1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
15 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
16 The I= interface field on outgoing lines is now after the H= remote
17 host field, same as incoming lines. There is a separate
18 outgoing_interface log selector which allows you to disable the
21 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
22 If not running log_selector +smtp_connection the mainlog would be held
23 open indefinitely after a "too many connections" event, including to a
24 deleted file after a log rotate. Leave the per net connection logging
25 leaving it open for efficiency as that will be quickly detected by the
26 check on the next write.
28 HS/01 Bug 1671: Fix post transport crash.
29 Processing the wait-<transport> messages could crash the delivery
30 process if the message IDs didn't exist for some reason. When
31 using 'split_spool_directory=yes' the construction of the spool
32 file name failed already, exposing the same netto behaviour.
34 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
35 mime_regex ACL conditions.
37 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
38 to DSN fail messages (bounces): remote IP, remote greeting, remote response
39 to HELO, local diagnostic string.
41 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
42 log line to debug. Even when configured with a tls authenticator many
43 client connections are expected to not authenticate in this way, so
44 an authenticate fail is not an error.
46 HS/02 Add the Exim version string to the process info. This way exiwhat
47 gives some more detail about the running daemon.
49 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
50 matter for fast-change records such as DNSBLs.
52 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
53 retry record, even if constant. There may be multiple transports with
54 different interface settings and the retry behaviour needs to be kept
57 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
59 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
61 JH/10 Bug 840: fix log_defer_output option of pipe transport
63 JH/11 Bug 830: use same host for all RCPTS of a message, even under
64 hosts_randomize. This matters a lot when combined with mua_wrapper.
66 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
67 ${quote_pgsql:<string>} operator.
69 JH/13 Bug 1708: avoid misaligned access in cached lookup.
71 JH/14 Change header file name for freeradius-client. Relevant if compiling
72 with Radius support; from the Gentoo tree and checked under Fedora.
74 JH/15 Bug 1712: Introduce $prdr_requested flag variable
76 JH/16 Bug 1714: Permit an empty string as expansion result for transport
77 option transport_filter, meaning no filtering.
79 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
81 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
82 defaults to "*" (all hosts). The variable is now available when not built
83 with TLS, default unset, mainly to enable keeping the testuite sane.
84 If a server certificate is not supplied (via tls_certificate) an error is
85 logged, and clients will find TLS connections fail on startup. Presumably
86 they will retry in-clear.
87 Packagers of Exim are strongly encouraged to create a server certificate
90 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
91 with the $config_file variable.
93 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
94 in transport context, after the attempt, and per-recipient. The latter type
95 is per host attempted. The event data is the error message, and the errno
96 information encodes the lookup type (A vs. MX) used for the (first) host,
97 and the trailing two digits of the smtp 4xx reponse.
99 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
100 to write to mainlog (or rejectlog, paniclog) in the window between file
101 creation and permissions/ownership being changed. Particularly affects
102 installations where exicyclog is run as root, rather than exim user;
103 result is that the running daemon panics and dies.
105 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
107 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
108 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
109 "pri" and "weight". Note that the previous implicit priority given by the
110 list order is no longer honoured.
112 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
115 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
116 by defining SUPPORT_SOCKS.
118 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
119 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
120 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
121 variables are renamed to proxy_{local,external}_{address,port}.
123 JH/27 Move Internationalisation support from Experimental to mainline, enabled
124 for a build by defining SUPPORT_I18N
126 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
127 of the query string, and make ${quote_redis:} do that quoting.
129 JH/29 Move Events support from Experimental to mainline, enabled by default
130 and removable for a build by defining DISABLE_EVENT.
132 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
134 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
135 cached by the daemon.
137 JH/32 Move Redis support from Experimental to mainline, enabled for a build
138 by defining LOOKUP_REDIS. The libhiredis library is required.
140 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
141 keys are given for lookup.
143 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
144 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
145 only supported when built with TLS support. The PolarSSL SHA routines
146 are still used when the TLS library is too old for convenient support.
148 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
149 openssl_options), for security. OpenSSL forces this from version 1.1.0
150 server-side so match that on older versions.
152 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
153 allocation for $value could be released as the expansion processing
154 concluded, but leaving the global pointer active for it.
156 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
157 and to use the domains and local_parts ACL conditions.
159 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
160 incorrectly not doubled on cutthrough transmission, hence seen as a
161 body-termination at the receiving system - resulting in truncated mails.
162 Commonly the sender saw a TCP-level error, and retransmitted the message
163 via the normal store-and-forward channel. This could result in duplicates
164 received - but deduplicating mailstores were liable to retain only the
165 initial truncated version.
167 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
169 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
171 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
172 we're in there, support oversigning also; bug 1309.
174 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
180 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
183 JH/02 The smtp transport option "multi_domain" is now expanded.
185 JH/03 The smtp transport now requests PRDR by default, if the server offers
188 JH/04 Certificate name checking on server certificates, when exim is a client,
189 is now done by default. The transport option tls_verify_cert_hostnames
190 can be used to disable this per-host. The build option
191 EXPERIMENTAL_CERTNAMES is withdrawn.
193 JH/05 The value of the tls_verify_certificates smtp transport and main options
194 default to the word "system" to access the system default CA bundle.
195 For GnuTLS, only version 3.0.20 or later.
197 JH/06 Verification of the server certificate for a TLS connection is now tried
198 (but not required) by default. The verification status is now logged by
199 default, for both outbound TLS and client-certificate supplying inbound
202 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
205 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
206 Status Notification (bounce) messages are now MIME format per RFC 3464.
207 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
208 under the control of the dsn_advertise_hosts option, and routers may
209 have a dsn_lasthop option.
211 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
212 default, modifiable by a malware= option. The list separator for
213 the options can now be changed in the usual way. Bug 68.
215 JH/10 The smtp_receive_timeout main option is now expanded before use.
217 JH/11 The incoming_interface log option now also enables logging of the
218 local interface on delivery outgoing connections.
220 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
221 if the interface and destination host and port all match.
223 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
226 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
227 Patch from Andrew Lewis.
229 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
230 now supports optional time-restrictions, weighting, and priority
231 modifiers per server. Patch originally by <rommer@active.by>.
233 JH/16 The spamd_address main option now supports a mixed list of local
234 and remote servers. Remote servers can be IPv6 addresses, and
235 specify a port-range.
237 JH/17 Bug 68: The spamd_address main option now supports an optional
238 timeout value per server.
240 JH/18 Bug 1581: Router and transport options headers_add/remove can
241 now have the list separator specified.
243 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
246 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
249 JH/21 Support for the A6 type of dns record is withdrawn.
251 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
252 rather than the verbs used.
254 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
255 from 255 to 1024 chars.
257 JH/24 Verification callouts now attempt to use TLS by default.
259 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
260 are generic router options now. The defaults didn't change.
262 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
263 Original patch from Alexander Shikoff, worked over by JH.
265 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
266 files if ZCAT_COMMAND is not executable.
268 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
270 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
272 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
273 Normally benign, it bites when the pair was led to by a CNAME;
274 modern usage is to not canoicalize the domain to a CNAME target
275 (and we were inconsistent anyway for A-only vs AAAA+A).
277 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
279 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
280 when evaluating $sender_host_dnssec.
282 JH/31 Check the HELO verification lookup for DNSSEC, adding new
283 $sender_helo_dnssec variable.
285 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
287 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
289 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
291 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
292 documented as working, but never had. Support all but $spam_report.
294 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
295 added for tls authenticator.
300 TL/01 When running the test suite, the README says that variables such as
301 no_msglog_check are global and can be placed anywhere in a specific
302 test's script, however it was observed that placement needed to be near
303 the beginning for it to behave that way. Changed the runtest perl
304 script to read through the entire script once to detect and set these
305 variables, reset to the beginning of the script, and then run through
306 the script parsing/test process like normal.
308 TL/02 The BSD's have an arc4random API. One of the functions to induce
309 adding randomness was arc4random_stir(), but it has been removed in
310 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
311 function when detected.
313 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
314 cause callback expansion.
316 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
317 syntax errors in an expansion can be treated as a string instead of
318 logging or causing an error, due to the internal use of bool_lax
319 instead of bool when processing it.
321 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
322 server certificates when making smtp deliveries.
324 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
326 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
328 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
330 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
331 Merged patch from Sebastian Wiedenroth.
333 JH/05 Fix results-pipe from transport process. Several recipients, combined
334 with certificate use, exposed issues where response data items split
335 over buffer boundaries were not parsed properly. This eventually
336 resulted in duplicates being sent. This issue only became common enough
337 to notice due to the introduction of conection certificate information,
338 the item size being so much larger. Found and fixed by Wolfgang Breyha.
340 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
341 size buffer was used, resulting in syntax errors when an expansion
344 JH/07 Add support for directories of certificates when compiled with a GnuTLS
345 version 3.3.6 or later.
347 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
348 is EXPERIMENTAL_EVENT, the main-configuration and transport options
349 both become "event_action", the variables become $event_name, $event_data
350 and $event_defer_errno. There is a new variable $verify_mode, usable in
351 routers, transports and related events. The tls:cert event is now also
352 raised for inbound connections, if the main configuration event_action
355 TL/06 In test suite, disable OCSP for old versions of openssl which contained
356 early OCSP support, but no stapling (appears to be less than 1.0.0).
358 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
359 server certificate names available under the smtp transport option
360 "tls_verify_cert_hostname" now do not permit multi-component wildcard
363 JH/10 Time-related extraction expansions from certificates now use the main
364 option "timezone" setting for output formatting, and are consistent
365 between OpenSSL and GnuTLS compilations. Bug 1541.
367 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
368 encoded parameter in the incoming message. Bug 1558.
370 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
371 include certificate info, eximon was claiming there were spoolfile
374 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
376 JH/14 Log delivery-related information more consistently, using the sequence
377 "H=<name> [<ip>]" wherever possible.
379 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
380 are problematic for Debian distribution, omit them from the release
383 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
385 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
388 JH/17 Fix a null-indirection in certextract expansions when a nondefault
389 output list separator was used.
394 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
395 checkers that were complaining about end of non-void function with no
398 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
399 This was a regression intruduced in 4.83 by another bugfix.
401 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
403 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
404 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
410 TF/01 Correctly close the server side of TLS when forking for delivery.
412 When a message was received over SMTP with TLS, Exim failed to clear up
413 the incoming connection properly after forking off the child process to
414 deliver the message. In some situations the subsequent outgoing
415 delivery connection happened to have the same fd number as the incoming
416 connection previously had. Exim would try to use TLS and fail, logging
417 a "Bad file descriptor" error.
419 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
420 utilities have not been installed.
422 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
423 temporary space as the ACL may create new global variables.
425 TL/01 LDAP support uses per connection or global context settings, depending
426 upon the detected version of the libraries at build time.
428 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
429 to extract and use the src ip:port in logging and expansions as if it
430 were a direct connection from the outside internet. PPv2 support was
431 updated based on HAProxy spec change in May 2014.
433 JH/02 Add ${listextract {number}{list}{success}{fail}}.
435 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
436 Properly escape header and check for NULL return.
438 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
441 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
443 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
444 characters in header names, implemented as a verify condition.
445 Contributed by Michael Fischer v. Mollard.
447 TL/05 Rename SPF condition results err_perm and err_temp to standardized
448 results permerror and temperror. Previous values are deprecated but
449 still accepted. In a future release, err_perm and err_temp will be
450 completely removed, which will be a backward incompatibility if the
451 ACL tests for either of these two old results. Patch contributed by
452 user bes-internal on the mailing list.
454 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
456 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
457 selectors, in both main and reject logs.
459 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
462 JH/07 Add malware type "sock" for talking to simple daemon.
464 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
466 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
467 routers/transports under cutthrough routing.
469 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
470 numbers. Touch up "bool" conditional to keep the same definition.
472 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
474 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
476 JH/12 Expand items in router/transport headers_add or headers_remove lists
477 individually rather than the list as a whole. Bug 1452.
479 Required for reasonable handling of multiple headers_ options when
480 they may be empty; requires that headers_remove items with embedded
481 colons must have them doubled (or the list-separator changed).
483 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
484 view the policy declared in the DMARC record. Currently, $dmarc_status
485 is a combined value of both the record presence and the result of the
488 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
490 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
491 dnslookup router and the smtp transport (applying to the forward
494 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
495 of ldap servers used for a specific lookup. Patch provided by Heiko
498 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
499 New variable $lookup_dnssec_authenticated for observability.
501 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
502 Patch submitted by Lars Timman.
504 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
506 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
507 Requires trusted mode and valid format message id, aborts otherwise.
508 Patch contributed by Heiko Schlichting.
510 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
511 certextract with support for various fields. Bug 1358.
513 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
514 is requested by default, modifiable by smtp transport option
517 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
518 operate on certificate variables to give certificate fingerprints
519 Also new ${sha256:cert_variable}.
521 JH/23 The PRDR feature is moved from being Experimental into the mainline.
523 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
524 Christian Aistleitner.
526 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
528 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
529 file. Patch from Wolfgang Breyha.
531 JH/25 Expand the coverage of the delivery $host and $host_address to
532 client authenticators run in verify callout. Bug 1476.
534 JH/26 Port service names are now accepted for tls_on_connect_ports, to
535 align with daemon_smtp_ports. Bug 72.
537 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
538 support and error reporting did not work properly.
540 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
541 and is readable. Patch from Andrew Colin Kissa.
543 TL/14 Enhance documentation of ${run expansion and how it parses the
544 commandline after expansion, particularly in the case when an
545 unquoted variable expansion results in an empty value.
547 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
549 PP/02 Fix internal collision of T_APL on systems which support RFC3123
550 by renaming away from it. Addresses GH issue 15, reported by
553 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
555 TL/15 SECURITY: prevent double expansion in math comparison functions
556 (can expand unsanitized data). Not remotely exploitable.
563 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
565 PP/02 Make -n do something, by making it not do something.
566 When combined with -bP, the name of an option is not output.
568 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
571 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
572 $sender_host_name and config options to manage this, and basic check
575 PP/05 DSCP support for outbound connections and control modifier for inbound.
577 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
578 (Only plugin which currently uses this is kerberos4, which nobody should
579 be using, but we should make it available and other future plugins might
580 conceivably use it, even though it would break NAT; stuff *should* be
581 using channel bindings instead).
583 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
584 name; added for Sendmail compatibility; requires admin caller.
585 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
586 just ignore it); requires trusted caller.
587 Also parse but ignore: -Ac -Am -X<logfile>
590 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
592 TL/02 Add +smtp_confirmation as a default logging option.
594 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
595 Patch by Magnus Holmgren from 2007-02-20.
597 TL/04 Bugzilla 1281 - Spec typo.
598 Bugzilla 1283 - Spec typo.
599 Bugzilla 1290 - Spec grammar fixes.
601 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
603 TL/06 Add Experimental DMARC support using libopendmarc libraries.
605 TL/07 Fix an out of order global option causing a segfault. Reported to dev
606 mailing list by by Dmitry Isaikin.
608 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
610 JH/02 Support "G" suffix to numbers in ${if comparisons.
612 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
614 NM/01 Bugzilla 1197 - Spec typo
615 Bugzilla 1196 - Spec examples corrections
617 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
619 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
620 gnutls_enable_pkcs11, but renamed to more accurately indicate its
623 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
624 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
626 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
627 "acl {{name}{arg}...}", and optional args on acl condition
630 JH/05 Permit multiple router/transport headers_add/remove lines.
632 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
634 JH/07 Avoid using a waiting database for a single-message-only transport.
635 Performance patch from Paul Fisher. Bugzilla 1262.
637 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
640 JH/09 Add $headers_added variable, with content from use of ACL modifier
641 add_header (but not yet added to the message). Bugzilla 199.
643 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
644 Pulled from Bugzilla 817 by Wolfgang Breyha.
646 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
648 (nb: this is the same fix as in Exim 4.80.1)
650 JH/11 Add A= logging on delivery lines, and a client_set_id option on
653 JH/12 Add optional authenticated_sender logging to A= and a log_selector
656 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
658 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
659 advertise SMTP AUTH mechanism to us, instead of a generic
660 protocol violation error. Also, make Exim more robust to bad
661 data from the Dovecot auth socket.
663 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
665 When a queue runner is handling a message, Exim first routes the
666 recipient addresses, during which it prunes them based on the retry
667 hints database. After that it attempts to deliver the message to
668 any remaining recipients. It then updates the hints database using
671 So if a recipient address works intermittently, it can get repeatedly
672 deferred at routing time. The retry hints record remains fresh so the
673 address never reaches the final cutoff time.
675 This is a fairly common occurrence when a user is bumping up against
676 their storage quota. Exim had some logic in its local delivery code
677 to deal with this. However it did not apply to per-recipient defers
678 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
680 This change adds a proper retry rule check during routing so that the
681 final cutoff time is checked against the message's age. We only do
682 this check if there is an address retry record and there is not a
683 domain retry record; this implies that previous attempts to handle
684 the address had the retry_use_local_parts option turned on. We use
685 this as an approximation for the destination being like a local
686 delivery, as in LMTP.
688 I suspect this new check makes the old local delivery cutoff check
689 redundant, but I have not verified this so I left the code in place.
691 TF/02 Correct gecos expansion when From: is a prefix of the username.
693 Test 0254 submits a message to Exim with the header
697 When I ran the test suite under the user fanf2, Exim expanded
698 the header to contain my full name, whereas it should have added
699 a Resent-Sender: header. It erroneously treats any prefix of the
700 username as equal to the username.
702 This change corrects that bug.
704 GF/01 DCC debug and logging tidyup
705 Error conditions log to paniclog rather than rejectlog.
706 Debug lines prefixed by "DCC: " to remove any ambiguity.
708 TF/03 Avoid unnecessary rebuilds of lookup-related code.
710 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
711 Bug spotted by Jeremy Harris; was flawed since initial commit.
712 Would have resulted in OCSP responses post-SNI triggering an Exim
713 NULL dereference and crash.
715 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
717 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
718 Bug detection, analysis and fix by Samuel Thibault.
719 Bugzilla 1331, Debian bug #698092.
721 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
723 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
724 Server implementation by Todd Lyons, client by JH.
725 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
726 config variable "prdr_enable" controls whether the server
727 advertises the facility. If the client requests PRDR a new
728 acl_data_smtp_prdr ACL is called once for each recipient, after
729 the body content is received and before the acl_smtp_data ACL.
730 The client is controlled by bolth of: a hosts_try_prdr option
731 on the smtp transport, and the server advertisement.
732 Default client logging of deliveries and rejections involving
733 PRDR are flagged with the string "PRDR".
735 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
736 fclose(). Diagnosis by Todd Lyons.
738 PP/17 Update configure.default to handle IPv6 localhost better.
739 Patch by Alain Williams (plus minor tweaks).
742 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
743 This is now consistent with GnuTLS, and is now documented: the
744 previous undocumented portable approach to treating the option as
745 unset was to force an expansion failure. That still works, and
746 an empty string is now equivalent.
748 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
749 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
750 not performing validation itself.
752 PP/20 Added force_command boolean option to pipe transport.
753 Patch from Nick Koston, of cPanel Inc.
755 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
758 TF/04 Added udpsend ACL modifer and hexquote expansion operator
760 PP/21 Fix eximon continuous updating with timestamped log-files.
761 Broken in a format-string cleanup in 4.80, missed when I repaired the
762 other false fix of the same issue.
763 Report and fix from Heiko Schlichting.
766 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
767 Report from Prashanth Katuri.
769 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
770 It's SecureTransport, so affects any MacOS clients which use the
771 system-integrated TLS libraries, including email clients.
773 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
774 using a MIME ACL for non-SMTP local injection.
775 Report and assistance in diagnosis by Warren Baker.
777 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
779 JH/16 Fix comparisons for 64b. Bugzilla 1385.
781 TL/09 Add expansion variable $authenticated_fail_id to keep track of
782 last id that failed so it may be referenced in subsequent ACL's.
784 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
787 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
788 ldap library initialization, allowing self-signed CA's to be
789 used. Also properly sets require_cert option later in code by
790 using NULL (global ldap config) instead of ldap handle (per
791 session). Bug diagnosis and testing by alxgomz.
793 TL/12 Enhanced documentation in the ratelimit.pl script provided in
794 the src/util/ subdirectory.
796 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
797 renamed to Transport Post Delivery Action by Jeremy Harris, as
800 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
801 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
802 redis_servers = needs to be configured which will be used by the redis
803 lookup. Patch from Warren Baker, of The Packet Hub.
805 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
807 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
808 hostname or reverse DNS when processing a host list. Used suggestions
809 from multiple comments on this bug.
811 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
813 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
814 Missed a few lines, added it to make the runtest require no keyboard
817 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
818 contains upper case chars. Make router use caseful_local_part.
820 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
821 support when GnuTLS has been built with p11-kit.
827 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
829 This, or similar/improved, will also be change PP/11 of 4.82.
835 PP/01 Handle short writes when writing local log-files.
836 In practice, only affects FreeBSD (8 onwards).
837 Bugzilla 1053, with thanks to Dmitry Isaikin.
839 NM/01 Bugzilla 949 - Documentation tweak
841 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
844 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
846 PP/02 Implemented gsasl authenticator.
848 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
850 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
851 `pkg-config foo` for cflags/libs.
853 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
854 with rest of GSASL and with heimdal_gssapi.
856 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
857 `pkg-config foo` for cflags/libs for the TLS implementation.
859 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
860 properties get this fed in as external SSF. A number of robustness
861 and debugging improvements to the cyrus_sasl authenticator.
863 PP/08 cyrus_sasl server now expands the server_realm option.
865 PP/09 Bugzilla 1214 - Log authentication information in reject log.
866 Patch by Jeremy Harris.
868 PP/10 Added dbmjz lookup type.
870 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
872 PP/12 MAIL args handles TAB as well as SP, for better interop with
873 non-compliant senders.
874 Analysis and variant patch by Todd Lyons.
876 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
877 Bug report from Lars Müller <lars@samba.org> (via SUSE),
878 Patch from Dirk Mueller <dmueller@suse.com>
880 PP/13 tls_peerdn now print-escaped for spool files.
881 Observed some $tls_peerdn in wild which contained \n, which resulted
882 in spool file corruption.
884 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
885 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
886 or write after TLS renegotiation, which otherwise led to messages
889 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
890 as a tracking header (ie: a signed header comes before the signature).
891 Patch from Wolfgang Breyha.
893 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
894 comma-sep list; embedded commas doubled.
896 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
898 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
900 Report and patch from Dmitry Banschikov.
902 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
903 Removed SSL_clear() after SSL_new() which led to protocol negotiation
904 failures. We appear to now support TLS1.1+ with Exim.
906 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
907 lets Exim select keys and certificates based upon TLS SNI from client.
908 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
909 before an outbound SMTP session. New log_selector, +tls_sni.
911 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
912 NULL dereference. Report and patch from Alun Jones.
914 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
915 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
916 Not seeing resolver debug output on NetBSD, but suspect this is a
917 resolver implementation change.
919 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
920 Left warnings. Added "eximon gdb" invocation mode.
922 PP/21 Defaulting "accept_8bitmime" to true, not false.
924 PP/22 Added -bw for inetd wait mode support.
926 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
927 locate the relevant includes and libraries. Made this the default.
929 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
930 Bugzilla 1246, report and most of solution from Tomasz Kusy.
932 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
933 This may cause build issues on older platforms.
935 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
936 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
937 gnutls_require_mac & gnutls_require_protocols (no longer supported).
938 Added SNI support via GnuTLS too.
939 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
941 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
943 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
944 Applied second patch from Janne, implementing suggestion to default
945 multiple-strings-in-record handling to match SPF spec.
947 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
949 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
950 read-only, out of scope).
951 Patch from Wolfgang Breyha, report from Stuart Northfield.
953 PP/29 Fix three issues highlighted by clang analyser static analysis.
954 Only crash-plausible issue would require the Cambridge-specific
955 iplookup router and a misconfiguration.
956 Report from Marcin Mirosław.
958 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
960 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
961 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
962 As part of this, removing so much warning spew let me fix some minor
963 real issues in debug logging.
965 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
966 assignment on my part. Fixed.
968 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
969 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
970 Janne Snabb (who went above and beyond: thank you).
972 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
973 string otherwise requires a connection and a bunch more work and it's
974 relatively easy to get wrong. Should also expose TLS library linkage
977 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
978 64-bit ${eval} (JH/03).
980 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
981 GNU libc to support some of the 64-bit stuff, should not lead to
982 conflicts. Defined before os.h is pulled in, so if a given platform
983 needs to override this, it can.
985 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
986 protection layer was required, which is not implemented.
987 Bugzilla 1254, patch from Wolfgang Breyha.
989 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
990 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
991 tls_dhparam take prime identifiers. Also unbreak combination of
992 OpenSSL+DH_params+TLSSNI.
994 PP/39 Disable SSLv2 by default in OpenSSL support.
1000 PP/01 Solaris build fix for Oracle's LDAP libraries.
1001 Bugzilla 1109, patch from Stephen Usher.
1003 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1005 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1008 TF/02 Fix a couple more cases where we did not log the error message
1009 when unlink() failed. See also change 4.74-TF/03.
1011 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1012 lock up or crash if it happened to be inside a call to libc when it
1013 got a SIGUSR1 from exiwhat.
1015 The SIGUSR1 handler appends the current process status to the process
1016 log which is later printed by exiwhat. It used to use the general
1017 purpose logging code to do this, but several functions it calls are
1018 not safe for signals.
1020 The new output code in the SIGUSR1 handler is specific to the process
1021 log, and simple enough that it's easy to inspect for signal safety.
1022 Removing some special cases also simplifies the general logging code.
1023 Removing the spurious timestamps from the process log simplifies
1026 TF/04 Improved ratelimit ACL condition.
1028 The /noupdate option has been deprecated in favour of /readonly which
1029 has clearer semantics. The /leaky, /strict, and /readonly update modes
1030 are mutually exclusive. The update mode is no longer included in the
1031 database key; it just determines when the database is updated. (This
1032 means that when you upgrde Exim will forget old rate measurements.)
1034 Exim now checks that the per_* options are used with an update mode that
1035 makes sense for the current ACL. For example, when Exim is processing a
1036 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1037 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1038 must specify per_mail/readonly. If you omit the update mode it defaults to
1039 /leaky where that makes sense (as before) or /readonly where required.
1041 The /noupdate option is now undocumented but still supported for
1042 backwards compatibility. It is equivalent to /readonly except that in
1043 ACLs where /readonly is required you may specify /leaky/noupdate or
1044 /strict/noupdate which are treated the same as /readonly.
1046 A useful new feature is the /count= option. This is a generalization
1047 of the per_byte option, so that you can measure the throughput of other
1048 aggregate values. For example, the per_byte option is now equivalent
1049 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1051 The per_rcpt option has been generalized using the /count= mechanism
1052 (though it's more complicated than the per_byte equivalence). When it is
1053 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1054 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1055 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1056 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1057 Note that using per_rcpt with a non-readonly update mode in more than
1058 one ACL will cause the recipients to be double-counted. (The per_mail
1059 and per_byte options don't have this problem.)
1061 The handling of very low rates has changed slightly. If the computed rate
1062 is less than the event's count (usually one) then this event is the first
1063 after a long gap. In this case the rate is set to the same as this event's
1064 count, so that the first message of a spam run is counted properly.
1066 The major new feature is a mechanism for counting the rate of unique
1067 events. The new per_addr option counts the number of different
1068 recipients that someone has sent messages to in the last time period. It
1069 behaves like per_rcpt if all the recipient addresses are different, but
1070 duplicate recipient addresses do not increase the measured rate. Like
1071 the /count= option this is a general mechanism, so the per_addr option
1072 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1073 example, measure the rate that a client uses different sender addresses
1074 with the options per_mail/unique=$sender_address. There are further
1075 details in the main documentation.
1077 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1079 TF/06 Removed a few PCRE remnants.
1081 TF/07 Automatically extract Exim's version number from tags in the git
1082 repository when doing development or release builds.
1084 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1085 Bugzilla 879. Patch from Paul Fisher.
1087 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1088 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1091 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1093 PP/05 Variable $av_failed, true if the AV scanner deferred.
1094 Bugzilla 1078. Patch from John Horne.
1096 PP/06 Stop make process more reliably on build failure.
1097 Bugzilla 1087. Patch from Heiko Schlittermann.
1099 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1100 Bugzilla 1089. Patch from Heiko Schlittermann.
1102 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1103 Bugzilla 1131. Patch from Holger Weiß.
1105 PP/09 Handle IPv6 addresses with SPF.
1106 Bugzilla 860. Patch from Wolfgang Breyha.
1108 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1110 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1113 PP/11 match_* no longer expand right-hand-side by default.
1114 New compile-time build option, EXPAND_LISTMATCH_RHS.
1115 New expansion conditions, "inlist", "inlisti".
1117 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1119 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1121 PP/14 fix log_write() format string regression from TF/03.
1122 Bugzilla 1152. Patch from Dmitry Isaikin.
1128 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1130 PP/02 Harmonised TLS library version reporting; only show if debugging.
1131 Layout now matches that introduced for other libraries in 4.74 PP/03.
1133 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1135 PP/04 New "dns_use_edns0" global option.
1137 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1140 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1141 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1143 TK/01 Updated PolarSSL code to 0.14.2.
1144 Bugzilla 1097. Patch from Andreas Metzler.
1146 PP/07 Catch divide-by-zero in ${eval:...}.
1147 Fixes bugzilla 1102.
1149 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1152 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1153 format-string attack -- SECURITY: remote arbitrary code execution.
1155 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1156 time unintentionally subject to list matching rules, letting the header
1157 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1158 arbitrary string expansion). This allowed for information disclosure.
1160 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1161 INT_MIN/-1 -- value coerced to INT_MAX.
1167 NM/01 Workround for PCRE version dependency in version reporting
1170 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1171 This fixes portability to compilers other than gcc, notably
1172 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1174 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1175 makefiles for portability to HP-UX and POSIX correctness.
1177 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1178 Also via indented variable definition in the Makefile.
1179 (Debugging by Oliver Heesakkers).
1181 PP/02 Restore caching of spamd results with expanded spamd_address.
1182 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1184 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1185 Improves build reliability. Fix from: Frank Elsner
1187 NM/02 Fix wide character breakage in the rfc2047 coding
1188 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1190 NM/03 Allow underscore in dnslist lookups
1191 Fixes bug 1026. Patch from Graeme Fowler
1193 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1194 Code patches from Adam Ciarcinski of NetBSD.
1196 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1199 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1200 is logged, to avoid truncation. Patch from John Horne.
1202 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1203 Patch from Jakob Hirsch.
1205 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1206 SQL string expansion failure details.
1207 Patch from Andrey Oktyabrski.
1209 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1210 Patch from Simon Arlott.
1212 PP/09 New lookups functionality failed to compile on old gcc which rejects
1213 extern declarations in function scope.
1214 Patch from Oliver Fleischmann
1216 PP/10 Use sig_atomic_t for flags set from signal handlers.
1217 Check getgroups() return and improve debugging.
1218 Fixed developed for diagnosis in bug 927 (which turned out to be
1221 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1222 Patch from Mark Zealey.
1224 PP/12 Bugzilla 1056: Improved spamd server selection.
1225 Patch from Mark Zealey.
1227 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1228 Based on patch from Heiko Schlittermann.
1230 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1231 Patch from Uwe Doering, sign-off by Michael Haardt.
1233 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1234 variable declaration deep within a block. Bug and patch from
1237 PP/15 lookups-Makefile IRIX compatibilty coercion.
1239 PP/16 Make DISABLE_DKIM build knob functional.
1241 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1242 Patch by Simon Arlott
1244 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1245 variable argument macros. Our copy now differs from upstream.
1251 TF/01 Failure to get a lock on a hints database can have serious
1252 consequences so log it to the panic log.
1254 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1255 controlled using the smtp_confirmation log selector.
1257 TF/03 Include the error message when we fail to unlink a spool file.
1259 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1260 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1261 for maintaining out-of-tree patches for some time.
1263 PP/01 Bugzilla 139: Documentation and portability issues.
1264 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1265 Handle per-OS dynamic-module compilation flags.
1267 PP/02 Let /dev/null have normal permissions.
1268 The 4.73 fixes were a little too stringent and complained about the
1269 permissions on /dev/null. Exempt it from some checks.
1270 Reported by Andreas M. Kirchwitz.
1272 PP/03 Report version information for many libraries, including
1273 Exim version information for dynamically loaded libraries. Created
1274 version.h, now support a version extension string for distributors
1275 who patch heavily. Dynamic module ABI change.
1277 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1278 privilege escalation vulnerability whereby the Exim run-time user
1279 can cause root to append content of the attacker's choosing to
1282 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1285 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1286 If dropping privileges for untrusted macros, we disabled normal logging
1287 on the basis that it would fail; for the Exim run-time user, this is not
1288 the case, and it resulted in successful deliveries going unlogged.
1289 Fixed. Reported by Andreas Metzler.
1295 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1296 only prepend for the Resent-* case. Fixes regression introduced in
1297 Exim 4.70 by NM/22 for Bugzilla 607.
1299 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1300 increasing numbers of administrators be bitten by this.
1302 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1304 PP/03 Bugzilla 994: added openssl_options main configuration option.
1306 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1308 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1310 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1312 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1313 without a peer certificate, leading to a segfault because of an
1314 assumption that peers always have certificates. Be a little more
1315 paranoid. Problem reported by Martin Tscholak.
1317 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1318 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1319 NB: ClamAV planning to remove STREAM in "middle of 2010".
1320 CL also introduces -bmalware, various -d+acl logging additions and
1321 more caution in buffer sizes.
1323 PP/09 Implemented reverse_ip expansion operator.
1325 PP/10 Bugzilla 937: provide a "debug" ACL control.
1327 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1329 PP/12 Bugzilla 973: Implement --version.
1331 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1333 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1335 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1337 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1338 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1339 ignore trailing whitespace.
1341 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1343 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1346 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1347 Notification from Dr Andrew Aitchison.
1349 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1350 ExtendedDetectionInfo response format.
1351 Notification from John Horne.
1353 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1356 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1357 XSL and documented dependency on system catalogs, with examples of how
1360 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1363 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1364 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1365 configuration file which is writeable by the Exim user or group.
1367 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1368 of configuration files to cover files specified with the -C option if
1369 they are going to be used with root privileges, not just the default
1372 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1373 option (effectively making it always true).
1375 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1376 files to be used while preserving root privileges.
1378 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1379 that rogue child processes cannot use them.
1381 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1382 run-time user, instead of root.
1384 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1385 Exim run-time user without dropping privileges.
1387 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1388 result string, instead of calling string_vformat() twice with the same
1391 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1392 for other users. Others should always drop root privileges if they use
1393 -C on the command line, even for a whitelisted configure file.
1395 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1397 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1403 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1404 $data_path, and $header_path variables; fixed documentation bugs and
1407 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1408 exipick to access non-standard spools, including the "frozen" queue
1411 NM/01 Bugzilla 965: Support mysql stored procedures.
1412 Patch from Alain Williams
1414 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1416 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1417 Patch from Andreas Metzler
1419 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1420 Patch from Kirill Miazine
1422 NM/05 Bugzilla 671: Added umask to procmail example.
1424 JJ/03 installed exipick 20100323.0, fixing doc bug
1426 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1427 directory. Notification and patch from Dan Rosenberg.
1429 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1431 TK/02 Improve log output when DKIM signing operation fails.
1433 MH/01 Treat the transport option dkim_domain as a colon separated
1434 list, not as a single string, and sign the message with each element,
1435 omitting multiple occurences of the same signer.
1437 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1438 Bugzilla 985, 986. Patch by Simon Arlott
1440 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1441 Patch by Simon Arlott
1443 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1444 MBX locking. Notification from Dan Rosenberg.
1450 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1452 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1454 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1456 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1458 NM/04 Bugzilla 915: Flag broken perl installation during build.
1464 TK/01 Added patch by Johannes Berg that expands the main option
1465 "spamd_address" if it starts with a dollar sign.
1467 TK/02 Write list of recipients to X-Envelope-Sender header when building
1468 the mbox-format spool file for content scanning (suggested by Jakob
1471 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1472 (http://www.dcc-servers.net/) support via dccifd. Activated by
1473 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1475 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1476 by Mark Daniel Reidel <mr@df.eu>.
1478 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1479 When building exim an external PCRE library is now needed -
1480 PCRE is a system library on the majority of modern systems.
1481 See entry on PCRE_LIBS in EDITME file.
1483 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1484 conversation. Added nologin parameter to request.
1485 Patch contributed by Kirill Miazine.
1487 TF/01 Do not log submission mode rewrites if they do not change the address.
1489 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1491 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1492 log files in place. Contributed by Roberto Lima.
1494 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1496 TF/03 Bugzilla 615: When checking the local_parts router precondition
1497 after a local_part_suffix or local_part_prefix option, Exim now
1498 does not use the address's named list lookup cache, since this
1499 contains cached lookups for the whole local part.
1501 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1502 Robert Millan. Documentation is in experimental-spec.txt.
1504 TF/04 Bugzilla 668: Fix parallel build (make -j).
1506 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1508 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1509 Patch provided by Jan Srzednicki.
1511 TF/05 Leading white space used to be stripped from $spam_report which
1512 wrecked the formatting. Now it is preserved.
1514 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1515 that they are available at delivery time.
1517 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1519 TF/08 TLS error reporting now respects the incoming_interface and
1520 incoming_port log selectors.
1522 TF/09 Produce a more useful error message if an SMTP transport's hosts
1523 setting expands to an empty string.
1525 NM/06 Bugzilla 744: EXPN did not work under TLS.
1526 Patch provided by Phil Pennock.
1528 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1529 Patch provided by Richard Godbee.
1531 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1532 acl_smtp_notquit, added index entry.
1534 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1535 Patch provided by Eugene Bujak.
1537 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1538 accept(). Patch provided by Maxim Dounin.
1540 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1541 Patch provided by Phil Pennock.
1543 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1545 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1546 Patch provided by Brad "anomie" Jorsch.
1548 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1549 Patch provided by Dean Brooks.
1551 TK/05 Add native DKIM support (does not depend on external libraries).
1553 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1554 Patch provided by Graeme Fowler.
1556 NM/16 Bugzilla 851: Documentation example syntax fix.
1558 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1560 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1563 NM/19 Bugzilla 745: TLS version reporting.
1564 Patch provided by Phil Pennock.
1566 NM/20 Bugzilla 167: bool: condition support.
1567 Patch provided by Phil Pennock.
1569 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1570 clients. Patch provided by Phil Pennock.
1572 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1573 Patch provided by Brad "anomie" Jorsch.
1575 NM/23 Bugzilla 687: Fix misparses in eximstats.
1576 Patch provided by Heiko Schlittermann.
1578 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1579 Patch provided by Heiko Schlittermann.
1581 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1582 plus update to original patch.
1584 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1586 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1587 Patch provided by David Brownlee.
1589 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1591 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1593 NM/30 Bugzilla 888: TLS documentation bugfixes.
1595 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1597 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1598 Unlike the original bugzilla I have changed all shell scripts in src tree.
1600 NM/33 Bugzilla 898: Transport filter timeout fix.
1601 Patch by Todd Rinaldo.
1603 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1604 Patch by Serge Demonchaux.
1606 NM/35 Bugzilla 39: Base64 decode bug fixes.
1607 Patch by Jakob Hirsch.
1609 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1611 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1613 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1615 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1621 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1622 ALT-N's libdkim that I have put here:
1623 http://duncanthrax.net/exim-experimental/
1625 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1626 were called 'true' and it seems that C99 defines that as a reserved
1627 keyword to be used with 'bool' variable types. That means you could
1628 not include C99-style headers which use bools without triggering
1629 build errors in sieve.c.
1631 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1632 as mailq or other aliases. Changed the --help handling significantly
1633 to do whats expected. exim_usage() emits usage/help information.
1635 SC/01 Added the -bylocaldomain option to eximstats.
1637 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1639 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1641 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1647 PH/01 Another patch from the Sieve maintainer.
1649 PH/02 When an IPv6 address is converted to a string for single-key lookup
1650 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1651 dots are used instead of colons so that keys in lsearch files need not
1652 contain colons. This was done some time before quoting was made available
1653 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1654 (notated using the quote facility) so as to distinguish them from IPv4
1655 keys. This meant that lookups for IP addresses in host lists did not work
1656 for iplsearch lookups.
1658 This has been fixed by arranging for IPv6 addresses to be expressed with
1659 colons if the lookup type is iplsearch. This is not incompatible, because
1660 previously such lookups could never work.
1662 The situation is now rather anomolous, since one *can* have colons in
1663 ordinary lsearch keys. However, making the change in all cases is
1664 incompatible and would probably break a number of configurations.
1666 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1669 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1670 conversion specification without a maximum field width, thereby enabling
1671 a rogue spamd server to cause a buffer overflow. While nobody in their
1672 right mind would setup Exim to query an untrusted spamd server, an
1673 attacker that gains access to a server running spamd could potentially
1674 exploit this vulnerability to run arbitrary code as the Exim user.
1676 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1677 $primary_hostname instead of what libspf2 thinks the hosts name is.
1679 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1680 a directory entry by the name of the lookup key. Previously, if a
1681 symlink pointed to a non-existing file or a file in a directory that
1682 Exim lacked permissions to read, a lookup for a key matching that
1683 symlink would fail. Now it is enough that a matching directory entry
1684 exists, symlink or not. (Bugzilla 503.)
1686 PH/03 The body_linecount and body_zerocount variables are now exported in the
1689 PH/04 Added the $dnslist_matched variable.
1691 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1692 This means they are set thereafter only if the connection becomes
1695 PH/06 Added the client_condition to authenticators so that some can be skipped
1696 by clients under certain conditions.
1698 PH/07 The error message for a badly-placed control=no_multiline_responses left
1699 "_responses" off the end of the name.
1701 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1703 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1704 (without spaces) instead of just copying the configuration text.
1706 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1708 PH/11 Added $max_received_linelength.
1710 PH/12 Added +ignore_defer and +include_defer to host lists.
1712 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1713 way in which PCRE > 7.0 is built.
1715 PH/14 Implemented queue_only_load_latch.
1717 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1718 MAIL command. The effect was to mangle the value on 64-bit systems.
1720 PH/16 Another patch from the Sieve maintainer.
1722 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1724 PH/18 If a system quota error occurred while trying to create the file for
1725 a maildir delivery, the message "Mailbox is full" was not appended to the
1726 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1727 only to a quota excession during the actual writing of the file.
1729 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1730 characters?) which causes problems in log lines. The DN values are now
1731 passed through string_printing() before being added to log lines.
1733 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1734 and InterBase are left for another time.)
1736 PH/21 Added message_body_newlines option.
1738 PH/22 Guard against possible overflow in moan_check_errorcopy().
1740 PH/23 POSIX allows open() to be a macro; guard against that.
1742 PH/24 If the recipient of an error message contained an @ in the local part
1743 (suitably quoted, of course), incorrect values were put in $domain and
1744 $local_part during the evaluation of errors_copy.
1750 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1751 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1754 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1755 issue a MAIL command.
1757 PH/02 In an ACL statement such as
1759 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1761 if a client was not listed at all, or was listed with a value other than
1762 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1763 the condition was not true (as it should be), so access was not denied.
1764 The bug was that the ! inversion was incorrectly passed on to the second
1765 item. This has been fixed.
1767 PH/03 Added additional dnslists conditions == and =& which are different from
1768 = and & when the dns lookup returns more than one IP address.
1770 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1771 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1773 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1774 FSYNC, which compiles an option called disable_fsync that allows for
1775 bypassing fsync(). The documentation is heavily laced with warnings.
1777 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1779 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1780 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1781 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1782 including adding "make clean"; (3) Added -fPIC when compiling the test
1783 dynamically loaded module, to get rid of a warning.
1785 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1786 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1787 = 0s. The bug is otherwise harmless.
1789 PH/07 There was a bug in the dovecot authenticator such that the value of
1790 $auth1 could be overwritten, and so not correctly preserved, after a
1791 successful authentication. This usually meant that the value preserved by
1792 the server_setid option was incorrect.
1794 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1796 PH/09 Installed PCRE release 7.0.
1798 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1799 run for batched SMTP input. It is now run at the start of every message
1800 in the batch. While fixing this I discovered that the process information
1801 (output by running exiwhat) was not always getting set for -bs and -bS
1802 input. This is fixed, and it now also says "batched" for BSMTP.
1804 PH/11 Added control=no_pipelining.
1806 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1807 patch, slightly modified), and move the expansion of helo_data till after
1808 the connection is made in the smtp transport (so it can use these
1811 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1813 PH/14 Added log_selector = +pid.
1815 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1817 PH/16 Add ${if forany and ${if forall.
1819 PH/17 Added dsn_from option to vary the From: line in DSNs.
1821 PH/18 Flush SMTP output before performing a callout, unless control =
1822 no_callout_flush is set.
1824 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1825 was true (the default) a successful delivery failed to delete the retry
1826 item, thus causing premature timeout of the address. The bug is now
1829 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1831 PH/21 Long custom messages for fakedefer and fakereject are now split up
1832 into multiline reponses in the same way that messages for "deny" and
1833 other ACL rejections are.
1835 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1836 with slight modification.
1838 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1839 draft, changing the syntax and factoring some duplicate code".
1841 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1842 for deliveries of the second and subsequent messages over the same SMTP
1845 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1846 ${reduce, with only minor "tidies".
1848 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1850 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1851 expansion side effects.
1853 PH/27 When a message times out after an over-quota error from an Exim-imposed
1854 quota, the bounce message says "mailbox is full". This message was not
1855 being given when it was a system quota that was exceeded. It now should
1858 MH/03 Made $recipients available in local_scan(). local_scan() already has
1859 better access to the recipient list through recipients_list[], but
1860 $recipients can be useful in postmaster-provided expansion strings.
1862 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1863 in the case of a MAIL command with additional options following the
1864 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1865 were accidentally chopped off.
1867 PH/29 SMTP synchronization checks are implemented when a command is read -
1868 there is a check that no more input is waiting when there shouldn't be
1869 any. However, for some commands, a delay in an ACL can mean that it is
1870 some time before the response is written. In this time, more input might
1871 arrive, invalidly. So now there are extra checks after an ACL has run for
1872 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1873 pipelining has not been advertised.
1875 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1877 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1878 returned in the SMTP response when smtp_return_error_details was false.
1879 This has been fixed.
1881 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1882 instead of the C I/O that was originally supplied, because problems were
1883 reported on Solaris.
1885 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1886 Exim which did not show up earlier: it was assuming that a call to
1887 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1888 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1889 was a macro that became an assignment, so it seemed to work. This has
1890 changed to a proper function call with a void return, hence the compile
1891 error. Exim's code has been fixed.
1893 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1896 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1898 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1900 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1901 because it turns out to be too big or there is a timeout) there is no
1902 "Completed" line in the log. When some messages of this type were
1903 selected by exigrep, they were listed as "not completed". Others were
1904 picked up by some special patterns. I have improved the selection
1905 criteria to be more general.
1907 PH/38 The host_find_failed option in the manualroute router can now be set
1908 to "ignore", to completely ignore a host whose IP address cannot be
1909 found. If all hosts are ignored, the behaviour is controlled by the new
1910 host_all_ignored option.
1912 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1913 homing or because of multiple MX records with /mx) generated more than
1914 one IP address, and the following item turned out to be the local host,
1915 all the secondary addresses of the first item were incorrectly removed
1916 from the list, along with the local host and any following hosts (which
1917 is what is supposed to happen).
1919 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1920 whoever called Exim into the -H file. In the case of the daemon it was
1921 behaving confusingly. When first started, it used values for whoever
1922 started the daemon, but after a SIGHUP it used the Exim user (because it
1923 calls itself on a restart). I have changed the code so that it now always
1926 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1927 message are rejected with the same error (e.g. no authentication or bad
1928 sender address), and a DATA command is nevertheless sent (as can happen
1929 with PIPELINING or a stupid MUA), the error message that was given to the
1930 RCPT commands is included in the rejection of the DATA command. This is
1931 intended to be helpful for MUAs that show only the final error to their
1934 PH/42 Another patch from the Sieve maintainer.
1936 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1937 Eximstats - Fixed some broken HTML links and added missing column headers
1939 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1940 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1942 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1944 PH/43 Yet another patch from the Sieve maintainer.
1946 PH/44 I found a way to check for a TCP/IP connection going away before sending
1947 the response to the final '.' that terminates a message, but only in the
1948 case where the client has not sent further data following the '.'
1949 (unfortunately, this is allowed). However, in many cases there won't be
1950 any further data because there won't be any more messages to send. A call
1951 to select() can be used: if it shows that the input is "ready", there is
1952 either input waiting, or the socket has been closed. An attempt to read
1953 the next input character can distinguish the two cases. Previously, Exim
1954 would have sent an OK response which the client would never have see.
1955 This could lead to message repetition. This fix should cure that, at
1956 least in a lot of common cases.
1958 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1959 advertised in response to EHLO.
1965 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1966 fixed by 4.65/MH/01 (is this a record?) are fixed:
1968 (i) An empty string was always treated as zero by the numeric comparison
1969 operators. This behaviour has been restored.
1971 (ii) It is documented that the numeric comparison operators always treat
1972 their arguments as decimal numbers. This was broken in that numbers
1973 starting with 0 were being interpreted as octal.
1975 While fixing these problems I realized that there was another issue that
1976 hadn't been noticed. Values of message_size_limit (both the global option
1977 and the transport option) were treated as octal if they started with 0.
1978 The documentation was vague. These values are now always treated as
1979 decimal, and I will make that clear in the documentation.
1985 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1986 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1989 MH/01 Don't check that the operands of numeric comparison operators are
1990 integers when their expansion is in "skipping" mode (fixes bug
1991 introduced by 4.64-PH/07).
1993 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1994 child addresses, Exim now panics and dies. Previously, because the count
1995 is held in a short int, deliveries were likely to be lost. As such a
1996 large number of recipients for a single message is ridiculous
1997 (performance will be very, very poor), I have chosen to impose a limit
1998 rather than extend the field.
2004 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2005 leftover -K file (the existence of which was triggered by #402).
2006 While we were at it, introduced process PID as part of the -K
2007 filename. This should rule out race conditions when creating
2010 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2011 processing considerably. Previous code took too long for large mails,
2012 triggering a timeout which in turn triggers #401.
2014 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2015 in the DK code in transports.c. sendfile() is not really portable,
2016 hence the _LINUX specificness.
2018 TF/01 In the add_headers option to the mail command in an Exim filter,
2019 there was a bug that Exim would claim a syntax error in any
2020 header after the first one which had an odd number of characters
2023 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2024 callout verification, Exim cached a "reject" for the entire domain. This
2025 is correct for most verifications, but it is not correct for a recipient
2026 verification with use_sender or use_postmaster set, because in that case
2027 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2028 case of MAIL FROM:<> rejection from other early rejections (e.g.
2029 rejection of HELO). When verifying a recipient using a non-null MAIL
2030 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2031 Whatever the result of the callout, the value of the domain cache is
2032 left unchanged (for any other kind of callout, getting as far as trying
2033 RCPT means that the domain itself is ok).
2035 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2038 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2039 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2040 session, the daemon ignores ECONNECT errors and logs others; it now
2041 ignores EPIPE as well.
2043 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2044 (quoted-printable decoding).
2046 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2047 later the small subsequent patch to fix an introduced bug.
2049 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2051 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2053 PH/08 An error is now given if message_size_limit is specified negative.
2055 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2056 to be given (somewhat) arbitrary names.
2058 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2061 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2062 miscellaneous code fixes
2064 PH/10 Added the log_reject_target ACL modifier to specify where to log
2067 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2068 hostname. This is wrong, because it relates to the incoming message (and
2069 probably the interface on which it is arriving) and not to the outgoing
2070 callout (which could be using a different interface). This has been
2071 changed to use the value of the helo_data option from the smtp transport
2072 instead - this is what is used when a message is actually being sent. If
2073 there is no remote transport (possible with a router that sets up host
2074 addresses), $smtp_active_hostname is used.
2076 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2077 tweaks were necessary in order to get it to work (see also 21 below):
2078 (a) The code assumed that strncpy() returns a negative number on buffer
2079 overflow, which isn't the case. Replaced with Exim's string_format()
2081 (b) There were several signed/unsigned issues. I just did the minimum
2082 hacking in of casts. There is scope for a larger refactoring.
2083 (c) The code used strcasecmp() which is not a standard C function.
2084 Replaced with Exim's strcmpic() function.
2085 (d) The code set only $1; it now sets $auth1 as well.
2086 (e) A simple test gave the error "authentication client didn't specify
2087 service in request". It would seem that Dovecot has changed its
2088 interface. Fortunately there's a specification; I followed it and
2089 changed what the client sends and it appears to be working now.
2091 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2094 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2095 address A is aliased to B and C, where B exists and C does not. Without
2096 -v the output is "A verified" because verification stops after a
2097 successful redirection if more than one address is generated. However,
2098 with -v the child addresses are also verified. Exim was outputting "A
2099 failed to verify" and then showing the successful verification for C,
2100 with its parentage. It now outputs "B failed to verify", showing B's
2101 parentage before showing the successful verification of C.
2103 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2104 look up a TXT record in a specific list after matching in a combined
2107 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2108 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2109 they consult the DNS. I had assumed they would set it the way they
2110 wanted; and indeed my experiments on Linux seem to show that in some
2111 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2112 To be on the safe side, however, I have now made the interface to
2113 host_find_byname() similar to host_find_bydns(), with an argument
2114 containing the DNS resolver options. The host_find_byname() function now
2115 sets these options at its start, just as host_find_bydns() does. The smtp
2116 transport options dns_qualify_single and dns_search_parents are passed to
2117 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2118 of host_find_byname() use the default settings of RES_DEFNAMES
2119 (qualify_single) but not RES_DNSRCH (search_parents).
2121 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2122 spool_read_header() do less string testing, by means of a preliminary
2123 switch on the second character of optional "-foo" lines. (This is
2124 overdue, caused by the large number of possibilities that now exist.
2125 Originally there were few.) While I was there, I also converted the
2126 str(n)cmp tests so they don't re-test the leading "-" and the first
2127 character, in the hope this might squeeze out yet more improvement.
2129 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2130 flag allowing group syntax was set by the header_syntax check but not
2131 turned off, possible causing trouble later; (2) The flag was not being
2132 set at all for the header_verify test, causing "group"-style headers to
2133 be rejected. I have now set it in this case, and also caused header_
2134 verify to ignore an empty address taken from a group. While doing this, I
2135 came across some other cases where the code for allowing group syntax
2136 while scanning a header line wasn't quite right (mostly, not resetting
2137 the flag correctly in the right place). These bugs could have caused
2138 trouble for malformed header lines. I hope it is now all correct.
2140 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2141 with the "reply" argument non-NULL. The code, however (which originally
2142 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2143 but it didn't always do it. This confused somebody who was copying the
2144 code for some other use. I have removed all the tests.
2146 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2147 feature that was used to support insecure browsers during the U.S. crypto
2148 embargo. It requires special client support, and Exim is probably the
2149 only MTA that supported it -- and would never use it because real RSA is
2150 always available. This code has been removed, because it had the bad
2151 effect of slowing Exim down by computing (never used) parameters for the
2152 RSA_EXPORT functionality.
2154 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2155 authenticator to fail if there's a tab character in the incoming data
2156 (there should never be unless someone is messing about, as it's supposed
2157 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2158 if the connection is using TLS or if the remote IP is the same as the
2159 local IP, and the "valid-client-cert option" if a client certificate has
2162 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2163 authenticators. This can be used for authorization after authentication
2164 succeeds. (In the case of plaintext, it servers for both authentication
2167 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2168 if any retry times were supplied.
2170 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2171 connection, where there is no client IP address to check. In this
2172 situation, the verify now always succeeds.
2174 PH/25 Applied John Jetmore's -Mset patch.
2176 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2178 PH/27 In a string expansion for a processed (not raw) header when multiple
2179 headers of the same name were present, leading whitespace was being
2180 removed from all of them, but trailing whitespace was being removed only
2181 from the last one. Now trailing whitespace is removed from each header
2182 before concatenation. Completely empty headers in a concatenation (as
2183 before) are ignored.
2185 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2186 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2188 PH/29 [Removed. This was a change that I later backed out, and forgot to
2189 correct the ChangeLog entry (that I had efficiently created) before
2190 committing the later change.]
2192 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2193 address errors (4xx response to RCPT) over the same connection as other
2194 messages routed to the same hosts. Such deliveries are always "forced",
2195 so retry times are not inspected. This resulted in far too many retries
2196 for the affected addresses. The effect occurred only when there were more
2197 hosts than the hosts_max_try setting in the smtp transport when it had
2198 the 4xx errors. Those hosts that it had tried were not added to the list
2199 of hosts for which the message was waiting, so if all were tried, there
2200 was no problem. Two fixes have been applied:
2202 (i) If there are any address or message errors in an SMTP delivery, none
2203 of the hosts (tried or untried) are now added to the list of hosts
2204 for which the message is waiting, so the message should not be a
2205 candidate for sending over the same connection that was used for a
2206 successful delivery of some other message. This seems entirely
2207 reasonable: after all the message is NOT "waiting for some host".
2208 This is so "obvious" that I'm not sure why it wasn't done
2209 previously. Hope I haven't missed anything, but it can't do any
2210 harm, as the worst effect is to miss an optimization.
2212 (ii) If, despite (i), such a delivery is accidentally attempted, the
2213 routing retry time is respected, so at least it doesn't keep
2214 hammering the server.
2216 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2217 in ${readsocket because some servers need this prod.
2219 PH/32 Added some extra debug output when updating a wait-xxx database.
2221 PH/33 The hint "could be header name not terminated by colon", which has been
2222 given for certain expansion errors for a long time, was not being given
2223 for the ${if def:h_colon_omitted{... case.
2225 PH/34 The spec says: "With one important exception, whenever a domain list is
2226 being scanned, $domain contains the subject domain." There was at least
2227 one case where this was not true.
2229 PH/35 The error "getsockname() failed: connection reset by peer" was being
2230 written to the panic log as well as the main log, but it isn't really
2231 panic-worthy as it just means the connection died rather early on. I have
2232 removed the panic log writing for the ECONNRESET error when getsockname()
2235 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2236 runs only) independently of the message's sender address. This meant
2237 that, if the 4xx error was in fact related to the sender, a different
2238 message to the same recipient with a different sender could confuse
2239 things. In particualar, this can happen when sending to a greylisting
2240 server, but other circumstances could also provoke similar problems.
2241 I have changed the default so that the retry time for these errors is now
2242 based a combination of the sender and recipient addresses. This change
2243 can be overridden by setting address_retry_include_sender=false in the
2246 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2247 remote server are returned as part of bounce messages. This was not
2248 happening for LMTP over a pipe (the lmtp transport), but now it is the
2249 same for both kinds of LMTP.
2251 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2252 in header lines that were in fact CNAMEs. This is no longer the case.
2254 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2255 and queue runs started by the daemon processed all messages. This has
2256 been fixed so that -R and -S can now usefully be given with -q<time>.
2258 PH/40 Import PCRE release 6.7 (fixes some bugs).
2260 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2262 PH/42 Give an error if -q is specified more than once.
2264 PH/43 Renamed the variables $interface_address and $interface_port as
2265 $received_ip_address and $received_port, to make it clear that these
2266 values apply to message reception, and not to the outgoing interface when
2267 a message is delivered. (The old names remain recognized, of course.)
2269 PH/44 There was no timeout on the connect() call when using a Unix domain
2270 socket in the ${readsocket expansion. There now is.
2272 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2273 be meaningful with "accept".
2275 SC/01 Eximstats V1.43
2276 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2278 SC/02 Eximstats V1.44
2279 Use a glob alias rather than an array ref in the generated
2280 parser. This improves both readability and performance.
2282 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2283 Collect SpamAssassin and rejection statistics.
2284 Don't display local sender or destination tables unless
2285 there is data to show.
2286 Added average volumes into the top table text output.
2288 SC/04 Eximstats V1.46
2289 Collect data on the number of addresses (recipients)
2290 as well as the number of messages.
2292 SC/05 Eximstats V1.47
2293 Added 'Message too big' to the list of mail rejection
2294 reasons (thanks to Marco Gaiarin).
2296 SC/06 Eximstats V1.48
2297 Mainlog lines which have GMT offsets and are too short to
2298 have a flag are now skipped.
2300 SC/07 Eximstats V1.49 (Alain Williams)
2301 Added the -emptyok flag.
2303 SC/08 Eximstats V1.50
2304 Fixes for obtaining the IP address from reject messages.
2306 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2307 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2308 whitesspace changes from 4.64-PH/27
2310 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2313 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2314 are found, allow negative numbers in numeric criteria)
2316 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2318 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2319 to match changes made in 4.64-PH/43
2321 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2323 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2324 "domain" is actually an IP address, and force a failure. This locks out
2325 those revolvers/nameservers that support "A-for-A" lookups, in
2326 contravention of the specifications.
2328 PH/48 When a host name was looked up from an IP address, and the subsequent
2329 forward lookup of the name timed out, the host name was left in
2330 $sender_host_name, contrary to the specification.
2332 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2333 restricted to single-key lookups, Exim was not diagnosing an error if
2334 * or *@ was used with a query-style lookup.
2336 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2338 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2339 long ago, but noone interested enough thought of it. Let's just say that
2340 the "1.1" means that there are some new functions that weren't there at
2341 some point in the past.
2343 PH/51 Error processing for expansion failure of helo_data from an smtp
2344 transport during callout processing was broken.
2346 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2347 tested/used via the -bh/-bhc/-bs options.
2349 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2350 bug, fixed in subsequent PCRE releases).
2352 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2353 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2355 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2361 SC/01 Use a glob alias rather than an array ref in eximstats generated
2362 parser. This improves both readability and performance.
2364 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2365 Don't display local sender or destination tables in eximstats unless
2366 there is data to show.
2367 Added average volumes into the eximstats top table text output.
2369 SC/03 Collect data on the number of addresses (recipients) as well
2370 as the number of messages in eximstats.
2372 TF/01 Correct an error in the documentation for the redirect router. Exim
2373 does (usually) call initgroups() when daemonizing.
2375 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2376 with consistent privilege compared to when running as a daemon.
2378 TF/03 Note in the spec that $authenticated_id is not set for local
2379 submissions from trusted users.
2381 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2382 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2384 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2385 by adding some example configuration directives to the default
2386 configuration file. A little bit of work is required to uncomment the
2387 directives and define how usernames and passwords are checked, but
2388 there is now a framework to start from.
2390 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2391 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2392 without this. I don't know how relevant this is to other LDAP libraries.
2394 PH/02 Add the verb name to the "unknown ACL verb" error.
2396 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2398 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2400 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2401 directory not expanded when it should be if an expanded home directory
2402 was set for the address (which is overridden by the transport).
2404 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2407 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2408 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2409 because it is too late at that time, and has no effect.
2411 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2412 security issue with \' (bugzilla #107). I could not use the
2413 PQescapeStringConn() function, because it needs a PGconn value as one of
2416 PH/08 When testing addresses using -bt, indicate those final addresses that
2417 are duplicates that would not cause an additional delivery. At least one
2418 person was confused, thinking that -bt output corresponded to deliveries.
2419 (Suppressing duplicates isn't a good idea as you lose the information
2420 about possibly different redirections that led to the duplicates.)
2422 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2423 systems where poll() doesn't work, in particular OS X.
2425 PH/10 Added more information to debugging output for retry time not reached.
2427 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2428 operations in malware.c.
2430 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2433 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2434 syslog_duplication set false, log lines that would normally be written
2435 both the the main log and to the reject log were not written to syslog at
2438 PH/14 In the default configuration, change the use of "message" in ACL warn
2439 statements to "add_header".
2441 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2442 not followed by a command (e.g. "seen endif").
2444 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2445 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2448 PH/17 Added extra conditions to the default value of delay_warning_condition
2452 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2453 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2454 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2457 The Auto-Submitted: and various List- headers are standardised, whereas I
2458 don't think Precedence: ever was.
2460 PH/18 Refactored debugging code in route_finduser() to show more information,
2461 in particular, the error code if getpwnam() issues one.
2463 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2464 This is apparently needed in addition to the PH/07 change above to avoid
2465 any possible encoding problems.
2467 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2468 but not after initializing Perl.
2470 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2471 output them only if debugging. By default they are written stderr,
2472 apparently, which is not desirable.
2474 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2477 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2480 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2482 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2483 authenticated or an ident call has been made. Suppress the default
2484 values for $authenticated_id and $authenticated_sender (but permit -oMai
2485 and -oMas) when testing with -bh.
2487 PH/24 Re-jigged the order of the tests in the default configuration so that the
2488 tests for valid domains and recipients precede the DNS black list and CSA
2489 tests, on the grounds that those ones are more expensive.
2491 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2492 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2493 This bug exists in every version of Exim that I still have, right back to
2496 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2497 However, an attempt to turn on case-sensitivity in a regex key by
2498 including (?-i) didn't work because the subject string was already
2499 lowercased, and the effects were non-intuitive. It turns out that a
2500 one-line patch can be used to allow (?-i) to work as expected.
2506 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2507 other effects) broke the use of negated acl sub-conditions.
2509 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2512 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2513 "Deny" causes Exim to reject the incoming connection with a 554 error.
2514 Unfortunately, if there is a major crisis, such as a disk failure,
2515 tcp-wrappers gives "deny", whereas what one would like would be some
2516 kind of temporary error. A kludge has been added to help with this.
2517 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2518 554 error is used if errno is still zero or contains ENOENT (which occurs
2519 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2522 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2524 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2525 errors. Otherwise a message that provokes a temporary error (when other
2526 messages do not) can cause a whole host to time out.
2528 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2529 addresses were routed directly to files or pipes from a redirect router.
2530 File deliveries just didn't batch; pipe deliveries might have suffered
2533 PH/06 A failure to get a lock for a hints database would erroneously always say
2534 "Failed to get write lock", even when it was really a read lock.
2536 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2537 of 0600. This has been changed to use the value of the lockfile_mode
2538 option (which defaults to 0600).
2540 PH/08 Applied small patch from the Sieve maintainer.
2542 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2543 folder from quota calculations, a direct delivery into this folder messed
2544 up the contents of the maildirsize file. This was because the regex was
2545 used only to exclude .Trash (or whatever) when the size of the mailbox
2546 was calculated. There was no check that a delivery was happening into an
2547 excluded directory. This bug has been fixed by ignoring all quota
2548 processing for deliveries into excluded directories.
2550 PH/10 Added the maildirfolder_create_regex option to appendfile.
2556 PH/01 The code for finding all the local interface addresses on a FreeBSD
2557 system running IPv6 was broken. This may well have applied to all BSD
2558 systems, as well as to others that have similar system calls. The broken
2559 code found IPv4 interfaces correctly, but gave incorrect values for the
2560 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2561 that it would not match correctly against @[] and not recognize the IPv6
2564 PH/02 The ipliteral router was not recognizing addresses of the form user@
2565 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2567 PH/03 Added disable_ipv6.
2569 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2570 decoded form, because it is most often used to construct To: headers
2571 lines in autoreplies, and the decoded form may well be syntactically
2572 invalid. However, $reply_address has leading white space removed, and all
2573 newlines turned into spaces so that the autoreply transport does not
2576 PH/05 If group was specified without a user on a router, and no group or user
2577 was specified on a transport, the group from the router was ignored.
2579 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2580 for visible compile-time settings that can be used to change these
2581 numbers, for those that want even more. Backwards compatibility with old
2582 spool files has been maintained. However, going back to a previous Exim
2583 release will lost any variables that are in spool files.
2585 PH/07 Two small changes when running in the test harness: increase delay when
2586 passing a TCP/IP connection to a new process, in case the original
2587 process has to generate a bounce, and remove special handling of
2588 127.0.0.2 (sic), which is no longer necessary.
2590 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2591 be the same on different OS.
2593 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2596 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2597 whether --show-vars was specified or not
2599 JJ/02 exipick: Added support for new ACL variable spool format introduced
2602 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2603 syntactically invalid From: or Reply-to: line, and a filter used this to
2604 generate an autoreply, and therefore failed to obtain an address for the
2605 autoreply, Exim could try to deliver to a non-existent relative file
2606 name, causing unrelated and misleading errors. What now happens is that
2607 it logs this as a hard delivery error, but does not attempt to create a
2610 PH/11 The exinext utility has a -C option for testing purposes, but although
2611 the given file was scanned by exinext itself; it wasn't being passed on
2612 when Exim was called.
2614 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2615 an end-of-file indication when reading a command response.
2617 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2618 compiled. In many other places in Exim, IPv6 addresses are always
2619 recognized, so I have changed this. It also means that IPv4 domain
2620 literals of the form [IPV4:n.n.n.n] are now always recognized.
2622 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2623 used if the router is not running as root, for example, when verifying at
2624 ACL time, or when using -bh. The debugging output from this situation was
2625 non-existent - all you got was a failure to exec. I have made two
2628 (a) Failures to set uid/gid, the current directory, or a process leader
2629 in a subprocess such as that created by queryprogram now generate
2630 suitable debugging ouput when -d is set.
2632 (b) The queryprogram router detects when it is not running as root,
2633 outputs suitable debugging information if -d is set, and then runs
2634 the subprocess without attempting to change uid/gid.
2636 PH/15 Minor change to Makefile for building test_host (undocumented testing
2639 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2640 additional section of a DNS packet that returns MX or SRV records.
2641 Instead, it always explicitly searches for A/AAAA records. This avoids
2642 major problems that occur when a DNS server includes only records of one
2643 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2644 fixed another bug: if SRV records were looked up and the corresponding
2645 address records were *not* found in the additional section, the port
2646 values from the SRV records were lost.
2648 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2649 using the correct key (the original address) when searching the retry
2650 rules in order to find which one to use for generating the retry hint.
2652 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2653 adding the default one. Similarly, if it contains a Reply-To: header, the
2654 errors_reply_to option, if set, is not used.
2656 PH/19 When calculating a retry time, Exim used to measure the "time since
2657 failure" by looking at the "first failed" field in the retry record. Now
2658 it does not use this if it is later than than the arrival time of the
2659 message. Instead it uses the arrival time. This makes for better
2660 behaviour in cases where some deliveries succeed, thus re-setting the
2661 "first failed" field. An example is a quota failure for a huge message
2662 when small messages continue to be delivered. Without this change, the
2663 "time since failure" will always be short, possible causing more frequent
2664 delivery attempts for the huge message than are intended.
2665 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2667 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2668 $1, $2, $3) because the numerical variables can be reset during some
2669 expansion items (e.g. "match"), thereby losing the authentication data.
2671 PH/21 Make -bV show the size of off_t variables so that the test suite can
2672 decide whether to run tests for quotas > 2G.
2674 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2675 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2676 filecount value is greater than 2G or if a quota value is greater than 2G
2677 on a system where the size of off_t is not greater than 4, a panic error
2680 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2681 never match. The debug and -bh output now contains an explicit error
2682 message indicating a malformed IPv4 address or mask.
2684 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2685 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2686 PH/23 above applies.
2688 PH/25 Do not write to syslog when running in the test harness. The only
2689 occasion when this arises is a failure to open the main or panic logs
2690 (for which there is an explicit test).
2692 PH/26 Added the /no_tell option to "control=freeze".
2694 PH/27 If a host name lookup failed very early in a connection, for example, if
2695 the IP address matched host_lookup and the reverse lookup yielded a name
2696 that did not have a forward lookup, an error message of the form "no IP
2697 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2698 could be logged. Now it outputs the IP address instead of "NULL".
2700 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2701 allows the sender and the authenticated sender to be set when
2702 submitting a message from within Exim. Since child_open_exim() is
2703 documented for local_scan(), the new function should be too.
2705 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2706 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2707 results in an empty string is now treated as unset.
2709 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2711 PH/31 Added sender_verify_fail logging option.
2713 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2714 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2715 tidied the source and removed it altogether.
2717 PH/33 When a queue run was abandoned because the load average was too high, a
2718 log line was always written; now it is written only if the queue_run log
2719 selector is set. In addition, the log line for abandonment now contains
2720 information about the queue run such as the pid. This is always present
2721 in "start" and "stop" lines but was omitted from the "abandon" line.
2723 PH/34 Omit spaces between a header name and the colon in the error message that
2724 is given when verify = headers_syntax fails (if there are lots of them,
2725 the message gets confusing).
2727 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2728 names, as there are now some PTR records that contain slashes. This check
2729 is only to protect against broken name servers that fall over on strange
2730 characters, so the fact that it applies to all lookups doesn't matter.
2732 PH/36 Now that the new test suite is complete, we can remove some of the
2733 special code in Exim that was needed for the old test suite. For example,
2734 sorting DNS records because real resolvers return them in an arbitrary
2735 order. The new test suite's fake resolver always returns records in the
2738 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2739 bounces) except when queue_only is set, to avoid logging races between
2740 the different processes.
2742 PH/38 Panic-die if .include specifies a non-absolute path.
2744 PH/39 A tweak to the "H" retry rule from its user.
2746 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2747 a label. They prevented compilation on older perls.
2749 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2750 a warning to be raised on newish perls.
2752 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2753 on queue. Changes to match documented behaviour of showing count of
2754 messages matching specified criteria.
2756 PH/40 Changed the default ident timeout from 30s to 5s.
2758 PH/41 Added support for the use of login_cap features, on those BSD systems
2759 that have them, for controlling the resources used by pipe deliveries.
2761 PH/42 The content-scanning code uses fopen() to create files in which to put
2762 message data. Previously it was not paying any attention to the mode of
2763 the files. Exim runs with umask(0) because the rest of the code creates
2764 files with open(), and sets the required mode explicitly. Thus, these
2765 files were ending up world-writeable. This was not a big issue, because,
2766 being within the spool directory, they were not world-accessible. I have
2767 created a function called modefopen, which takes an additional mode
2768 argument. It sets umask(777), creates the file, chmods it to the required
2769 mode, then resets the umask. All the relevant calls to fopen() in the
2770 content scanning code have been changed to use this function.
2772 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2773 to 24 hours. This avoids potential overflow problems when processing G
2774 and H retry rules. I suspect nobody ever tinkers with this value.
2776 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2778 PH/45 When the plaintext authenticator is running as a client, the server's
2779 challenges are checked to ensure they are valid base64 strings. By
2780 default, the authentication attempt is cancelled if an invalid string is
2781 received. Setting client_ignore_invalid_base64 true ignores these errors.
2782 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2783 they are received. Thus, the responses can be made to depend on the
2784 challenges. If an invalid string is ignored, an empty string is placed in
2787 PH/46 Messages that are created by the autoreply transport now contains a
2788 References: header, in accordance with RFCs 2822 and 3834.
2790 PH/47 Added authenticated_sender_force to the smtp transport.
2792 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2794 PH/49 Installed latest patch from the Sieve maintainer.
2796 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2797 was also set, the appendfile transport was unnecessarily scanning a
2798 directory of message files (e.g. for maildir delivery) to find the count
2799 of files (along with the size), even though it did not need this
2800 information. It now does the scan only if it needs to find either the
2801 size of the count of files.
2803 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2805 PH/52 Two bugs concerned with error handling when the smtp transport is
2808 (i) Exim was not creating retry information for temporary errors given
2809 for individual recipients after the DATA command when the smtp transport
2810 was used in LMTP mode. This meant that they could be retried too
2811 frequently, and not timed out correctly.
2813 (ii) Exim was setting the flag that allows error details to be returned
2814 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2815 recipients that were returned after the DATA command.
2817 PH/53 This is related to PH/52, but is more general: for any failing address,
2818 when detailed error information was permitted to be returned to the
2819 sender, but the error was temporary, then after the final timeout, only
2820 "retry timeout exceeded" was returned. Now it returns the full error as
2821 well as "retry timeout exceeded".
2823 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2824 do this, and (what is worse) MTAs that accept it.
2826 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2827 will now be deprecated.
2829 PH/56 New os.c-cygwin from the Cygwin maintainer.
2831 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2832 formats (previously only available in exim formats via -bpr, -bpru,
2833 and -bpra. Now also available in native and exiqgrep formats)
2835 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2836 with very large, slow to parse queues
2838 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2840 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2842 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2843 responses to authentication challenges, though it was showing the
2844 challenges; (ii) I've removed the CR characters from the debug output for
2847 PH/58 Allow for the insertion of a newline as well as a space when a string
2848 is turned into more than one encoded-word during RFC 2047 encoding. The
2849 Sieve code now uses this.
2851 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2852 data_4xx, lost_connection, tls_required.
2854 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2855 message was being sent as an SMTP response.
2857 PH/61 Add -l and -k options to exicyclog.
2859 PH/62 When verifying, if an address was redirected to one new address, so that
2860 verification continued, and the new address failed or deferred after
2861 having set something in $address_data, the value of $address_data was not
2862 passed back to the ACL. This was different to the case when no
2863 redirection occurred. The value is now passed back in both cases.
2865 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2866 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2867 use login_cap.h, so on its own it isn't the distinguishing feature. The
2868 new name refers directly to the setclassresources() function.
2870 PH/65 Added configuration files for NetBSD3.
2872 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2874 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2875 is preferred over IPv4.
2877 PH/68 The bounce_return_message and bounce_return_body options were not being
2878 honoured for bounces generated during the reception of non-SMTP messages.
2879 In particular, this applied to messages rejected by the ACL. This bug has
2880 been fixed. However, if bounce_return_message is true and bounce_return_
2881 body is false, the headers that are returned for a non-SMTP message
2882 include only those that have been read before the error was detected.
2883 (In the case of an ACL rejection, they have all been read.)
2885 PH/69 The HTML version of the specification is now built in a directory called
2886 spec_html instead of spec.html, because the latter looks like a path with
2887 a MIME-type, and this confuses some software.
2889 PH/70 Catch two compiler warnings in sieve.c.
2891 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2892 function verify_get_ident() calls ip_connect() to connect a socket, but
2893 if the "connect()" function timed out, ip_connect() used to close the
2894 socket. However, verify_get_ident() also closes the socket later, and in
2895 between Exim writes to the log, which may get opened at this point. When
2896 the socket was closed in ip_connect(), the log could get the same file
2897 descriptor number as the socket. This naturally causes chaos. The fix is
2898 not to close the socket in ip_connect(); the socket should be closed by
2899 the function that creates it. There was only one place in the code where
2900 this was missing, in the iplookup router, which I don't think anybody now
2901 uses, but I've fixed it anyway.
2903 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2904 well as to direct DNS lookups. Otherwise the handling of names in host
2905 lists is inconsistent and therefore confusing.
2911 PH/01 Two changes to the default runtime configuration:
2913 (1) Move the checks for relay_from_hosts and authenticated clients from
2914 after to before the (commented out) DNS black list checks.
2916 (2) Add control=submission to the relay_from_hosts and authenticated
2917 clients checks, on the grounds that messages accepted by these
2918 statements are most likely to be submissions.
2920 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2922 (1) Generate an error if the third argument for the ${prvs expansion is
2925 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2928 (3) Reset the variables that are obtained from the first argument of
2929 ${prvscheck and used in the second argument before leaving the code,
2930 because their memory is reclaimed, so using them afterwards may do
2933 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2934 one (it's much easier than Tom thought :-).
2936 (5) Because of (4), we can now allow for the use of $prvscheck_result
2937 inside the third argument.
2939 PH/03 For some reason, the default setting of PATH when running a command from
2940 a pipe transport was just "/usr/bin". I have changed it to
2943 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2944 anything to be listed in the output from -bV.
2946 PH/05 When a filter generated an autoreply, the entire To: header line was
2947 quoted in the delivery log line, like this:
2949 => >A.N.Other <ano@some.domain> <original@ddress> ...
2951 This has been changed so that it extracts the operative address. There
2952 may be more than one such address. If so, they are comma-separated, like
2955 => >ano@some.domain,ona@other.domain <original@ddress> ...
2957 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2958 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2959 not being looked up in the rDNS to get a host name, Exim was showing the
2960 IP address twice in Received: lines, even though the IP addresses were
2961 identical. For example:
2963 Received: from [1.2.3.4] (helo=[1.2.3.4])
2965 However, if the real host name was known, it was omitting the HELO data
2966 if it matched the actual IP address. This has been tidied up so that it
2967 doesn't show the same IP address twice.
2969 PH/07 When both +timestamp and +memory debugging was on, the value given by
2970 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2971 called by the debug printing, thereby overwriting the timestamp buffer.
2972 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2974 PH/08 When the original message was included in an autoreply transport, it
2975 always said "this is a copy of the message, including all the headers",
2976 even if body_only or headers_only was set. It now gives an appropriate
2979 PH/09 Applied a patch from the Sieve maintainer which:
2981 o fixes some comments
2982 o adds the (disabled) notify extension core
2983 o adds some debug output for the result of if/elsif tests
2984 o points to the current vacation draft in the documentation
2985 and documents the missing references header update
2989 o fixes a bug in processing the envelope test (when testing
2990 multiple envelope elements, the last element determinted the
2993 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2994 Electronic Mail") by including:
2996 Auto-submitted: auto-generated
2998 in the messages that it generates (bounce messages and others, such as
2999 warnings). In the case of bounce messages for non-SMTP mesages, there was
3000 also a typo: it was using "Auto_submitted" (underscore instead of
3001 hyphen). Since every message generated by Exim is necessarily in response
3002 to another message, thes have all been changed to:
3004 Auto-Submitted: auto-replied
3006 in accordance with these statements in the RFC:
3008 The auto-replied keyword:
3010 - SHOULD be used on messages sent in direct response to another
3011 message by an automatic process,
3013 - MUST NOT be used on manually-generated messages,
3015 - MAY be used on Delivery Status Notifications (DSNs) and Message
3016 Disposition Notifications (MDNs),
3018 - MUST NOT be used on messages generated by automatic or periodic
3019 processes, except for messages which are automatic responses to
3022 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3023 to the default Received: header definition.
3025 PH/12 Added log selector acl_warn_skipped (default on).
3027 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3028 variables because (a) they are in the wrong storage pool and (b) even if
3029 they were copied, it wouldn't work properly because of the caching.
3031 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3032 checking when decoding. Apparently there are clients that generate
3033 overlong encoded strings. Why am I not surprised?
3035 PH/15 If the first argument of "${if match_address" was not empty, but did not
3036 contain an "@" character, Exim crashed. Now it writes a panic log message
3037 and treats the condition as false.
3039 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3041 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3042 extension "envelope-auth". The code is finished and in agreement with
3043 other implementations, but there is no documentation so far and in fact,
3044 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3045 not changing the active code.
3047 Print executed "if" and "elsif" statements when debugging is used. This
3048 helps a great deal to understand what a filter does.
3050 Document more things not specified clearly in RFC3028. I had all this
3051 sorted out, when out of a sudden new issues came to my mind. Oops."
3053 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3056 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3057 canonical form (as documented). However, after a host name lookup from
3058 the IP address, check_host() was doing a simple string comparison with
3059 addresses acquired from the DNS when checking that the found name did
3060 have the original IP as one of its addresses. Since any found IPv6
3061 addresses are likely to be in abbreviated form, the comparison could
3062 fail. Luckily, there already exists a function for doing the comparison
3063 by converting both addresses to binary, so now that is used instead of
3064 the text comparison.
3066 PH/20 There was another similar case to PH/19, when a complete host name was
3067 given in a host list; looking up its IP address could give an abbreviated
3068 form, whereas the current host's name might or might not be abbreviated.
3069 The same fix has been applied.
3075 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3076 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3079 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3080 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3082 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3084 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3085 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3086 Also, for "Auto-Submitted", treat anything other than "no" as
3087 non-personal, in accordance with RFC 3834. (Previously it treated
3088 anything starting "auto-" as non-personal.)
3090 TF/01 The control=submission/name=... option had a problem with syntax
3091 errors if the name included a slash character. The /name= option
3092 now slurps the rest of the string, so it can include any characters
3093 but it must come last in the list of options (after /sender_retain
3096 PH/05 Some modifications to the interface to the fake nameserver for the new
3104 TK/01 Added the "success_on_redirect" address verification option. See
3105 NewStuff for rationale and an example.
3107 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3109 PH/02 Patch to exigrep to allow it to work on syslog lines.
3111 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3112 fread() to skip over the body file's header line, because in Cygwin the
3113 header line is locked and is inaccessible.
3115 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3116 co-exist for some time) to make it clear that it is the Exim ID that is
3117 referenced, not the Message-ID: header line.
3119 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3120 string_format() function, because snprintf() does not exist on all
3123 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3124 ${stat: expansion item.
3126 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3127 protocol synchronization error", to keep the pedants happy.
3129 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3130 well as for IRIX systems, when gcc is being used. See the host.c source
3133 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3135 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3138 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3139 transport and to the smtp transport in LMTP mode.
3141 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3143 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3144 run a filter in a subprocess. This could lead to confusion in subsequent
3145 lookups in the parent process. There should also be a search_tidyup() at
3146 the end of the subprocess.
3148 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3149 only if the host matched helo_try_verify_hosts, which caused the
3150 verification to occur when the EHLO/HELO command was issued. The ACL just
3151 tested the remembered result. Now, if a previous verification attempt has
3152 not happened, "verify = helo" does it there and then.
3154 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3156 TK/03 Fix log output including CR from clamd.
3158 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3159 exist provoked a memory error which could cause a segfault.
3161 PH/15 Installed PCRE 6.2
3163 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3165 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3166 of the problem. Specifically, suggested +O2 rather than +O1 for the
3169 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3171 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3172 record, the port was not correctly logged when the outgoing_port log
3173 selector was set (it logged the transort's default port).
3175 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3176 fallback_hosts, and "hosts" in the smtp transport.
3178 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3179 host errors such as "Connection refused".
3181 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3182 authentication with radiusclient 0.4.9:
3184 - Error returned from rc_read_config was caught wrongly
3185 - Username/password not passed on to radius server due to wrong length.
3187 The presumption is that some radiusclient API changes for 4.51/PH/17
3188 were not taken care of correctly. The code is still untested by me (my
3189 Linux distribution still has 0.3.2 of radiusclient), but it was
3190 contributed by a Radius user.
3192 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3193 expanding the "port" option of the smtp transport.
3195 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3196 while reading a MIME header. Thanks to Tom Hughes for a patch.
3198 PH/24 Include config.h inside local_scan.h so that configuration settings are
3201 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3202 that in an ACL for RCPT (for example), you can examine exactly what was
3205 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3206 commands, but it was not correctly comparing the address with the actual
3207 client host address. Thus, it would show the EHLO address in Received:
3208 header lines when this was not necessary.
3210 PH/27 Added the % operator to ${eval:}.
3212 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3213 it should be ignoring failures (because with -C, for example, it has lost
3214 privilege). It wasn't ignoring creation failures other than "already
3217 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3220 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3221 because an input line was too long, either on its own, or by virtue of
3222 too many continuations, the temporary file was not being removed, and the
3223 return code was incorrect.
3225 PH/31 Missing "BOOL" in function definition in filtertest.c.
3227 PH/32 Applied Sieve patches from the maintainer.
3229 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3231 PH/33 Added "verify = not_blind".
3233 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3234 Local/Makefile (with some defaults set). These are used in built scripts
3235 such as exicyclog, but they have never been used in the exim_install
3236 script (though there are many overriding facilities there). I have
3237 arranged that the exim_install script now takes note of these two
3240 PH/35 Installed configuration files for Dragonfly.
3242 PH/36 When a locally submitted message by a trusted user did not contain a
3243 From: header, and the sender address was obtained from -f or from an SMTP
3244 MAIL command, and the trusted user did not use -F to supply a sender
3245 name, $originator_name was incorrectly used when constructing a From:
3246 header. Furthermore, $originator_name was used for submission mode
3247 messages from external hosts without From: headers in a similar way,
3248 which is clearly wrong.
3250 PH/37 Added control=suppress_local_fixups.
3252 PH/38 When log_selector = +received_sender was set, and the addition of the
3253 sender made the log line's construction buffer exactly full, or one byte
3254 less than full, an overflow happened when the terminating "\n" was
3257 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3258 when the result of a list match is failure because a DNS lookup failed.
3260 PH/40 RM_COMMAND is now used in the building process.
3262 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3263 the "build-* directories that it finds.
3265 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3266 address in a domain literal was a prefix of an interface address.
3268 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3269 when verifying a sender address, unless rewrite_headers is false.
3271 PH/44 Wrote a long comment about why errors_to addresses are verified as
3272 recipients, not senders.
3274 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3275 the ratelimit ACL was added.
3277 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3279 PH/46 Added extra information about PostgreSQL errors to the error string.
3281 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3282 avoiding the need to install special zones in a real server. This is
3283 backwards compatible; if it can't find the fake resolver, it drops back.
3284 Thus, both old and new test suites can be run.
3286 TF/02 Added util/ratelimit.pl
3288 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3289 clock is set back in time.
3291 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3292 Candler <B.Candler@pobox.com>.
3294 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3295 OK for addresses that are the result of redirections.
3297 PH/48 A number of further additions for the benefit of the new test suite,
3298 including a fake gethostbyname() that interfaces to the fake DNS resolver
3301 TF/06 The fix for widen_domains has also been applied to qualify_single and
3302 search_parents which are the other dnslookup options that can cause
3305 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3308 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3310 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3311 service" if it encountered an MX record with an empty target hostname.
3312 The message is now "an MX or SRV record indicated no SMTP service".
3314 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3315 if the DNS of the sending site is misconfigured. This is quite a
3316 common situation. This change restores the behaviour of treating a
3317 helo verification defer as a failure.
3319 PH/51 If self=fail was set on a router, the bounce message did not include the
3320 actual error message.
3326 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3328 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3329 command itself ended in error, the underlying message about the transport
3330 filter timeout was being overwritten with the pipe command error. Now the
3331 underlying error message should be appended to the second error message.
3333 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3335 PH/02 Reduce the amount of output that "make" produces by default. Full output
3336 can still be requested.
3338 PH/03 The warning log line about a condition test deferring for a "warn" verb
3339 was being output only once per connection, rather than after each
3340 occurrence (because it was using the same function as for successful
3341 "warn" verbs). This seems wrong, so I have changed it.
3343 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3344 it should not have, which might have caused a crash in the right
3345 circumstances, but probably never did.
3347 PH/04 Installed a modified version of Tony Finch's patch to make submission
3348 mode fix the return path as well as the Sender: header line, and to
3349 add a /name= option so that you can make the user's friendly name appear
3352 TF/03 Added the control = fakedefer ACL modifier.
3354 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3355 Mark Lowes for thorough testing.
3357 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3359 TK/03 Merged latest SRS patch from Miles Wilton.
3361 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3362 with the definition in sysexits.h (which is #included earlier).
3363 Fortunately, Exim does not actually use EX_OK. The code used to try to
3364 preserve the sysexits.h value, by assumimg that macro definitions were
3365 scanned for macro replacements. I have been disabused of this notion,
3366 so now the code just undefines EX_OK before #including unistd.h.
3368 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3369 in the smtp transport. When a block could not be written in a single
3370 write() function, the timeout was being re-applied to each part-write.
3371 This seems wrong - if the receiver was accepting one byte at a time it
3372 would take for ever. The timeout is now adjusted when this happens. It
3373 doesn't have to be particularly precise.
3375 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3376 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3378 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3379 without a domain if the check to <postmaster@domain> fails.
3381 SC/01 Eximstats: added -xls and the ability to specify output files
3382 (patch written by Frank Heydlauf).
3384 SC/02 Eximstats: use FileHandles for outputing results.
3386 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3388 SC/04 Eximstats: fixed display of large numbers with -nvr option
3390 SC/05 Eximstats: fixed merging of reports with empty tables.
3392 SC/06 Eximstats: added the -include_original_destination flag
3394 SC/07 Eximstats: removed tabs and trailing whitespace.
3396 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3398 TK/06 MBOX spool code: Add real "From " MBOX separator line
3399 so the .eml file is really in mbox format (even though
3400 most programs do not really care). Patch from Alex Miller.
3402 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3403 The latter is generated from $received_to and is only set if the
3404 message has one envelope recipient. SA can use these headers,
3405 obviously out-of-the-box. Patch from Alex Miller.
3407 PH/08 The ${def test on a variable was returning false if the variable's
3408 value was "0", contrary to what the specification has always said!
3409 The result should be true unless the variable is empty.
3411 PH/09 The syntax error of a character other than { following "${if
3412 def:variable_name" (after optional whitespace) was not being diagnosed.
3413 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3414 accidental colon was present, for example, could give incorrect results.
3416 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3417 result is used (not including appendfile, where other changes are about
3420 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3421 This involved changing a lot of size variables from int to off_t. It
3422 should work with maildirs and everything.
3424 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3425 spamd dying while we are connected to it.
3427 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3430 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3431 function for BDB 4.3.
3433 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3435 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3436 the sources for PCRE have been re-arranged and more files are now
3439 PH/15 The code I had for printing potentially long long variables in PH/11
3440 above was not the best (it lost precision). The length of off_t variables
3441 is now inspected at build time, and an appropriate printing format (%ld
3442 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3443 to be "long long int" or "long int". This is needed for the internal
3444 formatting function string_vformat().
3446 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3447 the configuration file to be ":syslog", then the script "guesses" where
3448 the logs files are, rather than using the compiled in default. In our
3449 case the guess is not the same as the compiled default, so the script
3450 suddenly stopped working when I started to use syslog. The patch checks
3451 to see if log_file_path is "". If so, it attempts to read it from exim
3452 with no configuration file to get the compiled in version, before it
3453 falls back to the previous guessing code."
3455 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3456 implementing BATV in an Exim configuration. See NewStuff for the gory
3459 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3460 Makefile that are specific to HP-UX.
3462 PH/18 If the "use_postmaster" option was set for a recipient callout together
3463 with the "random" option, the postmaster address was used as the MAIL
3464 FROM address for the random test, but not for the subsequent recipient
3465 test. It is now used for both.
3467 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3468 patch removes a few documentation additions to RFC 3028, because the
3469 latest draft now contains them. It adds the new en;ascii-case comparator
3470 and a new error check for 8bit text in MIME parts. Comparator and
3471 require names are now matched exactly. I enabled the subaddress
3472 extension, but it is not well tested yet (read: it works for me)."
3474 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3475 rework some of the code of TK/09 above to avoid the hardwired use of
3476 "%lld" and "long long". Replaced the call to snprintf() with a call to
3479 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3480 records point to non-existent hosts", "retry timeout exceeded", and
3481 "retry time not reached for any host after a long failure period".
3483 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3484 experimental DomainKeys support:
3486 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3487 (2) On an error such as an illegally used "control", the wrong name for
3488 the control was given.
3490 These problems did NOT occur unless DomainKeys support was compiled.
3492 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3494 PH/24 Added ${if match_ip condition.
3496 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3497 databases so that it will be absolutely obvious if a crash occurs in the
3498 DB library. This is a regular occurrence (often caused by mis-matched
3501 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3502 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3503 picked up on a user's system that detects such things. There doesn't seem
3504 to be a gcc warning option for this - only an attribute that has to be
3505 put on the function's prototype. It seems that in Fedora Core 4 they have
3506 set this on a number of new functions. No doubt there will be more in due
3509 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3510 specify a transport. However, if an address that was verified by such a
3511 router was the subject of a callout, Exim crashed because it tried to
3512 read the rcpt_include_affixes from the non-existent transport. Now it
3513 just assumes that the setting of that option is false. This bug was
3514 introduced by 4.51/PH/31.
3516 PH/28 Changed -d+all to exclude +memory, because that information is very
3517 rarely of interest, but it makes the output a lot bigger. People tend to
3518 do -d+all out of habit.
3520 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3521 code in os-type was giving problems when libc.so lives in lib64, like on
3524 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3525 aren't the modern standard, and it seems that some systems' include files
3526 don't always have them. Exim was already checking for some of the newer
3527 ones like T_AAAA, and defining it itself. I've added checks for all the
3528 record types that Exim uses.
3530 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3531 not automatically generating a new one, as it is supposed to. This
3532 prevented TLS from working. If the file did exist, but contained invalid
3533 data, a new version was generated, as expected. It was only the case of a
3534 non-existent file that was broken.
3536 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3537 with a change in libdomainkeys > 0.64.
3539 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3540 from DNS. If the selector record carries the flag, it now has
3541 precedence over the domain-wide flag.
3543 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3545 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3546 the use of a port name that isn't defined in /etc/services) occurred, the
3547 message was deferred as in a normal delivery, and thus remained on the
3548 spool, instead of being failed because of the mua_wrapper setting. This
3549 is now fixed, and I tidied up some of the mua_wrapper messages at the
3552 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3553 the messages in a hash of arrays rather than using individual hashes.
3554 This is a bit cleaner and results in dramatic memory savings, albeit
3555 at a slight CPU cost.
3557 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3558 as requested by Marc Sherman.
3560 SC/10 Eximstats: added histograms for user specified patterns as requested
3563 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3565 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3566 fopen() in the content-scanning modules that did not already have it.
3572 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3573 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3575 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3577 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3579 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3580 to test Sieve filters that use "vacation".
3582 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3583 that changes the way the GnuTLS parameters are stored in the cache file.
3584 The new format can be generated externally. For backward compatibility,
3585 if the data in the cache doesn't make sense, Exim assumes it has read an
3586 old-format file, and it generates new data and writes a new file. This
3587 means that you can't go back to an older release without removing the
3590 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3591 work if there are any delivery delays because "one_time" forces the
3592 parent to be marked "delivered", so its unseen clone is never tried
3593 again. For this reason, Exim now forbids the simultaneous setting of
3596 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3597 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3598 entry if you want to know the details. The fix, however, neglected to
3599 consider the case when local delivery batching is involved. The test for
3600 "previously delivered" was not happening when checking to see if an
3601 address could be batched with a previous (undelivered) one; under
3602 certain circumstances this could lead to multiple deliveries to the same
3605 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3606 in its include files, and this causes problems building Exim.
3608 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3609 header_syntax) but Exim was just ignoring anything given after a slash.
3610 In particular, this caused confusion with an attempt to use "verify =
3611 reverse_host_lookup/defer_ok". An error is now given when options are
3612 supplied for verify items that do not have them. (Maybe reverse_host_
3613 lookup should have a defer_ok option, but that's a different point.)
3615 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3616 defined by RFC 821) to 2048, because there were problems with some AUTH
3617 commands, and RFC 1869 says the size should be increased for extended
3618 SMTP commands that take arguments.
3620 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3623 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3624 "unknown" error; now it says that the functionality isn't in the binary.
3626 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3627 an address' error message when a string expansion fails (syntax or
3628 whatever). Otherwise the password may appear in the log. Following change
3629 PH/42 below, there is no longer a chance of it appearing in a bounce
3632 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3634 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3635 of its addresses was ever tried. (Bugzilla bug #2.)
3637 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3638 the result incorrectly in the debug output. (It correctly added a newline
3639 to what was transported.)
3641 TF/01 Added $received_time.
3643 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3644 commented out examples of how to interface to a virus scanner and to
3645 SpamAssassin. Also added commented examples of av_scanner and
3646 spamd_address settings.
3648 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3649 and controls are allowed in which ACLs. There were a couple of minor
3650 errors. Some of the entries in the conditions table (which is a table of
3651 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3652 negation of where the condition IS allowed.
3654 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3656 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3657 header file does not have a version number, so I've had to invent a new
3658 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3659 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3660 radiusclient), but it was contributed by a Radius user.
3662 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3663 files or directories, for OpenSSL.
3665 PH/19 When an Exim process that is running as root has to create an Exim log
3666 file, it does so in a subprocess that runs as exim:exim so as to get the
3667 ownership right at creation (otherwise, other Exim processes might see
3668 the file with the wrong ownership). There was no test for failure of this
3669 fork() call, which would lead to the process getting stuck as it waited
3670 for a non-existent subprocess. Forks do occasionally fail when resources
3671 run out. I reviewed all the other calls to fork(); they all seem to check
3674 PH/20 When checking for unexpected SMTP input at connect time (before writing
3675 the banner), Exim was not dealing correctly with a non-positive return
3676 from the read() function. If the client had disconnected by this time,
3677 the result was a log entry for a synchronization error with an empty
3678 string after "input=" when read() returned zero. If read() returned -1
3679 (an event I could not check), uninitialized data bytes were printed.
3680 There were reports of junk text (parts of files, etc) appearing after
3683 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3685 PH/22 Added support for macro redefinition, and (re)definition in between
3686 driver and ACL definitions.
3688 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3689 forgetting to use the resulting value; it was using the unexpanded value.
3691 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3692 hadn't been configured. The fix is from Juergen Kreileder, who
3693 understands it better than I do:
3695 "Here's what I see happening with three configured cyrus_sasl
3696 authenticators configured (plain, login, cram-md5):
3698 On startup auth_cyrus_sasl_init() gets called for each of these.
3699 This means three calls to sasl_listmech() without a specified mech_list.
3700 => SASL tests which mechs of all available mechs actually work
3701 => three warnings about OTP not working
3702 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3704 With the patch, sasl_listmech() also gets called three times. But now
3705 SASL's mech_list option is set to the server_mech specified in the the
3706 authenticator. Or in other words, the answer from sasl_listmech()
3707 gets limited to just the mech you're testing for (which is different
3709 => the return list contains just 'plain' or 'login', 'cram-md5' or
3710 nothing depending on the value of ob->server_mech.
3712 I've just tested the patch: Authentication still works fine,
3713 unavailable mechs specified in the exim configuration are still
3714 caught, and the auth.log warnings about OTP are gone."
3716 PH/25 When debugging is enabled, the contents of the command line are added
3717 to the debugging output, even when log_selector=+arguments is not
3720 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3721 answer is "GNU", and only if the return is "GNU/something" is the answer
3724 PH/27 $acl_verify_message is now set immediately after the failure of a
3725 verification in an ACL, and so is available in subsequent modifiers. In
3726 particular, the message can be preserved by coding like this:
3728 warn !verify = sender
3729 set acl_m0 = $acl_verify_message
3731 Previously, $acl_verify_message was set only while expanding "message"
3732 and "log_message" when a very denied access.
3734 PH/28 Modified OS/os.c-Linux with
3736 -#ifndef OS_LOAD_AVERAGE
3737 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3739 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3740 nomenclature these days.)
3742 PH/29 Installed patch from the Sieve maintainer that adds the options
3743 sieve_useraddress and sieve_subaddress to the redirect router.
3745 PH/30 In these circumstances:
3746 . Two addresses routed to the same list of hosts;
3747 . First host does not offer TLS;
3748 . First host accepts first address;
3749 . First host gives temporary error to second address;
3750 . Second host offers TLS and a TLS session is established;
3751 . Second host accepts second address.
3752 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3753 and peerdn, if requested) that were in fact used only for the second
3756 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3757 attention to any local part prefix or suffix that was matched by the
3758 router that accepted the address. It now behaves in the same way as it
3759 does for delivery: the affixes are removed from the local part unless
3760 rcpt_include_affixes is set on the transport.
3762 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3763 timeout during the DATA phase of an incoming message.
3765 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3766 applied a patch sanctioned by the Sieve maintainer.
3768 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3769 the uid or gid is negative. A case of a negative gid caused this to be
3770 noticed. The fix allows for either to be negative.
3772 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3773 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3776 PH/36 The change PH/12 above was broken. Fixed it.
3778 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3779 the grounds that routing the same address twice would always produce the
3780 same answer. This might have been true once, but it is certainly no
3781 longer true now. Routing a child address may depend on the previous
3782 routing that produced that child. Some complicated redirection strategies
3783 went wrong when messages had multiple recipients, and made Exim's
3784 behaviour dependent on the order in which the addresses were given.
3786 I have moved the duplicate checking until after the routing is complete.
3787 Exim scans the addresses that are assigned to local and remote
3788 transports, and removes any duplicates. This means that more work will be
3789 done, as duplicates will always all be routed, but duplicates are
3790 presumably rare, so I don't expect this is of any significance.
3792 For deliveries to pipes, files, and autoreplies, the duplicate checking
3793 still happens during the routing process, since they are not going to be
3796 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3797 It corrects a timeout issue with spamd. This is Ian's comment: "The
3798 background is that sometimes spamd either never reads data from a
3799 connection it has accepted, or it never writes response data. The exiscan
3800 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3801 blindly assumes that writes won't block so it may never time out."
3803 PH/39 Allow G after quota size as well as K and M.
3805 PH/40 The value set for $authenticated_id in an authenticator may not contain
3806 binary zeroes or newlines because the value is written to log lines and
3807 to spool files. There was no check on this. Now the value is run through
3808 the string_printing() function so that such characters are converted to
3809 printable escape sequences.
3811 PH/41 $message_linecount is a new variable that contains the total number of
3812 lines in the message. Compare $body_linecount, which is the count for the
3815 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3816 bounce and delay warning messages, except in certain special cases, which
3819 (a) An SMTP error message from a remote host;
3820 (b) A message specified in a :fail: redirection;
3821 (c) A message specified in a "fail" command in a system filter;
3822 (d) A message specified in a FAIL return from the queryprogram router;
3823 (e) A message specified by the cannot_route_message router option.
3825 In these cases only, Exim does include the error details in bounce and
3826 warning messages. There are also a few cases where bland messages such
3827 as "unrouteable address" or "local delivery error" are given.
3829 PH/43 $value is now also set for the "else" part of a ${run expansion.
3831 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3832 being worked on, but at least Exim now implements the latest version to
3835 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3836 process to complete was treated as a delivery failure, a timeout while
3837 writing the message to the pipe was logged, but erroneously treated as a
3838 successful delivery. Such timeouts include transport filter timeouts. For
3839 consistency with the overall process timeout, these timeouts are now
3840 treated as errors, giving rise to delivery failures by default. However,
3841 there is now a new Boolean option for the pipe transport called
3842 timeout_defer, which, if set TRUE, converts the failures into defers for
3843 both kinds of timeout. A transport filter timeout is now identified in
3846 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3847 systems where "make" and "gmake" are different, calling "gmake" at top
3848 level broke things. I've arranged for the value of $(MAKE) to be passed
3849 from the Makefile to this script so that it can call the same version of
3853 A note about Exim versions 4.44 and 4.50
3854 ----------------------------------------
3856 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3857 changes of various kinds. As a consequence, a big documentation update was
3858 needed. This delayed the release for rather longer than seemed good, especially
3859 in the light of a couple of (minor) security issues. Therefore, the changes
3860 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3861 release. So 4.44 and 4.50 are in effect two different branches that both start
3864 I have left the 4.50 change log unchanged; it contains all the changes since
3865 4.43. The change log for 4.44 is below; many of its items are identical to
3866 those for 4.50. This seems to be the most sensible way to preserve the
3867 historical information.
3873 1. Minor wording change to the doc/README.SIEVE file.
3875 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3876 computation of the current number of files was incorrect.
3878 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3879 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3882 4. Give more explanation in the error message when the command for a transport
3883 filter fails to execute.
3885 5. There are several places where Exim runs a non-Exim command in a
3886 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3887 was being done only for the command run by the queryprogram router. It is
3888 now done for all such subprocesses. The other cases are: ${run, transport
3889 filters, and the commands run by the lmtp and pipe transports.
3891 6. Added CONFIGURE_GROUP build-time option.
3893 7. Some older OS have a limit of 256 on the maximum number of file
3894 descriptors. Exim was using setrlimit() to set 1000 as a large value
3895 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3896 systems. I've change it so that if it can't get 1000, it tries for 256.
3898 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3899 was an oversight, and furthermore, ever since the addition of extra
3900 controls (e.g. 4.43/32), the checks on when to allow different forms of
3901 "control" were broken. There should now be diagnostics for all cases when a
3902 control that does not make sense is encountered.
3904 9. Added the /retain_sender option to "control=submission".
3906 10. $recipients is now available in the predata ACL (oversight).
3908 11. Tidy the search cache before the fork to do a delivery from a message
3909 received from the command line. Otherwise the child will trigger a lookup
3910 failure and thereby defer the delivery if it tries to use (for example) a
3911 cached ldap connection that the parent has called unbind on.
3913 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3914 of $address_data from the recipient verification was clobbered by the
3915 sender verification.
3917 13. The value of address_data from a sender verification is now available in
3918 $sender_address_data in subsequent conditions in the ACL statement.
3920 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3922 15. Added a new option "connect=<time>" to callout options, to set a different
3925 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3926 was its contents. (It was OK if the option was not defined at all.)
3928 17. A "Completed" log line is now written for messages that are removed from
3929 the spool by the -Mrm option.
3931 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3932 information about exactly what failed.
3934 19. Added -dd to debug only the daemon process.
3936 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3937 handles timeouts, both on the server side and network timeouts. Renamed the
3938 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3940 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3941 It is now set to "smtps".
3943 22. $host_address is now set to the target address during the checking of
3944 ignore_target_hosts.
3946 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3947 being passed; this would have caused $sender_host_name to have been used if
3948 matching the list had actually called for a host name (not very likely,
3949 since this list is usually IP addresses). A host name is now passed as
3952 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3953 code that specifies a non-restarting handler (typically sigaction() in
3954 modern systems) in an attempt to fix a rare and obscure crash bug.
3956 25. Narrowed the window for a race in the daemon that could cause it to ignore
3957 SIGCHLD signals. This is not a major problem, because they are used only to
3958 wake it up if nothing else does.
3960 26. A malformed maildirsize file could cause Exim to calculate negative values
3961 for the mailbox size or file count. Odd effects could occur as a result.
3962 The maildirsize information is now recalculated if the size or filecount
3965 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3966 support for a long time. Removed HAVE_SYS_VFS_H.
3968 28. Installed the latest version of exipick from John Jetmore.
3970 29. In an address list, if the pattern was not a regular expression, an empty
3971 subject address (from a bounce message) matched only if the pattern was an
3972 empty string. Non-empty patterns were not even tested. This was the wrong
3973 because it is perfectly reasonable to use an empty address as part of a
3974 database query. An empty address is now tested by patterns that are
3975 lookups. However, all the other forms of pattern expect the subject to
3976 contain a local part and a domain, and therefore, for them, an empty
3977 address still always fails if the pattern is not itself empty.
3979 30. Exim went into a mad DNS loop when attempting to do a callout where the
3980 host was specified on an smtp transport, and looking it up yielded more
3981 than one IP address.
3983 31. Re-factored the code for checking spool and log partition space into a
3984 function that finds that data and another that does the check. The former
3985 is then used to implement four new variables: $spool_space, $log_space,
3986 $spool_inodes, and $log_inodes.
3988 32. The RFC2047 encoding function was originally intended for short strings
3989 such as real names; it was not keeping to the 75-character limit for
3990 encoded words that the RFC imposes. It now respects the limit, and
3991 generates multiple encoded words if necessary. To be on the safe side, I
3992 have increased the buffer size for the ${rfc2047: expansion operator from
3995 33. It is now permitted to omit both strings after an "if" condition; if the
3996 condition is true, the result is "true". As before, when the second string
3997 is omitted, a false condition yields an empty string. This makes it less
3998 cumbersome to write custom ACL and router conditions.
4000 34. Failure to deliver a bounce message always caused it to be frozen, even if
4001 there was an errors_to setting on the router. The errors_to setting is now
4004 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4005 canonical form (fully expanded) before being placed in
4006 $sender_host_address.
4008 36. The table in the code that translates DNS record types into text (T_A to
4009 "A" for instance) was missing entries for NS and CNAME. It is just possible
4010 that this could have caused confusion if both these types were looked up
4011 for the same domain, because the text type is used as part of Exim's
4012 per-process caching. But the chance of anyone hitting this buglet seems
4015 37. The dnsdb lookup has been extended in a number of ways.
4017 (1) There is a new type, "zns", which walks up the domain tree until it
4018 finds some nameserver records. It should be used with care.
4020 (2) There is a new type, "mxh", which is like "mx" except that it returns
4021 just the host names, not the priorities.
4023 (3) It is now possible to give a list of domains (or IP addresses) to be
4024 looked up. The behaviour when one of the lookups defers can be
4025 controlled by a keyword.
4027 (4) It is now possible to specify the separator character for use when
4028 multiple records are returned.
4030 38. The dnslists ACL condition has been extended: it is now possible to supply
4031 a list of IP addresses and/or domains to be looked up in a particular DNS
4034 39. Added log_selector=+queue_time_overall.
4036 40. When running the queue in the test harness, wait just a tad after forking a
4037 delivery process, to get repeatability of debugging output.
4039 41. Include certificate and key file names in error message when GnuTLS fails
4040 to set them up, because the GnuTLS error message doesn't include the name
4041 of the failing file when there is a problem reading it.
4043 42. Allow both -bf and -bF in the same test run.
4045 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4047 44. The "Exiscan patch" is now merged into the mainline Exim source.
4049 45. Sometimes the final signoff response after QUIT could fail to get
4050 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4051 < 0 before doing a fflush(). This bug looks as though it goes back to the
4052 introduction of TLS in release 3.20, but "sometimes" must have been rare
4053 because the tests only now provoked it.
4055 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4056 (this can affect the format of dates).
4058 47. exim_tidydb, when checking for the continued existence of a message for
4059 which it has found a message-specific retry record, was not finding
4060 messages that were in split spool directories. Consequently, it was
4061 deleting retry records that should have stayed in existence.
4063 48. Steve fixed some bugs in eximstats.
4065 49. The SPA authentication driver was not abandoning authentication and moving
4066 on to the next authenticator when an expansion was forced to fail,
4067 contradicting the general specification for all authenticators. Instead it
4068 was generating a temporary error. It now behaves as specified.
4070 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4071 (the order specifies the preference for clients). The order is now AES256,
4072 AES128, 3DES, ARCFOUR128.
4074 51. Small patch to Sieve code - explicitly set From: when generating an
4077 52. Exim crashed if a remote delivery caused a very long error message to be
4078 recorded - for instance if somebody sent an entire SpamAssassin report back
4079 as a large number of 550 error lines. This bug was coincidentally fixed by
4080 increasing the size of one of Exim's internal buffers (big_buffer) that
4081 happened as part of the Exiscan merge. However, to be on the safe side, I
4082 have made the code more robust (and fixed the comments that describe what
4085 53. Now that there can be additional text after "Completed" in log lines (if
4086 the queue_time_overall log selector is set), a one-byte patch to exigrep
4087 was needed to allow it to recognize "Completed" as not the last thing in
4090 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4091 patch that reportedly fixes this has been added. I am not expert enough to
4092 create a test for it. This is what the patch creator wrote:
4094 "I found a little strange behaviour of ldap code when working with
4095 Windows 2003 AD Domain, where users was placed in more than one
4096 Organization Units. When I tried to give exim partial DN, the exit code
4097 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4098 But simultaneously result of request was absolutely normal ldap result,
4099 so I produce this patch..."
4101 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4102 REFERENCE, so I have modified the code to exclude the patch when that macro
4105 55. Some experimental protocols are using DNS PTR records for new purposes. The
4106 keys for these records are domain names, not reversed IP addresses. The
4107 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4108 leaves it alone. Component reversal etc. now happens only for IP addresses.
4111 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4113 57. Double the size of the debug message buffer (to 2048) so that more of very
4114 long debug lines gets shown.
4116 58. The exicyclog utility now does better if the number of log files to keep
4117 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4119 59. Two changes related to the smtp_active_hostname option:
4121 (1) $smtp_active_hostname is now available as a variable.
4122 (2) The default for smtp_banner uses $smtp_active_hostname instead
4123 of $primary_hostname.
4125 60. The host_aton() function is supposed to be passed a string that is known
4126 to be a valid IP address. However, in the case of IPv6 addresses, it was
4127 not checking this. This is a hostage to fortune. Exim now panics and dies
4128 if the condition is not met. A case was found where this could be provoked
4129 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4130 components; fortuitously, this particular loophole had already been fixed
4131 by change 4.50/55 above.
4133 If there are any other similar loopholes, the new check in host_aton()
4134 itself should stop them being exploited. The report I received stated that
4135 data on the command line could provoke the exploit when Exim was running as
4136 exim, but did not say which command line option was involved. All I could
4137 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4138 running as the user.
4141 61. There was a buffer overflow vulnerability in the SPA authentication code
4142 (which came originally from the Samba project). I have added a test to the
4143 spa_base64_to_bits() function which I hope fixes it.
4146 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4147 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4149 63. The daemon start-up calls getloadavg() while still root for those OS that
4150 need the first call to be done as root, but it missed one case: when
4151 deliver_queue_load_max is set with deliver_drop_privilege. This is
4152 necessary for the benefit of the queue runner, because there is no re-exec
4153 when deliver_drop_privilege is set.
4155 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4156 This has been fixed.
4158 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4159 was in use, was not putting the data itself into the right store pool;
4160 consequently, it could be overwritten for a subsequent message in the same
4161 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4164 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4166 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4167 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4168 the function were treating the return as a boolean value, which happened to
4169 work because 0=false and not-0=true, but is not correct code.
4171 68. The host_aton() function was not handling scoped IPv6 addresses (those
4172 with, for example, "%eth0" on the end) correctly.
4174 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4175 negated items (that is, ~something) in unsigned ints. Some compilers
4176 apparently mutter when there is no cast.
4178 70. If an address verification called from an ACL failed, and did not produce a
4179 user-specific message (i.e. there was only a "system" message), nothing was
4180 put in $acl_verify_message. In this situation, it now puts the system
4183 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4184 session; change 4.31/43 added the unwanted input to the log line - except
4185 that it did not do this in the start of session case. It now does.
4187 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4188 This is wrong and can cause the other end to generate a synchronization
4189 error if it is another Exim or anything else that does the synchronization
4190 check. A QUIT command is no longer sent after a timeout.
4192 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4193 during host lookups.
4195 74. The defer_ok option of callout verification was not working if it was used
4196 when verifying addresses in header lines, that is, for this case:
4198 verify = header_sender/callout=defer_ok
4200 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4201 those file descriptors could be used for SMTP connections. If anything
4202 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4203 could be sent to the SMTP client, causing chaos. The daemon now opens
4204 stdin, stdout, and stderr to /dev/null when it puts itself into the
4207 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4208 log by default. The user can override this with suitable Perl magic.
4210 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4211 the log message when discard triggers, was not working for the DATA ACL or
4212 for the non-SMTP ACL.
4214 78. Error message wording change in sieve.c.
4216 79. If smtp_accept_max_per_host was set, the number of connections could be
4217 restricted to fewer than expected, because the daemon was trying to set up
4218 a new connection before checking whether the processes handling previous
4219 connections had finished. The check for completed processes is now done
4220 earlier. On busy systems, this bug wouldn't be noticed because something
4221 else would have woken the daemon, and it would have reaped the completed
4224 80. If a message was submitted locally by a user whose login name contained one
4225 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4226 It caused a spool format error. I have fixed the spool reading code. A
4227 related problem was that the "from" clause in the Received: line became
4228 illegal because of the space(s). It is now covered by ${quote_local_part.
4230 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4233 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4235 83. Patch from Sieve maintainer to fix unterminated string problem in
4236 "vacation" handling.
4238 84. Some minor changes to the Linux configuration files to help with other
4239 OS variants using glibc.
4241 85. One more patch for Sieve to update vacation handling to latest spec.
4244 ----------------------------------------------------
4245 See the note above about the 4.44 and 4.50 releases.
4246 ----------------------------------------------------
4252 1. Change 4.43/35 introduced a bug that caused file counts to be
4253 incorrectly computed when quota_filecount was set in an appendfile
4256 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4257 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4260 3. Give more explanation in the error message when the command for a transport
4261 filter fails to execute.
4263 4. There are several places where Exim runs a non-Exim command in a
4264 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4265 was being done only for the command run by the queryprogram router. It is
4266 now done for all such subprocesses. The other cases are: ${run, transport
4267 filters, and the commands run by the lmtp and pipe transports.
4269 5. Some older OS have a limit of 256 on the maximum number of file
4270 descriptors. Exim was using setrlimit() to set 1000 as a large value
4271 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4272 systems. I've change it so that if it can't get 1000, it tries for 256.
4274 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4275 was an oversight, and furthermore, ever since the addition of extra
4276 controls (e.g. 4.43/32), the checks on when to allow different forms of
4277 "control" were broken. There should now be diagnostics for all cases when a
4278 control that does not make sense is encountered.
4280 7. $recipients is now available in the predata ACL (oversight).
4282 8. Tidy the search cache before the fork to do a delivery from a message
4283 received from the command line. Otherwise the child will trigger a lookup
4284 failure and thereby defer the delivery if it tries to use (for example) a
4285 cached ldap connection that the parent has called unbind on.
4287 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4288 of $address_data from the recipient verification was clobbered by the
4289 sender verification.
4291 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4292 was its contents. (It was OK if the option was not defined at all.)
4294 11. A "Completed" log line is now written for messages that are removed from
4295 the spool by the -Mrm option.
4297 12. $host_address is now set to the target address during the checking of
4298 ignore_target_hosts.
4300 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4301 being passed; this would have caused $sender_host_name to have been used if
4302 matching the list had actually called for a host name (not very likely,
4303 since this list is usually IP addresses). A host name is now passed as
4306 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4307 code that specifies a non-restarting handler (typically sigaction() in
4308 modern systems) in an attempt to fix a rare and obscure crash bug.
4310 15. Narrowed the window for a race in the daemon that could cause it to ignore
4311 SIGCHLD signals. This is not a major problem, because they are used only to
4312 wake it up if nothing else does.
4314 16. A malformed maildirsize file could cause Exim to calculate negative values
4315 for the mailbox size or file count. Odd effects could occur as a result.
4316 The maildirsize information is now recalculated if the size or filecount
4319 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4320 support for a long time. Removed HAVE_SYS_VFS_H.
4322 18. Updated exipick to current release from John Jetmore.
4324 19. Allow an empty sender to be matched against a lookup in an address list.
4325 Previously the only cases considered were a regular expression, or an
4328 20. Exim went into a mad DNS lookup loop when doing a callout where the
4329 host was specified on the transport, if the DNS lookup yielded more than
4332 21. The RFC2047 encoding function was originally intended for short strings
4333 such as real names; it was not keeping to the 75-character limit for
4334 encoded words that the RFC imposes. It now respects the limit, and
4335 generates multiple encoded words if necessary. To be on the safe side, I
4336 have increased the buffer size for the ${rfc2047: expansion operator from
4339 22. Failure to deliver a bounce message always caused it to be frozen, even if
4340 there was an errors_to setting on the router. The errors_to setting is now
4343 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4344 canonical form (fully expanded) before being placed in
4345 $sender_host_address.
4347 24. Updated eximstats to version 1.33
4349 25. Include certificate and key file names in error message when GnuTLS fails
4350 to set them up, because the GnuTLS error message doesn't include the name
4351 of the failing file when there is a problem reading it.
4353 26. Expand error message when OpenSSL has problems setting up cert/key files.
4356 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4357 (this can affect the format of dates).
4359 28. exim_tidydb, when checking for the continued existence of a message for
4360 which it has found a message-specific retry record, was not finding
4361 messages that were in split spool directories. Consequently, it was
4362 deleting retry records that should have stayed in existence.
4364 29. eximstats updated to version 1.35
4365 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4366 1.35 - bugfix such that pie charts by volume are generated correctly
4368 30. The SPA authentication driver was not abandoning authentication and moving
4369 on to the next authenticator when an expansion was forced to fail,
4370 contradicting the general specification for all authenticators. Instead it
4371 was generating a temporary error. It now behaves as specified.
4373 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4374 (the order specifies the preference for clients). The order is now AES256,
4375 AES128, 3DES, ARCFOUR128.
4377 31. Small patch to Sieve code - explicitly set From: when generating an
4380 32. Exim crashed if a remote delivery caused a very long error message to be
4381 recorded - for instance if somebody sent an entire SpamAssassin report back
4382 as a large number of 550 error lines. This bug was coincidentally fixed by
4383 increasing the size of one of Exim's internal buffers (big_buffer) that
4384 happened as part of the Exiscan merge. However, to be on the safe side, I
4385 have made the code more robust (and fixed the comments that describe what
4388 33. Some experimental protocols are using DNS PTR records for new purposes. The
4389 keys for these records are domain names, not reversed IP addresses. The
4390 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4391 leaves it alone. Component reversal etc. now happens only for IP addresses.
4394 34. The host_aton() function is supposed to be passed a string that is known
4395 to be a valid IP address. However, in the case of IPv6 addresses, it was
4396 not checking this. This is a hostage to fortune. Exim now panics and dies
4397 if the condition is not met. A case was found where this could be provoked
4398 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4399 components; fortuitously, this particular loophole had already been fixed
4400 by change 4.50/55 or 4.44/33 above.
4402 If there are any other similar loopholes, the new check in host_aton()
4403 itself should stop them being exploited. The report I received stated that
4404 data on the command line could provoke the exploit when Exim was running as
4405 exim, but did not say which command line option was involved. All I could
4406 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4407 running as the user.
4410 35. There was a buffer overflow vulnerability in the SPA authentication code
4411 (which came originally from the Samba project). I have added a test to the
4412 spa_base64_to_bits() function which I hope fixes it.
4415 36. The daemon start-up calls getloadavg() while still root for those OS that
4416 need the first call to be done as root, but it missed one case: when
4417 deliver_queue_load_max is set with deliver_drop_privilege. This is
4418 necessary for the benefit of the queue runner, because there is no re-exec
4419 when deliver_drop_privilege is set.
4421 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4422 was in use, was not putting the data itself into the right store pool;
4423 consequently, it could be overwritten for a subsequent message in the same
4424 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4427 38. Sometimes the final signoff response after QUIT could fail to get
4428 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4429 < 0 before doing a fflush(). This bug looks as though it goes back to the
4430 introduction of TLS in release 3.20, but "sometimes" must have been rare
4431 because the tests only now provoked it.
4437 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4438 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4439 Now it returns an integer. A number of calls were still expecting a T/F
4440 return. Fortuitously, in all cases, the tests worked in OK situations,
4441 which is the norm. However, things would have gone wrong on any write
4442 failures on the smtp file descriptor. This function is used when sending
4443 messages over SMTP and also when doing verify callouts.
4445 2. When Exim is called to do synchronous delivery of a locally submitted
4446 message (the -odf or -odi options), it no longer closes stderr before doing
4449 3. Implemented the mua_wrapper option.
4451 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4453 5. Implemented the functions header_remove(), header_testname(),
4454 header_add_at_position(), and receive_remove_recipient(), and exported them
4457 6. If an ACL "warn" statement specified the addition of headers, Exim already
4458 inserted X-ACL-Warn: at the start if there was no header name. However, it
4459 was not making this test for the second and subsequent header lines if
4460 there were newlines in the string. This meant that an invalid header could
4461 be inserted if Exim was badly configured.
4463 7. Allow an ACL "warn" statement to add header lines at the start or after all
4464 the Received: headers, as well as at the end.
4466 8. Added the rcpt_4xx retry error code.
4468 9. Added postmaster_mailfrom=xxx to callout verification option.
4470 10. Added mailfrom=xxxx to the callout verification option, for verify=
4473 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4474 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4476 12. Inserted some casts to stop certain compilers warning when using pointer
4477 differences as field lengths or precisions in printf-type calls (mostly
4478 affecting debugging statements).
4480 13. Added optional readline() support for -be (dynamically loaded).
4482 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4483 same clock tick as a message's arrival, so that its received time was the
4484 same as the "first fail" time on the retry record, and that message
4485 remained on the queue past the ultimate address timeout, every queue runner
4486 would try a delivery (because it was past the ultimate address timeout) but
4487 after another failure, the ultimate address timeout, which should have then
4488 bounced the address, did not kick in. This was a "< instead of <=" error;
4489 in most cases the first failure would have been in the next clock tick
4490 after the received time, and all would be well.
4492 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4493 being recognized when the domain list was tested by the match_domain
4494 condition in an expansion string.
4496 16. Added the ${str2b64: operator.
4498 17. Exim was always calling setrlimit() to set a large limit for the number of
4499 processes, without checking whether the existing limit was already
4500 adequate. (It did check for the limit on file descriptors.) Furthermore,
4501 errors from getrlimit() and setrlimit() were being ignored. Now they are
4502 logged to the main and panic logs, but Exim does carry on, to try to do its
4503 job under whatever limits there are.
4505 18. Imported PCRE 5.0.
4507 19. Trivial typo in log message " temporarily refused connection" (the leading
4510 20. If the log selector return_path_on_delivery was set and an address was
4511 redirected to /dev/null, the delivery process crashed because it assumed
4512 that a return path would always be set for a "successful" delivery. In this
4513 case, the whole delivery is bypassed as an optimization, and therefore no
4516 21. Internal re-arrangement: the function for sending a challenge and reading
4517 a response while authentication was assuming a zero-terminated challenge
4518 string. It's now changed to take a pointer and a length, to allow for
4519 binary data in such strings.
4521 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4523 23. Exim was not respecting finduser_retries when seeking the login of the
4524 uid under which it was called; it was always trying 10 times. (The default
4525 setting of finduser_retries is zero.) Also, it was sleeping after the final
4526 failure, which is pointless.
4528 24. Implemented tls_on_connect_ports.
4530 25. Implemented acl_smtp_predata.
4532 26. If the domain in control=submission is set empty, Exim assumes that the
4533 authenticated id is a complete email address when it generates From: or
4534 Sender: header lines.
4536 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4537 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4538 chown and chgrp in /bin and hostname in /usr/bin.
4540 28. Exim was keeping the "process log" file open after each use, just as it
4541 does for the main log. This opens the possibility of it remaining open for
4542 long periods when the USR1 signal hits a daemon. Occasional processlog
4543 errors were reported, that could have been caused by this. Anyway, it seems
4544 much more sensible not to leave this file open at all, so that is what now
4547 29. The long-running daemon process does not normally write to the log once it
4548 has entered its main loop, and it closes the log before doing so. This is
4549 so that log files can straightforwardly be renamed and moved. However,
4550 there are a couple of unusual error situations where the daemon does write
4551 log entries, and I had neglected to close the log afterwards.
4553 30. The text of an SMTP error response that was received during a remote
4554 delivery was being truncated at 512 bytes. This is too short for some of
4555 the long messages that one sometimes sees. I've increased the limit to
4558 31. It is now possible to make retry rules that apply only when a message has a
4559 specific sender, in particular, an empty sender.
4561 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4562 it possible to be selective about when SMTP synchronization is enforced.
4564 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4566 32. Implemented hosts_connection_nolog.
4568 33. Added an ACL for QUIT.
4570 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4573 35. Added mailbox_size and mailbox_filecount to appendfile.
4575 36. Added control = no_multiline_responses to ACLs.
4577 37. There was a bug in the logic of the code that waits for the clock to tick
4578 in the case where the clock went backwards by a substantial amount such
4579 that the microsecond fraction of "now" was more than the microsecond
4580 fraction of "then" (but the whole seconds number was less).
4582 38. Added support for the libradius Radius client library this is found on
4583 FreeBSD (previously only the radiusclient library was supported).
4589 1. When certain lookups returned multiple values in the form name=value, the
4590 quoting of the values was not always being done properly. Specifically:
4591 (a) If the value started with a double quote, but contained no whitespace,
4593 (b) If the value contained whitespace other than a space character (i.e.
4594 tabs or newlines or carriage returns) it was not quoted.
4595 This fix has been applied to the mysql and pgsql lookups by writing a
4596 separate quoting function and calling it from the lookup code. The fix
4597 should probably also be applied to nisplus, ibase and oracle lookups, but
4598 since I cannot test any of those, I have not disturbed their existing code.
4600 2. A hit in the callout cache for a specific address caused a log line with no
4601 reason for rejecting RCPT. Now it says "Previous (cached) callout
4602 verification failure".
4604 3. There was an off-by-one bug in the queryprogram router. An over-long
4605 return line was truncated at 256 instead of 255 characters, thereby
4606 overflowing its buffer with the terminating zero. As well as fixing this, I
4607 have increased the buffer size to 1024 (and made a note to document this).
4609 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4610 when Exim is waiting for an SMTP response from a remote server, Exim
4611 restarts its select() call on the socket, thereby resetting its timeout.
4612 This is not a problem when such interrupts are rare. Somebody set up a cron
4613 job to run exiwhat every 2 minutes, which is less than the normal select()
4614 timeout (5 or 10 minutes). This meant that the select() timeout never
4615 kicked in because it was always reset. I have fixed this by comparing the
4616 time when an interrupt arrives with the time at the start of the first call
4617 to select(). If more time than the timeout has elapsed, the interrupt is
4618 treated as a timeout.
4620 5. Some internal re-factoring in preparation for the addition of Sieve
4621 extensions (by MH). In particular, the "personal" test is moved to a
4622 separate function, and given an option for scanning Cc: and Bcc: (which is
4623 not set for Exim filters).
4625 6. When Exim created an email address using the login of the caller as the
4626 local part (e.g. when creating a From: or Sender: header line), it was not
4627 quoting the local part when it contained special characters such as @.
4629 7. Installed new OpenBSD configuration files.
4631 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4632 try to make them clearer.
4634 9. Callout options, other than the timeout value, were being ignored when
4635 verifying sender addresses in header lines. For example, when using
4637 verify = header_sender/callout=no_cache
4639 the cache was (incorrectly) being used.
4641 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4642 only the Cygwin environment.
4644 11. When return_path_on_delivery was set as a log selector, if different remote
4645 addresses in the same message used different return paths and parallel
4646 remote delivery occurred, the wrong values would sometimes be logged.
4647 (Whenever a remote delivery process finished, the return path value from
4648 the most recently started remote delivery process was logged.)
4650 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4651 lines when AUTH and/or TLS are in use. This is the "received protocol"
4652 field. Exim used to use "asmtp" for authenticated SMTP, without any
4653 indication (in the protocol name) for TLS use. Now it follows the RFC and
4654 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4655 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4656 names appear in log lines as well as in Received: header lines.
4658 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4659 extensions, and comparison tests, and to fix some bugs.
4661 14. Changes to the "personal" filter test:
4663 (1) The test was buggy in that it was just doing the equivalent of
4664 "contains" tests on header lines. For example, if a user's address was
4665 anne@some.where, the "personal" test would incorrectly be true for
4667 To: susanne@some.where
4669 This test is now done by extracting each address from the header in turn,
4670 and checking the entire address. Other tests that are part of "personal"
4671 are now done using regular expressions (for example, to check local parts
4672 of addresses in From: header lines).
4674 (2) The list of non-personal local parts in From: addresses has been
4675 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4676 taken from the Sieve specification recommendations.
4678 (3) If the message contains any header line starting with "List-" it is
4679 treated as non-personal.
4681 (4) The test for "circular" in the Subject: header line has been removed
4682 because it now seems ill-conceived.
4684 15. Minor typos in src/EDITME comments corrected.
4686 16. Installed latest exipick from John Jetmore.
4688 17. If headers_add on a router specified a text string that was too long for
4689 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4690 of string_sprintf() is now avoided.
4692 18. $message_body_size was not set (it was always zero) when running the DATA
4693 ACL and the local_scan() function.
4695 19. For the "mail" command in an Exim filter, no default was being set for
4696 the once_repeat time, causing a random time value to be used if "once" was
4697 specified. (If the value happened to be <= 0, no repeat happened.) The
4698 default is now 0s, meaning "never repeat". The "vacation" command was OK
4699 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4700 (I found it when inspecting the code).
4702 20. There is now an overall timeout for performing a callout verification. It
4703 defaults to 4 times the callout timeout, which applies to individual SMTP
4704 commands during the callout. The overall timeout applies when there is more
4705 than one host that can be tried. The timeout is checked before trying the
4706 next host. This prevents very long delays if there are a large number of
4707 hosts and all are timing out (e.g. when the network connections are timing
4708 out). The value of the overall timeout can be changed by specifying an
4709 additional sub-option for "callout", called "maxwait". For example:
4711 verify = sender/callout=5s,maxwait=20s
4713 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4714 to the end before writing, but this should make it even safer).
4716 22. Exim was forgetting that it had advertised PIPELINING for the second and
4717 subsequent messages on an SMTP connection. It was also not resetting its
4718 memory on STARTTLS and an internal HELO.
4720 23. When Exim logs an SMTP synchronization error within a session, it now
4721 records whether PIPELINING has been advertised or not.
4723 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4724 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4725 rather than long int.
4727 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4729 26. Added the never_mail option to autoreply.
4735 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4736 crash if the getsockname() call failed; this can happen if a connection is
4737 closed very soon after it is established. The problem was simply in the
4738 order in which certain operations were done, causing Exim to try to write
4739 to the SMTP stream before it had set up the file descriptor. The bug has
4740 been fixed by making things happen in the correct order.
4746 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4747 before the connection was closed, thus losing the rejection response.
4749 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4750 some early Solaris releases, but causes trouble in current releases where
4751 socklen_t is defined.
4753 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4756 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4759 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4760 that releasing the top of it at the end releases what was used for sub-
4761 expansions (unless the block got too big). However, discard this block if
4762 the first thing is a variable or header, so that we can use its block when
4763 it is dynamic (useful for very large $message_headers, for example).
4765 6. Lookups now cache *every* query, not just the most recent. A new, separate
4766 store pool is used for this. It can be recovered when all lookup caches are
4767 flushed. Lookups now release memory at the end of their result strings.
4768 This has involved some general refactoring of the lookup sources.
4770 7. Some code has been added to the store_xxx() functions to reduce the amount
4771 of flapping under certain conditions.
4773 8. log_incoming_interface used to affect only the <= reception log lines. Now
4774 it causes the local interface and port to be added to several more SMTP log
4775 lines, for example "SMTP connection from", and rejection lines.
4777 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4779 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4781 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4782 could be overwritten at the end of the current message (or the start of a
4783 new message if it was set in a HELO ACL). The value is now preserved for
4784 the duration of the SMTP connection.
4786 12. If a transport had a headers_rewrite setting, and a matching header line
4787 contained an unqualified address, that address was qualified, even if it
4788 did not match any rewriting rules. The underlying bug was that the values
4789 of the flags that permit the existence of unqualified sender and recipient
4790 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4791 non-local messages, and by -bnq for local messages) were not being
4792 preserved with the message after it was received.
4794 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4795 "next input=" as part of the text comprising the host identity instead of
4796 the correct text. The code was using the same buffer for two different
4797 strings. However, depending on which order the printing function evaluated
4798 its arguments, the bug did not always show up. Under Linux, for example, my
4799 test suite worked just fine.
4801 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4802 doesn't work with some older versions of Perl. It has been changed to "my",
4803 which in any case is probably the better facility to use.
4805 15. A really picky compiler found some instances of statements for creating
4806 error messages that either had too many or two few arguments for the format
4809 16. The size of the buffer for calls to the DNS resolver has been increased
4810 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4811 for addresses that have a lot of PTR records. This alleviates a problem; it
4812 does not fully solve it.
4814 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4815 buffer now truncates the list and logs the incident, which is the same
4816 action as happens when Exim is looking up a host name and its aliases.
4817 Previously in this situation something unpredictable would happen;
4818 sometimes it was "internal error: store_reset failed".
4820 18. If a server dropped the connection unexpectedly when an Exim client was
4821 using GnuTLS and trying to read a response, the client delivery process
4822 crashed while trying to generate an error log message.
4824 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4825 string, for example:
4827 warn message = H1: something\nH2: something
4829 the text was added as a single header line from Exim's point of view
4830 though it ended up OK in the delivered message. However, searching for the
4831 second and subsequent header lines using $h_h2: did not work. This has been
4832 fixed. Similarly, if a system filter added multiple headers in this way,
4833 the routers could not see them.
4835 20. Expanded the error message when iplsearch is called with an invalid key to
4836 suggest using net-iplsearch in a host list.
4838 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4839 ACLs are no longer actually imposed (and a message to that effect is
4842 22. If a "gecos" field in a passwd entry contained escaped characters, in
4843 particular, if it contained a \" sequence, Exim got it wrong when building
4844 a From: or a Sender: header from that name. A second bug also caused
4845 incorrect handling when an unquoted " was present following a character
4846 that needed quoting.
4848 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4849 was not being matched caselessly.
4851 24. Arranged for all hyphens in the exim.8 source to be escaped with
4854 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4855 properly. Recipient callout cache records were still being keyed to include
4856 the sender, even when use_sender was set false. This led to far more
4857 callouts that were necessary. The sender is no longer included in the key
4858 when use_sender is false.
4860 26. Added "control = submission" modifier to ACLs.
4862 27. Added the ${base62d: operator to decode base 62 numbers.
4864 28. dnsdb lookups can now access SRV records.
4866 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4867 the configuration file.
4869 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4870 (-v) mode. This makes the output for a verbose queue run more intelligible.
4872 31. Added a use_postmaster feature to recipient callouts.
4874 32. Added the $body_zerocount variable, containing the number of binary zero
4875 bytes in the message body.
4877 33. The time of last modification of the "new" subdirectory is now used as the
4878 "mailbox time last read" when there is a quota error for a maildir
4881 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4883 35. Added +ignore_unknown as a special item in host lists.
4885 36. Code for decoding IPv6 addresses in host lists is now included, even if
4886 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4887 address was recognized as an IP address, but was then not correctly decoded
4888 into binary, causing unexpected and incorrect effects when compared with
4895 1. Very minor rewording of debugging text in manualroute to say "list of
4896 hosts" instead of "hostlist".
4898 2. If verify=header_syntax was set, and a header line with an unqualified
4899 address (no domain) and a large number of spaces between the end of the
4900 name and the colon was received, the reception process suffered a buffer
4901 overflow, and (when I tested it) crashed. This was caused by some obsolete
4902 code that should have been removed. The fix is to remove it!
4904 3. When running in the test harness, delay a bit after writing a bounce
4905 message to get a bit more predictability in the log output.
4907 4. Added a call to search_tidyup() just before forking a reception process. In
4908 theory, someone could use a lookup in the expansion of smtp_accept_max_
4909 per_host which, without the tidyup, could leave open a database connection.
4911 5. Added the variables $recipient_data and $sender_data which get set from a
4912 lookup success in an ACL "recipients" or "senders" condition, or a router
4913 "senders" option, similar to $domain_data and $local_part_data.
4915 6. Moved the writing of debug_print from before to after the "senders" test
4918 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4919 problems for message scanning, either using a data ACL, or using
4920 local_scan() because the Received: header was not generated till after they
4921 were called (in order to set the time as the time of reception completion).
4922 I have revised the way this works. The header is now generated after the
4923 body is received, but before the ACL or local_scan() are called. After they
4924 are run, the timestamp in the header is updated.
4930 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4931 before starting a queue runner without re-exec. This happened only when
4932 deliver_drop_privilege was set or when the Exim user was set to root. The
4933 effect of the bug was that timeouts during subsequent deliveries caused
4934 crashes instead of being properly handled. The handler is now left at its
4935 default (and expected) setting.
4937 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4938 message, again when deliver_drop_privilege is set or Exim is run as root.
4939 The bug described in (1) was not present in this case, but the tidying up
4940 of the other signals was missing. I have made the two cases consistent.
4942 3. The ignore_target_hosts setting on a manualroute router was being ignored
4943 for hosts that were looked up using the /MX notation.
4945 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4948 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4949 operated on the sender address. After changing the $sender_address to <>
4950 for the sender address verify, Exim was re-instated it as the original
4951 (before rewriting) address, but remembering that it had rewritten it, so it
4952 wasn't rewriting it again. This bug also had the effect of breaking the
4953 sender address verification caching when the sender address was rewritten.
4955 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4956 This has been changed so that if the ip literal address matches
4957 ignore_target_hosts, the router declines.
4959 7. Added expansion conditions match_domain, match_address, and match_local_
4960 part (NOT match_host).
4962 8. The placeholder for the Received: header didn't have a length field set.
4964 9. Added code to Exim itself and to exim_lock to test for a specific race
4965 condition that could lead to file corruption when using MBX delivery. The
4966 issue is with the lockfile that is created in /tmp. If this file is removed
4967 after a process has opened it but before that process has acquired a lock,
4968 there is the potential for a second process to recreate the file and also
4969 acquire a lock. This could lead to two Exim processes writing to the file
4970 at the same time. The added code performs the same test as UW imapd; it
4971 checks after acquiring the lock that its file descriptor still refers to
4972 the same named file.
4974 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4975 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4978 11. Added the smtp_active_hostname option. If used, this will typically be made
4979 to depend on the incoming interface address. Because $interface_address is
4980 not set up until the daemon has forked a reception process, error responses
4981 that can happen earlier (such as "too many connections") no longer contain
4984 12. If an expansion in a condition on a "warn" statement fails because a lookup
4985 defers, the "warn" statement is abandoned, and the next ACL statement is
4986 processed. Previously this caused the whole ACL to be aborted.
4988 13. Added the iplsearch lookup type.
4990 14. Added ident_timeout as a log selector.
4992 15. Added tls_certificate_verified as a log selector.
4994 16. Added a global option tls_require_ciphers (compare the smtp transport
4995 option of the same name). This controls incoming TLS connections.
4997 17. I finally figured out how to make tls_require_ciphers do a similar thing
4998 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4999 before starting the TLS session.
5001 18. Tabs are now shown as \t in -bP output.
5003 19. If the log selector return_path_on_delivery was set, Exim crashed when
5004 bouncing a message because it had too many Received: header lines.
5006 20. If two routers both had headers_remove settings, and the first one included
5007 a superfluous trailing colon, the final name in the first list and the
5008 first name in the second list were incorrectly joined into one item (with a
5009 colon in the middle).
5015 1. Added -C and -D options to the exinext utility, mainly to make it easier
5016 to include in the automated testing, but these could be helpful when
5017 multiple configurations are in use.
5019 2. The exinext utility was not formatting the output nicely when there was
5020 an alternate port involved in the retry record key, nor when there was a
5021 message id as well (for retries that were specific to a specific message
5022 and a specific host). It was also confused by IPv6 addresses, because of
5023 the additional colons they contain. I have fixed the IPv4 problem, and
5024 patched it up to do a reasonable job for IPv6.
5026 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5027 delivery, the log line now contains "pipelined" if PIPELINING was used.
5029 4. An SMTP transport process used to panic and die if the bind() call to set
5030 an explicit outgoing interface failed. This has been changed; it is now
5031 treated in the same way as a connect() failure.
5033 5. A reference to $sender_host_name in the part of a conditional expansion
5034 that was being skipped was still causing a DNS lookup. This no longer
5037 6. The def: expansion condition was not recognizing references to header lines
5038 that used bh_ and bheader_.
5040 7. Added the _cache feature to named lists.
5042 8. The code for checking quota_filecount in the appendfile transport was
5043 allowing one more file than it should have been.
5045 9. For compatibility with Sendmail, the command line option
5053 and sets the incoming protocol and host name (for trusted callers). The
5054 host name and its colon can be omitted when only the protocol is to be set.
5055 Note the Exim already has two private options, -pd and -ps, that refer to
5056 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5057 "s", but I don't think that's a major issue.
5059 10. A number of refactoring changes to the code, none of which should affect
5062 (a) The number of logging options was getting close to filling up the
5063 32-bit word that was used as a bit map. I have split them into two classes:
5064 those that are passed in the argument to log_write(), and those that are
5065 only ever tested independently outside of that function. These are now in
5066 separate 32-bit words, so there is plenty of room for expansion again.
5067 There is no change in the user interface or the logging behaviour.
5069 (b) When building, for example, log lines, the code previously used a
5070 macro that called string_cat() twice, in order to add two strings. This is
5071 not really sufficiently general. Furthermore, there was one instance where
5072 it was actually wrong because one of the argument was used twice, and in
5073 one call a function was used. (As it happened, calling the function twice
5074 did not affect the overall behaviour.) The macro has been replaced by a
5075 function that can join an arbitrary number of extra strings onto a growing
5078 (c) The code for expansion conditions now uses a table and a binary chop
5079 instead of a serial search (which was left over from when there were very
5080 few conditions). Also, it now recognizes conditions like "pam" even when
5081 the relevant support is not compiled in: a suitably worded error message is
5082 given if an attempt is made to use such a condition.
5084 11. Added ${time_interval:xxxxx}.
5086 12. A bug was causing one of the ddress fields not to be passed back correctly
5087 from remote delivery subprocesses. The field in question was not being
5088 subsequently used, so this caused to problems in practice.
5090 13. Added new log selectors queue_time and deliver_time.
5092 14. Might have fixed a bug in maildirsizefile handling that threw up
5093 "unexpected character" debug warnings, and recalculated the data
5094 unnecessarily. In any case, I expanded the warning message to give more
5097 15. Added the message "Restricted characters in address" to the statements in
5098 the default ACL that block characters like @ and % in local parts.
5100 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5101 Three changes have been made:
5103 (a) There was a serious bug; a negative response to MAIL caused the whole
5104 recipient domain to be cached as invalid, thereby blocking all messages
5105 to all local parts at the same domain, from all senders. This bug has
5106 been fixed. The domain is no longer cached after a negative response to
5107 MAIL if the sender used is not empty.
5109 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5112 (c) A new callout option, "use_sender" has been added for people who want
5113 the modified behaviour.
5119 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5122 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5123 indeed breaks things for older releases.
5125 3. Added additional logging to the case where there is a problem reading data
5126 from a filter that is running in a subprocess using a pipe, in order to
5127 try to track down a specific problem.
5129 4. Testing facility fudge: when running in the test harness and attempting
5130 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5131 getting "No route to host". Convert this to a timeout.
5133 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5136 6. Some OS don't have socklen_t but use size_t instead. This affects the
5137 fifth argument of getsockopt() amongst other things. This is now
5138 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5139 can be set for individual OS. I have set it for SunOS5, OSF1, and
5140 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5141 some earlier ones do not.
5143 7. Change 4.30/15 was not doing the test caselessly.
5145 8. The standard form for an IPv6 address literal was being rejected by address
5146 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5147 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5148 this, as well as the form without the "IPv6" on the front (but only when
5149 address literals are enabled, of course).
5151 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5153 10. Exim crashed if a message with an empty sender address specified by -f
5154 encountered a router with an errors_to setting. This could be provoked only
5155 by a command such as
5159 where an empty string was supplied; "<>" did not hit this bug.
5161 11. Installed PCRE release 4.5.
5163 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5164 remained set. It is now erased.
5166 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5167 times from message ids (which are base 36 rather than the normal 62).
5169 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5170 were being counted as actual protocol errors, and logged if the log
5171 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5172 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5173 and DATA following a set of rejected RCPTs do not count as protocol errors.
5174 In other words, Exim assumes they were pipelined, though this may not
5175 actually be the case. Of course, in all cases the client gets an
5176 appropriate error code.
5178 15. If a lookup fails in an ACL condition, a message about the failure may
5179 be available; it is used if testing the ACL cannot continue, because most
5180 such messages specify what the cause of the deferral is. However, some
5181 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5182 that caused an old message to be retained and used if a later statement
5183 caused a defer, replacing the real cause of the deferral.
5185 16. If an IP address had so many PTR records that the DNS lookup buffer
5186 was not large enough to hold them, Exim could crash while trying to process
5187 the truncated data. It now detects and logs this case.
5189 17. Further to 4.21/58, another change has been made: if (and only if) the
5190 first line of a message (the first header line) ends with CRLF, a bare LF
5191 in a subsequent header line has a space inserted after it, so as not to
5192 terminate the header.
5194 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5195 unnecessarily, used atoi() instead of strtol(), and didn't check the
5196 termination when getting file sizes from file names by regex.
5198 19. Completely re-implemented the support for maildirsize files, in the light
5199 of a number of problems with the previous contributed implementation
5200 (4.30/29). In particular:
5202 . If the quota is zero, the maildirsize file is maintained, but no quota is
5205 . If the maildir directory does not exist, it is created before any attempt
5206 to write a maildirsize file.
5208 . The quota value in the file is just a cache; if the quota is changed in
5209 the transport, the new value overrides.
5211 . A regular expression is available for excluding directories from the
5214 20. The autoreply transport checks the characters in options that define the
5215 message's headers; it allows continued headers, but it was checking with
5216 isspace() after an embedded newline instead of explicitly looking for a
5219 21. If all the "regular" hosts to which an address was routed had passed their
5220 expiry times, and had not reached their retry times, the address was
5221 bounced, even if fallback hosts were defined. Now Exim should go on to try
5224 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5225 equivalent code in the SMTP transport. Some hosts send humungous responses
5226 to HELO/EHLO, more than 1024 it seems.
5228 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5229 gives compiler warnings in some environments. I've now done it "properly",
5232 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5233 (because of problems with the built-in one) was declared to return uschar *
5234 instead of char *, causing compiler failure.
5236 25. Fixed a file descriptor leak when processing alias/forward files.
5238 26. Fixed a minor format string issue in dbfn.c.
5240 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5242 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5243 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5246 29. When a sender address is verified, it is cached, to save repeating the test
5247 when there is more than one recipient in a message. However, when the
5248 verification involves a callout, it is possible for different callout
5249 options to be set for different recipients. It is too complicated to keep
5250 track of this in the cache, so now Exim always runs a verification when a
5251 callout is required, relying on the callout cache for the optimization.
5252 The overhead is duplication of the address routing, but this should not be
5255 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5256 to be verified with callout=postmaster, and the main callout worked but the
5257 postmaster check failed, the verification correctly failed. However, if a
5258 subsequent RCPT command asked for sender verification *without* the
5259 postmaster check, incorrect caching caused this verification also to fail,
5262 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5263 it was not caching the DNS options (qualify_single, search_parents) that
5264 were used when the lookup failed. A subsequent lookup with different
5265 options therefore always gave the same answer, though there were cases
5266 where it should not have. (Example: a "domains = !$mx_any" option on a
5267 dnslookup router: the "domains" option is always processed without any
5268 widening, but the router might have qualify_single set.) Now Exim uses the
5269 cached value only when the same options are set.
5271 32. Added John Jetmore's "exipick" utility to the distribution.
5273 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5274 than a timeout (e.g. a certificate is required, and is not provided), an
5275 Exim server now closes the connection immediately. Previously it waited for
5276 the client to close - but if the client is SSL, it seems that they each
5277 wait for each other, leading to a delay before one of them times out.
5279 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5280 maintained 0.8.x compatibility because I don't think many are using it, and
5281 it is clearly obsolete.
5283 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5286 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5287 client certificate was expired. A simple patch fixes this, though I don't
5288 understand the full logic of why the verify callback is called multiple
5291 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5292 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5293 which causes problems with some clients (such as the Certicom SSL Plus
5294 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5295 disables the coutermeasure allowing Eudora to connect."
5297 38. Exim was not checking that a write() to a log file succeeded. This could
5298 lead to Bad Things if a log got too big, in particular if it hit a file
5299 size limit. Exim now panics and dies if it cannot write to a log file, just
5300 as it does if it cannot open a log file.
5302 39. Modified OS/Makefile-Linux so that it now contains
5304 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5306 The two -D definitions ensure that Exim is compiled with large file
5307 support, which makes it possible to handle log files that are bigger than
5310 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5311 instance) a domain was checked against a named list that involved a lookup,
5312 causing $domain_data to be set, then another domain was checked against the
5313 same list, then the first domain was re-checked, the value of $domain_data
5314 after the final check could be wrong. In particular, if the second check
5315 failed, it could be set empty. This bug probably also applied to
5318 41. The strip_trailing_dot option was not being applied to the address given
5319 with the -f command-line option.
5321 42. The code for reading a message's header from the spool was incrementing
5322 $received_count, but never initializing it. This meant that the value was
5323 incorrect (doubled) while delivering a message in the same process in which
5324 it was received. In the most common configuration of Exim, this never
5325 happens - a fresh exec is done - but it can happen when
5326 deliver_drop_privilege is set.
5328 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5329 it now includes up to 150 characters of the unexpected data in the log
5332 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5333 and building data strings. The size of both of these buffers was 10 000
5334 bytes - far larger than anybody would *ever* want, thought I. Needless to
5335 say, somebody hit the limit. I have increased the maximum line length to
5336 20 000 and the maximum data length of concatenated lines to 100 000. I have
5337 also fixed two bugs, because there was no checking on these buffers. Tsk,
5338 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5339 buffer is too small.
5341 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5342 lsearch lookups. Now it does.
5344 46. When parsing a route_list item in a manualroute router, a fixed-length
5345 buffer was used for the list of hosts. I made this 1024 bytes long,
5346 thinking that nobody would ever have a list of hosts that long. Wrong.
5347 Somebody had a whole pile of complicated expansion conditions, and the
5348 string was silently truncated, leading to an expansion error. It turns out
5349 that it is easier to change to an unlimited length (owing to other changes
5350 that have happened since this code was originally written) than to build
5351 structure for giving a limitation error. The length of the item that
5352 expands into the list of hosts is now unlimited.
5354 47. The lsearch lookup could not handle data where the length of text line was
5355 more than 4095 characters. Such lines were truncated, leading to shortened
5356 data being returned. It should now handle lines of any length.
5358 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5359 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5362 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5363 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5364 no longer happens in comment lines. A list of replacements is now placed
5365 at the head of all of the source files, except those whose only change is
5366 to replace PERL_COMMAND in the very first #! line.
5368 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5369 messages on the queue, with a bottom-up merge sort, using code contributed
5370 by Michael Haardt. This should make operations like -bp somewhat faster on
5371 large queues. It won't affect queue runners, except when queue_run_in_order
5374 51. Installed eximstats 1.31 in the distribution.
5376 52. Added support for SRV lookups to the dnslookup router.
5378 53. If an ACL referred to $message_body or $message_body_end, the value was not
5379 reset for any messages that followed in the same SMTP session.
5381 54. The store-handling optimization for building very long strings was not
5382 differentiating between the different store pools. I don't think this
5383 actually made any difference in practice, but I've tidied it.
5385 55. While running the routers to verify a sender address, $sender_address
5386 was still set to the sender address. This is wrong, because when routing to
5387 send a bounce to the sender, it would be empty. Therefore, I have changed
5388 it so that, while verifying a sender address, $sender_address is set to <>.
5389 (There is no change to what happens when verifying a recipient address.)
5391 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5392 target A or AAAA records (if not already returned) without resetting the
5393 qualify_single or search_parents options of the DNS resolver. These are
5394 inappropriate in this case because the targets of MX and SRV records must
5395 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5396 target that, when qualified, matched something in the local domain. These
5397 two options are now turned off when doing these lookups.
5399 57. It seems that at least some releases of Reiserfs (which does not have the
5400 concept of a fixed number of inodes) returns zero and not -1 for the
5401 number of available inodes. This interacted badly with check_spool_inodes,
5402 which assumed that -1 was the "no such thing" setting. What I have done is
5403 to check that the total number of inodes is greater than zero before doing
5404 the test of how many are available.
5406 58. When a "warn" ACL statement has a log_message modifier, the message is
5407 remembered, and not repeated. This is to avoid a lot of repetition when a
5408 message has many recipients that cause the same warning to be written.
5409 Howewer, Exim was preserving the list of already written lines for an
5410 entire SMTP session, which doesn't seem right. The memory is now reset if a
5411 new message is started.
5413 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5414 debugging output unless log_rewrite was also set.
5416 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5417 of (int)(handle) when we know that handle contains (void *)(-1).
5419 61. The Exim daemon panic-logs an error return when it closes the incoming
5420 connection. However "connection reset by peer" seems to be common, and
5421 isn't really an error worthy of noting specially, so that particular error
5424 62. When Exim is trying to find all the local interfaces, it used to panic and
5425 die if the ioctl to get the interface flags failed. However, it seems that
5426 on at least one OS (Solaris 9) it is possible to have an interface that is
5427 included in the list of interfaces, but for which you get a failure error
5428 for this call. This happens when the interface is not "plumbed" into a
5429 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5430 failure of the "get flags" call assumes that the interface is down.
5432 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5433 makes life easier for people who are doing arithmetic on fields extracted
5434 from dates, where you often get leading zeros that should not be
5435 interpreted as octal.
5437 64. Added qualify_domain to the redirect router, to override the global
5440 65. If a pathologically long header line contained very many addresses (the
5441 report of this problem mentioned 10 000) and each of them was rewritten,
5442 Exim could use up a very large amount of memory. (It kept on making new
5443 copies of the header line as it rewrote, and never released the old ones.)
5444 At the expense of a bit more processing, the header rewriting function has
5445 been changed so that it no longer eats memory in this way.
5447 66. The generation of the Received: header has been moved from the time that a
5448 message starts to be received, to the time that it finishes. The timestamp
5449 in the Received: header should now be very close to that of the <= log
5450 line. There are two side-effects of this change:
5452 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5453 logged header lines no longer include the local Received: line, because
5454 it has not yet been created. The same applies to a copy of the message
5455 that is returned to a non-SMTP sender when a message is rejected.
5457 (b) When a filter file is tested using -bf, no additional Received: header
5458 is added to the test message. After some thought, I decided that this
5461 This change does not affect the value of $received_for. It is still set
5462 after address rewriting, but before local_scan() is called.
5464 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5466 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5467 gave an unhelpful panic error message, and a defer error. I have managed to
5468 change this behaviour so that it now rejects any supplied certificate,
5469 which seems right, as the list of acceptable certificates is empty.
5471 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5472 gave an unhelpful defer error. I have not managed to make this reject any
5473 supplied certificates, but the error message it gives is "no certificate
5474 supplied", which is not helpful.
5476 70. exigrep's output now also includes lines that are not associated with any
5477 message, but which match the given pattern. Implemented by a patch from
5478 Martin Sluka, which also tidied up the Perl a bit.
5480 71. Recipient callout verification, like sender verification, was using <> in
5481 the MAIL FROM command. This isn't really the right thing, since the actual
5482 sender may affect whether the remote host accepts the recipient or not. I
5483 have changed it to use the actual sender in the callout; this means that
5484 the cache record is now keyed on a recipient/sender pair, not just the
5485 recipient address. There doesn't seem to be a real danger of callout loops,
5486 since a callout by the remote host to check the sender would use <>.
5487 [SEE ABOVE: changed after hitting problems.]
5489 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5490 temporary errors. However, in the case of such a code being given after
5491 the end of a data transmission (i.e. after ".") Exim was failing to write
5492 a retry record for the message. (Yes, there was some broken host that was
5493 actually sending 8xx at this point.)
5495 73. An unknown lookup type in a host list could cause Exim to panic-die when
5496 the list was checked. (An example that provoked this was putting <; in the
5497 middle of a list instead of at the start.) If this happened during a DATA
5498 ACL check, a -D file could be left lying around. This kind of configuration
5499 error no longer causes Exim to die; instead it causes a defer errror. The
5500 incident is still logged to the main and panic logs.
5502 74. Buglet left over from Exim 3 conversion. The message "too many messages
5503 in one connection" was written to the rejectlog but not the mainlog, except
5504 when address rewriting (yes!) was being logged.
5506 75. Added write_rejectlog option.
5508 76. When a system filter was run not as root (that is, when system_filter_user
5509 was set), the values of the $n variables were not being returned to the
5510 main process; thus, they were not subsequently available in the $sn
5513 77. Added +return_path_on_delivery log selector.
5515 78. A connection timeout was being treated differently from recipients deferred
5516 when testing hosts_max_try with a message that was older than the host's
5517 retry timeout. (The host should not be counted, thus allowing all hosts to
5518 be tried at least once before bouncing.) This may have been the cause of an
5519 occasionally reported bug whereby a message would remain on the queue
5520 longer than the retry timeout, but would be bounced if a delivery was
5521 forced. I say "may" because I never totally pinned down the problem;
5522 setting up timeout/retry tests is difficult. See also the next item.
5524 79. The ultimate address timeout was not being applied to errors that involved
5525 a combination of host plus message (for example, a timeout on a MAIL
5526 command). When an address resolved to a number of possible hosts, and they
5527 were not all tried for each delivery (e.g. because of hosts_max_try), a
5528 message could remain on the queue longer than the retry timeout.
5530 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5533 81. Fixed an obscure SMTP outgoing bug which required at least the following
5534 conditions: (a) there was another message waiting for the same server;
5535 (b) the server returned 5xx to all RCPT commands in the first message so
5536 that the message was not completed; (c) the server dropped the connection
5537 or gave a negative response to the RSET that Exim sends to abort the
5538 transaction. The observed case was a dropped connection after DATA that had
5539 been sent in pipelining mode. That is, the server had advertised PIPELINING
5540 but was not implementing it correctly. The effect of the bug was incorrect
5541 behaviour, such as trying another host, and this could lead to a crash.
5547 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5548 and daemon.c were passed as pointers to ints; they should have been
5549 pointers to socklen_t variables (which are typically unsigned ints).
5551 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5554 3. Fixed a really odd bug that affected only the testing scheme; patching a
5555 certain fixed string in the binary changed the value of another string that
5556 happened to be identical to the end of the original first string.
5558 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5559 name", it returns that address as the IP address. On some operating
5560 systems (e.g. Solaris), it also passes back the IP address string as the
5561 "host name". However, on others (e.g. Linux), it passes back an empty
5562 string. Exim wasn't checking for this, and was changing the host name to an
5563 empty string, assuming it had been canonicized.
5565 5. Although rare, it is permitted to have more than one PTR record for a given
5566 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5567 all the names associated with an address, because they do in Solaris.
5568 However, it seems that they do not in Linux for data that comes from the
5569 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5570 I found this out when I moved to a new Linux workstation and tried to run
5571 the Exim test suite.
5573 To get round this problem I have changed the code so that it now does its
5574 own call to the DNS to look up PTR records when searching for a host name.
5575 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5576 addresses that are only in /etc/hosts are still found.
5578 This behaviour is, however, controlled by an option called host_lookup_
5579 order, which defaults to "bydns:byaddr". If people want to use the other
5580 order, or indeed, just use one or the other means of lookup, they can
5581 specify it in this variable.
5583 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5584 some operating systems, this comes back from gethostbyaddr() as an empty
5585 string, and this is what Exim used to test for. However, it seems that in
5586 other systems, "." is yielded. Exim now tests for this case too.
5588 7. The values of check_spool_space and check_log_space are now held internally
5589 as a number of kilobytes instead of an absolute number of bytes. If a
5590 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5591 kilobyte. This means that much larger values can be stored.
5593 8. Exim monitor: an attempt to get the action menu when not actually pointing
5594 at a message produces an empty menu entitled "No message selected". This
5595 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5596 no entries in it ("Shell widget menu has zero width and/or height"). So I
5597 have added a single, blank menu entry in this case.
5599 9. Added ${quote_local_part.
5601 10. MIME decoding is now applied to the contents of Subject: header lines when
5604 11. Now that a reference to $sender_host_address automatically causes a reverse
5605 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5606 host lookup before query-style lookups in lists that might use this
5607 variable. This has therefore been abolished, and the "net-" prefix is no
5608 longer necessary for query-style lookups.
5610 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5611 have been a typo for LFLAGS, so it has been changed.
5613 13. The install script calls Exim with "-C /dev/null" in order to find the
5614 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5615 to be output. Howeve, since Exim outputs its version number before the
5616 error, it didn't break the script. It just looked ugly. I fixed this by
5617 always allowing "-C /dev/null" if the caller is root.
5619 14. Ignore overlarge ACL variable number when reading spool file - insurance
5620 against a later release with more variables having written the file.
5622 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5623 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5624 this, as well as the form without the "IPv6" on the front.
5626 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5627 OS/Makefile-Darwin file.
5629 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5630 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5632 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5633 variable. However, some OS use ints for inodes. I've added cast to long int
5634 to get rid of the compiler warning.
5636 19. I had forgotten to lock out "/../" in configuration file names when
5637 ALT_CONFIG_PREFIX was set.
5639 20. Routers used for verification do not need to specify transports. However,
5640 if such a router generated a host list, and callout was configured, Exim
5641 crashed, because it could not find a port number from the (non-existent)
5642 transport. It now assumes port 25 in this circumstance.
5644 21. Added the -t option to exigrep.
5646 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5647 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5648 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5649 them is set, LOOKUP_LSEARCH is forced.
5651 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5653 24. Added sender and host information to the "rejected by local_scan()" log
5654 line; previously there was no indication of these.
5656 25. Added .include_if_exists.
5658 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5659 Linux. It turns out that not all file systems support this. Apparently some
5660 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5661 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5662 not supported on the file descriptor, is now ignored when Exim is trying to
5663 sync a directory. This applies only to Linux.
5665 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5667 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5668 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5671 29. Added support for maildirsize files from supplied patch (modified a bit).
5673 30. The use of :fail: followed by an empty string could lead Exim to respond to
5674 sender verification failures with (e.g.):
5676 550 Verification failed for <xxx>
5677 550 Sender verify failed
5679 where the first response line was missing the '-' that indicates it is not
5680 the final line of the response.
5682 31. The loop for finding the name of the user that called Exim had a hardwired
5683 limit of 10; it now uses the value of finduser_retries, which is used for
5684 all other user lookups.
5686 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5689 33. Exim was neglecting to zero errno before one call of strtol() when
5690 expanding a string and expecting an integer value. On some systems this
5691 resulted in spurious "integer overflow" errors. Also, it was casting the
5692 result into an int without checking.
5694 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5695 did not work. The code looks as if it has *never* worked, though it appears
5696 to have been documented since at least releast 1.62. I have made it work.
5698 35. The "timeout_DNS" error in retry rules, also documented since at least
5699 1.62, also never worked. As it isn't clear exactly what this means, and
5700 clearly it isn't a major issue, I have abolished the feature by treating it
5701 as "timeout", and writing a warning to the main and panic logs.
5703 36. The display of retry rules for -brt wasn't always showing the error code
5706 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5707 timeout_connect_A, timeout_connect_MX.
5709 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5710 to the empty sender.
5712 39. The daemon was not analysing the content of -oX till after it had closed
5713 stderr and disconnected from the controlling terminal. This meant that any
5714 syntax errors were only noted on the panic log, and the return code from
5715 the command was 0. By re-arranging the code a little, I've made the
5716 decoding happen first, so such errors now appear on stderr, and the return
5717 code is 1. However, the actual setting up of the sockets still happens in
5718 the disconnected process, so errors there are still only recorded on the
5721 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5722 connections (as happens on some IP stacks) was logged at start up time as
5723 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5724 it from "IPv6 and IPv4", which means that two separate sockets are being
5727 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5728 says whether AF_INET or AF_INET6 was passed as an argument.
5730 42. Exiwhat output was messed up when time zones were included in log
5733 43. Exiwhat now gives more information about the daemon's listening ports,
5734 and whether -tls-on-connect was used.
5736 44. The "port" option of the smtp transport is now expanded.
5738 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5739 silently ignored. Now an error message is written to the main and panic
5742 46. There's a new ACL modifier called "logwrite" which writes to a log file
5743 as soon as it is encountered.
5745 47. Added $local_user_uid and $local_user_gid at routing time.
5747 48. Exim crashed when trying to verify a sender address that was being
5750 49. Exim was recognizing only a space character after ".include". It now also
5751 recognizes a tab character.
5753 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5754 extracting the relevant information from the specification. The man page no
5755 longer contains scrambled data for the -d option, and I've added a section
5756 at the front about calling Exim under different names.
5758 51. Added "extra_headers" argument to the "mail" command in filter files.
5760 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5763 53. Installed eximstats 1.29.
5765 54. Added transport_filter_timeout as a generic transport option.
5767 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5768 Cc: header lines. This was required by RFC 822, but it not required by RFC
5771 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5772 incoming messages that did not have them. Now it does so only if the
5773 message originates locally, that is, if there is no associated remote host
5774 address. When Resent- header lines are present, this applies to the Resent-
5775 lines rather than the non-Resent- lines.
5777 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5778 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5780 58. Messages for configuration errors now include the name of the main
5781 configuration file - useful now that there may be more than one file in a
5782 list (.included file names were always shown).
5784 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5785 for those rare installations that do not start the daemon as root or run it
5786 setuid root. I've cut out the call to initgroups() if the daemon is not
5789 60. The Exim user and group can now be bound into the binary as text strings
5790 that are looked up at the start of Exim's processing.
5792 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5794 62. Added $mailstore_basename variable.
5796 63. Installed patch to sieve.c from Michael Haardt.
5798 64. When Exim failed to open the panic log after failing to open the main log,
5799 the original message it was trying to log was written to stderr and debug
5800 output, but if they were not available (the usual case in production), it
5801 was lost. Now it is written to syslog before the two lines that record the
5802 failures to open the logs.
5804 65. Users' Exim filters run in subprocesses under the user's uid. It is
5805 possible for a "deliver" command or an alias in a "personal" command to
5806 provoke an address rewrite. If logging of address rewriting is configured,
5807 this fails because the process is not running as root or exim. There may be
5808 a better way of dealing with this, but for the moment (because 4.30 needs
5809 to be released), I have disabled address rewrite logging when running a
5810 filter in a non-root, non-exim process.
5816 1. The buildconfig auxiliary program wasn't quoting the value set for
5817 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5818 not defined. This bug was masked in 4.22 by the effect that was fixed in
5821 2. Some messages that were rejected after a message id was allocated were
5822 shown as "incomplete" by exigrep. It no longer does this for messages that
5823 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5825 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5826 have allow_domain_literals set, the ID did not get logged in the <= line.
5827 Domain literals are now always recognized in Message-ID: header lines.
5829 4. The first argument for a ${extract expansion item is the key name or field
5830 number. Leading and trailing spaces in this item were not being ignored,
5831 causing some misleading effects.
5833 5. When deliver_drop_privilege was set, single queue runner processes started
5834 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5835 same command in the process it spins off) were not dropping privilege.
5837 6. When the daemon running as "exim" started a queue runner, it always
5838 re-executed Exim in the spun-off process. This is a waste of effort when
5839 deliver_drop_privilege is set. The new process now just calls the
5840 queue-runner function directly.
5846 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5849 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5850 ignored. Though the use of -oP was forcing the writing of a pid file, it
5851 was always written to the default place.
5853 3. If the message "no IP address found for host xxxx" is generated during
5854 incoming verification, it is now followed by identification of the incoming
5855 connection (so you can more easily find what provoked it).
5857 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5859 5. Added some features to "harden" Exim a bit more against certain attacks:
5861 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5862 be put in Local/Makefile. This is like the never_users runtime option,
5863 but it cannot be overridden. The default setting is "root".
5865 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5866 prefix string with which any file named in a -C command line option
5869 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5870 is retained for -C and -D only if the caller of Exim is root. Without
5871 it, the exim user may also use -C and -D and retain privilege.
5873 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5874 command line option is disabled.
5876 6. Macro names set by the -D option must start with an upper case letter, just
5877 like macro names defined in the configuration file.
5879 7. Added "dereference=" facility to LDAP.
5881 8. Two instances of the typo "uknown" in the source files are fixed.
5883 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5884 the Configure-Makefile script screwed up while processing it.
5886 10. Incorporated PCRE 4.4.
5888 11. The SMTP synchronization check was not operating right at the start of an
5889 SMTP session. For example, it could not catch a HELO sent before the client
5890 waited for the greeting. There is now a check for outstanding input at the
5891 point when the greeting is written. Because of the duplex, asynchronous
5892 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5893 way, but not yet received, when the check is performed.
5895 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5896 on TCP/IP sockets, because this apparently causes some broken clients to
5899 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5900 unchanged) from the Cygwin maintainer.
5902 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5903 was supported instead of testing for mbx. Effectively a typo.
5905 15. The spa authenticator server code was not checking that the input it
5906 received was valid base64.
5908 16. The debug output line for the "set" modifier in ACLs was not showing the
5909 name of the variable that was being set.
5911 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5913 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5914 lookup on its own. Something else was needed to trigger the lookup. For
5915 example, a match in host_lookup or the need for a host name in a host list.
5916 Now, if $sender_host_name is referenced and the host name has not yet been
5917 looked up, a lookup is performed. If the lookup fails, the variable remains
5918 empty, and $host_lookup_failed is set to "1".
5920 19. Added "eqi" as a case-independent comparison operator.
5922 20. The saslauthd authentication condition could segfault if neither service
5923 nor realm was specified.
5925 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5926 error message that was logged was misleading, and incoming SMTP
5927 connections were dropped. The message is now more accurate, and temporary
5928 errors are given to SMTP connections.
5930 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5931 (except RSET) with a 421 error, until QUIT is received. However, it was
5932 failing to send a response to QUIT.
5934 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5935 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5936 ACL is now run after the helo_try_verify_hosts code.
5938 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5939 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5940 case-independent, so other case variants are also recognized). Apparently
5941 some systems use these upper case variants.
5943 25. If more than two messages were waiting for the same host, and a transport
5944 filter was specified for the transport, Exim sent two messages over the
5945 same TCP/IP connection, and then failed with "socket operation on non-
5946 socket" when it tried to send the third.
5948 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5950 27. The extern definition of crypt16() in expand.c was not being excluded when
5951 the OS had its own crypt16() function.
5953 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5954 as a preferred synonym for return_size_limit, both as an option and as an
5957 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5959 30. Changed the default configuration ACL to relax the local part checking rule
5960 for addresses that are not in any local domains. For these addresses,
5961 slashes and pipe symbols are allowed within local parts, but the sequence
5962 /../ is explicitly forbidden.
5964 31. SPA server authentication was not clearing the challenge buffer before
5967 32. log_message in a "warn" ACL statement was writing to the reject log as
5968 well as to the main log, which contradicts the documentation and doesn't
5969 seem right (because no rejection is happening). So I have stopped it.
5971 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5972 However, I am unable to do any testing of this.
5974 34. Fixed an infelicity in the appendfile transport. When checking directories
5975 for a mailbox, to see if any needed to be created, it was accidentally
5976 using path names with one or more superfluous leading slashes; tracing
5977 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5979 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5980 contents are added to the log line that is written for every discarded
5981 recipient. (Previously a log_message setting was ignored.)
5983 36. The ${quote: operator now quotes the string if it is empty.
5985 37. The install script runs exim in order to find its version number. If for
5986 some reason other than non-existence or emptiness, which it checks, it
5987 could not run './exim', it was installing it with an empty version number,
5988 i.e. as "exim-". This error state is now caught, and the installation is
5991 38. An argument was missing from the function that creates an error message
5992 when Exim fails to connect to the socket for saslauthd authentication.
5993 This could cause Exim to crash, or give a corrupted message.
5995 39. Added isip, isip4, and isip6 to ${if conditions.
5997 40. The ACL variables $acl_xx are now saved with the message, and can be
5998 accessed later in routers, transports, and filters.
6000 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6001 strings in the file are not string-expanded.
6003 42. If a MAIL command specified a SIZE value that was too large to fit into an
6004 int variable, the check against message_size_limit failed. Such values are
6005 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6006 day this will have to be increased, but I don't think I want to be around
6007 when emails are that large.
6014 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6015 iconv() is not standard in FreeBSD.
6017 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6018 IPv6 enabled. The observed symptom was a segmentation fault on return from
6019 the function os_common_find_running_interfaces() in src/os.c.
6021 3. In the check_special_case() function in daemon.c I had used "errno" as an
6022 argument name, which causes warnings on some systems. This was basically a
6023 typo, since it was named "eno" in the comments!
6025 4. The code that waits for the clock to tick (at a resolution of some fraction
6026 of a second) so as to ensure message-id uniqueness was always waiting for
6027 at least one whole tick, when it could have waited for less. [This is
6028 almost certainly not relevant at current processor speeds, where it is
6029 unlikely to ever wait at all. But we try to future-proof.]
6031 5. The function that sleeps for a time interval that includes fractions of a
6032 second contained a race. It did not block SIGALRM between setting the
6033 timer, and suspending (a couple of lines later). If the interval was short
6034 and the sigsuspend() was delayed until after it had expired, the suspension
6035 never ended. On busy systems this could lead to processes getting stuck for
6038 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6039 process, before it forks any delivery processes. The open lookup caching
6040 mechanism meant that the open file or database connection was passed into
6041 the delivery process. The problem was that delivery processes always tidy
6042 up cached lookup data. This could cause a problem for the next delivery
6043 process started by the queue runner, because the external queue runner
6044 process does not know about the closure. So the next delivery process
6045 still has data in the lookup cache. In the case of a file lookup, there was
6046 no problem because closing a file descriptor in a subprocess doesn't affect
6047 the parent. However, if the lookup was caching a connection to a database,
6048 the connection was closed, and the second delivery process was likely to
6049 see errors such as "PGSQL: query failed: server closed the connection
6050 unexpectedly". The problem has been fixed by closing all cached lookups
6051 in a queue runner before running a delivery process.
6053 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6054 seem to have the "const" qualifier which it has on other OS. I've
6057 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6058 *of the same type* (client or server) with the same public name that an
6059 error should be diagnosed.
6061 9. When Exim looked up a host name for an IP address, but failed to find the
6062 original IP address when looking up the host name (a safety check), it
6063 output the message "<ip address> does not match any IP for NULL", which was
6064 confusing, to say the least. The bug was that the host name should have
6065 appeared instead of "NULL".
6067 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6068 user that is built into the binary, and the -C or -D options is used, root
6069 privilege is dropped before the configuration file is read. In addition,
6070 logging is switched to stderr instead of the normal log files. If the
6071 configuration then re-defines the Exim user, the unprivileged environment
6072 is probably not what is expected, so Exim logs a panic warning message (but
6075 However, if deliver_drop_privilege is set, the unprivileged state may well
6076 be exactly what is intended, so the warning has been cut out in that case,
6077 and Exim is allowed to try to write to its normal log files.
6083 1. smtp_return_error_details was not giving details for temporary sender
6084 or receiver verification errors.
6086 2. Diagnose a configuration error if two authenticators have the same public
6089 3. Exim used not to create the message log file for a message until the first
6090 delivery attempt. This could be confusing when incoming messages were held
6091 for policy or load reasons. The message log file is now created at the time
6092 the message is received, and an initial "Received" line is written to it.
6094 4. The automatically generated man page for command line options had a minor
6095 bug that caused no ill effects; however, a more serious problem was that
6096 the procedure for building the man page automatically didn't always
6097 operate. Consequently, release 4.20 contains an out-of-date version. This
6098 shouldn't happen again.
6100 5. When building Exim with embedded Perl support, the script that builds the
6101 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6102 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6104 6. The freeze_tell option was not being used for messages that were frozen on
6105 arrival, either by an ACL or by local_scan().
6107 7. Added the smtp_incomplete_transaction log selector.
6109 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6110 was accepting AUTH without a new EHLO.
6112 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6113 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6116 10. Logging of TCP/IP connections (when configured) now happens in the main
6117 daemon process instead of the child process, so that the TCP/IP connection
6118 count is more accurate (but it can never be perfect).
6120 11. The use of "drop" in a nested ACL was not being handled correctly in the
6121 outer ACL. Now, if condition failure induced by the nested "drop" causes
6122 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6123 or "require"), the connection is dropped.
6125 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6126 that yield "discard" can now be used with an "accept" or a "discard" verb,
6127 but an error is generated for any others (because I can't see a useful way
6128 to define what should happen).
6130 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6131 are now processed in the same way as lines in the Exim configuration file.
6132 In particular, continuation lines are supported.
6134 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6136 15. Added -ti meaning -t -i.
6138 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6139 domains, and warn by logging if others are found.
6141 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6142 returned from GIFCONF when Exim is trying to find the list of interfaces on
6143 a host. The code in os.c has been modified to copy each ifreq to an aligned
6144 structure in all cases.
6146 Also, in some cases, the returned ifreq's were being copied to a 'struct
6147 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6148 means the last couple of bytes of an IPv6 address could be chopped if the
6149 ifreq contained only a normal sockaddr (14 bytes storage).
6151 18. Named domain lists were not supported in the hosts_treat_as_local option.
6152 An entry such as +xxxx was not recognized, and was treated as a literal
6155 19. Ensure that header lines added by a DATA ACL are included in the reject log
6156 if the ACL subsequently rejects the message.
6158 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6159 MD5 (which is deprecated).
6161 21. When testing a filter file using -bf, Exim was writing a message when it
6162 took the sender from a "From " line in the message, but it was not doing so
6163 when it took $return_path from a Return-Path: header line. It now does.
6165 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6166 with a valid header line field name (a series of printing characters
6167 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6169 23. Changed "disc" in the source to "disk" to conform to the documentation and
6170 the book and for uniformity.
6172 24. Ignore Sendmail's -Ooption=value command line item.
6174 25. When execve() failed while trying to run a command in a pipe transport,
6175 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6176 could be confused with a return value of 69 from the command itself. This
6177 has been changed to 127, the value the shell returns if it is asked to run
6178 a non-existent command. The wording for the related log line suggests a
6179 non-existent command as the problem.
6181 26. If received_header_text expands to an empty string, do not add a Received:
6182 header line to the message. (Well, it adds a token one on the spool, but
6183 marks it "old" so that it doesn't get used or transmitted.)
6185 27. Installed eximstats 1.28 (addition of -nt option).
6187 28. There was no check for failure on the call to getsockname() in the daemon
6188 code. This can fail if there is a shortage of resources on the system, with
6189 ENOMEM, for example. A temporary error is now given on failure.
6191 29. Contrary to the C standard, it seems that in some environments, the
6192 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6193 program. Exim now does this explicitly; it affects the formatting of
6194 timestamps using strftime().
6196 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6197 complaints. I've now initialized all the variables, to avoid this.
6199 32. Header lines added by a system filter were not being "seen" during
6200 transport-time rewrites.
6202 33. The info_callback() function passed to OpenSSL is set up with type void
6203 (*)(SSL *, int, int), as described somewhere. However, when calling the
6204 function (actually a macro) that sets it up, the type void(*)() is
6205 expected. I've put in a cast to prevent warnings from picky compilers.
6207 34. If a DNS black list lookup found a CNAME record, but there were no A
6208 records associated with the domain it pointed at, Exim crashed.
6210 35. If a DNS black list lookup returned more than one A record, Exim ignored
6211 all but the first. It now scans all returned addresses if a particular IP
6212 value is being sought. In this situation, the contents of the
6213 $dnslist_value variable are a list of all the addresses, separated by a
6216 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6217 to accept a host name and all its aliases if the forward lookup for any of
6218 them yielded the IP address of the incoming connection. Now it accepts only
6219 those names whose forward lookup yields the correct IP address. Any other
6220 names are discarded. This closes a loophole whereby a rogue DNS
6221 administrator could create reverse DNS records to break through a
6222 wildcarded host restriction in an ACL.
6224 37. If a user filter or a system filter that ran in a subprocess used any of
6225 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6226 the wrong values were passed to the pipe command ($thisaddress had the
6227 value of $0, $0 had the value of $1, etc). This bug was introduced by
6228 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6230 38. Improved the line breaking for long SMTP error messages from ACLs.
6231 Previously, if there was no break point between 40 and 75 characters, Exim
6232 left the rest of the message alone. Two changes have been made: (a) I've
6233 reduced the minimum length to 35 characters; (b) if it can't find a break
6234 point between 35 and 75 characters, it looks ahead and uses the first one
6235 that it finds. This may give the occasional overlong line, but at least the
6236 remaining text gets split now.
6238 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6239 file descriptors might be low, and that setting 1000 would always raise it.
6240 It turns out that in some environments, the limit is already over 1000 and
6241 that lowering it causes trouble. So now Exim takes care not to decrease it.
6243 40. When delivering a message, the value of $return_path is set to $sender_
6244 address at the start of routing (routers may change the value). By an
6245 oversight, this default was not being set up when an address was tested by
6246 -bt or -bv, which affected the outcome if any router or filter referred to
6249 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6250 log only when "message" or "log_message" are set. However, if one of the
6251 conditions was an address verification, or a call to a nested ACL, the
6252 messages generated by the underlying test were being passed through. This
6253 no longer happens. The underlying message is available in $acl_verify_
6254 message for both "message" and "log_message" expansions, so it can be
6255 passed through if needed.
6257 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6258 new expansion $bh_ to give the encoded byte string without charset
6259 translation. Translation happens only if iconv() is available; HAVE_ICONV
6260 indicates this at build time. HEADERS_CHARSET gives the charset to
6261 translate to; headers_charset can change it in the configuration, and
6262 "headers charset" can change it in an individual filter file.
6264 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6265 that creates RFC 2047 encoded "words" labels them as that charset instead
6266 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6267 expansion operator; (ii) when Exim creates a From: line for a local
6268 message; (iii) when a header line is rewritten to include a "phrase" part.
6270 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6271 buggy, causing it to skip the first lines of messages whose message ID
6272 ended in 'D'. This would not have bitten before Exim release 4.14, because
6273 message IDs were unlikely to end in 'D' before then. The effect was to have
6274 incorrect size information for certain domains.
6276 45. #include "config.h" was missing at the start of the crypt16.c module. This
6277 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6280 46. If there was a timeout during a "random" callout check, Exim treated it as
6281 a failure of the random address, and carried on sending RSET and the real
6282 address. If the delay was just some slowness somewhere, the response to the
6283 original RCPT would be taken as a response to RSET and so on, causing
6284 mayhem of various kinds.
6286 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6287 when I implemented it. It didn't allow for the fact that some option values
6288 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6289 the right test for positive values.
6291 48. Domain names in DNS records are case-independent. Exim always looks them up
6292 in lower case. Some resolvers return domain names in exactly the case they
6293 appear in the zone file, that is, they may contain uppercase letters. Not
6294 all resolvers do this - some return always lower case. Exim was treating a
6295 change of case by a resolver as a change of domain, similar to a widening
6296 of a domain abbreviation. This triggered its re-routing code and so it was
6297 trying to route what was effectively the same domain again. This normally
6298 caused routing to fail (because the router wouldn't handle the domain
6299 twice). Now Exim checks for this case specially, and just changes the
6300 casing of the domain that it ultimately uses when it transmits the message
6303 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6306 50. If a filter generated a file delivery with a non-absolute name (possible if
6307 no home directory exists for the router), the forbid_file option was not
6310 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6311 the existing equality matching.
6313 52. Exim was using ints instead of ino_t variables in some places where it was
6314 dealing with inode numbers.
6316 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6317 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6318 called TMPDIR, and if it finds it is different, it changes its value.
6320 54. The smtp_printf() function is now made available to local_scan() so
6321 additional output lines can be written before returning. There is also an
6322 smtp_fflush() function to enable the detection of a dropped connection.
6323 The variables smtp_input and smtp_batched_input are exported to
6326 55. Changed the default runtime configuration: the message "Unknown user"
6327 has been removed from the ACL, and instead placed on the localuser router,
6328 using the cannot_route_message feature. This means that any verification
6329 failures that generate their own messages won't get overridden. Similarly,
6330 the "Unrouteable address" message that was in the ACL for unverifiable
6331 relay addresses has also been removed.
6333 56. Added hosts_avoid_esmtp to the smtp transport.
6335 57. The exicyclog script was not checking for the esoteric option
6336 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6337 will work only if exicyclog is run under the appropriate euid.
6339 58. Following a discussion on the list, the rules by which Exim recognises line
6340 endings on incoming messages have been changed. The -dropcr and drop_cr
6341 options are now no-ops, retained only for backwards compatibility. The
6342 following line terminators are recognized: LF CRLF CR. However, special
6343 processing applies to CR:
6345 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6346 nor a local message in the state where . is a terminator.
6348 (ii) If a bare CR is encountered in a header line, an extra space is added
6349 after the line terminator so as not to end the header. The reasoning
6350 behind this is that bare CRs in header lines are most likely either
6351 to be mistakes, or people trying to play silly games.
6353 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6354 was being incorrectly given as 18 bytes larger than it should have been.
6355 This is a VOB (very old bug).
6357 60. This may never have affected anything current, but just in case it has:
6358 When the local host is found other than at the start of a list of hosts,
6359 the local host, those with the same MX, and any that follow, are discarded.
6360 When the list in question was part of a longer list of hosts, the following
6361 hosts (not currently being processed) were also being discarded. This no
6362 longer happens. I'm not sure if this situation could ever has previously
6365 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6368 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6369 header. This is something that is recommended in a new Internet Draft, and
6370 is something that is documented as being done by Sendmail. There are two
6371 possible values. For messages generated by the autoreply transport, Exim
6374 Auto-Submitted: auto-replied
6376 whereas for all other generated messages (e.g. bounces) it adds
6378 Auto-Submitted: auto-generated
6380 63. The "personal" condition in filters now includes a test for the
6381 Auto-Submitted: header. If it contains the string "auto-" the message it
6382 not considered personal.
6384 64. Added rcpt_include_affixes as a generic transport option.
6386 65. Added queue_only_override (default true).
6388 66. Added the syslog_duplication option.
6390 67. If what should have been the first header line of a message consisted of
6391 a space followed by a colon, Exim was mis-interpreting it as a header line.
6392 It isn't of course - it is syntactically invalid and should therefore be
6393 treated as the start of the message body. The misbehaviour could have
6394 caused a number of strange effects, including loss of data in subsequent
6395 header lines, and spool format errors.
6397 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6398 client host had authenticated. This control can now be exercised by an ACL
6399 for more flexibility.
6401 69. By default, callouts do not happen when testing with -bh. There is now a
6402 variant, -bhc, which does actually run the callout code, including
6403 consulting and updating the callout cache.
6405 70. Added support for saslauthd authentication, courtesy of Alexander
6408 71. If statvfs() failed on the spool or log directories while checking their
6409 size for availability, Exim confusingly gave the error "space shortage".
6410 Furthermore, in debugging mode it crashed with a floating point exception.
6411 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6412 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6413 serious problem, Exim now writes to the main and panic logs when this
6414 happens, with details of the failure. It then refuses to accept the
6415 incoming message, giving the message "spool directory problem" or "log
6416 directory problem" with a 421 code for SMTP messages.
6418 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6419 0, 1, and 2 exist, because some OS complain for execs without them (see
6420 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6421 descriptors. However, the code omitted to check that the open succeeded,
6422 causing mysterious errors if for some reason the permissions on /dev/null
6423 got screwed. Now Exim writes a message to the main and panic logs, and
6424 bombs out if it can't open /dev/null.
6426 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6427 interact so that it is all more flexible. It is supposed to remain
6428 backwards compatible. Also added extra_local_interfaces.
6430 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6431 to bomb out with an assertion failure - to the client this appears as a
6432 connection drop. This problem occurs in the part of the code that was taken
6433 from the Samba project. Fortunately, the assertion is in a very simple
6434 function, so I have fixed this by reproducing the function inline in the
6435 one place where it is called, and arranging for authentication to fail
6436 instead of killing the process with assert().
6438 75. The SPA client code was not working when the server requested OEM rather
6439 than Unicode encoding.
6441 76. Added code to make require_files with a specific uid setting more usable in
6442 the case where statting the file as root fails - usually a non-root-mounted
6443 NFS file system. When this happens and the failure is EACCES, Exim now
6444 forks a subprocess and does the per-uid checking as the relevant uid.
6446 77. Added process_log_path.
6448 78. If log_file_path was not explicitly set, a setting of check_log_space or
6449 check_log_inodes was ignored.
6451 79. If a space check for the spool or log partitions fails, the incident is now
6452 logged. Of course, in the latter case the data may get lost...
6454 80. Added the %p formatting code to string_format() so that it can be used to
6455 print addresses in debug_print(). Adjusted all the address printing in the
6456 debugging in store.c to use %p rather than %d.
6458 81. There was a concern that a line of code in smtp_in.c could overflow a
6459 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6460 initially expressed, the concern was not well-founded, because trailing
6461 spaces are removed early. However, if the trailing spaces were followed by
6462 a NULL, they did not get removed, so the overflow was possible. Two fixes
6465 (a) I re-wrote the offending code in a cleaner fashion.
6466 (b) If an incoming SMTP command contains a NULL character, it is rejected
6469 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6470 runs initgroups(), so that if the Exim user is in any additional groups,
6471 they will be used during message reception.
6477 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob