1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
8 supports it and a size is available (ie. the sending peer gave us one).
13 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
14 and 3.4.4 - once the server is enabled to respond to an OCSP request
15 it does even when not requested, resulting in a stapling non-aware
16 client dropping the TLS connection.
18 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
19 support variable-length bit vectors. No functional change.
21 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
22 The I= interface field on outgoing lines is now after the H= remote
23 host field, same as incoming lines. There is a separate
24 outgoing_interface log selector which allows you to disable the
27 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
28 If not running log_selector +smtp_connection the mainlog would be held
29 open indefinitely after a "too many connections" event, including to a
30 deleted file after a log rotate. Leave the per net connection logging
31 leaving it open for efficiency as that will be quickly detected by the
32 check on the next write.
34 HS/01 Bug 1671: Fix post transport crash.
35 Processing the wait-<transport> messages could crash the delivery
36 process if the message IDs didn't exist for some reason. When
37 using 'split_spool_directory=yes' the construction of the spool
38 file name failed already, exposing the same netto behaviour.
40 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
41 mime_regex ACL conditions.
43 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
44 to DSN fail messages (bounces): remote IP, remote greeting, remote response
45 to HELO, local diagnostic string.
47 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
48 log line to debug. Even when configured with a tls authenticator many
49 client connections are expected to not authenticate in this way, so
50 an authenticate fail is not an error.
52 HS/02 Add the Exim version string to the process info. This way exiwhat
53 gives some more detail about the running daemon.
55 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
56 matter for fast-change records such as DNSBLs.
58 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
59 retry record, even if constant. There may be multiple transports with
60 different interface settings and the retry behaviour needs to be kept
63 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
65 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
67 JH/10 Bug 840: fix log_defer_output option of pipe transport
69 JH/11 Bug 830: use same host for all RCPTS of a message, even under
70 hosts_randomize. This matters a lot when combined with mua_wrapper.
72 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
73 ${quote_pgsql:<string>} operator.
75 JH/13 Bug 1708: avoid misaligned access in cached lookup.
77 JH/14 Change header file name for freeradius-client. Relevant if compiling
78 with Radius support; from the Gentoo tree and checked under Fedora.
80 JH/15 Bug 1712: Introduce $prdr_requested flag variable
82 JH/16 Bug 1714: Permit an empty string as expansion result for transport
83 option transport_filter, meaning no filtering.
85 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
87 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
88 defaults to "*" (all hosts). The variable is now available when not built
89 with TLS, default unset, mainly to enable keeping the testuite sane.
90 If a server certificate is not supplied (via tls_certificate) an error is
91 logged, and clients will find TLS connections fail on startup. Presumably
92 they will retry in-clear.
93 Packagers of Exim are strongly encouraged to create a server certificate
96 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
97 with the $config_file variable.
99 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
100 in transport context, after the attempt, and per-recipient. The latter type
101 is per host attempted. The event data is the error message, and the errno
102 information encodes the lookup type (A vs. MX) used for the (first) host,
103 and the trailing two digits of the smtp 4xx reponse.
105 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
106 to write to mainlog (or rejectlog, paniclog) in the window between file
107 creation and permissions/ownership being changed. Particularly affects
108 installations where exicyclog is run as root, rather than exim user;
109 result is that the running daemon panics and dies.
111 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
113 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
114 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
115 "pri" and "weight". Note that the previous implicit priority given by the
116 list order is no longer honoured.
118 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
121 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
122 by defining SUPPORT_SOCKS.
124 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
125 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
126 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
127 variables are renamed to proxy_{local,external}_{address,port}.
129 JH/27 Move Internationalisation support from Experimental to mainline, enabled
130 for a build by defining SUPPORT_I18N
132 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
133 of the query string, and make ${quote_redis:} do that quoting.
135 JH/29 Move Events support from Experimental to mainline, enabled by default
136 and removable for a build by defining DISABLE_EVENT.
138 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
140 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
141 cached by the daemon.
143 JH/32 Move Redis support from Experimental to mainline, enabled for a build
144 by defining LOOKUP_REDIS. The libhiredis library is required.
146 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
147 keys are given for lookup.
149 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
150 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
151 only supported when built with TLS support. The PolarSSL SHA routines
152 are still used when the TLS library is too old for convenient support.
154 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
155 openssl_options), for security. OpenSSL forces this from version 1.1.0
156 server-side so match that on older versions.
158 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
159 allocation for $value could be released as the expansion processing
160 concluded, but leaving the global pointer active for it.
162 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
163 and to use the domains and local_parts ACL conditions.
165 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
166 incorrectly not doubled on cutthrough transmission, hence seen as a
167 body-termination at the receiving system - resulting in truncated mails.
168 Commonly the sender saw a TCP-level error, and retransmitted the message
169 via the normal store-and-forward channel. This could result in duplicates
170 received - but deduplicating mailstores were liable to retain only the
171 initial truncated version.
173 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
175 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
177 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
178 we're in there, support oversigning also; bug 1309.
180 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
182 HS/04 Add support for keep_environment and add_environment options.
184 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
185 either intentional arithmetic overflow during PRNG, or testing config-
188 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
189 delivery resulted in actual delivery. Cancel cutthrough before DATA
192 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
193 rejects a recipient: pass the reject to the originator.
195 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
196 Many were false-positives and ignorable, but it's worth fixing the
199 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
200 for the new environment-manipulation done at startup. Move the routines
201 from being local to tls.c to being global via the os.c file.
203 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
204 an extract embedded as result-arg for a map, the first arg for extract
205 is unavailable so we cannot tell if this is a numbered or keyed
206 extraction. Accept either.
212 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
215 JH/02 The smtp transport option "multi_domain" is now expanded.
217 JH/03 The smtp transport now requests PRDR by default, if the server offers
220 JH/04 Certificate name checking on server certificates, when exim is a client,
221 is now done by default. The transport option tls_verify_cert_hostnames
222 can be used to disable this per-host. The build option
223 EXPERIMENTAL_CERTNAMES is withdrawn.
225 JH/05 The value of the tls_verify_certificates smtp transport and main options
226 default to the word "system" to access the system default CA bundle.
227 For GnuTLS, only version 3.0.20 or later.
229 JH/06 Verification of the server certificate for a TLS connection is now tried
230 (but not required) by default. The verification status is now logged by
231 default, for both outbound TLS and client-certificate supplying inbound
234 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
237 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
238 Status Notification (bounce) messages are now MIME format per RFC 3464.
239 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
240 under the control of the dsn_advertise_hosts option, and routers may
241 have a dsn_lasthop option.
243 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
244 default, modifiable by a malware= option. The list separator for
245 the options can now be changed in the usual way. Bug 68.
247 JH/10 The smtp_receive_timeout main option is now expanded before use.
249 JH/11 The incoming_interface log option now also enables logging of the
250 local interface on delivery outgoing connections.
252 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
253 if the interface and destination host and port all match.
255 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
258 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
259 Patch from Andrew Lewis.
261 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
262 now supports optional time-restrictions, weighting, and priority
263 modifiers per server. Patch originally by <rommer@active.by>.
265 JH/16 The spamd_address main option now supports a mixed list of local
266 and remote servers. Remote servers can be IPv6 addresses, and
267 specify a port-range.
269 JH/17 Bug 68: The spamd_address main option now supports an optional
270 timeout value per server.
272 JH/18 Bug 1581: Router and transport options headers_add/remove can
273 now have the list separator specified.
275 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
278 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
281 JH/21 Support for the A6 type of dns record is withdrawn.
283 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
284 rather than the verbs used.
286 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
287 from 255 to 1024 chars.
289 JH/24 Verification callouts now attempt to use TLS by default.
291 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
292 are generic router options now. The defaults didn't change.
294 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
295 Original patch from Alexander Shikoff, worked over by JH.
297 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
298 files if ZCAT_COMMAND is not executable.
300 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
302 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
304 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
305 Normally benign, it bites when the pair was led to by a CNAME;
306 modern usage is to not canoicalize the domain to a CNAME target
307 (and we were inconsistent anyway for A-only vs AAAA+A).
309 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
311 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
312 when evaluating $sender_host_dnssec.
314 JH/31 Check the HELO verification lookup for DNSSEC, adding new
315 $sender_helo_dnssec variable.
317 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
319 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
321 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
323 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
324 documented as working, but never had. Support all but $spam_report.
326 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
327 added for tls authenticator.
332 TL/01 When running the test suite, the README says that variables such as
333 no_msglog_check are global and can be placed anywhere in a specific
334 test's script, however it was observed that placement needed to be near
335 the beginning for it to behave that way. Changed the runtest perl
336 script to read through the entire script once to detect and set these
337 variables, reset to the beginning of the script, and then run through
338 the script parsing/test process like normal.
340 TL/02 The BSD's have an arc4random API. One of the functions to induce
341 adding randomness was arc4random_stir(), but it has been removed in
342 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
343 function when detected.
345 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
346 cause callback expansion.
348 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
349 syntax errors in an expansion can be treated as a string instead of
350 logging or causing an error, due to the internal use of bool_lax
351 instead of bool when processing it.
353 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
354 server certificates when making smtp deliveries.
356 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
358 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
360 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
362 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
363 Merged patch from Sebastian Wiedenroth.
365 JH/05 Fix results-pipe from transport process. Several recipients, combined
366 with certificate use, exposed issues where response data items split
367 over buffer boundaries were not parsed properly. This eventually
368 resulted in duplicates being sent. This issue only became common enough
369 to notice due to the introduction of conection certificate information,
370 the item size being so much larger. Found and fixed by Wolfgang Breyha.
372 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
373 size buffer was used, resulting in syntax errors when an expansion
376 JH/07 Add support for directories of certificates when compiled with a GnuTLS
377 version 3.3.6 or later.
379 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
380 is EXPERIMENTAL_EVENT, the main-configuration and transport options
381 both become "event_action", the variables become $event_name, $event_data
382 and $event_defer_errno. There is a new variable $verify_mode, usable in
383 routers, transports and related events. The tls:cert event is now also
384 raised for inbound connections, if the main configuration event_action
387 TL/06 In test suite, disable OCSP for old versions of openssl which contained
388 early OCSP support, but no stapling (appears to be less than 1.0.0).
390 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
391 server certificate names available under the smtp transport option
392 "tls_verify_cert_hostname" now do not permit multi-component wildcard
395 JH/10 Time-related extraction expansions from certificates now use the main
396 option "timezone" setting for output formatting, and are consistent
397 between OpenSSL and GnuTLS compilations. Bug 1541.
399 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
400 encoded parameter in the incoming message. Bug 1558.
402 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
403 include certificate info, eximon was claiming there were spoolfile
406 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
408 JH/14 Log delivery-related information more consistently, using the sequence
409 "H=<name> [<ip>]" wherever possible.
411 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
412 are problematic for Debian distribution, omit them from the release
415 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
417 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
420 JH/17 Fix a null-indirection in certextract expansions when a nondefault
421 output list separator was used.
426 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
427 checkers that were complaining about end of non-void function with no
430 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
431 This was a regression intruduced in 4.83 by another bugfix.
433 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
435 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
436 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
442 TF/01 Correctly close the server side of TLS when forking for delivery.
444 When a message was received over SMTP with TLS, Exim failed to clear up
445 the incoming connection properly after forking off the child process to
446 deliver the message. In some situations the subsequent outgoing
447 delivery connection happened to have the same fd number as the incoming
448 connection previously had. Exim would try to use TLS and fail, logging
449 a "Bad file descriptor" error.
451 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
452 utilities have not been installed.
454 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
455 temporary space as the ACL may create new global variables.
457 TL/01 LDAP support uses per connection or global context settings, depending
458 upon the detected version of the libraries at build time.
460 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
461 to extract and use the src ip:port in logging and expansions as if it
462 were a direct connection from the outside internet. PPv2 support was
463 updated based on HAProxy spec change in May 2014.
465 JH/02 Add ${listextract {number}{list}{success}{fail}}.
467 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
468 Properly escape header and check for NULL return.
470 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
473 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
475 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
476 characters in header names, implemented as a verify condition.
477 Contributed by Michael Fischer v. Mollard.
479 TL/05 Rename SPF condition results err_perm and err_temp to standardized
480 results permerror and temperror. Previous values are deprecated but
481 still accepted. In a future release, err_perm and err_temp will be
482 completely removed, which will be a backward incompatibility if the
483 ACL tests for either of these two old results. Patch contributed by
484 user bes-internal on the mailing list.
486 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
488 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
489 selectors, in both main and reject logs.
491 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
494 JH/07 Add malware type "sock" for talking to simple daemon.
496 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
498 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
499 routers/transports under cutthrough routing.
501 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
502 numbers. Touch up "bool" conditional to keep the same definition.
504 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
506 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
508 JH/12 Expand items in router/transport headers_add or headers_remove lists
509 individually rather than the list as a whole. Bug 1452.
511 Required for reasonable handling of multiple headers_ options when
512 they may be empty; requires that headers_remove items with embedded
513 colons must have them doubled (or the list-separator changed).
515 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
516 view the policy declared in the DMARC record. Currently, $dmarc_status
517 is a combined value of both the record presence and the result of the
520 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
522 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
523 dnslookup router and the smtp transport (applying to the forward
526 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
527 of ldap servers used for a specific lookup. Patch provided by Heiko
530 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
531 New variable $lookup_dnssec_authenticated for observability.
533 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
534 Patch submitted by Lars Timman.
536 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
538 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
539 Requires trusted mode and valid format message id, aborts otherwise.
540 Patch contributed by Heiko Schlichting.
542 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
543 certextract with support for various fields. Bug 1358.
545 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
546 is requested by default, modifiable by smtp transport option
549 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
550 operate on certificate variables to give certificate fingerprints
551 Also new ${sha256:cert_variable}.
553 JH/23 The PRDR feature is moved from being Experimental into the mainline.
555 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
556 Christian Aistleitner.
558 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
560 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
561 file. Patch from Wolfgang Breyha.
563 JH/25 Expand the coverage of the delivery $host and $host_address to
564 client authenticators run in verify callout. Bug 1476.
566 JH/26 Port service names are now accepted for tls_on_connect_ports, to
567 align with daemon_smtp_ports. Bug 72.
569 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
570 support and error reporting did not work properly.
572 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
573 and is readable. Patch from Andrew Colin Kissa.
575 TL/14 Enhance documentation of ${run expansion and how it parses the
576 commandline after expansion, particularly in the case when an
577 unquoted variable expansion results in an empty value.
579 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
581 PP/02 Fix internal collision of T_APL on systems which support RFC3123
582 by renaming away from it. Addresses GH issue 15, reported by
585 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
587 TL/15 SECURITY: prevent double expansion in math comparison functions
588 (can expand unsanitized data). Not remotely exploitable.
595 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
597 PP/02 Make -n do something, by making it not do something.
598 When combined with -bP, the name of an option is not output.
600 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
603 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
604 $sender_host_name and config options to manage this, and basic check
607 PP/05 DSCP support for outbound connections and control modifier for inbound.
609 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
610 (Only plugin which currently uses this is kerberos4, which nobody should
611 be using, but we should make it available and other future plugins might
612 conceivably use it, even though it would break NAT; stuff *should* be
613 using channel bindings instead).
615 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
616 name; added for Sendmail compatibility; requires admin caller.
617 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
618 just ignore it); requires trusted caller.
619 Also parse but ignore: -Ac -Am -X<logfile>
622 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
624 TL/02 Add +smtp_confirmation as a default logging option.
626 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
627 Patch by Magnus Holmgren from 2007-02-20.
629 TL/04 Bugzilla 1281 - Spec typo.
630 Bugzilla 1283 - Spec typo.
631 Bugzilla 1290 - Spec grammar fixes.
633 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
635 TL/06 Add Experimental DMARC support using libopendmarc libraries.
637 TL/07 Fix an out of order global option causing a segfault. Reported to dev
638 mailing list by by Dmitry Isaikin.
640 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
642 JH/02 Support "G" suffix to numbers in ${if comparisons.
644 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
646 NM/01 Bugzilla 1197 - Spec typo
647 Bugzilla 1196 - Spec examples corrections
649 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
651 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
652 gnutls_enable_pkcs11, but renamed to more accurately indicate its
655 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
656 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
658 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
659 "acl {{name}{arg}...}", and optional args on acl condition
662 JH/05 Permit multiple router/transport headers_add/remove lines.
664 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
666 JH/07 Avoid using a waiting database for a single-message-only transport.
667 Performance patch from Paul Fisher. Bugzilla 1262.
669 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
672 JH/09 Add $headers_added variable, with content from use of ACL modifier
673 add_header (but not yet added to the message). Bugzilla 199.
675 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
676 Pulled from Bugzilla 817 by Wolfgang Breyha.
678 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
680 (nb: this is the same fix as in Exim 4.80.1)
682 JH/11 Add A= logging on delivery lines, and a client_set_id option on
685 JH/12 Add optional authenticated_sender logging to A= and a log_selector
688 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
690 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
691 advertise SMTP AUTH mechanism to us, instead of a generic
692 protocol violation error. Also, make Exim more robust to bad
693 data from the Dovecot auth socket.
695 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
697 When a queue runner is handling a message, Exim first routes the
698 recipient addresses, during which it prunes them based on the retry
699 hints database. After that it attempts to deliver the message to
700 any remaining recipients. It then updates the hints database using
703 So if a recipient address works intermittently, it can get repeatedly
704 deferred at routing time. The retry hints record remains fresh so the
705 address never reaches the final cutoff time.
707 This is a fairly common occurrence when a user is bumping up against
708 their storage quota. Exim had some logic in its local delivery code
709 to deal with this. However it did not apply to per-recipient defers
710 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
712 This change adds a proper retry rule check during routing so that the
713 final cutoff time is checked against the message's age. We only do
714 this check if there is an address retry record and there is not a
715 domain retry record; this implies that previous attempts to handle
716 the address had the retry_use_local_parts option turned on. We use
717 this as an approximation for the destination being like a local
718 delivery, as in LMTP.
720 I suspect this new check makes the old local delivery cutoff check
721 redundant, but I have not verified this so I left the code in place.
723 TF/02 Correct gecos expansion when From: is a prefix of the username.
725 Test 0254 submits a message to Exim with the header
729 When I ran the test suite under the user fanf2, Exim expanded
730 the header to contain my full name, whereas it should have added
731 a Resent-Sender: header. It erroneously treats any prefix of the
732 username as equal to the username.
734 This change corrects that bug.
736 GF/01 DCC debug and logging tidyup
737 Error conditions log to paniclog rather than rejectlog.
738 Debug lines prefixed by "DCC: " to remove any ambiguity.
740 TF/03 Avoid unnecessary rebuilds of lookup-related code.
742 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
743 Bug spotted by Jeremy Harris; was flawed since initial commit.
744 Would have resulted in OCSP responses post-SNI triggering an Exim
745 NULL dereference and crash.
747 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
749 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
750 Bug detection, analysis and fix by Samuel Thibault.
751 Bugzilla 1331, Debian bug #698092.
753 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
755 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
756 Server implementation by Todd Lyons, client by JH.
757 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
758 config variable "prdr_enable" controls whether the server
759 advertises the facility. If the client requests PRDR a new
760 acl_data_smtp_prdr ACL is called once for each recipient, after
761 the body content is received and before the acl_smtp_data ACL.
762 The client is controlled by bolth of: a hosts_try_prdr option
763 on the smtp transport, and the server advertisement.
764 Default client logging of deliveries and rejections involving
765 PRDR are flagged with the string "PRDR".
767 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
768 fclose(). Diagnosis by Todd Lyons.
770 PP/17 Update configure.default to handle IPv6 localhost better.
771 Patch by Alain Williams (plus minor tweaks).
774 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
775 This is now consistent with GnuTLS, and is now documented: the
776 previous undocumented portable approach to treating the option as
777 unset was to force an expansion failure. That still works, and
778 an empty string is now equivalent.
780 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
781 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
782 not performing validation itself.
784 PP/20 Added force_command boolean option to pipe transport.
785 Patch from Nick Koston, of cPanel Inc.
787 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
790 TF/04 Added udpsend ACL modifer and hexquote expansion operator
792 PP/21 Fix eximon continuous updating with timestamped log-files.
793 Broken in a format-string cleanup in 4.80, missed when I repaired the
794 other false fix of the same issue.
795 Report and fix from Heiko Schlichting.
798 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
799 Report from Prashanth Katuri.
801 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
802 It's SecureTransport, so affects any MacOS clients which use the
803 system-integrated TLS libraries, including email clients.
805 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
806 using a MIME ACL for non-SMTP local injection.
807 Report and assistance in diagnosis by Warren Baker.
809 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
811 JH/16 Fix comparisons for 64b. Bugzilla 1385.
813 TL/09 Add expansion variable $authenticated_fail_id to keep track of
814 last id that failed so it may be referenced in subsequent ACL's.
816 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
819 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
820 ldap library initialization, allowing self-signed CA's to be
821 used. Also properly sets require_cert option later in code by
822 using NULL (global ldap config) instead of ldap handle (per
823 session). Bug diagnosis and testing by alxgomz.
825 TL/12 Enhanced documentation in the ratelimit.pl script provided in
826 the src/util/ subdirectory.
828 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
829 renamed to Transport Post Delivery Action by Jeremy Harris, as
832 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
833 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
834 redis_servers = needs to be configured which will be used by the redis
835 lookup. Patch from Warren Baker, of The Packet Hub.
837 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
839 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
840 hostname or reverse DNS when processing a host list. Used suggestions
841 from multiple comments on this bug.
843 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
845 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
846 Missed a few lines, added it to make the runtest require no keyboard
849 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
850 contains upper case chars. Make router use caseful_local_part.
852 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
853 support when GnuTLS has been built with p11-kit.
859 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
861 This, or similar/improved, will also be change PP/11 of 4.82.
867 PP/01 Handle short writes when writing local log-files.
868 In practice, only affects FreeBSD (8 onwards).
869 Bugzilla 1053, with thanks to Dmitry Isaikin.
871 NM/01 Bugzilla 949 - Documentation tweak
873 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
876 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
878 PP/02 Implemented gsasl authenticator.
880 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
882 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
883 `pkg-config foo` for cflags/libs.
885 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
886 with rest of GSASL and with heimdal_gssapi.
888 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
889 `pkg-config foo` for cflags/libs for the TLS implementation.
891 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
892 properties get this fed in as external SSF. A number of robustness
893 and debugging improvements to the cyrus_sasl authenticator.
895 PP/08 cyrus_sasl server now expands the server_realm option.
897 PP/09 Bugzilla 1214 - Log authentication information in reject log.
898 Patch by Jeremy Harris.
900 PP/10 Added dbmjz lookup type.
902 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
904 PP/12 MAIL args handles TAB as well as SP, for better interop with
905 non-compliant senders.
906 Analysis and variant patch by Todd Lyons.
908 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
909 Bug report from Lars Müller <lars@samba.org> (via SUSE),
910 Patch from Dirk Mueller <dmueller@suse.com>
912 PP/13 tls_peerdn now print-escaped for spool files.
913 Observed some $tls_peerdn in wild which contained \n, which resulted
914 in spool file corruption.
916 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
917 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
918 or write after TLS renegotiation, which otherwise led to messages
921 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
922 as a tracking header (ie: a signed header comes before the signature).
923 Patch from Wolfgang Breyha.
925 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
926 comma-sep list; embedded commas doubled.
928 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
930 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
932 Report and patch from Dmitry Banschikov.
934 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
935 Removed SSL_clear() after SSL_new() which led to protocol negotiation
936 failures. We appear to now support TLS1.1+ with Exim.
938 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
939 lets Exim select keys and certificates based upon TLS SNI from client.
940 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
941 before an outbound SMTP session. New log_selector, +tls_sni.
943 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
944 NULL dereference. Report and patch from Alun Jones.
946 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
947 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
948 Not seeing resolver debug output on NetBSD, but suspect this is a
949 resolver implementation change.
951 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
952 Left warnings. Added "eximon gdb" invocation mode.
954 PP/21 Defaulting "accept_8bitmime" to true, not false.
956 PP/22 Added -bw for inetd wait mode support.
958 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
959 locate the relevant includes and libraries. Made this the default.
961 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
962 Bugzilla 1246, report and most of solution from Tomasz Kusy.
964 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
965 This may cause build issues on older platforms.
967 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
968 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
969 gnutls_require_mac & gnutls_require_protocols (no longer supported).
970 Added SNI support via GnuTLS too.
971 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
973 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
975 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
976 Applied second patch from Janne, implementing suggestion to default
977 multiple-strings-in-record handling to match SPF spec.
979 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
981 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
982 read-only, out of scope).
983 Patch from Wolfgang Breyha, report from Stuart Northfield.
985 PP/29 Fix three issues highlighted by clang analyser static analysis.
986 Only crash-plausible issue would require the Cambridge-specific
987 iplookup router and a misconfiguration.
988 Report from Marcin Mirosław.
990 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
992 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
993 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
994 As part of this, removing so much warning spew let me fix some minor
995 real issues in debug logging.
997 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
998 assignment on my part. Fixed.
1000 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1001 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1002 Janne Snabb (who went above and beyond: thank you).
1004 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1005 string otherwise requires a connection and a bunch more work and it's
1006 relatively easy to get wrong. Should also expose TLS library linkage
1009 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1010 64-bit ${eval} (JH/03).
1012 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1013 GNU libc to support some of the 64-bit stuff, should not lead to
1014 conflicts. Defined before os.h is pulled in, so if a given platform
1015 needs to override this, it can.
1017 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1018 protection layer was required, which is not implemented.
1019 Bugzilla 1254, patch from Wolfgang Breyha.
1021 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1022 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1023 tls_dhparam take prime identifiers. Also unbreak combination of
1024 OpenSSL+DH_params+TLSSNI.
1026 PP/39 Disable SSLv2 by default in OpenSSL support.
1032 PP/01 Solaris build fix for Oracle's LDAP libraries.
1033 Bugzilla 1109, patch from Stephen Usher.
1035 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1037 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1040 TF/02 Fix a couple more cases where we did not log the error message
1041 when unlink() failed. See also change 4.74-TF/03.
1043 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1044 lock up or crash if it happened to be inside a call to libc when it
1045 got a SIGUSR1 from exiwhat.
1047 The SIGUSR1 handler appends the current process status to the process
1048 log which is later printed by exiwhat. It used to use the general
1049 purpose logging code to do this, but several functions it calls are
1050 not safe for signals.
1052 The new output code in the SIGUSR1 handler is specific to the process
1053 log, and simple enough that it's easy to inspect for signal safety.
1054 Removing some special cases also simplifies the general logging code.
1055 Removing the spurious timestamps from the process log simplifies
1058 TF/04 Improved ratelimit ACL condition.
1060 The /noupdate option has been deprecated in favour of /readonly which
1061 has clearer semantics. The /leaky, /strict, and /readonly update modes
1062 are mutually exclusive. The update mode is no longer included in the
1063 database key; it just determines when the database is updated. (This
1064 means that when you upgrde Exim will forget old rate measurements.)
1066 Exim now checks that the per_* options are used with an update mode that
1067 makes sense for the current ACL. For example, when Exim is processing a
1068 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1069 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1070 must specify per_mail/readonly. If you omit the update mode it defaults to
1071 /leaky where that makes sense (as before) or /readonly where required.
1073 The /noupdate option is now undocumented but still supported for
1074 backwards compatibility. It is equivalent to /readonly except that in
1075 ACLs where /readonly is required you may specify /leaky/noupdate or
1076 /strict/noupdate which are treated the same as /readonly.
1078 A useful new feature is the /count= option. This is a generalization
1079 of the per_byte option, so that you can measure the throughput of other
1080 aggregate values. For example, the per_byte option is now equivalent
1081 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1083 The per_rcpt option has been generalized using the /count= mechanism
1084 (though it's more complicated than the per_byte equivalence). When it is
1085 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1086 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1087 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1088 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1089 Note that using per_rcpt with a non-readonly update mode in more than
1090 one ACL will cause the recipients to be double-counted. (The per_mail
1091 and per_byte options don't have this problem.)
1093 The handling of very low rates has changed slightly. If the computed rate
1094 is less than the event's count (usually one) then this event is the first
1095 after a long gap. In this case the rate is set to the same as this event's
1096 count, so that the first message of a spam run is counted properly.
1098 The major new feature is a mechanism for counting the rate of unique
1099 events. The new per_addr option counts the number of different
1100 recipients that someone has sent messages to in the last time period. It
1101 behaves like per_rcpt if all the recipient addresses are different, but
1102 duplicate recipient addresses do not increase the measured rate. Like
1103 the /count= option this is a general mechanism, so the per_addr option
1104 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1105 example, measure the rate that a client uses different sender addresses
1106 with the options per_mail/unique=$sender_address. There are further
1107 details in the main documentation.
1109 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1111 TF/06 Removed a few PCRE remnants.
1113 TF/07 Automatically extract Exim's version number from tags in the git
1114 repository when doing development or release builds.
1116 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1117 Bugzilla 879. Patch from Paul Fisher.
1119 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1120 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1123 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1125 PP/05 Variable $av_failed, true if the AV scanner deferred.
1126 Bugzilla 1078. Patch from John Horne.
1128 PP/06 Stop make process more reliably on build failure.
1129 Bugzilla 1087. Patch from Heiko Schlittermann.
1131 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1132 Bugzilla 1089. Patch from Heiko Schlittermann.
1134 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1135 Bugzilla 1131. Patch from Holger Weiß.
1137 PP/09 Handle IPv6 addresses with SPF.
1138 Bugzilla 860. Patch from Wolfgang Breyha.
1140 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1142 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1145 PP/11 match_* no longer expand right-hand-side by default.
1146 New compile-time build option, EXPAND_LISTMATCH_RHS.
1147 New expansion conditions, "inlist", "inlisti".
1149 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1151 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1153 PP/14 fix log_write() format string regression from TF/03.
1154 Bugzilla 1152. Patch from Dmitry Isaikin.
1160 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1162 PP/02 Harmonised TLS library version reporting; only show if debugging.
1163 Layout now matches that introduced for other libraries in 4.74 PP/03.
1165 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1167 PP/04 New "dns_use_edns0" global option.
1169 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1172 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1173 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1175 TK/01 Updated PolarSSL code to 0.14.2.
1176 Bugzilla 1097. Patch from Andreas Metzler.
1178 PP/07 Catch divide-by-zero in ${eval:...}.
1179 Fixes bugzilla 1102.
1181 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1184 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1185 format-string attack -- SECURITY: remote arbitrary code execution.
1187 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1188 time unintentionally subject to list matching rules, letting the header
1189 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1190 arbitrary string expansion). This allowed for information disclosure.
1192 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1193 INT_MIN/-1 -- value coerced to INT_MAX.
1199 NM/01 Workround for PCRE version dependency in version reporting
1202 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1203 This fixes portability to compilers other than gcc, notably
1204 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1206 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1207 makefiles for portability to HP-UX and POSIX correctness.
1209 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1210 Also via indented variable definition in the Makefile.
1211 (Debugging by Oliver Heesakkers).
1213 PP/02 Restore caching of spamd results with expanded spamd_address.
1214 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1216 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1217 Improves build reliability. Fix from: Frank Elsner
1219 NM/02 Fix wide character breakage in the rfc2047 coding
1220 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1222 NM/03 Allow underscore in dnslist lookups
1223 Fixes bug 1026. Patch from Graeme Fowler
1225 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1226 Code patches from Adam Ciarcinski of NetBSD.
1228 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1231 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1232 is logged, to avoid truncation. Patch from John Horne.
1234 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1235 Patch from Jakob Hirsch.
1237 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1238 SQL string expansion failure details.
1239 Patch from Andrey Oktyabrski.
1241 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1242 Patch from Simon Arlott.
1244 PP/09 New lookups functionality failed to compile on old gcc which rejects
1245 extern declarations in function scope.
1246 Patch from Oliver Fleischmann
1248 PP/10 Use sig_atomic_t for flags set from signal handlers.
1249 Check getgroups() return and improve debugging.
1250 Fixed developed for diagnosis in bug 927 (which turned out to be
1253 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1254 Patch from Mark Zealey.
1256 PP/12 Bugzilla 1056: Improved spamd server selection.
1257 Patch from Mark Zealey.
1259 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1260 Based on patch from Heiko Schlittermann.
1262 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1263 Patch from Uwe Doering, sign-off by Michael Haardt.
1265 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1266 variable declaration deep within a block. Bug and patch from
1269 PP/15 lookups-Makefile IRIX compatibilty coercion.
1271 PP/16 Make DISABLE_DKIM build knob functional.
1273 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1274 Patch by Simon Arlott
1276 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1277 variable argument macros. Our copy now differs from upstream.
1283 TF/01 Failure to get a lock on a hints database can have serious
1284 consequences so log it to the panic log.
1286 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1287 controlled using the smtp_confirmation log selector.
1289 TF/03 Include the error message when we fail to unlink a spool file.
1291 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1292 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1293 for maintaining out-of-tree patches for some time.
1295 PP/01 Bugzilla 139: Documentation and portability issues.
1296 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1297 Handle per-OS dynamic-module compilation flags.
1299 PP/02 Let /dev/null have normal permissions.
1300 The 4.73 fixes were a little too stringent and complained about the
1301 permissions on /dev/null. Exempt it from some checks.
1302 Reported by Andreas M. Kirchwitz.
1304 PP/03 Report version information for many libraries, including
1305 Exim version information for dynamically loaded libraries. Created
1306 version.h, now support a version extension string for distributors
1307 who patch heavily. Dynamic module ABI change.
1309 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1310 privilege escalation vulnerability whereby the Exim run-time user
1311 can cause root to append content of the attacker's choosing to
1314 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1317 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1318 If dropping privileges for untrusted macros, we disabled normal logging
1319 on the basis that it would fail; for the Exim run-time user, this is not
1320 the case, and it resulted in successful deliveries going unlogged.
1321 Fixed. Reported by Andreas Metzler.
1327 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1328 only prepend for the Resent-* case. Fixes regression introduced in
1329 Exim 4.70 by NM/22 for Bugzilla 607.
1331 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1332 increasing numbers of administrators be bitten by this.
1334 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1336 PP/03 Bugzilla 994: added openssl_options main configuration option.
1338 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1340 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1342 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1344 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1345 without a peer certificate, leading to a segfault because of an
1346 assumption that peers always have certificates. Be a little more
1347 paranoid. Problem reported by Martin Tscholak.
1349 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1350 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1351 NB: ClamAV planning to remove STREAM in "middle of 2010".
1352 CL also introduces -bmalware, various -d+acl logging additions and
1353 more caution in buffer sizes.
1355 PP/09 Implemented reverse_ip expansion operator.
1357 PP/10 Bugzilla 937: provide a "debug" ACL control.
1359 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1361 PP/12 Bugzilla 973: Implement --version.
1363 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1365 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1367 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1369 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1370 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1371 ignore trailing whitespace.
1373 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1375 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1378 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1379 Notification from Dr Andrew Aitchison.
1381 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1382 ExtendedDetectionInfo response format.
1383 Notification from John Horne.
1385 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1388 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1389 XSL and documented dependency on system catalogs, with examples of how
1392 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1395 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1396 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1397 configuration file which is writeable by the Exim user or group.
1399 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1400 of configuration files to cover files specified with the -C option if
1401 they are going to be used with root privileges, not just the default
1404 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1405 option (effectively making it always true).
1407 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1408 files to be used while preserving root privileges.
1410 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1411 that rogue child processes cannot use them.
1413 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1414 run-time user, instead of root.
1416 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1417 Exim run-time user without dropping privileges.
1419 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1420 result string, instead of calling string_vformat() twice with the same
1423 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1424 for other users. Others should always drop root privileges if they use
1425 -C on the command line, even for a whitelisted configure file.
1427 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1429 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1435 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1436 $data_path, and $header_path variables; fixed documentation bugs and
1439 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1440 exipick to access non-standard spools, including the "frozen" queue
1443 NM/01 Bugzilla 965: Support mysql stored procedures.
1444 Patch from Alain Williams
1446 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1448 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1449 Patch from Andreas Metzler
1451 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1452 Patch from Kirill Miazine
1454 NM/05 Bugzilla 671: Added umask to procmail example.
1456 JJ/03 installed exipick 20100323.0, fixing doc bug
1458 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1459 directory. Notification and patch from Dan Rosenberg.
1461 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1463 TK/02 Improve log output when DKIM signing operation fails.
1465 MH/01 Treat the transport option dkim_domain as a colon separated
1466 list, not as a single string, and sign the message with each element,
1467 omitting multiple occurences of the same signer.
1469 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1470 Bugzilla 985, 986. Patch by Simon Arlott
1472 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1473 Patch by Simon Arlott
1475 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1476 MBX locking. Notification from Dan Rosenberg.
1482 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1484 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1486 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1488 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1490 NM/04 Bugzilla 915: Flag broken perl installation during build.
1496 TK/01 Added patch by Johannes Berg that expands the main option
1497 "spamd_address" if it starts with a dollar sign.
1499 TK/02 Write list of recipients to X-Envelope-Sender header when building
1500 the mbox-format spool file for content scanning (suggested by Jakob
1503 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1504 (http://www.dcc-servers.net/) support via dccifd. Activated by
1505 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1507 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1508 by Mark Daniel Reidel <mr@df.eu>.
1510 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1511 When building exim an external PCRE library is now needed -
1512 PCRE is a system library on the majority of modern systems.
1513 See entry on PCRE_LIBS in EDITME file.
1515 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1516 conversation. Added nologin parameter to request.
1517 Patch contributed by Kirill Miazine.
1519 TF/01 Do not log submission mode rewrites if they do not change the address.
1521 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1523 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1524 log files in place. Contributed by Roberto Lima.
1526 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1528 TF/03 Bugzilla 615: When checking the local_parts router precondition
1529 after a local_part_suffix or local_part_prefix option, Exim now
1530 does not use the address's named list lookup cache, since this
1531 contains cached lookups for the whole local part.
1533 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1534 Robert Millan. Documentation is in experimental-spec.txt.
1536 TF/04 Bugzilla 668: Fix parallel build (make -j).
1538 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1540 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1541 Patch provided by Jan Srzednicki.
1543 TF/05 Leading white space used to be stripped from $spam_report which
1544 wrecked the formatting. Now it is preserved.
1546 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1547 that they are available at delivery time.
1549 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1551 TF/08 TLS error reporting now respects the incoming_interface and
1552 incoming_port log selectors.
1554 TF/09 Produce a more useful error message if an SMTP transport's hosts
1555 setting expands to an empty string.
1557 NM/06 Bugzilla 744: EXPN did not work under TLS.
1558 Patch provided by Phil Pennock.
1560 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1561 Patch provided by Richard Godbee.
1563 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1564 acl_smtp_notquit, added index entry.
1566 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1567 Patch provided by Eugene Bujak.
1569 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1570 accept(). Patch provided by Maxim Dounin.
1572 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1573 Patch provided by Phil Pennock.
1575 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1577 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1578 Patch provided by Brad "anomie" Jorsch.
1580 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1581 Patch provided by Dean Brooks.
1583 TK/05 Add native DKIM support (does not depend on external libraries).
1585 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1586 Patch provided by Graeme Fowler.
1588 NM/16 Bugzilla 851: Documentation example syntax fix.
1590 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1592 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1595 NM/19 Bugzilla 745: TLS version reporting.
1596 Patch provided by Phil Pennock.
1598 NM/20 Bugzilla 167: bool: condition support.
1599 Patch provided by Phil Pennock.
1601 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1602 clients. Patch provided by Phil Pennock.
1604 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1605 Patch provided by Brad "anomie" Jorsch.
1607 NM/23 Bugzilla 687: Fix misparses in eximstats.
1608 Patch provided by Heiko Schlittermann.
1610 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1611 Patch provided by Heiko Schlittermann.
1613 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1614 plus update to original patch.
1616 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1618 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1619 Patch provided by David Brownlee.
1621 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1623 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1625 NM/30 Bugzilla 888: TLS documentation bugfixes.
1627 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1629 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1630 Unlike the original bugzilla I have changed all shell scripts in src tree.
1632 NM/33 Bugzilla 898: Transport filter timeout fix.
1633 Patch by Todd Rinaldo.
1635 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1636 Patch by Serge Demonchaux.
1638 NM/35 Bugzilla 39: Base64 decode bug fixes.
1639 Patch by Jakob Hirsch.
1641 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1643 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1645 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1647 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1653 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1654 ALT-N's libdkim that I have put here:
1655 http://duncanthrax.net/exim-experimental/
1657 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1658 were called 'true' and it seems that C99 defines that as a reserved
1659 keyword to be used with 'bool' variable types. That means you could
1660 not include C99-style headers which use bools without triggering
1661 build errors in sieve.c.
1663 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1664 as mailq or other aliases. Changed the --help handling significantly
1665 to do whats expected. exim_usage() emits usage/help information.
1667 SC/01 Added the -bylocaldomain option to eximstats.
1669 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1671 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1673 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1679 PH/01 Another patch from the Sieve maintainer.
1681 PH/02 When an IPv6 address is converted to a string for single-key lookup
1682 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1683 dots are used instead of colons so that keys in lsearch files need not
1684 contain colons. This was done some time before quoting was made available
1685 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1686 (notated using the quote facility) so as to distinguish them from IPv4
1687 keys. This meant that lookups for IP addresses in host lists did not work
1688 for iplsearch lookups.
1690 This has been fixed by arranging for IPv6 addresses to be expressed with
1691 colons if the lookup type is iplsearch. This is not incompatible, because
1692 previously such lookups could never work.
1694 The situation is now rather anomolous, since one *can* have colons in
1695 ordinary lsearch keys. However, making the change in all cases is
1696 incompatible and would probably break a number of configurations.
1698 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1701 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1702 conversion specification without a maximum field width, thereby enabling
1703 a rogue spamd server to cause a buffer overflow. While nobody in their
1704 right mind would setup Exim to query an untrusted spamd server, an
1705 attacker that gains access to a server running spamd could potentially
1706 exploit this vulnerability to run arbitrary code as the Exim user.
1708 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1709 $primary_hostname instead of what libspf2 thinks the hosts name is.
1711 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1712 a directory entry by the name of the lookup key. Previously, if a
1713 symlink pointed to a non-existing file or a file in a directory that
1714 Exim lacked permissions to read, a lookup for a key matching that
1715 symlink would fail. Now it is enough that a matching directory entry
1716 exists, symlink or not. (Bugzilla 503.)
1718 PH/03 The body_linecount and body_zerocount variables are now exported in the
1721 PH/04 Added the $dnslist_matched variable.
1723 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1724 This means they are set thereafter only if the connection becomes
1727 PH/06 Added the client_condition to authenticators so that some can be skipped
1728 by clients under certain conditions.
1730 PH/07 The error message for a badly-placed control=no_multiline_responses left
1731 "_responses" off the end of the name.
1733 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1735 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1736 (without spaces) instead of just copying the configuration text.
1738 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1740 PH/11 Added $max_received_linelength.
1742 PH/12 Added +ignore_defer and +include_defer to host lists.
1744 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1745 way in which PCRE > 7.0 is built.
1747 PH/14 Implemented queue_only_load_latch.
1749 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1750 MAIL command. The effect was to mangle the value on 64-bit systems.
1752 PH/16 Another patch from the Sieve maintainer.
1754 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1756 PH/18 If a system quota error occurred while trying to create the file for
1757 a maildir delivery, the message "Mailbox is full" was not appended to the
1758 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1759 only to a quota excession during the actual writing of the file.
1761 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1762 characters?) which causes problems in log lines. The DN values are now
1763 passed through string_printing() before being added to log lines.
1765 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1766 and InterBase are left for another time.)
1768 PH/21 Added message_body_newlines option.
1770 PH/22 Guard against possible overflow in moan_check_errorcopy().
1772 PH/23 POSIX allows open() to be a macro; guard against that.
1774 PH/24 If the recipient of an error message contained an @ in the local part
1775 (suitably quoted, of course), incorrect values were put in $domain and
1776 $local_part during the evaluation of errors_copy.
1782 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1783 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1786 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1787 issue a MAIL command.
1789 PH/02 In an ACL statement such as
1791 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1793 if a client was not listed at all, or was listed with a value other than
1794 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1795 the condition was not true (as it should be), so access was not denied.
1796 The bug was that the ! inversion was incorrectly passed on to the second
1797 item. This has been fixed.
1799 PH/03 Added additional dnslists conditions == and =& which are different from
1800 = and & when the dns lookup returns more than one IP address.
1802 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1803 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1805 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1806 FSYNC, which compiles an option called disable_fsync that allows for
1807 bypassing fsync(). The documentation is heavily laced with warnings.
1809 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1811 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1812 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1813 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1814 including adding "make clean"; (3) Added -fPIC when compiling the test
1815 dynamically loaded module, to get rid of a warning.
1817 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1818 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1819 = 0s. The bug is otherwise harmless.
1821 PH/07 There was a bug in the dovecot authenticator such that the value of
1822 $auth1 could be overwritten, and so not correctly preserved, after a
1823 successful authentication. This usually meant that the value preserved by
1824 the server_setid option was incorrect.
1826 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1828 PH/09 Installed PCRE release 7.0.
1830 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1831 run for batched SMTP input. It is now run at the start of every message
1832 in the batch. While fixing this I discovered that the process information
1833 (output by running exiwhat) was not always getting set for -bs and -bS
1834 input. This is fixed, and it now also says "batched" for BSMTP.
1836 PH/11 Added control=no_pipelining.
1838 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1839 patch, slightly modified), and move the expansion of helo_data till after
1840 the connection is made in the smtp transport (so it can use these
1843 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1845 PH/14 Added log_selector = +pid.
1847 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1849 PH/16 Add ${if forany and ${if forall.
1851 PH/17 Added dsn_from option to vary the From: line in DSNs.
1853 PH/18 Flush SMTP output before performing a callout, unless control =
1854 no_callout_flush is set.
1856 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1857 was true (the default) a successful delivery failed to delete the retry
1858 item, thus causing premature timeout of the address. The bug is now
1861 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1863 PH/21 Long custom messages for fakedefer and fakereject are now split up
1864 into multiline reponses in the same way that messages for "deny" and
1865 other ACL rejections are.
1867 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1868 with slight modification.
1870 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1871 draft, changing the syntax and factoring some duplicate code".
1873 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1874 for deliveries of the second and subsequent messages over the same SMTP
1877 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1878 ${reduce, with only minor "tidies".
1880 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1882 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1883 expansion side effects.
1885 PH/27 When a message times out after an over-quota error from an Exim-imposed
1886 quota, the bounce message says "mailbox is full". This message was not
1887 being given when it was a system quota that was exceeded. It now should
1890 MH/03 Made $recipients available in local_scan(). local_scan() already has
1891 better access to the recipient list through recipients_list[], but
1892 $recipients can be useful in postmaster-provided expansion strings.
1894 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1895 in the case of a MAIL command with additional options following the
1896 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1897 were accidentally chopped off.
1899 PH/29 SMTP synchronization checks are implemented when a command is read -
1900 there is a check that no more input is waiting when there shouldn't be
1901 any. However, for some commands, a delay in an ACL can mean that it is
1902 some time before the response is written. In this time, more input might
1903 arrive, invalidly. So now there are extra checks after an ACL has run for
1904 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1905 pipelining has not been advertised.
1907 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1909 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1910 returned in the SMTP response when smtp_return_error_details was false.
1911 This has been fixed.
1913 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1914 instead of the C I/O that was originally supplied, because problems were
1915 reported on Solaris.
1917 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1918 Exim which did not show up earlier: it was assuming that a call to
1919 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1920 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1921 was a macro that became an assignment, so it seemed to work. This has
1922 changed to a proper function call with a void return, hence the compile
1923 error. Exim's code has been fixed.
1925 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1928 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1930 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1932 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1933 because it turns out to be too big or there is a timeout) there is no
1934 "Completed" line in the log. When some messages of this type were
1935 selected by exigrep, they were listed as "not completed". Others were
1936 picked up by some special patterns. I have improved the selection
1937 criteria to be more general.
1939 PH/38 The host_find_failed option in the manualroute router can now be set
1940 to "ignore", to completely ignore a host whose IP address cannot be
1941 found. If all hosts are ignored, the behaviour is controlled by the new
1942 host_all_ignored option.
1944 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1945 homing or because of multiple MX records with /mx) generated more than
1946 one IP address, and the following item turned out to be the local host,
1947 all the secondary addresses of the first item were incorrectly removed
1948 from the list, along with the local host and any following hosts (which
1949 is what is supposed to happen).
1951 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1952 whoever called Exim into the -H file. In the case of the daemon it was
1953 behaving confusingly. When first started, it used values for whoever
1954 started the daemon, but after a SIGHUP it used the Exim user (because it
1955 calls itself on a restart). I have changed the code so that it now always
1958 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1959 message are rejected with the same error (e.g. no authentication or bad
1960 sender address), and a DATA command is nevertheless sent (as can happen
1961 with PIPELINING or a stupid MUA), the error message that was given to the
1962 RCPT commands is included in the rejection of the DATA command. This is
1963 intended to be helpful for MUAs that show only the final error to their
1966 PH/42 Another patch from the Sieve maintainer.
1968 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1969 Eximstats - Fixed some broken HTML links and added missing column headers
1971 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1972 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1974 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1976 PH/43 Yet another patch from the Sieve maintainer.
1978 PH/44 I found a way to check for a TCP/IP connection going away before sending
1979 the response to the final '.' that terminates a message, but only in the
1980 case where the client has not sent further data following the '.'
1981 (unfortunately, this is allowed). However, in many cases there won't be
1982 any further data because there won't be any more messages to send. A call
1983 to select() can be used: if it shows that the input is "ready", there is
1984 either input waiting, or the socket has been closed. An attempt to read
1985 the next input character can distinguish the two cases. Previously, Exim
1986 would have sent an OK response which the client would never have see.
1987 This could lead to message repetition. This fix should cure that, at
1988 least in a lot of common cases.
1990 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1991 advertised in response to EHLO.
1997 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1998 fixed by 4.65/MH/01 (is this a record?) are fixed:
2000 (i) An empty string was always treated as zero by the numeric comparison
2001 operators. This behaviour has been restored.
2003 (ii) It is documented that the numeric comparison operators always treat
2004 their arguments as decimal numbers. This was broken in that numbers
2005 starting with 0 were being interpreted as octal.
2007 While fixing these problems I realized that there was another issue that
2008 hadn't been noticed. Values of message_size_limit (both the global option
2009 and the transport option) were treated as octal if they started with 0.
2010 The documentation was vague. These values are now always treated as
2011 decimal, and I will make that clear in the documentation.
2017 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2018 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2021 MH/01 Don't check that the operands of numeric comparison operators are
2022 integers when their expansion is in "skipping" mode (fixes bug
2023 introduced by 4.64-PH/07).
2025 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2026 child addresses, Exim now panics and dies. Previously, because the count
2027 is held in a short int, deliveries were likely to be lost. As such a
2028 large number of recipients for a single message is ridiculous
2029 (performance will be very, very poor), I have chosen to impose a limit
2030 rather than extend the field.
2036 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2037 leftover -K file (the existence of which was triggered by #402).
2038 While we were at it, introduced process PID as part of the -K
2039 filename. This should rule out race conditions when creating
2042 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2043 processing considerably. Previous code took too long for large mails,
2044 triggering a timeout which in turn triggers #401.
2046 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2047 in the DK code in transports.c. sendfile() is not really portable,
2048 hence the _LINUX specificness.
2050 TF/01 In the add_headers option to the mail command in an Exim filter,
2051 there was a bug that Exim would claim a syntax error in any
2052 header after the first one which had an odd number of characters
2055 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2056 callout verification, Exim cached a "reject" for the entire domain. This
2057 is correct for most verifications, but it is not correct for a recipient
2058 verification with use_sender or use_postmaster set, because in that case
2059 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2060 case of MAIL FROM:<> rejection from other early rejections (e.g.
2061 rejection of HELO). When verifying a recipient using a non-null MAIL
2062 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2063 Whatever the result of the callout, the value of the domain cache is
2064 left unchanged (for any other kind of callout, getting as far as trying
2065 RCPT means that the domain itself is ok).
2067 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2070 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2071 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2072 session, the daemon ignores ECONNECT errors and logs others; it now
2073 ignores EPIPE as well.
2075 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2076 (quoted-printable decoding).
2078 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2079 later the small subsequent patch to fix an introduced bug.
2081 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2083 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2085 PH/08 An error is now given if message_size_limit is specified negative.
2087 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2088 to be given (somewhat) arbitrary names.
2090 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2093 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2094 miscellaneous code fixes
2096 PH/10 Added the log_reject_target ACL modifier to specify where to log
2099 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2100 hostname. This is wrong, because it relates to the incoming message (and
2101 probably the interface on which it is arriving) and not to the outgoing
2102 callout (which could be using a different interface). This has been
2103 changed to use the value of the helo_data option from the smtp transport
2104 instead - this is what is used when a message is actually being sent. If
2105 there is no remote transport (possible with a router that sets up host
2106 addresses), $smtp_active_hostname is used.
2108 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2109 tweaks were necessary in order to get it to work (see also 21 below):
2110 (a) The code assumed that strncpy() returns a negative number on buffer
2111 overflow, which isn't the case. Replaced with Exim's string_format()
2113 (b) There were several signed/unsigned issues. I just did the minimum
2114 hacking in of casts. There is scope for a larger refactoring.
2115 (c) The code used strcasecmp() which is not a standard C function.
2116 Replaced with Exim's strcmpic() function.
2117 (d) The code set only $1; it now sets $auth1 as well.
2118 (e) A simple test gave the error "authentication client didn't specify
2119 service in request". It would seem that Dovecot has changed its
2120 interface. Fortunately there's a specification; I followed it and
2121 changed what the client sends and it appears to be working now.
2123 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2126 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2127 address A is aliased to B and C, where B exists and C does not. Without
2128 -v the output is "A verified" because verification stops after a
2129 successful redirection if more than one address is generated. However,
2130 with -v the child addresses are also verified. Exim was outputting "A
2131 failed to verify" and then showing the successful verification for C,
2132 with its parentage. It now outputs "B failed to verify", showing B's
2133 parentage before showing the successful verification of C.
2135 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2136 look up a TXT record in a specific list after matching in a combined
2139 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2140 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2141 they consult the DNS. I had assumed they would set it the way they
2142 wanted; and indeed my experiments on Linux seem to show that in some
2143 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2144 To be on the safe side, however, I have now made the interface to
2145 host_find_byname() similar to host_find_bydns(), with an argument
2146 containing the DNS resolver options. The host_find_byname() function now
2147 sets these options at its start, just as host_find_bydns() does. The smtp
2148 transport options dns_qualify_single and dns_search_parents are passed to
2149 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2150 of host_find_byname() use the default settings of RES_DEFNAMES
2151 (qualify_single) but not RES_DNSRCH (search_parents).
2153 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2154 spool_read_header() do less string testing, by means of a preliminary
2155 switch on the second character of optional "-foo" lines. (This is
2156 overdue, caused by the large number of possibilities that now exist.
2157 Originally there were few.) While I was there, I also converted the
2158 str(n)cmp tests so they don't re-test the leading "-" and the first
2159 character, in the hope this might squeeze out yet more improvement.
2161 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2162 flag allowing group syntax was set by the header_syntax check but not
2163 turned off, possible causing trouble later; (2) The flag was not being
2164 set at all for the header_verify test, causing "group"-style headers to
2165 be rejected. I have now set it in this case, and also caused header_
2166 verify to ignore an empty address taken from a group. While doing this, I
2167 came across some other cases where the code for allowing group syntax
2168 while scanning a header line wasn't quite right (mostly, not resetting
2169 the flag correctly in the right place). These bugs could have caused
2170 trouble for malformed header lines. I hope it is now all correct.
2172 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2173 with the "reply" argument non-NULL. The code, however (which originally
2174 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2175 but it didn't always do it. This confused somebody who was copying the
2176 code for some other use. I have removed all the tests.
2178 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2179 feature that was used to support insecure browsers during the U.S. crypto
2180 embargo. It requires special client support, and Exim is probably the
2181 only MTA that supported it -- and would never use it because real RSA is
2182 always available. This code has been removed, because it had the bad
2183 effect of slowing Exim down by computing (never used) parameters for the
2184 RSA_EXPORT functionality.
2186 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2187 authenticator to fail if there's a tab character in the incoming data
2188 (there should never be unless someone is messing about, as it's supposed
2189 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2190 if the connection is using TLS or if the remote IP is the same as the
2191 local IP, and the "valid-client-cert option" if a client certificate has
2194 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2195 authenticators. This can be used for authorization after authentication
2196 succeeds. (In the case of plaintext, it servers for both authentication
2199 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2200 if any retry times were supplied.
2202 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2203 connection, where there is no client IP address to check. In this
2204 situation, the verify now always succeeds.
2206 PH/25 Applied John Jetmore's -Mset patch.
2208 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2210 PH/27 In a string expansion for a processed (not raw) header when multiple
2211 headers of the same name were present, leading whitespace was being
2212 removed from all of them, but trailing whitespace was being removed only
2213 from the last one. Now trailing whitespace is removed from each header
2214 before concatenation. Completely empty headers in a concatenation (as
2215 before) are ignored.
2217 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2218 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2220 PH/29 [Removed. This was a change that I later backed out, and forgot to
2221 correct the ChangeLog entry (that I had efficiently created) before
2222 committing the later change.]
2224 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2225 address errors (4xx response to RCPT) over the same connection as other
2226 messages routed to the same hosts. Such deliveries are always "forced",
2227 so retry times are not inspected. This resulted in far too many retries
2228 for the affected addresses. The effect occurred only when there were more
2229 hosts than the hosts_max_try setting in the smtp transport when it had
2230 the 4xx errors. Those hosts that it had tried were not added to the list
2231 of hosts for which the message was waiting, so if all were tried, there
2232 was no problem. Two fixes have been applied:
2234 (i) If there are any address or message errors in an SMTP delivery, none
2235 of the hosts (tried or untried) are now added to the list of hosts
2236 for which the message is waiting, so the message should not be a
2237 candidate for sending over the same connection that was used for a
2238 successful delivery of some other message. This seems entirely
2239 reasonable: after all the message is NOT "waiting for some host".
2240 This is so "obvious" that I'm not sure why it wasn't done
2241 previously. Hope I haven't missed anything, but it can't do any
2242 harm, as the worst effect is to miss an optimization.
2244 (ii) If, despite (i), such a delivery is accidentally attempted, the
2245 routing retry time is respected, so at least it doesn't keep
2246 hammering the server.
2248 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2249 in ${readsocket because some servers need this prod.
2251 PH/32 Added some extra debug output when updating a wait-xxx database.
2253 PH/33 The hint "could be header name not terminated by colon", which has been
2254 given for certain expansion errors for a long time, was not being given
2255 for the ${if def:h_colon_omitted{... case.
2257 PH/34 The spec says: "With one important exception, whenever a domain list is
2258 being scanned, $domain contains the subject domain." There was at least
2259 one case where this was not true.
2261 PH/35 The error "getsockname() failed: connection reset by peer" was being
2262 written to the panic log as well as the main log, but it isn't really
2263 panic-worthy as it just means the connection died rather early on. I have
2264 removed the panic log writing for the ECONNRESET error when getsockname()
2267 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2268 runs only) independently of the message's sender address. This meant
2269 that, if the 4xx error was in fact related to the sender, a different
2270 message to the same recipient with a different sender could confuse
2271 things. In particualar, this can happen when sending to a greylisting
2272 server, but other circumstances could also provoke similar problems.
2273 I have changed the default so that the retry time for these errors is now
2274 based a combination of the sender and recipient addresses. This change
2275 can be overridden by setting address_retry_include_sender=false in the
2278 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2279 remote server are returned as part of bounce messages. This was not
2280 happening for LMTP over a pipe (the lmtp transport), but now it is the
2281 same for both kinds of LMTP.
2283 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2284 in header lines that were in fact CNAMEs. This is no longer the case.
2286 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2287 and queue runs started by the daemon processed all messages. This has
2288 been fixed so that -R and -S can now usefully be given with -q<time>.
2290 PH/40 Import PCRE release 6.7 (fixes some bugs).
2292 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2294 PH/42 Give an error if -q is specified more than once.
2296 PH/43 Renamed the variables $interface_address and $interface_port as
2297 $received_ip_address and $received_port, to make it clear that these
2298 values apply to message reception, and not to the outgoing interface when
2299 a message is delivered. (The old names remain recognized, of course.)
2301 PH/44 There was no timeout on the connect() call when using a Unix domain
2302 socket in the ${readsocket expansion. There now is.
2304 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2305 be meaningful with "accept".
2307 SC/01 Eximstats V1.43
2308 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2310 SC/02 Eximstats V1.44
2311 Use a glob alias rather than an array ref in the generated
2312 parser. This improves both readability and performance.
2314 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2315 Collect SpamAssassin and rejection statistics.
2316 Don't display local sender or destination tables unless
2317 there is data to show.
2318 Added average volumes into the top table text output.
2320 SC/04 Eximstats V1.46
2321 Collect data on the number of addresses (recipients)
2322 as well as the number of messages.
2324 SC/05 Eximstats V1.47
2325 Added 'Message too big' to the list of mail rejection
2326 reasons (thanks to Marco Gaiarin).
2328 SC/06 Eximstats V1.48
2329 Mainlog lines which have GMT offsets and are too short to
2330 have a flag are now skipped.
2332 SC/07 Eximstats V1.49 (Alain Williams)
2333 Added the -emptyok flag.
2335 SC/08 Eximstats V1.50
2336 Fixes for obtaining the IP address from reject messages.
2338 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2339 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2340 whitesspace changes from 4.64-PH/27
2342 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2345 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2346 are found, allow negative numbers in numeric criteria)
2348 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2350 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2351 to match changes made in 4.64-PH/43
2353 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2355 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2356 "domain" is actually an IP address, and force a failure. This locks out
2357 those revolvers/nameservers that support "A-for-A" lookups, in
2358 contravention of the specifications.
2360 PH/48 When a host name was looked up from an IP address, and the subsequent
2361 forward lookup of the name timed out, the host name was left in
2362 $sender_host_name, contrary to the specification.
2364 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2365 restricted to single-key lookups, Exim was not diagnosing an error if
2366 * or *@ was used with a query-style lookup.
2368 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2370 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2371 long ago, but noone interested enough thought of it. Let's just say that
2372 the "1.1" means that there are some new functions that weren't there at
2373 some point in the past.
2375 PH/51 Error processing for expansion failure of helo_data from an smtp
2376 transport during callout processing was broken.
2378 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2379 tested/used via the -bh/-bhc/-bs options.
2381 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2382 bug, fixed in subsequent PCRE releases).
2384 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2385 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2387 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2393 SC/01 Use a glob alias rather than an array ref in eximstats generated
2394 parser. This improves both readability and performance.
2396 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2397 Don't display local sender or destination tables in eximstats unless
2398 there is data to show.
2399 Added average volumes into the eximstats top table text output.
2401 SC/03 Collect data on the number of addresses (recipients) as well
2402 as the number of messages in eximstats.
2404 TF/01 Correct an error in the documentation for the redirect router. Exim
2405 does (usually) call initgroups() when daemonizing.
2407 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2408 with consistent privilege compared to when running as a daemon.
2410 TF/03 Note in the spec that $authenticated_id is not set for local
2411 submissions from trusted users.
2413 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2414 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2416 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2417 by adding some example configuration directives to the default
2418 configuration file. A little bit of work is required to uncomment the
2419 directives and define how usernames and passwords are checked, but
2420 there is now a framework to start from.
2422 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2423 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2424 without this. I don't know how relevant this is to other LDAP libraries.
2426 PH/02 Add the verb name to the "unknown ACL verb" error.
2428 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2430 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2432 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2433 directory not expanded when it should be if an expanded home directory
2434 was set for the address (which is overridden by the transport).
2436 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2439 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2440 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2441 because it is too late at that time, and has no effect.
2443 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2444 security issue with \' (bugzilla #107). I could not use the
2445 PQescapeStringConn() function, because it needs a PGconn value as one of
2448 PH/08 When testing addresses using -bt, indicate those final addresses that
2449 are duplicates that would not cause an additional delivery. At least one
2450 person was confused, thinking that -bt output corresponded to deliveries.
2451 (Suppressing duplicates isn't a good idea as you lose the information
2452 about possibly different redirections that led to the duplicates.)
2454 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2455 systems where poll() doesn't work, in particular OS X.
2457 PH/10 Added more information to debugging output for retry time not reached.
2459 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2460 operations in malware.c.
2462 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2465 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2466 syslog_duplication set false, log lines that would normally be written
2467 both the the main log and to the reject log were not written to syslog at
2470 PH/14 In the default configuration, change the use of "message" in ACL warn
2471 statements to "add_header".
2473 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2474 not followed by a command (e.g. "seen endif").
2476 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2477 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2480 PH/17 Added extra conditions to the default value of delay_warning_condition
2484 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2485 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2486 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2489 The Auto-Submitted: and various List- headers are standardised, whereas I
2490 don't think Precedence: ever was.
2492 PH/18 Refactored debugging code in route_finduser() to show more information,
2493 in particular, the error code if getpwnam() issues one.
2495 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2496 This is apparently needed in addition to the PH/07 change above to avoid
2497 any possible encoding problems.
2499 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2500 but not after initializing Perl.
2502 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2503 output them only if debugging. By default they are written stderr,
2504 apparently, which is not desirable.
2506 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2509 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2512 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2514 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2515 authenticated or an ident call has been made. Suppress the default
2516 values for $authenticated_id and $authenticated_sender (but permit -oMai
2517 and -oMas) when testing with -bh.
2519 PH/24 Re-jigged the order of the tests in the default configuration so that the
2520 tests for valid domains and recipients precede the DNS black list and CSA
2521 tests, on the grounds that those ones are more expensive.
2523 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2524 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2525 This bug exists in every version of Exim that I still have, right back to
2528 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2529 However, an attempt to turn on case-sensitivity in a regex key by
2530 including (?-i) didn't work because the subject string was already
2531 lowercased, and the effects were non-intuitive. It turns out that a
2532 one-line patch can be used to allow (?-i) to work as expected.
2538 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2539 other effects) broke the use of negated acl sub-conditions.
2541 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2544 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2545 "Deny" causes Exim to reject the incoming connection with a 554 error.
2546 Unfortunately, if there is a major crisis, such as a disk failure,
2547 tcp-wrappers gives "deny", whereas what one would like would be some
2548 kind of temporary error. A kludge has been added to help with this.
2549 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2550 554 error is used if errno is still zero or contains ENOENT (which occurs
2551 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2554 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2556 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2557 errors. Otherwise a message that provokes a temporary error (when other
2558 messages do not) can cause a whole host to time out.
2560 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2561 addresses were routed directly to files or pipes from a redirect router.
2562 File deliveries just didn't batch; pipe deliveries might have suffered
2565 PH/06 A failure to get a lock for a hints database would erroneously always say
2566 "Failed to get write lock", even when it was really a read lock.
2568 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2569 of 0600. This has been changed to use the value of the lockfile_mode
2570 option (which defaults to 0600).
2572 PH/08 Applied small patch from the Sieve maintainer.
2574 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2575 folder from quota calculations, a direct delivery into this folder messed
2576 up the contents of the maildirsize file. This was because the regex was
2577 used only to exclude .Trash (or whatever) when the size of the mailbox
2578 was calculated. There was no check that a delivery was happening into an
2579 excluded directory. This bug has been fixed by ignoring all quota
2580 processing for deliveries into excluded directories.
2582 PH/10 Added the maildirfolder_create_regex option to appendfile.
2588 PH/01 The code for finding all the local interface addresses on a FreeBSD
2589 system running IPv6 was broken. This may well have applied to all BSD
2590 systems, as well as to others that have similar system calls. The broken
2591 code found IPv4 interfaces correctly, but gave incorrect values for the
2592 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2593 that it would not match correctly against @[] and not recognize the IPv6
2596 PH/02 The ipliteral router was not recognizing addresses of the form user@
2597 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2599 PH/03 Added disable_ipv6.
2601 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2602 decoded form, because it is most often used to construct To: headers
2603 lines in autoreplies, and the decoded form may well be syntactically
2604 invalid. However, $reply_address has leading white space removed, and all
2605 newlines turned into spaces so that the autoreply transport does not
2608 PH/05 If group was specified without a user on a router, and no group or user
2609 was specified on a transport, the group from the router was ignored.
2611 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2612 for visible compile-time settings that can be used to change these
2613 numbers, for those that want even more. Backwards compatibility with old
2614 spool files has been maintained. However, going back to a previous Exim
2615 release will lost any variables that are in spool files.
2617 PH/07 Two small changes when running in the test harness: increase delay when
2618 passing a TCP/IP connection to a new process, in case the original
2619 process has to generate a bounce, and remove special handling of
2620 127.0.0.2 (sic), which is no longer necessary.
2622 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2623 be the same on different OS.
2625 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2628 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2629 whether --show-vars was specified or not
2631 JJ/02 exipick: Added support for new ACL variable spool format introduced
2634 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2635 syntactically invalid From: or Reply-to: line, and a filter used this to
2636 generate an autoreply, and therefore failed to obtain an address for the
2637 autoreply, Exim could try to deliver to a non-existent relative file
2638 name, causing unrelated and misleading errors. What now happens is that
2639 it logs this as a hard delivery error, but does not attempt to create a
2642 PH/11 The exinext utility has a -C option for testing purposes, but although
2643 the given file was scanned by exinext itself; it wasn't being passed on
2644 when Exim was called.
2646 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2647 an end-of-file indication when reading a command response.
2649 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2650 compiled. In many other places in Exim, IPv6 addresses are always
2651 recognized, so I have changed this. It also means that IPv4 domain
2652 literals of the form [IPV4:n.n.n.n] are now always recognized.
2654 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2655 used if the router is not running as root, for example, when verifying at
2656 ACL time, or when using -bh. The debugging output from this situation was
2657 non-existent - all you got was a failure to exec. I have made two
2660 (a) Failures to set uid/gid, the current directory, or a process leader
2661 in a subprocess such as that created by queryprogram now generate
2662 suitable debugging ouput when -d is set.
2664 (b) The queryprogram router detects when it is not running as root,
2665 outputs suitable debugging information if -d is set, and then runs
2666 the subprocess without attempting to change uid/gid.
2668 PH/15 Minor change to Makefile for building test_host (undocumented testing
2671 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2672 additional section of a DNS packet that returns MX or SRV records.
2673 Instead, it always explicitly searches for A/AAAA records. This avoids
2674 major problems that occur when a DNS server includes only records of one
2675 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2676 fixed another bug: if SRV records were looked up and the corresponding
2677 address records were *not* found in the additional section, the port
2678 values from the SRV records were lost.
2680 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2681 using the correct key (the original address) when searching the retry
2682 rules in order to find which one to use for generating the retry hint.
2684 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2685 adding the default one. Similarly, if it contains a Reply-To: header, the
2686 errors_reply_to option, if set, is not used.
2688 PH/19 When calculating a retry time, Exim used to measure the "time since
2689 failure" by looking at the "first failed" field in the retry record. Now
2690 it does not use this if it is later than than the arrival time of the
2691 message. Instead it uses the arrival time. This makes for better
2692 behaviour in cases where some deliveries succeed, thus re-setting the
2693 "first failed" field. An example is a quota failure for a huge message
2694 when small messages continue to be delivered. Without this change, the
2695 "time since failure" will always be short, possible causing more frequent
2696 delivery attempts for the huge message than are intended.
2697 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2699 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2700 $1, $2, $3) because the numerical variables can be reset during some
2701 expansion items (e.g. "match"), thereby losing the authentication data.
2703 PH/21 Make -bV show the size of off_t variables so that the test suite can
2704 decide whether to run tests for quotas > 2G.
2706 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2707 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2708 filecount value is greater than 2G or if a quota value is greater than 2G
2709 on a system where the size of off_t is not greater than 4, a panic error
2712 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2713 never match. The debug and -bh output now contains an explicit error
2714 message indicating a malformed IPv4 address or mask.
2716 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2717 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2718 PH/23 above applies.
2720 PH/25 Do not write to syslog when running in the test harness. The only
2721 occasion when this arises is a failure to open the main or panic logs
2722 (for which there is an explicit test).
2724 PH/26 Added the /no_tell option to "control=freeze".
2726 PH/27 If a host name lookup failed very early in a connection, for example, if
2727 the IP address matched host_lookup and the reverse lookup yielded a name
2728 that did not have a forward lookup, an error message of the form "no IP
2729 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2730 could be logged. Now it outputs the IP address instead of "NULL".
2732 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2733 allows the sender and the authenticated sender to be set when
2734 submitting a message from within Exim. Since child_open_exim() is
2735 documented for local_scan(), the new function should be too.
2737 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2738 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2739 results in an empty string is now treated as unset.
2741 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2743 PH/31 Added sender_verify_fail logging option.
2745 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2746 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2747 tidied the source and removed it altogether.
2749 PH/33 When a queue run was abandoned because the load average was too high, a
2750 log line was always written; now it is written only if the queue_run log
2751 selector is set. In addition, the log line for abandonment now contains
2752 information about the queue run such as the pid. This is always present
2753 in "start" and "stop" lines but was omitted from the "abandon" line.
2755 PH/34 Omit spaces between a header name and the colon in the error message that
2756 is given when verify = headers_syntax fails (if there are lots of them,
2757 the message gets confusing).
2759 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2760 names, as there are now some PTR records that contain slashes. This check
2761 is only to protect against broken name servers that fall over on strange
2762 characters, so the fact that it applies to all lookups doesn't matter.
2764 PH/36 Now that the new test suite is complete, we can remove some of the
2765 special code in Exim that was needed for the old test suite. For example,
2766 sorting DNS records because real resolvers return them in an arbitrary
2767 order. The new test suite's fake resolver always returns records in the
2770 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2771 bounces) except when queue_only is set, to avoid logging races between
2772 the different processes.
2774 PH/38 Panic-die if .include specifies a non-absolute path.
2776 PH/39 A tweak to the "H" retry rule from its user.
2778 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2779 a label. They prevented compilation on older perls.
2781 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2782 a warning to be raised on newish perls.
2784 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2785 on queue. Changes to match documented behaviour of showing count of
2786 messages matching specified criteria.
2788 PH/40 Changed the default ident timeout from 30s to 5s.
2790 PH/41 Added support for the use of login_cap features, on those BSD systems
2791 that have them, for controlling the resources used by pipe deliveries.
2793 PH/42 The content-scanning code uses fopen() to create files in which to put
2794 message data. Previously it was not paying any attention to the mode of
2795 the files. Exim runs with umask(0) because the rest of the code creates
2796 files with open(), and sets the required mode explicitly. Thus, these
2797 files were ending up world-writeable. This was not a big issue, because,
2798 being within the spool directory, they were not world-accessible. I have
2799 created a function called modefopen, which takes an additional mode
2800 argument. It sets umask(777), creates the file, chmods it to the required
2801 mode, then resets the umask. All the relevant calls to fopen() in the
2802 content scanning code have been changed to use this function.
2804 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2805 to 24 hours. This avoids potential overflow problems when processing G
2806 and H retry rules. I suspect nobody ever tinkers with this value.
2808 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2810 PH/45 When the plaintext authenticator is running as a client, the server's
2811 challenges are checked to ensure they are valid base64 strings. By
2812 default, the authentication attempt is cancelled if an invalid string is
2813 received. Setting client_ignore_invalid_base64 true ignores these errors.
2814 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2815 they are received. Thus, the responses can be made to depend on the
2816 challenges. If an invalid string is ignored, an empty string is placed in
2819 PH/46 Messages that are created by the autoreply transport now contains a
2820 References: header, in accordance with RFCs 2822 and 3834.
2822 PH/47 Added authenticated_sender_force to the smtp transport.
2824 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2826 PH/49 Installed latest patch from the Sieve maintainer.
2828 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2829 was also set, the appendfile transport was unnecessarily scanning a
2830 directory of message files (e.g. for maildir delivery) to find the count
2831 of files (along with the size), even though it did not need this
2832 information. It now does the scan only if it needs to find either the
2833 size of the count of files.
2835 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2837 PH/52 Two bugs concerned with error handling when the smtp transport is
2840 (i) Exim was not creating retry information for temporary errors given
2841 for individual recipients after the DATA command when the smtp transport
2842 was used in LMTP mode. This meant that they could be retried too
2843 frequently, and not timed out correctly.
2845 (ii) Exim was setting the flag that allows error details to be returned
2846 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2847 recipients that were returned after the DATA command.
2849 PH/53 This is related to PH/52, but is more general: for any failing address,
2850 when detailed error information was permitted to be returned to the
2851 sender, but the error was temporary, then after the final timeout, only
2852 "retry timeout exceeded" was returned. Now it returns the full error as
2853 well as "retry timeout exceeded".
2855 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2856 do this, and (what is worse) MTAs that accept it.
2858 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2859 will now be deprecated.
2861 PH/56 New os.c-cygwin from the Cygwin maintainer.
2863 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2864 formats (previously only available in exim formats via -bpr, -bpru,
2865 and -bpra. Now also available in native and exiqgrep formats)
2867 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2868 with very large, slow to parse queues
2870 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2872 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2874 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2875 responses to authentication challenges, though it was showing the
2876 challenges; (ii) I've removed the CR characters from the debug output for
2879 PH/58 Allow for the insertion of a newline as well as a space when a string
2880 is turned into more than one encoded-word during RFC 2047 encoding. The
2881 Sieve code now uses this.
2883 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2884 data_4xx, lost_connection, tls_required.
2886 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2887 message was being sent as an SMTP response.
2889 PH/61 Add -l and -k options to exicyclog.
2891 PH/62 When verifying, if an address was redirected to one new address, so that
2892 verification continued, and the new address failed or deferred after
2893 having set something in $address_data, the value of $address_data was not
2894 passed back to the ACL. This was different to the case when no
2895 redirection occurred. The value is now passed back in both cases.
2897 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2898 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2899 use login_cap.h, so on its own it isn't the distinguishing feature. The
2900 new name refers directly to the setclassresources() function.
2902 PH/65 Added configuration files for NetBSD3.
2904 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2906 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2907 is preferred over IPv4.
2909 PH/68 The bounce_return_message and bounce_return_body options were not being
2910 honoured for bounces generated during the reception of non-SMTP messages.
2911 In particular, this applied to messages rejected by the ACL. This bug has
2912 been fixed. However, if bounce_return_message is true and bounce_return_
2913 body is false, the headers that are returned for a non-SMTP message
2914 include only those that have been read before the error was detected.
2915 (In the case of an ACL rejection, they have all been read.)
2917 PH/69 The HTML version of the specification is now built in a directory called
2918 spec_html instead of spec.html, because the latter looks like a path with
2919 a MIME-type, and this confuses some software.
2921 PH/70 Catch two compiler warnings in sieve.c.
2923 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2924 function verify_get_ident() calls ip_connect() to connect a socket, but
2925 if the "connect()" function timed out, ip_connect() used to close the
2926 socket. However, verify_get_ident() also closes the socket later, and in
2927 between Exim writes to the log, which may get opened at this point. When
2928 the socket was closed in ip_connect(), the log could get the same file
2929 descriptor number as the socket. This naturally causes chaos. The fix is
2930 not to close the socket in ip_connect(); the socket should be closed by
2931 the function that creates it. There was only one place in the code where
2932 this was missing, in the iplookup router, which I don't think anybody now
2933 uses, but I've fixed it anyway.
2935 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2936 well as to direct DNS lookups. Otherwise the handling of names in host
2937 lists is inconsistent and therefore confusing.
2943 PH/01 Two changes to the default runtime configuration:
2945 (1) Move the checks for relay_from_hosts and authenticated clients from
2946 after to before the (commented out) DNS black list checks.
2948 (2) Add control=submission to the relay_from_hosts and authenticated
2949 clients checks, on the grounds that messages accepted by these
2950 statements are most likely to be submissions.
2952 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2954 (1) Generate an error if the third argument for the ${prvs expansion is
2957 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2960 (3) Reset the variables that are obtained from the first argument of
2961 ${prvscheck and used in the second argument before leaving the code,
2962 because their memory is reclaimed, so using them afterwards may do
2965 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2966 one (it's much easier than Tom thought :-).
2968 (5) Because of (4), we can now allow for the use of $prvscheck_result
2969 inside the third argument.
2971 PH/03 For some reason, the default setting of PATH when running a command from
2972 a pipe transport was just "/usr/bin". I have changed it to
2975 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2976 anything to be listed in the output from -bV.
2978 PH/05 When a filter generated an autoreply, the entire To: header line was
2979 quoted in the delivery log line, like this:
2981 => >A.N.Other <ano@some.domain> <original@ddress> ...
2983 This has been changed so that it extracts the operative address. There
2984 may be more than one such address. If so, they are comma-separated, like
2987 => >ano@some.domain,ona@other.domain <original@ddress> ...
2989 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2990 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2991 not being looked up in the rDNS to get a host name, Exim was showing the
2992 IP address twice in Received: lines, even though the IP addresses were
2993 identical. For example:
2995 Received: from [1.2.3.4] (helo=[1.2.3.4])
2997 However, if the real host name was known, it was omitting the HELO data
2998 if it matched the actual IP address. This has been tidied up so that it
2999 doesn't show the same IP address twice.
3001 PH/07 When both +timestamp and +memory debugging was on, the value given by
3002 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3003 called by the debug printing, thereby overwriting the timestamp buffer.
3004 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3006 PH/08 When the original message was included in an autoreply transport, it
3007 always said "this is a copy of the message, including all the headers",
3008 even if body_only or headers_only was set. It now gives an appropriate
3011 PH/09 Applied a patch from the Sieve maintainer which:
3013 o fixes some comments
3014 o adds the (disabled) notify extension core
3015 o adds some debug output for the result of if/elsif tests
3016 o points to the current vacation draft in the documentation
3017 and documents the missing references header update
3021 o fixes a bug in processing the envelope test (when testing
3022 multiple envelope elements, the last element determinted the
3025 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3026 Electronic Mail") by including:
3028 Auto-submitted: auto-generated
3030 in the messages that it generates (bounce messages and others, such as
3031 warnings). In the case of bounce messages for non-SMTP mesages, there was
3032 also a typo: it was using "Auto_submitted" (underscore instead of
3033 hyphen). Since every message generated by Exim is necessarily in response
3034 to another message, thes have all been changed to:
3036 Auto-Submitted: auto-replied
3038 in accordance with these statements in the RFC:
3040 The auto-replied keyword:
3042 - SHOULD be used on messages sent in direct response to another
3043 message by an automatic process,
3045 - MUST NOT be used on manually-generated messages,
3047 - MAY be used on Delivery Status Notifications (DSNs) and Message
3048 Disposition Notifications (MDNs),
3050 - MUST NOT be used on messages generated by automatic or periodic
3051 processes, except for messages which are automatic responses to
3054 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3055 to the default Received: header definition.
3057 PH/12 Added log selector acl_warn_skipped (default on).
3059 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3060 variables because (a) they are in the wrong storage pool and (b) even if
3061 they were copied, it wouldn't work properly because of the caching.
3063 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3064 checking when decoding. Apparently there are clients that generate
3065 overlong encoded strings. Why am I not surprised?
3067 PH/15 If the first argument of "${if match_address" was not empty, but did not
3068 contain an "@" character, Exim crashed. Now it writes a panic log message
3069 and treats the condition as false.
3071 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3073 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3074 extension "envelope-auth". The code is finished and in agreement with
3075 other implementations, but there is no documentation so far and in fact,
3076 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3077 not changing the active code.
3079 Print executed "if" and "elsif" statements when debugging is used. This
3080 helps a great deal to understand what a filter does.
3082 Document more things not specified clearly in RFC3028. I had all this
3083 sorted out, when out of a sudden new issues came to my mind. Oops."
3085 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3088 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3089 canonical form (as documented). However, after a host name lookup from
3090 the IP address, check_host() was doing a simple string comparison with
3091 addresses acquired from the DNS when checking that the found name did
3092 have the original IP as one of its addresses. Since any found IPv6
3093 addresses are likely to be in abbreviated form, the comparison could
3094 fail. Luckily, there already exists a function for doing the comparison
3095 by converting both addresses to binary, so now that is used instead of
3096 the text comparison.
3098 PH/20 There was another similar case to PH/19, when a complete host name was
3099 given in a host list; looking up its IP address could give an abbreviated
3100 form, whereas the current host's name might or might not be abbreviated.
3101 The same fix has been applied.
3107 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3108 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3111 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3112 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3114 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3116 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3117 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3118 Also, for "Auto-Submitted", treat anything other than "no" as
3119 non-personal, in accordance with RFC 3834. (Previously it treated
3120 anything starting "auto-" as non-personal.)
3122 TF/01 The control=submission/name=... option had a problem with syntax
3123 errors if the name included a slash character. The /name= option
3124 now slurps the rest of the string, so it can include any characters
3125 but it must come last in the list of options (after /sender_retain
3128 PH/05 Some modifications to the interface to the fake nameserver for the new
3136 TK/01 Added the "success_on_redirect" address verification option. See
3137 NewStuff for rationale and an example.
3139 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3141 PH/02 Patch to exigrep to allow it to work on syslog lines.
3143 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3144 fread() to skip over the body file's header line, because in Cygwin the
3145 header line is locked and is inaccessible.
3147 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3148 co-exist for some time) to make it clear that it is the Exim ID that is
3149 referenced, not the Message-ID: header line.
3151 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3152 string_format() function, because snprintf() does not exist on all
3155 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3156 ${stat: expansion item.
3158 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3159 protocol synchronization error", to keep the pedants happy.
3161 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3162 well as for IRIX systems, when gcc is being used. See the host.c source
3165 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3167 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3170 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3171 transport and to the smtp transport in LMTP mode.
3173 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3175 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3176 run a filter in a subprocess. This could lead to confusion in subsequent
3177 lookups in the parent process. There should also be a search_tidyup() at
3178 the end of the subprocess.
3180 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3181 only if the host matched helo_try_verify_hosts, which caused the
3182 verification to occur when the EHLO/HELO command was issued. The ACL just
3183 tested the remembered result. Now, if a previous verification attempt has
3184 not happened, "verify = helo" does it there and then.
3186 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3188 TK/03 Fix log output including CR from clamd.
3190 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3191 exist provoked a memory error which could cause a segfault.
3193 PH/15 Installed PCRE 6.2
3195 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3197 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3198 of the problem. Specifically, suggested +O2 rather than +O1 for the
3201 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3203 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3204 record, the port was not correctly logged when the outgoing_port log
3205 selector was set (it logged the transort's default port).
3207 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3208 fallback_hosts, and "hosts" in the smtp transport.
3210 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3211 host errors such as "Connection refused".
3213 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3214 authentication with radiusclient 0.4.9:
3216 - Error returned from rc_read_config was caught wrongly
3217 - Username/password not passed on to radius server due to wrong length.
3219 The presumption is that some radiusclient API changes for 4.51/PH/17
3220 were not taken care of correctly. The code is still untested by me (my
3221 Linux distribution still has 0.3.2 of radiusclient), but it was
3222 contributed by a Radius user.
3224 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3225 expanding the "port" option of the smtp transport.
3227 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3228 while reading a MIME header. Thanks to Tom Hughes for a patch.
3230 PH/24 Include config.h inside local_scan.h so that configuration settings are
3233 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3234 that in an ACL for RCPT (for example), you can examine exactly what was
3237 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3238 commands, but it was not correctly comparing the address with the actual
3239 client host address. Thus, it would show the EHLO address in Received:
3240 header lines when this was not necessary.
3242 PH/27 Added the % operator to ${eval:}.
3244 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3245 it should be ignoring failures (because with -C, for example, it has lost
3246 privilege). It wasn't ignoring creation failures other than "already
3249 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3252 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3253 because an input line was too long, either on its own, or by virtue of
3254 too many continuations, the temporary file was not being removed, and the
3255 return code was incorrect.
3257 PH/31 Missing "BOOL" in function definition in filtertest.c.
3259 PH/32 Applied Sieve patches from the maintainer.
3261 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3263 PH/33 Added "verify = not_blind".
3265 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3266 Local/Makefile (with some defaults set). These are used in built scripts
3267 such as exicyclog, but they have never been used in the exim_install
3268 script (though there are many overriding facilities there). I have
3269 arranged that the exim_install script now takes note of these two
3272 PH/35 Installed configuration files for Dragonfly.
3274 PH/36 When a locally submitted message by a trusted user did not contain a
3275 From: header, and the sender address was obtained from -f or from an SMTP
3276 MAIL command, and the trusted user did not use -F to supply a sender
3277 name, $originator_name was incorrectly used when constructing a From:
3278 header. Furthermore, $originator_name was used for submission mode
3279 messages from external hosts without From: headers in a similar way,
3280 which is clearly wrong.
3282 PH/37 Added control=suppress_local_fixups.
3284 PH/38 When log_selector = +received_sender was set, and the addition of the
3285 sender made the log line's construction buffer exactly full, or one byte
3286 less than full, an overflow happened when the terminating "\n" was
3289 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3290 when the result of a list match is failure because a DNS lookup failed.
3292 PH/40 RM_COMMAND is now used in the building process.
3294 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3295 the "build-* directories that it finds.
3297 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3298 address in a domain literal was a prefix of an interface address.
3300 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3301 when verifying a sender address, unless rewrite_headers is false.
3303 PH/44 Wrote a long comment about why errors_to addresses are verified as
3304 recipients, not senders.
3306 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3307 the ratelimit ACL was added.
3309 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3311 PH/46 Added extra information about PostgreSQL errors to the error string.
3313 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3314 avoiding the need to install special zones in a real server. This is
3315 backwards compatible; if it can't find the fake resolver, it drops back.
3316 Thus, both old and new test suites can be run.
3318 TF/02 Added util/ratelimit.pl
3320 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3321 clock is set back in time.
3323 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3324 Candler <B.Candler@pobox.com>.
3326 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3327 OK for addresses that are the result of redirections.
3329 PH/48 A number of further additions for the benefit of the new test suite,
3330 including a fake gethostbyname() that interfaces to the fake DNS resolver
3333 TF/06 The fix for widen_domains has also been applied to qualify_single and
3334 search_parents which are the other dnslookup options that can cause
3337 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3340 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3342 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3343 service" if it encountered an MX record with an empty target hostname.
3344 The message is now "an MX or SRV record indicated no SMTP service".
3346 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3347 if the DNS of the sending site is misconfigured. This is quite a
3348 common situation. This change restores the behaviour of treating a
3349 helo verification defer as a failure.
3351 PH/51 If self=fail was set on a router, the bounce message did not include the
3352 actual error message.
3358 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3360 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3361 command itself ended in error, the underlying message about the transport
3362 filter timeout was being overwritten with the pipe command error. Now the
3363 underlying error message should be appended to the second error message.
3365 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3367 PH/02 Reduce the amount of output that "make" produces by default. Full output
3368 can still be requested.
3370 PH/03 The warning log line about a condition test deferring for a "warn" verb
3371 was being output only once per connection, rather than after each
3372 occurrence (because it was using the same function as for successful
3373 "warn" verbs). This seems wrong, so I have changed it.
3375 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3376 it should not have, which might have caused a crash in the right
3377 circumstances, but probably never did.
3379 PH/04 Installed a modified version of Tony Finch's patch to make submission
3380 mode fix the return path as well as the Sender: header line, and to
3381 add a /name= option so that you can make the user's friendly name appear
3384 TF/03 Added the control = fakedefer ACL modifier.
3386 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3387 Mark Lowes for thorough testing.
3389 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3391 TK/03 Merged latest SRS patch from Miles Wilton.
3393 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3394 with the definition in sysexits.h (which is #included earlier).
3395 Fortunately, Exim does not actually use EX_OK. The code used to try to
3396 preserve the sysexits.h value, by assumimg that macro definitions were
3397 scanned for macro replacements. I have been disabused of this notion,
3398 so now the code just undefines EX_OK before #including unistd.h.
3400 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3401 in the smtp transport. When a block could not be written in a single
3402 write() function, the timeout was being re-applied to each part-write.
3403 This seems wrong - if the receiver was accepting one byte at a time it
3404 would take for ever. The timeout is now adjusted when this happens. It
3405 doesn't have to be particularly precise.
3407 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3408 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3410 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3411 without a domain if the check to <postmaster@domain> fails.
3413 SC/01 Eximstats: added -xls and the ability to specify output files
3414 (patch written by Frank Heydlauf).
3416 SC/02 Eximstats: use FileHandles for outputing results.
3418 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3420 SC/04 Eximstats: fixed display of large numbers with -nvr option
3422 SC/05 Eximstats: fixed merging of reports with empty tables.
3424 SC/06 Eximstats: added the -include_original_destination flag
3426 SC/07 Eximstats: removed tabs and trailing whitespace.
3428 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3430 TK/06 MBOX spool code: Add real "From " MBOX separator line
3431 so the .eml file is really in mbox format (even though
3432 most programs do not really care). Patch from Alex Miller.
3434 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3435 The latter is generated from $received_to and is only set if the
3436 message has one envelope recipient. SA can use these headers,
3437 obviously out-of-the-box. Patch from Alex Miller.
3439 PH/08 The ${def test on a variable was returning false if the variable's
3440 value was "0", contrary to what the specification has always said!
3441 The result should be true unless the variable is empty.
3443 PH/09 The syntax error of a character other than { following "${if
3444 def:variable_name" (after optional whitespace) was not being diagnosed.
3445 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3446 accidental colon was present, for example, could give incorrect results.
3448 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3449 result is used (not including appendfile, where other changes are about
3452 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3453 This involved changing a lot of size variables from int to off_t. It
3454 should work with maildirs and everything.
3456 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3457 spamd dying while we are connected to it.
3459 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3462 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3463 function for BDB 4.3.
3465 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3467 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3468 the sources for PCRE have been re-arranged and more files are now
3471 PH/15 The code I had for printing potentially long long variables in PH/11
3472 above was not the best (it lost precision). The length of off_t variables
3473 is now inspected at build time, and an appropriate printing format (%ld
3474 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3475 to be "long long int" or "long int". This is needed for the internal
3476 formatting function string_vformat().
3478 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3479 the configuration file to be ":syslog", then the script "guesses" where
3480 the logs files are, rather than using the compiled in default. In our
3481 case the guess is not the same as the compiled default, so the script
3482 suddenly stopped working when I started to use syslog. The patch checks
3483 to see if log_file_path is "". If so, it attempts to read it from exim
3484 with no configuration file to get the compiled in version, before it
3485 falls back to the previous guessing code."
3487 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3488 implementing BATV in an Exim configuration. See NewStuff for the gory
3491 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3492 Makefile that are specific to HP-UX.
3494 PH/18 If the "use_postmaster" option was set for a recipient callout together
3495 with the "random" option, the postmaster address was used as the MAIL
3496 FROM address for the random test, but not for the subsequent recipient
3497 test. It is now used for both.
3499 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3500 patch removes a few documentation additions to RFC 3028, because the
3501 latest draft now contains them. It adds the new en;ascii-case comparator
3502 and a new error check for 8bit text in MIME parts. Comparator and
3503 require names are now matched exactly. I enabled the subaddress
3504 extension, but it is not well tested yet (read: it works for me)."
3506 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3507 rework some of the code of TK/09 above to avoid the hardwired use of
3508 "%lld" and "long long". Replaced the call to snprintf() with a call to
3511 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3512 records point to non-existent hosts", "retry timeout exceeded", and
3513 "retry time not reached for any host after a long failure period".
3515 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3516 experimental DomainKeys support:
3518 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3519 (2) On an error such as an illegally used "control", the wrong name for
3520 the control was given.
3522 These problems did NOT occur unless DomainKeys support was compiled.
3524 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3526 PH/24 Added ${if match_ip condition.
3528 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3529 databases so that it will be absolutely obvious if a crash occurs in the
3530 DB library. This is a regular occurrence (often caused by mis-matched
3533 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3534 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3535 picked up on a user's system that detects such things. There doesn't seem
3536 to be a gcc warning option for this - only an attribute that has to be
3537 put on the function's prototype. It seems that in Fedora Core 4 they have
3538 set this on a number of new functions. No doubt there will be more in due
3541 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3542 specify a transport. However, if an address that was verified by such a
3543 router was the subject of a callout, Exim crashed because it tried to
3544 read the rcpt_include_affixes from the non-existent transport. Now it
3545 just assumes that the setting of that option is false. This bug was
3546 introduced by 4.51/PH/31.
3548 PH/28 Changed -d+all to exclude +memory, because that information is very
3549 rarely of interest, but it makes the output a lot bigger. People tend to
3550 do -d+all out of habit.
3552 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3553 code in os-type was giving problems when libc.so lives in lib64, like on
3556 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3557 aren't the modern standard, and it seems that some systems' include files
3558 don't always have them. Exim was already checking for some of the newer
3559 ones like T_AAAA, and defining it itself. I've added checks for all the
3560 record types that Exim uses.
3562 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3563 not automatically generating a new one, as it is supposed to. This
3564 prevented TLS from working. If the file did exist, but contained invalid
3565 data, a new version was generated, as expected. It was only the case of a
3566 non-existent file that was broken.
3568 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3569 with a change in libdomainkeys > 0.64.
3571 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3572 from DNS. If the selector record carries the flag, it now has
3573 precedence over the domain-wide flag.
3575 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3577 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3578 the use of a port name that isn't defined in /etc/services) occurred, the
3579 message was deferred as in a normal delivery, and thus remained on the
3580 spool, instead of being failed because of the mua_wrapper setting. This
3581 is now fixed, and I tidied up some of the mua_wrapper messages at the
3584 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3585 the messages in a hash of arrays rather than using individual hashes.
3586 This is a bit cleaner and results in dramatic memory savings, albeit
3587 at a slight CPU cost.
3589 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3590 as requested by Marc Sherman.
3592 SC/10 Eximstats: added histograms for user specified patterns as requested
3595 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3597 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3598 fopen() in the content-scanning modules that did not already have it.
3604 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3605 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3607 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3609 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3611 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3612 to test Sieve filters that use "vacation".
3614 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3615 that changes the way the GnuTLS parameters are stored in the cache file.
3616 The new format can be generated externally. For backward compatibility,
3617 if the data in the cache doesn't make sense, Exim assumes it has read an
3618 old-format file, and it generates new data and writes a new file. This
3619 means that you can't go back to an older release without removing the
3622 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3623 work if there are any delivery delays because "one_time" forces the
3624 parent to be marked "delivered", so its unseen clone is never tried
3625 again. For this reason, Exim now forbids the simultaneous setting of
3628 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3629 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3630 entry if you want to know the details. The fix, however, neglected to
3631 consider the case when local delivery batching is involved. The test for
3632 "previously delivered" was not happening when checking to see if an
3633 address could be batched with a previous (undelivered) one; under
3634 certain circumstances this could lead to multiple deliveries to the same
3637 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3638 in its include files, and this causes problems building Exim.
3640 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3641 header_syntax) but Exim was just ignoring anything given after a slash.
3642 In particular, this caused confusion with an attempt to use "verify =
3643 reverse_host_lookup/defer_ok". An error is now given when options are
3644 supplied for verify items that do not have them. (Maybe reverse_host_
3645 lookup should have a defer_ok option, but that's a different point.)
3647 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3648 defined by RFC 821) to 2048, because there were problems with some AUTH
3649 commands, and RFC 1869 says the size should be increased for extended
3650 SMTP commands that take arguments.
3652 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3655 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3656 "unknown" error; now it says that the functionality isn't in the binary.
3658 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3659 an address' error message when a string expansion fails (syntax or
3660 whatever). Otherwise the password may appear in the log. Following change
3661 PH/42 below, there is no longer a chance of it appearing in a bounce
3664 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3666 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3667 of its addresses was ever tried. (Bugzilla bug #2.)
3669 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3670 the result incorrectly in the debug output. (It correctly added a newline
3671 to what was transported.)
3673 TF/01 Added $received_time.
3675 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3676 commented out examples of how to interface to a virus scanner and to
3677 SpamAssassin. Also added commented examples of av_scanner and
3678 spamd_address settings.
3680 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3681 and controls are allowed in which ACLs. There were a couple of minor
3682 errors. Some of the entries in the conditions table (which is a table of
3683 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3684 negation of where the condition IS allowed.
3686 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3688 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3689 header file does not have a version number, so I've had to invent a new
3690 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3691 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3692 radiusclient), but it was contributed by a Radius user.
3694 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3695 files or directories, for OpenSSL.
3697 PH/19 When an Exim process that is running as root has to create an Exim log
3698 file, it does so in a subprocess that runs as exim:exim so as to get the
3699 ownership right at creation (otherwise, other Exim processes might see
3700 the file with the wrong ownership). There was no test for failure of this
3701 fork() call, which would lead to the process getting stuck as it waited
3702 for a non-existent subprocess. Forks do occasionally fail when resources
3703 run out. I reviewed all the other calls to fork(); they all seem to check
3706 PH/20 When checking for unexpected SMTP input at connect time (before writing
3707 the banner), Exim was not dealing correctly with a non-positive return
3708 from the read() function. If the client had disconnected by this time,
3709 the result was a log entry for a synchronization error with an empty
3710 string after "input=" when read() returned zero. If read() returned -1
3711 (an event I could not check), uninitialized data bytes were printed.
3712 There were reports of junk text (parts of files, etc) appearing after
3715 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3717 PH/22 Added support for macro redefinition, and (re)definition in between
3718 driver and ACL definitions.
3720 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3721 forgetting to use the resulting value; it was using the unexpanded value.
3723 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3724 hadn't been configured. The fix is from Juergen Kreileder, who
3725 understands it better than I do:
3727 "Here's what I see happening with three configured cyrus_sasl
3728 authenticators configured (plain, login, cram-md5):
3730 On startup auth_cyrus_sasl_init() gets called for each of these.
3731 This means three calls to sasl_listmech() without a specified mech_list.
3732 => SASL tests which mechs of all available mechs actually work
3733 => three warnings about OTP not working
3734 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3736 With the patch, sasl_listmech() also gets called three times. But now
3737 SASL's mech_list option is set to the server_mech specified in the the
3738 authenticator. Or in other words, the answer from sasl_listmech()
3739 gets limited to just the mech you're testing for (which is different
3741 => the return list contains just 'plain' or 'login', 'cram-md5' or
3742 nothing depending on the value of ob->server_mech.
3744 I've just tested the patch: Authentication still works fine,
3745 unavailable mechs specified in the exim configuration are still
3746 caught, and the auth.log warnings about OTP are gone."
3748 PH/25 When debugging is enabled, the contents of the command line are added
3749 to the debugging output, even when log_selector=+arguments is not
3752 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3753 answer is "GNU", and only if the return is "GNU/something" is the answer
3756 PH/27 $acl_verify_message is now set immediately after the failure of a
3757 verification in an ACL, and so is available in subsequent modifiers. In
3758 particular, the message can be preserved by coding like this:
3760 warn !verify = sender
3761 set acl_m0 = $acl_verify_message
3763 Previously, $acl_verify_message was set only while expanding "message"
3764 and "log_message" when a very denied access.
3766 PH/28 Modified OS/os.c-Linux with
3768 -#ifndef OS_LOAD_AVERAGE
3769 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3771 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3772 nomenclature these days.)
3774 PH/29 Installed patch from the Sieve maintainer that adds the options
3775 sieve_useraddress and sieve_subaddress to the redirect router.
3777 PH/30 In these circumstances:
3778 . Two addresses routed to the same list of hosts;
3779 . First host does not offer TLS;
3780 . First host accepts first address;
3781 . First host gives temporary error to second address;
3782 . Second host offers TLS and a TLS session is established;
3783 . Second host accepts second address.
3784 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3785 and peerdn, if requested) that were in fact used only for the second
3788 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3789 attention to any local part prefix or suffix that was matched by the
3790 router that accepted the address. It now behaves in the same way as it
3791 does for delivery: the affixes are removed from the local part unless
3792 rcpt_include_affixes is set on the transport.
3794 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3795 timeout during the DATA phase of an incoming message.
3797 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3798 applied a patch sanctioned by the Sieve maintainer.
3800 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3801 the uid or gid is negative. A case of a negative gid caused this to be
3802 noticed. The fix allows for either to be negative.
3804 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3805 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3808 PH/36 The change PH/12 above was broken. Fixed it.
3810 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3811 the grounds that routing the same address twice would always produce the
3812 same answer. This might have been true once, but it is certainly no
3813 longer true now. Routing a child address may depend on the previous
3814 routing that produced that child. Some complicated redirection strategies
3815 went wrong when messages had multiple recipients, and made Exim's
3816 behaviour dependent on the order in which the addresses were given.
3818 I have moved the duplicate checking until after the routing is complete.
3819 Exim scans the addresses that are assigned to local and remote
3820 transports, and removes any duplicates. This means that more work will be
3821 done, as duplicates will always all be routed, but duplicates are
3822 presumably rare, so I don't expect this is of any significance.
3824 For deliveries to pipes, files, and autoreplies, the duplicate checking
3825 still happens during the routing process, since they are not going to be
3828 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3829 It corrects a timeout issue with spamd. This is Ian's comment: "The
3830 background is that sometimes spamd either never reads data from a
3831 connection it has accepted, or it never writes response data. The exiscan
3832 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3833 blindly assumes that writes won't block so it may never time out."
3835 PH/39 Allow G after quota size as well as K and M.
3837 PH/40 The value set for $authenticated_id in an authenticator may not contain
3838 binary zeroes or newlines because the value is written to log lines and
3839 to spool files. There was no check on this. Now the value is run through
3840 the string_printing() function so that such characters are converted to
3841 printable escape sequences.
3843 PH/41 $message_linecount is a new variable that contains the total number of
3844 lines in the message. Compare $body_linecount, which is the count for the
3847 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3848 bounce and delay warning messages, except in certain special cases, which
3851 (a) An SMTP error message from a remote host;
3852 (b) A message specified in a :fail: redirection;
3853 (c) A message specified in a "fail" command in a system filter;
3854 (d) A message specified in a FAIL return from the queryprogram router;
3855 (e) A message specified by the cannot_route_message router option.
3857 In these cases only, Exim does include the error details in bounce and
3858 warning messages. There are also a few cases where bland messages such
3859 as "unrouteable address" or "local delivery error" are given.
3861 PH/43 $value is now also set for the "else" part of a ${run expansion.
3863 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3864 being worked on, but at least Exim now implements the latest version to
3867 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3868 process to complete was treated as a delivery failure, a timeout while
3869 writing the message to the pipe was logged, but erroneously treated as a
3870 successful delivery. Such timeouts include transport filter timeouts. For
3871 consistency with the overall process timeout, these timeouts are now
3872 treated as errors, giving rise to delivery failures by default. However,
3873 there is now a new Boolean option for the pipe transport called
3874 timeout_defer, which, if set TRUE, converts the failures into defers for
3875 both kinds of timeout. A transport filter timeout is now identified in
3878 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3879 systems where "make" and "gmake" are different, calling "gmake" at top
3880 level broke things. I've arranged for the value of $(MAKE) to be passed
3881 from the Makefile to this script so that it can call the same version of
3885 A note about Exim versions 4.44 and 4.50
3886 ----------------------------------------
3888 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3889 changes of various kinds. As a consequence, a big documentation update was
3890 needed. This delayed the release for rather longer than seemed good, especially
3891 in the light of a couple of (minor) security issues. Therefore, the changes
3892 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3893 release. So 4.44 and 4.50 are in effect two different branches that both start
3896 I have left the 4.50 change log unchanged; it contains all the changes since
3897 4.43. The change log for 4.44 is below; many of its items are identical to
3898 those for 4.50. This seems to be the most sensible way to preserve the
3899 historical information.
3905 1. Minor wording change to the doc/README.SIEVE file.
3907 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3908 computation of the current number of files was incorrect.
3910 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3911 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3914 4. Give more explanation in the error message when the command for a transport
3915 filter fails to execute.
3917 5. There are several places where Exim runs a non-Exim command in a
3918 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3919 was being done only for the command run by the queryprogram router. It is
3920 now done for all such subprocesses. The other cases are: ${run, transport
3921 filters, and the commands run by the lmtp and pipe transports.
3923 6. Added CONFIGURE_GROUP build-time option.
3925 7. Some older OS have a limit of 256 on the maximum number of file
3926 descriptors. Exim was using setrlimit() to set 1000 as a large value
3927 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3928 systems. I've change it so that if it can't get 1000, it tries for 256.
3930 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3931 was an oversight, and furthermore, ever since the addition of extra
3932 controls (e.g. 4.43/32), the checks on when to allow different forms of
3933 "control" were broken. There should now be diagnostics for all cases when a
3934 control that does not make sense is encountered.
3936 9. Added the /retain_sender option to "control=submission".
3938 10. $recipients is now available in the predata ACL (oversight).
3940 11. Tidy the search cache before the fork to do a delivery from a message
3941 received from the command line. Otherwise the child will trigger a lookup
3942 failure and thereby defer the delivery if it tries to use (for example) a
3943 cached ldap connection that the parent has called unbind on.
3945 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3946 of $address_data from the recipient verification was clobbered by the
3947 sender verification.
3949 13. The value of address_data from a sender verification is now available in
3950 $sender_address_data in subsequent conditions in the ACL statement.
3952 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3954 15. Added a new option "connect=<time>" to callout options, to set a different
3957 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3958 was its contents. (It was OK if the option was not defined at all.)
3960 17. A "Completed" log line is now written for messages that are removed from
3961 the spool by the -Mrm option.
3963 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3964 information about exactly what failed.
3966 19. Added -dd to debug only the daemon process.
3968 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3969 handles timeouts, both on the server side and network timeouts. Renamed the
3970 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3972 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3973 It is now set to "smtps".
3975 22. $host_address is now set to the target address during the checking of
3976 ignore_target_hosts.
3978 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3979 being passed; this would have caused $sender_host_name to have been used if
3980 matching the list had actually called for a host name (not very likely,
3981 since this list is usually IP addresses). A host name is now passed as
3984 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3985 code that specifies a non-restarting handler (typically sigaction() in
3986 modern systems) in an attempt to fix a rare and obscure crash bug.
3988 25. Narrowed the window for a race in the daemon that could cause it to ignore
3989 SIGCHLD signals. This is not a major problem, because they are used only to
3990 wake it up if nothing else does.
3992 26. A malformed maildirsize file could cause Exim to calculate negative values
3993 for the mailbox size or file count. Odd effects could occur as a result.
3994 The maildirsize information is now recalculated if the size or filecount
3997 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3998 support for a long time. Removed HAVE_SYS_VFS_H.
4000 28. Installed the latest version of exipick from John Jetmore.
4002 29. In an address list, if the pattern was not a regular expression, an empty
4003 subject address (from a bounce message) matched only if the pattern was an
4004 empty string. Non-empty patterns were not even tested. This was the wrong
4005 because it is perfectly reasonable to use an empty address as part of a
4006 database query. An empty address is now tested by patterns that are
4007 lookups. However, all the other forms of pattern expect the subject to
4008 contain a local part and a domain, and therefore, for them, an empty
4009 address still always fails if the pattern is not itself empty.
4011 30. Exim went into a mad DNS loop when attempting to do a callout where the
4012 host was specified on an smtp transport, and looking it up yielded more
4013 than one IP address.
4015 31. Re-factored the code for checking spool and log partition space into a
4016 function that finds that data and another that does the check. The former
4017 is then used to implement four new variables: $spool_space, $log_space,
4018 $spool_inodes, and $log_inodes.
4020 32. The RFC2047 encoding function was originally intended for short strings
4021 such as real names; it was not keeping to the 75-character limit for
4022 encoded words that the RFC imposes. It now respects the limit, and
4023 generates multiple encoded words if necessary. To be on the safe side, I
4024 have increased the buffer size for the ${rfc2047: expansion operator from
4027 33. It is now permitted to omit both strings after an "if" condition; if the
4028 condition is true, the result is "true". As before, when the second string
4029 is omitted, a false condition yields an empty string. This makes it less
4030 cumbersome to write custom ACL and router conditions.
4032 34. Failure to deliver a bounce message always caused it to be frozen, even if
4033 there was an errors_to setting on the router. The errors_to setting is now
4036 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4037 canonical form (fully expanded) before being placed in
4038 $sender_host_address.
4040 36. The table in the code that translates DNS record types into text (T_A to
4041 "A" for instance) was missing entries for NS and CNAME. It is just possible
4042 that this could have caused confusion if both these types were looked up
4043 for the same domain, because the text type is used as part of Exim's
4044 per-process caching. But the chance of anyone hitting this buglet seems
4047 37. The dnsdb lookup has been extended in a number of ways.
4049 (1) There is a new type, "zns", which walks up the domain tree until it
4050 finds some nameserver records. It should be used with care.
4052 (2) There is a new type, "mxh", which is like "mx" except that it returns
4053 just the host names, not the priorities.
4055 (3) It is now possible to give a list of domains (or IP addresses) to be
4056 looked up. The behaviour when one of the lookups defers can be
4057 controlled by a keyword.
4059 (4) It is now possible to specify the separator character for use when
4060 multiple records are returned.
4062 38. The dnslists ACL condition has been extended: it is now possible to supply
4063 a list of IP addresses and/or domains to be looked up in a particular DNS
4066 39. Added log_selector=+queue_time_overall.
4068 40. When running the queue in the test harness, wait just a tad after forking a
4069 delivery process, to get repeatability of debugging output.
4071 41. Include certificate and key file names in error message when GnuTLS fails
4072 to set them up, because the GnuTLS error message doesn't include the name
4073 of the failing file when there is a problem reading it.
4075 42. Allow both -bf and -bF in the same test run.
4077 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4079 44. The "Exiscan patch" is now merged into the mainline Exim source.
4081 45. Sometimes the final signoff response after QUIT could fail to get
4082 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4083 < 0 before doing a fflush(). This bug looks as though it goes back to the
4084 introduction of TLS in release 3.20, but "sometimes" must have been rare
4085 because the tests only now provoked it.
4087 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4088 (this can affect the format of dates).
4090 47. exim_tidydb, when checking for the continued existence of a message for
4091 which it has found a message-specific retry record, was not finding
4092 messages that were in split spool directories. Consequently, it was
4093 deleting retry records that should have stayed in existence.
4095 48. Steve fixed some bugs in eximstats.
4097 49. The SPA authentication driver was not abandoning authentication and moving
4098 on to the next authenticator when an expansion was forced to fail,
4099 contradicting the general specification for all authenticators. Instead it
4100 was generating a temporary error. It now behaves as specified.
4102 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4103 (the order specifies the preference for clients). The order is now AES256,
4104 AES128, 3DES, ARCFOUR128.
4106 51. Small patch to Sieve code - explicitly set From: when generating an
4109 52. Exim crashed if a remote delivery caused a very long error message to be
4110 recorded - for instance if somebody sent an entire SpamAssassin report back
4111 as a large number of 550 error lines. This bug was coincidentally fixed by
4112 increasing the size of one of Exim's internal buffers (big_buffer) that
4113 happened as part of the Exiscan merge. However, to be on the safe side, I
4114 have made the code more robust (and fixed the comments that describe what
4117 53. Now that there can be additional text after "Completed" in log lines (if
4118 the queue_time_overall log selector is set), a one-byte patch to exigrep
4119 was needed to allow it to recognize "Completed" as not the last thing in
4122 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4123 patch that reportedly fixes this has been added. I am not expert enough to
4124 create a test for it. This is what the patch creator wrote:
4126 "I found a little strange behaviour of ldap code when working with
4127 Windows 2003 AD Domain, where users was placed in more than one
4128 Organization Units. When I tried to give exim partial DN, the exit code
4129 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4130 But simultaneously result of request was absolutely normal ldap result,
4131 so I produce this patch..."
4133 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4134 REFERENCE, so I have modified the code to exclude the patch when that macro
4137 55. Some experimental protocols are using DNS PTR records for new purposes. The
4138 keys for these records are domain names, not reversed IP addresses. The
4139 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4140 leaves it alone. Component reversal etc. now happens only for IP addresses.
4143 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4145 57. Double the size of the debug message buffer (to 2048) so that more of very
4146 long debug lines gets shown.
4148 58. The exicyclog utility now does better if the number of log files to keep
4149 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4151 59. Two changes related to the smtp_active_hostname option:
4153 (1) $smtp_active_hostname is now available as a variable.
4154 (2) The default for smtp_banner uses $smtp_active_hostname instead
4155 of $primary_hostname.
4157 60. The host_aton() function is supposed to be passed a string that is known
4158 to be a valid IP address. However, in the case of IPv6 addresses, it was
4159 not checking this. This is a hostage to fortune. Exim now panics and dies
4160 if the condition is not met. A case was found where this could be provoked
4161 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4162 components; fortuitously, this particular loophole had already been fixed
4163 by change 4.50/55 above.
4165 If there are any other similar loopholes, the new check in host_aton()
4166 itself should stop them being exploited. The report I received stated that
4167 data on the command line could provoke the exploit when Exim was running as
4168 exim, but did not say which command line option was involved. All I could
4169 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4170 running as the user.
4173 61. There was a buffer overflow vulnerability in the SPA authentication code
4174 (which came originally from the Samba project). I have added a test to the
4175 spa_base64_to_bits() function which I hope fixes it.
4178 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4179 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4181 63. The daemon start-up calls getloadavg() while still root for those OS that
4182 need the first call to be done as root, but it missed one case: when
4183 deliver_queue_load_max is set with deliver_drop_privilege. This is
4184 necessary for the benefit of the queue runner, because there is no re-exec
4185 when deliver_drop_privilege is set.
4187 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4188 This has been fixed.
4190 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4191 was in use, was not putting the data itself into the right store pool;
4192 consequently, it could be overwritten for a subsequent message in the same
4193 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4196 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4198 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4199 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4200 the function were treating the return as a boolean value, which happened to
4201 work because 0=false and not-0=true, but is not correct code.
4203 68. The host_aton() function was not handling scoped IPv6 addresses (those
4204 with, for example, "%eth0" on the end) correctly.
4206 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4207 negated items (that is, ~something) in unsigned ints. Some compilers
4208 apparently mutter when there is no cast.
4210 70. If an address verification called from an ACL failed, and did not produce a
4211 user-specific message (i.e. there was only a "system" message), nothing was
4212 put in $acl_verify_message. In this situation, it now puts the system
4215 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4216 session; change 4.31/43 added the unwanted input to the log line - except
4217 that it did not do this in the start of session case. It now does.
4219 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4220 This is wrong and can cause the other end to generate a synchronization
4221 error if it is another Exim or anything else that does the synchronization
4222 check. A QUIT command is no longer sent after a timeout.
4224 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4225 during host lookups.
4227 74. The defer_ok option of callout verification was not working if it was used
4228 when verifying addresses in header lines, that is, for this case:
4230 verify = header_sender/callout=defer_ok
4232 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4233 those file descriptors could be used for SMTP connections. If anything
4234 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4235 could be sent to the SMTP client, causing chaos. The daemon now opens
4236 stdin, stdout, and stderr to /dev/null when it puts itself into the
4239 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4240 log by default. The user can override this with suitable Perl magic.
4242 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4243 the log message when discard triggers, was not working for the DATA ACL or
4244 for the non-SMTP ACL.
4246 78. Error message wording change in sieve.c.
4248 79. If smtp_accept_max_per_host was set, the number of connections could be
4249 restricted to fewer than expected, because the daemon was trying to set up
4250 a new connection before checking whether the processes handling previous
4251 connections had finished. The check for completed processes is now done
4252 earlier. On busy systems, this bug wouldn't be noticed because something
4253 else would have woken the daemon, and it would have reaped the completed
4256 80. If a message was submitted locally by a user whose login name contained one
4257 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4258 It caused a spool format error. I have fixed the spool reading code. A
4259 related problem was that the "from" clause in the Received: line became
4260 illegal because of the space(s). It is now covered by ${quote_local_part.
4262 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4265 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4267 83. Patch from Sieve maintainer to fix unterminated string problem in
4268 "vacation" handling.
4270 84. Some minor changes to the Linux configuration files to help with other
4271 OS variants using glibc.
4273 85. One more patch for Sieve to update vacation handling to latest spec.
4276 ----------------------------------------------------
4277 See the note above about the 4.44 and 4.50 releases.
4278 ----------------------------------------------------
4284 1. Change 4.43/35 introduced a bug that caused file counts to be
4285 incorrectly computed when quota_filecount was set in an appendfile
4288 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4289 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4292 3. Give more explanation in the error message when the command for a transport
4293 filter fails to execute.
4295 4. There are several places where Exim runs a non-Exim command in a
4296 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4297 was being done only for the command run by the queryprogram router. It is
4298 now done for all such subprocesses. The other cases are: ${run, transport
4299 filters, and the commands run by the lmtp and pipe transports.
4301 5. Some older OS have a limit of 256 on the maximum number of file
4302 descriptors. Exim was using setrlimit() to set 1000 as a large value
4303 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4304 systems. I've change it so that if it can't get 1000, it tries for 256.
4306 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4307 was an oversight, and furthermore, ever since the addition of extra
4308 controls (e.g. 4.43/32), the checks on when to allow different forms of
4309 "control" were broken. There should now be diagnostics for all cases when a
4310 control that does not make sense is encountered.
4312 7. $recipients is now available in the predata ACL (oversight).
4314 8. Tidy the search cache before the fork to do a delivery from a message
4315 received from the command line. Otherwise the child will trigger a lookup
4316 failure and thereby defer the delivery if it tries to use (for example) a
4317 cached ldap connection that the parent has called unbind on.
4319 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4320 of $address_data from the recipient verification was clobbered by the
4321 sender verification.
4323 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4324 was its contents. (It was OK if the option was not defined at all.)
4326 11. A "Completed" log line is now written for messages that are removed from
4327 the spool by the -Mrm option.
4329 12. $host_address is now set to the target address during the checking of
4330 ignore_target_hosts.
4332 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4333 being passed; this would have caused $sender_host_name to have been used if
4334 matching the list had actually called for a host name (not very likely,
4335 since this list is usually IP addresses). A host name is now passed as
4338 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4339 code that specifies a non-restarting handler (typically sigaction() in
4340 modern systems) in an attempt to fix a rare and obscure crash bug.
4342 15. Narrowed the window for a race in the daemon that could cause it to ignore
4343 SIGCHLD signals. This is not a major problem, because they are used only to
4344 wake it up if nothing else does.
4346 16. A malformed maildirsize file could cause Exim to calculate negative values
4347 for the mailbox size or file count. Odd effects could occur as a result.
4348 The maildirsize information is now recalculated if the size or filecount
4351 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4352 support for a long time. Removed HAVE_SYS_VFS_H.
4354 18. Updated exipick to current release from John Jetmore.
4356 19. Allow an empty sender to be matched against a lookup in an address list.
4357 Previously the only cases considered were a regular expression, or an
4360 20. Exim went into a mad DNS lookup loop when doing a callout where the
4361 host was specified on the transport, if the DNS lookup yielded more than
4364 21. The RFC2047 encoding function was originally intended for short strings
4365 such as real names; it was not keeping to the 75-character limit for
4366 encoded words that the RFC imposes. It now respects the limit, and
4367 generates multiple encoded words if necessary. To be on the safe side, I
4368 have increased the buffer size for the ${rfc2047: expansion operator from
4371 22. Failure to deliver a bounce message always caused it to be frozen, even if
4372 there was an errors_to setting on the router. The errors_to setting is now
4375 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4376 canonical form (fully expanded) before being placed in
4377 $sender_host_address.
4379 24. Updated eximstats to version 1.33
4381 25. Include certificate and key file names in error message when GnuTLS fails
4382 to set them up, because the GnuTLS error message doesn't include the name
4383 of the failing file when there is a problem reading it.
4385 26. Expand error message when OpenSSL has problems setting up cert/key files.
4388 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4389 (this can affect the format of dates).
4391 28. exim_tidydb, when checking for the continued existence of a message for
4392 which it has found a message-specific retry record, was not finding
4393 messages that were in split spool directories. Consequently, it was
4394 deleting retry records that should have stayed in existence.
4396 29. eximstats updated to version 1.35
4397 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4398 1.35 - bugfix such that pie charts by volume are generated correctly
4400 30. The SPA authentication driver was not abandoning authentication and moving
4401 on to the next authenticator when an expansion was forced to fail,
4402 contradicting the general specification for all authenticators. Instead it
4403 was generating a temporary error. It now behaves as specified.
4405 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4406 (the order specifies the preference for clients). The order is now AES256,
4407 AES128, 3DES, ARCFOUR128.
4409 31. Small patch to Sieve code - explicitly set From: when generating an
4412 32. Exim crashed if a remote delivery caused a very long error message to be
4413 recorded - for instance if somebody sent an entire SpamAssassin report back
4414 as a large number of 550 error lines. This bug was coincidentally fixed by
4415 increasing the size of one of Exim's internal buffers (big_buffer) that
4416 happened as part of the Exiscan merge. However, to be on the safe side, I
4417 have made the code more robust (and fixed the comments that describe what
4420 33. Some experimental protocols are using DNS PTR records for new purposes. The
4421 keys for these records are domain names, not reversed IP addresses. The
4422 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4423 leaves it alone. Component reversal etc. now happens only for IP addresses.
4426 34. The host_aton() function is supposed to be passed a string that is known
4427 to be a valid IP address. However, in the case of IPv6 addresses, it was
4428 not checking this. This is a hostage to fortune. Exim now panics and dies
4429 if the condition is not met. A case was found where this could be provoked
4430 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4431 components; fortuitously, this particular loophole had already been fixed
4432 by change 4.50/55 or 4.44/33 above.
4434 If there are any other similar loopholes, the new check in host_aton()
4435 itself should stop them being exploited. The report I received stated that
4436 data on the command line could provoke the exploit when Exim was running as
4437 exim, but did not say which command line option was involved. All I could
4438 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4439 running as the user.
4442 35. There was a buffer overflow vulnerability in the SPA authentication code
4443 (which came originally from the Samba project). I have added a test to the
4444 spa_base64_to_bits() function which I hope fixes it.
4447 36. The daemon start-up calls getloadavg() while still root for those OS that
4448 need the first call to be done as root, but it missed one case: when
4449 deliver_queue_load_max is set with deliver_drop_privilege. This is
4450 necessary for the benefit of the queue runner, because there is no re-exec
4451 when deliver_drop_privilege is set.
4453 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4454 was in use, was not putting the data itself into the right store pool;
4455 consequently, it could be overwritten for a subsequent message in the same
4456 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4459 38. Sometimes the final signoff response after QUIT could fail to get
4460 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4461 < 0 before doing a fflush(). This bug looks as though it goes back to the
4462 introduction of TLS in release 3.20, but "sometimes" must have been rare
4463 because the tests only now provoked it.
4469 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4470 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4471 Now it returns an integer. A number of calls were still expecting a T/F
4472 return. Fortuitously, in all cases, the tests worked in OK situations,
4473 which is the norm. However, things would have gone wrong on any write
4474 failures on the smtp file descriptor. This function is used when sending
4475 messages over SMTP and also when doing verify callouts.
4477 2. When Exim is called to do synchronous delivery of a locally submitted
4478 message (the -odf or -odi options), it no longer closes stderr before doing
4481 3. Implemented the mua_wrapper option.
4483 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4485 5. Implemented the functions header_remove(), header_testname(),
4486 header_add_at_position(), and receive_remove_recipient(), and exported them
4489 6. If an ACL "warn" statement specified the addition of headers, Exim already
4490 inserted X-ACL-Warn: at the start if there was no header name. However, it
4491 was not making this test for the second and subsequent header lines if
4492 there were newlines in the string. This meant that an invalid header could
4493 be inserted if Exim was badly configured.
4495 7. Allow an ACL "warn" statement to add header lines at the start or after all
4496 the Received: headers, as well as at the end.
4498 8. Added the rcpt_4xx retry error code.
4500 9. Added postmaster_mailfrom=xxx to callout verification option.
4502 10. Added mailfrom=xxxx to the callout verification option, for verify=
4505 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4506 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4508 12. Inserted some casts to stop certain compilers warning when using pointer
4509 differences as field lengths or precisions in printf-type calls (mostly
4510 affecting debugging statements).
4512 13. Added optional readline() support for -be (dynamically loaded).
4514 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4515 same clock tick as a message's arrival, so that its received time was the
4516 same as the "first fail" time on the retry record, and that message
4517 remained on the queue past the ultimate address timeout, every queue runner
4518 would try a delivery (because it was past the ultimate address timeout) but
4519 after another failure, the ultimate address timeout, which should have then
4520 bounced the address, did not kick in. This was a "< instead of <=" error;
4521 in most cases the first failure would have been in the next clock tick
4522 after the received time, and all would be well.
4524 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4525 being recognized when the domain list was tested by the match_domain
4526 condition in an expansion string.
4528 16. Added the ${str2b64: operator.
4530 17. Exim was always calling setrlimit() to set a large limit for the number of
4531 processes, without checking whether the existing limit was already
4532 adequate. (It did check for the limit on file descriptors.) Furthermore,
4533 errors from getrlimit() and setrlimit() were being ignored. Now they are
4534 logged to the main and panic logs, but Exim does carry on, to try to do its
4535 job under whatever limits there are.
4537 18. Imported PCRE 5.0.
4539 19. Trivial typo in log message " temporarily refused connection" (the leading
4542 20. If the log selector return_path_on_delivery was set and an address was
4543 redirected to /dev/null, the delivery process crashed because it assumed
4544 that a return path would always be set for a "successful" delivery. In this
4545 case, the whole delivery is bypassed as an optimization, and therefore no
4548 21. Internal re-arrangement: the function for sending a challenge and reading
4549 a response while authentication was assuming a zero-terminated challenge
4550 string. It's now changed to take a pointer and a length, to allow for
4551 binary data in such strings.
4553 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4555 23. Exim was not respecting finduser_retries when seeking the login of the
4556 uid under which it was called; it was always trying 10 times. (The default
4557 setting of finduser_retries is zero.) Also, it was sleeping after the final
4558 failure, which is pointless.
4560 24. Implemented tls_on_connect_ports.
4562 25. Implemented acl_smtp_predata.
4564 26. If the domain in control=submission is set empty, Exim assumes that the
4565 authenticated id is a complete email address when it generates From: or
4566 Sender: header lines.
4568 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4569 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4570 chown and chgrp in /bin and hostname in /usr/bin.
4572 28. Exim was keeping the "process log" file open after each use, just as it
4573 does for the main log. This opens the possibility of it remaining open for
4574 long periods when the USR1 signal hits a daemon. Occasional processlog
4575 errors were reported, that could have been caused by this. Anyway, it seems
4576 much more sensible not to leave this file open at all, so that is what now
4579 29. The long-running daemon process does not normally write to the log once it
4580 has entered its main loop, and it closes the log before doing so. This is
4581 so that log files can straightforwardly be renamed and moved. However,
4582 there are a couple of unusual error situations where the daemon does write
4583 log entries, and I had neglected to close the log afterwards.
4585 30. The text of an SMTP error response that was received during a remote
4586 delivery was being truncated at 512 bytes. This is too short for some of
4587 the long messages that one sometimes sees. I've increased the limit to
4590 31. It is now possible to make retry rules that apply only when a message has a
4591 specific sender, in particular, an empty sender.
4593 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4594 it possible to be selective about when SMTP synchronization is enforced.
4596 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4598 32. Implemented hosts_connection_nolog.
4600 33. Added an ACL for QUIT.
4602 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4605 35. Added mailbox_size and mailbox_filecount to appendfile.
4607 36. Added control = no_multiline_responses to ACLs.
4609 37. There was a bug in the logic of the code that waits for the clock to tick
4610 in the case where the clock went backwards by a substantial amount such
4611 that the microsecond fraction of "now" was more than the microsecond
4612 fraction of "then" (but the whole seconds number was less).
4614 38. Added support for the libradius Radius client library this is found on
4615 FreeBSD (previously only the radiusclient library was supported).
4621 1. When certain lookups returned multiple values in the form name=value, the
4622 quoting of the values was not always being done properly. Specifically:
4623 (a) If the value started with a double quote, but contained no whitespace,
4625 (b) If the value contained whitespace other than a space character (i.e.
4626 tabs or newlines or carriage returns) it was not quoted.
4627 This fix has been applied to the mysql and pgsql lookups by writing a
4628 separate quoting function and calling it from the lookup code. The fix
4629 should probably also be applied to nisplus, ibase and oracle lookups, but
4630 since I cannot test any of those, I have not disturbed their existing code.
4632 2. A hit in the callout cache for a specific address caused a log line with no
4633 reason for rejecting RCPT. Now it says "Previous (cached) callout
4634 verification failure".
4636 3. There was an off-by-one bug in the queryprogram router. An over-long
4637 return line was truncated at 256 instead of 255 characters, thereby
4638 overflowing its buffer with the terminating zero. As well as fixing this, I
4639 have increased the buffer size to 1024 (and made a note to document this).
4641 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4642 when Exim is waiting for an SMTP response from a remote server, Exim
4643 restarts its select() call on the socket, thereby resetting its timeout.
4644 This is not a problem when such interrupts are rare. Somebody set up a cron
4645 job to run exiwhat every 2 minutes, which is less than the normal select()
4646 timeout (5 or 10 minutes). This meant that the select() timeout never
4647 kicked in because it was always reset. I have fixed this by comparing the
4648 time when an interrupt arrives with the time at the start of the first call
4649 to select(). If more time than the timeout has elapsed, the interrupt is
4650 treated as a timeout.
4652 5. Some internal re-factoring in preparation for the addition of Sieve
4653 extensions (by MH). In particular, the "personal" test is moved to a
4654 separate function, and given an option for scanning Cc: and Bcc: (which is
4655 not set for Exim filters).
4657 6. When Exim created an email address using the login of the caller as the
4658 local part (e.g. when creating a From: or Sender: header line), it was not
4659 quoting the local part when it contained special characters such as @.
4661 7. Installed new OpenBSD configuration files.
4663 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4664 try to make them clearer.
4666 9. Callout options, other than the timeout value, were being ignored when
4667 verifying sender addresses in header lines. For example, when using
4669 verify = header_sender/callout=no_cache
4671 the cache was (incorrectly) being used.
4673 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4674 only the Cygwin environment.
4676 11. When return_path_on_delivery was set as a log selector, if different remote
4677 addresses in the same message used different return paths and parallel
4678 remote delivery occurred, the wrong values would sometimes be logged.
4679 (Whenever a remote delivery process finished, the return path value from
4680 the most recently started remote delivery process was logged.)
4682 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4683 lines when AUTH and/or TLS are in use. This is the "received protocol"
4684 field. Exim used to use "asmtp" for authenticated SMTP, without any
4685 indication (in the protocol name) for TLS use. Now it follows the RFC and
4686 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4687 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4688 names appear in log lines as well as in Received: header lines.
4690 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4691 extensions, and comparison tests, and to fix some bugs.
4693 14. Changes to the "personal" filter test:
4695 (1) The test was buggy in that it was just doing the equivalent of
4696 "contains" tests on header lines. For example, if a user's address was
4697 anne@some.where, the "personal" test would incorrectly be true for
4699 To: susanne@some.where
4701 This test is now done by extracting each address from the header in turn,
4702 and checking the entire address. Other tests that are part of "personal"
4703 are now done using regular expressions (for example, to check local parts
4704 of addresses in From: header lines).
4706 (2) The list of non-personal local parts in From: addresses has been
4707 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4708 taken from the Sieve specification recommendations.
4710 (3) If the message contains any header line starting with "List-" it is
4711 treated as non-personal.
4713 (4) The test for "circular" in the Subject: header line has been removed
4714 because it now seems ill-conceived.
4716 15. Minor typos in src/EDITME comments corrected.
4718 16. Installed latest exipick from John Jetmore.
4720 17. If headers_add on a router specified a text string that was too long for
4721 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4722 of string_sprintf() is now avoided.
4724 18. $message_body_size was not set (it was always zero) when running the DATA
4725 ACL and the local_scan() function.
4727 19. For the "mail" command in an Exim filter, no default was being set for
4728 the once_repeat time, causing a random time value to be used if "once" was
4729 specified. (If the value happened to be <= 0, no repeat happened.) The
4730 default is now 0s, meaning "never repeat". The "vacation" command was OK
4731 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4732 (I found it when inspecting the code).
4734 20. There is now an overall timeout for performing a callout verification. It
4735 defaults to 4 times the callout timeout, which applies to individual SMTP
4736 commands during the callout. The overall timeout applies when there is more
4737 than one host that can be tried. The timeout is checked before trying the
4738 next host. This prevents very long delays if there are a large number of
4739 hosts and all are timing out (e.g. when the network connections are timing
4740 out). The value of the overall timeout can be changed by specifying an
4741 additional sub-option for "callout", called "maxwait". For example:
4743 verify = sender/callout=5s,maxwait=20s
4745 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4746 to the end before writing, but this should make it even safer).
4748 22. Exim was forgetting that it had advertised PIPELINING for the second and
4749 subsequent messages on an SMTP connection. It was also not resetting its
4750 memory on STARTTLS and an internal HELO.
4752 23. When Exim logs an SMTP synchronization error within a session, it now
4753 records whether PIPELINING has been advertised or not.
4755 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4756 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4757 rather than long int.
4759 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4761 26. Added the never_mail option to autoreply.
4767 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4768 crash if the getsockname() call failed; this can happen if a connection is
4769 closed very soon after it is established. The problem was simply in the
4770 order in which certain operations were done, causing Exim to try to write
4771 to the SMTP stream before it had set up the file descriptor. The bug has
4772 been fixed by making things happen in the correct order.
4778 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4779 before the connection was closed, thus losing the rejection response.
4781 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4782 some early Solaris releases, but causes trouble in current releases where
4783 socklen_t is defined.
4785 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4788 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4791 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4792 that releasing the top of it at the end releases what was used for sub-
4793 expansions (unless the block got too big). However, discard this block if
4794 the first thing is a variable or header, so that we can use its block when
4795 it is dynamic (useful for very large $message_headers, for example).
4797 6. Lookups now cache *every* query, not just the most recent. A new, separate
4798 store pool is used for this. It can be recovered when all lookup caches are
4799 flushed. Lookups now release memory at the end of their result strings.
4800 This has involved some general refactoring of the lookup sources.
4802 7. Some code has been added to the store_xxx() functions to reduce the amount
4803 of flapping under certain conditions.
4805 8. log_incoming_interface used to affect only the <= reception log lines. Now
4806 it causes the local interface and port to be added to several more SMTP log
4807 lines, for example "SMTP connection from", and rejection lines.
4809 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4811 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4813 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4814 could be overwritten at the end of the current message (or the start of a
4815 new message if it was set in a HELO ACL). The value is now preserved for
4816 the duration of the SMTP connection.
4818 12. If a transport had a headers_rewrite setting, and a matching header line
4819 contained an unqualified address, that address was qualified, even if it
4820 did not match any rewriting rules. The underlying bug was that the values
4821 of the flags that permit the existence of unqualified sender and recipient
4822 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4823 non-local messages, and by -bnq for local messages) were not being
4824 preserved with the message after it was received.
4826 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4827 "next input=" as part of the text comprising the host identity instead of
4828 the correct text. The code was using the same buffer for two different
4829 strings. However, depending on which order the printing function evaluated
4830 its arguments, the bug did not always show up. Under Linux, for example, my
4831 test suite worked just fine.
4833 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4834 doesn't work with some older versions of Perl. It has been changed to "my",
4835 which in any case is probably the better facility to use.
4837 15. A really picky compiler found some instances of statements for creating
4838 error messages that either had too many or two few arguments for the format
4841 16. The size of the buffer for calls to the DNS resolver has been increased
4842 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4843 for addresses that have a lot of PTR records. This alleviates a problem; it
4844 does not fully solve it.
4846 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4847 buffer now truncates the list and logs the incident, which is the same
4848 action as happens when Exim is looking up a host name and its aliases.
4849 Previously in this situation something unpredictable would happen;
4850 sometimes it was "internal error: store_reset failed".
4852 18. If a server dropped the connection unexpectedly when an Exim client was
4853 using GnuTLS and trying to read a response, the client delivery process
4854 crashed while trying to generate an error log message.
4856 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4857 string, for example:
4859 warn message = H1: something\nH2: something
4861 the text was added as a single header line from Exim's point of view
4862 though it ended up OK in the delivered message. However, searching for the
4863 second and subsequent header lines using $h_h2: did not work. This has been
4864 fixed. Similarly, if a system filter added multiple headers in this way,
4865 the routers could not see them.
4867 20. Expanded the error message when iplsearch is called with an invalid key to
4868 suggest using net-iplsearch in a host list.
4870 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4871 ACLs are no longer actually imposed (and a message to that effect is
4874 22. If a "gecos" field in a passwd entry contained escaped characters, in
4875 particular, if it contained a \" sequence, Exim got it wrong when building
4876 a From: or a Sender: header from that name. A second bug also caused
4877 incorrect handling when an unquoted " was present following a character
4878 that needed quoting.
4880 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4881 was not being matched caselessly.
4883 24. Arranged for all hyphens in the exim.8 source to be escaped with
4886 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4887 properly. Recipient callout cache records were still being keyed to include
4888 the sender, even when use_sender was set false. This led to far more
4889 callouts that were necessary. The sender is no longer included in the key
4890 when use_sender is false.
4892 26. Added "control = submission" modifier to ACLs.
4894 27. Added the ${base62d: operator to decode base 62 numbers.
4896 28. dnsdb lookups can now access SRV records.
4898 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4899 the configuration file.
4901 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4902 (-v) mode. This makes the output for a verbose queue run more intelligible.
4904 31. Added a use_postmaster feature to recipient callouts.
4906 32. Added the $body_zerocount variable, containing the number of binary zero
4907 bytes in the message body.
4909 33. The time of last modification of the "new" subdirectory is now used as the
4910 "mailbox time last read" when there is a quota error for a maildir
4913 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4915 35. Added +ignore_unknown as a special item in host lists.
4917 36. Code for decoding IPv6 addresses in host lists is now included, even if
4918 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4919 address was recognized as an IP address, but was then not correctly decoded
4920 into binary, causing unexpected and incorrect effects when compared with
4927 1. Very minor rewording of debugging text in manualroute to say "list of
4928 hosts" instead of "hostlist".
4930 2. If verify=header_syntax was set, and a header line with an unqualified
4931 address (no domain) and a large number of spaces between the end of the
4932 name and the colon was received, the reception process suffered a buffer
4933 overflow, and (when I tested it) crashed. This was caused by some obsolete
4934 code that should have been removed. The fix is to remove it!
4936 3. When running in the test harness, delay a bit after writing a bounce
4937 message to get a bit more predictability in the log output.
4939 4. Added a call to search_tidyup() just before forking a reception process. In
4940 theory, someone could use a lookup in the expansion of smtp_accept_max_
4941 per_host which, without the tidyup, could leave open a database connection.
4943 5. Added the variables $recipient_data and $sender_data which get set from a
4944 lookup success in an ACL "recipients" or "senders" condition, or a router
4945 "senders" option, similar to $domain_data and $local_part_data.
4947 6. Moved the writing of debug_print from before to after the "senders" test
4950 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4951 problems for message scanning, either using a data ACL, or using
4952 local_scan() because the Received: header was not generated till after they
4953 were called (in order to set the time as the time of reception completion).
4954 I have revised the way this works. The header is now generated after the
4955 body is received, but before the ACL or local_scan() are called. After they
4956 are run, the timestamp in the header is updated.
4962 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4963 before starting a queue runner without re-exec. This happened only when
4964 deliver_drop_privilege was set or when the Exim user was set to root. The
4965 effect of the bug was that timeouts during subsequent deliveries caused
4966 crashes instead of being properly handled. The handler is now left at its
4967 default (and expected) setting.
4969 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4970 message, again when deliver_drop_privilege is set or Exim is run as root.
4971 The bug described in (1) was not present in this case, but the tidying up
4972 of the other signals was missing. I have made the two cases consistent.
4974 3. The ignore_target_hosts setting on a manualroute router was being ignored
4975 for hosts that were looked up using the /MX notation.
4977 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4980 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4981 operated on the sender address. After changing the $sender_address to <>
4982 for the sender address verify, Exim was re-instated it as the original
4983 (before rewriting) address, but remembering that it had rewritten it, so it
4984 wasn't rewriting it again. This bug also had the effect of breaking the
4985 sender address verification caching when the sender address was rewritten.
4987 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4988 This has been changed so that if the ip literal address matches
4989 ignore_target_hosts, the router declines.
4991 7. Added expansion conditions match_domain, match_address, and match_local_
4992 part (NOT match_host).
4994 8. The placeholder for the Received: header didn't have a length field set.
4996 9. Added code to Exim itself and to exim_lock to test for a specific race
4997 condition that could lead to file corruption when using MBX delivery. The
4998 issue is with the lockfile that is created in /tmp. If this file is removed
4999 after a process has opened it but before that process has acquired a lock,
5000 there is the potential for a second process to recreate the file and also
5001 acquire a lock. This could lead to two Exim processes writing to the file
5002 at the same time. The added code performs the same test as UW imapd; it
5003 checks after acquiring the lock that its file descriptor still refers to
5004 the same named file.
5006 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5007 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5010 11. Added the smtp_active_hostname option. If used, this will typically be made
5011 to depend on the incoming interface address. Because $interface_address is
5012 not set up until the daemon has forked a reception process, error responses
5013 that can happen earlier (such as "too many connections") no longer contain
5016 12. If an expansion in a condition on a "warn" statement fails because a lookup
5017 defers, the "warn" statement is abandoned, and the next ACL statement is
5018 processed. Previously this caused the whole ACL to be aborted.
5020 13. Added the iplsearch lookup type.
5022 14. Added ident_timeout as a log selector.
5024 15. Added tls_certificate_verified as a log selector.
5026 16. Added a global option tls_require_ciphers (compare the smtp transport
5027 option of the same name). This controls incoming TLS connections.
5029 17. I finally figured out how to make tls_require_ciphers do a similar thing
5030 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5031 before starting the TLS session.
5033 18. Tabs are now shown as \t in -bP output.
5035 19. If the log selector return_path_on_delivery was set, Exim crashed when
5036 bouncing a message because it had too many Received: header lines.
5038 20. If two routers both had headers_remove settings, and the first one included
5039 a superfluous trailing colon, the final name in the first list and the
5040 first name in the second list were incorrectly joined into one item (with a
5041 colon in the middle).
5047 1. Added -C and -D options to the exinext utility, mainly to make it easier
5048 to include in the automated testing, but these could be helpful when
5049 multiple configurations are in use.
5051 2. The exinext utility was not formatting the output nicely when there was
5052 an alternate port involved in the retry record key, nor when there was a
5053 message id as well (for retries that were specific to a specific message
5054 and a specific host). It was also confused by IPv6 addresses, because of
5055 the additional colons they contain. I have fixed the IPv4 problem, and
5056 patched it up to do a reasonable job for IPv6.
5058 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5059 delivery, the log line now contains "pipelined" if PIPELINING was used.
5061 4. An SMTP transport process used to panic and die if the bind() call to set
5062 an explicit outgoing interface failed. This has been changed; it is now
5063 treated in the same way as a connect() failure.
5065 5. A reference to $sender_host_name in the part of a conditional expansion
5066 that was being skipped was still causing a DNS lookup. This no longer
5069 6. The def: expansion condition was not recognizing references to header lines
5070 that used bh_ and bheader_.
5072 7. Added the _cache feature to named lists.
5074 8. The code for checking quota_filecount in the appendfile transport was
5075 allowing one more file than it should have been.
5077 9. For compatibility with Sendmail, the command line option
5085 and sets the incoming protocol and host name (for trusted callers). The
5086 host name and its colon can be omitted when only the protocol is to be set.
5087 Note the Exim already has two private options, -pd and -ps, that refer to
5088 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5089 "s", but I don't think that's a major issue.
5091 10. A number of refactoring changes to the code, none of which should affect
5094 (a) The number of logging options was getting close to filling up the
5095 32-bit word that was used as a bit map. I have split them into two classes:
5096 those that are passed in the argument to log_write(), and those that are
5097 only ever tested independently outside of that function. These are now in
5098 separate 32-bit words, so there is plenty of room for expansion again.
5099 There is no change in the user interface or the logging behaviour.
5101 (b) When building, for example, log lines, the code previously used a
5102 macro that called string_cat() twice, in order to add two strings. This is
5103 not really sufficiently general. Furthermore, there was one instance where
5104 it was actually wrong because one of the argument was used twice, and in
5105 one call a function was used. (As it happened, calling the function twice
5106 did not affect the overall behaviour.) The macro has been replaced by a
5107 function that can join an arbitrary number of extra strings onto a growing
5110 (c) The code for expansion conditions now uses a table and a binary chop
5111 instead of a serial search (which was left over from when there were very
5112 few conditions). Also, it now recognizes conditions like "pam" even when
5113 the relevant support is not compiled in: a suitably worded error message is
5114 given if an attempt is made to use such a condition.
5116 11. Added ${time_interval:xxxxx}.
5118 12. A bug was causing one of the ddress fields not to be passed back correctly
5119 from remote delivery subprocesses. The field in question was not being
5120 subsequently used, so this caused to problems in practice.
5122 13. Added new log selectors queue_time and deliver_time.
5124 14. Might have fixed a bug in maildirsizefile handling that threw up
5125 "unexpected character" debug warnings, and recalculated the data
5126 unnecessarily. In any case, I expanded the warning message to give more
5129 15. Added the message "Restricted characters in address" to the statements in
5130 the default ACL that block characters like @ and % in local parts.
5132 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5133 Three changes have been made:
5135 (a) There was a serious bug; a negative response to MAIL caused the whole
5136 recipient domain to be cached as invalid, thereby blocking all messages
5137 to all local parts at the same domain, from all senders. This bug has
5138 been fixed. The domain is no longer cached after a negative response to
5139 MAIL if the sender used is not empty.
5141 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5144 (c) A new callout option, "use_sender" has been added for people who want
5145 the modified behaviour.
5151 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5154 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5155 indeed breaks things for older releases.
5157 3. Added additional logging to the case where there is a problem reading data
5158 from a filter that is running in a subprocess using a pipe, in order to
5159 try to track down a specific problem.
5161 4. Testing facility fudge: when running in the test harness and attempting
5162 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5163 getting "No route to host". Convert this to a timeout.
5165 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5168 6. Some OS don't have socklen_t but use size_t instead. This affects the
5169 fifth argument of getsockopt() amongst other things. This is now
5170 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5171 can be set for individual OS. I have set it for SunOS5, OSF1, and
5172 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5173 some earlier ones do not.
5175 7. Change 4.30/15 was not doing the test caselessly.
5177 8. The standard form for an IPv6 address literal was being rejected by address
5178 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5179 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5180 this, as well as the form without the "IPv6" on the front (but only when
5181 address literals are enabled, of course).
5183 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5185 10. Exim crashed if a message with an empty sender address specified by -f
5186 encountered a router with an errors_to setting. This could be provoked only
5187 by a command such as
5191 where an empty string was supplied; "<>" did not hit this bug.
5193 11. Installed PCRE release 4.5.
5195 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5196 remained set. It is now erased.
5198 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5199 times from message ids (which are base 36 rather than the normal 62).
5201 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5202 were being counted as actual protocol errors, and logged if the log
5203 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5204 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5205 and DATA following a set of rejected RCPTs do not count as protocol errors.
5206 In other words, Exim assumes they were pipelined, though this may not
5207 actually be the case. Of course, in all cases the client gets an
5208 appropriate error code.
5210 15. If a lookup fails in an ACL condition, a message about the failure may
5211 be available; it is used if testing the ACL cannot continue, because most
5212 such messages specify what the cause of the deferral is. However, some
5213 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5214 that caused an old message to be retained and used if a later statement
5215 caused a defer, replacing the real cause of the deferral.
5217 16. If an IP address had so many PTR records that the DNS lookup buffer
5218 was not large enough to hold them, Exim could crash while trying to process
5219 the truncated data. It now detects and logs this case.
5221 17. Further to 4.21/58, another change has been made: if (and only if) the
5222 first line of a message (the first header line) ends with CRLF, a bare LF
5223 in a subsequent header line has a space inserted after it, so as not to
5224 terminate the header.
5226 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5227 unnecessarily, used atoi() instead of strtol(), and didn't check the
5228 termination when getting file sizes from file names by regex.
5230 19. Completely re-implemented the support for maildirsize files, in the light
5231 of a number of problems with the previous contributed implementation
5232 (4.30/29). In particular:
5234 . If the quota is zero, the maildirsize file is maintained, but no quota is
5237 . If the maildir directory does not exist, it is created before any attempt
5238 to write a maildirsize file.
5240 . The quota value in the file is just a cache; if the quota is changed in
5241 the transport, the new value overrides.
5243 . A regular expression is available for excluding directories from the
5246 20. The autoreply transport checks the characters in options that define the
5247 message's headers; it allows continued headers, but it was checking with
5248 isspace() after an embedded newline instead of explicitly looking for a
5251 21. If all the "regular" hosts to which an address was routed had passed their
5252 expiry times, and had not reached their retry times, the address was
5253 bounced, even if fallback hosts were defined. Now Exim should go on to try
5256 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5257 equivalent code in the SMTP transport. Some hosts send humungous responses
5258 to HELO/EHLO, more than 1024 it seems.
5260 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5261 gives compiler warnings in some environments. I've now done it "properly",
5264 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5265 (because of problems with the built-in one) was declared to return uschar *
5266 instead of char *, causing compiler failure.
5268 25. Fixed a file descriptor leak when processing alias/forward files.
5270 26. Fixed a minor format string issue in dbfn.c.
5272 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5274 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5275 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5278 29. When a sender address is verified, it is cached, to save repeating the test
5279 when there is more than one recipient in a message. However, when the
5280 verification involves a callout, it is possible for different callout
5281 options to be set for different recipients. It is too complicated to keep
5282 track of this in the cache, so now Exim always runs a verification when a
5283 callout is required, relying on the callout cache for the optimization.
5284 The overhead is duplication of the address routing, but this should not be
5287 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5288 to be verified with callout=postmaster, and the main callout worked but the
5289 postmaster check failed, the verification correctly failed. However, if a
5290 subsequent RCPT command asked for sender verification *without* the
5291 postmaster check, incorrect caching caused this verification also to fail,
5294 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5295 it was not caching the DNS options (qualify_single, search_parents) that
5296 were used when the lookup failed. A subsequent lookup with different
5297 options therefore always gave the same answer, though there were cases
5298 where it should not have. (Example: a "domains = !$mx_any" option on a
5299 dnslookup router: the "domains" option is always processed without any
5300 widening, but the router might have qualify_single set.) Now Exim uses the
5301 cached value only when the same options are set.
5303 32. Added John Jetmore's "exipick" utility to the distribution.
5305 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5306 than a timeout (e.g. a certificate is required, and is not provided), an
5307 Exim server now closes the connection immediately. Previously it waited for
5308 the client to close - but if the client is SSL, it seems that they each
5309 wait for each other, leading to a delay before one of them times out.
5311 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5312 maintained 0.8.x compatibility because I don't think many are using it, and
5313 it is clearly obsolete.
5315 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5318 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5319 client certificate was expired. A simple patch fixes this, though I don't
5320 understand the full logic of why the verify callback is called multiple
5323 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5324 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5325 which causes problems with some clients (such as the Certicom SSL Plus
5326 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5327 disables the coutermeasure allowing Eudora to connect."
5329 38. Exim was not checking that a write() to a log file succeeded. This could
5330 lead to Bad Things if a log got too big, in particular if it hit a file
5331 size limit. Exim now panics and dies if it cannot write to a log file, just
5332 as it does if it cannot open a log file.
5334 39. Modified OS/Makefile-Linux so that it now contains
5336 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5338 The two -D definitions ensure that Exim is compiled with large file
5339 support, which makes it possible to handle log files that are bigger than
5342 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5343 instance) a domain was checked against a named list that involved a lookup,
5344 causing $domain_data to be set, then another domain was checked against the
5345 same list, then the first domain was re-checked, the value of $domain_data
5346 after the final check could be wrong. In particular, if the second check
5347 failed, it could be set empty. This bug probably also applied to
5350 41. The strip_trailing_dot option was not being applied to the address given
5351 with the -f command-line option.
5353 42. The code for reading a message's header from the spool was incrementing
5354 $received_count, but never initializing it. This meant that the value was
5355 incorrect (doubled) while delivering a message in the same process in which
5356 it was received. In the most common configuration of Exim, this never
5357 happens - a fresh exec is done - but it can happen when
5358 deliver_drop_privilege is set.
5360 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5361 it now includes up to 150 characters of the unexpected data in the log
5364 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5365 and building data strings. The size of both of these buffers was 10 000
5366 bytes - far larger than anybody would *ever* want, thought I. Needless to
5367 say, somebody hit the limit. I have increased the maximum line length to
5368 20 000 and the maximum data length of concatenated lines to 100 000. I have
5369 also fixed two bugs, because there was no checking on these buffers. Tsk,
5370 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5371 buffer is too small.
5373 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5374 lsearch lookups. Now it does.
5376 46. When parsing a route_list item in a manualroute router, a fixed-length
5377 buffer was used for the list of hosts. I made this 1024 bytes long,
5378 thinking that nobody would ever have a list of hosts that long. Wrong.
5379 Somebody had a whole pile of complicated expansion conditions, and the
5380 string was silently truncated, leading to an expansion error. It turns out
5381 that it is easier to change to an unlimited length (owing to other changes
5382 that have happened since this code was originally written) than to build
5383 structure for giving a limitation error. The length of the item that
5384 expands into the list of hosts is now unlimited.
5386 47. The lsearch lookup could not handle data where the length of text line was
5387 more than 4095 characters. Such lines were truncated, leading to shortened
5388 data being returned. It should now handle lines of any length.
5390 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5391 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5394 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5395 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5396 no longer happens in comment lines. A list of replacements is now placed
5397 at the head of all of the source files, except those whose only change is
5398 to replace PERL_COMMAND in the very first #! line.
5400 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5401 messages on the queue, with a bottom-up merge sort, using code contributed
5402 by Michael Haardt. This should make operations like -bp somewhat faster on
5403 large queues. It won't affect queue runners, except when queue_run_in_order
5406 51. Installed eximstats 1.31 in the distribution.
5408 52. Added support for SRV lookups to the dnslookup router.
5410 53. If an ACL referred to $message_body or $message_body_end, the value was not
5411 reset for any messages that followed in the same SMTP session.
5413 54. The store-handling optimization for building very long strings was not
5414 differentiating between the different store pools. I don't think this
5415 actually made any difference in practice, but I've tidied it.
5417 55. While running the routers to verify a sender address, $sender_address
5418 was still set to the sender address. This is wrong, because when routing to
5419 send a bounce to the sender, it would be empty. Therefore, I have changed
5420 it so that, while verifying a sender address, $sender_address is set to <>.
5421 (There is no change to what happens when verifying a recipient address.)
5423 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5424 target A or AAAA records (if not already returned) without resetting the
5425 qualify_single or search_parents options of the DNS resolver. These are
5426 inappropriate in this case because the targets of MX and SRV records must
5427 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5428 target that, when qualified, matched something in the local domain. These
5429 two options are now turned off when doing these lookups.
5431 57. It seems that at least some releases of Reiserfs (which does not have the
5432 concept of a fixed number of inodes) returns zero and not -1 for the
5433 number of available inodes. This interacted badly with check_spool_inodes,
5434 which assumed that -1 was the "no such thing" setting. What I have done is
5435 to check that the total number of inodes is greater than zero before doing
5436 the test of how many are available.
5438 58. When a "warn" ACL statement has a log_message modifier, the message is
5439 remembered, and not repeated. This is to avoid a lot of repetition when a
5440 message has many recipients that cause the same warning to be written.
5441 Howewer, Exim was preserving the list of already written lines for an
5442 entire SMTP session, which doesn't seem right. The memory is now reset if a
5443 new message is started.
5445 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5446 debugging output unless log_rewrite was also set.
5448 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5449 of (int)(handle) when we know that handle contains (void *)(-1).
5451 61. The Exim daemon panic-logs an error return when it closes the incoming
5452 connection. However "connection reset by peer" seems to be common, and
5453 isn't really an error worthy of noting specially, so that particular error
5456 62. When Exim is trying to find all the local interfaces, it used to panic and
5457 die if the ioctl to get the interface flags failed. However, it seems that
5458 on at least one OS (Solaris 9) it is possible to have an interface that is
5459 included in the list of interfaces, but for which you get a failure error
5460 for this call. This happens when the interface is not "plumbed" into a
5461 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5462 failure of the "get flags" call assumes that the interface is down.
5464 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5465 makes life easier for people who are doing arithmetic on fields extracted
5466 from dates, where you often get leading zeros that should not be
5467 interpreted as octal.
5469 64. Added qualify_domain to the redirect router, to override the global
5472 65. If a pathologically long header line contained very many addresses (the
5473 report of this problem mentioned 10 000) and each of them was rewritten,
5474 Exim could use up a very large amount of memory. (It kept on making new
5475 copies of the header line as it rewrote, and never released the old ones.)
5476 At the expense of a bit more processing, the header rewriting function has
5477 been changed so that it no longer eats memory in this way.
5479 66. The generation of the Received: header has been moved from the time that a
5480 message starts to be received, to the time that it finishes. The timestamp
5481 in the Received: header should now be very close to that of the <= log
5482 line. There are two side-effects of this change:
5484 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5485 logged header lines no longer include the local Received: line, because
5486 it has not yet been created. The same applies to a copy of the message
5487 that is returned to a non-SMTP sender when a message is rejected.
5489 (b) When a filter file is tested using -bf, no additional Received: header
5490 is added to the test message. After some thought, I decided that this
5493 This change does not affect the value of $received_for. It is still set
5494 after address rewriting, but before local_scan() is called.
5496 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5498 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5499 gave an unhelpful panic error message, and a defer error. I have managed to
5500 change this behaviour so that it now rejects any supplied certificate,
5501 which seems right, as the list of acceptable certificates is empty.
5503 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5504 gave an unhelpful defer error. I have not managed to make this reject any
5505 supplied certificates, but the error message it gives is "no certificate
5506 supplied", which is not helpful.
5508 70. exigrep's output now also includes lines that are not associated with any
5509 message, but which match the given pattern. Implemented by a patch from
5510 Martin Sluka, which also tidied up the Perl a bit.
5512 71. Recipient callout verification, like sender verification, was using <> in
5513 the MAIL FROM command. This isn't really the right thing, since the actual
5514 sender may affect whether the remote host accepts the recipient or not. I
5515 have changed it to use the actual sender in the callout; this means that
5516 the cache record is now keyed on a recipient/sender pair, not just the
5517 recipient address. There doesn't seem to be a real danger of callout loops,
5518 since a callout by the remote host to check the sender would use <>.
5519 [SEE ABOVE: changed after hitting problems.]
5521 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5522 temporary errors. However, in the case of such a code being given after
5523 the end of a data transmission (i.e. after ".") Exim was failing to write
5524 a retry record for the message. (Yes, there was some broken host that was
5525 actually sending 8xx at this point.)
5527 73. An unknown lookup type in a host list could cause Exim to panic-die when
5528 the list was checked. (An example that provoked this was putting <; in the
5529 middle of a list instead of at the start.) If this happened during a DATA
5530 ACL check, a -D file could be left lying around. This kind of configuration
5531 error no longer causes Exim to die; instead it causes a defer errror. The
5532 incident is still logged to the main and panic logs.
5534 74. Buglet left over from Exim 3 conversion. The message "too many messages
5535 in one connection" was written to the rejectlog but not the mainlog, except
5536 when address rewriting (yes!) was being logged.
5538 75. Added write_rejectlog option.
5540 76. When a system filter was run not as root (that is, when system_filter_user
5541 was set), the values of the $n variables were not being returned to the
5542 main process; thus, they were not subsequently available in the $sn
5545 77. Added +return_path_on_delivery log selector.
5547 78. A connection timeout was being treated differently from recipients deferred
5548 when testing hosts_max_try with a message that was older than the host's
5549 retry timeout. (The host should not be counted, thus allowing all hosts to
5550 be tried at least once before bouncing.) This may have been the cause of an
5551 occasionally reported bug whereby a message would remain on the queue
5552 longer than the retry timeout, but would be bounced if a delivery was
5553 forced. I say "may" because I never totally pinned down the problem;
5554 setting up timeout/retry tests is difficult. See also the next item.
5556 79. The ultimate address timeout was not being applied to errors that involved
5557 a combination of host plus message (for example, a timeout on a MAIL
5558 command). When an address resolved to a number of possible hosts, and they
5559 were not all tried for each delivery (e.g. because of hosts_max_try), a
5560 message could remain on the queue longer than the retry timeout.
5562 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5565 81. Fixed an obscure SMTP outgoing bug which required at least the following
5566 conditions: (a) there was another message waiting for the same server;
5567 (b) the server returned 5xx to all RCPT commands in the first message so
5568 that the message was not completed; (c) the server dropped the connection
5569 or gave a negative response to the RSET that Exim sends to abort the
5570 transaction. The observed case was a dropped connection after DATA that had
5571 been sent in pipelining mode. That is, the server had advertised PIPELINING
5572 but was not implementing it correctly. The effect of the bug was incorrect
5573 behaviour, such as trying another host, and this could lead to a crash.
5579 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5580 and daemon.c were passed as pointers to ints; they should have been
5581 pointers to socklen_t variables (which are typically unsigned ints).
5583 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5586 3. Fixed a really odd bug that affected only the testing scheme; patching a
5587 certain fixed string in the binary changed the value of another string that
5588 happened to be identical to the end of the original first string.
5590 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5591 name", it returns that address as the IP address. On some operating
5592 systems (e.g. Solaris), it also passes back the IP address string as the
5593 "host name". However, on others (e.g. Linux), it passes back an empty
5594 string. Exim wasn't checking for this, and was changing the host name to an
5595 empty string, assuming it had been canonicized.
5597 5. Although rare, it is permitted to have more than one PTR record for a given
5598 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5599 all the names associated with an address, because they do in Solaris.
5600 However, it seems that they do not in Linux for data that comes from the
5601 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5602 I found this out when I moved to a new Linux workstation and tried to run
5603 the Exim test suite.
5605 To get round this problem I have changed the code so that it now does its
5606 own call to the DNS to look up PTR records when searching for a host name.
5607 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5608 addresses that are only in /etc/hosts are still found.
5610 This behaviour is, however, controlled by an option called host_lookup_
5611 order, which defaults to "bydns:byaddr". If people want to use the other
5612 order, or indeed, just use one or the other means of lookup, they can
5613 specify it in this variable.
5615 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5616 some operating systems, this comes back from gethostbyaddr() as an empty
5617 string, and this is what Exim used to test for. However, it seems that in
5618 other systems, "." is yielded. Exim now tests for this case too.
5620 7. The values of check_spool_space and check_log_space are now held internally
5621 as a number of kilobytes instead of an absolute number of bytes. If a
5622 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5623 kilobyte. This means that much larger values can be stored.
5625 8. Exim monitor: an attempt to get the action menu when not actually pointing
5626 at a message produces an empty menu entitled "No message selected". This
5627 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5628 no entries in it ("Shell widget menu has zero width and/or height"). So I
5629 have added a single, blank menu entry in this case.
5631 9. Added ${quote_local_part.
5633 10. MIME decoding is now applied to the contents of Subject: header lines when
5636 11. Now that a reference to $sender_host_address automatically causes a reverse
5637 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5638 host lookup before query-style lookups in lists that might use this
5639 variable. This has therefore been abolished, and the "net-" prefix is no
5640 longer necessary for query-style lookups.
5642 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5643 have been a typo for LFLAGS, so it has been changed.
5645 13. The install script calls Exim with "-C /dev/null" in order to find the
5646 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5647 to be output. Howeve, since Exim outputs its version number before the
5648 error, it didn't break the script. It just looked ugly. I fixed this by
5649 always allowing "-C /dev/null" if the caller is root.
5651 14. Ignore overlarge ACL variable number when reading spool file - insurance
5652 against a later release with more variables having written the file.
5654 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5655 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5656 this, as well as the form without the "IPv6" on the front.
5658 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5659 OS/Makefile-Darwin file.
5661 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5662 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5664 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5665 variable. However, some OS use ints for inodes. I've added cast to long int
5666 to get rid of the compiler warning.
5668 19. I had forgotten to lock out "/../" in configuration file names when
5669 ALT_CONFIG_PREFIX was set.
5671 20. Routers used for verification do not need to specify transports. However,
5672 if such a router generated a host list, and callout was configured, Exim
5673 crashed, because it could not find a port number from the (non-existent)
5674 transport. It now assumes port 25 in this circumstance.
5676 21. Added the -t option to exigrep.
5678 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5679 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5680 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5681 them is set, LOOKUP_LSEARCH is forced.
5683 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5685 24. Added sender and host information to the "rejected by local_scan()" log
5686 line; previously there was no indication of these.
5688 25. Added .include_if_exists.
5690 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5691 Linux. It turns out that not all file systems support this. Apparently some
5692 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5693 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5694 not supported on the file descriptor, is now ignored when Exim is trying to
5695 sync a directory. This applies only to Linux.
5697 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5699 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5700 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5703 29. Added support for maildirsize files from supplied patch (modified a bit).
5705 30. The use of :fail: followed by an empty string could lead Exim to respond to
5706 sender verification failures with (e.g.):
5708 550 Verification failed for <xxx>
5709 550 Sender verify failed
5711 where the first response line was missing the '-' that indicates it is not
5712 the final line of the response.
5714 31. The loop for finding the name of the user that called Exim had a hardwired
5715 limit of 10; it now uses the value of finduser_retries, which is used for
5716 all other user lookups.
5718 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5721 33. Exim was neglecting to zero errno before one call of strtol() when
5722 expanding a string and expecting an integer value. On some systems this
5723 resulted in spurious "integer overflow" errors. Also, it was casting the
5724 result into an int without checking.
5726 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5727 did not work. The code looks as if it has *never* worked, though it appears
5728 to have been documented since at least releast 1.62. I have made it work.
5730 35. The "timeout_DNS" error in retry rules, also documented since at least
5731 1.62, also never worked. As it isn't clear exactly what this means, and
5732 clearly it isn't a major issue, I have abolished the feature by treating it
5733 as "timeout", and writing a warning to the main and panic logs.
5735 36. The display of retry rules for -brt wasn't always showing the error code
5738 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5739 timeout_connect_A, timeout_connect_MX.
5741 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5742 to the empty sender.
5744 39. The daemon was not analysing the content of -oX till after it had closed
5745 stderr and disconnected from the controlling terminal. This meant that any
5746 syntax errors were only noted on the panic log, and the return code from
5747 the command was 0. By re-arranging the code a little, I've made the
5748 decoding happen first, so such errors now appear on stderr, and the return
5749 code is 1. However, the actual setting up of the sockets still happens in
5750 the disconnected process, so errors there are still only recorded on the
5753 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5754 connections (as happens on some IP stacks) was logged at start up time as
5755 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5756 it from "IPv6 and IPv4", which means that two separate sockets are being
5759 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5760 says whether AF_INET or AF_INET6 was passed as an argument.
5762 42. Exiwhat output was messed up when time zones were included in log
5765 43. Exiwhat now gives more information about the daemon's listening ports,
5766 and whether -tls-on-connect was used.
5768 44. The "port" option of the smtp transport is now expanded.
5770 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5771 silently ignored. Now an error message is written to the main and panic
5774 46. There's a new ACL modifier called "logwrite" which writes to a log file
5775 as soon as it is encountered.
5777 47. Added $local_user_uid and $local_user_gid at routing time.
5779 48. Exim crashed when trying to verify a sender address that was being
5782 49. Exim was recognizing only a space character after ".include". It now also
5783 recognizes a tab character.
5785 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5786 extracting the relevant information from the specification. The man page no
5787 longer contains scrambled data for the -d option, and I've added a section
5788 at the front about calling Exim under different names.
5790 51. Added "extra_headers" argument to the "mail" command in filter files.
5792 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5795 53. Installed eximstats 1.29.
5797 54. Added transport_filter_timeout as a generic transport option.
5799 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5800 Cc: header lines. This was required by RFC 822, but it not required by RFC
5803 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5804 incoming messages that did not have them. Now it does so only if the
5805 message originates locally, that is, if there is no associated remote host
5806 address. When Resent- header lines are present, this applies to the Resent-
5807 lines rather than the non-Resent- lines.
5809 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5810 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5812 58. Messages for configuration errors now include the name of the main
5813 configuration file - useful now that there may be more than one file in a
5814 list (.included file names were always shown).
5816 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5817 for those rare installations that do not start the daemon as root or run it
5818 setuid root. I've cut out the call to initgroups() if the daemon is not
5821 60. The Exim user and group can now be bound into the binary as text strings
5822 that are looked up at the start of Exim's processing.
5824 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5826 62. Added $mailstore_basename variable.
5828 63. Installed patch to sieve.c from Michael Haardt.
5830 64. When Exim failed to open the panic log after failing to open the main log,
5831 the original message it was trying to log was written to stderr and debug
5832 output, but if they were not available (the usual case in production), it
5833 was lost. Now it is written to syslog before the two lines that record the
5834 failures to open the logs.
5836 65. Users' Exim filters run in subprocesses under the user's uid. It is
5837 possible for a "deliver" command or an alias in a "personal" command to
5838 provoke an address rewrite. If logging of address rewriting is configured,
5839 this fails because the process is not running as root or exim. There may be
5840 a better way of dealing with this, but for the moment (because 4.30 needs
5841 to be released), I have disabled address rewrite logging when running a
5842 filter in a non-root, non-exim process.
5848 1. The buildconfig auxiliary program wasn't quoting the value set for
5849 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5850 not defined. This bug was masked in 4.22 by the effect that was fixed in
5853 2. Some messages that were rejected after a message id was allocated were
5854 shown as "incomplete" by exigrep. It no longer does this for messages that
5855 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5857 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5858 have allow_domain_literals set, the ID did not get logged in the <= line.
5859 Domain literals are now always recognized in Message-ID: header lines.
5861 4. The first argument for a ${extract expansion item is the key name or field
5862 number. Leading and trailing spaces in this item were not being ignored,
5863 causing some misleading effects.
5865 5. When deliver_drop_privilege was set, single queue runner processes started
5866 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5867 same command in the process it spins off) were not dropping privilege.
5869 6. When the daemon running as "exim" started a queue runner, it always
5870 re-executed Exim in the spun-off process. This is a waste of effort when
5871 deliver_drop_privilege is set. The new process now just calls the
5872 queue-runner function directly.
5878 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5881 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5882 ignored. Though the use of -oP was forcing the writing of a pid file, it
5883 was always written to the default place.
5885 3. If the message "no IP address found for host xxxx" is generated during
5886 incoming verification, it is now followed by identification of the incoming
5887 connection (so you can more easily find what provoked it).
5889 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5891 5. Added some features to "harden" Exim a bit more against certain attacks:
5893 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5894 be put in Local/Makefile. This is like the never_users runtime option,
5895 but it cannot be overridden. The default setting is "root".
5897 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5898 prefix string with which any file named in a -C command line option
5901 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5902 is retained for -C and -D only if the caller of Exim is root. Without
5903 it, the exim user may also use -C and -D and retain privilege.
5905 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5906 command line option is disabled.
5908 6. Macro names set by the -D option must start with an upper case letter, just
5909 like macro names defined in the configuration file.
5911 7. Added "dereference=" facility to LDAP.
5913 8. Two instances of the typo "uknown" in the source files are fixed.
5915 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5916 the Configure-Makefile script screwed up while processing it.
5918 10. Incorporated PCRE 4.4.
5920 11. The SMTP synchronization check was not operating right at the start of an
5921 SMTP session. For example, it could not catch a HELO sent before the client
5922 waited for the greeting. There is now a check for outstanding input at the
5923 point when the greeting is written. Because of the duplex, asynchronous
5924 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5925 way, but not yet received, when the check is performed.
5927 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5928 on TCP/IP sockets, because this apparently causes some broken clients to
5931 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5932 unchanged) from the Cygwin maintainer.
5934 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5935 was supported instead of testing for mbx. Effectively a typo.
5937 15. The spa authenticator server code was not checking that the input it
5938 received was valid base64.
5940 16. The debug output line for the "set" modifier in ACLs was not showing the
5941 name of the variable that was being set.
5943 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5945 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5946 lookup on its own. Something else was needed to trigger the lookup. For
5947 example, a match in host_lookup or the need for a host name in a host list.
5948 Now, if $sender_host_name is referenced and the host name has not yet been
5949 looked up, a lookup is performed. If the lookup fails, the variable remains
5950 empty, and $host_lookup_failed is set to "1".
5952 19. Added "eqi" as a case-independent comparison operator.
5954 20. The saslauthd authentication condition could segfault if neither service
5955 nor realm was specified.
5957 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5958 error message that was logged was misleading, and incoming SMTP
5959 connections were dropped. The message is now more accurate, and temporary
5960 errors are given to SMTP connections.
5962 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5963 (except RSET) with a 421 error, until QUIT is received. However, it was
5964 failing to send a response to QUIT.
5966 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5967 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5968 ACL is now run after the helo_try_verify_hosts code.
5970 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5971 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5972 case-independent, so other case variants are also recognized). Apparently
5973 some systems use these upper case variants.
5975 25. If more than two messages were waiting for the same host, and a transport
5976 filter was specified for the transport, Exim sent two messages over the
5977 same TCP/IP connection, and then failed with "socket operation on non-
5978 socket" when it tried to send the third.
5980 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5982 27. The extern definition of crypt16() in expand.c was not being excluded when
5983 the OS had its own crypt16() function.
5985 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5986 as a preferred synonym for return_size_limit, both as an option and as an
5989 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5991 30. Changed the default configuration ACL to relax the local part checking rule
5992 for addresses that are not in any local domains. For these addresses,
5993 slashes and pipe symbols are allowed within local parts, but the sequence
5994 /../ is explicitly forbidden.
5996 31. SPA server authentication was not clearing the challenge buffer before
5999 32. log_message in a "warn" ACL statement was writing to the reject log as
6000 well as to the main log, which contradicts the documentation and doesn't
6001 seem right (because no rejection is happening). So I have stopped it.
6003 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6004 However, I am unable to do any testing of this.
6006 34. Fixed an infelicity in the appendfile transport. When checking directories
6007 for a mailbox, to see if any needed to be created, it was accidentally
6008 using path names with one or more superfluous leading slashes; tracing
6009 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6011 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6012 contents are added to the log line that is written for every discarded
6013 recipient. (Previously a log_message setting was ignored.)
6015 36. The ${quote: operator now quotes the string if it is empty.
6017 37. The install script runs exim in order to find its version number. If for
6018 some reason other than non-existence or emptiness, which it checks, it
6019 could not run './exim', it was installing it with an empty version number,
6020 i.e. as "exim-". This error state is now caught, and the installation is
6023 38. An argument was missing from the function that creates an error message
6024 when Exim fails to connect to the socket for saslauthd authentication.
6025 This could cause Exim to crash, or give a corrupted message.
6027 39. Added isip, isip4, and isip6 to ${if conditions.
6029 40. The ACL variables $acl_xx are now saved with the message, and can be
6030 accessed later in routers, transports, and filters.
6032 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6033 strings in the file are not string-expanded.
6035 42. If a MAIL command specified a SIZE value that was too large to fit into an
6036 int variable, the check against message_size_limit failed. Such values are
6037 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6038 day this will have to be increased, but I don't think I want to be around
6039 when emails are that large.
6046 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6047 iconv() is not standard in FreeBSD.
6049 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6050 IPv6 enabled. The observed symptom was a segmentation fault on return from
6051 the function os_common_find_running_interfaces() in src/os.c.
6053 3. In the check_special_case() function in daemon.c I had used "errno" as an
6054 argument name, which causes warnings on some systems. This was basically a
6055 typo, since it was named "eno" in the comments!
6057 4. The code that waits for the clock to tick (at a resolution of some fraction
6058 of a second) so as to ensure message-id uniqueness was always waiting for
6059 at least one whole tick, when it could have waited for less. [This is
6060 almost certainly not relevant at current processor speeds, where it is
6061 unlikely to ever wait at all. But we try to future-proof.]
6063 5. The function that sleeps for a time interval that includes fractions of a
6064 second contained a race. It did not block SIGALRM between setting the
6065 timer, and suspending (a couple of lines later). If the interval was short
6066 and the sigsuspend() was delayed until after it had expired, the suspension
6067 never ended. On busy systems this could lead to processes getting stuck for
6070 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6071 process, before it forks any delivery processes. The open lookup caching
6072 mechanism meant that the open file or database connection was passed into
6073 the delivery process. The problem was that delivery processes always tidy
6074 up cached lookup data. This could cause a problem for the next delivery
6075 process started by the queue runner, because the external queue runner
6076 process does not know about the closure. So the next delivery process
6077 still has data in the lookup cache. In the case of a file lookup, there was
6078 no problem because closing a file descriptor in a subprocess doesn't affect
6079 the parent. However, if the lookup was caching a connection to a database,
6080 the connection was closed, and the second delivery process was likely to
6081 see errors such as "PGSQL: query failed: server closed the connection
6082 unexpectedly". The problem has been fixed by closing all cached lookups
6083 in a queue runner before running a delivery process.
6085 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6086 seem to have the "const" qualifier which it has on other OS. I've
6089 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6090 *of the same type* (client or server) with the same public name that an
6091 error should be diagnosed.
6093 9. When Exim looked up a host name for an IP address, but failed to find the
6094 original IP address when looking up the host name (a safety check), it
6095 output the message "<ip address> does not match any IP for NULL", which was
6096 confusing, to say the least. The bug was that the host name should have
6097 appeared instead of "NULL".
6099 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6100 user that is built into the binary, and the -C or -D options is used, root
6101 privilege is dropped before the configuration file is read. In addition,
6102 logging is switched to stderr instead of the normal log files. If the
6103 configuration then re-defines the Exim user, the unprivileged environment
6104 is probably not what is expected, so Exim logs a panic warning message (but
6107 However, if deliver_drop_privilege is set, the unprivileged state may well
6108 be exactly what is intended, so the warning has been cut out in that case,
6109 and Exim is allowed to try to write to its normal log files.
6115 1. smtp_return_error_details was not giving details for temporary sender
6116 or receiver verification errors.
6118 2. Diagnose a configuration error if two authenticators have the same public
6121 3. Exim used not to create the message log file for a message until the first
6122 delivery attempt. This could be confusing when incoming messages were held
6123 for policy or load reasons. The message log file is now created at the time
6124 the message is received, and an initial "Received" line is written to it.
6126 4. The automatically generated man page for command line options had a minor
6127 bug that caused no ill effects; however, a more serious problem was that
6128 the procedure for building the man page automatically didn't always
6129 operate. Consequently, release 4.20 contains an out-of-date version. This
6130 shouldn't happen again.
6132 5. When building Exim with embedded Perl support, the script that builds the
6133 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6134 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6136 6. The freeze_tell option was not being used for messages that were frozen on
6137 arrival, either by an ACL or by local_scan().
6139 7. Added the smtp_incomplete_transaction log selector.
6141 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6142 was accepting AUTH without a new EHLO.
6144 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6145 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6148 10. Logging of TCP/IP connections (when configured) now happens in the main
6149 daemon process instead of the child process, so that the TCP/IP connection
6150 count is more accurate (but it can never be perfect).
6152 11. The use of "drop" in a nested ACL was not being handled correctly in the
6153 outer ACL. Now, if condition failure induced by the nested "drop" causes
6154 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6155 or "require"), the connection is dropped.
6157 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6158 that yield "discard" can now be used with an "accept" or a "discard" verb,
6159 but an error is generated for any others (because I can't see a useful way
6160 to define what should happen).
6162 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6163 are now processed in the same way as lines in the Exim configuration file.
6164 In particular, continuation lines are supported.
6166 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6168 15. Added -ti meaning -t -i.
6170 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6171 domains, and warn by logging if others are found.
6173 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6174 returned from GIFCONF when Exim is trying to find the list of interfaces on
6175 a host. The code in os.c has been modified to copy each ifreq to an aligned
6176 structure in all cases.
6178 Also, in some cases, the returned ifreq's were being copied to a 'struct
6179 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6180 means the last couple of bytes of an IPv6 address could be chopped if the
6181 ifreq contained only a normal sockaddr (14 bytes storage).
6183 18. Named domain lists were not supported in the hosts_treat_as_local option.
6184 An entry such as +xxxx was not recognized, and was treated as a literal
6187 19. Ensure that header lines added by a DATA ACL are included in the reject log
6188 if the ACL subsequently rejects the message.
6190 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6191 MD5 (which is deprecated).
6193 21. When testing a filter file using -bf, Exim was writing a message when it
6194 took the sender from a "From " line in the message, but it was not doing so
6195 when it took $return_path from a Return-Path: header line. It now does.
6197 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6198 with a valid header line field name (a series of printing characters
6199 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6201 23. Changed "disc" in the source to "disk" to conform to the documentation and
6202 the book and for uniformity.
6204 24. Ignore Sendmail's -Ooption=value command line item.
6206 25. When execve() failed while trying to run a command in a pipe transport,
6207 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6208 could be confused with a return value of 69 from the command itself. This
6209 has been changed to 127, the value the shell returns if it is asked to run
6210 a non-existent command. The wording for the related log line suggests a
6211 non-existent command as the problem.
6213 26. If received_header_text expands to an empty string, do not add a Received:
6214 header line to the message. (Well, it adds a token one on the spool, but
6215 marks it "old" so that it doesn't get used or transmitted.)
6217 27. Installed eximstats 1.28 (addition of -nt option).
6219 28. There was no check for failure on the call to getsockname() in the daemon
6220 code. This can fail if there is a shortage of resources on the system, with
6221 ENOMEM, for example. A temporary error is now given on failure.
6223 29. Contrary to the C standard, it seems that in some environments, the
6224 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6225 program. Exim now does this explicitly; it affects the formatting of
6226 timestamps using strftime().
6228 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6229 complaints. I've now initialized all the variables, to avoid this.
6231 32. Header lines added by a system filter were not being "seen" during
6232 transport-time rewrites.
6234 33. The info_callback() function passed to OpenSSL is set up with type void
6235 (*)(SSL *, int, int), as described somewhere. However, when calling the
6236 function (actually a macro) that sets it up, the type void(*)() is
6237 expected. I've put in a cast to prevent warnings from picky compilers.
6239 34. If a DNS black list lookup found a CNAME record, but there were no A
6240 records associated with the domain it pointed at, Exim crashed.
6242 35. If a DNS black list lookup returned more than one A record, Exim ignored
6243 all but the first. It now scans all returned addresses if a particular IP
6244 value is being sought. In this situation, the contents of the
6245 $dnslist_value variable are a list of all the addresses, separated by a
6248 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6249 to accept a host name and all its aliases if the forward lookup for any of
6250 them yielded the IP address of the incoming connection. Now it accepts only
6251 those names whose forward lookup yields the correct IP address. Any other
6252 names are discarded. This closes a loophole whereby a rogue DNS
6253 administrator could create reverse DNS records to break through a
6254 wildcarded host restriction in an ACL.
6256 37. If a user filter or a system filter that ran in a subprocess used any of
6257 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6258 the wrong values were passed to the pipe command ($thisaddress had the
6259 value of $0, $0 had the value of $1, etc). This bug was introduced by
6260 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6262 38. Improved the line breaking for long SMTP error messages from ACLs.
6263 Previously, if there was no break point between 40 and 75 characters, Exim
6264 left the rest of the message alone. Two changes have been made: (a) I've
6265 reduced the minimum length to 35 characters; (b) if it can't find a break
6266 point between 35 and 75 characters, it looks ahead and uses the first one
6267 that it finds. This may give the occasional overlong line, but at least the
6268 remaining text gets split now.
6270 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6271 file descriptors might be low, and that setting 1000 would always raise it.
6272 It turns out that in some environments, the limit is already over 1000 and
6273 that lowering it causes trouble. So now Exim takes care not to decrease it.
6275 40. When delivering a message, the value of $return_path is set to $sender_
6276 address at the start of routing (routers may change the value). By an
6277 oversight, this default was not being set up when an address was tested by
6278 -bt or -bv, which affected the outcome if any router or filter referred to
6281 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6282 log only when "message" or "log_message" are set. However, if one of the
6283 conditions was an address verification, or a call to a nested ACL, the
6284 messages generated by the underlying test were being passed through. This
6285 no longer happens. The underlying message is available in $acl_verify_
6286 message for both "message" and "log_message" expansions, so it can be
6287 passed through if needed.
6289 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6290 new expansion $bh_ to give the encoded byte string without charset
6291 translation. Translation happens only if iconv() is available; HAVE_ICONV
6292 indicates this at build time. HEADERS_CHARSET gives the charset to
6293 translate to; headers_charset can change it in the configuration, and
6294 "headers charset" can change it in an individual filter file.
6296 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6297 that creates RFC 2047 encoded "words" labels them as that charset instead
6298 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6299 expansion operator; (ii) when Exim creates a From: line for a local
6300 message; (iii) when a header line is rewritten to include a "phrase" part.
6302 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6303 buggy, causing it to skip the first lines of messages whose message ID
6304 ended in 'D'. This would not have bitten before Exim release 4.14, because
6305 message IDs were unlikely to end in 'D' before then. The effect was to have
6306 incorrect size information for certain domains.
6308 45. #include "config.h" was missing at the start of the crypt16.c module. This
6309 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6312 46. If there was a timeout during a "random" callout check, Exim treated it as
6313 a failure of the random address, and carried on sending RSET and the real
6314 address. If the delay was just some slowness somewhere, the response to the
6315 original RCPT would be taken as a response to RSET and so on, causing
6316 mayhem of various kinds.
6318 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6319 when I implemented it. It didn't allow for the fact that some option values
6320 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6321 the right test for positive values.
6323 48. Domain names in DNS records are case-independent. Exim always looks them up
6324 in lower case. Some resolvers return domain names in exactly the case they
6325 appear in the zone file, that is, they may contain uppercase letters. Not
6326 all resolvers do this - some return always lower case. Exim was treating a
6327 change of case by a resolver as a change of domain, similar to a widening
6328 of a domain abbreviation. This triggered its re-routing code and so it was
6329 trying to route what was effectively the same domain again. This normally
6330 caused routing to fail (because the router wouldn't handle the domain
6331 twice). Now Exim checks for this case specially, and just changes the
6332 casing of the domain that it ultimately uses when it transmits the message
6335 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6338 50. If a filter generated a file delivery with a non-absolute name (possible if
6339 no home directory exists for the router), the forbid_file option was not
6342 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6343 the existing equality matching.
6345 52. Exim was using ints instead of ino_t variables in some places where it was
6346 dealing with inode numbers.
6348 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6349 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6350 called TMPDIR, and if it finds it is different, it changes its value.
6352 54. The smtp_printf() function is now made available to local_scan() so
6353 additional output lines can be written before returning. There is also an
6354 smtp_fflush() function to enable the detection of a dropped connection.
6355 The variables smtp_input and smtp_batched_input are exported to
6358 55. Changed the default runtime configuration: the message "Unknown user"
6359 has been removed from the ACL, and instead placed on the localuser router,
6360 using the cannot_route_message feature. This means that any verification
6361 failures that generate their own messages won't get overridden. Similarly,
6362 the "Unrouteable address" message that was in the ACL for unverifiable
6363 relay addresses has also been removed.
6365 56. Added hosts_avoid_esmtp to the smtp transport.
6367 57. The exicyclog script was not checking for the esoteric option
6368 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6369 will work only if exicyclog is run under the appropriate euid.
6371 58. Following a discussion on the list, the rules by which Exim recognises line
6372 endings on incoming messages have been changed. The -dropcr and drop_cr
6373 options are now no-ops, retained only for backwards compatibility. The
6374 following line terminators are recognized: LF CRLF CR. However, special
6375 processing applies to CR:
6377 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6378 nor a local message in the state where . is a terminator.
6380 (ii) If a bare CR is encountered in a header line, an extra space is added
6381 after the line terminator so as not to end the header. The reasoning
6382 behind this is that bare CRs in header lines are most likely either
6383 to be mistakes, or people trying to play silly games.
6385 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6386 was being incorrectly given as 18 bytes larger than it should have been.
6387 This is a VOB (very old bug).
6389 60. This may never have affected anything current, but just in case it has:
6390 When the local host is found other than at the start of a list of hosts,
6391 the local host, those with the same MX, and any that follow, are discarded.
6392 When the list in question was part of a longer list of hosts, the following
6393 hosts (not currently being processed) were also being discarded. This no
6394 longer happens. I'm not sure if this situation could ever has previously
6397 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6400 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6401 header. This is something that is recommended in a new Internet Draft, and
6402 is something that is documented as being done by Sendmail. There are two
6403 possible values. For messages generated by the autoreply transport, Exim
6406 Auto-Submitted: auto-replied
6408 whereas for all other generated messages (e.g. bounces) it adds
6410 Auto-Submitted: auto-generated
6412 63. The "personal" condition in filters now includes a test for the
6413 Auto-Submitted: header. If it contains the string "auto-" the message it
6414 not considered personal.
6416 64. Added rcpt_include_affixes as a generic transport option.
6418 65. Added queue_only_override (default true).
6420 66. Added the syslog_duplication option.
6422 67. If what should have been the first header line of a message consisted of
6423 a space followed by a colon, Exim was mis-interpreting it as a header line.
6424 It isn't of course - it is syntactically invalid and should therefore be
6425 treated as the start of the message body. The misbehaviour could have
6426 caused a number of strange effects, including loss of data in subsequent
6427 header lines, and spool format errors.
6429 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6430 client host had authenticated. This control can now be exercised by an ACL
6431 for more flexibility.
6433 69. By default, callouts do not happen when testing with -bh. There is now a
6434 variant, -bhc, which does actually run the callout code, including
6435 consulting and updating the callout cache.
6437 70. Added support for saslauthd authentication, courtesy of Alexander
6440 71. If statvfs() failed on the spool or log directories while checking their
6441 size for availability, Exim confusingly gave the error "space shortage".
6442 Furthermore, in debugging mode it crashed with a floating point exception.
6443 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6444 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6445 serious problem, Exim now writes to the main and panic logs when this
6446 happens, with details of the failure. It then refuses to accept the
6447 incoming message, giving the message "spool directory problem" or "log
6448 directory problem" with a 421 code for SMTP messages.
6450 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6451 0, 1, and 2 exist, because some OS complain for execs without them (see
6452 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6453 descriptors. However, the code omitted to check that the open succeeded,
6454 causing mysterious errors if for some reason the permissions on /dev/null
6455 got screwed. Now Exim writes a message to the main and panic logs, and
6456 bombs out if it can't open /dev/null.
6458 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6459 interact so that it is all more flexible. It is supposed to remain
6460 backwards compatible. Also added extra_local_interfaces.
6462 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6463 to bomb out with an assertion failure - to the client this appears as a
6464 connection drop. This problem occurs in the part of the code that was taken
6465 from the Samba project. Fortunately, the assertion is in a very simple
6466 function, so I have fixed this by reproducing the function inline in the
6467 one place where it is called, and arranging for authentication to fail
6468 instead of killing the process with assert().
6470 75. The SPA client code was not working when the server requested OEM rather
6471 than Unicode encoding.
6473 76. Added code to make require_files with a specific uid setting more usable in
6474 the case where statting the file as root fails - usually a non-root-mounted
6475 NFS file system. When this happens and the failure is EACCES, Exim now
6476 forks a subprocess and does the per-uid checking as the relevant uid.
6478 77. Added process_log_path.
6480 78. If log_file_path was not explicitly set, a setting of check_log_space or
6481 check_log_inodes was ignored.
6483 79. If a space check for the spool or log partitions fails, the incident is now
6484 logged. Of course, in the latter case the data may get lost...
6486 80. Added the %p formatting code to string_format() so that it can be used to
6487 print addresses in debug_print(). Adjusted all the address printing in the
6488 debugging in store.c to use %p rather than %d.
6490 81. There was a concern that a line of code in smtp_in.c could overflow a
6491 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6492 initially expressed, the concern was not well-founded, because trailing
6493 spaces are removed early. However, if the trailing spaces were followed by
6494 a NULL, they did not get removed, so the overflow was possible. Two fixes
6497 (a) I re-wrote the offending code in a cleaner fashion.
6498 (b) If an incoming SMTP command contains a NULL character, it is rejected
6501 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6502 runs initgroups(), so that if the Exim user is in any additional groups,
6503 they will be used during message reception.
6509 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob