3f69ca8
JJ
1838JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1839 whether --show-vars was specified or not
1840
1841JJ/02 exipick: Added support for new ACL variable spool format introduced
1842 in 4.61-PH/06
1843
424a1c63
PH
1844PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1845 syntactically invalid From: or Reply-to: line, and a filter used this to
1846 generate an autoreply, and therefore failed to obtain an address for the
1847 autoreply, Exim could try to deliver to a non-existent relative file
1848 name, causing unrelated and misleading errors. What now happens is that
1849 it logs this as a hard delivery error, but does not attempt to create a
1850 bounce message.
1851
7a100415
PH
1852PH/11 The exinext utility has a -C option for testing purposes, but although
1853 the given file was scanned by exinext itself; it wasn't being passed on
1854 when Exim was called.
1855
19b9dc85
PH
1856PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1857 an end-of-file indication when reading a command response.
1858
309bd837
PH
1859PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1860 compiled. In many other places in Exim, IPv6 addresses are always
1861 recognized, so I have changed this. It also means that IPv4 domain
1862 literals of the form [IPV4:n.n.n.n] are now always recognized.
1863
59e82a2a
PH
1864PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1865 used if the router is not running as root, for example, when verifying at
1866 ACL time, or when using -bh. The debugging output from this situation was
1867 non-existent - all you got was a failure to exec. I have made two
1868 changes:
1869
1870 (a) Failures to set uid/gid, the current directory, or a process leader
1871 in a subprocess such as that created by queryprogram now generate
1872 suitable debugging ouput when -d is set.
1873
1874 (b) The queryprogram router detects when it is not running as root,
1875 outputs suitable debugging information if -d is set, and then runs
1876 the subprocess without attempting to change uid/gid.
1877
9edc04ce
PH
1878PH/15 Minor change to Makefile for building test_host (undocumented testing
1879 feature).
1880
1349e1e5
PH
1881PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1882 additional section of a DNS packet that returns MX or SRV records.
1883 Instead, it always explicitly searches for A/AAAA records. This avoids
1884 major problems that occur when a DNS server includes only records of one
1885 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1886 fixed another bug: if SRV records were looked up and the corresponding
1887 address records were *not* found in the additional section, the port
1888 values from the SRV records were lost.
1889
ea49d0e1
PH
1890PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1891 using the correct key (the original address) when searching the retry
1892 rules in order to find which one to use for generating the retry hint.
1893
064a94c9
PH
1894PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1895 adding the default one. Similarly, if it contains a Reply-To: header, the
1896 errors_reply_to option, if set, is not used.
1897
727071f8
PH
1898PH/19 When calculating a retry time, Exim used to measure the "time since
1899 failure" by looking at the "first failed" field in the retry record. Now
1900 it does not use this if it is later than than the arrival time of the
1901 message. Instead it uses the arrival time. This makes for better
1902 behaviour in cases where some deliveries succeed, thus re-setting the
1903 "first failed" field. An example is a quota failure for a huge message
1904 when small messages continue to be delivered. Without this change, the
1905 "time since failure" will always be short, possible causing more frequent
1906 delivery attempts for the huge message than are intended.
dd16e114 1907 [Note: This change was subsequently modified - see PH/04 for 4.62.]
727071f8 1908
f78eb7c6
PH
1909PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1910 $1, $2, $3) because the numerical variables can be reset during some
1911 expansion items (e.g. "match"), thereby losing the authentication data.
1912
21c28500
PH
1913PH/21 Make -bV show the size of off_t variables so that the test suite can
1914 decide whether to run tests for quotas > 2G.
1915
1916PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1917 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1918 filecount value is greater than 2G or if a quota value is greater than 2G
1919 on a system where the size of off_t is not greater than 4, a panic error
1920 is given.
1921
1688f43b
PH
1922PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1923 never match. The debug and -bh output now contains an explicit error
1924 message indicating a malformed IPv4 address or mask.
1925
1926PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1927 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1928 PH/23 above applies.
1929
9675b384
PH
1930PH/25 Do not write to syslog when running in the test harness. The only
1931 occasion when this arises is a failure to open the main or panic logs
1932 (for which there is an explicit test).
1933
6a3f1455
PH
1934PH/26 Added the /no_tell option to "control=freeze".
1935
dac79d3e
PH
1936PH/27 If a host name lookup failed very early in a connection, for example, if
1937 the IP address matched host_lookup and the reverse lookup yielded a name
1938 that did not have a forward lookup, an error message of the form "no IP
1939 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1940 could be logged. Now it outputs the IP address instead of "NULL".
1349e1e5 1941
5977a0b3
PH
1942PH/28 An enabling patch from MH: add new function child_open_exim2() which
1943 allows the sender and the authenticated sender to be set when
1944 submitting a message from within Exim. Since child_open_exim() is
1945 documented for local_scan(), the new function should be too.
1946
c91535f3
PH
1947PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1948 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1949 results in an empty string is now treated as unset.
1950
0d46a8c8
PH
1951PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1952
278c6e6c
PH
1953PH/31 Added sender_verify_fail logging option.
1954
2cbb4081
PH
1955PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1956 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1957 tidied the source and removed it altogether.
1958
3eef829e
PH
1959PH/33 When a queue run was abandoned because the load average was too high, a
1960 log line was always written; now it is written only if the queue_run log
1961 selector is set. In addition, the log line for abandonment now contains
1962 information about the queue run such as the pid. This is always present
1963 in "start" and "stop" lines but was omitted from the "abandon" line.
1964
1ab95fa6
PH
1965PH/34 Omit spaces between a header name and the colon in the error message that
1966 is given when verify = headers_syntax fails (if there are lots of them,
1967 the message gets confusing).
1968
230205fc
PH
1969PH/35 Change the default for dns_check_names_pattern to allow slashes within
1970 names, as there are now some PTR records that contain slashes. This check
1971 is only to protect against broken name servers that fall over on strange
1972 characters, so the fact that it applies to all lookups doesn't matter.
1973
75e0e026
PH
1974PH/36 Now that the new test suite is complete, we can remove some of the
1975 special code in Exim that was needed for the old test suite. For example,
1976 sorting DNS records because real resolvers return them in an arbitrary
1977 order. The new test suite's fake resolver always returns records in the
1978 same order.
1979
1980PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1981 bounces) except when queue_only is set, to avoid logging races between
1982 the different processes.
1983
145396a6
PH
1984PH/38 Panic-die if .include specifies a non-absolute path.
1985
3cd34f13
PH
1986PH/39 A tweak to the "H" retry rule from its user.
1987
11121d3d
JJ
1988JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1989 a label. They prevented compilation on older perls.
1990
1991JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1992 a warning to be raised on newish perls.
1993
1994JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1995 on queue. Changes to match documented behaviour of showing count of
1996 messages matching specified criteria.
1997
8def5aaf
PH
1998PH/40 Changed the default ident timeout from 30s to 5s.
1999
929ba01c
PH
2000PH/41 Added support for the use of login_cap features, on those BSD systems
2001 that have them, for controlling the resources used by pipe deliveries.
2002
2632889e
PH
2003PH/42 The content-scanning code uses fopen() to create files in which to put
2004 message data. Previously it was not paying any attention to the mode of
2005 the files. Exim runs with umask(0) because the rest of the code creates
2006 files with open(), and sets the required mode explicitly. Thus, these
2007 files were ending up world-writeable. This was not a big issue, because,
2008 being within the spool directory, they were not world-accessible. I have
2009 created a function called modefopen, which takes an additional mode
2010 argument. It sets umask(777), creates the file, chmods it to the required
2011 mode, then resets the umask. All the relevant calls to fopen() in the
2012 content scanning code have been changed to use this function.
2013
944a9c55
PH
2014PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2015 to 24 hours. This avoids potential overflow problems when processing G
2016 and H retry rules. I suspect nobody ever tinkers with this value.
2017
4a23603b
PH
2018PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2019
4730f942
PH
2020PH/45 When the plaintext authenticator is running as a client, the server's
2021 challenges are checked to ensure they are valid base64 strings. By
2022 default, the authentication attempt is cancelled if an invalid string is
2023 received. Setting client_ignore_invalid_base64 true ignores these errors.
2024 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2025 they are received. Thus, the responses can be made to depend on the
2026 challenges. If an invalid string is ignored, an empty string is placed in
2027 the variable.
2028
30dba1e6
PH
2029PH/46 Messages that are created by the autoreply transport now contains a
2030 References: header, in accordance with RFCs 2822 and 3834.
2031
382afc6b
PH
2032PH/47 Added authenticated_sender_force to the smtp transport.
2033
a86229cf
PH
2034PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2035
50c99ba6
PH
2036PH/49 Installed latest patch from the Sieve maintainer.
2037
d35e429d
PH
2038PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2039 was also set, the appendfile transport was unnecessarily scanning a
2040 directory of message files (e.g. for maildir delivery) to find the count
2041 of files (along with the size), even though it did not need this
2042 information. It now does the scan only if it needs to find either the
2043 size of the count of files.
2044
f90d018c
PH
2045PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2046
75def545
PH
2047PH/52 Two bugs concerned with error handling when the smtp transport is
2048 used in LMTP mode:
2049
2050 (i) Exim was not creating retry information for temporary errors given
2051 for individual recipients after the DATA command when the smtp transport
2052 was used in LMTP mode. This meant that they could be retried too
2053 frequently, and not timed out correctly.
2054
2055 (ii) Exim was setting the flag that allows error details to be returned
2056 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2057 recipients that were returned after the DATA command.
2058
2059PH/53 This is related to PH/52, but is more general: for any failing address,
2060 when detailed error information was permitted to be returned to the
2061 sender, but the error was temporary, then after the final timeout, only
2062 "retry timeout exceeded" was returned. Now it returns the full error as
2063 well as "retry timeout exceeded".
2064
c46782ef
PH
2065PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2066 do this, and (what is worse) MTAs that accept it.
2067
71fafd95
PH
2068PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2069 will now be deprecated.
2070
2c5db4fd
PH
2071PH/56 New os.c-cygwin from the Cygwin maintainer.
2072
9cf6b11a
JJ
2073JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2074 formats (previously only available in exim formats via -bpr, -bpru,
2075 and -bpra. Now also available in native and exiqgrep formats)
2076
2077JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2078 with very large, slow to parse queues
2079
2080JJ/08 exipick: added ! as generic prefix to negate any criteria format
2081
2082JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2083
898d150f
PH
2084PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2085 responses to authentication challenges, though it was showing the
2086 challenges; (ii) I've removed the CR characters from the debug output for
2087 SMTP output lines.
2088
46218253
PH
2089PH/58 Allow for the insertion of a newline as well as a space when a string
2090 is turned into more than one encoded-word during RFC 2047 encoding. The
2091 Sieve code now uses this.
2092
e97957bc
PH
2093PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2094 data_4xx, lost_connection, tls_required.
2095
81e509d7
PH
2096PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2097 message was being sent as an SMTP response.
2098
3d240ff7
PH
2099PH/61 Add -l and -k options to exicyclog.
2100
b37c4101
PH
2101PH/62 When verifying, if an address was redirected to one new address, so that
2102 verification continued, and the new address failed or deferred after
2103 having set something in $address_data, the value of $address_data was not
2104 passed back to the ACL. This was different to the case when no
2105 redirection occurred. The value is now passed back in both cases.
2106
79378e0f
PH
2107PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2108 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2109 use login_cap.h, so on its own it isn't the distinguishing feature. The
2110 new name refers directly to the setclassresources() function.
2111
e49c7bb4
PH
2112PH/65 Added configuration files for NetBSD3.
2113
d114ec46
PH
2114PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2115
f3d7df6c
PH
2116PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2117 is preferred over IPv4.
2118
715ab376
PH
2119PH/68 The bounce_return_message and bounce_return_body options were not being
2120 honoured for bounces generated during the reception of non-SMTP messages.
2121 In particular, this applied to messages rejected by the ACL. This bug has
2122 been fixed. However, if bounce_return_message is true and bounce_return_
2123 body is false, the headers that are returned for a non-SMTP message
2124 include only those that have been read before the error was detected.
2125 (In the case of an ACL rejection, they have all been read.)
2126
6b31b150
PH
2127PH/69 The HTML version of the specification is now built in a directory called
2128 spec_html instead of spec.html, because the latter looks like a path with
2129 a MIME-type, and this confuses some software.
2130
2131PH/70 Catch two compiler warnings in sieve.c.
2132
d515a917
PH
2133PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2134 function verify_get_ident() calls ip_connect() to connect a socket, but
2135 if the "connect()" function timed out, ip_connect() used to close the
2136 socket. However, verify_get_ident() also closes the socket later, and in
2137 between Exim writes to the log, which may get opened at this point. When
2138 the socket was closed in ip_connect(), the log could get the same file
2139 descriptor number as the socket. This naturally causes chaos. The fix is
2140 not to close the socket in ip_connect(); the socket should be closed by
2141 the function that creates it. There was only one place in the code where
2142 this was missing, in the iplookup router, which I don't think anybody now
2143 uses, but I've fixed it anyway.
2144
9b8fadde
PH
2145PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2146 well as to direct DNS lookups. Otherwise the handling of names in host
2147 lists is inconsistent and therefore confusing.
2148
214e2000 2149
5de37277
PH
2150Exim version 4.60
2151-----------------
2152
cc38ddbf
PH
2153PH/01 Two changes to the default runtime configuration:
2154
2155 (1) Move the checks for relay_from_hosts and authenticated clients from
2156 after to before the (commented out) DNS black list checks.
2157
2158 (2) Add control=submission to the relay_from_hosts and authenticated
2159 clients checks, on the grounds that messages accepted by these
2160 statements are most likely to be submissions.
5de37277 2161
72fdd6ae
PH
2162PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2163
2164 (1) Generate an error if the third argument for the ${prvs expansion is
2165 not a single digit.
2166
2167 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2168 string.
2169
2170 (3) Reset the variables that are obtained from the first argument of
2171 ${prvscheck and used in the second argument before leaving the code,
2172 because their memory is reclaimed, so using them afterwards may do
2173 silly things.
2174
2175 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2176 one (it's much easier than Tom thought :-).
2177
2178 (5) Because of (4), we can now allow for the use of $prvscheck_result
2179 inside the third argument.
cb9328de 2180
cb741023
PH
2181PH/03 For some reason, the default setting of PATH when running a command from
2182 a pipe transport was just "/usr/bin". I have changed it to
2183 "/bin:/usr/bin".
2184
f174f16e
PH
2185PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2186 anything to be listed in the output from -bV.
b2f5a032 2187
c25242d7
PH
2188PH/05 When a filter generated an autoreply, the entire To: header line was
2189 quoted in the delivery log line, like this:
2190
2191 => >A.N.Other <ano@some.domain> <original@ddress> ...
2192
2193 This has been changed so that it extracts the operative address. There
2194 may be more than one such address. If so, they are comma-separated, like
2195 this:
2196
2197 => >ano@some.domain,ona@other.domain <original@ddress> ...
2198
82c19f95
PH
2199PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2200 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2201 not being looked up in the rDNS to get a host name, Exim was showing the
2202 IP address twice in Received: lines, even though the IP addresses were
2203 identical. For example:
2204
2205 Received: from [1.2.3.4] (helo=[1.2.3.4])
2206
2207 However, if the real host name was known, it was omitting the HELO data
2208 if it matched the actual IP address. This has been tidied up so that it
2209 doesn't show the same IP address twice.
2210
d7ffbc12
PH
2211PH/07 When both +timestamp and +memory debugging was on, the value given by
2212 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2213 called by the debug printing, thereby overwriting the timestamp buffer.
2214 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2215
9f526266
PH
2216PH/08 When the original message was included in an autoreply transport, it
2217 always said "this is a copy of the message, including all the headers",
2218 even if body_only or headers_only was set. It now gives an appropriate
2219 message.
2220
87fcc8b9
PH
2221PH/09 Applied a patch from the Sieve maintainer which:
2222
2223 o fixes some comments
2224 o adds the (disabled) notify extension core
2225 o adds some debug output for the result of if/elsif tests
2226 o points to the current vacation draft in the documentation
2227 and documents the missing references header update
2228
2229 and most important:
2230
2231 o fixes a bug in processing the envelope test (when testing
2232 multiple envelope elements, the last element determinted the
2233 result)
2234
456682f5
PH
2235PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2236 Electronic Mail") by including:
2237
2238 Auto-submitted: auto-generated
2239
2240 in the messages that it generates (bounce messages and others, such as
2241 warnings). In the case of bounce messages for non-SMTP mesages, there was
2242 also a typo: it was using "Auto_submitted" (underscore instead of
2243 hyphen). Since every message generated by Exim is necessarily in response
2244 to another message, thes have all been changed to:
2245
2246 Auto-Submitted: auto-replied
2247
2248 in accordance with these statements in the RFC:
2249
2250 The auto-replied keyword:
2251
2252 - SHOULD be used on messages sent in direct response to another
2253 message by an automatic process,
2254
2255 - MUST NOT be used on manually-generated messages,
2256
2257 - MAY be used on Delivery Status Notifications (DSNs) and Message
2258 Disposition Notifications (MDNs),
2259
2260 - MUST NOT be used on messages generated by automatic or periodic
2261 processes, except for messages which are automatic responses to
2262 other messages.
2263
3e46c1aa
PH
2264PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2265 to the default Received: header definition.
456682f5 2266
49826d12
PH
2267PH/12 Added log selector acl_warn_skipped (default on).
2268
eba0c039
PH
2269PH/13 After a successful wildlsearch lookup, discard the values of numeric
2270 variables because (a) they are in the wrong storage pool and (b) even if
2271 they were copied, it wouldn't work properly because of the caching.
2272
a0d6ba8a
PH
2273PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2274 checking when decoding. Apparently there are clients that generate
2275 overlong encoded strings. Why am I not surprised?
2276
f0917727
PH
2277PH/15 If the first argument of "${if match_address" was not empty, but did not
2278 contain an "@" character, Exim crashed. Now it writes a panic log message
2279 and treats the condition as false.
2280
096fee00
PH
2281PH/16 In autoreply, treat an empty string for "once" the same as unset.
2282
024bd3c2
PH
2283PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2284 extension "envelope-auth". The code is finished and in agreement with
2285 other implementations, but there is no documentation so far and in fact,
2286 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2287 not changing the active code.
2288
2289 Print executed "if" and "elsif" statements when debugging is used. This
2290 helps a great deal to understand what a filter does.
2291
2292 Document more things not specified clearly in RFC3028. I had all this
2293 sorted out, when out of a sudden new issues came to my mind. Oops."
2294
df199fec
PH
2295PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2296 (Bugzilla #53).
2297
d27f1df3
PH
2298PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2299 canonical form (as documented). However, after a host name lookup from
2300 the IP address, check_host() was doing a simple string comparison with
2301 addresses acquired from the DNS when checking that the found name did
2302 have the original IP as one of its addresses. Since any found IPv6
2303 addresses are likely to be in abbreviated form, the comparison could
2304 fail. Luckily, there already exists a function for doing the comparison
2305 by converting both addresses to binary, so now that is used instead of
2306 the text comparison.
2307
96776534
PH
2308PH/20 There was another similar case to PH/19, when a complete host name was
2309 given in a host list; looking up its IP address could give an abbreviated
2310 form, whereas the current host's name might or might not be abbreviated.
2311 The same fix has been applied.
2312
5de37277 2313
9a799bc0
PH
2314Exim version 4.54
2315-----------------
2316
2317PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2318 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2319 It now does.
2320
99a4b039
PH
2321PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2322 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2323
2324PH/03 Typo: missing ".o" in src/pcre/Makefile.
2325
4b233853
PH
2326PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2327 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2328 Also, for "Auto-Submitted", treat anything other than "no" as
2329 non-personal, in accordance with RFC 3834. (Previously it treated
2330 anything starting "auto-" as non-personal.)
2331
8857ccfd
PH
2332TF/01 The control=submission/name=... option had a problem with syntax
2333 errors if the name included a slash character. The /name= option
2334 now slurps the rest of the string, so it can include any characters
2335 but it must come last in the list of options (after /sender_retain
2336 or /domain=).
2337
433a2980
PH
2338PH/05 Some modifications to the interface to the fake nameserver for the new
2339 testing suite.
2340
3e46c1aa 2341
9a799bc0 2342
e3a311ba
TK
2343Exim version 4.53
2344-----------------
2345
2346TK/01 Added the "success_on_redirect" address verification option. See
2347 NewStuff for rationale and an example.
2348
13b685f9
PH
2349PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2350
395ff96d
PH
2351PH/02 Patch to exigrep to allow it to work on syslog lines.
2352
5b68f6e4
PH
2353PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2354 fread() to skip over the body file's header line, because in Cygwin the
2355 header line is locked and is inaccessible.
2356
1ab52c69
PH
2357PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2358 co-exist for some time) to make it clear that it is the Exim ID that is
2359 referenced, not the Message-ID: header line.
2360
b07e6aa3
PH
2361PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2362 string_format() function, because snprintf() does not exist on all
2363 operating systems.
2364
254e032f
PH
2365PH/06 The use of forbid_filter_existstest now also locks out the use of the
2366 ${stat: expansion item.
2367
3af76a81
PH
2368PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2369 protocol synchronization error", to keep the pedants happy.
2370
2548ba04
PH
2371PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2372 well as for IRIX systems, when gcc is being used. See the host.c source
2373 file for comments.
2374
b6c6011d
PH
2375PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2376
cf39cf57
PH
2377PH/10 Named domain lists were not working if used in a queue_smtp_domains
2378 setting.
2379
f1513293
PH
2380PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2381 transport and to the smtp transport in LMTP mode.
2382
727549a4
PH
2383TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2384
af46795e
PH
2385PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2386 run a filter in a subprocess. This could lead to confusion in subsequent
2387 lookups in the parent process. There should also be a search_tidyup() at
2388 the end of the subprocess.
2389
d7b47fd0
PH
2390PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2391 only if the host matched helo_try_verify_hosts, which caused the
2392 verification to occur when the EHLO/HELO command was issued. The ACL just
2393 tested the remembered result. Now, if a previous verification attempt has
2394 not happened, "verify = helo" does it there and then.
2395
ee744174
JJ
2396JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2397
b582ab87
PH
2398TK/03 Fix log output including CR from clamd.
2399
41a13e0a
PH
2400PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2401 exist provoked a memory error which could cause a segfault.
2402
f625cc5a
PH
2403PH/15 Installed PCRE 6.2
2404
2405PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2406
21f7af35
PH
2407PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2408 of the problem. Specifically, suggested +O2 rather than +O1 for the
2409 HP-UX compiler.
2410
31480e42
PH
2411PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2412
2d280592
PH
2413PH/20 If a delivery was routed to a non-standard port by means of an SRV
2414 record, the port was not correctly logged when the outgoing_port log
2415 selector was set (it logged the transort's default port).
2416
7cd1141b
PH
2417PH/21 Added support for host-specific ports to manualroute, queryprogram,
2418 fallback_hosts, and "hosts" in the smtp transport.
2419
2420PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2421 host errors such as "Connection refused".
2422
750af86e
PH
2423PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2424 authentication with radiusclient 0.4.9:
2425
2426 - Error returned from rc_read_config was caught wrongly
2427 - Username/password not passed on to radius server due to wrong length.
2428
2429 The presumption is that some radiusclient API changes for 4.51/PH/17
2430 were not taken care of correctly. The code is still untested by me (my
2431 Linux distribution still has 0.3.2 of radiusclient), but it was
2432 contributed by a Radius user.
2433
2434PH/24 When doing a callout, the value of $domain wasn't set correctly when
2435 expanding the "port" option of the smtp transport.
2436
4304270b
TK
2437TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2438 while reading a MIME header. Thanks to Tom Hughes for a patch.
2439
750af86e
PH
2440PH/24 Include config.h inside local_scan.h so that configuration settings are
2441 available.
2442
64ffc24f
PH
2443PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2444 that in an ACL for RCPT (for example), you can examine exactly what was
2445 received.
2446
5dd9625b
PH
2447PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2448 commands, but it was not correctly comparing the address with the actual
2449 client host address. Thus, it would show the EHLO address in Received:
2450 header lines when this was not necessary.
2451
5591031b
PH
2452PH/27 Added the % operator to ${eval:}.
2453
ba18e66a
PH
2454PH/28 Exim tries to create and chdir to its spool directory when it starts;
2455 it should be ignoring failures (because with -C, for example, it has lost
2456 privilege). It wasn't ignoring creation failures other than "already
2457 exists".
2458
9cec981f
PH
2459PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2460 -bV or -d is used.
2461
aa2b5c79
PH
2462PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2463 because an input line was too long, either on its own, or by virtue of
1509d3a8
PH
2464 too many continuations, the temporary file was not being removed, and the
2465 return code was incorrect.
aa2b5c79 2466
48a53b7f
PH
2467PH/31 Missing "BOOL" in function definition in filtertest.c.
2468
1c59d63b
PH
2469PH/32 Applied Sieve patches from the maintainer.
2470
671012da
TK
2471TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2472
1509d3a8
PH
2473PH/33 Added "verify = not_blind".
2474
2475PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2476 Local/Makefile (with some defaults set). These are used in built scripts
2477 such as exicyclog, but they have never been used in the exim_install
2478 script (though there are many overriding facilities there). I have
2479 arranged that the exim_install script now takes note of these two
2480 settings.
2481
2482PH/35 Installed configuration files for Dragonfly.
2483
2fe1a124
PH
2484PH/36 When a locally submitted message by a trusted user did not contain a
2485 From: header, and the sender address was obtained from -f or from an SMTP
2486 MAIL command, and the trusted user did not use -F to supply a sender
2487 name, $originator_name was incorrectly used when constructing a From:
2488 header. Furthermore, $originator_name was used for submission mode
2489 messages from external hosts without From: headers in a similar way,
2490 which is clearly wrong.
2491
8800895a
PH
2492PH/37 Added control=suppress_local_fixups.
2493
ccfdb010
PH
2494PH/38 When log_selector = +received_sender was set, and the addition of the
2495 sender made the log line's construction buffer exactly full, or one byte
2496 less than full, an overflow happened when the terminating "\n" was
2497 subsequently added.
2498
1130bfb0
PH
2499PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2500 when the result of a list match is failure because a DNS lookup failed.
2501
ebcb507f
PH
2502PH/40 RM_COMMAND is now used in the building process.
2503
c35e155c
PH
2504PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2505 the "build-* directories that it finds.
2506
95d1f782
PH
2507PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2508 address in a domain literal was a prefix of an interface address.
2509
fd6de02e
PH
2510PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2511 when verifying a sender address, unless rewrite_headers is false.
2512
58de37c5
PH
2513PH/44 Wrote a long comment about why errors_to addresses are verified as
2514 recipients, not senders.
2515
261cf466
TF
2516TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2517 the ratelimit ACL was added.
2518
3ee512ff
PH
2519PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2520
e08c430f
PH
2521PH/46 Added extra information about PostgreSQL errors to the error string.
2522
bef5a11f
PH
2523PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2524 avoiding the need to install special zones in a real server. This is
2525 backwards compatible; if it can't find the fake resolver, it drops back.
2526 Thus, both old and new test suites can be run.
2527
7546de58
TF
2528TF/02 Added util/ratelimit.pl
2529
e5d5a95f
TF
2530TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2531 clock is set back in time.
2532
2e88a017
TF
2533TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2534 Candler <B.Candler@pobox.com>.
2535
a5f65aa4
TF
2536TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2537 OK for addresses that are the result of redirections.
2538
e7726cbf
PH
2539PH/48 A number of further additions for the benefit of the new test suite,
2540 including a fake gethostbyname() that interfaces to the fake DNS resolver
2541 (see PH/47 above).
2542
a7fdad5b
TF
2543TF/06 The fix for widen_domains has also been applied to qualify_single and
2544 search_parents which are the other dnslookup options that can cause
2545 header rewrites.
2546
6af56900
PH
2547PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2548 type ("H").
2549
0925ede6
PH
2550PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2551
66afa403
TF
2552TF/07 Exim produced the error message "an SRV record indicated no SMTP
2553 service" if it encountered an MX record with an empty target hostname.
2554 The message is now "an MX or SRV record indicated no SMTP service".
2555
0154e85a
TF
2556TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2557 if the DNS of the sending site is misconfigured. This is quite a
2558 common situation. This change restores the behaviour of treating a
2559 helo verification defer as a failure.
2560
16f12c76
PH
2561PH/51 If self=fail was set on a router, the bounce message did not include the
2562 actual error message.
2563
bbe902f0 2564
e5a9dba6
PH
2565Exim version 4.52
2566-----------------
2567
2568TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2569
22c3b60b
PH
2570PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2571 command itself ended in error, the underlying message about the transport
2572 filter timeout was being overwritten with the pipe command error. Now the
2573 underlying error message should be appended to the second error message.
2574
06a9b4b5
PH
2575TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2576
c1ac6996
PH
2577PH/02 Reduce the amount of output that "make" produces by default. Full output
2578 can still be requested.
2579
9c7a242c
PH
2580PH/03 The warning log line about a condition test deferring for a "warn" verb
2581 was being output only once per connection, rather than after each
2582 occurrence (because it was using the same function as for successful
2583 "warn" verbs). This seems wrong, so I have changed it.
2584
87ba3f5f
PH
2585TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2586 it should not have, which might have caused a crash in the right
2587 circumstances, but probably never did.
2588
2589PH/04 Installed a modified version of Tony Finch's patch to make submission
2590 mode fix the return path as well as the Sender: header line, and to
2591 add a /name= option so that you can make the user's friendly name appear
2592 in the header line.
2593
29aba418
TF
2594TF/03 Added the control = fakedefer ACL modifier.
2595
fe0dab11
TF
2596TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2597 Mark Lowes for thorough testing.
870f6ba8 2598
11d337a4
TK
2599TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2600
2601TK/03 Merged latest SRS patch from Miles Wilton.
2602
415c8f3b
PH
2603PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2604 with the definition in sysexits.h (which is #included earlier).
2605 Fortunately, Exim does not actually use EX_OK. The code used to try to
2606 preserve the sysexits.h value, by assumimg that macro definitions were
2607 scanned for macro replacements. I have been disabused of this notion,
2608 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 2609
958541e9
PH
2610PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2611 in the smtp transport. When a block could not be written in a single
2612 write() function, the timeout was being re-applied to each part-write.
2613 This seems wrong - if the receiver was accepting one byte at a time it
2614 would take for ever. The timeout is now adjusted when this happens. It
2615 doesn't have to be particularly precise.
2616
c206415f
TK
2617TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2618 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2619
2a4be8f9
PH
2620PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2621 without a domain if the check to <postmaster@domain> fails.
2622
1cba11c5
SC
2623SC/01 Eximstats: added -xls and the ability to specify output files
2624 (patch written by Frank Heydlauf).
2625
2626SC/02 Eximstats: use FileHandles for outputing results.
2627
2628SC/03 Eximstats: allow any combination of xls, txt, and html output.
2629
2630SC/04 Eximstats: fixed display of large numbers with -nvr option
2631
2632SC/05 Eximstats: fixed merging of reports with empty tables.
2633
2634SC/06 Eximstats: added the -include_original_destination flag
2635
2636SC/07 Eximstats: removed tabs and trailing whitespace.
2637
1005d00e
TK
2638TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2639
2640TK/06 MBOX spool code: Add real "From " MBOX separator line
2641 so the .eml file is really in mbox format (even though
2642 most programs do not really care). Patch from Alex Miller.
2643
2644TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2645 The latter is generated from $received_to and is only set if the
2646 message has one envelope recipient. SA can use these headers,
2647 obviously out-of-the-box. Patch from Alex Miller.
2648
9b4768fa
PH
2649PH/08 The ${def test on a variable was returning false if the variable's
2650 value was "0", contrary to what the specification has always said!
2651 The result should be true unless the variable is empty.
2652
2653PH/09 The syntax error of a character other than { following "${if
2654 def:variable_name" (after optional whitespace) was not being diagnosed.
2655 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2656 accidental colon was present, for example, could give incorrect results.
2657
0d7eb84a
PH
2658PH/10 Tidied the code in a number of places where the st_size field of a stat()
2659 result is used (not including appendfile, where other changes are about
2660 to be made).
2661
2662PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2663 This involved changing a lot of size variables from int to off_t. It
2664 should work with maildirs and everything.
2665
40727bee
TK
2666TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2667 spamd dying while we are connected to it.
2668
554d2369
TF
2669TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2670 <jgh@wizmail.org>
2671
1f922db1
PH
2672PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2673 function for BDB 4.3.
2674
ef213c3b
PH
2675PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2676
8ac170f3
PH
2677PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2678 the sources for PCRE have been re-arranged and more files are now
2679 involved.
2680
b1c749bb
PH
2681PH/15 The code I had for printing potentially long long variables in PH/11
2682 above was not the best (it lost precision). The length of off_t variables
2683 is now inspected at build time, and an appropriate printing format (%ld
c6c2dc1d
PH
2684 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2685 to be "long long int" or "long int". This is needed for the internal
2686 formatting function string_vformat().
b1c749bb 2687
4aac9b49
PH
2688PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2689 the configuration file to be ":syslog", then the script "guesses" where
2690 the logs files are, rather than using the compiled in default. In our
2691 case the guess is not the same as the compiled default, so the script
2692 suddenly stopped working when I started to use syslog. The patch checks
2693 to see if log_file_path is "". If so, it attempts to read it from exim
2694 with no configuration file to get the compiled in version, before it
2695 falls back to the previous guessing code."
2696
294520c8
TK
2697TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2698 implementing BATV in an Exim configuration. See NewStuff for the gory
2699 details.
2700
5bd022fe
PH
2701PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2702 Makefile that are specific to HP-UX.
2703
90e9ce59
PH
2704PH/18 If the "use_postmaster" option was set for a recipient callout together
2705 with the "random" option, the postmaster address was used as the MAIL
2706 FROM address for the random test, but not for the subsequent recipient
2707 test. It is now used for both.
2708
5ea81592
PH
2709PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2710 patch removes a few documentation additions to RFC 3028, because the
2711 latest draft now contains them. It adds the new en;ascii-case comparator
2712 and a new error check for 8bit text in MIME parts. Comparator and
2713 require names are now matched exactly. I enabled the subaddress
2714 extension, but it is not well tested yet (read: it works for me)."
2715
c6c2dc1d
PH
2716PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2717 rework some of the code of TK/09 above to avoid the hardwired use of
2718 "%lld" and "long long". Replaced the call to snprintf() with a call to
2719 string_vformat().
2720
fffffe4c
PH
2721PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2722 records point to non-existent hosts", "retry timeout exceeded", and
2723 "retry time not reached for any host after a long failure period".
ca02eafb 2724
9a26b6b2
PH
2725PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2726 experimental DomainKeys support:
2727
2728 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2729 (2) On an error such as an illegally used "control", the wrong name for
2730 the control was given.
2731
2732 These problems did NOT occur unless DomainKeys support was compiled.
2733
4aee0225
PH
2734PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2735
32d668a5
PH
2736PH/24 Added ${if match_ip condition.
2737
8187c3f3
PH
2738PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2739 databases so that it will be absolutely obvious if a crash occurs in the
2740 DB library. This is a regular occurrence (often caused by mis-matched
2741 db.h files).
2742
ff790e47 2743PH/26 Insert a lot of missing (void) casts for functions such as chown(),
f1e894f3
PH
2744 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2745 picked up on a user's system that detects such things. There doesn't seem
2746 to be a gcc warning option for this - only an attribute that has to be
2747 put on the function's prototype. It seems that in Fedora Core 4 they have
2748 set this on a number of new functions. No doubt there will be more in due
2749 course.
ff790e47 2750
5417f6d1
PH
2751PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2752 specify a transport. However, if an address that was verified by such a
2753 router was the subject of a callout, Exim crashed because it tried to
2754 read the rcpt_include_affixes from the non-existent transport. Now it
2755 just assumes that the setting of that option is false. This bug was
2756 introduced by 4.51/PH/31.
2757
59cf8544
PH
2758PH/28 Changed -d+all to exclude +memory, because that information is very
2759 rarely of interest, but it makes the output a lot bigger. People tend to
2760 do -d+all out of habit.
2761
e7ad8a65
PH
2762PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2763 code in os-type was giving problems when libc.so lives in lib64, like on
2764 x86_64 Fedora Core.
2765
ade42478
PH
2766PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2767 aren't the modern standard, and it seems that some systems' include files
2768 don't always have them. Exim was already checking for some of the newer
2769 ones like T_AAAA, and defining it itself. I've added checks for all the
2770 record types that Exim uses.
2771
182ad5cf
PH
2772PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2773 not automatically generating a new one, as it is supposed to. This
2774 prevented TLS from working. If the file did exist, but contained invalid
2775 data, a new version was generated, as expected. It was only the case of a
2776 non-existent file that was broken.
2777
b0d9fc80
TK
2778TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2779 with a change in libdomainkeys > 0.64.
2780
2781TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2782 from DNS. If the selector record carries the flag, it now has
2783 precedence over the domain-wide flag.
2784
2785TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2786
47c7a64a
PH
2787PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2788 the use of a port name that isn't defined in /etc/services) occurred, the
2789 message was deferred as in a normal delivery, and thus remained on the
2790 spool, instead of being failed because of the mua_wrapper setting. This
2791 is now fixed, and I tidied up some of the mua_wrapper messages at the
2792 same time.
2793
a388bce4
SC
2794SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2795 the messages in a hash of arrays rather than using individual hashes.
2796 This is a bit cleaner and results in dramatic memory savings, albeit
2797 at a slight CPU cost.
2798
2799SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2800 as requested by Marc Sherman.
2801
2802SC/10 Eximstats: added histograms for user specified patterns as requested
2803 by Marc Sherman.
2804
0793e4ed
SC
2805SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2806
c58b88df
PH
2807PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2808 fopen() in the content-scanning modules that did not already have it.
2809
e7ad8a65 2810
7982096b
PH
2811Exim version 4.51
2812-----------------
2813
1a46a8c5
PH
2814TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2815 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2816
2f079f46 2817TK/02 Fix ACL "control" statement not being available in MIME ACL.
1a46a8c5
PH
2818
2819TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2820
2821PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2822 to test Sieve filters that use "vacation".
2823
2824PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2825 that changes the way the GnuTLS parameters are stored in the cache file.
2826 The new format can be generated externally. For backward compatibility,
2827 if the data in the cache doesn't make sense, Exim assumes it has read an
2828 old-format file, and it generates new data and writes a new file. This
2829 means that you can't go back to an older release without removing the
2830 file.
2831
2832PH/03 A redirect router that has both "unseen" and "one_time" set does not
2833 work if there are any delivery delays because "one_time" forces the
2834 parent to be marked "delivered", so its unseen clone is never tried
2835 again. For this reason, Exim now forbids the simultaneous setting of
2836 these two options.
2837
2838PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2839 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2840 entry if you want to know the details. The fix, however, neglected to
2841 consider the case when local delivery batching is involved. The test for
2842 "previously delivered" was not happening when checking to see if an
2843 address could be batched with a previous (undelivered) one; under
2844 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 2845 address.
1a46a8c5
PH
2846
2847PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2848 in its include files, and this causes problems building Exim.
2849
2850PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2851 header_syntax) but Exim was just ignoring anything given after a slash.
2852 In particular, this caused confusion with an attempt to use "verify =
2853 reverse_host_lookup/defer_ok". An error is now given when options are
2854 supplied for verify items that do not have them. (Maybe reverse_host_
2855 lookup should have a defer_ok option, but that's a different point.)
2856
2857PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2858 defined by RFC 821) to 2048, because there were problems with some AUTH
2859 commands, and RFC 1869 says the size should be increased for extended
2860 SMTP commands that take arguments.
2861
2862PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2863 Finch).
2864
2865PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2866 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 2867
49c2d5ea
PH
2868PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2869 an address' error message when a string expansion fails (syntax or
f331f3b6
PH
2870 whatever). Otherwise the password may appear in the log. Following change
2871 PH/42 below, there is no longer a chance of it appearing in a bounce
2872 message.
49c2d5ea 2873
bf759a8b
PH
2874PH/11 Installed exipick version 20050225.0 from John Jetmore.
2875
83364d30
PH
2876PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2877 of its addresses was ever tried. (Bugzilla bug #2.)
2878
7999bbd7
PH
2879PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2880 the result incorrectly in the debug output. (It correctly added a newline
2881 to what was transported.)
2882
7dbf77c9
PH
2883TF/01 Added $received_time.
2884
74e0617f
PH
2885PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2886 commented out examples of how to interface to a virus scanner and to
2887 SpamAssassin. Also added commented examples of av_scanner and
2888 spamd_address settings.
2889
2f079f46
PH
2890PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2891 and controls are allowed in which ACLs. There were a couple of minor
2892 errors. Some of the entries in the conditions table (which is a table of
2893 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2894 negation of where the condition IS allowed.
2895
8c841523
PH
2896PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2897
7766a4f0
PH
2898PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2899 header file does not have a version number, so I've had to invent a new
2900 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2901 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2902 radiusclient), but it was contributed by a Radius user.
2903
8b417f2c
PH
2904PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2905 files or directories, for OpenSSL.
2906
901f42cb
PH
2907PH/19 When an Exim process that is running as root has to create an Exim log
2908 file, it does so in a subprocess that runs as exim:exim so as to get the
2909 ownership right at creation (otherwise, other Exim processes might see
2910 the file with the wrong ownership). There was no test for failure of this
2911 fork() call, which would lead to the process getting stuck as it waited
2912 for a non-existent subprocess. Forks do occasionally fail when resources
2913 run out. I reviewed all the other calls to fork(); they all seem to check
2914 for failure.
2915
f9b9210e
PH
2916PH/20 When checking for unexpected SMTP input at connect time (before writing
2917 the banner), Exim was not dealing correctly with a non-positive return
2918 from the read() function. If the client had disconnected by this time,
2919 the result was a log entry for a synchronization error with an empty
2920 string after "input=" when read() returned zero. If read() returned -1
2921 (an event I could not check), uninitialized data bytes were printed.
2922 There were reports of junk text (parts of files, etc) appearing after
2923 "input=".
2924
54cdb463
PH
2925PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2926
cf00dad6
PH
2927PH/22 Added support for macro redefinition, and (re)definition in between
2928 driver and ACL definitions.
2929
acb1b346
PH
2930PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2931 forgetting to use the resulting value; it was using the unexpanded value.
2932
c5ddb310
PH
2933PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2934 hadn't been configured. The fix is from Juergen Kreileder, who
2935 understands it better than I do:
2936
2937 "Here's what I see happening with three configured cyrus_sasl
2938 authenticators configured (plain, login, cram-md5):
2939
2940 On startup auth_cyrus_sasl_init() gets called for each of these.
2941 This means three calls to sasl_listmech() without a specified mech_list.
2942 => SASL tests which mechs of all available mechs actually work
2943 => three warnings about OTP not working
2944 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2945
2946 With the patch, sasl_listmech() also gets called three times. But now
2947 SASL's mech_list option is set to the server_mech specified in the the
2948 authenticator. Or in other words, the answer from sasl_listmech()
2949 gets limited to just the mech you're testing for (which is different
2950 for each call.)
2951 => the return list contains just 'plain' or 'login', 'cram-md5' or
2952 nothing depending on the value of ob->server_mech.
2953
2954 I've just tested the patch: Authentication still works fine,
2955 unavailable mechs specified in the exim configuration are still
2956 caught, and the auth.log warnings about OTP are gone."
2957
31619da6
PH
2958PH/25 When debugging is enabled, the contents of the command line are added
2959 to the debugging output, even when log_selector=+arguments is not
2960 specified.
2961
bebaf0fc
PH
2962PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2963 answer is "GNU", and only if the return is "GNU/something" is the answer
2964 "Linux".
2965
475fe28a
PH
2966PH/27 $acl_verify_message is now set immediately after the failure of a
2967 verification in an ACL, and so is available in subsequent modifiers. In
2968 particular, the message can be preserved by coding like this:
2969
2970 warn !verify = sender
2971 set acl_m0 = $acl_verify_message
2972
2973 Previously, $acl_verify_message was set only while expanding "message"
2974 and "log_message" when a very denied access.
2975
7e8bec7a
PH
2976PH/28 Modified OS/os.c-Linux with
2977
2978 -#ifndef OS_LOAD_AVERAGE
2979 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2980
2981 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2982 nomenclature these days.)
2983
e4a89c47
PH
2984PH/29 Installed patch from the Sieve maintainer that adds the options
2985 sieve_useraddress and sieve_subaddress to the redirect router.
2986
5ca2a9a1
PH
2987PH/30 In these circumstances:
2988 . Two addresses routed to the same list of hosts;
2989 . First host does not offer TLS;
2990 . First host accepts first address;
2991 . First host gives temporary error to second address;
2992 . Second host offers TLS and a TLS session is established;
2993 . Second host accepts second address.
2994 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2995 and peerdn, if requested) that were in fact used only for the second
2996 address.
7e8bec7a 2997
c688b954
PH
2998PH/31 When doing a callout as part of verifying an address, Exim was not paying
2999 attention to any local part prefix or suffix that was matched by the
3000 router that accepted the address. It now behaves in the same way as it
3001 does for delivery: the affixes are removed from the local part unless
3002 rcpt_include_affixes is set on the transport.
3003
fed77020
PH
3004PH/32 Add the sender address, as F=<...>, to the log line when logging a
3005 timeout during the DATA phase of an incoming message.
3006
7fe1560f
PH
3007PH/33 Sieve envelope tests were broken for match types other than :is. I have
3008 applied a patch sanctioned by the Sieve maintainer.
c688b954 3009
ebb6e6d5
PH
3010PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3011 the uid or gid is negative. A case of a negative gid caused this to be
3012 noticed. The fix allows for either to be negative.
3013
9c4e8f60
PH
3014PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3015 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3016 overlooked.
3017
3018PH/36 The change PH/12 above was broken. Fixed it.
3019
d7174846
PH
3020PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3021 the grounds that routing the same address twice would always produce the
3022 same answer. This might have been true once, but it is certainly no
3023 longer true now. Routing a child address may depend on the previous
3024 routing that produced that child. Some complicated redirection strategies
3025 went wrong when messages had multiple recipients, and made Exim's
3026 behaviour dependent on the order in which the addresses were given.
3027
3028 I have moved the duplicate checking until after the routing is complete.
3029 Exim scans the addresses that are assigned to local and remote
3030 transports, and removes any duplicates. This means that more work will be
3031 done, as duplicates will always all be routed, but duplicates are
3032 presumably rare, so I don't expect this is of any significance.
3033
3034 For deliveries to pipes, files, and autoreplies, the duplicate checking
3035 still happens during the routing process, since they are not going to be
3036 routed further.
3037
cfe75fc3
PH
3038PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3039 It corrects a timeout issue with spamd. This is Ian's comment: "The
3040 background is that sometimes spamd either never reads data from a
3041 connection it has accepted, or it never writes response data. The exiscan
3042 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3043 blindly assumes that writes won't block so it may never time out."
3044
be22d70e
PH
3045PH/39 Allow G after quota size as well as K and M.
3046
0612b098
PH
3047PH/40 The value set for $authenticated_id in an authenticator may not contain
3048 binary zeroes or newlines because the value is written to log lines and
3049 to spool files. There was no check on this. Now the value is run through
3050 the string_printing() function so that such characters are converted to
3051 printable escape sequences.
3052
2e0c1448
PH
3053PH/41 $message_linecount is a new variable that contains the total number of
3054 lines in the message. Compare $body_linecount, which is the count for the
3055 body only.
3056
447d236c
PH
3057PH/42 Exim no longer gives details of delivery errors for specific addresses in
3058 bounce and delay warning messages, except in certain special cases, which
3059 are as follows:
3060
3061 (a) An SMTP error message from a remote host;
3062 (b) A message specified in a :fail: redirection;
3063 (c) A message specified in a "fail" command in a system filter;
3064 (d) A message specified in a FAIL return from the queryprogram router;
3065 (e) A message specified by the cannot_route_message router option.
3066
3067 In these cases only, Exim does include the error details in bounce and
3068 warning messages. There are also a few cases where bland messages such
3069 as "unrouteable address" or "local delivery error" are given.
3070
d20976dc
PH
3071PH/43 $value is now also set for the "else" part of a ${run expansion.
3072
f656d135
PH
3073PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3074 being worked on, but at least Exim now implements the latest version to
3075 play with."
3076
2e2a30b4
PH
3077PH/45 In a pipe transport, although a timeout while waiting for the pipe
3078 process to complete was treated as a delivery failure, a timeout while
3079 writing the message to the pipe was logged, but erroneously treated as a
3080 successful delivery. Such timeouts include transport filter timeouts. For
3081 consistency with the overall process timeout, these timeouts are now
3082 treated as errors, giving rise to delivery failures by default. However,
3083 there is now a new Boolean option for the pipe transport called
3084 timeout_defer, which, if set TRUE, converts the failures into defers for
3085 both kinds of timeout. A transport filter timeout is now identified in
3086 the log output.
3087
9176e9f0
PH
3088PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3089 systems where "make" and "gmake" are different, calling "gmake" at top
3090 level broke things. I've arranged for the value of $(MAKE) to be passed
3091 from the Makefile to this script so that it can call the same version of
3092 "make".
3093
7982096b 3094
bbe902f0
PH
3095A note about Exim versions 4.44 and 4.50
3096----------------------------------------
3097
3098Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3099changes of various kinds. As a consequence, a big documentation update was
3100needed. This delayed the release for rather longer than seemed good, especially
3101in the light of a couple of (minor) security issues. Therefore, the changes
3102that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3103release. So 4.44 and 4.50 are in effect two different branches that both start
3104from 4.43.
3105
3106I have left the 4.50 change log unchanged; it contains all the changes since
31074.43. The change log for 4.44 is below; many of its items are identical to
3108those for 4.50. This seems to be the most sensible way to preserve the
3109historical information.
3110
3111
f7b63901 3112Exim version 4.50
495ae4b0
PH
3113-----------------
3114
5fe762f6
PH
3115 1. Minor wording change to the doc/README.SIEVE file.
3116
139059f6 3117 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 3118 computation of the current number of files was incorrect.
495ae4b0 3119
7086e875
PH
3120 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3121 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3122 place.
3123
35af9f61
PH
3124 4. Give more explanation in the error message when the command for a transport
3125 filter fails to execute.
3126
b668c215
PH
3127 5. There are several places where Exim runs a non-Exim command in a
3128 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3129 was being done only for the command run by the queryprogram router. It is
3130 now done for all such subprocesses. The other cases are: ${run, transport
3131 filters, and the commands run by the lmtp and pipe transports.
3132
a494b1e1
PH
3133 6. Added CONFIGURE_GROUP build-time option.
3134
3135 7. Some older OS have a limit of 256 on the maximum number of file
3136 descriptors. Exim was using setrlimit() to set 1000 as a large value
3137 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3138 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 3139
c5fcb476
PH
3140 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3141 was an oversight, and furthermore, ever since the addition of extra
3142 controls (e.g. 4.43/32), the checks on when to allow different forms of
3143 "control" were broken. There should now be diagnostics for all cases when a
3144 control that does not make sense is encountered.
3145
69358f02
PH
3146 9. Added the /retain_sender option to "control=submission".
3147
5be20824
PH
314810. $recipients is now available in the predata ACL (oversight).
3149
eb2c0248
PH
315011. Tidy the search cache before the fork to do a delivery from a message
3151 received from the command line. Otherwise the child will trigger a lookup
3152 failure and thereby defer the delivery if it tries to use (for example) a
3153 cached ldap connection that the parent has called unbind on.
3154
2a3eea10
PH
315512. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3156 of $address_data from the recipient verification was clobbered by the
3157 sender verification.
3158
315913. The value of address_data from a sender verification is now available in
3160 $sender_address_data in subsequent conditions in the ACL statement.
3161
23c7ff99
PH
316214. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3163
4deaf07d
PH
316415. Added a new option "connect=<time>" to callout options, to set a different
3165 connection timeout.
3166
926e1192
PH
316716. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3168 was its contents. (It was OK if the option was not defined at all.)
3169
650edc6f
PH
317017. A "Completed" log line is now written for messages that are removed from
3171 the spool by the -Mrm option.
3172
2c7db3f5
PH
317318. New variables $sender_verify_failure and $recipient_verify_failure contain
3174 information about exactly what failed.
3175
3d235903
PH
317619. Added -dd to debug only the daemon process.
3177
7c7ad977
PH
317820. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3179 handles timeouts, both on the server side and network timeouts. Renamed the
3180 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3181
981756db
PH
318221. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3183 It is now set to "smtps".
3184
d4eb88df
PH
318522. $host_address is now set to the target address during the checking of
3186 ignore_target_hosts.
3187
318823. When checking ignore_target_hosts for an ipliteral router, no host name was
3189 being passed; this would have caused $sender_host_name to have been used if
3190 matching the list had actually called for a host name (not very likely,
3191 since this list is usually IP addresses). A host name is now passed as
3192 "[x.x.x.x]".
3193
7d468ab8
PH
319424. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3195 code that specifies a non-restarting handler (typically sigaction() in
3196 modern systems) in an attempt to fix a rare and obscure crash bug.
3197
319825. Narrowed the window for a race in the daemon that could cause it to ignore
3199 SIGCHLD signals. This is not a major problem, because they are used only to
3200 wake it up if nothing else does.
3201
62c0818f
PH
320226. A malformed maildirsize file could cause Exim to calculate negative values
3203 for the mailbox size or file count. Odd effects could occur as a result.
3204 The maildirsize information is now recalculated if the size or filecount
3205 end up negative.
3206
26034054
PH
320727. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3208 support for a long time. Removed HAVE_SYS_VFS_H.
3209
af66f652
PH
321028. Installed the latest version of exipick from John Jetmore.
3211
90af77f4
PH
321229. In an address list, if the pattern was not a regular expression, an empty
3213 subject address (from a bounce message) matched only if the pattern was an
3214 empty string. Non-empty patterns were not even tested. This was the wrong
3215 because it is perfectly reasonable to use an empty address as part of a
3216 database query. An empty address is now tested by patterns that are
3217 lookups. However, all the other forms of pattern expect the subject to
3218 contain a local part and a domain, and therefore, for them, an empty
3219 address still always fails if the pattern is not itself empty.
3220
d8ef3577
PH
322130. Exim went into a mad DNS loop when attempting to do a callout where the
3222 host was specified on an smtp transport, and looking it up yielded more
3223 than one IP address.
3224
5cb8cbc6
PH
322531. Re-factored the code for checking spool and log partition space into a
3226 function that finds that data and another that does the check. The former
3227 is then used to implement four new variables: $spool_space, $log_space,
3228 $spool_inodes, and $log_inodes.
3229
14702f5b
PH
323032. The RFC2047 encoding function was originally intended for short strings
3231 such as real names; it was not keeping to the 75-character limit for
3232 encoded words that the RFC imposes. It now respects the limit, and
3233 generates multiple encoded words if necessary. To be on the safe side, I
3234 have increased the buffer size for the ${rfc2047: expansion operator from
3235 1024 to 2048 bytes.
3236
063b1e99
PH
323733. It is now permitted to omit both strings after an "if" condition; if the
3238 condition is true, the result is "true". As before, when the second string
3239 is omitted, a false condition yields an empty string. This makes it less
3240 cumbersome to write custom ACL and router conditions.
3241
652e1b65
PH
324234. Failure to deliver a bounce message always caused it to be frozen, even if
3243 there was an errors_to setting on the router. The errors_to setting is now
3244 respected.
3245
6f0c9a4f
PH
324635. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3247 canonical form (fully expanded) before being placed in
3248 $sender_host_address.
3249
33397d19
PH
325036. The table in the code that translates DNS record types into text (T_A to
3251 "A" for instance) was missing entries for NS and CNAME. It is just possible
3252 that this could have caused confusion if both these types were looked up
3253 for the same domain, because the text type is used as part of Exim's
3254 per-process caching. But the chance of anyone hitting this buglet seems
3255 very small.
3256
7bb56e1f
PH
325737. The dnsdb lookup has been extended in a number of ways.
3258
3259 (1) There is a new type, "zns", which walks up the domain tree until it
3260 finds some nameserver records. It should be used with care.
3261
ea3bc19b
PH
3262 (2) There is a new type, "mxh", which is like "mx" except that it returns
3263 just the host names, not the priorities.
3264
3265 (3) It is now possible to give a list of domains (or IP addresses) to be
ff4dbb19
PH
3266 looked up. The behaviour when one of the lookups defers can be
3267 controlled by a keyword.
7bb56e1f 3268
ea3bc19b 3269 (4) It is now possible to specify the separator character for use when
7bb56e1f 3270 multiple records are returned.
33397d19 3271
0bcb2a0e
PH
327238. The dnslists ACL condition has been extended: it is now possible to supply
3273 a list of IP addresses and/or domains to be looked up in a particular DNS
3274 domain.
3275
2ac0e484
PH
327639. Added log_selector=+queue_time_overall.
3277
4e1fde53
PH
327840. When running the queue in the test harness, wait just a tad after forking a
3279 delivery process, to get repeatability of debugging output.
3280
de365ded
PH
328141. Include certificate and key file names in error message when GnuTLS fails
3282 to set them up, because the GnuTLS error message doesn't include the name
3283 of the failing file when there is a problem reading it.
3284
f05da2e8
PH
328542. Allow both -bf and -bF in the same test run.
3286
d6453af2
PH
328743. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3288
f7b63901
PH
328944. The "Exiscan patch" is now merged into the mainline Exim source.
3290
329145. Sometimes the final signoff response after QUIT could fail to get
3292 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3293 < 0 before doing a fflush(). This bug looks as though it goes back to the
3294 introduction of TLS in release 3.20, but "sometimes" must have been rare
3295 because the tests only now provoked it.
3296
a444213a
PH
329746. Reset the locale to "C" after calling embedded Perl, in case it was changed
3298 (this can affect the format of dates).
3299
0ec020ea
PH
330047. exim_tidydb, when checking for the continued existence of a message for
3301 which it has found a message-specific retry record, was not finding
3302 messages that were in split spool directories. Consequently, it was
3303 deleting retry records that should have stayed in existence.
3304
b1206957
PH
330548. Steve fixed some bugs in eximstats.
3306
330749. The SPA authentication driver was not abandoning authentication and moving
3308 on to the next authenticator when an expansion was forced to fail,
3309 contradicting the general specification for all authenticators. Instead it
3310 was generating a temporary error. It now behaves as specified.
3311
26dd5a95
PH
331250. The default ordering of permitted cipher suites for GnuTLS was pessimal
3313 (the order specifies the preference for clients). The order is now AES256,
3314 AES128, 3DES, ARCFOUR128.
3315
343b2385
PH
331651. Small patch to Sieve code - explicitly set From: when generating an
3317 autoreply.
3318
1c5466b9
PH
331952. Exim crashed if a remote delivery caused a very long error message to be
3320 recorded - for instance if somebody sent an entire SpamAssassin report back
3321 as a large number of 550 error lines. This bug was coincidentally fixed by
3322 increasing the size of one of Exim's internal buffers (big_buffer) that
3323 happened as part of the Exiscan merge. However, to be on the safe side, I
3324 have made the code more robust (and fixed the comments that describe what
3325 is going on).
3326
55ee9ee3
PH
332753. Now that there can be additional text after "Completed" in log lines (if
3328 the queue_time_overall log selector is set), a one-byte patch to exigrep
3329 was needed to allow it to recognize "Completed" as not the last thing in
3330 the line.
3331
d38f8232
PH
333254. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3333 patch that reportedly fixes this has been added. I am not expert enough to
3334 create a test for it. This is what the patch creator wrote:
3335
3336 "I found a little strange behaviour of ldap code when working with
3337 Windows 2003 AD Domain, where users was placed in more than one
3338 Organization Units. When I tried to give exim partial DN, the exit code
3339 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3340 But simultaneously result of request was absolutely normal ldap result,
3341 so I produce this patch..."
3342
3295e65b
PH
3343 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3344 REFERENCE, so I have modified the code to exclude the patch when that macro
3345 is not defined.
3346
7102e136
PH
334755. Some experimental protocols are using DNS PTR records for new purposes. The
3348 keys for these records are domain names, not reversed IP addresses. The
b975ba52
PH
3349 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3350 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 3351 CAN-2005-0021
7102e136 3352
3ca0ba97
PH
335356. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3354
c2bcbe20
PH
335557. Double the size of the debug message buffer (to 2048) so that more of very
3356 long debug lines gets shown.
3357
18ce445d
PH
335858. The exicyclog utility now does better if the number of log files to keep
3359 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3360
1f5b4c3d
PH
336159. Two changes related to the smtp_active_hostname option:
3362
3363 (1) $smtp_active_hostname is now available as a variable.
3364 (2) The default for smtp_banner uses $smtp_active_hostname instead
3365 of $primary_hostname.
3366
b975ba52
PH
336760. The host_aton() function is supposed to be passed a string that is known
3368 to be a valid IP address. However, in the case of IPv6 addresses, it was
3369 not checking this. This is a hostage to fortune. Exim now panics and dies
3370 if the condition is not met. A case was found where this could be provoked
85b87bc2
PH
3371 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3372 components; fortuitously, this particular loophole had already been fixed
3373 by change 4.50/55 above.
3374
3375 If there are any other similar loopholes, the new check in host_aton()
3376 itself should stop them being exploited. The report I received stated that
3377 data on the command line could provoke the exploit when Exim was running as
3378 exim, but did not say which command line option was involved. All I could
3379 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3380 running as th