1 Change log file for Exim from version 4.21
2 -------------------------------------------
11 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
12 and 3.4.4 - once the server is enabled to respond to an OCSP request
13 it does even when not requested, resulting in a stapling non-aware
14 client dropping the TLS connection.
16 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
17 support variable-length bit vectors. No functional change.
19 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
20 The I= interface field on outgoing lines is now after the H= remote
21 host field, same as incoming lines. There is a separate
22 outgoing_interface log selector which allows you to disable the
25 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
26 If not running log_selector +smtp_connection the mainlog would be held
27 open indefinitely after a "too many connections" event, including to a
28 deleted file after a log rotate. Leave the per net connection logging
29 leaving it open for efficiency as that will be quickly detected by the
30 check on the next write.
32 HS/01 Bug 1671: Fix post transport crash.
33 Processing the wait-<transport> messages could crash the delivery
34 process if the message IDs didn't exist for some reason. When
35 using 'split_spool_directory=yes' the construction of the spool
36 file name failed already, exposing the same netto behaviour.
38 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
39 mime_regex ACL conditions.
41 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
42 to DSN fail messages (bounces): remote IP, remote greeting, remote response
43 to HELO, local diagnostic string.
45 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
46 log line to debug. Even when configured with a tls authenticator many
47 client connections are expected to not authenticate in this way, so
48 an authenticate fail is not an error.
50 HS/02 Add the Exim version string to the process info. This way exiwhat
51 gives some more detail about the running daemon.
53 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
54 matter for fast-change records such as DNSBLs.
56 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
57 retry record, even if constant. There may be multiple transports with
58 different interface settings and the retry behaviour needs to be kept
61 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
63 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
65 JH/10 Bug 840: fix log_defer_output option of pipe transport
67 JH/11 Bug 830: use same host for all RCPTS of a message, even under
68 hosts_randomize. This matters a lot when combined with mua_wrapper.
70 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
71 ${quote_pgsql:<string>} operator.
73 JH/13 Bug 1708: avoid misaligned access in cached lookup.
75 JH/14 Change header file name for freeradius-client. Relevant if compiling
76 with Radius support; from the Gentoo tree and checked under Fedora.
78 JH/15 Bug 1712: Introduce $prdr_requested flag variable
80 JH/16 Bug 1714: Permit an empty string as expansion result for transport
81 option transport_filter, meaning no filtering.
83 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
85 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
86 defaults to "*" (all hosts). The variable is now available when not built
87 with TLS, default unset, mainly to enable keeping the testuite sane.
88 If a server certificate is not supplied (via tls_certificate) an error is
89 logged, and clients will find TLS connections fail on startup. Presumably
90 they will retry in-clear.
91 Packagers of Exim are strongly encouraged to create a server certificate
94 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
95 with the $config_file variable.
97 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
98 in transport context, after the attempt, and per-recipient. The latter type
99 is per host attempted. The event data is the error message, and the errno
100 information encodes the lookup type (A vs. MX) used for the (first) host,
101 and the trailing two digits of the smtp 4xx reponse.
103 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
104 to write to mainlog (or rejectlog, paniclog) in the window between file
105 creation and permissions/ownership being changed. Particularly affects
106 installations where exicyclog is run as root, rather than exim user;
107 result is that the running daemon panics and dies.
109 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
111 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
112 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
113 "pri" and "weight". Note that the previous implicit priority given by the
114 list order is no longer honoured.
116 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
119 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
120 by defining SUPPORT_SOCKS.
122 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
123 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
124 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
125 variables are renamed to proxy_{local,external}_{address,port}.
127 JH/27 Move Internationalisation support from Experimental to mainline, enabled
128 for a build by defining SUPPORT_I18N
130 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
131 of the query string, and make ${quote_redis:} do that quoting.
133 JH/29 Move Events support from Experimental to mainline, enabled by default
134 and removable for a build by defining DISABLE_EVENT.
136 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
138 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
139 cached by the daemon.
141 JH/32 Move Redis support from Experimental to mainline, enabled for a build
142 by defining LOOKUP_REDIS. The libhiredis library is required.
144 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
145 keys are given for lookup.
147 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
148 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
149 only supported when built with TLS support. The PolarSSL SHA routines
150 are still used when the TLS library is too old for convenient support.
152 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
153 openssl_options), for security. OpenSSL forces this from version 1.1.0
154 server-side so match that on older versions.
156 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
157 allocation for $value could be released as the expansion processing
158 concluded, but leaving the global pointer active for it.
160 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
161 and to use the domains and local_parts ACL conditions.
163 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
164 incorrectly not doubled on cutthrough transmission, hence seen as a
165 body-termination at the receiving system - resulting in truncated mails.
166 Commonly the sender saw a TCP-level error, and retransmitted the message
167 via the normal store-and-forward channel. This could result in duplicates
168 received - but deduplicating mailstores were liable to retain only the
169 initial truncated version.
171 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
173 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
175 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
176 we're in there, support oversigning also; bug 1309.
178 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
180 HS/04 Add support for keep_environment and add_environment options.
182 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
183 either intentional arithmetic overflow during PRNG, or testing config-
186 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
187 delivery resulted in actual delivery. Cancel cutthrough before DATA
190 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
191 rejects a recipient: pass the reject to the originator.
193 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
194 Many were false-positives and ignorable, but it's worth fixing the
197 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
198 for the new environment-manipulation done at startup. Move the routines
199 from being local to tls.c to being global via the os.c file.
201 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
202 an extract embedded as result-arg for a map, the first arg for extract
203 is unavailable so we cannot tell if this is a numbered or keyed
204 extraction. Accept either.
210 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
213 JH/02 The smtp transport option "multi_domain" is now expanded.
215 JH/03 The smtp transport now requests PRDR by default, if the server offers
218 JH/04 Certificate name checking on server certificates, when exim is a client,
219 is now done by default. The transport option tls_verify_cert_hostnames
220 can be used to disable this per-host. The build option
221 EXPERIMENTAL_CERTNAMES is withdrawn.
223 JH/05 The value of the tls_verify_certificates smtp transport and main options
224 default to the word "system" to access the system default CA bundle.
225 For GnuTLS, only version 3.0.20 or later.
227 JH/06 Verification of the server certificate for a TLS connection is now tried
228 (but not required) by default. The verification status is now logged by
229 default, for both outbound TLS and client-certificate supplying inbound
232 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
235 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
236 Status Notification (bounce) messages are now MIME format per RFC 3464.
237 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
238 under the control of the dsn_advertise_hosts option, and routers may
239 have a dsn_lasthop option.
241 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
242 default, modifiable by a malware= option. The list separator for
243 the options can now be changed in the usual way. Bug 68.
245 JH/10 The smtp_receive_timeout main option is now expanded before use.
247 JH/11 The incoming_interface log option now also enables logging of the
248 local interface on delivery outgoing connections.
250 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
251 if the interface and destination host and port all match.
253 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
256 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
257 Patch from Andrew Lewis.
259 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
260 now supports optional time-restrictions, weighting, and priority
261 modifiers per server. Patch originally by <rommer@active.by>.
263 JH/16 The spamd_address main option now supports a mixed list of local
264 and remote servers. Remote servers can be IPv6 addresses, and
265 specify a port-range.
267 JH/17 Bug 68: The spamd_address main option now supports an optional
268 timeout value per server.
270 JH/18 Bug 1581: Router and transport options headers_add/remove can
271 now have the list separator specified.
273 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
276 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
279 JH/21 Support for the A6 type of dns record is withdrawn.
281 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
282 rather than the verbs used.
284 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
285 from 255 to 1024 chars.
287 JH/24 Verification callouts now attempt to use TLS by default.
289 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
290 are generic router options now. The defaults didn't change.
292 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
293 Original patch from Alexander Shikoff, worked over by JH.
295 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
296 files if ZCAT_COMMAND is not executable.
298 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
300 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
302 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
303 Normally benign, it bites when the pair was led to by a CNAME;
304 modern usage is to not canoicalize the domain to a CNAME target
305 (and we were inconsistent anyway for A-only vs AAAA+A).
307 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
309 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
310 when evaluating $sender_host_dnssec.
312 JH/31 Check the HELO verification lookup for DNSSEC, adding new
313 $sender_helo_dnssec variable.
315 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
317 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
319 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
321 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
322 documented as working, but never had. Support all but $spam_report.
324 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
325 added for tls authenticator.
330 TL/01 When running the test suite, the README says that variables such as
331 no_msglog_check are global and can be placed anywhere in a specific
332 test's script, however it was observed that placement needed to be near
333 the beginning for it to behave that way. Changed the runtest perl
334 script to read through the entire script once to detect and set these
335 variables, reset to the beginning of the script, and then run through
336 the script parsing/test process like normal.
338 TL/02 The BSD's have an arc4random API. One of the functions to induce
339 adding randomness was arc4random_stir(), but it has been removed in
340 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
341 function when detected.
343 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
344 cause callback expansion.
346 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
347 syntax errors in an expansion can be treated as a string instead of
348 logging or causing an error, due to the internal use of bool_lax
349 instead of bool when processing it.
351 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
352 server certificates when making smtp deliveries.
354 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
356 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
358 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
360 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
361 Merged patch from Sebastian Wiedenroth.
363 JH/05 Fix results-pipe from transport process. Several recipients, combined
364 with certificate use, exposed issues where response data items split
365 over buffer boundaries were not parsed properly. This eventually
366 resulted in duplicates being sent. This issue only became common enough
367 to notice due to the introduction of conection certificate information,
368 the item size being so much larger. Found and fixed by Wolfgang Breyha.
370 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
371 size buffer was used, resulting in syntax errors when an expansion
374 JH/07 Add support for directories of certificates when compiled with a GnuTLS
375 version 3.3.6 or later.
377 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
378 is EXPERIMENTAL_EVENT, the main-configuration and transport options
379 both become "event_action", the variables become $event_name, $event_data
380 and $event_defer_errno. There is a new variable $verify_mode, usable in
381 routers, transports and related events. The tls:cert event is now also
382 raised for inbound connections, if the main configuration event_action
385 TL/06 In test suite, disable OCSP for old versions of openssl which contained
386 early OCSP support, but no stapling (appears to be less than 1.0.0).
388 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
389 server certificate names available under the smtp transport option
390 "tls_verify_cert_hostname" now do not permit multi-component wildcard
393 JH/10 Time-related extraction expansions from certificates now use the main
394 option "timezone" setting for output formatting, and are consistent
395 between OpenSSL and GnuTLS compilations. Bug 1541.
397 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
398 encoded parameter in the incoming message. Bug 1558.
400 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
401 include certificate info, eximon was claiming there were spoolfile
404 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
406 JH/14 Log delivery-related information more consistently, using the sequence
407 "H=<name> [<ip>]" wherever possible.
409 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
410 are problematic for Debian distribution, omit them from the release
413 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
415 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
418 JH/17 Fix a null-indirection in certextract expansions when a nondefault
419 output list separator was used.
424 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
425 checkers that were complaining about end of non-void function with no
428 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
429 This was a regression intruduced in 4.83 by another bugfix.
431 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
433 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
434 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
440 TF/01 Correctly close the server side of TLS when forking for delivery.
442 When a message was received over SMTP with TLS, Exim failed to clear up
443 the incoming connection properly after forking off the child process to
444 deliver the message. In some situations the subsequent outgoing
445 delivery connection happened to have the same fd number as the incoming
446 connection previously had. Exim would try to use TLS and fail, logging
447 a "Bad file descriptor" error.
449 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
450 utilities have not been installed.
452 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
453 temporary space as the ACL may create new global variables.
455 TL/01 LDAP support uses per connection or global context settings, depending
456 upon the detected version of the libraries at build time.
458 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
459 to extract and use the src ip:port in logging and expansions as if it
460 were a direct connection from the outside internet. PPv2 support was
461 updated based on HAProxy spec change in May 2014.
463 JH/02 Add ${listextract {number}{list}{success}{fail}}.
465 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
466 Properly escape header and check for NULL return.
468 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
471 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
473 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
474 characters in header names, implemented as a verify condition.
475 Contributed by Michael Fischer v. Mollard.
477 TL/05 Rename SPF condition results err_perm and err_temp to standardized
478 results permerror and temperror. Previous values are deprecated but
479 still accepted. In a future release, err_perm and err_temp will be
480 completely removed, which will be a backward incompatibility if the
481 ACL tests for either of these two old results. Patch contributed by
482 user bes-internal on the mailing list.
484 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
486 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
487 selectors, in both main and reject logs.
489 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
492 JH/07 Add malware type "sock" for talking to simple daemon.
494 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
496 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
497 routers/transports under cutthrough routing.
499 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
500 numbers. Touch up "bool" conditional to keep the same definition.
502 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
504 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
506 JH/12 Expand items in router/transport headers_add or headers_remove lists
507 individually rather than the list as a whole. Bug 1452.
509 Required for reasonable handling of multiple headers_ options when
510 they may be empty; requires that headers_remove items with embedded
511 colons must have them doubled (or the list-separator changed).
513 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
514 view the policy declared in the DMARC record. Currently, $dmarc_status
515 is a combined value of both the record presence and the result of the
518 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
520 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
521 dnslookup router and the smtp transport (applying to the forward
524 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
525 of ldap servers used for a specific lookup. Patch provided by Heiko
528 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
529 New variable $lookup_dnssec_authenticated for observability.
531 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
532 Patch submitted by Lars Timman.
534 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
536 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
537 Requires trusted mode and valid format message id, aborts otherwise.
538 Patch contributed by Heiko Schlichting.
540 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
541 certextract with support for various fields. Bug 1358.
543 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
544 is requested by default, modifiable by smtp transport option
547 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
548 operate on certificate variables to give certificate fingerprints
549 Also new ${sha256:cert_variable}.
551 JH/23 The PRDR feature is moved from being Experimental into the mainline.
553 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
554 Christian Aistleitner.
556 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
558 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
559 file. Patch from Wolfgang Breyha.
561 JH/25 Expand the coverage of the delivery $host and $host_address to
562 client authenticators run in verify callout. Bug 1476.
564 JH/26 Port service names are now accepted for tls_on_connect_ports, to
565 align with daemon_smtp_ports. Bug 72.
567 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
568 support and error reporting did not work properly.
570 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
571 and is readable. Patch from Andrew Colin Kissa.
573 TL/14 Enhance documentation of ${run expansion and how it parses the
574 commandline after expansion, particularly in the case when an
575 unquoted variable expansion results in an empty value.
577 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
579 PP/02 Fix internal collision of T_APL on systems which support RFC3123
580 by renaming away from it. Addresses GH issue 15, reported by
583 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
585 TL/15 SECURITY: prevent double expansion in math comparison functions
586 (can expand unsanitized data). Not remotely exploitable.
593 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
595 PP/02 Make -n do something, by making it not do something.
596 When combined with -bP, the name of an option is not output.
598 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
601 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
602 $sender_host_name and config options to manage this, and basic check
605 PP/05 DSCP support for outbound connections and control modifier for inbound.
607 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
608 (Only plugin which currently uses this is kerberos4, which nobody should
609 be using, but we should make it available and other future plugins might
610 conceivably use it, even though it would break NAT; stuff *should* be
611 using channel bindings instead).
613 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
614 name; added for Sendmail compatibility; requires admin caller.
615 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
616 just ignore it); requires trusted caller.
617 Also parse but ignore: -Ac -Am -X<logfile>
620 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
622 TL/02 Add +smtp_confirmation as a default logging option.
624 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
625 Patch by Magnus Holmgren from 2007-02-20.
627 TL/04 Bugzilla 1281 - Spec typo.
628 Bugzilla 1283 - Spec typo.
629 Bugzilla 1290 - Spec grammar fixes.
631 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
633 TL/06 Add Experimental DMARC support using libopendmarc libraries.
635 TL/07 Fix an out of order global option causing a segfault. Reported to dev
636 mailing list by by Dmitry Isaikin.
638 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
640 JH/02 Support "G" suffix to numbers in ${if comparisons.
642 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
644 NM/01 Bugzilla 1197 - Spec typo
645 Bugzilla 1196 - Spec examples corrections
647 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
649 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
650 gnutls_enable_pkcs11, but renamed to more accurately indicate its
653 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
654 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
656 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
657 "acl {{name}{arg}...}", and optional args on acl condition
660 JH/05 Permit multiple router/transport headers_add/remove lines.
662 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
664 JH/07 Avoid using a waiting database for a single-message-only transport.
665 Performance patch from Paul Fisher. Bugzilla 1262.
667 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
670 JH/09 Add $headers_added variable, with content from use of ACL modifier
671 add_header (but not yet added to the message). Bugzilla 199.
673 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
674 Pulled from Bugzilla 817 by Wolfgang Breyha.
676 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
678 (nb: this is the same fix as in Exim 4.80.1)
680 JH/11 Add A= logging on delivery lines, and a client_set_id option on
683 JH/12 Add optional authenticated_sender logging to A= and a log_selector
686 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
688 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
689 advertise SMTP AUTH mechanism to us, instead of a generic
690 protocol violation error. Also, make Exim more robust to bad
691 data from the Dovecot auth socket.
693 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
695 When a queue runner is handling a message, Exim first routes the
696 recipient addresses, during which it prunes them based on the retry
697 hints database. After that it attempts to deliver the message to
698 any remaining recipients. It then updates the hints database using
701 So if a recipient address works intermittently, it can get repeatedly
702 deferred at routing time. The retry hints record remains fresh so the
703 address never reaches the final cutoff time.
705 This is a fairly common occurrence when a user is bumping up against
706 their storage quota. Exim had some logic in its local delivery code
707 to deal with this. However it did not apply to per-recipient defers
708 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
710 This change adds a proper retry rule check during routing so that the
711 final cutoff time is checked against the message's age. We only do
712 this check if there is an address retry record and there is not a
713 domain retry record; this implies that previous attempts to handle
714 the address had the retry_use_local_parts option turned on. We use
715 this as an approximation for the destination being like a local
716 delivery, as in LMTP.
718 I suspect this new check makes the old local delivery cutoff check
719 redundant, but I have not verified this so I left the code in place.
721 TF/02 Correct gecos expansion when From: is a prefix of the username.
723 Test 0254 submits a message to Exim with the header
727 When I ran the test suite under the user fanf2, Exim expanded
728 the header to contain my full name, whereas it should have added
729 a Resent-Sender: header. It erroneously treats any prefix of the
730 username as equal to the username.
732 This change corrects that bug.
734 GF/01 DCC debug and logging tidyup
735 Error conditions log to paniclog rather than rejectlog.
736 Debug lines prefixed by "DCC: " to remove any ambiguity.
738 TF/03 Avoid unnecessary rebuilds of lookup-related code.
740 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
741 Bug spotted by Jeremy Harris; was flawed since initial commit.
742 Would have resulted in OCSP responses post-SNI triggering an Exim
743 NULL dereference and crash.
745 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
747 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
748 Bug detection, analysis and fix by Samuel Thibault.
749 Bugzilla 1331, Debian bug #698092.
751 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
753 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
754 Server implementation by Todd Lyons, client by JH.
755 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
756 config variable "prdr_enable" controls whether the server
757 advertises the facility. If the client requests PRDR a new
758 acl_data_smtp_prdr ACL is called once for each recipient, after
759 the body content is received and before the acl_smtp_data ACL.
760 The client is controlled by bolth of: a hosts_try_prdr option
761 on the smtp transport, and the server advertisement.
762 Default client logging of deliveries and rejections involving
763 PRDR are flagged with the string "PRDR".
765 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
766 fclose(). Diagnosis by Todd Lyons.
768 PP/17 Update configure.default to handle IPv6 localhost better.
769 Patch by Alain Williams (plus minor tweaks).
772 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
773 This is now consistent with GnuTLS, and is now documented: the
774 previous undocumented portable approach to treating the option as
775 unset was to force an expansion failure. That still works, and
776 an empty string is now equivalent.
778 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
779 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
780 not performing validation itself.
782 PP/20 Added force_command boolean option to pipe transport.
783 Patch from Nick Koston, of cPanel Inc.
785 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
788 TF/04 Added udpsend ACL modifer and hexquote expansion operator
790 PP/21 Fix eximon continuous updating with timestamped log-files.
791 Broken in a format-string cleanup in 4.80, missed when I repaired the
792 other false fix of the same issue.
793 Report and fix from Heiko Schlichting.
796 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
797 Report from Prashanth Katuri.
799 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
800 It's SecureTransport, so affects any MacOS clients which use the
801 system-integrated TLS libraries, including email clients.
803 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
804 using a MIME ACL for non-SMTP local injection.
805 Report and assistance in diagnosis by Warren Baker.
807 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
809 JH/16 Fix comparisons for 64b. Bugzilla 1385.
811 TL/09 Add expansion variable $authenticated_fail_id to keep track of
812 last id that failed so it may be referenced in subsequent ACL's.
814 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
817 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
818 ldap library initialization, allowing self-signed CA's to be
819 used. Also properly sets require_cert option later in code by
820 using NULL (global ldap config) instead of ldap handle (per
821 session). Bug diagnosis and testing by alxgomz.
823 TL/12 Enhanced documentation in the ratelimit.pl script provided in
824 the src/util/ subdirectory.
826 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
827 renamed to Transport Post Delivery Action by Jeremy Harris, as
830 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
831 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
832 redis_servers = needs to be configured which will be used by the redis
833 lookup. Patch from Warren Baker, of The Packet Hub.
835 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
837 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
838 hostname or reverse DNS when processing a host list. Used suggestions
839 from multiple comments on this bug.
841 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
843 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
844 Missed a few lines, added it to make the runtest require no keyboard
847 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
848 contains upper case chars. Make router use caseful_local_part.
850 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
851 support when GnuTLS has been built with p11-kit.
857 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
859 This, or similar/improved, will also be change PP/11 of 4.82.
865 PP/01 Handle short writes when writing local log-files.
866 In practice, only affects FreeBSD (8 onwards).
867 Bugzilla 1053, with thanks to Dmitry Isaikin.
869 NM/01 Bugzilla 949 - Documentation tweak
871 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
874 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
876 PP/02 Implemented gsasl authenticator.
878 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
880 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
881 `pkg-config foo` for cflags/libs.
883 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
884 with rest of GSASL and with heimdal_gssapi.
886 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
887 `pkg-config foo` for cflags/libs for the TLS implementation.
889 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
890 properties get this fed in as external SSF. A number of robustness
891 and debugging improvements to the cyrus_sasl authenticator.
893 PP/08 cyrus_sasl server now expands the server_realm option.
895 PP/09 Bugzilla 1214 - Log authentication information in reject log.
896 Patch by Jeremy Harris.
898 PP/10 Added dbmjz lookup type.
900 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
902 PP/12 MAIL args handles TAB as well as SP, for better interop with
903 non-compliant senders.
904 Analysis and variant patch by Todd Lyons.
906 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
907 Bug report from Lars Müller <lars@samba.org> (via SUSE),
908 Patch from Dirk Mueller <dmueller@suse.com>
910 PP/13 tls_peerdn now print-escaped for spool files.
911 Observed some $tls_peerdn in wild which contained \n, which resulted
912 in spool file corruption.
914 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
915 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
916 or write after TLS renegotiation, which otherwise led to messages
919 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
920 as a tracking header (ie: a signed header comes before the signature).
921 Patch from Wolfgang Breyha.
923 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
924 comma-sep list; embedded commas doubled.
926 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
928 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
930 Report and patch from Dmitry Banschikov.
932 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
933 Removed SSL_clear() after SSL_new() which led to protocol negotiation
934 failures. We appear to now support TLS1.1+ with Exim.
936 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
937 lets Exim select keys and certificates based upon TLS SNI from client.
938 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
939 before an outbound SMTP session. New log_selector, +tls_sni.
941 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
942 NULL dereference. Report and patch from Alun Jones.
944 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
945 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
946 Not seeing resolver debug output on NetBSD, but suspect this is a
947 resolver implementation change.
949 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
950 Left warnings. Added "eximon gdb" invocation mode.
952 PP/21 Defaulting "accept_8bitmime" to true, not false.
954 PP/22 Added -bw for inetd wait mode support.
956 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
957 locate the relevant includes and libraries. Made this the default.
959 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
960 Bugzilla 1246, report and most of solution from Tomasz Kusy.
962 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
963 This may cause build issues on older platforms.
965 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
966 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
967 gnutls_require_mac & gnutls_require_protocols (no longer supported).
968 Added SNI support via GnuTLS too.
969 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
971 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
973 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
974 Applied second patch from Janne, implementing suggestion to default
975 multiple-strings-in-record handling to match SPF spec.
977 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
979 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
980 read-only, out of scope).
981 Patch from Wolfgang Breyha, report from Stuart Northfield.
983 PP/29 Fix three issues highlighted by clang analyser static analysis.
984 Only crash-plausible issue would require the Cambridge-specific
985 iplookup router and a misconfiguration.
986 Report from Marcin Mirosław.
988 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
990 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
991 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
992 As part of this, removing so much warning spew let me fix some minor
993 real issues in debug logging.
995 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
996 assignment on my part. Fixed.
998 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
999 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1000 Janne Snabb (who went above and beyond: thank you).
1002 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1003 string otherwise requires a connection and a bunch more work and it's
1004 relatively easy to get wrong. Should also expose TLS library linkage
1007 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1008 64-bit ${eval} (JH/03).
1010 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1011 GNU libc to support some of the 64-bit stuff, should not lead to
1012 conflicts. Defined before os.h is pulled in, so if a given platform
1013 needs to override this, it can.
1015 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1016 protection layer was required, which is not implemented.
1017 Bugzilla 1254, patch from Wolfgang Breyha.
1019 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1020 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1021 tls_dhparam take prime identifiers. Also unbreak combination of
1022 OpenSSL+DH_params+TLSSNI.
1024 PP/39 Disable SSLv2 by default in OpenSSL support.
1030 PP/01 Solaris build fix for Oracle's LDAP libraries.
1031 Bugzilla 1109, patch from Stephen Usher.
1033 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1035 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1038 TF/02 Fix a couple more cases where we did not log the error message
1039 when unlink() failed. See also change 4.74-TF/03.
1041 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1042 lock up or crash if it happened to be inside a call to libc when it
1043 got a SIGUSR1 from exiwhat.
1045 The SIGUSR1 handler appends the current process status to the process
1046 log which is later printed by exiwhat. It used to use the general
1047 purpose logging code to do this, but several functions it calls are
1048 not safe for signals.
1050 The new output code in the SIGUSR1 handler is specific to the process
1051 log, and simple enough that it's easy to inspect for signal safety.
1052 Removing some special cases also simplifies the general logging code.
1053 Removing the spurious timestamps from the process log simplifies
1056 TF/04 Improved ratelimit ACL condition.
1058 The /noupdate option has been deprecated in favour of /readonly which
1059 has clearer semantics. The /leaky, /strict, and /readonly update modes
1060 are mutually exclusive. The update mode is no longer included in the
1061 database key; it just determines when the database is updated. (This
1062 means that when you upgrde Exim will forget old rate measurements.)
1064 Exim now checks that the per_* options are used with an update mode that
1065 makes sense for the current ACL. For example, when Exim is processing a
1066 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1067 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1068 must specify per_mail/readonly. If you omit the update mode it defaults to
1069 /leaky where that makes sense (as before) or /readonly where required.
1071 The /noupdate option is now undocumented but still supported for
1072 backwards compatibility. It is equivalent to /readonly except that in
1073 ACLs where /readonly is required you may specify /leaky/noupdate or
1074 /strict/noupdate which are treated the same as /readonly.
1076 A useful new feature is the /count= option. This is a generalization
1077 of the per_byte option, so that you can measure the throughput of other
1078 aggregate values. For example, the per_byte option is now equivalent
1079 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1081 The per_rcpt option has been generalized using the /count= mechanism
1082 (though it's more complicated than the per_byte equivalence). When it is
1083 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1084 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1085 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1086 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1087 Note that using per_rcpt with a non-readonly update mode in more than
1088 one ACL will cause the recipients to be double-counted. (The per_mail
1089 and per_byte options don't have this problem.)
1091 The handling of very low rates has changed slightly. If the computed rate
1092 is less than the event's count (usually one) then this event is the first
1093 after a long gap. In this case the rate is set to the same as this event's
1094 count, so that the first message of a spam run is counted properly.
1096 The major new feature is a mechanism for counting the rate of unique
1097 events. The new per_addr option counts the number of different
1098 recipients that someone has sent messages to in the last time period. It
1099 behaves like per_rcpt if all the recipient addresses are different, but
1100 duplicate recipient addresses do not increase the measured rate. Like
1101 the /count= option this is a general mechanism, so the per_addr option
1102 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1103 example, measure the rate that a client uses different sender addresses
1104 with the options per_mail/unique=$sender_address. There are further
1105 details in the main documentation.
1107 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1109 TF/06 Removed a few PCRE remnants.
1111 TF/07 Automatically extract Exim's version number from tags in the git
1112 repository when doing development or release builds.
1114 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1115 Bugzilla 879. Patch from Paul Fisher.
1117 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1118 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1121 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1123 PP/05 Variable $av_failed, true if the AV scanner deferred.
1124 Bugzilla 1078. Patch from John Horne.
1126 PP/06 Stop make process more reliably on build failure.
1127 Bugzilla 1087. Patch from Heiko Schlittermann.
1129 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1130 Bugzilla 1089. Patch from Heiko Schlittermann.
1132 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1133 Bugzilla 1131. Patch from Holger Weiß.
1135 PP/09 Handle IPv6 addresses with SPF.
1136 Bugzilla 860. Patch from Wolfgang Breyha.
1138 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1140 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1143 PP/11 match_* no longer expand right-hand-side by default.
1144 New compile-time build option, EXPAND_LISTMATCH_RHS.
1145 New expansion conditions, "inlist", "inlisti".
1147 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1149 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1151 PP/14 fix log_write() format string regression from TF/03.
1152 Bugzilla 1152. Patch from Dmitry Isaikin.
1158 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1160 PP/02 Harmonised TLS library version reporting; only show if debugging.
1161 Layout now matches that introduced for other libraries in 4.74 PP/03.
1163 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1165 PP/04 New "dns_use_edns0" global option.
1167 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1170 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1171 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1173 TK/01 Updated PolarSSL code to 0.14.2.
1174 Bugzilla 1097. Patch from Andreas Metzler.
1176 PP/07 Catch divide-by-zero in ${eval:...}.
1177 Fixes bugzilla 1102.
1179 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1182 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1183 format-string attack -- SECURITY: remote arbitrary code execution.
1185 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1186 time unintentionally subject to list matching rules, letting the header
1187 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1188 arbitrary string expansion). This allowed for information disclosure.
1190 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1191 INT_MIN/-1 -- value coerced to INT_MAX.
1197 NM/01 Workround for PCRE version dependency in version reporting
1200 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1201 This fixes portability to compilers other than gcc, notably
1202 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1204 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1205 makefiles for portability to HP-UX and POSIX correctness.
1207 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1208 Also via indented variable definition in the Makefile.
1209 (Debugging by Oliver Heesakkers).
1211 PP/02 Restore caching of spamd results with expanded spamd_address.
1212 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1214 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1215 Improves build reliability. Fix from: Frank Elsner
1217 NM/02 Fix wide character breakage in the rfc2047 coding
1218 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1220 NM/03 Allow underscore in dnslist lookups
1221 Fixes bug 1026. Patch from Graeme Fowler
1223 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1224 Code patches from Adam Ciarcinski of NetBSD.
1226 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1229 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1230 is logged, to avoid truncation. Patch from John Horne.
1232 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1233 Patch from Jakob Hirsch.
1235 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1236 SQL string expansion failure details.
1237 Patch from Andrey Oktyabrski.
1239 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1240 Patch from Simon Arlott.
1242 PP/09 New lookups functionality failed to compile on old gcc which rejects
1243 extern declarations in function scope.
1244 Patch from Oliver Fleischmann
1246 PP/10 Use sig_atomic_t for flags set from signal handlers.
1247 Check getgroups() return and improve debugging.
1248 Fixed developed for diagnosis in bug 927 (which turned out to be
1251 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1252 Patch from Mark Zealey.
1254 PP/12 Bugzilla 1056: Improved spamd server selection.
1255 Patch from Mark Zealey.
1257 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1258 Based on patch from Heiko Schlittermann.
1260 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1261 Patch from Uwe Doering, sign-off by Michael Haardt.
1263 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1264 variable declaration deep within a block. Bug and patch from
1267 PP/15 lookups-Makefile IRIX compatibilty coercion.
1269 PP/16 Make DISABLE_DKIM build knob functional.
1271 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1272 Patch by Simon Arlott
1274 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1275 variable argument macros. Our copy now differs from upstream.
1281 TF/01 Failure to get a lock on a hints database can have serious
1282 consequences so log it to the panic log.
1284 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1285 controlled using the smtp_confirmation log selector.
1287 TF/03 Include the error message when we fail to unlink a spool file.
1289 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1290 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1291 for maintaining out-of-tree patches for some time.
1293 PP/01 Bugzilla 139: Documentation and portability issues.
1294 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1295 Handle per-OS dynamic-module compilation flags.
1297 PP/02 Let /dev/null have normal permissions.
1298 The 4.73 fixes were a little too stringent and complained about the
1299 permissions on /dev/null. Exempt it from some checks.
1300 Reported by Andreas M. Kirchwitz.
1302 PP/03 Report version information for many libraries, including
1303 Exim version information for dynamically loaded libraries. Created
1304 version.h, now support a version extension string for distributors
1305 who patch heavily. Dynamic module ABI change.
1307 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1308 privilege escalation vulnerability whereby the Exim run-time user
1309 can cause root to append content of the attacker's choosing to
1312 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1315 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1316 If dropping privileges for untrusted macros, we disabled normal logging
1317 on the basis that it would fail; for the Exim run-time user, this is not
1318 the case, and it resulted in successful deliveries going unlogged.
1319 Fixed. Reported by Andreas Metzler.
1325 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1326 only prepend for the Resent-* case. Fixes regression introduced in
1327 Exim 4.70 by NM/22 for Bugzilla 607.
1329 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1330 increasing numbers of administrators be bitten by this.
1332 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1334 PP/03 Bugzilla 994: added openssl_options main configuration option.
1336 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1338 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1340 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1342 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1343 without a peer certificate, leading to a segfault because of an
1344 assumption that peers always have certificates. Be a little more
1345 paranoid. Problem reported by Martin Tscholak.
1347 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1348 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1349 NB: ClamAV planning to remove STREAM in "middle of 2010".
1350 CL also introduces -bmalware, various -d+acl logging additions and
1351 more caution in buffer sizes.
1353 PP/09 Implemented reverse_ip expansion operator.
1355 PP/10 Bugzilla 937: provide a "debug" ACL control.
1357 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1359 PP/12 Bugzilla 973: Implement --version.
1361 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1363 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1365 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1367 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1368 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1369 ignore trailing whitespace.
1371 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1373 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1376 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1377 Notification from Dr Andrew Aitchison.
1379 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1380 ExtendedDetectionInfo response format.
1381 Notification from John Horne.
1383 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1386 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1387 XSL and documented dependency on system catalogs, with examples of how
1390 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1393 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1394 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1395 configuration file which is writeable by the Exim user or group.
1397 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1398 of configuration files to cover files specified with the -C option if
1399 they are going to be used with root privileges, not just the default
1402 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1403 option (effectively making it always true).
1405 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1406 files to be used while preserving root privileges.
1408 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1409 that rogue child processes cannot use them.
1411 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1412 run-time user, instead of root.
1414 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1415 Exim run-time user without dropping privileges.
1417 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1418 result string, instead of calling string_vformat() twice with the same
1421 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1422 for other users. Others should always drop root privileges if they use
1423 -C on the command line, even for a whitelisted configure file.
1425 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1427 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1433 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1434 $data_path, and $header_path variables; fixed documentation bugs and
1437 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1438 exipick to access non-standard spools, including the "frozen" queue
1441 NM/01 Bugzilla 965: Support mysql stored procedures.
1442 Patch from Alain Williams
1444 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1446 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1447 Patch from Andreas Metzler
1449 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1450 Patch from Kirill Miazine
1452 NM/05 Bugzilla 671: Added umask to procmail example.
1454 JJ/03 installed exipick 20100323.0, fixing doc bug
1456 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1457 directory. Notification and patch from Dan Rosenberg.
1459 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1461 TK/02 Improve log output when DKIM signing operation fails.
1463 MH/01 Treat the transport option dkim_domain as a colon separated
1464 list, not as a single string, and sign the message with each element,
1465 omitting multiple occurences of the same signer.
1467 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1468 Bugzilla 985, 986. Patch by Simon Arlott
1470 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1471 Patch by Simon Arlott
1473 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1474 MBX locking. Notification from Dan Rosenberg.
1480 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1482 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1484 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1486 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1488 NM/04 Bugzilla 915: Flag broken perl installation during build.
1494 TK/01 Added patch by Johannes Berg that expands the main option
1495 "spamd_address" if it starts with a dollar sign.
1497 TK/02 Write list of recipients to X-Envelope-Sender header when building
1498 the mbox-format spool file for content scanning (suggested by Jakob
1501 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1502 (http://www.dcc-servers.net/) support via dccifd. Activated by
1503 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1505 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1506 by Mark Daniel Reidel <mr@df.eu>.
1508 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1509 When building exim an external PCRE library is now needed -
1510 PCRE is a system library on the majority of modern systems.
1511 See entry on PCRE_LIBS in EDITME file.
1513 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1514 conversation. Added nologin parameter to request.
1515 Patch contributed by Kirill Miazine.
1517 TF/01 Do not log submission mode rewrites if they do not change the address.
1519 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1521 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1522 log files in place. Contributed by Roberto Lima.
1524 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1526 TF/03 Bugzilla 615: When checking the local_parts router precondition
1527 after a local_part_suffix or local_part_prefix option, Exim now
1528 does not use the address's named list lookup cache, since this
1529 contains cached lookups for the whole local part.
1531 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1532 Robert Millan. Documentation is in experimental-spec.txt.
1534 TF/04 Bugzilla 668: Fix parallel build (make -j).
1536 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1538 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1539 Patch provided by Jan Srzednicki.
1541 TF/05 Leading white space used to be stripped from $spam_report which
1542 wrecked the formatting. Now it is preserved.
1544 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1545 that they are available at delivery time.
1547 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1549 TF/08 TLS error reporting now respects the incoming_interface and
1550 incoming_port log selectors.
1552 TF/09 Produce a more useful error message if an SMTP transport's hosts
1553 setting expands to an empty string.
1555 NM/06 Bugzilla 744: EXPN did not work under TLS.
1556 Patch provided by Phil Pennock.
1558 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1559 Patch provided by Richard Godbee.
1561 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1562 acl_smtp_notquit, added index entry.
1564 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1565 Patch provided by Eugene Bujak.
1567 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1568 accept(). Patch provided by Maxim Dounin.
1570 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1571 Patch provided by Phil Pennock.
1573 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1575 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1576 Patch provided by Brad "anomie" Jorsch.
1578 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1579 Patch provided by Dean Brooks.
1581 TK/05 Add native DKIM support (does not depend on external libraries).
1583 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1584 Patch provided by Graeme Fowler.
1586 NM/16 Bugzilla 851: Documentation example syntax fix.
1588 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1590 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1593 NM/19 Bugzilla 745: TLS version reporting.
1594 Patch provided by Phil Pennock.
1596 NM/20 Bugzilla 167: bool: condition support.
1597 Patch provided by Phil Pennock.
1599 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1600 clients. Patch provided by Phil Pennock.
1602 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1603 Patch provided by Brad "anomie" Jorsch.
1605 NM/23 Bugzilla 687: Fix misparses in eximstats.
1606 Patch provided by Heiko Schlittermann.
1608 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1609 Patch provided by Heiko Schlittermann.
1611 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1612 plus update to original patch.
1614 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1616 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1617 Patch provided by David Brownlee.
1619 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1621 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1623 NM/30 Bugzilla 888: TLS documentation bugfixes.
1625 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1627 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1628 Unlike the original bugzilla I have changed all shell scripts in src tree.
1630 NM/33 Bugzilla 898: Transport filter timeout fix.
1631 Patch by Todd Rinaldo.
1633 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1634 Patch by Serge Demonchaux.
1636 NM/35 Bugzilla 39: Base64 decode bug fixes.
1637 Patch by Jakob Hirsch.
1639 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1641 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1643 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1645 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1651 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1652 ALT-N's libdkim that I have put here:
1653 http://duncanthrax.net/exim-experimental/
1655 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1656 were called 'true' and it seems that C99 defines that as a reserved
1657 keyword to be used with 'bool' variable types. That means you could
1658 not include C99-style headers which use bools without triggering
1659 build errors in sieve.c.
1661 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1662 as mailq or other aliases. Changed the --help handling significantly
1663 to do whats expected. exim_usage() emits usage/help information.
1665 SC/01 Added the -bylocaldomain option to eximstats.
1667 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1669 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1671 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1677 PH/01 Another patch from the Sieve maintainer.
1679 PH/02 When an IPv6 address is converted to a string for single-key lookup
1680 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1681 dots are used instead of colons so that keys in lsearch files need not
1682 contain colons. This was done some time before quoting was made available
1683 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1684 (notated using the quote facility) so as to distinguish them from IPv4
1685 keys. This meant that lookups for IP addresses in host lists did not work
1686 for iplsearch lookups.
1688 This has been fixed by arranging for IPv6 addresses to be expressed with
1689 colons if the lookup type is iplsearch. This is not incompatible, because
1690 previously such lookups could never work.
1692 The situation is now rather anomolous, since one *can* have colons in
1693 ordinary lsearch keys. However, making the change in all cases is
1694 incompatible and would probably break a number of configurations.
1696 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1699 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1700 conversion specification without a maximum field width, thereby enabling
1701 a rogue spamd server to cause a buffer overflow. While nobody in their
1702 right mind would setup Exim to query an untrusted spamd server, an
1703 attacker that gains access to a server running spamd could potentially
1704 exploit this vulnerability to run arbitrary code as the Exim user.
1706 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1707 $primary_hostname instead of what libspf2 thinks the hosts name is.
1709 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1710 a directory entry by the name of the lookup key. Previously, if a
1711 symlink pointed to a non-existing file or a file in a directory that
1712 Exim lacked permissions to read, a lookup for a key matching that
1713 symlink would fail. Now it is enough that a matching directory entry
1714 exists, symlink or not. (Bugzilla 503.)
1716 PH/03 The body_linecount and body_zerocount variables are now exported in the
1719 PH/04 Added the $dnslist_matched variable.
1721 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1722 This means they are set thereafter only if the connection becomes
1725 PH/06 Added the client_condition to authenticators so that some can be skipped
1726 by clients under certain conditions.
1728 PH/07 The error message for a badly-placed control=no_multiline_responses left
1729 "_responses" off the end of the name.
1731 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1733 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1734 (without spaces) instead of just copying the configuration text.
1736 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1738 PH/11 Added $max_received_linelength.
1740 PH/12 Added +ignore_defer and +include_defer to host lists.
1742 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1743 way in which PCRE > 7.0 is built.
1745 PH/14 Implemented queue_only_load_latch.
1747 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1748 MAIL command. The effect was to mangle the value on 64-bit systems.
1750 PH/16 Another patch from the Sieve maintainer.
1752 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1754 PH/18 If a system quota error occurred while trying to create the file for
1755 a maildir delivery, the message "Mailbox is full" was not appended to the
1756 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1757 only to a quota excession during the actual writing of the file.
1759 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1760 characters?) which causes problems in log lines. The DN values are now
1761 passed through string_printing() before being added to log lines.
1763 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1764 and InterBase are left for another time.)
1766 PH/21 Added message_body_newlines option.
1768 PH/22 Guard against possible overflow in moan_check_errorcopy().
1770 PH/23 POSIX allows open() to be a macro; guard against that.
1772 PH/24 If the recipient of an error message contained an @ in the local part
1773 (suitably quoted, of course), incorrect values were put in $domain and
1774 $local_part during the evaluation of errors_copy.
1780 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1781 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1784 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1785 issue a MAIL command.
1787 PH/02 In an ACL statement such as
1789 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1791 if a client was not listed at all, or was listed with a value other than
1792 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1793 the condition was not true (as it should be), so access was not denied.
1794 The bug was that the ! inversion was incorrectly passed on to the second
1795 item. This has been fixed.
1797 PH/03 Added additional dnslists conditions == and =& which are different from
1798 = and & when the dns lookup returns more than one IP address.
1800 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1801 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1803 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1804 FSYNC, which compiles an option called disable_fsync that allows for
1805 bypassing fsync(). The documentation is heavily laced with warnings.
1807 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1809 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1810 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1811 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1812 including adding "make clean"; (3) Added -fPIC when compiling the test
1813 dynamically loaded module, to get rid of a warning.
1815 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1816 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1817 = 0s. The bug is otherwise harmless.
1819 PH/07 There was a bug in the dovecot authenticator such that the value of
1820 $auth1 could be overwritten, and so not correctly preserved, after a
1821 successful authentication. This usually meant that the value preserved by
1822 the server_setid option was incorrect.
1824 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1826 PH/09 Installed PCRE release 7.0.
1828 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1829 run for batched SMTP input. It is now run at the start of every message
1830 in the batch. While fixing this I discovered that the process information
1831 (output by running exiwhat) was not always getting set for -bs and -bS
1832 input. This is fixed, and it now also says "batched" for BSMTP.
1834 PH/11 Added control=no_pipelining.
1836 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1837 patch, slightly modified), and move the expansion of helo_data till after
1838 the connection is made in the smtp transport (so it can use these
1841 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1843 PH/14 Added log_selector = +pid.
1845 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1847 PH/16 Add ${if forany and ${if forall.
1849 PH/17 Added dsn_from option to vary the From: line in DSNs.
1851 PH/18 Flush SMTP output before performing a callout, unless control =
1852 no_callout_flush is set.
1854 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1855 was true (the default) a successful delivery failed to delete the retry
1856 item, thus causing premature timeout of the address. The bug is now
1859 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1861 PH/21 Long custom messages for fakedefer and fakereject are now split up
1862 into multiline reponses in the same way that messages for "deny" and
1863 other ACL rejections are.
1865 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1866 with slight modification.
1868 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1869 draft, changing the syntax and factoring some duplicate code".
1871 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1872 for deliveries of the second and subsequent messages over the same SMTP
1875 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1876 ${reduce, with only minor "tidies".
1878 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1880 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1881 expansion side effects.
1883 PH/27 When a message times out after an over-quota error from an Exim-imposed
1884 quota, the bounce message says "mailbox is full". This message was not
1885 being given when it was a system quota that was exceeded. It now should
1888 MH/03 Made $recipients available in local_scan(). local_scan() already has
1889 better access to the recipient list through recipients_list[], but
1890 $recipients can be useful in postmaster-provided expansion strings.
1892 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1893 in the case of a MAIL command with additional options following the
1894 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1895 were accidentally chopped off.
1897 PH/29 SMTP synchronization checks are implemented when a command is read -
1898 there is a check that no more input is waiting when there shouldn't be
1899 any. However, for some commands, a delay in an ACL can mean that it is
1900 some time before the response is written. In this time, more input might
1901 arrive, invalidly. So now there are extra checks after an ACL has run for
1902 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1903 pipelining has not been advertised.
1905 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1907 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1908 returned in the SMTP response when smtp_return_error_details was false.
1909 This has been fixed.
1911 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1912 instead of the C I/O that was originally supplied, because problems were
1913 reported on Solaris.
1915 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1916 Exim which did not show up earlier: it was assuming that a call to
1917 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1918 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1919 was a macro that became an assignment, so it seemed to work. This has
1920 changed to a proper function call with a void return, hence the compile
1921 error. Exim's code has been fixed.
1923 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1926 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1928 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1930 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1931 because it turns out to be too big or there is a timeout) there is no
1932 "Completed" line in the log. When some messages of this type were
1933 selected by exigrep, they were listed as "not completed". Others were
1934 picked up by some special patterns. I have improved the selection
1935 criteria to be more general.
1937 PH/38 The host_find_failed option in the manualroute router can now be set
1938 to "ignore", to completely ignore a host whose IP address cannot be
1939 found. If all hosts are ignored, the behaviour is controlled by the new
1940 host_all_ignored option.
1942 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1943 homing or because of multiple MX records with /mx) generated more than
1944 one IP address, and the following item turned out to be the local host,
1945 all the secondary addresses of the first item were incorrectly removed
1946 from the list, along with the local host and any following hosts (which
1947 is what is supposed to happen).
1949 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1950 whoever called Exim into the -H file. In the case of the daemon it was
1951 behaving confusingly. When first started, it used values for whoever
1952 started the daemon, but after a SIGHUP it used the Exim user (because it
1953 calls itself on a restart). I have changed the code so that it now always
1956 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1957 message are rejected with the same error (e.g. no authentication or bad
1958 sender address), and a DATA command is nevertheless sent (as can happen
1959 with PIPELINING or a stupid MUA), the error message that was given to the
1960 RCPT commands is included in the rejection of the DATA command. This is
1961 intended to be helpful for MUAs that show only the final error to their
1964 PH/42 Another patch from the Sieve maintainer.
1966 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1967 Eximstats - Fixed some broken HTML links and added missing column headers
1969 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1970 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1972 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1974 PH/43 Yet another patch from the Sieve maintainer.
1976 PH/44 I found a way to check for a TCP/IP connection going away before sending
1977 the response to the final '.' that terminates a message, but only in the
1978 case where the client has not sent further data following the '.'
1979 (unfortunately, this is allowed). However, in many cases there won't be
1980 any further data because there won't be any more messages to send. A call
1981 to select() can be used: if it shows that the input is "ready", there is
1982 either input waiting, or the socket has been closed. An attempt to read
1983 the next input character can distinguish the two cases. Previously, Exim
1984 would have sent an OK response which the client would never have see.
1985 This could lead to message repetition. This fix should cure that, at
1986 least in a lot of common cases.
1988 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1989 advertised in response to EHLO.
1995 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1996 fixed by 4.65/MH/01 (is this a record?) are fixed:
1998 (i) An empty string was always treated as zero by the numeric comparison
1999 operators. This behaviour has been restored.
2001 (ii) It is documented that the numeric comparison operators always treat
2002 their arguments as decimal numbers. This was broken in that numbers
2003 starting with 0 were being interpreted as octal.
2005 While fixing these problems I realized that there was another issue that
2006 hadn't been noticed. Values of message_size_limit (both the global option
2007 and the transport option) were treated as octal if they started with 0.
2008 The documentation was vague. These values are now always treated as
2009 decimal, and I will make that clear in the documentation.
2015 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2016 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2019 MH/01 Don't check that the operands of numeric comparison operators are
2020 integers when their expansion is in "skipping" mode (fixes bug
2021 introduced by 4.64-PH/07).
2023 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2024 child addresses, Exim now panics and dies. Previously, because the count
2025 is held in a short int, deliveries were likely to be lost. As such a
2026 large number of recipients for a single message is ridiculous
2027 (performance will be very, very poor), I have chosen to impose a limit
2028 rather than extend the field.
2034 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2035 leftover -K file (the existence of which was triggered by #402).
2036 While we were at it, introduced process PID as part of the -K
2037 filename. This should rule out race conditions when creating
2040 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2041 processing considerably. Previous code took too long for large mails,
2042 triggering a timeout which in turn triggers #401.
2044 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2045 in the DK code in transports.c. sendfile() is not really portable,
2046 hence the _LINUX specificness.
2048 TF/01 In the add_headers option to the mail command in an Exim filter,
2049 there was a bug that Exim would claim a syntax error in any
2050 header after the first one which had an odd number of characters
2053 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2054 callout verification, Exim cached a "reject" for the entire domain. This
2055 is correct for most verifications, but it is not correct for a recipient
2056 verification with use_sender or use_postmaster set, because in that case
2057 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2058 case of MAIL FROM:<> rejection from other early rejections (e.g.
2059 rejection of HELO). When verifying a recipient using a non-null MAIL
2060 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2061 Whatever the result of the callout, the value of the domain cache is
2062 left unchanged (for any other kind of callout, getting as far as trying
2063 RCPT means that the domain itself is ok).
2065 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2068 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2069 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2070 session, the daemon ignores ECONNECT errors and logs others; it now
2071 ignores EPIPE as well.
2073 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2074 (quoted-printable decoding).
2076 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2077 later the small subsequent patch to fix an introduced bug.
2079 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2081 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2083 PH/08 An error is now given if message_size_limit is specified negative.
2085 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2086 to be given (somewhat) arbitrary names.
2088 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2091 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2092 miscellaneous code fixes
2094 PH/10 Added the log_reject_target ACL modifier to specify where to log
2097 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2098 hostname. This is wrong, because it relates to the incoming message (and
2099 probably the interface on which it is arriving) and not to the outgoing
2100 callout (which could be using a different interface). This has been
2101 changed to use the value of the helo_data option from the smtp transport
2102 instead - this is what is used when a message is actually being sent. If
2103 there is no remote transport (possible with a router that sets up host
2104 addresses), $smtp_active_hostname is used.
2106 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2107 tweaks were necessary in order to get it to work (see also 21 below):
2108 (a) The code assumed that strncpy() returns a negative number on buffer
2109 overflow, which isn't the case. Replaced with Exim's string_format()
2111 (b) There were several signed/unsigned issues. I just did the minimum
2112 hacking in of casts. There is scope for a larger refactoring.
2113 (c) The code used strcasecmp() which is not a standard C function.
2114 Replaced with Exim's strcmpic() function.
2115 (d) The code set only $1; it now sets $auth1 as well.
2116 (e) A simple test gave the error "authentication client didn't specify
2117 service in request". It would seem that Dovecot has changed its
2118 interface. Fortunately there's a specification; I followed it and
2119 changed what the client sends and it appears to be working now.
2121 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2124 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2125 address A is aliased to B and C, where B exists and C does not. Without
2126 -v the output is "A verified" because verification stops after a
2127 successful redirection if more than one address is generated. However,
2128 with -v the child addresses are also verified. Exim was outputting "A
2129 failed to verify" and then showing the successful verification for C,
2130 with its parentage. It now outputs "B failed to verify", showing B's
2131 parentage before showing the successful verification of C.
2133 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2134 look up a TXT record in a specific list after matching in a combined
2137 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2138 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2139 they consult the DNS. I had assumed they would set it the way they
2140 wanted; and indeed my experiments on Linux seem to show that in some
2141 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2142 To be on the safe side, however, I have now made the interface to
2143 host_find_byname() similar to host_find_bydns(), with an argument
2144 containing the DNS resolver options. The host_find_byname() function now
2145 sets these options at its start, just as host_find_bydns() does. The smtp
2146 transport options dns_qualify_single and dns_search_parents are passed to
2147 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2148 of host_find_byname() use the default settings of RES_DEFNAMES
2149 (qualify_single) but not RES_DNSRCH (search_parents).
2151 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2152 spool_read_header() do less string testing, by means of a preliminary
2153 switch on the second character of optional "-foo" lines. (This is
2154 overdue, caused by the large number of possibilities that now exist.
2155 Originally there were few.) While I was there, I also converted the
2156 str(n)cmp tests so they don't re-test the leading "-" and the first
2157 character, in the hope this might squeeze out yet more improvement.
2159 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2160 flag allowing group syntax was set by the header_syntax check but not
2161 turned off, possible causing trouble later; (2) The flag was not being
2162 set at all for the header_verify test, causing "group"-style headers to
2163 be rejected. I have now set it in this case, and also caused header_
2164 verify to ignore an empty address taken from a group. While doing this, I
2165 came across some other cases where the code for allowing group syntax
2166 while scanning a header line wasn't quite right (mostly, not resetting
2167 the flag correctly in the right place). These bugs could have caused
2168 trouble for malformed header lines. I hope it is now all correct.
2170 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2171 with the "reply" argument non-NULL. The code, however (which originally
2172 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2173 but it didn't always do it. This confused somebody who was copying the
2174 code for some other use. I have removed all the tests.
2176 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2177 feature that was used to support insecure browsers during the U.S. crypto
2178 embargo. It requires special client support, and Exim is probably the
2179 only MTA that supported it -- and would never use it because real RSA is
2180 always available. This code has been removed, because it had the bad
2181 effect of slowing Exim down by computing (never used) parameters for the
2182 RSA_EXPORT functionality.
2184 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2185 authenticator to fail if there's a tab character in the incoming data
2186 (there should never be unless someone is messing about, as it's supposed
2187 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2188 if the connection is using TLS or if the remote IP is the same as the
2189 local IP, and the "valid-client-cert option" if a client certificate has
2192 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2193 authenticators. This can be used for authorization after authentication
2194 succeeds. (In the case of plaintext, it servers for both authentication
2197 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2198 if any retry times were supplied.
2200 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2201 connection, where there is no client IP address to check. In this
2202 situation, the verify now always succeeds.
2204 PH/25 Applied John Jetmore's -Mset patch.
2206 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2208 PH/27 In a string expansion for a processed (not raw) header when multiple
2209 headers of the same name were present, leading whitespace was being
2210 removed from all of them, but trailing whitespace was being removed only
2211 from the last one. Now trailing whitespace is removed from each header
2212 before concatenation. Completely empty headers in a concatenation (as
2213 before) are ignored.
2215 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2216 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2218 PH/29 [Removed. This was a change that I later backed out, and forgot to
2219 correct the ChangeLog entry (that I had efficiently created) before
2220 committing the later change.]
2222 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2223 address errors (4xx response to RCPT) over the same connection as other
2224 messages routed to the same hosts. Such deliveries are always "forced",
2225 so retry times are not inspected. This resulted in far too many retries
2226 for the affected addresses. The effect occurred only when there were more
2227 hosts than the hosts_max_try setting in the smtp transport when it had
2228 the 4xx errors. Those hosts that it had tried were not added to the list
2229 of hosts for which the message was waiting, so if all were tried, there
2230 was no problem. Two fixes have been applied:
2232 (i) If there are any address or message errors in an SMTP delivery, none
2233 of the hosts (tried or untried) are now added to the list of hosts
2234 for which the message is waiting, so the message should not be a
2235 candidate for sending over the same connection that was used for a
2236 successful delivery of some other message. This seems entirely
2237 reasonable: after all the message is NOT "waiting for some host".
2238 This is so "obvious" that I'm not sure why it wasn't done
2239 previously. Hope I haven't missed anything, but it can't do any
2240 harm, as the worst effect is to miss an optimization.
2242 (ii) If, despite (i), such a delivery is accidentally attempted, the
2243 routing retry time is respected, so at least it doesn't keep
2244 hammering the server.
2246 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2247 in ${readsocket because some servers need this prod.
2249 PH/32 Added some extra debug output when updating a wait-xxx database.
2251 PH/33 The hint "could be header name not terminated by colon", which has been
2252 given for certain expansion errors for a long time, was not being given
2253 for the ${if def:h_colon_omitted{... case.
2255 PH/34 The spec says: "With one important exception, whenever a domain list is
2256 being scanned, $domain contains the subject domain." There was at least
2257 one case where this was not true.
2259 PH/35 The error "getsockname() failed: connection reset by peer" was being
2260 written to the panic log as well as the main log, but it isn't really
2261 panic-worthy as it just means the connection died rather early on. I have
2262 removed the panic log writing for the ECONNRESET error when getsockname()
2265 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2266 runs only) independently of the message's sender address. This meant
2267 that, if the 4xx error was in fact related to the sender, a different
2268 message to the same recipient with a different sender could confuse
2269 things. In particualar, this can happen when sending to a greylisting
2270 server, but other circumstances could also provoke similar problems.
2271 I have changed the default so that the retry time for these errors is now
2272 based a combination of the sender and recipient addresses. This change
2273 can be overridden by setting address_retry_include_sender=false in the
2276 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2277 remote server are returned as part of bounce messages. This was not
2278 happening for LMTP over a pipe (the lmtp transport), but now it is the
2279 same for both kinds of LMTP.
2281 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2282 in header lines that were in fact CNAMEs. This is no longer the case.
2284 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2285 and queue runs started by the daemon processed all messages. This has
2286 been fixed so that -R and -S can now usefully be given with -q<time>.
2288 PH/40 Import PCRE release 6.7 (fixes some bugs).
2290 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2292 PH/42 Give an error if -q is specified more than once.
2294 PH/43 Renamed the variables $interface_address and $interface_port as
2295 $received_ip_address and $received_port, to make it clear that these
2296 values apply to message reception, and not to the outgoing interface when
2297 a message is delivered. (The old names remain recognized, of course.)
2299 PH/44 There was no timeout on the connect() call when using a Unix domain
2300 socket in the ${readsocket expansion. There now is.
2302 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2303 be meaningful with "accept".
2305 SC/01 Eximstats V1.43
2306 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2308 SC/02 Eximstats V1.44
2309 Use a glob alias rather than an array ref in the generated
2310 parser. This improves both readability and performance.
2312 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2313 Collect SpamAssassin and rejection statistics.
2314 Don't display local sender or destination tables unless
2315 there is data to show.
2316 Added average volumes into the top table text output.
2318 SC/04 Eximstats V1.46
2319 Collect data on the number of addresses (recipients)
2320 as well as the number of messages.
2322 SC/05 Eximstats V1.47
2323 Added 'Message too big' to the list of mail rejection
2324 reasons (thanks to Marco Gaiarin).
2326 SC/06 Eximstats V1.48
2327 Mainlog lines which have GMT offsets and are too short to
2328 have a flag are now skipped.
2330 SC/07 Eximstats V1.49 (Alain Williams)
2331 Added the -emptyok flag.
2333 SC/08 Eximstats V1.50
2334 Fixes for obtaining the IP address from reject messages.
2336 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2337 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2338 whitesspace changes from 4.64-PH/27
2340 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2343 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2344 are found, allow negative numbers in numeric criteria)
2346 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2348 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2349 to match changes made in 4.64-PH/43
2351 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2353 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2354 "domain" is actually an IP address, and force a failure. This locks out
2355 those revolvers/nameservers that support "A-for-A" lookups, in
2356 contravention of the specifications.
2358 PH/48 When a host name was looked up from an IP address, and the subsequent
2359 forward lookup of the name timed out, the host name was left in
2360 $sender_host_name, contrary to the specification.
2362 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2363 restricted to single-key lookups, Exim was not diagnosing an error if
2364 * or *@ was used with a query-style lookup.
2366 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2368 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2369 long ago, but noone interested enough thought of it. Let's just say that
2370 the "1.1" means that there are some new functions that weren't there at
2371 some point in the past.
2373 PH/51 Error processing for expansion failure of helo_data from an smtp
2374 transport during callout processing was broken.
2376 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2377 tested/used via the -bh/-bhc/-bs options.
2379 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2380 bug, fixed in subsequent PCRE releases).
2382 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2383 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2385 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2391 SC/01 Use a glob alias rather than an array ref in eximstats generated
2392 parser. This improves both readability and performance.
2394 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2395 Don't display local sender or destination tables in eximstats unless
2396 there is data to show.
2397 Added average volumes into the eximstats top table text output.
2399 SC/03 Collect data on the number of addresses (recipients) as well
2400 as the number of messages in eximstats.
2402 TF/01 Correct an error in the documentation for the redirect router. Exim
2403 does (usually) call initgroups() when daemonizing.
2405 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2406 with consistent privilege compared to when running as a daemon.
2408 TF/03 Note in the spec that $authenticated_id is not set for local
2409 submissions from trusted users.
2411 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2412 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2414 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2415 by adding some example configuration directives to the default
2416 configuration file. A little bit of work is required to uncomment the
2417 directives and define how usernames and passwords are checked, but
2418 there is now a framework to start from.
2420 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2421 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2422 without this. I don't know how relevant this is to other LDAP libraries.
2424 PH/02 Add the verb name to the "unknown ACL verb" error.
2426 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2428 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2430 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2431 directory not expanded when it should be if an expanded home directory
2432 was set for the address (which is overridden by the transport).
2434 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2437 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2438 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2439 because it is too late at that time, and has no effect.
2441 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2442 security issue with \' (bugzilla #107). I could not use the
2443 PQescapeStringConn() function, because it needs a PGconn value as one of
2446 PH/08 When testing addresses using -bt, indicate those final addresses that
2447 are duplicates that would not cause an additional delivery. At least one
2448 person was confused, thinking that -bt output corresponded to deliveries.
2449 (Suppressing duplicates isn't a good idea as you lose the information
2450 about possibly different redirections that led to the duplicates.)
2452 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2453 systems where poll() doesn't work, in particular OS X.
2455 PH/10 Added more information to debugging output for retry time not reached.
2457 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2458 operations in malware.c.
2460 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2463 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2464 syslog_duplication set false, log lines that would normally be written
2465 both the the main log and to the reject log were not written to syslog at
2468 PH/14 In the default configuration, change the use of "message" in ACL warn
2469 statements to "add_header".
2471 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2472 not followed by a command (e.g. "seen endif").
2474 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2475 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2478 PH/17 Added extra conditions to the default value of delay_warning_condition
2482 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2483 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2484 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2487 The Auto-Submitted: and various List- headers are standardised, whereas I
2488 don't think Precedence: ever was.
2490 PH/18 Refactored debugging code in route_finduser() to show more information,
2491 in particular, the error code if getpwnam() issues one.
2493 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2494 This is apparently needed in addition to the PH/07 change above to avoid
2495 any possible encoding problems.
2497 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2498 but not after initializing Perl.
2500 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2501 output them only if debugging. By default they are written stderr,
2502 apparently, which is not desirable.
2504 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2507 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2510 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2512 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2513 authenticated or an ident call has been made. Suppress the default
2514 values for $authenticated_id and $authenticated_sender (but permit -oMai
2515 and -oMas) when testing with -bh.
2517 PH/24 Re-jigged the order of the tests in the default configuration so that the
2518 tests for valid domains and recipients precede the DNS black list and CSA
2519 tests, on the grounds that those ones are more expensive.
2521 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2522 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2523 This bug exists in every version of Exim that I still have, right back to
2526 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2527 However, an attempt to turn on case-sensitivity in a regex key by
2528 including (?-i) didn't work because the subject string was already
2529 lowercased, and the effects were non-intuitive. It turns out that a
2530 one-line patch can be used to allow (?-i) to work as expected.
2536 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2537 other effects) broke the use of negated acl sub-conditions.
2539 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2542 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2543 "Deny" causes Exim to reject the incoming connection with a 554 error.
2544 Unfortunately, if there is a major crisis, such as a disk failure,
2545 tcp-wrappers gives "deny", whereas what one would like would be some
2546 kind of temporary error. A kludge has been added to help with this.
2547 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2548 554 error is used if errno is still zero or contains ENOENT (which occurs
2549 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2552 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2554 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2555 errors. Otherwise a message that provokes a temporary error (when other
2556 messages do not) can cause a whole host to time out.
2558 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2559 addresses were routed directly to files or pipes from a redirect router.
2560 File deliveries just didn't batch; pipe deliveries might have suffered
2563 PH/06 A failure to get a lock for a hints database would erroneously always say
2564 "Failed to get write lock", even when it was really a read lock.
2566 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2567 of 0600. This has been changed to use the value of the lockfile_mode
2568 option (which defaults to 0600).
2570 PH/08 Applied small patch from the Sieve maintainer.
2572 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2573 folder from quota calculations, a direct delivery into this folder messed
2574 up the contents of the maildirsize file. This was because the regex was
2575 used only to exclude .Trash (or whatever) when the size of the mailbox
2576 was calculated. There was no check that a delivery was happening into an
2577 excluded directory. This bug has been fixed by ignoring all quota
2578 processing for deliveries into excluded directories.
2580 PH/10 Added the maildirfolder_create_regex option to appendfile.
2586 PH/01 The code for finding all the local interface addresses on a FreeBSD
2587 system running IPv6 was broken. This may well have applied to all BSD
2588 systems, as well as to others that have similar system calls. The broken
2589 code found IPv4 interfaces correctly, but gave incorrect values for the
2590 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2591 that it would not match correctly against @[] and not recognize the IPv6
2594 PH/02 The ipliteral router was not recognizing addresses of the form user@
2595 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2597 PH/03 Added disable_ipv6.
2599 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2600 decoded form, because it is most often used to construct To: headers
2601 lines in autoreplies, and the decoded form may well be syntactically
2602 invalid. However, $reply_address has leading white space removed, and all
2603 newlines turned into spaces so that the autoreply transport does not
2606 PH/05 If group was specified without a user on a router, and no group or user
2607 was specified on a transport, the group from the router was ignored.
2609 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2610 for visible compile-time settings that can be used to change these
2611 numbers, for those that want even more. Backwards compatibility with old
2612 spool files has been maintained. However, going back to a previous Exim
2613 release will lost any variables that are in spool files.
2615 PH/07 Two small changes when running in the test harness: increase delay when
2616 passing a TCP/IP connection to a new process, in case the original
2617 process has to generate a bounce, and remove special handling of
2618 127.0.0.2 (sic), which is no longer necessary.
2620 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2621 be the same on different OS.
2623 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2626 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2627 whether --show-vars was specified or not
2629 JJ/02 exipick: Added support for new ACL variable spool format introduced
2632 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2633 syntactically invalid From: or Reply-to: line, and a filter used this to
2634 generate an autoreply, and therefore failed to obtain an address for the
2635 autoreply, Exim could try to deliver to a non-existent relative file
2636 name, causing unrelated and misleading errors. What now happens is that
2637 it logs this as a hard delivery error, but does not attempt to create a
2640 PH/11 The exinext utility has a -C option for testing purposes, but although
2641 the given file was scanned by exinext itself; it wasn't being passed on
2642 when Exim was called.
2644 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2645 an end-of-file indication when reading a command response.
2647 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2648 compiled. In many other places in Exim, IPv6 addresses are always
2649 recognized, so I have changed this. It also means that IPv4 domain
2650 literals of the form [IPV4:n.n.n.n] are now always recognized.
2652 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2653 used if the router is not running as root, for example, when verifying at
2654 ACL time, or when using -bh. The debugging output from this situation was
2655 non-existent - all you got was a failure to exec. I have made two
2658 (a) Failures to set uid/gid, the current directory, or a process leader
2659 in a subprocess such as that created by queryprogram now generate
2660 suitable debugging ouput when -d is set.
2662 (b) The queryprogram router detects when it is not running as root,
2663 outputs suitable debugging information if -d is set, and then runs
2664 the subprocess without attempting to change uid/gid.
2666 PH/15 Minor change to Makefile for building test_host (undocumented testing
2669 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2670 additional section of a DNS packet that returns MX or SRV records.
2671 Instead, it always explicitly searches for A/AAAA records. This avoids
2672 major problems that occur when a DNS server includes only records of one
2673 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2674 fixed another bug: if SRV records were looked up and the corresponding
2675 address records were *not* found in the additional section, the port
2676 values from the SRV records were lost.
2678 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2679 using the correct key (the original address) when searching the retry
2680 rules in order to find which one to use for generating the retry hint.
2682 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2683 adding the default one. Similarly, if it contains a Reply-To: header, the
2684 errors_reply_to option, if set, is not used.
2686 PH/19 When calculating a retry time, Exim used to measure the "time since
2687 failure" by looking at the "first failed" field in the retry record. Now
2688 it does not use this if it is later than than the arrival time of the
2689 message. Instead it uses the arrival time. This makes for better
2690 behaviour in cases where some deliveries succeed, thus re-setting the
2691 "first failed" field. An example is a quota failure for a huge message
2692 when small messages continue to be delivered. Without this change, the
2693 "time since failure" will always be short, possible causing more frequent
2694 delivery attempts for the huge message than are intended.
2695 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2697 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2698 $1, $2, $3) because the numerical variables can be reset during some
2699 expansion items (e.g. "match"), thereby losing the authentication data.
2701 PH/21 Make -bV show the size of off_t variables so that the test suite can
2702 decide whether to run tests for quotas > 2G.
2704 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2705 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2706 filecount value is greater than 2G or if a quota value is greater than 2G
2707 on a system where the size of off_t is not greater than 4, a panic error
2710 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2711 never match. The debug and -bh output now contains an explicit error
2712 message indicating a malformed IPv4 address or mask.
2714 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2715 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2716 PH/23 above applies.
2718 PH/25 Do not write to syslog when running in the test harness. The only
2719 occasion when this arises is a failure to open the main or panic logs
2720 (for which there is an explicit test).
2722 PH/26 Added the /no_tell option to "control=freeze".
2724 PH/27 If a host name lookup failed very early in a connection, for example, if
2725 the IP address matched host_lookup and the reverse lookup yielded a name
2726 that did not have a forward lookup, an error message of the form "no IP
2727 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2728 could be logged. Now it outputs the IP address instead of "NULL".
2730 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2731 allows the sender and the authenticated sender to be set when
2732 submitting a message from within Exim. Since child_open_exim() is
2733 documented for local_scan(), the new function should be too.
2735 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2736 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2737 results in an empty string is now treated as unset.
2739 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2741 PH/31 Added sender_verify_fail logging option.
2743 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2744 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2745 tidied the source and removed it altogether.
2747 PH/33 When a queue run was abandoned because the load average was too high, a
2748 log line was always written; now it is written only if the queue_run log
2749 selector is set. In addition, the log line for abandonment now contains
2750 information about the queue run such as the pid. This is always present
2751 in "start" and "stop" lines but was omitted from the "abandon" line.
2753 PH/34 Omit spaces between a header name and the colon in the error message that
2754 is given when verify = headers_syntax fails (if there are lots of them,
2755 the message gets confusing).
2757 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2758 names, as there are now some PTR records that contain slashes. This check
2759 is only to protect against broken name servers that fall over on strange
2760 characters, so the fact that it applies to all lookups doesn't matter.
2762 PH/36 Now that the new test suite is complete, we can remove some of the
2763 special code in Exim that was needed for the old test suite. For example,
2764 sorting DNS records because real resolvers return them in an arbitrary
2765 order. The new test suite's fake resolver always returns records in the
2768 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2769 bounces) except when queue_only is set, to avoid logging races between
2770 the different processes.
2772 PH/38 Panic-die if .include specifies a non-absolute path.
2774 PH/39 A tweak to the "H" retry rule from its user.
2776 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2777 a label. They prevented compilation on older perls.
2779 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2780 a warning to be raised on newish perls.
2782 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2783 on queue. Changes to match documented behaviour of showing count of
2784 messages matching specified criteria.
2786 PH/40 Changed the default ident timeout from 30s to 5s.
2788 PH/41 Added support for the use of login_cap features, on those BSD systems
2789 that have them, for controlling the resources used by pipe deliveries.
2791 PH/42 The content-scanning code uses fopen() to create files in which to put
2792 message data. Previously it was not paying any attention to the mode of
2793 the files. Exim runs with umask(0) because the rest of the code creates
2794 files with open(), and sets the required mode explicitly. Thus, these
2795 files were ending up world-writeable. This was not a big issue, because,
2796 being within the spool directory, they were not world-accessible. I have
2797 created a function called modefopen, which takes an additional mode
2798 argument. It sets umask(777), creates the file, chmods it to the required
2799 mode, then resets the umask. All the relevant calls to fopen() in the
2800 content scanning code have been changed to use this function.
2802 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2803 to 24 hours. This avoids potential overflow problems when processing G
2804 and H retry rules. I suspect nobody ever tinkers with this value.
2806 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2808 PH/45 When the plaintext authenticator is running as a client, the server's
2809 challenges are checked to ensure they are valid base64 strings. By
2810 default, the authentication attempt is cancelled if an invalid string is
2811 received. Setting client_ignore_invalid_base64 true ignores these errors.
2812 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2813 they are received. Thus, the responses can be made to depend on the
2814 challenges. If an invalid string is ignored, an empty string is placed in
2817 PH/46 Messages that are created by the autoreply transport now contains a
2818 References: header, in accordance with RFCs 2822 and 3834.
2820 PH/47 Added authenticated_sender_force to the smtp transport.
2822 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2824 PH/49 Installed latest patch from the Sieve maintainer.
2826 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2827 was also set, the appendfile transport was unnecessarily scanning a
2828 directory of message files (e.g. for maildir delivery) to find the count
2829 of files (along with the size), even though it did not need this
2830 information. It now does the scan only if it needs to find either the
2831 size of the count of files.
2833 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2835 PH/52 Two bugs concerned with error handling when the smtp transport is
2838 (i) Exim was not creating retry information for temporary errors given
2839 for individual recipients after the DATA command when the smtp transport
2840 was used in LMTP mode. This meant that they could be retried too
2841 frequently, and not timed out correctly.
2843 (ii) Exim was setting the flag that allows error details to be returned
2844 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2845 recipients that were returned after the DATA command.
2847 PH/53 This is related to PH/52, but is more general: for any failing address,
2848 when detailed error information was permitted to be returned to the
2849 sender, but the error was temporary, then after the final timeout, only
2850 "retry timeout exceeded" was returned. Now it returns the full error as
2851 well as "retry timeout exceeded".
2853 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2854 do this, and (what is worse) MTAs that accept it.
2856 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2857 will now be deprecated.
2859 PH/56 New os.c-cygwin from the Cygwin maintainer.
2861 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2862 formats (previously only available in exim formats via -bpr, -bpru,
2863 and -bpra. Now also available in native and exiqgrep formats)
2865 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2866 with very large, slow to parse queues
2868 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2870 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2872 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2873 responses to authentication challenges, though it was showing the
2874 challenges; (ii) I've removed the CR characters from the debug output for
2877 PH/58 Allow for the insertion of a newline as well as a space when a string
2878 is turned into more than one encoded-word during RFC 2047 encoding. The
2879 Sieve code now uses this.
2881 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2882 data_4xx, lost_connection, tls_required.
2884 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2885 message was being sent as an SMTP response.
2887 PH/61 Add -l and -k options to exicyclog.
2889 PH/62 When verifying, if an address was redirected to one new address, so that
2890 verification continued, and the new address failed or deferred after
2891 having set something in $address_data, the value of $address_data was not
2892 passed back to the ACL. This was different to the case when no
2893 redirection occurred. The value is now passed back in both cases.
2895 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2896 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2897 use login_cap.h, so on its own it isn't the distinguishing feature. The
2898 new name refers directly to the setclassresources() function.
2900 PH/65 Added configuration files for NetBSD3.
2902 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2904 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2905 is preferred over IPv4.
2907 PH/68 The bounce_return_message and bounce_return_body options were not being
2908 honoured for bounces generated during the reception of non-SMTP messages.
2909 In particular, this applied to messages rejected by the ACL. This bug has
2910 been fixed. However, if bounce_return_message is true and bounce_return_
2911 body is false, the headers that are returned for a non-SMTP message
2912 include only those that have been read before the error was detected.
2913 (In the case of an ACL rejection, they have all been read.)
2915 PH/69 The HTML version of the specification is now built in a directory called
2916 spec_html instead of spec.html, because the latter looks like a path with
2917 a MIME-type, and this confuses some software.
2919 PH/70 Catch two compiler warnings in sieve.c.
2921 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2922 function verify_get_ident() calls ip_connect() to connect a socket, but
2923 if the "connect()" function timed out, ip_connect() used to close the
2924 socket. However, verify_get_ident() also closes the socket later, and in
2925 between Exim writes to the log, which may get opened at this point. When
2926 the socket was closed in ip_connect(), the log could get the same file
2927 descriptor number as the socket. This naturally causes chaos. The fix is
2928 not to close the socket in ip_connect(); the socket should be closed by
2929 the function that creates it. There was only one place in the code where
2930 this was missing, in the iplookup router, which I don't think anybody now
2931 uses, but I've fixed it anyway.
2933 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2934 well as to direct DNS lookups. Otherwise the handling of names in host
2935 lists is inconsistent and therefore confusing.
2941 PH/01 Two changes to the default runtime configuration:
2943 (1) Move the checks for relay_from_hosts and authenticated clients from
2944 after to before the (commented out) DNS black list checks.
2946 (2) Add control=submission to the relay_from_hosts and authenticated
2947 clients checks, on the grounds that messages accepted by these
2948 statements are most likely to be submissions.
2950 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2952 (1) Generate an error if the third argument for the ${prvs expansion is
2955 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2958 (3) Reset the variables that are obtained from the first argument of
2959 ${prvscheck and used in the second argument before leaving the code,
2960 because their memory is reclaimed, so using them afterwards may do
2963 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2964 one (it's much easier than Tom thought :-).
2966 (5) Because of (4), we can now allow for the use of $prvscheck_result
2967 inside the third argument.
2969 PH/03 For some reason, the default setting of PATH when running a command from
2970 a pipe transport was just "/usr/bin". I have changed it to
2973 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2974 anything to be listed in the output from -bV.
2976 PH/05 When a filter generated an autoreply, the entire To: header line was
2977 quoted in the delivery log line, like this:
2979 => >A.N.Other <ano@some.domain> <original@ddress> ...
2981 This has been changed so that it extracts the operative address. There
2982 may be more than one such address. If so, they are comma-separated, like
2985 => >ano@some.domain,ona@other.domain <original@ddress> ...
2987 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2988 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2989 not being looked up in the rDNS to get a host name, Exim was showing the
2990 IP address twice in Received: lines, even though the IP addresses were
2991 identical. For example:
2993 Received: from [1.2.3.4] (helo=[1.2.3.4])
2995 However, if the real host name was known, it was omitting the HELO data
2996 if it matched the actual IP address. This has been tidied up so that it
2997 doesn't show the same IP address twice.
2999 PH/07 When both +timestamp and +memory debugging was on, the value given by
3000 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3001 called by the debug printing, thereby overwriting the timestamp buffer.
3002 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3004 PH/08 When the original message was included in an autoreply transport, it
3005 always said "this is a copy of the message, including all the headers",
3006 even if body_only or headers_only was set. It now gives an appropriate
3009 PH/09 Applied a patch from the Sieve maintainer which:
3011 o fixes some comments
3012 o adds the (disabled) notify extension core
3013 o adds some debug output for the result of if/elsif tests
3014 o points to the current vacation draft in the documentation
3015 and documents the missing references header update
3019 o fixes a bug in processing the envelope test (when testing
3020 multiple envelope elements, the last element determinted the
3023 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3024 Electronic Mail") by including:
3026 Auto-submitted: auto-generated
3028 in the messages that it generates (bounce messages and others, such as
3029 warnings). In the case of bounce messages for non-SMTP mesages, there was
3030 also a typo: it was using "Auto_submitted" (underscore instead of
3031 hyphen). Since every message generated by Exim is necessarily in response
3032 to another message, thes have all been changed to:
3034 Auto-Submitted: auto-replied
3036 in accordance with these statements in the RFC:
3038 The auto-replied keyword:
3040 - SHOULD be used on messages sent in direct response to another
3041 message by an automatic process,
3043 - MUST NOT be used on manually-generated messages,
3045 - MAY be used on Delivery Status Notifications (DSNs) and Message
3046 Disposition Notifications (MDNs),
3048 - MUST NOT be used on messages generated by automatic or periodic
3049 processes, except for messages which are automatic responses to
3052 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3053 to the default Received: header definition.
3055 PH/12 Added log selector acl_warn_skipped (default on).
3057 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3058 variables because (a) they are in the wrong storage pool and (b) even if
3059 they were copied, it wouldn't work properly because of the caching.
3061 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3062 checking when decoding. Apparently there are clients that generate
3063 overlong encoded strings. Why am I not surprised?
3065 PH/15 If the first argument of "${if match_address" was not empty, but did not
3066 contain an "@" character, Exim crashed. Now it writes a panic log message
3067 and treats the condition as false.
3069 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3071 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3072 extension "envelope-auth". The code is finished and in agreement with
3073 other implementations, but there is no documentation so far and in fact,
3074 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3075 not changing the active code.
3077 Print executed "if" and "elsif" statements when debugging is used. This
3078 helps a great deal to understand what a filter does.
3080 Document more things not specified clearly in RFC3028. I had all this
3081 sorted out, when out of a sudden new issues came to my mind. Oops."
3083 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3086 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3087 canonical form (as documented). However, after a host name lookup from
3088 the IP address, check_host() was doing a simple string comparison with
3089 addresses acquired from the DNS when checking that the found name did
3090 have the original IP as one of its addresses. Since any found IPv6
3091 addresses are likely to be in abbreviated form, the comparison could
3092 fail. Luckily, there already exists a function for doing the comparison
3093 by converting both addresses to binary, so now that is used instead of
3094 the text comparison.
3096 PH/20 There was another similar case to PH/19, when a complete host name was
3097 given in a host list; looking up its IP address could give an abbreviated
3098 form, whereas the current host's name might or might not be abbreviated.
3099 The same fix has been applied.
3105 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3106 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3109 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3110 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3112 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3114 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3115 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3116 Also, for "Auto-Submitted", treat anything other than "no" as
3117 non-personal, in accordance with RFC 3834. (Previously it treated
3118 anything starting "auto-" as non-personal.)
3120 TF/01 The control=submission/name=... option had a problem with syntax
3121 errors if the name included a slash character. The /name= option
3122 now slurps the rest of the string, so it can include any characters
3123 but it must come last in the list of options (after /sender_retain
3126 PH/05 Some modifications to the interface to the fake nameserver for the new
3134 TK/01 Added the "success_on_redirect" address verification option. See
3135 NewStuff for rationale and an example.
3137 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3139 PH/02 Patch to exigrep to allow it to work on syslog lines.
3141 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3142 fread() to skip over the body file's header line, because in Cygwin the
3143 header line is locked and is inaccessible.
3145 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3146 co-exist for some time) to make it clear that it is the Exim ID that is
3147 referenced, not the Message-ID: header line.
3149 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3150 string_format() function, because snprintf() does not exist on all
3153 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3154 ${stat: expansion item.
3156 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3157 protocol synchronization error", to keep the pedants happy.
3159 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3160 well as for IRIX systems, when gcc is being used. See the host.c source
3163 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3165 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3168 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3169 transport and to the smtp transport in LMTP mode.
3171 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3173 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3174 run a filter in a subprocess. This could lead to confusion in subsequent
3175 lookups in the parent process. There should also be a search_tidyup() at
3176 the end of the subprocess.
3178 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3179 only if the host matched helo_try_verify_hosts, which caused the
3180 verification to occur when the EHLO/HELO command was issued. The ACL just
3181 tested the remembered result. Now, if a previous verification attempt has
3182 not happened, "verify = helo" does it there and then.
3184 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3186 TK/03 Fix log output including CR from clamd.
3188 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3189 exist provoked a memory error which could cause a segfault.
3191 PH/15 Installed PCRE 6.2
3193 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3195 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3196 of the problem. Specifically, suggested +O2 rather than +O1 for the
3199 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3201 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3202 record, the port was not correctly logged when the outgoing_port log
3203 selector was set (it logged the transort's default port).
3205 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3206 fallback_hosts, and "hosts" in the smtp transport.
3208 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3209 host errors such as "Connection refused".
3211 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3212 authentication with radiusclient 0.4.9:
3214 - Error returned from rc_read_config was caught wrongly
3215 - Username/password not passed on to radius server due to wrong length.
3217 The presumption is that some radiusclient API changes for 4.51/PH/17
3218 were not taken care of correctly. The code is still untested by me (my
3219 Linux distribution still has 0.3.2 of radiusclient), but it was
3220 contributed by a Radius user.
3222 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3223 expanding the "port" option of the smtp transport.
3225 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3226 while reading a MIME header. Thanks to Tom Hughes for a patch.
3228 PH/24 Include config.h inside local_scan.h so that configuration settings are
3231 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3232 that in an ACL for RCPT (for example), you can examine exactly what was
3235 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3236 commands, but it was not correctly comparing the address with the actual
3237 client host address. Thus, it would show the EHLO address in Received:
3238 header lines when this was not necessary.
3240 PH/27 Added the % operator to ${eval:}.
3242 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3243 it should be ignoring failures (because with -C, for example, it has lost
3244 privilege). It wasn't ignoring creation failures other than "already
3247 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3250 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3251 because an input line was too long, either on its own, or by virtue of
3252 too many continuations, the temporary file was not being removed, and the
3253 return code was incorrect.
3255 PH/31 Missing "BOOL" in function definition in filtertest.c.
3257 PH/32 Applied Sieve patches from the maintainer.
3259 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3261 PH/33 Added "verify = not_blind".
3263 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3264 Local/Makefile (with some defaults set). These are used in built scripts
3265 such as exicyclog, but they have never been used in the exim_install
3266 script (though there are many overriding facilities there). I have
3267 arranged that the exim_install script now takes note of these two
3270 PH/35 Installed configuration files for Dragonfly.
3272 PH/36 When a locally submitted message by a trusted user did not contain a
3273 From: header, and the sender address was obtained from -f or from an SMTP
3274 MAIL command, and the trusted user did not use -F to supply a sender
3275 name, $originator_name was incorrectly used when constructing a From:
3276 header. Furthermore, $originator_name was used for submission mode
3277 messages from external hosts without From: headers in a similar way,
3278 which is clearly wrong.
3280 PH/37 Added control=suppress_local_fixups.
3282 PH/38 When log_selector = +received_sender was set, and the addition of the
3283 sender made the log line's construction buffer exactly full, or one byte
3284 less than full, an overflow happened when the terminating "\n" was
3287 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3288 when the result of a list match is failure because a DNS lookup failed.
3290 PH/40 RM_COMMAND is now used in the building process.
3292 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3293 the "build-* directories that it finds.
3295 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3296 address in a domain literal was a prefix of an interface address.
3298 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3299 when verifying a sender address, unless rewrite_headers is false.
3301 PH/44 Wrote a long comment about why errors_to addresses are verified as
3302 recipients, not senders.
3304 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3305 the ratelimit ACL was added.
3307 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3309 PH/46 Added extra information about PostgreSQL errors to the error string.
3311 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3312 avoiding the need to install special zones in a real server. This is
3313 backwards compatible; if it can't find the fake resolver, it drops back.
3314 Thus, both old and new test suites can be run.
3316 TF/02 Added util/ratelimit.pl
3318 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3319 clock is set back in time.
3321 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3322 Candler <B.Candler@pobox.com>.
3324 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3325 OK for addresses that are the result of redirections.
3327 PH/48 A number of further additions for the benefit of the new test suite,
3328 including a fake gethostbyname() that interfaces to the fake DNS resolver
3331 TF/06 The fix for widen_domains has also been applied to qualify_single and
3332 search_parents which are the other dnslookup options that can cause
3335 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3338 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3340 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3341 service" if it encountered an MX record with an empty target hostname.
3342 The message is now "an MX or SRV record indicated no SMTP service".
3344 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3345 if the DNS of the sending site is misconfigured. This is quite a
3346 common situation. This change restores the behaviour of treating a
3347 helo verification defer as a failure.
3349 PH/51 If self=fail was set on a router, the bounce message did not include the
3350 actual error message.
3356 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3358 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3359 command itself ended in error, the underlying message about the transport
3360 filter timeout was being overwritten with the pipe command error. Now the
3361 underlying error message should be appended to the second error message.
3363 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3365 PH/02 Reduce the amount of output that "make" produces by default. Full output
3366 can still be requested.
3368 PH/03 The warning log line about a condition test deferring for a "warn" verb
3369 was being output only once per connection, rather than after each
3370 occurrence (because it was using the same function as for successful
3371 "warn" verbs). This seems wrong, so I have changed it.
3373 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3374 it should not have, which might have caused a crash in the right
3375 circumstances, but probably never did.
3377 PH/04 Installed a modified version of Tony Finch's patch to make submission
3378 mode fix the return path as well as the Sender: header line, and to
3379 add a /name= option so that you can make the user's friendly name appear
3382 TF/03 Added the control = fakedefer ACL modifier.
3384 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3385 Mark Lowes for thorough testing.
3387 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3389 TK/03 Merged latest SRS patch from Miles Wilton.
3391 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3392 with the definition in sysexits.h (which is #included earlier).
3393 Fortunately, Exim does not actually use EX_OK. The code used to try to
3394 preserve the sysexits.h value, by assumimg that macro definitions were
3395 scanned for macro replacements. I have been disabused of this notion,
3396 so now the code just undefines EX_OK before #including unistd.h.
3398 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3399 in the smtp transport. When a block could not be written in a single
3400 write() function, the timeout was being re-applied to each part-write.
3401 This seems wrong - if the receiver was accepting one byte at a time it
3402 would take for ever. The timeout is now adjusted when this happens. It
3403 doesn't have to be particularly precise.
3405 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3406 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3408 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3409 without a domain if the check to <postmaster@domain> fails.
3411 SC/01 Eximstats: added -xls and the ability to specify output files
3412 (patch written by Frank Heydlauf).
3414 SC/02 Eximstats: use FileHandles for outputing results.
3416 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3418 SC/04 Eximstats: fixed display of large numbers with -nvr option
3420 SC/05 Eximstats: fixed merging of reports with empty tables.
3422 SC/06 Eximstats: added the -include_original_destination flag
3424 SC/07 Eximstats: removed tabs and trailing whitespace.
3426 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3428 TK/06 MBOX spool code: Add real "From " MBOX separator line
3429 so the .eml file is really in mbox format (even though
3430 most programs do not really care). Patch from Alex Miller.
3432 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3433 The latter is generated from $received_to and is only set if the
3434 message has one envelope recipient. SA can use these headers,
3435 obviously out-of-the-box. Patch from Alex Miller.
3437 PH/08 The ${def test on a variable was returning false if the variable's
3438 value was "0", contrary to what the specification has always said!
3439 The result should be true unless the variable is empty.
3441 PH/09 The syntax error of a character other than { following "${if
3442 def:variable_name" (after optional whitespace) was not being diagnosed.
3443 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3444 accidental colon was present, for example, could give incorrect results.
3446 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3447 result is used (not including appendfile, where other changes are about
3450 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3451 This involved changing a lot of size variables from int to off_t. It
3452 should work with maildirs and everything.
3454 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3455 spamd dying while we are connected to it.
3457 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3460 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3461 function for BDB 4.3.
3463 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3465 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3466 the sources for PCRE have been re-arranged and more files are now
3469 PH/15 The code I had for printing potentially long long variables in PH/11
3470 above was not the best (it lost precision). The length of off_t variables
3471 is now inspected at build time, and an appropriate printing format (%ld
3472 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3473 to be "long long int" or "long int". This is needed for the internal
3474 formatting function string_vformat().
3476 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3477 the configuration file to be ":syslog", then the script "guesses" where
3478 the logs files are, rather than using the compiled in default. In our
3479 case the guess is not the same as the compiled default, so the script
3480 suddenly stopped working when I started to use syslog. The patch checks
3481 to see if log_file_path is "". If so, it attempts to read it from exim
3482 with no configuration file to get the compiled in version, before it
3483 falls back to the previous guessing code."
3485 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3486 implementing BATV in an Exim configuration. See NewStuff for the gory
3489 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3490 Makefile that are specific to HP-UX.
3492 PH/18 If the "use_postmaster" option was set for a recipient callout together
3493 with the "random" option, the postmaster address was used as the MAIL
3494 FROM address for the random test, but not for the subsequent recipient
3495 test. It is now used for both.
3497 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3498 patch removes a few documentation additions to RFC 3028, because the
3499 latest draft now contains them. It adds the new en;ascii-case comparator
3500 and a new error check for 8bit text in MIME parts. Comparator and
3501 require names are now matched exactly. I enabled the subaddress
3502 extension, but it is not well tested yet (read: it works for me)."
3504 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3505 rework some of the code of TK/09 above to avoid the hardwired use of
3506 "%lld" and "long long". Replaced the call to snprintf() with a call to
3509 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3510 records point to non-existent hosts", "retry timeout exceeded", and
3511 "retry time not reached for any host after a long failure period".
3513 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3514 experimental DomainKeys support:
3516 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3517 (2) On an error such as an illegally used "control", the wrong name for
3518 the control was given.
3520 These problems did NOT occur unless DomainKeys support was compiled.
3522 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3524 PH/24 Added ${if match_ip condition.
3526 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3527 databases so that it will be absolutely obvious if a crash occurs in the
3528 DB library. This is a regular occurrence (often caused by mis-matched
3531 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3532 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3533 picked up on a user's system that detects such things. There doesn't seem
3534 to be a gcc warning option for this - only an attribute that has to be
3535 put on the function's prototype. It seems that in Fedora Core 4 they have
3536 set this on a number of new functions. No doubt there will be more in due
3539 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3540 specify a transport. However, if an address that was verified by such a
3541 router was the subject of a callout, Exim crashed because it tried to
3542 read the rcpt_include_affixes from the non-existent transport. Now it
3543 just assumes that the setting of that option is false. This bug was
3544 introduced by 4.51/PH/31.
3546 PH/28 Changed -d+all to exclude +memory, because that information is very
3547 rarely of interest, but it makes the output a lot bigger. People tend to
3548 do -d+all out of habit.
3550 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3551 code in os-type was giving problems when libc.so lives in lib64, like on
3554 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3555 aren't the modern standard, and it seems that some systems' include files
3556 don't always have them. Exim was already checking for some of the newer
3557 ones like T_AAAA, and defining it itself. I've added checks for all the
3558 record types that Exim uses.
3560 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3561 not automatically generating a new one, as it is supposed to. This
3562 prevented TLS from working. If the file did exist, but contained invalid
3563 data, a new version was generated, as expected. It was only the case of a
3564 non-existent file that was broken.
3566 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3567 with a change in libdomainkeys > 0.64.
3569 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3570 from DNS. If the selector record carries the flag, it now has
3571 precedence over the domain-wide flag.
3573 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3575 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3576 the use of a port name that isn't defined in /etc/services) occurred, the
3577 message was deferred as in a normal delivery, and thus remained on the
3578 spool, instead of being failed because of the mua_wrapper setting. This
3579 is now fixed, and I tidied up some of the mua_wrapper messages at the
3582 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3583 the messages in a hash of arrays rather than using individual hashes.
3584 This is a bit cleaner and results in dramatic memory savings, albeit
3585 at a slight CPU cost.
3587 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3588 as requested by Marc Sherman.
3590 SC/10 Eximstats: added histograms for user specified patterns as requested
3593 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3595 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3596 fopen() in the content-scanning modules that did not already have it.
3602 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3603 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3605 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3607 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3609 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3610 to test Sieve filters that use "vacation".
3612 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3613 that changes the way the GnuTLS parameters are stored in the cache file.
3614 The new format can be generated externally. For backward compatibility,
3615 if the data in the cache doesn't make sense, Exim assumes it has read an
3616 old-format file, and it generates new data and writes a new file. This
3617 means that you can't go back to an older release without removing the
3620 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3621 work if there are any delivery delays because "one_time" forces the
3622 parent to be marked "delivered", so its unseen clone is never tried
3623 again. For this reason, Exim now forbids the simultaneous setting of
3626 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3627 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3628 entry if you want to know the details. The fix, however, neglected to
3629 consider the case when local delivery batching is involved. The test for
3630 "previously delivered" was not happening when checking to see if an
3631 address could be batched with a previous (undelivered) one; under
3632 certain circumstances this could lead to multiple deliveries to the same
3635 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3636 in its include files, and this causes problems building Exim.
3638 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3639 header_syntax) but Exim was just ignoring anything given after a slash.
3640 In particular, this caused confusion with an attempt to use "verify =
3641 reverse_host_lookup/defer_ok". An error is now given when options are
3642 supplied for verify items that do not have them. (Maybe reverse_host_
3643 lookup should have a defer_ok option, but that's a different point.)
3645 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3646 defined by RFC 821) to 2048, because there were problems with some AUTH
3647 commands, and RFC 1869 says the size should be increased for extended
3648 SMTP commands that take arguments.
3650 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3653 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3654 "unknown" error; now it says that the functionality isn't in the binary.
3656 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3657 an address' error message when a string expansion fails (syntax or
3658 whatever). Otherwise the password may appear in the log. Following change
3659 PH/42 below, there is no longer a chance of it appearing in a bounce
3662 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3664 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3665 of its addresses was ever tried. (Bugzilla bug #2.)
3667 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3668 the result incorrectly in the debug output. (It correctly added a newline
3669 to what was transported.)
3671 TF/01 Added $received_time.
3673 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3674 commented out examples of how to interface to a virus scanner and to
3675 SpamAssassin. Also added commented examples of av_scanner and
3676 spamd_address settings.
3678 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3679 and controls are allowed in which ACLs. There were a couple of minor
3680 errors. Some of the entries in the conditions table (which is a table of
3681 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3682 negation of where the condition IS allowed.
3684 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3686 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3687 header file does not have a version number, so I've had to invent a new
3688 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3689 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3690 radiusclient), but it was contributed by a Radius user.
3692 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3693 files or directories, for OpenSSL.
3695 PH/19 When an Exim process that is running as root has to create an Exim log
3696 file, it does so in a subprocess that runs as exim:exim so as to get the
3697 ownership right at creation (otherwise, other Exim processes might see
3698 the file with the wrong ownership). There was no test for failure of this
3699 fork() call, which would lead to the process getting stuck as it waited
3700 for a non-existent subprocess. Forks do occasionally fail when resources
3701 run out. I reviewed all the other calls to fork(); they all seem to check
3704 PH/20 When checking for unexpected SMTP input at connect time (before writing
3705 the banner), Exim was not dealing correctly with a non-positive return
3706 from the read() function. If the client had disconnected by this time,
3707 the result was a log entry for a synchronization error with an empty
3708 string after "input=" when read() returned zero. If read() returned -1
3709 (an event I could not check), uninitialized data bytes were printed.
3710 There were reports of junk text (parts of files, etc) appearing after
3713 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3715 PH/22 Added support for macro redefinition, and (re)definition in between
3716 driver and ACL definitions.
3718 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3719 forgetting to use the resulting value; it was using the unexpanded value.
3721 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3722 hadn't been configured. The fix is from Juergen Kreileder, who
3723 understands it better than I do:
3725 "Here's what I see happening with three configured cyrus_sasl
3726 authenticators configured (plain, login, cram-md5):
3728 On startup auth_cyrus_sasl_init() gets called for each of these.
3729 This means three calls to sasl_listmech() without a specified mech_list.
3730 => SASL tests which mechs of all available mechs actually work
3731 => three warnings about OTP not working
3732 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3734 With the patch, sasl_listmech() also gets called three times. But now
3735 SASL's mech_list option is set to the server_mech specified in the the
3736 authenticator. Or in other words, the answer from sasl_listmech()
3737 gets limited to just the mech you're testing for (which is different
3739 => the return list contains just 'plain' or 'login', 'cram-md5' or
3740 nothing depending on the value of ob->server_mech.
3742 I've just tested the patch: Authentication still works fine,
3743 unavailable mechs specified in the exim configuration are still
3744 caught, and the auth.log warnings about OTP are gone."
3746 PH/25 When debugging is enabled, the contents of the command line are added
3747 to the debugging output, even when log_selector=+arguments is not
3750 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3751 answer is "GNU", and only if the return is "GNU/something" is the answer
3754 PH/27 $acl_verify_message is now set immediately after the failure of a
3755 verification in an ACL, and so is available in subsequent modifiers. In
3756 particular, the message can be preserved by coding like this:
3758 warn !verify = sender
3759 set acl_m0 = $acl_verify_message
3761 Previously, $acl_verify_message was set only while expanding "message"
3762 and "log_message" when a very denied access.
3764 PH/28 Modified OS/os.c-Linux with
3766 -#ifndef OS_LOAD_AVERAGE
3767 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3769 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3770 nomenclature these days.)
3772 PH/29 Installed patch from the Sieve maintainer that adds the options
3773 sieve_useraddress and sieve_subaddress to the redirect router.
3775 PH/30 In these circumstances:
3776 . Two addresses routed to the same list of hosts;
3777 . First host does not offer TLS;
3778 . First host accepts first address;
3779 . First host gives temporary error to second address;
3780 . Second host offers TLS and a TLS session is established;
3781 . Second host accepts second address.
3782 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3783 and peerdn, if requested) that were in fact used only for the second
3786 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3787 attention to any local part prefix or suffix that was matched by the
3788 router that accepted the address. It now behaves in the same way as it
3789 does for delivery: the affixes are removed from the local part unless
3790 rcpt_include_affixes is set on the transport.
3792 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3793 timeout during the DATA phase of an incoming message.
3795 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3796 applied a patch sanctioned by the Sieve maintainer.
3798 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3799 the uid or gid is negative. A case of a negative gid caused this to be
3800 noticed. The fix allows for either to be negative.
3802 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3803 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3806 PH/36 The change PH/12 above was broken. Fixed it.
3808 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3809 the grounds that routing the same address twice would always produce the
3810 same answer. This might have been true once, but it is certainly no
3811 longer true now. Routing a child address may depend on the previous
3812 routing that produced that child. Some complicated redirection strategies
3813 went wrong when messages had multiple recipients, and made Exim's
3814 behaviour dependent on the order in which the addresses were given.
3816 I have moved the duplicate checking until after the routing is complete.
3817 Exim scans the addresses that are assigned to local and remote
3818 transports, and removes any duplicates. This means that more work will be
3819 done, as duplicates will always all be routed, but duplicates are
3820 presumably rare, so I don't expect this is of any significance.
3822 For deliveries to pipes, files, and autoreplies, the duplicate checking
3823 still happens during the routing process, since they are not going to be
3826 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3827 It corrects a timeout issue with spamd. This is Ian's comment: "The
3828 background is that sometimes spamd either never reads data from a
3829 connection it has accepted, or it never writes response data. The exiscan
3830 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3831 blindly assumes that writes won't block so it may never time out."
3833 PH/39 Allow G after quota size as well as K and M.
3835 PH/40 The value set for $authenticated_id in an authenticator may not contain
3836 binary zeroes or newlines because the value is written to log lines and
3837 to spool files. There was no check on this. Now the value is run through
3838 the string_printing() function so that such characters are converted to
3839 printable escape sequences.
3841 PH/41 $message_linecount is a new variable that contains the total number of
3842 lines in the message. Compare $body_linecount, which is the count for the
3845 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3846 bounce and delay warning messages, except in certain special cases, which
3849 (a) An SMTP error message from a remote host;
3850 (b) A message specified in a :fail: redirection;
3851 (c) A message specified in a "fail" command in a system filter;
3852 (d) A message specified in a FAIL return from the queryprogram router;
3853 (e) A message specified by the cannot_route_message router option.
3855 In these cases only, Exim does include the error details in bounce and
3856 warning messages. There are also a few cases where bland messages such
3857 as "unrouteable address" or "local delivery error" are given.
3859 PH/43 $value is now also set for the "else" part of a ${run expansion.
3861 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3862 being worked on, but at least Exim now implements the latest version to
3865 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3866 process to complete was treated as a delivery failure, a timeout while
3867 writing the message to the pipe was logged, but erroneously treated as a
3868 successful delivery. Such timeouts include transport filter timeouts. For
3869 consistency with the overall process timeout, these timeouts are now
3870 treated as errors, giving rise to delivery failures by default. However,
3871 there is now a new Boolean option for the pipe transport called
3872 timeout_defer, which, if set TRUE, converts the failures into defers for
3873 both kinds of timeout. A transport filter timeout is now identified in
3876 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3877 systems where "make" and "gmake" are different, calling "gmake" at top
3878 level broke things. I've arranged for the value of $(MAKE) to be passed
3879 from the Makefile to this script so that it can call the same version of
3883 A note about Exim versions 4.44 and 4.50
3884 ----------------------------------------
3886 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3887 changes of various kinds. As a consequence, a big documentation update was
3888 needed. This delayed the release for rather longer than seemed good, especially
3889 in the light of a couple of (minor) security issues. Therefore, the changes
3890 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3891 release. So 4.44 and 4.50 are in effect two different branches that both start
3894 I have left the 4.50 change log unchanged; it contains all the changes since
3895 4.43. The change log for 4.44 is below; many of its items are identical to
3896 those for 4.50. This seems to be the most sensible way to preserve the
3897 historical information.
3903 1. Minor wording change to the doc/README.SIEVE file.
3905 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3906 computation of the current number of files was incorrect.
3908 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3909 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3912 4. Give more explanation in the error message when the command for a transport
3913 filter fails to execute.
3915 5. There are several places where Exim runs a non-Exim command in a
3916 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3917 was being done only for the command run by the queryprogram router. It is
3918 now done for all such subprocesses. The other cases are: ${run, transport
3919 filters, and the commands run by the lmtp and pipe transports.
3921 6. Added CONFIGURE_GROUP build-time option.
3923 7. Some older OS have a limit of 256 on the maximum number of file
3924 descriptors. Exim was using setrlimit() to set 1000 as a large value
3925 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3926 systems. I've change it so that if it can't get 1000, it tries for 256.
3928 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3929 was an oversight, and furthermore, ever since the addition of extra
3930 controls (e.g. 4.43/32), the checks on when to allow different forms of
3931 "control" were broken. There should now be diagnostics for all cases when a
3932 control that does not make sense is encountered.
3934 9. Added the /retain_sender option to "control=submission".
3936 10. $recipients is now available in the predata ACL (oversight).
3938 11. Tidy the search cache before the fork to do a delivery from a message
3939 received from the command line. Otherwise the child will trigger a lookup
3940 failure and thereby defer the delivery if it tries to use (for example) a
3941 cached ldap connection that the parent has called unbind on.
3943 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3944 of $address_data from the recipient verification was clobbered by the
3945 sender verification.
3947 13. The value of address_data from a sender verification is now available in
3948 $sender_address_data in subsequent conditions in the ACL statement.
3950 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3952 15. Added a new option "connect=<time>" to callout options, to set a different
3955 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3956 was its contents. (It was OK if the option was not defined at all.)
3958 17. A "Completed" log line is now written for messages that are removed from
3959 the spool by the -Mrm option.
3961 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3962 information about exactly what failed.
3964 19. Added -dd to debug only the daemon process.
3966 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3967 handles timeouts, both on the server side and network timeouts. Renamed the
3968 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3970 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3971 It is now set to "smtps".
3973 22. $host_address is now set to the target address during the checking of
3974 ignore_target_hosts.
3976 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3977 being passed; this would have caused $sender_host_name to have been used if
3978 matching the list had actually called for a host name (not very likely,
3979 since this list is usually IP addresses). A host name is now passed as
3982 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3983 code that specifies a non-restarting handler (typically sigaction() in
3984 modern systems) in an attempt to fix a rare and obscure crash bug.
3986 25. Narrowed the window for a race in the daemon that could cause it to ignore
3987 SIGCHLD signals. This is not a major problem, because they are used only to
3988 wake it up if nothing else does.
3990 26. A malformed maildirsize file could cause Exim to calculate negative values
3991 for the mailbox size or file count. Odd effects could occur as a result.
3992 The maildirsize information is now recalculated if the size or filecount
3995 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3996 support for a long time. Removed HAVE_SYS_VFS_H.
3998 28. Installed the latest version of exipick from John Jetmore.
4000 29. In an address list, if the pattern was not a regular expression, an empty
4001 subject address (from a bounce message) matched only if the pattern was an
4002 empty string. Non-empty patterns were not even tested. This was the wrong
4003 because it is perfectly reasonable to use an empty address as part of a
4004 database query. An empty address is now tested by patterns that are
4005 lookups. However, all the other forms of pattern expect the subject to
4006 contain a local part and a domain, and therefore, for them, an empty
4007 address still always fails if the pattern is not itself empty.
4009 30. Exim went into a mad DNS loop when attempting to do a callout where the
4010 host was specified on an smtp transport, and looking it up yielded more
4011 than one IP address.
4013 31. Re-factored the code for checking spool and log partition space into a
4014 function that finds that data and another that does the check. The former
4015 is then used to implement four new variables: $spool_space, $log_space,
4016 $spool_inodes, and $log_inodes.
4018 32. The RFC2047 encoding function was originally intended for short strings
4019 such as real names; it was not keeping to the 75-character limit for
4020 encoded words that the RFC imposes. It now respects the limit, and
4021 generates multiple encoded words if necessary. To be on the safe side, I
4022 have increased the buffer size for the ${rfc2047: expansion operator from
4025 33. It is now permitted to omit both strings after an "if" condition; if the
4026 condition is true, the result is "true". As before, when the second string
4027 is omitted, a false condition yields an empty string. This makes it less
4028 cumbersome to write custom ACL and router conditions.
4030 34. Failure to deliver a bounce message always caused it to be frozen, even if
4031 there was an errors_to setting on the router. The errors_to setting is now
4034 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4035 canonical form (fully expanded) before being placed in
4036 $sender_host_address.
4038 36. The table in the code that translates DNS record types into text (T_A to
4039 "A" for instance) was missing entries for NS and CNAME. It is just possible
4040 that this could have caused confusion if both these types were looked up
4041 for the same domain, because the text type is used as part of Exim's
4042 per-process caching. But the chance of anyone hitting this buglet seems
4045 37. The dnsdb lookup has been extended in a number of ways.
4047 (1) There is a new type, "zns", which walks up the domain tree until it
4048 finds some nameserver records. It should be used with care.
4050 (2) There is a new type, "mxh", which is like "mx" except that it returns
4051 just the host names, not the priorities.
4053 (3) It is now possible to give a list of domains (or IP addresses) to be
4054 looked up. The behaviour when one of the lookups defers can be
4055 controlled by a keyword.
4057 (4) It is now possible to specify the separator character for use when
4058 multiple records are returned.
4060 38. The dnslists ACL condition has been extended: it is now possible to supply
4061 a list of IP addresses and/or domains to be looked up in a particular DNS
4064 39. Added log_selector=+queue_time_overall.
4066 40. When running the queue in the test harness, wait just a tad after forking a
4067 delivery process, to get repeatability of debugging output.
4069 41. Include certificate and key file names in error message when GnuTLS fails
4070 to set them up, because the GnuTLS error message doesn't include the name
4071 of the failing file when there is a problem reading it.
4073 42. Allow both -bf and -bF in the same test run.
4075 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4077 44. The "Exiscan patch" is now merged into the mainline Exim source.
4079 45. Sometimes the final signoff response after QUIT could fail to get
4080 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4081 < 0 before doing a fflush(). This bug looks as though it goes back to the
4082 introduction of TLS in release 3.20, but "sometimes" must have been rare
4083 because the tests only now provoked it.
4085 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4086 (this can affect the format of dates).
4088 47. exim_tidydb, when checking for the continued existence of a message for
4089 which it has found a message-specific retry record, was not finding
4090 messages that were in split spool directories. Consequently, it was
4091 deleting retry records that should have stayed in existence.
4093 48. Steve fixed some bugs in eximstats.
4095 49. The SPA authentication driver was not abandoning authentication and moving
4096 on to the next authenticator when an expansion was forced to fail,
4097 contradicting the general specification for all authenticators. Instead it
4098 was generating a temporary error. It now behaves as specified.
4100 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4101 (the order specifies the preference for clients). The order is now AES256,
4102 AES128, 3DES, ARCFOUR128.
4104 51. Small patch to Sieve code - explicitly set From: when generating an
4107 52. Exim crashed if a remote delivery caused a very long error message to be
4108 recorded - for instance if somebody sent an entire SpamAssassin report back
4109 as a large number of 550 error lines. This bug was coincidentally fixed by
4110 increasing the size of one of Exim's internal buffers (big_buffer) that
4111 happened as part of the Exiscan merge. However, to be on the safe side, I
4112 have made the code more robust (and fixed the comments that describe what
4115 53. Now that there can be additional text after "Completed" in log lines (if
4116 the queue_time_overall log selector is set), a one-byte patch to exigrep
4117 was needed to allow it to recognize "Completed" as not the last thing in
4120 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4121 patch that reportedly fixes this has been added. I am not expert enough to
4122 create a test for it. This is what the patch creator wrote:
4124 "I found a little strange behaviour of ldap code when working with
4125 Windows 2003 AD Domain, where users was placed in more than one
4126 Organization Units. When I tried to give exim partial DN, the exit code
4127 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4128 But simultaneously result of request was absolutely normal ldap result,
4129 so I produce this patch..."
4131 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4132 REFERENCE, so I have modified the code to exclude the patch when that macro
4135 55. Some experimental protocols are using DNS PTR records for new purposes. The
4136 keys for these records are domain names, not reversed IP addresses. The
4137 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4138 leaves it alone. Component reversal etc. now happens only for IP addresses.
4141 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4143 57. Double the size of the debug message buffer (to 2048) so that more of very
4144 long debug lines gets shown.
4146 58. The exicyclog utility now does better if the number of log files to keep
4147 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4149 59. Two changes related to the smtp_active_hostname option:
4151 (1) $smtp_active_hostname is now available as a variable.
4152 (2) The default for smtp_banner uses $smtp_active_hostname instead
4153 of $primary_hostname.
4155 60. The host_aton() function is supposed to be passed a string that is known
4156 to be a valid IP address. However, in the case of IPv6 addresses, it was
4157 not checking this. This is a hostage to fortune. Exim now panics and dies
4158 if the condition is not met. A case was found where this could be provoked
4159 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4160 components; fortuitously, this particular loophole had already been fixed
4161 by change 4.50/55 above.
4163 If there are any other similar loopholes, the new check in host_aton()
4164 itself should stop them being exploited. The report I received stated that
4165 data on the command line could provoke the exploit when Exim was running as
4166 exim, but did not say which command line option was involved. All I could
4167 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4168 running as the user.
4171 61. There was a buffer overflow vulnerability in the SPA authentication code
4172 (which came originally from the Samba project). I have added a test to the
4173 spa_base64_to_bits() function which I hope fixes it.
4176 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4177 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4179 63. The daemon start-up calls getloadavg() while still root for those OS that
4180 need the first call to be done as root, but it missed one case: when
4181 deliver_queue_load_max is set with deliver_drop_privilege. This is
4182 necessary for the benefit of the queue runner, because there is no re-exec
4183 when deliver_drop_privilege is set.
4185 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4186 This has been fixed.
4188 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4189 was in use, was not putting the data itself into the right store pool;
4190 consequently, it could be overwritten for a subsequent message in the same
4191 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4194 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4196 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4197 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4198 the function were treating the return as a boolean value, which happened to
4199 work because 0=false and not-0=true, but is not correct code.
4201 68. The host_aton() function was not handling scoped IPv6 addresses (those
4202 with, for example, "%eth0" on the end) correctly.
4204 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4205 negated items (that is, ~something) in unsigned ints. Some compilers
4206 apparently mutter when there is no cast.
4208 70. If an address verification called from an ACL failed, and did not produce a
4209 user-specific message (i.e. there was only a "system" message), nothing was
4210 put in $acl_verify_message. In this situation, it now puts the system
4213 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4214 session; change 4.31/43 added the unwanted input to the log line - except
4215 that it did not do this in the start of session case. It now does.
4217 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4218 This is wrong and can cause the other end to generate a synchronization
4219 error if it is another Exim or anything else that does the synchronization
4220 check. A QUIT command is no longer sent after a timeout.
4222 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4223 during host lookups.
4225 74. The defer_ok option of callout verification was not working if it was used
4226 when verifying addresses in header lines, that is, for this case:
4228 verify = header_sender/callout=defer_ok
4230 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4231 those file descriptors could be used for SMTP connections. If anything
4232 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4233 could be sent to the SMTP client, causing chaos. The daemon now opens
4234 stdin, stdout, and stderr to /dev/null when it puts itself into the
4237 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4238 log by default. The user can override this with suitable Perl magic.
4240 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4241 the log message when discard triggers, was not working for the DATA ACL or
4242 for the non-SMTP ACL.
4244 78. Error message wording change in sieve.c.
4246 79. If smtp_accept_max_per_host was set, the number of connections could be
4247 restricted to fewer than expected, because the daemon was trying to set up
4248 a new connection before checking whether the processes handling previous
4249 connections had finished. The check for completed processes is now done
4250 earlier. On busy systems, this bug wouldn't be noticed because something
4251 else would have woken the daemon, and it would have reaped the completed
4254 80. If a message was submitted locally by a user whose login name contained one
4255 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4256 It caused a spool format error. I have fixed the spool reading code. A
4257 related problem was that the "from" clause in the Received: line became
4258 illegal because of the space(s). It is now covered by ${quote_local_part.
4260 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4263 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4265 83. Patch from Sieve maintainer to fix unterminated string problem in
4266 "vacation" handling.
4268 84. Some minor changes to the Linux configuration files to help with other
4269 OS variants using glibc.
4271 85. One more patch for Sieve to update vacation handling to latest spec.
4274 ----------------------------------------------------
4275 See the note above about the 4.44 and 4.50 releases.
4276 ----------------------------------------------------
4282 1. Change 4.43/35 introduced a bug that caused file counts to be
4283 incorrectly computed when quota_filecount was set in an appendfile
4286 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4287 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4290 3. Give more explanation in the error message when the command for a transport
4291 filter fails to execute.
4293 4. There are several places where Exim runs a non-Exim command in a
4294 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4295 was being done only for the command run by the queryprogram router. It is
4296 now done for all such subprocesses. The other cases are: ${run, transport
4297 filters, and the commands run by the lmtp and pipe transports.
4299 5. Some older OS have a limit of 256 on the maximum number of file
4300 descriptors. Exim was using setrlimit() to set 1000 as a large value
4301 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4302 systems. I've change it so that if it can't get 1000, it tries for 256.
4304 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4305 was an oversight, and furthermore, ever since the addition of extra
4306 controls (e.g. 4.43/32), the checks on when to allow different forms of
4307 "control" were broken. There should now be diagnostics for all cases when a
4308 control that does not make sense is encountered.
4310 7. $recipients is now available in the predata ACL (oversight).
4312 8. Tidy the search cache before the fork to do a delivery from a message
4313 received from the command line. Otherwise the child will trigger a lookup
4314 failure and thereby defer the delivery if it tries to use (for example) a
4315 cached ldap connection that the parent has called unbind on.
4317 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4318 of $address_data from the recipient verification was clobbered by the
4319 sender verification.
4321 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4322 was its contents. (It was OK if the option was not defined at all.)
4324 11. A "Completed" log line is now written for messages that are removed from
4325 the spool by the -Mrm option.
4327 12. $host_address is now set to the target address during the checking of
4328 ignore_target_hosts.
4330 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4331 being passed; this would have caused $sender_host_name to have been used if
4332 matching the list had actually called for a host name (not very likely,
4333 since this list is usually IP addresses). A host name is now passed as
4336 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4337 code that specifies a non-restarting handler (typically sigaction() in
4338 modern systems) in an attempt to fix a rare and obscure crash bug.
4340 15. Narrowed the window for a race in the daemon that could cause it to ignore
4341 SIGCHLD signals. This is not a major problem, because they are used only to
4342 wake it up if nothing else does.
4344 16. A malformed maildirsize file could cause Exim to calculate negative values
4345 for the mailbox size or file count. Odd effects could occur as a result.
4346 The maildirsize information is now recalculated if the size or filecount
4349 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4350 support for a long time. Removed HAVE_SYS_VFS_H.
4352 18. Updated exipick to current release from John Jetmore.
4354 19. Allow an empty sender to be matched against a lookup in an address list.
4355 Previously the only cases considered were a regular expression, or an
4358 20. Exim went into a mad DNS lookup loop when doing a callout where the
4359 host was specified on the transport, if the DNS lookup yielded more than
4362 21. The RFC2047 encoding function was originally intended for short strings
4363 such as real names; it was not keeping to the 75-character limit for
4364 encoded words that the RFC imposes. It now respects the limit, and
4365 generates multiple encoded words if necessary. To be on the safe side, I
4366 have increased the buffer size for the ${rfc2047: expansion operator from
4369 22. Failure to deliver a bounce message always caused it to be frozen, even if
4370 there was an errors_to setting on the router. The errors_to setting is now
4373 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4374 canonical form (fully expanded) before being placed in
4375 $sender_host_address.
4377 24. Updated eximstats to version 1.33
4379 25. Include certificate and key file names in error message when GnuTLS fails
4380 to set them up, because the GnuTLS error message doesn't include the name
4381 of the failing file when there is a problem reading it.
4383 26. Expand error message when OpenSSL has problems setting up cert/key files.
4386 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4387 (this can affect the format of dates).
4389 28. exim_tidydb, when checking for the continued existence of a message for
4390 which it has found a message-specific retry record, was not finding
4391 messages that were in split spool directories. Consequently, it was
4392 deleting retry records that should have stayed in existence.
4394 29. eximstats updated to version 1.35
4395 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4396 1.35 - bugfix such that pie charts by volume are generated correctly
4398 30. The SPA authentication driver was not abandoning authentication and moving
4399 on to the next authenticator when an expansion was forced to fail,
4400 contradicting the general specification for all authenticators. Instead it
4401 was generating a temporary error. It now behaves as specified.
4403 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4404 (the order specifies the preference for clients). The order is now AES256,
4405 AES128, 3DES, ARCFOUR128.
4407 31. Small patch to Sieve code - explicitly set From: when generating an
4410 32. Exim crashed if a remote delivery caused a very long error message to be
4411 recorded - for instance if somebody sent an entire SpamAssassin report back
4412 as a large number of 550 error lines. This bug was coincidentally fixed by
4413 increasing the size of one of Exim's internal buffers (big_buffer) that
4414 happened as part of the Exiscan merge. However, to be on the safe side, I
4415 have made the code more robust (and fixed the comments that describe what
4418 33. Some experimental protocols are using DNS PTR records for new purposes. The
4419 keys for these records are domain names, not reversed IP addresses. The
4420 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4421 leaves it alone. Component reversal etc. now happens only for IP addresses.
4424 34. The host_aton() function is supposed to be passed a string that is known
4425 to be a valid IP address. However, in the case of IPv6 addresses, it was
4426 not checking this. This is a hostage to fortune. Exim now panics and dies
4427 if the condition is not met. A case was found where this could be provoked
4428 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4429 components; fortuitously, this particular loophole had already been fixed
4430 by change 4.50/55 or 4.44/33 above.
4432 If there are any other similar loopholes, the new check in host_aton()
4433 itself should stop them being exploited. The report I received stated that
4434 data on the command line could provoke the exploit when Exim was running as
4435 exim, but did not say which command line option was involved. All I could
4436 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4437 running as the user.
4440 35. There was a buffer overflow vulnerability in the SPA authentication code
4441 (which came originally from the Samba project). I have added a test to the
4442 spa_base64_to_bits() function which I hope fixes it.
4445 36. The daemon start-up calls getloadavg() while still root for those OS that
4446 need the first call to be done as root, but it missed one case: when
4447 deliver_queue_load_max is set with deliver_drop_privilege. This is
4448 necessary for the benefit of the queue runner, because there is no re-exec
4449 when deliver_drop_privilege is set.
4451 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4452 was in use, was not putting the data itself into the right store pool;
4453 consequently, it could be overwritten for a subsequent message in the same
4454 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4457 38. Sometimes the final signoff response after QUIT could fail to get
4458 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4459 < 0 before doing a fflush(). This bug looks as though it goes back to the
4460 introduction of TLS in release 3.20, but "sometimes" must have been rare
4461 because the tests only now provoked it.
4467 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4468 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4469 Now it returns an integer. A number of calls were still expecting a T/F
4470 return. Fortuitously, in all cases, the tests worked in OK situations,
4471 which is the norm. However, things would have gone wrong on any write
4472 failures on the smtp file descriptor. This function is used when sending
4473 messages over SMTP and also when doing verify callouts.
4475 2. When Exim is called to do synchronous delivery of a locally submitted
4476 message (the -odf or -odi options), it no longer closes stderr before doing
4479 3. Implemented the mua_wrapper option.
4481 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4483 5. Implemented the functions header_remove(), header_testname(),
4484 header_add_at_position(), and receive_remove_recipient(), and exported them
4487 6. If an ACL "warn" statement specified the addition of headers, Exim already
4488 inserted X-ACL-Warn: at the start if there was no header name. However, it
4489 was not making this test for the second and subsequent header lines if
4490 there were newlines in the string. This meant that an invalid header could
4491 be inserted if Exim was badly configured.
4493 7. Allow an ACL "warn" statement to add header lines at the start or after all
4494 the Received: headers, as well as at the end.
4496 8. Added the rcpt_4xx retry error code.
4498 9. Added postmaster_mailfrom=xxx to callout verification option.
4500 10. Added mailfrom=xxxx to the callout verification option, for verify=
4503 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4504 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4506 12. Inserted some casts to stop certain compilers warning when using pointer
4507 differences as field lengths or precisions in printf-type calls (mostly
4508 affecting debugging statements).
4510 13. Added optional readline() support for -be (dynamically loaded).
4512 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4513 same clock tick as a message's arrival, so that its received time was the
4514 same as the "first fail" time on the retry record, and that message
4515 remained on the queue past the ultimate address timeout, every queue runner
4516 would try a delivery (because it was past the ultimate address timeout) but
4517 after another failure, the ultimate address timeout, which should have then
4518 bounced the address, did not kick in. This was a "< instead of <=" error;
4519 in most cases the first failure would have been in the next clock tick
4520 after the received time, and all would be well.
4522 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4523 being recognized when the domain list was tested by the match_domain
4524 condition in an expansion string.
4526 16. Added the ${str2b64: operator.
4528 17. Exim was always calling setrlimit() to set a large limit for the number of
4529 processes, without checking whether the existing limit was already
4530 adequate. (It did check for the limit on file descriptors.) Furthermore,
4531 errors from getrlimit() and setrlimit() were being ignored. Now they are
4532 logged to the main and panic logs, but Exim does carry on, to try to do its
4533 job under whatever limits there are.
4535 18. Imported PCRE 5.0.
4537 19. Trivial typo in log message " temporarily refused connection" (the leading
4540 20. If the log selector return_path_on_delivery was set and an address was
4541 redirected to /dev/null, the delivery process crashed because it assumed
4542 that a return path would always be set for a "successful" delivery. In this
4543 case, the whole delivery is bypassed as an optimization, and therefore no
4546 21. Internal re-arrangement: the function for sending a challenge and reading
4547 a response while authentication was assuming a zero-terminated challenge
4548 string. It's now changed to take a pointer and a length, to allow for
4549 binary data in such strings.
4551 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4553 23. Exim was not respecting finduser_retries when seeking the login of the
4554 uid under which it was called; it was always trying 10 times. (The default
4555 setting of finduser_retries is zero.) Also, it was sleeping after the final
4556 failure, which is pointless.
4558 24. Implemented tls_on_connect_ports.
4560 25. Implemented acl_smtp_predata.
4562 26. If the domain in control=submission is set empty, Exim assumes that the
4563 authenticated id is a complete email address when it generates From: or
4564 Sender: header lines.
4566 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4567 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4568 chown and chgrp in /bin and hostname in /usr/bin.
4570 28. Exim was keeping the "process log" file open after each use, just as it
4571 does for the main log. This opens the possibility of it remaining open for
4572 long periods when the USR1 signal hits a daemon. Occasional processlog
4573 errors were reported, that could have been caused by this. Anyway, it seems
4574 much more sensible not to leave this file open at all, so that is what now
4577 29. The long-running daemon process does not normally write to the log once it
4578 has entered its main loop, and it closes the log before doing so. This is
4579 so that log files can straightforwardly be renamed and moved. However,
4580 there are a couple of unusual error situations where the daemon does write
4581 log entries, and I had neglected to close the log afterwards.
4583 30. The text of an SMTP error response that was received during a remote
4584 delivery was being truncated at 512 bytes. This is too short for some of
4585 the long messages that one sometimes sees. I've increased the limit to
4588 31. It is now possible to make retry rules that apply only when a message has a
4589 specific sender, in particular, an empty sender.
4591 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4592 it possible to be selective about when SMTP synchronization is enforced.
4594 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4596 32. Implemented hosts_connection_nolog.
4598 33. Added an ACL for QUIT.
4600 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4603 35. Added mailbox_size and mailbox_filecount to appendfile.
4605 36. Added control = no_multiline_responses to ACLs.
4607 37. There was a bug in the logic of the code that waits for the clock to tick
4608 in the case where the clock went backwards by a substantial amount such
4609 that the microsecond fraction of "now" was more than the microsecond
4610 fraction of "then" (but the whole seconds number was less).
4612 38. Added support for the libradius Radius client library this is found on
4613 FreeBSD (previously only the radiusclient library was supported).
4619 1. When certain lookups returned multiple values in the form name=value, the
4620 quoting of the values was not always being done properly. Specifically:
4621 (a) If the value started with a double quote, but contained no whitespace,
4623 (b) If the value contained whitespace other than a space character (i.e.
4624 tabs or newlines or carriage returns) it was not quoted.
4625 This fix has been applied to the mysql and pgsql lookups by writing a
4626 separate quoting function and calling it from the lookup code. The fix
4627 should probably also be applied to nisplus, ibase and oracle lookups, but
4628 since I cannot test any of those, I have not disturbed their existing code.
4630 2. A hit in the callout cache for a specific address caused a log line with no
4631 reason for rejecting RCPT. Now it says "Previous (cached) callout
4632 verification failure".
4634 3. There was an off-by-one bug in the queryprogram router. An over-long
4635 return line was truncated at 256 instead of 255 characters, thereby
4636 overflowing its buffer with the terminating zero. As well as fixing this, I
4637 have increased the buffer size to 1024 (and made a note to document this).
4639 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4640 when Exim is waiting for an SMTP response from a remote server, Exim
4641 restarts its select() call on the socket, thereby resetting its timeout.
4642 This is not a problem when such interrupts are rare. Somebody set up a cron
4643 job to run exiwhat every 2 minutes, which is less than the normal select()
4644 timeout (5 or 10 minutes). This meant that the select() timeout never
4645 kicked in because it was always reset. I have fixed this by comparing the
4646 time when an interrupt arrives with the time at the start of the first call
4647 to select(). If more time than the timeout has elapsed, the interrupt is
4648 treated as a timeout.
4650 5. Some internal re-factoring in preparation for the addition of Sieve
4651 extensions (by MH). In particular, the "personal" test is moved to a
4652 separate function, and given an option for scanning Cc: and Bcc: (which is
4653 not set for Exim filters).
4655 6. When Exim created an email address using the login of the caller as the
4656 local part (e.g. when creating a From: or Sender: header line), it was not
4657 quoting the local part when it contained special characters such as @.
4659 7. Installed new OpenBSD configuration files.
4661 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4662 try to make them clearer.
4664 9. Callout options, other than the timeout value, were being ignored when
4665 verifying sender addresses in header lines. For example, when using
4667 verify = header_sender/callout=no_cache
4669 the cache was (incorrectly) being used.
4671 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4672 only the Cygwin environment.
4674 11. When return_path_on_delivery was set as a log selector, if different remote
4675 addresses in the same message used different return paths and parallel
4676 remote delivery occurred, the wrong values would sometimes be logged.
4677 (Whenever a remote delivery process finished, the return path value from
4678 the most recently started remote delivery process was logged.)
4680 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4681 lines when AUTH and/or TLS are in use. This is the "received protocol"
4682 field. Exim used to use "asmtp" for authenticated SMTP, without any
4683 indication (in the protocol name) for TLS use. Now it follows the RFC and
4684 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4685 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4686 names appear in log lines as well as in Received: header lines.
4688 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4689 extensions, and comparison tests, and to fix some bugs.
4691 14. Changes to the "personal" filter test:
4693 (1) The test was buggy in that it was just doing the equivalent of
4694 "contains" tests on header lines. For example, if a user's address was
4695 anne@some.where, the "personal" test would incorrectly be true for
4697 To: susanne@some.where
4699 This test is now done by extracting each address from the header in turn,
4700 and checking the entire address. Other tests that are part of "personal"
4701 are now done using regular expressions (for example, to check local parts
4702 of addresses in From: header lines).
4704 (2) The list of non-personal local parts in From: addresses has been
4705 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4706 taken from the Sieve specification recommendations.
4708 (3) If the message contains any header line starting with "List-" it is
4709 treated as non-personal.
4711 (4) The test for "circular" in the Subject: header line has been removed
4712 because it now seems ill-conceived.
4714 15. Minor typos in src/EDITME comments corrected.
4716 16. Installed latest exipick from John Jetmore.
4718 17. If headers_add on a router specified a text string that was too long for
4719 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4720 of string_sprintf() is now avoided.
4722 18. $message_body_size was not set (it was always zero) when running the DATA
4723 ACL and the local_scan() function.
4725 19. For the "mail" command in an Exim filter, no default was being set for
4726 the once_repeat time, causing a random time value to be used if "once" was
4727 specified. (If the value happened to be <= 0, no repeat happened.) The
4728 default is now 0s, meaning "never repeat". The "vacation" command was OK
4729 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4730 (I found it when inspecting the code).
4732 20. There is now an overall timeout for performing a callout verification. It
4733 defaults to 4 times the callout timeout, which applies to individual SMTP
4734 commands during the callout. The overall timeout applies when there is more
4735 than one host that can be tried. The timeout is checked before trying the
4736 next host. This prevents very long delays if there are a large number of
4737 hosts and all are timing out (e.g. when the network connections are timing
4738 out). The value of the overall timeout can be changed by specifying an
4739 additional sub-option for "callout", called "maxwait". For example:
4741 verify = sender/callout=5s,maxwait=20s
4743 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4744 to the end before writing, but this should make it even safer).
4746 22. Exim was forgetting that it had advertised PIPELINING for the second and
4747 subsequent messages on an SMTP connection. It was also not resetting its
4748 memory on STARTTLS and an internal HELO.
4750 23. When Exim logs an SMTP synchronization error within a session, it now
4751 records whether PIPELINING has been advertised or not.
4753 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4754 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4755 rather than long int.
4757 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4759 26. Added the never_mail option to autoreply.
4765 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4766 crash if the getsockname() call failed; this can happen if a connection is
4767 closed very soon after it is established. The problem was simply in the
4768 order in which certain operations were done, causing Exim to try to write
4769 to the SMTP stream before it had set up the file descriptor. The bug has
4770 been fixed by making things happen in the correct order.
4776 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4777 before the connection was closed, thus losing the rejection response.
4779 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4780 some early Solaris releases, but causes trouble in current releases where
4781 socklen_t is defined.
4783 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4786 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4789 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4790 that releasing the top of it at the end releases what was used for sub-
4791 expansions (unless the block got too big). However, discard this block if
4792 the first thing is a variable or header, so that we can use its block when
4793 it is dynamic (useful for very large $message_headers, for example).
4795 6. Lookups now cache *every* query, not just the most recent. A new, separate
4796 store pool is used for this. It can be recovered when all lookup caches are
4797 flushed. Lookups now release memory at the end of their result strings.
4798 This has involved some general refactoring of the lookup sources.
4800 7. Some code has been added to the store_xxx() functions to reduce the amount
4801 of flapping under certain conditions.
4803 8. log_incoming_interface used to affect only the <= reception log lines. Now
4804 it causes the local interface and port to be added to several more SMTP log
4805 lines, for example "SMTP connection from", and rejection lines.
4807 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4809 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4811 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4812 could be overwritten at the end of the current message (or the start of a
4813 new message if it was set in a HELO ACL). The value is now preserved for
4814 the duration of the SMTP connection.
4816 12. If a transport had a headers_rewrite setting, and a matching header line
4817 contained an unqualified address, that address was qualified, even if it
4818 did not match any rewriting rules. The underlying bug was that the values
4819 of the flags that permit the existence of unqualified sender and recipient
4820 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4821 non-local messages, and by -bnq for local messages) were not being
4822 preserved with the message after it was received.
4824 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4825 "next input=" as part of the text comprising the host identity instead of
4826 the correct text. The code was using the same buffer for two different
4827 strings. However, depending on which order the printing function evaluated
4828 its arguments, the bug did not always show up. Under Linux, for example, my
4829 test suite worked just fine.
4831 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4832 doesn't work with some older versions of Perl. It has been changed to "my",
4833 which in any case is probably the better facility to use.
4835 15. A really picky compiler found some instances of statements for creating
4836 error messages that either had too many or two few arguments for the format
4839 16. The size of the buffer for calls to the DNS resolver has been increased
4840 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4841 for addresses that have a lot of PTR records. This alleviates a problem; it
4842 does not fully solve it.
4844 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4845 buffer now truncates the list and logs the incident, which is the same
4846 action as happens when Exim is looking up a host name and its aliases.
4847 Previously in this situation something unpredictable would happen;
4848 sometimes it was "internal error: store_reset failed".
4850 18. If a server dropped the connection unexpectedly when an Exim client was
4851 using GnuTLS and trying to read a response, the client delivery process
4852 crashed while trying to generate an error log message.
4854 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4855 string, for example:
4857 warn message = H1: something\nH2: something
4859 the text was added as a single header line from Exim's point of view
4860 though it ended up OK in the delivered message. However, searching for the
4861 second and subsequent header lines using $h_h2: did not work. This has been
4862 fixed. Similarly, if a system filter added multiple headers in this way,
4863 the routers could not see them.
4865 20. Expanded the error message when iplsearch is called with an invalid key to
4866 suggest using net-iplsearch in a host list.
4868 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4869 ACLs are no longer actually imposed (and a message to that effect is
4872 22. If a "gecos" field in a passwd entry contained escaped characters, in
4873 particular, if it contained a \" sequence, Exim got it wrong when building
4874 a From: or a Sender: header from that name. A second bug also caused
4875 incorrect handling when an unquoted " was present following a character
4876 that needed quoting.
4878 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4879 was not being matched caselessly.
4881 24. Arranged for all hyphens in the exim.8 source to be escaped with
4884 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4885 properly. Recipient callout cache records were still being keyed to include
4886 the sender, even when use_sender was set false. This led to far more
4887 callouts that were necessary. The sender is no longer included in the key
4888 when use_sender is false.
4890 26. Added "control = submission" modifier to ACLs.
4892 27. Added the ${base62d: operator to decode base 62 numbers.
4894 28. dnsdb lookups can now access SRV records.
4896 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4897 the configuration file.
4899 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4900 (-v) mode. This makes the output for a verbose queue run more intelligible.
4902 31. Added a use_postmaster feature to recipient callouts.
4904 32. Added the $body_zerocount variable, containing the number of binary zero
4905 bytes in the message body.
4907 33. The time of last modification of the "new" subdirectory is now used as the
4908 "mailbox time last read" when there is a quota error for a maildir
4911 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4913 35. Added +ignore_unknown as a special item in host lists.
4915 36. Code for decoding IPv6 addresses in host lists is now included, even if
4916 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4917 address was recognized as an IP address, but was then not correctly decoded
4918 into binary, causing unexpected and incorrect effects when compared with
4925 1. Very minor rewording of debugging text in manualroute to say "list of
4926 hosts" instead of "hostlist".
4928 2. If verify=header_syntax was set, and a header line with an unqualified
4929 address (no domain) and a large number of spaces between the end of the
4930 name and the colon was received, the reception process suffered a buffer
4931 overflow, and (when I tested it) crashed. This was caused by some obsolete
4932 code that should have been removed. The fix is to remove it!
4934 3. When running in the test harness, delay a bit after writing a bounce
4935 message to get a bit more predictability in the log output.
4937 4. Added a call to search_tidyup() just before forking a reception process. In
4938 theory, someone could use a lookup in the expansion of smtp_accept_max_
4939 per_host which, without the tidyup, could leave open a database connection.
4941 5. Added the variables $recipient_data and $sender_data which get set from a
4942 lookup success in an ACL "recipients" or "senders" condition, or a router
4943 "senders" option, similar to $domain_data and $local_part_data.
4945 6. Moved the writing of debug_print from before to after the "senders" test
4948 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4949 problems for message scanning, either using a data ACL, or using
4950 local_scan() because the Received: header was not generated till after they
4951 were called (in order to set the time as the time of reception completion).
4952 I have revised the way this works. The header is now generated after the
4953 body is received, but before the ACL or local_scan() are called. After they
4954 are run, the timestamp in the header is updated.
4960 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4961 before starting a queue runner without re-exec. This happened only when
4962 deliver_drop_privilege was set or when the Exim user was set to root. The
4963 effect of the bug was that timeouts during subsequent deliveries caused
4964 crashes instead of being properly handled. The handler is now left at its
4965 default (and expected) setting.
4967 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4968 message, again when deliver_drop_privilege is set or Exim is run as root.
4969 The bug described in (1) was not present in this case, but the tidying up
4970 of the other signals was missing. I have made the two cases consistent.
4972 3. The ignore_target_hosts setting on a manualroute router was being ignored
4973 for hosts that were looked up using the /MX notation.
4975 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4978 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4979 operated on the sender address. After changing the $sender_address to <>
4980 for the sender address verify, Exim was re-instated it as the original
4981 (before rewriting) address, but remembering that it had rewritten it, so it
4982 wasn't rewriting it again. This bug also had the effect of breaking the
4983 sender address verification caching when the sender address was rewritten.
4985 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4986 This has been changed so that if the ip literal address matches
4987 ignore_target_hosts, the router declines.
4989 7. Added expansion conditions match_domain, match_address, and match_local_
4990 part (NOT match_host).
4992 8. The placeholder for the Received: header didn't have a length field set.
4994 9. Added code to Exim itself and to exim_lock to test for a specific race
4995 condition that could lead to file corruption when using MBX delivery. The
4996 issue is with the lockfile that is created in /tmp. If this file is removed
4997 after a process has opened it but before that process has acquired a lock,
4998 there is the potential for a second process to recreate the file and also
4999 acquire a lock. This could lead to two Exim processes writing to the file
5000 at the same time. The added code performs the same test as UW imapd; it
5001 checks after acquiring the lock that its file descriptor still refers to
5002 the same named file.
5004 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5005 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5008 11. Added the smtp_active_hostname option. If used, this will typically be made
5009 to depend on the incoming interface address. Because $interface_address is
5010 not set up until the daemon has forked a reception process, error responses
5011 that can happen earlier (such as "too many connections") no longer contain
5014 12. If an expansion in a condition on a "warn" statement fails because a lookup
5015 defers, the "warn" statement is abandoned, and the next ACL statement is
5016 processed. Previously this caused the whole ACL to be aborted.
5018 13. Added the iplsearch lookup type.
5020 14. Added ident_timeout as a log selector.
5022 15. Added tls_certificate_verified as a log selector.
5024 16. Added a global option tls_require_ciphers (compare the smtp transport
5025 option of the same name). This controls incoming TLS connections.
5027 17. I finally figured out how to make tls_require_ciphers do a similar thing
5028 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5029 before starting the TLS session.
5031 18. Tabs are now shown as \t in -bP output.
5033 19. If the log selector return_path_on_delivery was set, Exim crashed when
5034 bouncing a message because it had too many Received: header lines.
5036 20. If two routers both had headers_remove settings, and the first one included
5037 a superfluous trailing colon, the final name in the first list and the
5038 first name in the second list were incorrectly joined into one item (with a
5039 colon in the middle).
5045 1. Added -C and -D options to the exinext utility, mainly to make it easier
5046 to include in the automated testing, but these could be helpful when
5047 multiple configurations are in use.
5049 2. The exinext utility was not formatting the output nicely when there was
5050 an alternate port involved in the retry record key, nor when there was a
5051 message id as well (for retries that were specific to a specific message
5052 and a specific host). It was also confused by IPv6 addresses, because of
5053 the additional colons they contain. I have fixed the IPv4 problem, and
5054 patched it up to do a reasonable job for IPv6.
5056 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5057 delivery, the log line now contains "pipelined" if PIPELINING was used.
5059 4. An SMTP transport process used to panic and die if the bind() call to set
5060 an explicit outgoing interface failed. This has been changed; it is now
5061 treated in the same way as a connect() failure.
5063 5. A reference to $sender_host_name in the part of a conditional expansion
5064 that was being skipped was still causing a DNS lookup. This no longer
5067 6. The def: expansion condition was not recognizing references to header lines
5068 that used bh_ and bheader_.
5070 7. Added the _cache feature to named lists.
5072 8. The code for checking quota_filecount in the appendfile transport was
5073 allowing one more file than it should have been.
5075 9. For compatibility with Sendmail, the command line option
5083 and sets the incoming protocol and host name (for trusted callers). The
5084 host name and its colon can be omitted when only the protocol is to be set.
5085 Note the Exim already has two private options, -pd and -ps, that refer to
5086 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5087 "s", but I don't think that's a major issue.
5089 10. A number of refactoring changes to the code, none of which should affect
5092 (a) The number of logging options was getting close to filling up the
5093 32-bit word that was used as a bit map. I have split them into two classes:
5094 those that are passed in the argument to log_write(), and those that are
5095 only ever tested independently outside of that function. These are now in
5096 separate 32-bit words, so there is plenty of room for expansion again.
5097 There is no change in the user interface or the logging behaviour.
5099 (b) When building, for example, log lines, the code previously used a
5100 macro that called string_cat() twice, in order to add two strings. This is
5101 not really sufficiently general. Furthermore, there was one instance where
5102 it was actually wrong because one of the argument was used twice, and in
5103 one call a function was used. (As it happened, calling the function twice
5104 did not affect the overall behaviour.) The macro has been replaced by a
5105 function that can join an arbitrary number of extra strings onto a growing
5108 (c) The code for expansion conditions now uses a table and a binary chop
5109 instead of a serial search (which was left over from when there were very
5110 few conditions). Also, it now recognizes conditions like "pam" even when
5111 the relevant support is not compiled in: a suitably worded error message is
5112 given if an attempt is made to use such a condition.
5114 11. Added ${time_interval:xxxxx}.
5116 12. A bug was causing one of the ddress fields not to be passed back correctly
5117 from remote delivery subprocesses. The field in question was not being
5118 subsequently used, so this caused to problems in practice.
5120 13. Added new log selectors queue_time and deliver_time.
5122 14. Might have fixed a bug in maildirsizefile handling that threw up
5123 "unexpected character" debug warnings, and recalculated the data
5124 unnecessarily. In any case, I expanded the warning message to give more
5127 15. Added the message "Restricted characters in address" to the statements in
5128 the default ACL that block characters like @ and % in local parts.
5130 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5131 Three changes have been made:
5133 (a) There was a serious bug; a negative response to MAIL caused the whole
5134 recipient domain to be cached as invalid, thereby blocking all messages
5135 to all local parts at the same domain, from all senders. This bug has
5136 been fixed. The domain is no longer cached after a negative response to
5137 MAIL if the sender used is not empty.
5139 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5142 (c) A new callout option, "use_sender" has been added for people who want
5143 the modified behaviour.
5149 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5152 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5153 indeed breaks things for older releases.
5155 3. Added additional logging to the case where there is a problem reading data
5156 from a filter that is running in a subprocess using a pipe, in order to
5157 try to track down a specific problem.
5159 4. Testing facility fudge: when running in the test harness and attempting
5160 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5161 getting "No route to host". Convert this to a timeout.
5163 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5166 6. Some OS don't have socklen_t but use size_t instead. This affects the
5167 fifth argument of getsockopt() amongst other things. This is now
5168 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5169 can be set for individual OS. I have set it for SunOS5, OSF1, and
5170 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5171 some earlier ones do not.
5173 7. Change 4.30/15 was not doing the test caselessly.
5175 8. The standard form for an IPv6 address literal was being rejected by address
5176 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5177 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5178 this, as well as the form without the "IPv6" on the front (but only when
5179 address literals are enabled, of course).
5181 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5183 10. Exim crashed if a message with an empty sender address specified by -f
5184 encountered a router with an errors_to setting. This could be provoked only
5185 by a command such as
5189 where an empty string was supplied; "<>" did not hit this bug.
5191 11. Installed PCRE release 4.5.
5193 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5194 remained set. It is now erased.
5196 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5197 times from message ids (which are base 36 rather than the normal 62).
5199 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5200 were being counted as actual protocol errors, and logged if the log
5201 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5202 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5203 and DATA following a set of rejected RCPTs do not count as protocol errors.
5204 In other words, Exim assumes they were pipelined, though this may not
5205 actually be the case. Of course, in all cases the client gets an
5206 appropriate error code.
5208 15. If a lookup fails in an ACL condition, a message about the failure may
5209 be available; it is used if testing the ACL cannot continue, because most
5210 such messages specify what the cause of the deferral is. However, some
5211 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5212 that caused an old message to be retained and used if a later statement
5213 caused a defer, replacing the real cause of the deferral.
5215 16. If an IP address had so many PTR records that the DNS lookup buffer
5216 was not large enough to hold them, Exim could crash while trying to process
5217 the truncated data. It now detects and logs this case.
5219 17. Further to 4.21/58, another change has been made: if (and only if) the
5220 first line of a message (the first header line) ends with CRLF, a bare LF
5221 in a subsequent header line has a space inserted after it, so as not to
5222 terminate the header.
5224 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5225 unnecessarily, used atoi() instead of strtol(), and didn't check the
5226 termination when getting file sizes from file names by regex.
5228 19. Completely re-implemented the support for maildirsize files, in the light
5229 of a number of problems with the previous contributed implementation
5230 (4.30/29). In particular:
5232 . If the quota is zero, the maildirsize file is maintained, but no quota is
5235 . If the maildir directory does not exist, it is created before any attempt
5236 to write a maildirsize file.
5238 . The quota value in the file is just a cache; if the quota is changed in
5239 the transport, the new value overrides.
5241 . A regular expression is available for excluding directories from the
5244 20. The autoreply transport checks the characters in options that define the
5245 message's headers; it allows continued headers, but it was checking with
5246 isspace() after an embedded newline instead of explicitly looking for a
5249 21. If all the "regular" hosts to which an address was routed had passed their
5250 expiry times, and had not reached their retry times, the address was
5251 bounced, even if fallback hosts were defined. Now Exim should go on to try
5254 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5255 equivalent code in the SMTP transport. Some hosts send humungous responses
5256 to HELO/EHLO, more than 1024 it seems.
5258 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5259 gives compiler warnings in some environments. I've now done it "properly",
5262 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5263 (because of problems with the built-in one) was declared to return uschar *
5264 instead of char *, causing compiler failure.
5266 25. Fixed a file descriptor leak when processing alias/forward files.
5268 26. Fixed a minor format string issue in dbfn.c.
5270 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5272 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5273 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5276 29. When a sender address is verified, it is cached, to save repeating the test
5277 when there is more than one recipient in a message. However, when the
5278 verification involves a callout, it is possible for different callout
5279 options to be set for different recipients. It is too complicated to keep
5280 track of this in the cache, so now Exim always runs a verification when a
5281 callout is required, relying on the callout cache for the optimization.
5282 The overhead is duplication of the address routing, but this should not be
5285 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5286 to be verified with callout=postmaster, and the main callout worked but the
5287 postmaster check failed, the verification correctly failed. However, if a
5288 subsequent RCPT command asked for sender verification *without* the
5289 postmaster check, incorrect caching caused this verification also to fail,
5292 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5293 it was not caching the DNS options (qualify_single, search_parents) that
5294 were used when the lookup failed. A subsequent lookup with different
5295 options therefore always gave the same answer, though there were cases
5296 where it should not have. (Example: a "domains = !$mx_any" option on a
5297 dnslookup router: the "domains" option is always processed without any
5298 widening, but the router might have qualify_single set.) Now Exim uses the
5299 cached value only when the same options are set.
5301 32. Added John Jetmore's "exipick" utility to the distribution.
5303 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5304 than a timeout (e.g. a certificate is required, and is not provided), an
5305 Exim server now closes the connection immediately. Previously it waited for
5306 the client to close - but if the client is SSL, it seems that they each
5307 wait for each other, leading to a delay before one of them times out.
5309 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5310 maintained 0.8.x compatibility because I don't think many are using it, and
5311 it is clearly obsolete.
5313 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5316 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5317 client certificate was expired. A simple patch fixes this, though I don't
5318 understand the full logic of why the verify callback is called multiple
5321 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5322 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5323 which causes problems with some clients (such as the Certicom SSL Plus
5324 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5325 disables the coutermeasure allowing Eudora to connect."
5327 38. Exim was not checking that a write() to a log file succeeded. This could
5328 lead to Bad Things if a log got too big, in particular if it hit a file
5329 size limit. Exim now panics and dies if it cannot write to a log file, just
5330 as it does if it cannot open a log file.
5332 39. Modified OS/Makefile-Linux so that it now contains
5334 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5336 The two -D definitions ensure that Exim is compiled with large file
5337 support, which makes it possible to handle log files that are bigger than
5340 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5341 instance) a domain was checked against a named list that involved a lookup,
5342 causing $domain_data to be set, then another domain was checked against the
5343 same list, then the first domain was re-checked, the value of $domain_data
5344 after the final check could be wrong. In particular, if the second check
5345 failed, it could be set empty. This bug probably also applied to
5348 41. The strip_trailing_dot option was not being applied to the address given
5349 with the -f command-line option.
5351 42. The code for reading a message's header from the spool was incrementing
5352 $received_count, but never initializing it. This meant that the value was
5353 incorrect (doubled) while delivering a message in the same process in which
5354 it was received. In the most common configuration of Exim, this never
5355 happens - a fresh exec is done - but it can happen when
5356 deliver_drop_privilege is set.
5358 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5359 it now includes up to 150 characters of the unexpected data in the log
5362 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5363 and building data strings. The size of both of these buffers was 10 000
5364 bytes - far larger than anybody would *ever* want, thought I. Needless to
5365 say, somebody hit the limit. I have increased the maximum line length to
5366 20 000 and the maximum data length of concatenated lines to 100 000. I have
5367 also fixed two bugs, because there was no checking on these buffers. Tsk,
5368 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5369 buffer is too small.
5371 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5372 lsearch lookups. Now it does.
5374 46. When parsing a route_list item in a manualroute router, a fixed-length
5375 buffer was used for the list of hosts. I made this 1024 bytes long,
5376 thinking that nobody would ever have a list of hosts that long. Wrong.
5377 Somebody had a whole pile of complicated expansion conditions, and the
5378 string was silently truncated, leading to an expansion error. It turns out
5379 that it is easier to change to an unlimited length (owing to other changes
5380 that have happened since this code was originally written) than to build
5381 structure for giving a limitation error. The length of the item that
5382 expands into the list of hosts is now unlimited.
5384 47. The lsearch lookup could not handle data where the length of text line was
5385 more than 4095 characters. Such lines were truncated, leading to shortened
5386 data being returned. It should now handle lines of any length.
5388 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5389 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5392 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5393 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5394 no longer happens in comment lines. A list of replacements is now placed
5395 at the head of all of the source files, except those whose only change is
5396 to replace PERL_COMMAND in the very first #! line.
5398 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5399 messages on the queue, with a bottom-up merge sort, using code contributed
5400 by Michael Haardt. This should make operations like -bp somewhat faster on
5401 large queues. It won't affect queue runners, except when queue_run_in_order
5404 51. Installed eximstats 1.31 in the distribution.
5406 52. Added support for SRV lookups to the dnslookup router.
5408 53. If an ACL referred to $message_body or $message_body_end, the value was not
5409 reset for any messages that followed in the same SMTP session.
5411 54. The store-handling optimization for building very long strings was not
5412 differentiating between the different store pools. I don't think this
5413 actually made any difference in practice, but I've tidied it.
5415 55. While running the routers to verify a sender address, $sender_address
5416 was still set to the sender address. This is wrong, because when routing to
5417 send a bounce to the sender, it would be empty. Therefore, I have changed
5418 it so that, while verifying a sender address, $sender_address is set to <>.
5419 (There is no change to what happens when verifying a recipient address.)
5421 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5422 target A or AAAA records (if not already returned) without resetting the
5423 qualify_single or search_parents options of the DNS resolver. These are
5424 inappropriate in this case because the targets of MX and SRV records must
5425 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5426 target that, when qualified, matched something in the local domain. These
5427 two options are now turned off when doing these lookups.
5429 57. It seems that at least some releases of Reiserfs (which does not have the
5430 concept of a fixed number of inodes) returns zero and not -1 for the
5431 number of available inodes. This interacted badly with check_spool_inodes,
5432 which assumed that -1 was the "no such thing" setting. What I have done is
5433 to check that the total number of inodes is greater than zero before doing
5434 the test of how many are available.
5436 58. When a "warn" ACL statement has a log_message modifier, the message is
5437 remembered, and not repeated. This is to avoid a lot of repetition when a
5438 message has many recipients that cause the same warning to be written.
5439 Howewer, Exim was preserving the list of already written lines for an
5440 entire SMTP session, which doesn't seem right. The memory is now reset if a
5441 new message is started.
5443 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5444 debugging output unless log_rewrite was also set.
5446 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5447 of (int)(handle) when we know that handle contains (void *)(-1).
5449 61. The Exim daemon panic-logs an error return when it closes the incoming
5450 connection. However "connection reset by peer" seems to be common, and
5451 isn't really an error worthy of noting specially, so that particular error
5454 62. When Exim is trying to find all the local interfaces, it used to panic and
5455 die if the ioctl to get the interface flags failed. However, it seems that
5456 on at least one OS (Solaris 9) it is possible to have an interface that is
5457 included in the list of interfaces, but for which you get a failure error
5458 for this call. This happens when the interface is not "plumbed" into a
5459 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5460 failure of the "get flags" call assumes that the interface is down.
5462 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5463 makes life easier for people who are doing arithmetic on fields extracted
5464 from dates, where you often get leading zeros that should not be
5465 interpreted as octal.
5467 64. Added qualify_domain to the redirect router, to override the global
5470 65. If a pathologically long header line contained very many addresses (the
5471 report of this problem mentioned 10 000) and each of them was rewritten,
5472 Exim could use up a very large amount of memory. (It kept on making new
5473 copies of the header line as it rewrote, and never released the old ones.)
5474 At the expense of a bit more processing, the header rewriting function has
5475 been changed so that it no longer eats memory in this way.
5477 66. The generation of the Received: header has been moved from the time that a
5478 message starts to be received, to the time that it finishes. The timestamp
5479 in the Received: header should now be very close to that of the <= log
5480 line. There are two side-effects of this change:
5482 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5483 logged header lines no longer include the local Received: line, because
5484 it has not yet been created. The same applies to a copy of the message
5485 that is returned to a non-SMTP sender when a message is rejected.
5487 (b) When a filter file is tested using -bf, no additional Received: header
5488 is added to the test message. After some thought, I decided that this
5491 This change does not affect the value of $received_for. It is still set
5492 after address rewriting, but before local_scan() is called.
5494 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5496 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5497 gave an unhelpful panic error message, and a defer error. I have managed to
5498 change this behaviour so that it now rejects any supplied certificate,
5499 which seems right, as the list of acceptable certificates is empty.
5501 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5502 gave an unhelpful defer error. I have not managed to make this reject any
5503 supplied certificates, but the error message it gives is "no certificate
5504 supplied", which is not helpful.
5506 70. exigrep's output now also includes lines that are not associated with any
5507 message, but which match the given pattern. Implemented by a patch from
5508 Martin Sluka, which also tidied up the Perl a bit.
5510 71. Recipient callout verification, like sender verification, was using <> in
5511 the MAIL FROM command. This isn't really the right thing, since the actual
5512 sender may affect whether the remote host accepts the recipient or not. I
5513 have changed it to use the actual sender in the callout; this means that
5514 the cache record is now keyed on a recipient/sender pair, not just the
5515 recipient address. There doesn't seem to be a real danger of callout loops,
5516 since a callout by the remote host to check the sender would use <>.
5517 [SEE ABOVE: changed after hitting problems.]
5519 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5520 temporary errors. However, in the case of such a code being given after
5521 the end of a data transmission (i.e. after ".") Exim was failing to write
5522 a retry record for the message. (Yes, there was some broken host that was
5523 actually sending 8xx at this point.)
5525 73. An unknown lookup type in a host list could cause Exim to panic-die when
5526 the list was checked. (An example that provoked this was putting <; in the
5527 middle of a list instead of at the start.) If this happened during a DATA
5528 ACL check, a -D file could be left lying around. This kind of configuration
5529 error no longer causes Exim to die; instead it causes a defer errror. The
5530 incident is still logged to the main and panic logs.
5532 74. Buglet left over from Exim 3 conversion. The message "too many messages
5533 in one connection" was written to the rejectlog but not the mainlog, except
5534 when address rewriting (yes!) was being logged.
5536 75. Added write_rejectlog option.
5538 76. When a system filter was run not as root (that is, when system_filter_user
5539 was set), the values of the $n variables were not being returned to the
5540 main process; thus, they were not subsequently available in the $sn
5543 77. Added +return_path_on_delivery log selector.
5545 78. A connection timeout was being treated differently from recipients deferred
5546 when testing hosts_max_try with a message that was older than the host's
5547 retry timeout. (The host should not be counted, thus allowing all hosts to
5548 be tried at least once before bouncing.) This may have been the cause of an
5549 occasionally reported bug whereby a message would remain on the queue
5550 longer than the retry timeout, but would be bounced if a delivery was
5551 forced. I say "may" because I never totally pinned down the problem;
5552 setting up timeout/retry tests is difficult. See also the next item.
5554 79. The ultimate address timeout was not being applied to errors that involved
5555 a combination of host plus message (for example, a timeout on a MAIL
5556 command). When an address resolved to a number of possible hosts, and they
5557 were not all tried for each delivery (e.g. because of hosts_max_try), a
5558 message could remain on the queue longer than the retry timeout.
5560 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5563 81. Fixed an obscure SMTP outgoing bug which required at least the following
5564 conditions: (a) there was another message waiting for the same server;
5565 (b) the server returned 5xx to all RCPT commands in the first message so
5566 that the message was not completed; (c) the server dropped the connection
5567 or gave a negative response to the RSET that Exim sends to abort the
5568 transaction. The observed case was a dropped connection after DATA that had
5569 been sent in pipelining mode. That is, the server had advertised PIPELINING
5570 but was not implementing it correctly. The effect of the bug was incorrect
5571 behaviour, such as trying another host, and this could lead to a crash.
5577 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5578 and daemon.c were passed as pointers to ints; they should have been
5579 pointers to socklen_t variables (which are typically unsigned ints).
5581 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5584 3. Fixed a really odd bug that affected only the testing scheme; patching a
5585 certain fixed string in the binary changed the value of another string that
5586 happened to be identical to the end of the original first string.
5588 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5589 name", it returns that address as the IP address. On some operating
5590 systems (e.g. Solaris), it also passes back the IP address string as the
5591 "host name". However, on others (e.g. Linux), it passes back an empty
5592 string. Exim wasn't checking for this, and was changing the host name to an
5593 empty string, assuming it had been canonicized.
5595 5. Although rare, it is permitted to have more than one PTR record for a given
5596 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5597 all the names associated with an address, because they do in Solaris.
5598 However, it seems that they do not in Linux for data that comes from the
5599 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5600 I found this out when I moved to a new Linux workstation and tried to run
5601 the Exim test suite.
5603 To get round this problem I have changed the code so that it now does its
5604 own call to the DNS to look up PTR records when searching for a host name.
5605 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5606 addresses that are only in /etc/hosts are still found.
5608 This behaviour is, however, controlled by an option called host_lookup_
5609 order, which defaults to "bydns:byaddr". If people want to use the other
5610 order, or indeed, just use one or the other means of lookup, they can
5611 specify it in this variable.
5613 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5614 some operating systems, this comes back from gethostbyaddr() as an empty
5615 string, and this is what Exim used to test for. However, it seems that in
5616 other systems, "." is yielded. Exim now tests for this case too.
5618 7. The values of check_spool_space and check_log_space are now held internally
5619 as a number of kilobytes instead of an absolute number of bytes. If a
5620 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5621 kilobyte. This means that much larger values can be stored.
5623 8. Exim monitor: an attempt to get the action menu when not actually pointing
5624 at a message produces an empty menu entitled "No message selected". This
5625 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5626 no entries in it ("Shell widget menu has zero width and/or height"). So I
5627 have added a single, blank menu entry in this case.
5629 9. Added ${quote_local_part.
5631 10. MIME decoding is now applied to the contents of Subject: header lines when
5634 11. Now that a reference to $sender_host_address automatically causes a reverse
5635 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5636 host lookup before query-style lookups in lists that might use this
5637 variable. This has therefore been abolished, and the "net-" prefix is no
5638 longer necessary for query-style lookups.
5640 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5641 have been a typo for LFLAGS, so it has been changed.
5643 13. The install script calls Exim with "-C /dev/null" in order to find the
5644 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5645 to be output. Howeve, since Exim outputs its version number before the
5646 error, it didn't break the script. It just looked ugly. I fixed this by
5647 always allowing "-C /dev/null" if the caller is root.
5649 14. Ignore overlarge ACL variable number when reading spool file - insurance
5650 against a later release with more variables having written the file.
5652 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5653 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5654 this, as well as the form without the "IPv6" on the front.
5656 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5657 OS/Makefile-Darwin file.
5659 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5660 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5662 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5663 variable. However, some OS use ints for inodes. I've added cast to long int
5664 to get rid of the compiler warning.
5666 19. I had forgotten to lock out "/../" in configuration file names when
5667 ALT_CONFIG_PREFIX was set.
5669 20. Routers used for verification do not need to specify transports. However,
5670 if such a router generated a host list, and callout was configured, Exim
5671 crashed, because it could not find a port number from the (non-existent)
5672 transport. It now assumes port 25 in this circumstance.
5674 21. Added the -t option to exigrep.
5676 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5677 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5678 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5679 them is set, LOOKUP_LSEARCH is forced.
5681 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5683 24. Added sender and host information to the "rejected by local_scan()" log
5684 line; previously there was no indication of these.
5686 25. Added .include_if_exists.
5688 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5689 Linux. It turns out that not all file systems support this. Apparently some
5690 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5691 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5692 not supported on the file descriptor, is now ignored when Exim is trying to
5693 sync a directory. This applies only to Linux.
5695 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5697 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5698 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5701 29. Added support for maildirsize files from supplied patch (modified a bit).
5703 30. The use of :fail: followed by an empty string could lead Exim to respond to
5704 sender verification failures with (e.g.):
5706 550 Verification failed for <xxx>
5707 550 Sender verify failed
5709 where the first response line was missing the '-' that indicates it is not
5710 the final line of the response.
5712 31. The loop for finding the name of the user that called Exim had a hardwired
5713 limit of 10; it now uses the value of finduser_retries, which is used for
5714 all other user lookups.
5716 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5719 33. Exim was neglecting to zero errno before one call of strtol() when
5720 expanding a string and expecting an integer value. On some systems this
5721 resulted in spurious "integer overflow" errors. Also, it was casting the
5722 result into an int without checking.
5724 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5725 did not work. The code looks as if it has *never* worked, though it appears
5726 to have been documented since at least releast 1.62. I have made it work.
5728 35. The "timeout_DNS" error in retry rules, also documented since at least
5729 1.62, also never worked. As it isn't clear exactly what this means, and
5730 clearly it isn't a major issue, I have abolished the feature by treating it
5731 as "timeout", and writing a warning to the main and panic logs.
5733 36. The display of retry rules for -brt wasn't always showing the error code
5736 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5737 timeout_connect_A, timeout_connect_MX.
5739 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5740 to the empty sender.
5742 39. The daemon was not analysing the content of -oX till after it had closed
5743 stderr and disconnected from the controlling terminal. This meant that any
5744 syntax errors were only noted on the panic log, and the return code from
5745 the command was 0. By re-arranging the code a little, I've made the
5746 decoding happen first, so such errors now appear on stderr, and the return
5747 code is 1. However, the actual setting up of the sockets still happens in
5748 the disconnected process, so errors there are still only recorded on the
5751 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5752 connections (as happens on some IP stacks) was logged at start up time as
5753 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5754 it from "IPv6 and IPv4", which means that two separate sockets are being
5757 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5758 says whether AF_INET or AF_INET6 was passed as an argument.
5760 42. Exiwhat output was messed up when time zones were included in log
5763 43. Exiwhat now gives more information about the daemon's listening ports,
5764 and whether -tls-on-connect was used.
5766 44. The "port" option of the smtp transport is now expanded.
5768 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5769 silently ignored. Now an error message is written to the main and panic
5772 46. There's a new ACL modifier called "logwrite" which writes to a log file
5773 as soon as it is encountered.
5775 47. Added $local_user_uid and $local_user_gid at routing time.
5777 48. Exim crashed when trying to verify a sender address that was being
5780 49. Exim was recognizing only a space character after ".include". It now also
5781 recognizes a tab character.
5783 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5784 extracting the relevant information from the specification. The man page no
5785 longer contains scrambled data for the -d option, and I've added a section
5786 at the front about calling Exim under different names.
5788 51. Added "extra_headers" argument to the "mail" command in filter files.
5790 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5793 53. Installed eximstats 1.29.
5795 54. Added transport_filter_timeout as a generic transport option.
5797 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5798 Cc: header lines. This was required by RFC 822, but it not required by RFC
5801 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5802 incoming messages that did not have them. Now it does so only if the
5803 message originates locally, that is, if there is no associated remote host
5804 address. When Resent- header lines are present, this applies to the Resent-
5805 lines rather than the non-Resent- lines.
5807 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5808 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5810 58. Messages for configuration errors now include the name of the main
5811 configuration file - useful now that there may be more than one file in a
5812 list (.included file names were always shown).
5814 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5815 for those rare installations that do not start the daemon as root or run it
5816 setuid root. I've cut out the call to initgroups() if the daemon is not
5819 60. The Exim user and group can now be bound into the binary as text strings
5820 that are looked up at the start of Exim's processing.
5822 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5824 62. Added $mailstore_basename variable.
5826 63. Installed patch to sieve.c from Michael Haardt.
5828 64. When Exim failed to open the panic log after failing to open the main log,
5829 the original message it was trying to log was written to stderr and debug
5830 output, but if they were not available (the usual case in production), it
5831 was lost. Now it is written to syslog before the two lines that record the
5832 failures to open the logs.
5834 65. Users' Exim filters run in subprocesses under the user's uid. It is
5835 possible for a "deliver" command or an alias in a "personal" command to
5836 provoke an address rewrite. If logging of address rewriting is configured,
5837 this fails because the process is not running as root or exim. There may be
5838 a better way of dealing with this, but for the moment (because 4.30 needs
5839 to be released), I have disabled address rewrite logging when running a
5840 filter in a non-root, non-exim process.
5846 1. The buildconfig auxiliary program wasn't quoting the value set for
5847 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5848 not defined. This bug was masked in 4.22 by the effect that was fixed in
5851 2. Some messages that were rejected after a message id was allocated were
5852 shown as "incomplete" by exigrep. It no longer does this for messages that
5853 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5855 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5856 have allow_domain_literals set, the ID did not get logged in the <= line.
5857 Domain literals are now always recognized in Message-ID: header lines.
5859 4. The first argument for a ${extract expansion item is the key name or field
5860 number. Leading and trailing spaces in this item were not being ignored,
5861 causing some misleading effects.
5863 5. When deliver_drop_privilege was set, single queue runner processes started
5864 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5865 same command in the process it spins off) were not dropping privilege.
5867 6. When the daemon running as "exim" started a queue runner, it always
5868 re-executed Exim in the spun-off process. This is a waste of effort when
5869 deliver_drop_privilege is set. The new process now just calls the
5870 queue-runner function directly.
5876 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5879 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5880 ignored. Though the use of -oP was forcing the writing of a pid file, it
5881 was always written to the default place.
5883 3. If the message "no IP address found for host xxxx" is generated during
5884 incoming verification, it is now followed by identification of the incoming
5885 connection (so you can more easily find what provoked it).
5887 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5889 5. Added some features to "harden" Exim a bit more against certain attacks:
5891 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5892 be put in Local/Makefile. This is like the never_users runtime option,
5893 but it cannot be overridden. The default setting is "root".
5895 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5896 prefix string with which any file named in a -C command line option
5899 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5900 is retained for -C and -D only if the caller of Exim is root. Without
5901 it, the exim user may also use -C and -D and retain privilege.
5903 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5904 command line option is disabled.
5906 6. Macro names set by the -D option must start with an upper case letter, just
5907 like macro names defined in the configuration file.
5909 7. Added "dereference=" facility to LDAP.
5911 8. Two instances of the typo "uknown" in the source files are fixed.
5913 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5914 the Configure-Makefile script screwed up while processing it.
5916 10. Incorporated PCRE 4.4.
5918 11. The SMTP synchronization check was not operating right at the start of an
5919 SMTP session. For example, it could not catch a HELO sent before the client
5920 waited for the greeting. There is now a check for outstanding input at the
5921 point when the greeting is written. Because of the duplex, asynchronous
5922 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5923 way, but not yet received, when the check is performed.
5925 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5926 on TCP/IP sockets, because this apparently causes some broken clients to
5929 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5930 unchanged) from the Cygwin maintainer.
5932 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5933 was supported instead of testing for mbx. Effectively a typo.
5935 15. The spa authenticator server code was not checking that the input it
5936 received was valid base64.
5938 16. The debug output line for the "set" modifier in ACLs was not showing the
5939 name of the variable that was being set.
5941 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5943 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5944 lookup on its own. Something else was needed to trigger the lookup. For
5945 example, a match in host_lookup or the need for a host name in a host list.
5946 Now, if $sender_host_name is referenced and the host name has not yet been
5947 looked up, a lookup is performed. If the lookup fails, the variable remains
5948 empty, and $host_lookup_failed is set to "1".
5950 19. Added "eqi" as a case-independent comparison operator.
5952 20. The saslauthd authentication condition could segfault if neither service
5953 nor realm was specified.
5955 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5956 error message that was logged was misleading, and incoming SMTP
5957 connections were dropped. The message is now more accurate, and temporary
5958 errors are given to SMTP connections.
5960 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5961 (except RSET) with a 421 error, until QUIT is received. However, it was
5962 failing to send a response to QUIT.
5964 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5965 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5966 ACL is now run after the helo_try_verify_hosts code.
5968 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5969 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5970 case-independent, so other case variants are also recognized). Apparently
5971 some systems use these upper case variants.
5973 25. If more than two messages were waiting for the same host, and a transport
5974 filter was specified for the transport, Exim sent two messages over the
5975 same TCP/IP connection, and then failed with "socket operation on non-
5976 socket" when it tried to send the third.
5978 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5980 27. The extern definition of crypt16() in expand.c was not being excluded when
5981 the OS had its own crypt16() function.
5983 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5984 as a preferred synonym for return_size_limit, both as an option and as an
5987 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5989 30. Changed the default configuration ACL to relax the local part checking rule
5990 for addresses that are not in any local domains. For these addresses,
5991 slashes and pipe symbols are allowed within local parts, but the sequence
5992 /../ is explicitly forbidden.
5994 31. SPA server authentication was not clearing the challenge buffer before
5997 32. log_message in a "warn" ACL statement was writing to the reject log as
5998 well as to the main log, which contradicts the documentation and doesn't
5999 seem right (because no rejection is happening). So I have stopped it.
6001 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6002 However, I am unable to do any testing of this.
6004 34. Fixed an infelicity in the appendfile transport. When checking directories
6005 for a mailbox, to see if any needed to be created, it was accidentally
6006 using path names with one or more superfluous leading slashes; tracing
6007 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6009 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6010 contents are added to the log line that is written for every discarded
6011 recipient. (Previously a log_message setting was ignored.)
6013 36. The ${quote: operator now quotes the string if it is empty.
6015 37. The install script runs exim in order to find its version number. If for
6016 some reason other than non-existence or emptiness, which it checks, it
6017 could not run './exim', it was installing it with an empty version number,
6018 i.e. as "exim-". This error state is now caught, and the installation is
6021 38. An argument was missing from the function that creates an error message
6022 when Exim fails to connect to the socket for saslauthd authentication.
6023 This could cause Exim to crash, or give a corrupted message.
6025 39. Added isip, isip4, and isip6 to ${if conditions.
6027 40. The ACL variables $acl_xx are now saved with the message, and can be
6028 accessed later in routers, transports, and filters.
6030 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6031 strings in the file are not string-expanded.
6033 42. If a MAIL command specified a SIZE value that was too large to fit into an
6034 int variable, the check against message_size_limit failed. Such values are
6035 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6036 day this will have to be increased, but I don't think I want to be around
6037 when emails are that large.
6044 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6045 iconv() is not standard in FreeBSD.
6047 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6048 IPv6 enabled. The observed symptom was a segmentation fault on return from
6049 the function os_common_find_running_interfaces() in src/os.c.
6051 3. In the check_special_case() function in daemon.c I had used "errno" as an
6052 argument name, which causes warnings on some systems. This was basically a
6053 typo, since it was named "eno" in the comments!
6055 4. The code that waits for the clock to tick (at a resolution of some fraction
6056 of a second) so as to ensure message-id uniqueness was always waiting for
6057 at least one whole tick, when it could have waited for less. [This is
6058 almost certainly not relevant at current processor speeds, where it is
6059 unlikely to ever wait at all. But we try to future-proof.]
6061 5. The function that sleeps for a time interval that includes fractions of a
6062 second contained a race. It did not block SIGALRM between setting the
6063 timer, and suspending (a couple of lines later). If the interval was short
6064 and the sigsuspend() was delayed until after it had expired, the suspension
6065 never ended. On busy systems this could lead to processes getting stuck for
6068 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6069 process, before it forks any delivery processes. The open lookup caching
6070 mechanism meant that the open file or database connection was passed into
6071 the delivery process. The problem was that delivery processes always tidy
6072 up cached lookup data. This could cause a problem for the next delivery
6073 process started by the queue runner, because the external queue runner
6074 process does not know about the closure. So the next delivery process
6075 still has data in the lookup cache. In the case of a file lookup, there was
6076 no problem because closing a file descriptor in a subprocess doesn't affect
6077 the parent. However, if the lookup was caching a connection to a database,
6078 the connection was closed, and the second delivery process was likely to
6079 see errors such as "PGSQL: query failed: server closed the connection
6080 unexpectedly". The problem has been fixed by closing all cached lookups
6081 in a queue runner before running a delivery process.
6083 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6084 seem to have the "const" qualifier which it has on other OS. I've
6087 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6088 *of the same type* (client or server) with the same public name that an
6089 error should be diagnosed.
6091 9. When Exim looked up a host name for an IP address, but failed to find the
6092 original IP address when looking up the host name (a safety check), it
6093 output the message "<ip address> does not match any IP for NULL", which was
6094 confusing, to say the least. The bug was that the host name should have
6095 appeared instead of "NULL".
6097 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6098 user that is built into the binary, and the -C or -D options is used, root
6099 privilege is dropped before the configuration file is read. In addition,
6100 logging is switched to stderr instead of the normal log files. If the
6101 configuration then re-defines the Exim user, the unprivileged environment
6102 is probably not what is expected, so Exim logs a panic warning message (but
6105 However, if deliver_drop_privilege is set, the unprivileged state may well
6106 be exactly what is intended, so the warning has been cut out in that case,
6107 and Exim is allowed to try to write to its normal log files.
6113 1. smtp_return_error_details was not giving details for temporary sender
6114 or receiver verification errors.
6116 2. Diagnose a configuration error if two authenticators have the same public
6119 3. Exim used not to create the message log file for a message until the first
6120 delivery attempt. This could be confusing when incoming messages were held
6121 for policy or load reasons. The message log file is now created at the time
6122 the message is received, and an initial "Received" line is written to it.
6124 4. The automatically generated man page for command line options had a minor
6125 bug that caused no ill effects; however, a more serious problem was that
6126 the procedure for building the man page automatically didn't always
6127 operate. Consequently, release 4.20 contains an out-of-date version. This
6128 shouldn't happen again.
6130 5. When building Exim with embedded Perl support, the script that builds the
6131 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6132 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6134 6. The freeze_tell option was not being used for messages that were frozen on
6135 arrival, either by an ACL or by local_scan().
6137 7. Added the smtp_incomplete_transaction log selector.
6139 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6140 was accepting AUTH without a new EHLO.
6142 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6143 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6146 10. Logging of TCP/IP connections (when configured) now happens in the main
6147 daemon process instead of the child process, so that the TCP/IP connection
6148 count is more accurate (but it can never be perfect).
6150 11. The use of "drop" in a nested ACL was not being handled correctly in the
6151 outer ACL. Now, if condition failure induced by the nested "drop" causes
6152 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6153 or "require"), the connection is dropped.
6155 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6156 that yield "discard" can now be used with an "accept" or a "discard" verb,
6157 but an error is generated for any others (because I can't see a useful way
6158 to define what should happen).
6160 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6161 are now processed in the same way as lines in the Exim configuration file.
6162 In particular, continuation lines are supported.
6164 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6166 15. Added -ti meaning -t -i.
6168 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6169 domains, and warn by logging if others are found.
6171 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6172 returned from GIFCONF when Exim is trying to find the list of interfaces on
6173 a host. The code in os.c has been modified to copy each ifreq to an aligned
6174 structure in all cases.
6176 Also, in some cases, the returned ifreq's were being copied to a 'struct
6177 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6178 means the last couple of bytes of an IPv6 address could be chopped if the
6179 ifreq contained only a normal sockaddr (14 bytes storage).
6181 18. Named domain lists were not supported in the hosts_treat_as_local option.
6182 An entry such as +xxxx was not recognized, and was treated as a literal
6185 19. Ensure that header lines added by a DATA ACL are included in the reject log
6186 if the ACL subsequently rejects the message.
6188 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6189 MD5 (which is deprecated).
6191 21. When testing a filter file using -bf, Exim was writing a message when it
6192 took the sender from a "From " line in the message, but it was not doing so
6193 when it took $return_path from a Return-Path: header line. It now does.
6195 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6196 with a valid header line field name (a series of printing characters
6197 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6199 23. Changed "disc" in the source to "disk" to conform to the documentation and
6200 the book and for uniformity.
6202 24. Ignore Sendmail's -Ooption=value command line item.
6204 25. When execve() failed while trying to run a command in a pipe transport,
6205 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6206 could be confused with a return value of 69 from the command itself. This
6207 has been changed to 127, the value the shell returns if it is asked to run
6208 a non-existent command. The wording for the related log line suggests a
6209 non-existent command as the problem.
6211 26. If received_header_text expands to an empty string, do not add a Received:
6212 header line to the message. (Well, it adds a token one on the spool, but
6213 marks it "old" so that it doesn't get used or transmitted.)
6215 27. Installed eximstats 1.28 (addition of -nt option).
6217 28. There was no check for failure on the call to getsockname() in the daemon
6218 code. This can fail if there is a shortage of resources on the system, with
6219 ENOMEM, for example. A temporary error is now given on failure.
6221 29. Contrary to the C standard, it seems that in some environments, the
6222 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6223 program. Exim now does this explicitly; it affects the formatting of
6224 timestamps using strftime().
6226 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6227 complaints. I've now initialized all the variables, to avoid this.
6229 32. Header lines added by a system filter were not being "seen" during
6230 transport-time rewrites.
6232 33. The info_callback() function passed to OpenSSL is set up with type void
6233 (*)(SSL *, int, int), as described somewhere. However, when calling the
6234 function (actually a macro) that sets it up, the type void(*)() is
6235 expected. I've put in a cast to prevent warnings from picky compilers.
6237 34. If a DNS black list lookup found a CNAME record, but there were no A
6238 records associated with the domain it pointed at, Exim crashed.
6240 35. If a DNS black list lookup returned more than one A record, Exim ignored
6241 all but the first. It now scans all returned addresses if a particular IP
6242 value is being sought. In this situation, the contents of the
6243 $dnslist_value variable are a list of all the addresses, separated by a
6246 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6247 to accept a host name and all its aliases if the forward lookup for any of
6248 them yielded the IP address of the incoming connection. Now it accepts only
6249 those names whose forward lookup yields the correct IP address. Any other
6250 names are discarded. This closes a loophole whereby a rogue DNS
6251 administrator could create reverse DNS records to break through a
6252 wildcarded host restriction in an ACL.
6254 37. If a user filter or a system filter that ran in a subprocess used any of
6255 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6256 the wrong values were passed to the pipe command ($thisaddress had the
6257 value of $0, $0 had the value of $1, etc). This bug was introduced by
6258 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6260 38. Improved the line breaking for long SMTP error messages from ACLs.
6261 Previously, if there was no break point between 40 and 75 characters, Exim
6262 left the rest of the message alone. Two changes have been made: (a) I've
6263 reduced the minimum length to 35 characters; (b) if it can't find a break
6264 point between 35 and 75 characters, it looks ahead and uses the first one
6265 that it finds. This may give the occasional overlong line, but at least the
6266 remaining text gets split now.
6268 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6269 file descriptors might be low, and that setting 1000 would always raise it.
6270 It turns out that in some environments, the limit is already over 1000 and
6271 that lowering it causes trouble. So now Exim takes care not to decrease it.
6273 40. When delivering a message, the value of $return_path is set to $sender_
6274 address at the start of routing (routers may change the value). By an
6275 oversight, this default was not being set up when an address was tested by
6276 -bt or -bv, which affected the outcome if any router or filter referred to
6279 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6280 log only when "message" or "log_message" are set. However, if one of the
6281 conditions was an address verification, or a call to a nested ACL, the
6282 messages generated by the underlying test were being passed through. This
6283 no longer happens. The underlying message is available in $acl_verify_
6284 message for both "message" and "log_message" expansions, so it can be
6285 passed through if needed.
6287 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6288 new expansion $bh_ to give the encoded byte string without charset
6289 translation. Translation happens only if iconv() is available; HAVE_ICONV
6290 indicates this at build time. HEADERS_CHARSET gives the charset to
6291 translate to; headers_charset can change it in the configuration, and
6292 "headers charset" can change it in an individual filter file.
6294 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6295 that creates RFC 2047 encoded "words" labels them as that charset instead
6296 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6297 expansion operator; (ii) when Exim creates a From: line for a local
6298 message; (iii) when a header line is rewritten to include a "phrase" part.
6300 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6301 buggy, causing it to skip the first lines of messages whose message ID
6302 ended in 'D'. This would not have bitten before Exim release 4.14, because
6303 message IDs were unlikely to end in 'D' before then. The effect was to have
6304 incorrect size information for certain domains.
6306 45. #include "config.h" was missing at the start of the crypt16.c module. This
6307 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6310 46. If there was a timeout during a "random" callout check, Exim treated it as
6311 a failure of the random address, and carried on sending RSET and the real
6312 address. If the delay was just some slowness somewhere, the response to the
6313 original RCPT would be taken as a response to RSET and so on, causing
6314 mayhem of various kinds.
6316 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6317 when I implemented it. It didn't allow for the fact that some option values
6318 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6319 the right test for positive values.
6321 48. Domain names in DNS records are case-independent. Exim always looks them up
6322 in lower case. Some resolvers return domain names in exactly the case they
6323 appear in the zone file, that is, they may contain uppercase letters. Not
6324 all resolvers do this - some return always lower case. Exim was treating a
6325 change of case by a resolver as a change of domain, similar to a widening
6326 of a domain abbreviation. This triggered its re-routing code and so it was
6327 trying to route what was effectively the same domain again. This normally
6328 caused routing to fail (because the router wouldn't handle the domain
6329 twice). Now Exim checks for this case specially, and just changes the
6330 casing of the domain that it ultimately uses when it transmits the message
6333 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6336 50. If a filter generated a file delivery with a non-absolute name (possible if
6337 no home directory exists for the router), the forbid_file option was not
6340 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6341 the existing equality matching.
6343 52. Exim was using ints instead of ino_t variables in some places where it was
6344 dealing with inode numbers.
6346 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6347 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6348 called TMPDIR, and if it finds it is different, it changes its value.
6350 54. The smtp_printf() function is now made available to local_scan() so
6351 additional output lines can be written before returning. There is also an
6352 smtp_fflush() function to enable the detection of a dropped connection.
6353 The variables smtp_input and smtp_batched_input are exported to
6356 55. Changed the default runtime configuration: the message "Unknown user"
6357 has been removed from the ACL, and instead placed on the localuser router,
6358 using the cannot_route_message feature. This means that any verification
6359 failures that generate their own messages won't get overridden. Similarly,
6360 the "Unrouteable address" message that was in the ACL for unverifiable
6361 relay addresses has also been removed.
6363 56. Added hosts_avoid_esmtp to the smtp transport.
6365 57. The exicyclog script was not checking for the esoteric option
6366 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6367 will work only if exicyclog is run under the appropriate euid.
6369 58. Following a discussion on the list, the rules by which Exim recognises line
6370 endings on incoming messages have been changed. The -dropcr and drop_cr
6371 options are now no-ops, retained only for backwards compatibility. The
6372 following line terminators are recognized: LF CRLF CR. However, special
6373 processing applies to CR:
6375 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6376 nor a local message in the state where . is a terminator.
6378 (ii) If a bare CR is encountered in a header line, an extra space is added
6379 after the line terminator so as not to end the header. The reasoning
6380 behind this is that bare CRs in header lines are most likely either
6381 to be mistakes, or people trying to play silly games.
6383 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6384 was being incorrectly given as 18 bytes larger than it should have been.
6385 This is a VOB (very old bug).
6387 60. This may never have affected anything current, but just in case it has:
6388 When the local host is found other than at the start of a list of hosts,
6389 the local host, those with the same MX, and any that follow, are discarded.
6390 When the list in question was part of a longer list of hosts, the following
6391 hosts (not currently being processed) were also being discarded. This no
6392 longer happens. I'm not sure if this situation could ever has previously
6395 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6398 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6399 header. This is something that is recommended in a new Internet Draft, and
6400 is something that is documented as being done by Sendmail. There are two
6401 possible values. For messages generated by the autoreply transport, Exim
6404 Auto-Submitted: auto-replied
6406 whereas for all other generated messages (e.g. bounces) it adds
6408 Auto-Submitted: auto-generated
6410 63. The "personal" condition in filters now includes a test for the
6411 Auto-Submitted: header. If it contains the string "auto-" the message it
6412 not considered personal.
6414 64. Added rcpt_include_affixes as a generic transport option.
6416 65. Added queue_only_override (default true).
6418 66. Added the syslog_duplication option.
6420 67. If what should have been the first header line of a message consisted of
6421 a space followed by a colon, Exim was mis-interpreting it as a header line.
6422 It isn't of course - it is syntactically invalid and should therefore be
6423 treated as the start of the message body. The misbehaviour could have
6424 caused a number of strange effects, including loss of data in subsequent
6425 header lines, and spool format errors.
6427 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6428 client host had authenticated. This control can now be exercised by an ACL
6429 for more flexibility.
6431 69. By default, callouts do not happen when testing with -bh. There is now a
6432 variant, -bhc, which does actually run the callout code, including
6433 consulting and updating the callout cache.
6435 70. Added support for saslauthd authentication, courtesy of Alexander
6438 71. If statvfs() failed on the spool or log directories while checking their
6439 size for availability, Exim confusingly gave the error "space shortage".
6440 Furthermore, in debugging mode it crashed with a floating point exception.
6441 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6442 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6443 serious problem, Exim now writes to the main and panic logs when this
6444 happens, with details of the failure. It then refuses to accept the
6445 incoming message, giving the message "spool directory problem" or "log
6446 directory problem" with a 421 code for SMTP messages.
6448 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6449 0, 1, and 2 exist, because some OS complain for execs without them (see
6450 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6451 descriptors. However, the code omitted to check that the open succeeded,
6452 causing mysterious errors if for some reason the permissions on /dev/null
6453 got screwed. Now Exim writes a message to the main and panic logs, and
6454 bombs out if it can't open /dev/null.
6456 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6457 interact so that it is all more flexible. It is supposed to remain
6458 backwards compatible. Also added extra_local_interfaces.
6460 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6461 to bomb out with an assertion failure - to the client this appears as a
6462 connection drop. This problem occurs in the part of the code that was taken
6463 from the Samba project. Fortunately, the assertion is in a very simple
6464 function, so I have fixed this by reproducing the function inline in the
6465 one place where it is called, and arranging for authentication to fail
6466 instead of killing the process with assert().
6468 75. The SPA client code was not working when the server requested OEM rather
6469 than Unicode encoding.
6471 76. Added code to make require_files with a specific uid setting more usable in
6472 the case where statting the file as root fails - usually a non-root-mounted
6473 NFS file system. When this happens and the failure is EACCES, Exim now
6474 forks a subprocess and does the per-uid checking as the relevant uid.
6476 77. Added process_log_path.
6478 78. If log_file_path was not explicitly set, a setting of check_log_space or
6479 check_log_inodes was ignored.
6481 79. If a space check for the spool or log partitions fails, the incident is now
6482 logged. Of course, in the latter case the data may get lost...
6484 80. Added the %p formatting code to string_format() so that it can be used to
6485 print addresses in debug_print(). Adjusted all the address printing in the
6486 debugging in store.c to use %p rather than %d.
6488 81. There was a concern that a line of code in smtp_in.c could overflow a
6489 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6490 initially expressed, the concern was not well-founded, because trailing
6491 spaces are removed early. However, if the trailing spaces were followed by
6492 a NULL, they did not get removed, so the overflow was possible. Two fixes
6495 (a) I re-wrote the offending code in a cleaner fashion.
6496 (b) If an incoming SMTP command contains a NULL character, it is rejected
6499 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6500 runs initgroups(), so that if the Exim user is in any additional groups,
6501 they will be used during message reception.
6507 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob