1 Change log file for Exim from version 4.21
2 -------------------------------------------
6 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
7 supports it and a size is available (ie. the sending peer gave us one).
9 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
10 years of being deprecated). The replacements are the ACLs
11 acl_smtp_mime and acl_not_smtp_mime.
13 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
14 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
15 or even an in-clear connection were permitted. Now, if the host lookup
16 was dnssec and dane was requested then the host is only used if the
17 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
18 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
19 if one fails this test.
20 This means that a poorly-configured remote DNS will make it incommunicado;
21 but it protects against a DNS-interception attack on it.
23 JH/04 Bug 1810: make continued-use of an open smtp transport connection
24 non-noisy when a race steals the message being considered.
29 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
30 and 3.4.4 - once the server is enabled to respond to an OCSP request
31 it does even when not requested, resulting in a stapling non-aware
32 client dropping the TLS connection.
34 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
35 support variable-length bit vectors. No functional change.
37 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
38 The I= interface field on outgoing lines is now after the H= remote
39 host field, same as incoming lines. There is a separate
40 outgoing_interface log selector which allows you to disable the
43 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
44 If not running log_selector +smtp_connection the mainlog would be held
45 open indefinitely after a "too many connections" event, including to a
46 deleted file after a log rotate. Leave the per net connection logging
47 leaving it open for efficiency as that will be quickly detected by the
48 check on the next write.
50 HS/01 Bug 1671: Fix post transport crash.
51 Processing the wait-<transport> messages could crash the delivery
52 process if the message IDs didn't exist for some reason. When
53 using 'split_spool_directory=yes' the construction of the spool
54 file name failed already, exposing the same netto behaviour.
56 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
57 mime_regex ACL conditions.
59 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
60 to DSN fail messages (bounces): remote IP, remote greeting, remote response
61 to HELO, local diagnostic string.
63 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
64 log line to debug. Even when configured with a tls authenticator many
65 client connections are expected to not authenticate in this way, so
66 an authenticate fail is not an error.
68 HS/02 Add the Exim version string to the process info. This way exiwhat
69 gives some more detail about the running daemon.
71 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
72 matter for fast-change records such as DNSBLs.
74 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
75 retry record, even if constant. There may be multiple transports with
76 different interface settings and the retry behaviour needs to be kept
79 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
81 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
83 JH/10 Bug 840: fix log_defer_output option of pipe transport
85 JH/11 Bug 830: use same host for all RCPTS of a message, even under
86 hosts_randomize. This matters a lot when combined with mua_wrapper.
88 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
89 ${quote_pgsql:<string>} operator.
91 JH/13 Bug 1708: avoid misaligned access in cached lookup.
93 JH/14 Change header file name for freeradius-client. Relevant if compiling
94 with Radius support; from the Gentoo tree and checked under Fedora.
96 JH/15 Bug 1712: Introduce $prdr_requested flag variable
98 JH/16 Bug 1714: Permit an empty string as expansion result for transport
99 option transport_filter, meaning no filtering.
101 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
103 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
104 defaults to "*" (all hosts). The variable is now available when not built
105 with TLS, default unset, mainly to enable keeping the testuite sane.
106 If a server certificate is not supplied (via tls_certificate) an error is
107 logged, and clients will find TLS connections fail on startup. Presumably
108 they will retry in-clear.
109 Packagers of Exim are strongly encouraged to create a server certificate
110 at installation time.
112 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
113 with the $config_file variable.
115 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
116 in transport context, after the attempt, and per-recipient. The latter type
117 is per host attempted. The event data is the error message, and the errno
118 information encodes the lookup type (A vs. MX) used for the (first) host,
119 and the trailing two digits of the smtp 4xx reponse.
121 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
122 to write to mainlog (or rejectlog, paniclog) in the window between file
123 creation and permissions/ownership being changed. Particularly affects
124 installations where exicyclog is run as root, rather than exim user;
125 result is that the running daemon panics and dies.
127 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
129 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
130 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
131 "pri" and "weight". Note that the previous implicit priority given by the
132 list order is no longer honoured.
134 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
137 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
138 by defining SUPPORT_SOCKS.
140 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
141 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
142 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
143 variables are renamed to proxy_{local,external}_{address,port}.
145 JH/27 Move Internationalisation support from Experimental to mainline, enabled
146 for a build by defining SUPPORT_I18N
148 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
149 of the query string, and make ${quote_redis:} do that quoting.
151 JH/29 Move Events support from Experimental to mainline, enabled by default
152 and removable for a build by defining DISABLE_EVENT.
154 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
156 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
157 cached by the daemon.
159 JH/32 Move Redis support from Experimental to mainline, enabled for a build
160 by defining LOOKUP_REDIS. The libhiredis library is required.
162 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
163 keys are given for lookup.
165 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
166 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
167 only supported when built with TLS support. The PolarSSL SHA routines
168 are still used when the TLS library is too old for convenient support.
170 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
171 openssl_options), for security. OpenSSL forces this from version 1.1.0
172 server-side so match that on older versions.
174 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
175 allocation for $value could be released as the expansion processing
176 concluded, but leaving the global pointer active for it.
178 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
179 and to use the domains and local_parts ACL conditions.
181 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
182 incorrectly not doubled on cutthrough transmission, hence seen as a
183 body-termination at the receiving system - resulting in truncated mails.
184 Commonly the sender saw a TCP-level error, and retransmitted the message
185 via the normal store-and-forward channel. This could result in duplicates
186 received - but deduplicating mailstores were liable to retain only the
187 initial truncated version.
189 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
191 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
193 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
194 we're in there, support oversigning also; bug 1309.
196 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
198 HS/04 Add support for keep_environment and add_environment options.
200 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
201 either intentional arithmetic overflow during PRNG, or testing config-
204 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
205 delivery resulted in actual delivery. Cancel cutthrough before DATA
208 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
209 rejects a recipient: pass the reject to the originator.
211 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
212 Many were false-positives and ignorable, but it's worth fixing the
215 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
216 for the new environment-manipulation done at startup. Move the routines
217 from being local to tls.c to being global via the os.c file.
219 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
220 an extract embedded as result-arg for a map, the first arg for extract
221 is unavailable so we cannot tell if this is a numbered or keyed
222 extraction. Accept either.
228 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
231 JH/02 The smtp transport option "multi_domain" is now expanded.
233 JH/03 The smtp transport now requests PRDR by default, if the server offers
236 JH/04 Certificate name checking on server certificates, when exim is a client,
237 is now done by default. The transport option tls_verify_cert_hostnames
238 can be used to disable this per-host. The build option
239 EXPERIMENTAL_CERTNAMES is withdrawn.
241 JH/05 The value of the tls_verify_certificates smtp transport and main options
242 default to the word "system" to access the system default CA bundle.
243 For GnuTLS, only version 3.0.20 or later.
245 JH/06 Verification of the server certificate for a TLS connection is now tried
246 (but not required) by default. The verification status is now logged by
247 default, for both outbound TLS and client-certificate supplying inbound
250 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
253 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
254 Status Notification (bounce) messages are now MIME format per RFC 3464.
255 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
256 under the control of the dsn_advertise_hosts option, and routers may
257 have a dsn_lasthop option.
259 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
260 default, modifiable by a malware= option. The list separator for
261 the options can now be changed in the usual way. Bug 68.
263 JH/10 The smtp_receive_timeout main option is now expanded before use.
265 JH/11 The incoming_interface log option now also enables logging of the
266 local interface on delivery outgoing connections.
268 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
269 if the interface and destination host and port all match.
271 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
274 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
275 Patch from Andrew Lewis.
277 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
278 now supports optional time-restrictions, weighting, and priority
279 modifiers per server. Patch originally by <rommer@active.by>.
281 JH/16 The spamd_address main option now supports a mixed list of local
282 and remote servers. Remote servers can be IPv6 addresses, and
283 specify a port-range.
285 JH/17 Bug 68: The spamd_address main option now supports an optional
286 timeout value per server.
288 JH/18 Bug 1581: Router and transport options headers_add/remove can
289 now have the list separator specified.
291 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
294 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
297 JH/21 Support for the A6 type of dns record is withdrawn.
299 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
300 rather than the verbs used.
302 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
303 from 255 to 1024 chars.
305 JH/24 Verification callouts now attempt to use TLS by default.
307 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
308 are generic router options now. The defaults didn't change.
310 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
311 Original patch from Alexander Shikoff, worked over by JH.
313 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
314 files if ZCAT_COMMAND is not executable.
316 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
318 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
320 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
321 Normally benign, it bites when the pair was led to by a CNAME;
322 modern usage is to not canoicalize the domain to a CNAME target
323 (and we were inconsistent anyway for A-only vs AAAA+A).
325 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
327 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
328 when evaluating $sender_host_dnssec.
330 JH/31 Check the HELO verification lookup for DNSSEC, adding new
331 $sender_helo_dnssec variable.
333 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
335 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
337 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
339 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
340 documented as working, but never had. Support all but $spam_report.
342 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
343 added for tls authenticator.
345 HS/03 Add perl_taintmode main config option
350 TL/01 When running the test suite, the README says that variables such as
351 no_msglog_check are global and can be placed anywhere in a specific
352 test's script, however it was observed that placement needed to be near
353 the beginning for it to behave that way. Changed the runtest perl
354 script to read through the entire script once to detect and set these
355 variables, reset to the beginning of the script, and then run through
356 the script parsing/test process like normal.
358 TL/02 The BSD's have an arc4random API. One of the functions to induce
359 adding randomness was arc4random_stir(), but it has been removed in
360 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
361 function when detected.
363 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
364 cause callback expansion.
366 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
367 syntax errors in an expansion can be treated as a string instead of
368 logging or causing an error, due to the internal use of bool_lax
369 instead of bool when processing it.
371 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
372 server certificates when making smtp deliveries.
374 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
376 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
378 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
380 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
381 Merged patch from Sebastian Wiedenroth.
383 JH/05 Fix results-pipe from transport process. Several recipients, combined
384 with certificate use, exposed issues where response data items split
385 over buffer boundaries were not parsed properly. This eventually
386 resulted in duplicates being sent. This issue only became common enough
387 to notice due to the introduction of conection certificate information,
388 the item size being so much larger. Found and fixed by Wolfgang Breyha.
390 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
391 size buffer was used, resulting in syntax errors when an expansion
394 JH/07 Add support for directories of certificates when compiled with a GnuTLS
395 version 3.3.6 or later.
397 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
398 is EXPERIMENTAL_EVENT, the main-configuration and transport options
399 both become "event_action", the variables become $event_name, $event_data
400 and $event_defer_errno. There is a new variable $verify_mode, usable in
401 routers, transports and related events. The tls:cert event is now also
402 raised for inbound connections, if the main configuration event_action
405 TL/06 In test suite, disable OCSP for old versions of openssl which contained
406 early OCSP support, but no stapling (appears to be less than 1.0.0).
408 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
409 server certificate names available under the smtp transport option
410 "tls_verify_cert_hostname" now do not permit multi-component wildcard
413 JH/10 Time-related extraction expansions from certificates now use the main
414 option "timezone" setting for output formatting, and are consistent
415 between OpenSSL and GnuTLS compilations. Bug 1541.
417 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
418 encoded parameter in the incoming message. Bug 1558.
420 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
421 include certificate info, eximon was claiming there were spoolfile
424 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
426 JH/14 Log delivery-related information more consistently, using the sequence
427 "H=<name> [<ip>]" wherever possible.
429 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
430 are problematic for Debian distribution, omit them from the release
433 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
435 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
438 JH/17 Fix a null-indirection in certextract expansions when a nondefault
439 output list separator was used.
444 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
445 checkers that were complaining about end of non-void function with no
448 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
449 This was a regression intruduced in 4.83 by another bugfix.
451 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
453 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
454 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
460 TF/01 Correctly close the server side of TLS when forking for delivery.
462 When a message was received over SMTP with TLS, Exim failed to clear up
463 the incoming connection properly after forking off the child process to
464 deliver the message. In some situations the subsequent outgoing
465 delivery connection happened to have the same fd number as the incoming
466 connection previously had. Exim would try to use TLS and fail, logging
467 a "Bad file descriptor" error.
469 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
470 utilities have not been installed.
472 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
473 temporary space as the ACL may create new global variables.
475 TL/01 LDAP support uses per connection or global context settings, depending
476 upon the detected version of the libraries at build time.
478 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
479 to extract and use the src ip:port in logging and expansions as if it
480 were a direct connection from the outside internet. PPv2 support was
481 updated based on HAProxy spec change in May 2014.
483 JH/02 Add ${listextract {number}{list}{success}{fail}}.
485 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
486 Properly escape header and check for NULL return.
488 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
491 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
493 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
494 characters in header names, implemented as a verify condition.
495 Contributed by Michael Fischer v. Mollard.
497 TL/05 Rename SPF condition results err_perm and err_temp to standardized
498 results permerror and temperror. Previous values are deprecated but
499 still accepted. In a future release, err_perm and err_temp will be
500 completely removed, which will be a backward incompatibility if the
501 ACL tests for either of these two old results. Patch contributed by
502 user bes-internal on the mailing list.
504 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
506 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
507 selectors, in both main and reject logs.
509 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
512 JH/07 Add malware type "sock" for talking to simple daemon.
514 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
516 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
517 routers/transports under cutthrough routing.
519 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
520 numbers. Touch up "bool" conditional to keep the same definition.
522 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
524 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
526 JH/12 Expand items in router/transport headers_add or headers_remove lists
527 individually rather than the list as a whole. Bug 1452.
529 Required for reasonable handling of multiple headers_ options when
530 they may be empty; requires that headers_remove items with embedded
531 colons must have them doubled (or the list-separator changed).
533 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
534 view the policy declared in the DMARC record. Currently, $dmarc_status
535 is a combined value of both the record presence and the result of the
538 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
540 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
541 dnslookup router and the smtp transport (applying to the forward
544 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
545 of ldap servers used for a specific lookup. Patch provided by Heiko
548 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
549 New variable $lookup_dnssec_authenticated for observability.
551 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
552 Patch submitted by Lars Timman.
554 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
556 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
557 Requires trusted mode and valid format message id, aborts otherwise.
558 Patch contributed by Heiko Schlichting.
560 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
561 certextract with support for various fields. Bug 1358.
563 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
564 is requested by default, modifiable by smtp transport option
567 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
568 operate on certificate variables to give certificate fingerprints
569 Also new ${sha256:cert_variable}.
571 JH/23 The PRDR feature is moved from being Experimental into the mainline.
573 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
574 Christian Aistleitner.
576 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
578 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
579 file. Patch from Wolfgang Breyha.
581 JH/25 Expand the coverage of the delivery $host and $host_address to
582 client authenticators run in verify callout. Bug 1476.
584 JH/26 Port service names are now accepted for tls_on_connect_ports, to
585 align with daemon_smtp_ports. Bug 72.
587 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
588 support and error reporting did not work properly.
590 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
591 and is readable. Patch from Andrew Colin Kissa.
593 TL/14 Enhance documentation of ${run expansion and how it parses the
594 commandline after expansion, particularly in the case when an
595 unquoted variable expansion results in an empty value.
597 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
599 PP/02 Fix internal collision of T_APL on systems which support RFC3123
600 by renaming away from it. Addresses GH issue 15, reported by
603 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
605 TL/15 SECURITY: prevent double expansion in math comparison functions
606 (can expand unsanitized data). Not remotely exploitable.
613 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
615 PP/02 Make -n do something, by making it not do something.
616 When combined with -bP, the name of an option is not output.
618 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
621 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
622 $sender_host_name and config options to manage this, and basic check
625 PP/05 DSCP support for outbound connections and control modifier for inbound.
627 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
628 (Only plugin which currently uses this is kerberos4, which nobody should
629 be using, but we should make it available and other future plugins might
630 conceivably use it, even though it would break NAT; stuff *should* be
631 using channel bindings instead).
633 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
634 name; added for Sendmail compatibility; requires admin caller.
635 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
636 just ignore it); requires trusted caller.
637 Also parse but ignore: -Ac -Am -X<logfile>
640 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
642 TL/02 Add +smtp_confirmation as a default logging option.
644 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
645 Patch by Magnus Holmgren from 2007-02-20.
647 TL/04 Bugzilla 1281 - Spec typo.
648 Bugzilla 1283 - Spec typo.
649 Bugzilla 1290 - Spec grammar fixes.
651 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
653 TL/06 Add Experimental DMARC support using libopendmarc libraries.
655 TL/07 Fix an out of order global option causing a segfault. Reported to dev
656 mailing list by by Dmitry Isaikin.
658 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
660 JH/02 Support "G" suffix to numbers in ${if comparisons.
662 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
664 NM/01 Bugzilla 1197 - Spec typo
665 Bugzilla 1196 - Spec examples corrections
667 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
669 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
670 gnutls_enable_pkcs11, but renamed to more accurately indicate its
673 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
674 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
676 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
677 "acl {{name}{arg}...}", and optional args on acl condition
680 JH/05 Permit multiple router/transport headers_add/remove lines.
682 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
684 JH/07 Avoid using a waiting database for a single-message-only transport.
685 Performance patch from Paul Fisher. Bugzilla 1262.
687 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
690 JH/09 Add $headers_added variable, with content from use of ACL modifier
691 add_header (but not yet added to the message). Bugzilla 199.
693 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
694 Pulled from Bugzilla 817 by Wolfgang Breyha.
696 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
698 (nb: this is the same fix as in Exim 4.80.1)
700 JH/11 Add A= logging on delivery lines, and a client_set_id option on
703 JH/12 Add optional authenticated_sender logging to A= and a log_selector
706 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
708 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
709 advertise SMTP AUTH mechanism to us, instead of a generic
710 protocol violation error. Also, make Exim more robust to bad
711 data from the Dovecot auth socket.
713 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
715 When a queue runner is handling a message, Exim first routes the
716 recipient addresses, during which it prunes them based on the retry
717 hints database. After that it attempts to deliver the message to
718 any remaining recipients. It then updates the hints database using
721 So if a recipient address works intermittently, it can get repeatedly
722 deferred at routing time. The retry hints record remains fresh so the
723 address never reaches the final cutoff time.
725 This is a fairly common occurrence when a user is bumping up against
726 their storage quota. Exim had some logic in its local delivery code
727 to deal with this. However it did not apply to per-recipient defers
728 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
730 This change adds a proper retry rule check during routing so that the
731 final cutoff time is checked against the message's age. We only do
732 this check if there is an address retry record and there is not a
733 domain retry record; this implies that previous attempts to handle
734 the address had the retry_use_local_parts option turned on. We use
735 this as an approximation for the destination being like a local
736 delivery, as in LMTP.
738 I suspect this new check makes the old local delivery cutoff check
739 redundant, but I have not verified this so I left the code in place.
741 TF/02 Correct gecos expansion when From: is a prefix of the username.
743 Test 0254 submits a message to Exim with the header
747 When I ran the test suite under the user fanf2, Exim expanded
748 the header to contain my full name, whereas it should have added
749 a Resent-Sender: header. It erroneously treats any prefix of the
750 username as equal to the username.
752 This change corrects that bug.
754 GF/01 DCC debug and logging tidyup
755 Error conditions log to paniclog rather than rejectlog.
756 Debug lines prefixed by "DCC: " to remove any ambiguity.
758 TF/03 Avoid unnecessary rebuilds of lookup-related code.
760 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
761 Bug spotted by Jeremy Harris; was flawed since initial commit.
762 Would have resulted in OCSP responses post-SNI triggering an Exim
763 NULL dereference and crash.
765 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
767 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
768 Bug detection, analysis and fix by Samuel Thibault.
769 Bugzilla 1331, Debian bug #698092.
771 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
773 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
774 Server implementation by Todd Lyons, client by JH.
775 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
776 config variable "prdr_enable" controls whether the server
777 advertises the facility. If the client requests PRDR a new
778 acl_data_smtp_prdr ACL is called once for each recipient, after
779 the body content is received and before the acl_smtp_data ACL.
780 The client is controlled by bolth of: a hosts_try_prdr option
781 on the smtp transport, and the server advertisement.
782 Default client logging of deliveries and rejections involving
783 PRDR are flagged with the string "PRDR".
785 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
786 fclose(). Diagnosis by Todd Lyons.
788 PP/17 Update configure.default to handle IPv6 localhost better.
789 Patch by Alain Williams (plus minor tweaks).
792 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
793 This is now consistent with GnuTLS, and is now documented: the
794 previous undocumented portable approach to treating the option as
795 unset was to force an expansion failure. That still works, and
796 an empty string is now equivalent.
798 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
799 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
800 not performing validation itself.
802 PP/20 Added force_command boolean option to pipe transport.
803 Patch from Nick Koston, of cPanel Inc.
805 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
808 TF/04 Added udpsend ACL modifer and hexquote expansion operator
810 PP/21 Fix eximon continuous updating with timestamped log-files.
811 Broken in a format-string cleanup in 4.80, missed when I repaired the
812 other false fix of the same issue.
813 Report and fix from Heiko Schlichting.
816 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
817 Report from Prashanth Katuri.
819 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
820 It's SecureTransport, so affects any MacOS clients which use the
821 system-integrated TLS libraries, including email clients.
823 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
824 using a MIME ACL for non-SMTP local injection.
825 Report and assistance in diagnosis by Warren Baker.
827 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
829 JH/16 Fix comparisons for 64b. Bugzilla 1385.
831 TL/09 Add expansion variable $authenticated_fail_id to keep track of
832 last id that failed so it may be referenced in subsequent ACL's.
834 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
837 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
838 ldap library initialization, allowing self-signed CA's to be
839 used. Also properly sets require_cert option later in code by
840 using NULL (global ldap config) instead of ldap handle (per
841 session). Bug diagnosis and testing by alxgomz.
843 TL/12 Enhanced documentation in the ratelimit.pl script provided in
844 the src/util/ subdirectory.
846 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
847 renamed to Transport Post Delivery Action by Jeremy Harris, as
850 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
851 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
852 redis_servers = needs to be configured which will be used by the redis
853 lookup. Patch from Warren Baker, of The Packet Hub.
855 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
857 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
858 hostname or reverse DNS when processing a host list. Used suggestions
859 from multiple comments on this bug.
861 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
863 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
864 Missed a few lines, added it to make the runtest require no keyboard
867 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
868 contains upper case chars. Make router use caseful_local_part.
870 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
871 support when GnuTLS has been built with p11-kit.
877 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
879 This, or similar/improved, will also be change PP/11 of 4.82.
885 PP/01 Handle short writes when writing local log-files.
886 In practice, only affects FreeBSD (8 onwards).
887 Bugzilla 1053, with thanks to Dmitry Isaikin.
889 NM/01 Bugzilla 949 - Documentation tweak
891 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
894 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
896 PP/02 Implemented gsasl authenticator.
898 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
900 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
901 `pkg-config foo` for cflags/libs.
903 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
904 with rest of GSASL and with heimdal_gssapi.
906 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
907 `pkg-config foo` for cflags/libs for the TLS implementation.
909 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
910 properties get this fed in as external SSF. A number of robustness
911 and debugging improvements to the cyrus_sasl authenticator.
913 PP/08 cyrus_sasl server now expands the server_realm option.
915 PP/09 Bugzilla 1214 - Log authentication information in reject log.
916 Patch by Jeremy Harris.
918 PP/10 Added dbmjz lookup type.
920 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
922 PP/12 MAIL args handles TAB as well as SP, for better interop with
923 non-compliant senders.
924 Analysis and variant patch by Todd Lyons.
926 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
927 Bug report from Lars Müller <lars@samba.org> (via SUSE),
928 Patch from Dirk Mueller <dmueller@suse.com>
930 PP/13 tls_peerdn now print-escaped for spool files.
931 Observed some $tls_peerdn in wild which contained \n, which resulted
932 in spool file corruption.
934 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
935 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
936 or write after TLS renegotiation, which otherwise led to messages
939 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
940 as a tracking header (ie: a signed header comes before the signature).
941 Patch from Wolfgang Breyha.
943 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
944 comma-sep list; embedded commas doubled.
946 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
948 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
950 Report and patch from Dmitry Banschikov.
952 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
953 Removed SSL_clear() after SSL_new() which led to protocol negotiation
954 failures. We appear to now support TLS1.1+ with Exim.
956 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
957 lets Exim select keys and certificates based upon TLS SNI from client.
958 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
959 before an outbound SMTP session. New log_selector, +tls_sni.
961 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
962 NULL dereference. Report and patch from Alun Jones.
964 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
965 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
966 Not seeing resolver debug output on NetBSD, but suspect this is a
967 resolver implementation change.
969 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
970 Left warnings. Added "eximon gdb" invocation mode.
972 PP/21 Defaulting "accept_8bitmime" to true, not false.
974 PP/22 Added -bw for inetd wait mode support.
976 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
977 locate the relevant includes and libraries. Made this the default.
979 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
980 Bugzilla 1246, report and most of solution from Tomasz Kusy.
982 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
983 This may cause build issues on older platforms.
985 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
986 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
987 gnutls_require_mac & gnutls_require_protocols (no longer supported).
988 Added SNI support via GnuTLS too.
989 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
991 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
993 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
994 Applied second patch from Janne, implementing suggestion to default
995 multiple-strings-in-record handling to match SPF spec.
997 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
999 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1000 read-only, out of scope).
1001 Patch from Wolfgang Breyha, report from Stuart Northfield.
1003 PP/29 Fix three issues highlighted by clang analyser static analysis.
1004 Only crash-plausible issue would require the Cambridge-specific
1005 iplookup router and a misconfiguration.
1006 Report from Marcin Mirosław.
1008 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1010 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1011 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1012 As part of this, removing so much warning spew let me fix some minor
1013 real issues in debug logging.
1015 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1016 assignment on my part. Fixed.
1018 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1019 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1020 Janne Snabb (who went above and beyond: thank you).
1022 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1023 string otherwise requires a connection and a bunch more work and it's
1024 relatively easy to get wrong. Should also expose TLS library linkage
1027 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1028 64-bit ${eval} (JH/03).
1030 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1031 GNU libc to support some of the 64-bit stuff, should not lead to
1032 conflicts. Defined before os.h is pulled in, so if a given platform
1033 needs to override this, it can.
1035 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1036 protection layer was required, which is not implemented.
1037 Bugzilla 1254, patch from Wolfgang Breyha.
1039 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1040 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1041 tls_dhparam take prime identifiers. Also unbreak combination of
1042 OpenSSL+DH_params+TLSSNI.
1044 PP/39 Disable SSLv2 by default in OpenSSL support.
1050 PP/01 Solaris build fix for Oracle's LDAP libraries.
1051 Bugzilla 1109, patch from Stephen Usher.
1053 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1055 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1058 TF/02 Fix a couple more cases where we did not log the error message
1059 when unlink() failed. See also change 4.74-TF/03.
1061 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1062 lock up or crash if it happened to be inside a call to libc when it
1063 got a SIGUSR1 from exiwhat.
1065 The SIGUSR1 handler appends the current process status to the process
1066 log which is later printed by exiwhat. It used to use the general
1067 purpose logging code to do this, but several functions it calls are
1068 not safe for signals.
1070 The new output code in the SIGUSR1 handler is specific to the process
1071 log, and simple enough that it's easy to inspect for signal safety.
1072 Removing some special cases also simplifies the general logging code.
1073 Removing the spurious timestamps from the process log simplifies
1076 TF/04 Improved ratelimit ACL condition.
1078 The /noupdate option has been deprecated in favour of /readonly which
1079 has clearer semantics. The /leaky, /strict, and /readonly update modes
1080 are mutually exclusive. The update mode is no longer included in the
1081 database key; it just determines when the database is updated. (This
1082 means that when you upgrde Exim will forget old rate measurements.)
1084 Exim now checks that the per_* options are used with an update mode that
1085 makes sense for the current ACL. For example, when Exim is processing a
1086 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1087 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1088 must specify per_mail/readonly. If you omit the update mode it defaults to
1089 /leaky where that makes sense (as before) or /readonly where required.
1091 The /noupdate option is now undocumented but still supported for
1092 backwards compatibility. It is equivalent to /readonly except that in
1093 ACLs where /readonly is required you may specify /leaky/noupdate or
1094 /strict/noupdate which are treated the same as /readonly.
1096 A useful new feature is the /count= option. This is a generalization
1097 of the per_byte option, so that you can measure the throughput of other
1098 aggregate values. For example, the per_byte option is now equivalent
1099 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1101 The per_rcpt option has been generalized using the /count= mechanism
1102 (though it's more complicated than the per_byte equivalence). When it is
1103 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1104 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1105 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1106 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1107 Note that using per_rcpt with a non-readonly update mode in more than
1108 one ACL will cause the recipients to be double-counted. (The per_mail
1109 and per_byte options don't have this problem.)
1111 The handling of very low rates has changed slightly. If the computed rate
1112 is less than the event's count (usually one) then this event is the first
1113 after a long gap. In this case the rate is set to the same as this event's
1114 count, so that the first message of a spam run is counted properly.
1116 The major new feature is a mechanism for counting the rate of unique
1117 events. The new per_addr option counts the number of different
1118 recipients that someone has sent messages to in the last time period. It
1119 behaves like per_rcpt if all the recipient addresses are different, but
1120 duplicate recipient addresses do not increase the measured rate. Like
1121 the /count= option this is a general mechanism, so the per_addr option
1122 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1123 example, measure the rate that a client uses different sender addresses
1124 with the options per_mail/unique=$sender_address. There are further
1125 details in the main documentation.
1127 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1129 TF/06 Removed a few PCRE remnants.
1131 TF/07 Automatically extract Exim's version number from tags in the git
1132 repository when doing development or release builds.
1134 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1135 Bugzilla 879. Patch from Paul Fisher.
1137 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1138 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1141 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1143 PP/05 Variable $av_failed, true if the AV scanner deferred.
1144 Bugzilla 1078. Patch from John Horne.
1146 PP/06 Stop make process more reliably on build failure.
1147 Bugzilla 1087. Patch from Heiko Schlittermann.
1149 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1150 Bugzilla 1089. Patch from Heiko Schlittermann.
1152 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1153 Bugzilla 1131. Patch from Holger Weiß.
1155 PP/09 Handle IPv6 addresses with SPF.
1156 Bugzilla 860. Patch from Wolfgang Breyha.
1158 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1160 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1163 PP/11 match_* no longer expand right-hand-side by default.
1164 New compile-time build option, EXPAND_LISTMATCH_RHS.
1165 New expansion conditions, "inlist", "inlisti".
1167 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1169 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1171 PP/14 fix log_write() format string regression from TF/03.
1172 Bugzilla 1152. Patch from Dmitry Isaikin.
1178 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1180 PP/02 Harmonised TLS library version reporting; only show if debugging.
1181 Layout now matches that introduced for other libraries in 4.74 PP/03.
1183 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1185 PP/04 New "dns_use_edns0" global option.
1187 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1190 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1191 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1193 TK/01 Updated PolarSSL code to 0.14.2.
1194 Bugzilla 1097. Patch from Andreas Metzler.
1196 PP/07 Catch divide-by-zero in ${eval:...}.
1197 Fixes bugzilla 1102.
1199 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1202 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1203 format-string attack -- SECURITY: remote arbitrary code execution.
1205 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1206 time unintentionally subject to list matching rules, letting the header
1207 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1208 arbitrary string expansion). This allowed for information disclosure.
1210 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1211 INT_MIN/-1 -- value coerced to INT_MAX.
1217 NM/01 Workround for PCRE version dependency in version reporting
1220 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1221 This fixes portability to compilers other than gcc, notably
1222 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1224 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1225 makefiles for portability to HP-UX and POSIX correctness.
1227 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1228 Also via indented variable definition in the Makefile.
1229 (Debugging by Oliver Heesakkers).
1231 PP/02 Restore caching of spamd results with expanded spamd_address.
1232 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1234 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1235 Improves build reliability. Fix from: Frank Elsner
1237 NM/02 Fix wide character breakage in the rfc2047 coding
1238 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1240 NM/03 Allow underscore in dnslist lookups
1241 Fixes bug 1026. Patch from Graeme Fowler
1243 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1244 Code patches from Adam Ciarcinski of NetBSD.
1246 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1249 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1250 is logged, to avoid truncation. Patch from John Horne.
1252 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1253 Patch from Jakob Hirsch.
1255 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1256 SQL string expansion failure details.
1257 Patch from Andrey Oktyabrski.
1259 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1260 Patch from Simon Arlott.
1262 PP/09 New lookups functionality failed to compile on old gcc which rejects
1263 extern declarations in function scope.
1264 Patch from Oliver Fleischmann
1266 PP/10 Use sig_atomic_t for flags set from signal handlers.
1267 Check getgroups() return and improve debugging.
1268 Fixed developed for diagnosis in bug 927 (which turned out to be
1271 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1272 Patch from Mark Zealey.
1274 PP/12 Bugzilla 1056: Improved spamd server selection.
1275 Patch from Mark Zealey.
1277 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1278 Based on patch from Heiko Schlittermann.
1280 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1281 Patch from Uwe Doering, sign-off by Michael Haardt.
1283 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1284 variable declaration deep within a block. Bug and patch from
1287 PP/15 lookups-Makefile IRIX compatibilty coercion.
1289 PP/16 Make DISABLE_DKIM build knob functional.
1291 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1292 Patch by Simon Arlott
1294 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1295 variable argument macros. Our copy now differs from upstream.
1301 TF/01 Failure to get a lock on a hints database can have serious
1302 consequences so log it to the panic log.
1304 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1305 controlled using the smtp_confirmation log selector.
1307 TF/03 Include the error message when we fail to unlink a spool file.
1309 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1310 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1311 for maintaining out-of-tree patches for some time.
1313 PP/01 Bugzilla 139: Documentation and portability issues.
1314 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1315 Handle per-OS dynamic-module compilation flags.
1317 PP/02 Let /dev/null have normal permissions.
1318 The 4.73 fixes were a little too stringent and complained about the
1319 permissions on /dev/null. Exempt it from some checks.
1320 Reported by Andreas M. Kirchwitz.
1322 PP/03 Report version information for many libraries, including
1323 Exim version information for dynamically loaded libraries. Created
1324 version.h, now support a version extension string for distributors
1325 who patch heavily. Dynamic module ABI change.
1327 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1328 privilege escalation vulnerability whereby the Exim run-time user
1329 can cause root to append content of the attacker's choosing to
1332 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1335 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1336 If dropping privileges for untrusted macros, we disabled normal logging
1337 on the basis that it would fail; for the Exim run-time user, this is not
1338 the case, and it resulted in successful deliveries going unlogged.
1339 Fixed. Reported by Andreas Metzler.
1345 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1346 only prepend for the Resent-* case. Fixes regression introduced in
1347 Exim 4.70 by NM/22 for Bugzilla 607.
1349 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1350 increasing numbers of administrators be bitten by this.
1352 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1354 PP/03 Bugzilla 994: added openssl_options main configuration option.
1356 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1358 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1360 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1362 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1363 without a peer certificate, leading to a segfault because of an
1364 assumption that peers always have certificates. Be a little more
1365 paranoid. Problem reported by Martin Tscholak.
1367 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1368 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1369 NB: ClamAV planning to remove STREAM in "middle of 2010".
1370 CL also introduces -bmalware, various -d+acl logging additions and
1371 more caution in buffer sizes.
1373 PP/09 Implemented reverse_ip expansion operator.
1375 PP/10 Bugzilla 937: provide a "debug" ACL control.
1377 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1379 PP/12 Bugzilla 973: Implement --version.
1381 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1383 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1385 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1387 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1388 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1389 ignore trailing whitespace.
1391 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1393 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1396 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1397 Notification from Dr Andrew Aitchison.
1399 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1400 ExtendedDetectionInfo response format.
1401 Notification from John Horne.
1403 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1406 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1407 XSL and documented dependency on system catalogs, with examples of how
1410 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1413 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1414 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1415 configuration file which is writeable by the Exim user or group.
1417 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1418 of configuration files to cover files specified with the -C option if
1419 they are going to be used with root privileges, not just the default
1422 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1423 option (effectively making it always true).
1425 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1426 files to be used while preserving root privileges.
1428 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1429 that rogue child processes cannot use them.
1431 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1432 run-time user, instead of root.
1434 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1435 Exim run-time user without dropping privileges.
1437 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1438 result string, instead of calling string_vformat() twice with the same
1441 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1442 for other users. Others should always drop root privileges if they use
1443 -C on the command line, even for a whitelisted configure file.
1445 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1447 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1453 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1454 $data_path, and $header_path variables; fixed documentation bugs and
1457 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1458 exipick to access non-standard spools, including the "frozen" queue
1461 NM/01 Bugzilla 965: Support mysql stored procedures.
1462 Patch from Alain Williams
1464 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1466 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1467 Patch from Andreas Metzler
1469 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1470 Patch from Kirill Miazine
1472 NM/05 Bugzilla 671: Added umask to procmail example.
1474 JJ/03 installed exipick 20100323.0, fixing doc bug
1476 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1477 directory. Notification and patch from Dan Rosenberg.
1479 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1481 TK/02 Improve log output when DKIM signing operation fails.
1483 MH/01 Treat the transport option dkim_domain as a colon separated
1484 list, not as a single string, and sign the message with each element,
1485 omitting multiple occurences of the same signer.
1487 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1488 Bugzilla 985, 986. Patch by Simon Arlott
1490 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1491 Patch by Simon Arlott
1493 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1494 MBX locking. Notification from Dan Rosenberg.
1500 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1502 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1504 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1506 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1508 NM/04 Bugzilla 915: Flag broken perl installation during build.
1514 TK/01 Added patch by Johannes Berg that expands the main option
1515 "spamd_address" if it starts with a dollar sign.
1517 TK/02 Write list of recipients to X-Envelope-Sender header when building
1518 the mbox-format spool file for content scanning (suggested by Jakob
1521 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1522 (http://www.dcc-servers.net/) support via dccifd. Activated by
1523 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1525 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1526 by Mark Daniel Reidel <mr@df.eu>.
1528 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1529 When building exim an external PCRE library is now needed -
1530 PCRE is a system library on the majority of modern systems.
1531 See entry on PCRE_LIBS in EDITME file.
1533 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1534 conversation. Added nologin parameter to request.
1535 Patch contributed by Kirill Miazine.
1537 TF/01 Do not log submission mode rewrites if they do not change the address.
1539 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1541 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1542 log files in place. Contributed by Roberto Lima.
1544 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1546 TF/03 Bugzilla 615: When checking the local_parts router precondition
1547 after a local_part_suffix or local_part_prefix option, Exim now
1548 does not use the address's named list lookup cache, since this
1549 contains cached lookups for the whole local part.
1551 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1552 Robert Millan. Documentation is in experimental-spec.txt.
1554 TF/04 Bugzilla 668: Fix parallel build (make -j).
1556 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1558 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1559 Patch provided by Jan Srzednicki.
1561 TF/05 Leading white space used to be stripped from $spam_report which
1562 wrecked the formatting. Now it is preserved.
1564 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1565 that they are available at delivery time.
1567 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1569 TF/08 TLS error reporting now respects the incoming_interface and
1570 incoming_port log selectors.
1572 TF/09 Produce a more useful error message if an SMTP transport's hosts
1573 setting expands to an empty string.
1575 NM/06 Bugzilla 744: EXPN did not work under TLS.
1576 Patch provided by Phil Pennock.
1578 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1579 Patch provided by Richard Godbee.
1581 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1582 acl_smtp_notquit, added index entry.
1584 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1585 Patch provided by Eugene Bujak.
1587 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1588 accept(). Patch provided by Maxim Dounin.
1590 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1591 Patch provided by Phil Pennock.
1593 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1595 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1596 Patch provided by Brad "anomie" Jorsch.
1598 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1599 Patch provided by Dean Brooks.
1601 TK/05 Add native DKIM support (does not depend on external libraries).
1603 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1604 Patch provided by Graeme Fowler.
1606 NM/16 Bugzilla 851: Documentation example syntax fix.
1608 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1610 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1613 NM/19 Bugzilla 745: TLS version reporting.
1614 Patch provided by Phil Pennock.
1616 NM/20 Bugzilla 167: bool: condition support.
1617 Patch provided by Phil Pennock.
1619 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1620 clients. Patch provided by Phil Pennock.
1622 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1623 Patch provided by Brad "anomie" Jorsch.
1625 NM/23 Bugzilla 687: Fix misparses in eximstats.
1626 Patch provided by Heiko Schlittermann.
1628 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1629 Patch provided by Heiko Schlittermann.
1631 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1632 plus update to original patch.
1634 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1636 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1637 Patch provided by David Brownlee.
1639 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1641 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1643 NM/30 Bugzilla 888: TLS documentation bugfixes.
1645 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1647 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1648 Unlike the original bugzilla I have changed all shell scripts in src tree.
1650 NM/33 Bugzilla 898: Transport filter timeout fix.
1651 Patch by Todd Rinaldo.
1653 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1654 Patch by Serge Demonchaux.
1656 NM/35 Bugzilla 39: Base64 decode bug fixes.
1657 Patch by Jakob Hirsch.
1659 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1661 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1663 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1665 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1671 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1672 ALT-N's libdkim that I have put here:
1673 http://duncanthrax.net/exim-experimental/
1675 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1676 were called 'true' and it seems that C99 defines that as a reserved
1677 keyword to be used with 'bool' variable types. That means you could
1678 not include C99-style headers which use bools without triggering
1679 build errors in sieve.c.
1681 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1682 as mailq or other aliases. Changed the --help handling significantly
1683 to do whats expected. exim_usage() emits usage/help information.
1685 SC/01 Added the -bylocaldomain option to eximstats.
1687 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1689 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1691 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1697 PH/01 Another patch from the Sieve maintainer.
1699 PH/02 When an IPv6 address is converted to a string for single-key lookup
1700 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1701 dots are used instead of colons so that keys in lsearch files need not
1702 contain colons. This was done some time before quoting was made available
1703 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1704 (notated using the quote facility) so as to distinguish them from IPv4
1705 keys. This meant that lookups for IP addresses in host lists did not work
1706 for iplsearch lookups.
1708 This has been fixed by arranging for IPv6 addresses to be expressed with
1709 colons if the lookup type is iplsearch. This is not incompatible, because
1710 previously such lookups could never work.
1712 The situation is now rather anomolous, since one *can* have colons in
1713 ordinary lsearch keys. However, making the change in all cases is
1714 incompatible and would probably break a number of configurations.
1716 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1719 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1720 conversion specification without a maximum field width, thereby enabling
1721 a rogue spamd server to cause a buffer overflow. While nobody in their
1722 right mind would setup Exim to query an untrusted spamd server, an
1723 attacker that gains access to a server running spamd could potentially
1724 exploit this vulnerability to run arbitrary code as the Exim user.
1726 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1727 $primary_hostname instead of what libspf2 thinks the hosts name is.
1729 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1730 a directory entry by the name of the lookup key. Previously, if a
1731 symlink pointed to a non-existing file or a file in a directory that
1732 Exim lacked permissions to read, a lookup for a key matching that
1733 symlink would fail. Now it is enough that a matching directory entry
1734 exists, symlink or not. (Bugzilla 503.)
1736 PH/03 The body_linecount and body_zerocount variables are now exported in the
1739 PH/04 Added the $dnslist_matched variable.
1741 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1742 This means they are set thereafter only if the connection becomes
1745 PH/06 Added the client_condition to authenticators so that some can be skipped
1746 by clients under certain conditions.
1748 PH/07 The error message for a badly-placed control=no_multiline_responses left
1749 "_responses" off the end of the name.
1751 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1753 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1754 (without spaces) instead of just copying the configuration text.
1756 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1758 PH/11 Added $max_received_linelength.
1760 PH/12 Added +ignore_defer and +include_defer to host lists.
1762 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1763 way in which PCRE > 7.0 is built.
1765 PH/14 Implemented queue_only_load_latch.
1767 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1768 MAIL command. The effect was to mangle the value on 64-bit systems.
1770 PH/16 Another patch from the Sieve maintainer.
1772 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1774 PH/18 If a system quota error occurred while trying to create the file for
1775 a maildir delivery, the message "Mailbox is full" was not appended to the
1776 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1777 only to a quota excession during the actual writing of the file.
1779 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1780 characters?) which causes problems in log lines. The DN values are now
1781 passed through string_printing() before being added to log lines.
1783 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1784 and InterBase are left for another time.)
1786 PH/21 Added message_body_newlines option.
1788 PH/22 Guard against possible overflow in moan_check_errorcopy().
1790 PH/23 POSIX allows open() to be a macro; guard against that.
1792 PH/24 If the recipient of an error message contained an @ in the local part
1793 (suitably quoted, of course), incorrect values were put in $domain and
1794 $local_part during the evaluation of errors_copy.
1800 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1801 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1804 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1805 issue a MAIL command.
1807 PH/02 In an ACL statement such as
1809 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1811 if a client was not listed at all, or was listed with a value other than
1812 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1813 the condition was not true (as it should be), so access was not denied.
1814 The bug was that the ! inversion was incorrectly passed on to the second
1815 item. This has been fixed.
1817 PH/03 Added additional dnslists conditions == and =& which are different from
1818 = and & when the dns lookup returns more than one IP address.
1820 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1821 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1823 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1824 FSYNC, which compiles an option called disable_fsync that allows for
1825 bypassing fsync(). The documentation is heavily laced with warnings.
1827 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1829 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1830 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1831 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1832 including adding "make clean"; (3) Added -fPIC when compiling the test
1833 dynamically loaded module, to get rid of a warning.
1835 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1836 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1837 = 0s. The bug is otherwise harmless.
1839 PH/07 There was a bug in the dovecot authenticator such that the value of
1840 $auth1 could be overwritten, and so not correctly preserved, after a
1841 successful authentication. This usually meant that the value preserved by
1842 the server_setid option was incorrect.
1844 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1846 PH/09 Installed PCRE release 7.0.
1848 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1849 run for batched SMTP input. It is now run at the start of every message
1850 in the batch. While fixing this I discovered that the process information
1851 (output by running exiwhat) was not always getting set for -bs and -bS
1852 input. This is fixed, and it now also says "batched" for BSMTP.
1854 PH/11 Added control=no_pipelining.
1856 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1857 patch, slightly modified), and move the expansion of helo_data till after
1858 the connection is made in the smtp transport (so it can use these
1861 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1863 PH/14 Added log_selector = +pid.
1865 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1867 PH/16 Add ${if forany and ${if forall.
1869 PH/17 Added dsn_from option to vary the From: line in DSNs.
1871 PH/18 Flush SMTP output before performing a callout, unless control =
1872 no_callout_flush is set.
1874 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1875 was true (the default) a successful delivery failed to delete the retry
1876 item, thus causing premature timeout of the address. The bug is now
1879 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1881 PH/21 Long custom messages for fakedefer and fakereject are now split up
1882 into multiline reponses in the same way that messages for "deny" and
1883 other ACL rejections are.
1885 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1886 with slight modification.
1888 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1889 draft, changing the syntax and factoring some duplicate code".
1891 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1892 for deliveries of the second and subsequent messages over the same SMTP
1895 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1896 ${reduce, with only minor "tidies".
1898 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1900 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1901 expansion side effects.
1903 PH/27 When a message times out after an over-quota error from an Exim-imposed
1904 quota, the bounce message says "mailbox is full". This message was not
1905 being given when it was a system quota that was exceeded. It now should
1908 MH/03 Made $recipients available in local_scan(). local_scan() already has
1909 better access to the recipient list through recipients_list[], but
1910 $recipients can be useful in postmaster-provided expansion strings.
1912 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1913 in the case of a MAIL command with additional options following the
1914 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1915 were accidentally chopped off.
1917 PH/29 SMTP synchronization checks are implemented when a command is read -
1918 there is a check that no more input is waiting when there shouldn't be
1919 any. However, for some commands, a delay in an ACL can mean that it is
1920 some time before the response is written. In this time, more input might
1921 arrive, invalidly. So now there are extra checks after an ACL has run for
1922 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1923 pipelining has not been advertised.
1925 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1927 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1928 returned in the SMTP response when smtp_return_error_details was false.
1929 This has been fixed.
1931 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1932 instead of the C I/O that was originally supplied, because problems were
1933 reported on Solaris.
1935 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1936 Exim which did not show up earlier: it was assuming that a call to
1937 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1938 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1939 was a macro that became an assignment, so it seemed to work. This has
1940 changed to a proper function call with a void return, hence the compile
1941 error. Exim's code has been fixed.
1943 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1946 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1948 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1950 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1951 because it turns out to be too big or there is a timeout) there is no
1952 "Completed" line in the log. When some messages of this type were
1953 selected by exigrep, they were listed as "not completed". Others were
1954 picked up by some special patterns. I have improved the selection
1955 criteria to be more general.
1957 PH/38 The host_find_failed option in the manualroute router can now be set
1958 to "ignore", to completely ignore a host whose IP address cannot be
1959 found. If all hosts are ignored, the behaviour is controlled by the new
1960 host_all_ignored option.
1962 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1963 homing or because of multiple MX records with /mx) generated more than
1964 one IP address, and the following item turned out to be the local host,
1965 all the secondary addresses of the first item were incorrectly removed
1966 from the list, along with the local host and any following hosts (which
1967 is what is supposed to happen).
1969 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1970 whoever called Exim into the -H file. In the case of the daemon it was
1971 behaving confusingly. When first started, it used values for whoever
1972 started the daemon, but after a SIGHUP it used the Exim user (because it
1973 calls itself on a restart). I have changed the code so that it now always
1976 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1977 message are rejected with the same error (e.g. no authentication or bad
1978 sender address), and a DATA command is nevertheless sent (as can happen
1979 with PIPELINING or a stupid MUA), the error message that was given to the
1980 RCPT commands is included in the rejection of the DATA command. This is
1981 intended to be helpful for MUAs that show only the final error to their
1984 PH/42 Another patch from the Sieve maintainer.
1986 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1987 Eximstats - Fixed some broken HTML links and added missing column headers
1989 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1990 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1992 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1994 PH/43 Yet another patch from the Sieve maintainer.
1996 PH/44 I found a way to check for a TCP/IP connection going away before sending
1997 the response to the final '.' that terminates a message, but only in the
1998 case where the client has not sent further data following the '.'
1999 (unfortunately, this is allowed). However, in many cases there won't be
2000 any further data because there won't be any more messages to send. A call
2001 to select() can be used: if it shows that the input is "ready", there is
2002 either input waiting, or the socket has been closed. An attempt to read
2003 the next input character can distinguish the two cases. Previously, Exim
2004 would have sent an OK response which the client would never have see.
2005 This could lead to message repetition. This fix should cure that, at
2006 least in a lot of common cases.
2008 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2009 advertised in response to EHLO.
2015 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2016 fixed by 4.65/MH/01 (is this a record?) are fixed:
2018 (i) An empty string was always treated as zero by the numeric comparison
2019 operators. This behaviour has been restored.
2021 (ii) It is documented that the numeric comparison operators always treat
2022 their arguments as decimal numbers. This was broken in that numbers
2023 starting with 0 were being interpreted as octal.
2025 While fixing these problems I realized that there was another issue that
2026 hadn't been noticed. Values of message_size_limit (both the global option
2027 and the transport option) were treated as octal if they started with 0.
2028 The documentation was vague. These values are now always treated as
2029 decimal, and I will make that clear in the documentation.
2035 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2036 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2039 MH/01 Don't check that the operands of numeric comparison operators are
2040 integers when their expansion is in "skipping" mode (fixes bug
2041 introduced by 4.64-PH/07).
2043 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2044 child addresses, Exim now panics and dies. Previously, because the count
2045 is held in a short int, deliveries were likely to be lost. As such a
2046 large number of recipients for a single message is ridiculous
2047 (performance will be very, very poor), I have chosen to impose a limit
2048 rather than extend the field.
2054 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2055 leftover -K file (the existence of which was triggered by #402).
2056 While we were at it, introduced process PID as part of the -K
2057 filename. This should rule out race conditions when creating
2060 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2061 processing considerably. Previous code took too long for large mails,
2062 triggering a timeout which in turn triggers #401.
2064 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2065 in the DK code in transports.c. sendfile() is not really portable,
2066 hence the _LINUX specificness.
2068 TF/01 In the add_headers option to the mail command in an Exim filter,
2069 there was a bug that Exim would claim a syntax error in any
2070 header after the first one which had an odd number of characters
2073 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2074 callout verification, Exim cached a "reject" for the entire domain. This
2075 is correct for most verifications, but it is not correct for a recipient
2076 verification with use_sender or use_postmaster set, because in that case
2077 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2078 case of MAIL FROM:<> rejection from other early rejections (e.g.
2079 rejection of HELO). When verifying a recipient using a non-null MAIL
2080 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2081 Whatever the result of the callout, the value of the domain cache is
2082 left unchanged (for any other kind of callout, getting as far as trying
2083 RCPT means that the domain itself is ok).
2085 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2088 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2089 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2090 session, the daemon ignores ECONNECT errors and logs others; it now
2091 ignores EPIPE as well.
2093 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2094 (quoted-printable decoding).
2096 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2097 later the small subsequent patch to fix an introduced bug.
2099 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2101 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2103 PH/08 An error is now given if message_size_limit is specified negative.
2105 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2106 to be given (somewhat) arbitrary names.
2108 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2111 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2112 miscellaneous code fixes
2114 PH/10 Added the log_reject_target ACL modifier to specify where to log
2117 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2118 hostname. This is wrong, because it relates to the incoming message (and
2119 probably the interface on which it is arriving) and not to the outgoing
2120 callout (which could be using a different interface). This has been
2121 changed to use the value of the helo_data option from the smtp transport
2122 instead - this is what is used when a message is actually being sent. If
2123 there is no remote transport (possible with a router that sets up host
2124 addresses), $smtp_active_hostname is used.
2126 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2127 tweaks were necessary in order to get it to work (see also 21 below):
2128 (a) The code assumed that strncpy() returns a negative number on buffer
2129 overflow, which isn't the case. Replaced with Exim's string_format()
2131 (b) There were several signed/unsigned issues. I just did the minimum
2132 hacking in of casts. There is scope for a larger refactoring.
2133 (c) The code used strcasecmp() which is not a standard C function.
2134 Replaced with Exim's strcmpic() function.
2135 (d) The code set only $1; it now sets $auth1 as well.
2136 (e) A simple test gave the error "authentication client didn't specify
2137 service in request". It would seem that Dovecot has changed its
2138 interface. Fortunately there's a specification; I followed it and
2139 changed what the client sends and it appears to be working now.
2141 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2144 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2145 address A is aliased to B and C, where B exists and C does not. Without
2146 -v the output is "A verified" because verification stops after a
2147 successful redirection if more than one address is generated. However,
2148 with -v the child addresses are also verified. Exim was outputting "A
2149 failed to verify" and then showing the successful verification for C,
2150 with its parentage. It now outputs "B failed to verify", showing B's
2151 parentage before showing the successful verification of C.
2153 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2154 look up a TXT record in a specific list after matching in a combined
2157 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2158 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2159 they consult the DNS. I had assumed they would set it the way they
2160 wanted; and indeed my experiments on Linux seem to show that in some
2161 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2162 To be on the safe side, however, I have now made the interface to
2163 host_find_byname() similar to host_find_bydns(), with an argument
2164 containing the DNS resolver options. The host_find_byname() function now
2165 sets these options at its start, just as host_find_bydns() does. The smtp
2166 transport options dns_qualify_single and dns_search_parents are passed to
2167 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2168 of host_find_byname() use the default settings of RES_DEFNAMES
2169 (qualify_single) but not RES_DNSRCH (search_parents).
2171 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2172 spool_read_header() do less string testing, by means of a preliminary
2173 switch on the second character of optional "-foo" lines. (This is
2174 overdue, caused by the large number of possibilities that now exist.
2175 Originally there were few.) While I was there, I also converted the
2176 str(n)cmp tests so they don't re-test the leading "-" and the first
2177 character, in the hope this might squeeze out yet more improvement.
2179 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2180 flag allowing group syntax was set by the header_syntax check but not
2181 turned off, possible causing trouble later; (2) The flag was not being
2182 set at all for the header_verify test, causing "group"-style headers to
2183 be rejected. I have now set it in this case, and also caused header_
2184 verify to ignore an empty address taken from a group. While doing this, I
2185 came across some other cases where the code for allowing group syntax
2186 while scanning a header line wasn't quite right (mostly, not resetting
2187 the flag correctly in the right place). These bugs could have caused
2188 trouble for malformed header lines. I hope it is now all correct.
2190 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2191 with the "reply" argument non-NULL. The code, however (which originally
2192 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2193 but it didn't always do it. This confused somebody who was copying the
2194 code for some other use. I have removed all the tests.
2196 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2197 feature that was used to support insecure browsers during the U.S. crypto
2198 embargo. It requires special client support, and Exim is probably the
2199 only MTA that supported it -- and would never use it because real RSA is
2200 always available. This code has been removed, because it had the bad
2201 effect of slowing Exim down by computing (never used) parameters for the
2202 RSA_EXPORT functionality.
2204 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2205 authenticator to fail if there's a tab character in the incoming data
2206 (there should never be unless someone is messing about, as it's supposed
2207 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2208 if the connection is using TLS or if the remote IP is the same as the
2209 local IP, and the "valid-client-cert option" if a client certificate has
2212 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2213 authenticators. This can be used for authorization after authentication
2214 succeeds. (In the case of plaintext, it servers for both authentication
2217 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2218 if any retry times were supplied.
2220 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2221 connection, where there is no client IP address to check. In this
2222 situation, the verify now always succeeds.
2224 PH/25 Applied John Jetmore's -Mset patch.
2226 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2228 PH/27 In a string expansion for a processed (not raw) header when multiple
2229 headers of the same name were present, leading whitespace was being
2230 removed from all of them, but trailing whitespace was being removed only
2231 from the last one. Now trailing whitespace is removed from each header
2232 before concatenation. Completely empty headers in a concatenation (as
2233 before) are ignored.
2235 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2236 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2238 PH/29 [Removed. This was a change that I later backed out, and forgot to
2239 correct the ChangeLog entry (that I had efficiently created) before
2240 committing the later change.]
2242 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2243 address errors (4xx response to RCPT) over the same connection as other
2244 messages routed to the same hosts. Such deliveries are always "forced",
2245 so retry times are not inspected. This resulted in far too many retries
2246 for the affected addresses. The effect occurred only when there were more
2247 hosts than the hosts_max_try setting in the smtp transport when it had
2248 the 4xx errors. Those hosts that it had tried were not added to the list
2249 of hosts for which the message was waiting, so if all were tried, there
2250 was no problem. Two fixes have been applied:
2252 (i) If there are any address or message errors in an SMTP delivery, none
2253 of the hosts (tried or untried) are now added to the list of hosts
2254 for which the message is waiting, so the message should not be a
2255 candidate for sending over the same connection that was used for a
2256 successful delivery of some other message. This seems entirely
2257 reasonable: after all the message is NOT "waiting for some host".
2258 This is so "obvious" that I'm not sure why it wasn't done
2259 previously. Hope I haven't missed anything, but it can't do any
2260 harm, as the worst effect is to miss an optimization.
2262 (ii) If, despite (i), such a delivery is accidentally attempted, the
2263 routing retry time is respected, so at least it doesn't keep
2264 hammering the server.
2266 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2267 in ${readsocket because some servers need this prod.
2269 PH/32 Added some extra debug output when updating a wait-xxx database.
2271 PH/33 The hint "could be header name not terminated by colon", which has been
2272 given for certain expansion errors for a long time, was not being given
2273 for the ${if def:h_colon_omitted{... case.
2275 PH/34 The spec says: "With one important exception, whenever a domain list is
2276 being scanned, $domain contains the subject domain." There was at least
2277 one case where this was not true.
2279 PH/35 The error "getsockname() failed: connection reset by peer" was being
2280 written to the panic log as well as the main log, but it isn't really
2281 panic-worthy as it just means the connection died rather early on. I have
2282 removed the panic log writing for the ECONNRESET error when getsockname()
2285 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2286 runs only) independently of the message's sender address. This meant
2287 that, if the 4xx error was in fact related to the sender, a different
2288 message to the same recipient with a different sender could confuse
2289 things. In particualar, this can happen when sending to a greylisting
2290 server, but other circumstances could also provoke similar problems.
2291 I have changed the default so that the retry time for these errors is now
2292 based a combination of the sender and recipient addresses. This change
2293 can be overridden by setting address_retry_include_sender=false in the
2296 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2297 remote server are returned as part of bounce messages. This was not
2298 happening for LMTP over a pipe (the lmtp transport), but now it is the
2299 same for both kinds of LMTP.
2301 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2302 in header lines that were in fact CNAMEs. This is no longer the case.
2304 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2305 and queue runs started by the daemon processed all messages. This has
2306 been fixed so that -R and -S can now usefully be given with -q<time>.
2308 PH/40 Import PCRE release 6.7 (fixes some bugs).
2310 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2312 PH/42 Give an error if -q is specified more than once.
2314 PH/43 Renamed the variables $interface_address and $interface_port as
2315 $received_ip_address and $received_port, to make it clear that these
2316 values apply to message reception, and not to the outgoing interface when
2317 a message is delivered. (The old names remain recognized, of course.)
2319 PH/44 There was no timeout on the connect() call when using a Unix domain
2320 socket in the ${readsocket expansion. There now is.
2322 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2323 be meaningful with "accept".
2325 SC/01 Eximstats V1.43
2326 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2328 SC/02 Eximstats V1.44
2329 Use a glob alias rather than an array ref in the generated
2330 parser. This improves both readability and performance.
2332 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2333 Collect SpamAssassin and rejection statistics.
2334 Don't display local sender or destination tables unless
2335 there is data to show.
2336 Added average volumes into the top table text output.
2338 SC/04 Eximstats V1.46
2339 Collect data on the number of addresses (recipients)
2340 as well as the number of messages.
2342 SC/05 Eximstats V1.47
2343 Added 'Message too big' to the list of mail rejection
2344 reasons (thanks to Marco Gaiarin).
2346 SC/06 Eximstats V1.48
2347 Mainlog lines which have GMT offsets and are too short to
2348 have a flag are now skipped.
2350 SC/07 Eximstats V1.49 (Alain Williams)
2351 Added the -emptyok flag.
2353 SC/08 Eximstats V1.50
2354 Fixes for obtaining the IP address from reject messages.
2356 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2357 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2358 whitesspace changes from 4.64-PH/27
2360 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2363 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2364 are found, allow negative numbers in numeric criteria)
2366 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2368 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2369 to match changes made in 4.64-PH/43
2371 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2373 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2374 "domain" is actually an IP address, and force a failure. This locks out
2375 those revolvers/nameservers that support "A-for-A" lookups, in
2376 contravention of the specifications.
2378 PH/48 When a host name was looked up from an IP address, and the subsequent
2379 forward lookup of the name timed out, the host name was left in
2380 $sender_host_name, contrary to the specification.
2382 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2383 restricted to single-key lookups, Exim was not diagnosing an error if
2384 * or *@ was used with a query-style lookup.
2386 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2388 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2389 long ago, but noone interested enough thought of it. Let's just say that
2390 the "1.1" means that there are some new functions that weren't there at
2391 some point in the past.
2393 PH/51 Error processing for expansion failure of helo_data from an smtp
2394 transport during callout processing was broken.
2396 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2397 tested/used via the -bh/-bhc/-bs options.
2399 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2400 bug, fixed in subsequent PCRE releases).
2402 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2403 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2405 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2411 SC/01 Use a glob alias rather than an array ref in eximstats generated
2412 parser. This improves both readability and performance.
2414 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2415 Don't display local sender or destination tables in eximstats unless
2416 there is data to show.
2417 Added average volumes into the eximstats top table text output.
2419 SC/03 Collect data on the number of addresses (recipients) as well
2420 as the number of messages in eximstats.
2422 TF/01 Correct an error in the documentation for the redirect router. Exim
2423 does (usually) call initgroups() when daemonizing.
2425 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2426 with consistent privilege compared to when running as a daemon.
2428 TF/03 Note in the spec that $authenticated_id is not set for local
2429 submissions from trusted users.
2431 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2432 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2434 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2435 by adding some example configuration directives to the default
2436 configuration file. A little bit of work is required to uncomment the
2437 directives and define how usernames and passwords are checked, but
2438 there is now a framework to start from.
2440 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2441 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2442 without this. I don't know how relevant this is to other LDAP libraries.
2444 PH/02 Add the verb name to the "unknown ACL verb" error.
2446 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2448 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2450 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2451 directory not expanded when it should be if an expanded home directory
2452 was set for the address (which is overridden by the transport).
2454 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2457 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2458 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2459 because it is too late at that time, and has no effect.
2461 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2462 security issue with \' (bugzilla #107). I could not use the
2463 PQescapeStringConn() function, because it needs a PGconn value as one of
2466 PH/08 When testing addresses using -bt, indicate those final addresses that
2467 are duplicates that would not cause an additional delivery. At least one
2468 person was confused, thinking that -bt output corresponded to deliveries.
2469 (Suppressing duplicates isn't a good idea as you lose the information
2470 about possibly different redirections that led to the duplicates.)
2472 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2473 systems where poll() doesn't work, in particular OS X.
2475 PH/10 Added more information to debugging output for retry time not reached.
2477 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2478 operations in malware.c.
2480 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2483 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2484 syslog_duplication set false, log lines that would normally be written
2485 both the the main log and to the reject log were not written to syslog at
2488 PH/14 In the default configuration, change the use of "message" in ACL warn
2489 statements to "add_header".
2491 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2492 not followed by a command (e.g. "seen endif").
2494 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2495 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2498 PH/17 Added extra conditions to the default value of delay_warning_condition
2502 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2503 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2504 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2507 The Auto-Submitted: and various List- headers are standardised, whereas I
2508 don't think Precedence: ever was.
2510 PH/18 Refactored debugging code in route_finduser() to show more information,
2511 in particular, the error code if getpwnam() issues one.
2513 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2514 This is apparently needed in addition to the PH/07 change above to avoid
2515 any possible encoding problems.
2517 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2518 but not after initializing Perl.
2520 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2521 output them only if debugging. By default they are written stderr,
2522 apparently, which is not desirable.
2524 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2527 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2530 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2532 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2533 authenticated or an ident call has been made. Suppress the default
2534 values for $authenticated_id and $authenticated_sender (but permit -oMai
2535 and -oMas) when testing with -bh.
2537 PH/24 Re-jigged the order of the tests in the default configuration so that the
2538 tests for valid domains and recipients precede the DNS black list and CSA
2539 tests, on the grounds that those ones are more expensive.
2541 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2542 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2543 This bug exists in every version of Exim that I still have, right back to
2546 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2547 However, an attempt to turn on case-sensitivity in a regex key by
2548 including (?-i) didn't work because the subject string was already
2549 lowercased, and the effects were non-intuitive. It turns out that a
2550 one-line patch can be used to allow (?-i) to work as expected.
2556 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2557 other effects) broke the use of negated acl sub-conditions.
2559 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2562 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2563 "Deny" causes Exim to reject the incoming connection with a 554 error.
2564 Unfortunately, if there is a major crisis, such as a disk failure,
2565 tcp-wrappers gives "deny", whereas what one would like would be some
2566 kind of temporary error. A kludge has been added to help with this.
2567 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2568 554 error is used if errno is still zero or contains ENOENT (which occurs
2569 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2572 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2574 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2575 errors. Otherwise a message that provokes a temporary error (when other
2576 messages do not) can cause a whole host to time out.
2578 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2579 addresses were routed directly to files or pipes from a redirect router.
2580 File deliveries just didn't batch; pipe deliveries might have suffered
2583 PH/06 A failure to get a lock for a hints database would erroneously always say
2584 "Failed to get write lock", even when it was really a read lock.
2586 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2587 of 0600. This has been changed to use the value of the lockfile_mode
2588 option (which defaults to 0600).
2590 PH/08 Applied small patch from the Sieve maintainer.
2592 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2593 folder from quota calculations, a direct delivery into this folder messed
2594 up the contents of the maildirsize file. This was because the regex was
2595 used only to exclude .Trash (or whatever) when the size of the mailbox
2596 was calculated. There was no check that a delivery was happening into an
2597 excluded directory. This bug has been fixed by ignoring all quota
2598 processing for deliveries into excluded directories.
2600 PH/10 Added the maildirfolder_create_regex option to appendfile.
2606 PH/01 The code for finding all the local interface addresses on a FreeBSD
2607 system running IPv6 was broken. This may well have applied to all BSD
2608 systems, as well as to others that have similar system calls. The broken
2609 code found IPv4 interfaces correctly, but gave incorrect values for the
2610 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2611 that it would not match correctly against @[] and not recognize the IPv6
2614 PH/02 The ipliteral router was not recognizing addresses of the form user@
2615 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2617 PH/03 Added disable_ipv6.
2619 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2620 decoded form, because it is most often used to construct To: headers
2621 lines in autoreplies, and the decoded form may well be syntactically
2622 invalid. However, $reply_address has leading white space removed, and all
2623 newlines turned into spaces so that the autoreply transport does not
2626 PH/05 If group was specified without a user on a router, and no group or user
2627 was specified on a transport, the group from the router was ignored.
2629 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2630 for visible compile-time settings that can be used to change these
2631 numbers, for those that want even more. Backwards compatibility with old
2632 spool files has been maintained. However, going back to a previous Exim
2633 release will lost any variables that are in spool files.
2635 PH/07 Two small changes when running in the test harness: increase delay when
2636 passing a TCP/IP connection to a new process, in case the original
2637 process has to generate a bounce, and remove special handling of
2638 127.0.0.2 (sic), which is no longer necessary.
2640 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2641 be the same on different OS.
2643 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2646 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2647 whether --show-vars was specified or not
2649 JJ/02 exipick: Added support for new ACL variable spool format introduced
2652 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2653 syntactically invalid From: or Reply-to: line, and a filter used this to
2654 generate an autoreply, and therefore failed to obtain an address for the
2655 autoreply, Exim could try to deliver to a non-existent relative file
2656 name, causing unrelated and misleading errors. What now happens is that
2657 it logs this as a hard delivery error, but does not attempt to create a
2660 PH/11 The exinext utility has a -C option for testing purposes, but although
2661 the given file was scanned by exinext itself; it wasn't being passed on
2662 when Exim was called.
2664 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2665 an end-of-file indication when reading a command response.
2667 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2668 compiled. In many other places in Exim, IPv6 addresses are always
2669 recognized, so I have changed this. It also means that IPv4 domain
2670 literals of the form [IPV4:n.n.n.n] are now always recognized.
2672 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2673 used if the router is not running as root, for example, when verifying at
2674 ACL time, or when using -bh. The debugging output from this situation was
2675 non-existent - all you got was a failure to exec. I have made two
2678 (a) Failures to set uid/gid, the current directory, or a process leader
2679 in a subprocess such as that created by queryprogram now generate
2680 suitable debugging ouput when -d is set.
2682 (b) The queryprogram router detects when it is not running as root,
2683 outputs suitable debugging information if -d is set, and then runs
2684 the subprocess without attempting to change uid/gid.
2686 PH/15 Minor change to Makefile for building test_host (undocumented testing
2689 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2690 additional section of a DNS packet that returns MX or SRV records.
2691 Instead, it always explicitly searches for A/AAAA records. This avoids
2692 major problems that occur when a DNS server includes only records of one
2693 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2694 fixed another bug: if SRV records were looked up and the corresponding
2695 address records were *not* found in the additional section, the port
2696 values from the SRV records were lost.
2698 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2699 using the correct key (the original address) when searching the retry
2700 rules in order to find which one to use for generating the retry hint.
2702 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2703 adding the default one. Similarly, if it contains a Reply-To: header, the
2704 errors_reply_to option, if set, is not used.
2706 PH/19 When calculating a retry time, Exim used to measure the "time since
2707 failure" by looking at the "first failed" field in the retry record. Now
2708 it does not use this if it is later than than the arrival time of the
2709 message. Instead it uses the arrival time. This makes for better
2710 behaviour in cases where some deliveries succeed, thus re-setting the
2711 "first failed" field. An example is a quota failure for a huge message
2712 when small messages continue to be delivered. Without this change, the
2713 "time since failure" will always be short, possible causing more frequent
2714 delivery attempts for the huge message than are intended.
2715 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2717 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2718 $1, $2, $3) because the numerical variables can be reset during some
2719 expansion items (e.g. "match"), thereby losing the authentication data.
2721 PH/21 Make -bV show the size of off_t variables so that the test suite can
2722 decide whether to run tests for quotas > 2G.
2724 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2725 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2726 filecount value is greater than 2G or if a quota value is greater than 2G
2727 on a system where the size of off_t is not greater than 4, a panic error
2730 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2731 never match. The debug and -bh output now contains an explicit error
2732 message indicating a malformed IPv4 address or mask.
2734 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2735 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2736 PH/23 above applies.
2738 PH/25 Do not write to syslog when running in the test harness. The only
2739 occasion when this arises is a failure to open the main or panic logs
2740 (for which there is an explicit test).
2742 PH/26 Added the /no_tell option to "control=freeze".
2744 PH/27 If a host name lookup failed very early in a connection, for example, if
2745 the IP address matched host_lookup and the reverse lookup yielded a name
2746 that did not have a forward lookup, an error message of the form "no IP
2747 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2748 could be logged. Now it outputs the IP address instead of "NULL".
2750 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2751 allows the sender and the authenticated sender to be set when
2752 submitting a message from within Exim. Since child_open_exim() is
2753 documented for local_scan(), the new function should be too.
2755 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2756 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2757 results in an empty string is now treated as unset.
2759 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2761 PH/31 Added sender_verify_fail logging option.
2763 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2764 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2765 tidied the source and removed it altogether.
2767 PH/33 When a queue run was abandoned because the load average was too high, a
2768 log line was always written; now it is written only if the queue_run log
2769 selector is set. In addition, the log line for abandonment now contains
2770 information about the queue run such as the pid. This is always present
2771 in "start" and "stop" lines but was omitted from the "abandon" line.
2773 PH/34 Omit spaces between a header name and the colon in the error message that
2774 is given when verify = headers_syntax fails (if there are lots of them,
2775 the message gets confusing).
2777 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2778 names, as there are now some PTR records that contain slashes. This check
2779 is only to protect against broken name servers that fall over on strange
2780 characters, so the fact that it applies to all lookups doesn't matter.
2782 PH/36 Now that the new test suite is complete, we can remove some of the
2783 special code in Exim that was needed for the old test suite. For example,
2784 sorting DNS records because real resolvers return them in an arbitrary
2785 order. The new test suite's fake resolver always returns records in the
2788 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2789 bounces) except when queue_only is set, to avoid logging races between
2790 the different processes.
2792 PH/38 Panic-die if .include specifies a non-absolute path.
2794 PH/39 A tweak to the "H" retry rule from its user.
2796 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2797 a label. They prevented compilation on older perls.
2799 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2800 a warning to be raised on newish perls.
2802 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2803 on queue. Changes to match documented behaviour of showing count of
2804 messages matching specified criteria.
2806 PH/40 Changed the default ident timeout from 30s to 5s.
2808 PH/41 Added support for the use of login_cap features, on those BSD systems
2809 that have them, for controlling the resources used by pipe deliveries.
2811 PH/42 The content-scanning code uses fopen() to create files in which to put
2812 message data. Previously it was not paying any attention to the mode of
2813 the files. Exim runs with umask(0) because the rest of the code creates
2814 files with open(), and sets the required mode explicitly. Thus, these
2815 files were ending up world-writeable. This was not a big issue, because,
2816 being within the spool directory, they were not world-accessible. I have
2817 created a function called modefopen, which takes an additional mode
2818 argument. It sets umask(777), creates the file, chmods it to the required
2819 mode, then resets the umask. All the relevant calls to fopen() in the
2820 content scanning code have been changed to use this function.
2822 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2823 to 24 hours. This avoids potential overflow problems when processing G
2824 and H retry rules. I suspect nobody ever tinkers with this value.
2826 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2828 PH/45 When the plaintext authenticator is running as a client, the server's
2829 challenges are checked to ensure they are valid base64 strings. By
2830 default, the authentication attempt is cancelled if an invalid string is
2831 received. Setting client_ignore_invalid_base64 true ignores these errors.
2832 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2833 they are received. Thus, the responses can be made to depend on the
2834 challenges. If an invalid string is ignored, an empty string is placed in
2837 PH/46 Messages that are created by the autoreply transport now contains a
2838 References: header, in accordance with RFCs 2822 and 3834.
2840 PH/47 Added authenticated_sender_force to the smtp transport.
2842 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2844 PH/49 Installed latest patch from the Sieve maintainer.
2846 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2847 was also set, the appendfile transport was unnecessarily scanning a
2848 directory of message files (e.g. for maildir delivery) to find the count
2849 of files (along with the size), even though it did not need this
2850 information. It now does the scan only if it needs to find either the
2851 size of the count of files.
2853 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2855 PH/52 Two bugs concerned with error handling when the smtp transport is
2858 (i) Exim was not creating retry information for temporary errors given
2859 for individual recipients after the DATA command when the smtp transport
2860 was used in LMTP mode. This meant that they could be retried too
2861 frequently, and not timed out correctly.
2863 (ii) Exim was setting the flag that allows error details to be returned
2864 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2865 recipients that were returned after the DATA command.
2867 PH/53 This is related to PH/52, but is more general: for any failing address,
2868 when detailed error information was permitted to be returned to the
2869 sender, but the error was temporary, then after the final timeout, only
2870 "retry timeout exceeded" was returned. Now it returns the full error as
2871 well as "retry timeout exceeded".
2873 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2874 do this, and (what is worse) MTAs that accept it.
2876 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2877 will now be deprecated.
2879 PH/56 New os.c-cygwin from the Cygwin maintainer.
2881 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2882 formats (previously only available in exim formats via -bpr, -bpru,
2883 and -bpra. Now also available in native and exiqgrep formats)
2885 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2886 with very large, slow to parse queues
2888 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2890 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2892 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2893 responses to authentication challenges, though it was showing the
2894 challenges; (ii) I've removed the CR characters from the debug output for
2897 PH/58 Allow for the insertion of a newline as well as a space when a string
2898 is turned into more than one encoded-word during RFC 2047 encoding. The
2899 Sieve code now uses this.
2901 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2902 data_4xx, lost_connection, tls_required.
2904 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2905 message was being sent as an SMTP response.
2907 PH/61 Add -l and -k options to exicyclog.
2909 PH/62 When verifying, if an address was redirected to one new address, so that
2910 verification continued, and the new address failed or deferred after
2911 having set something in $address_data, the value of $address_data was not
2912 passed back to the ACL. This was different to the case when no
2913 redirection occurred. The value is now passed back in both cases.
2915 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2916 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2917 use login_cap.h, so on its own it isn't the distinguishing feature. The
2918 new name refers directly to the setclassresources() function.
2920 PH/65 Added configuration files for NetBSD3.
2922 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2924 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2925 is preferred over IPv4.
2927 PH/68 The bounce_return_message and bounce_return_body options were not being
2928 honoured for bounces generated during the reception of non-SMTP messages.
2929 In particular, this applied to messages rejected by the ACL. This bug has
2930 been fixed. However, if bounce_return_message is true and bounce_return_
2931 body is false, the headers that are returned for a non-SMTP message
2932 include only those that have been read before the error was detected.
2933 (In the case of an ACL rejection, they have all been read.)
2935 PH/69 The HTML version of the specification is now built in a directory called
2936 spec_html instead of spec.html, because the latter looks like a path with
2937 a MIME-type, and this confuses some software.
2939 PH/70 Catch two compiler warnings in sieve.c.
2941 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2942 function verify_get_ident() calls ip_connect() to connect a socket, but
2943 if the "connect()" function timed out, ip_connect() used to close the
2944 socket. However, verify_get_ident() also closes the socket later, and in
2945 between Exim writes to the log, which may get opened at this point. When
2946 the socket was closed in ip_connect(), the log could get the same file
2947 descriptor number as the socket. This naturally causes chaos. The fix is
2948 not to close the socket in ip_connect(); the socket should be closed by
2949 the function that creates it. There was only one place in the code where
2950 this was missing, in the iplookup router, which I don't think anybody now
2951 uses, but I've fixed it anyway.
2953 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2954 well as to direct DNS lookups. Otherwise the handling of names in host
2955 lists is inconsistent and therefore confusing.
2961 PH/01 Two changes to the default runtime configuration:
2963 (1) Move the checks for relay_from_hosts and authenticated clients from
2964 after to before the (commented out) DNS black list checks.
2966 (2) Add control=submission to the relay_from_hosts and authenticated
2967 clients checks, on the grounds that messages accepted by these
2968 statements are most likely to be submissions.
2970 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2972 (1) Generate an error if the third argument for the ${prvs expansion is
2975 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2978 (3) Reset the variables that are obtained from the first argument of
2979 ${prvscheck and used in the second argument before leaving the code,
2980 because their memory is reclaimed, so using them afterwards may do
2983 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2984 one (it's much easier than Tom thought :-).
2986 (5) Because of (4), we can now allow for the use of $prvscheck_result
2987 inside the third argument.
2989 PH/03 For some reason, the default setting of PATH when running a command from
2990 a pipe transport was just "/usr/bin". I have changed it to
2993 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2994 anything to be listed in the output from -bV.
2996 PH/05 When a filter generated an autoreply, the entire To: header line was
2997 quoted in the delivery log line, like this:
2999 => >A.N.Other <ano@some.domain> <original@ddress> ...
3001 This has been changed so that it extracts the operative address. There
3002 may be more than one such address. If so, they are comma-separated, like
3005 => >ano@some.domain,ona@other.domain <original@ddress> ...
3007 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3008 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3009 not being looked up in the rDNS to get a host name, Exim was showing the
3010 IP address twice in Received: lines, even though the IP addresses were
3011 identical. For example:
3013 Received: from [1.2.3.4] (helo=[1.2.3.4])
3015 However, if the real host name was known, it was omitting the HELO data
3016 if it matched the actual IP address. This has been tidied up so that it
3017 doesn't show the same IP address twice.
3019 PH/07 When both +timestamp and +memory debugging was on, the value given by
3020 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3021 called by the debug printing, thereby overwriting the timestamp buffer.
3022 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3024 PH/08 When the original message was included in an autoreply transport, it
3025 always said "this is a copy of the message, including all the headers",
3026 even if body_only or headers_only was set. It now gives an appropriate
3029 PH/09 Applied a patch from the Sieve maintainer which:
3031 o fixes some comments
3032 o adds the (disabled) notify extension core
3033 o adds some debug output for the result of if/elsif tests
3034 o points to the current vacation draft in the documentation
3035 and documents the missing references header update
3039 o fixes a bug in processing the envelope test (when testing
3040 multiple envelope elements, the last element determinted the
3043 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3044 Electronic Mail") by including:
3046 Auto-submitted: auto-generated
3048 in the messages that it generates (bounce messages and others, such as
3049 warnings). In the case of bounce messages for non-SMTP mesages, there was
3050 also a typo: it was using "Auto_submitted" (underscore instead of
3051 hyphen). Since every message generated by Exim is necessarily in response
3052 to another message, thes have all been changed to:
3054 Auto-Submitted: auto-replied
3056 in accordance with these statements in the RFC:
3058 The auto-replied keyword:
3060 - SHOULD be used on messages sent in direct response to another
3061 message by an automatic process,
3063 - MUST NOT be used on manually-generated messages,
3065 - MAY be used on Delivery Status Notifications (DSNs) and Message
3066 Disposition Notifications (MDNs),
3068 - MUST NOT be used on messages generated by automatic or periodic
3069 processes, except for messages which are automatic responses to
3072 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3073 to the default Received: header definition.
3075 PH/12 Added log selector acl_warn_skipped (default on).
3077 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3078 variables because (a) they are in the wrong storage pool and (b) even if
3079 they were copied, it wouldn't work properly because of the caching.
3081 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3082 checking when decoding. Apparently there are clients that generate
3083 overlong encoded strings. Why am I not surprised?
3085 PH/15 If the first argument of "${if match_address" was not empty, but did not
3086 contain an "@" character, Exim crashed. Now it writes a panic log message
3087 and treats the condition as false.
3089 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3091 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3092 extension "envelope-auth". The code is finished and in agreement with
3093 other implementations, but there is no documentation so far and in fact,
3094 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3095 not changing the active code.
3097 Print executed "if" and "elsif" statements when debugging is used. This
3098 helps a great deal to understand what a filter does.
3100 Document more things not specified clearly in RFC3028. I had all this
3101 sorted out, when out of a sudden new issues came to my mind. Oops."
3103 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3106 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3107 canonical form (as documented). However, after a host name lookup from
3108 the IP address, check_host() was doing a simple string comparison with
3109 addresses acquired from the DNS when checking that the found name did
3110 have the original IP as one of its addresses. Since any found IPv6
3111 addresses are likely to be in abbreviated form, the comparison could
3112 fail. Luckily, there already exists a function for doing the comparison
3113 by converting both addresses to binary, so now that is used instead of
3114 the text comparison.
3116 PH/20 There was another similar case to PH/19, when a complete host name was
3117 given in a host list; looking up its IP address could give an abbreviated
3118 form, whereas the current host's name might or might not be abbreviated.
3119 The same fix has been applied.
3125 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3126 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3129 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3130 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3132 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3134 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3135 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3136 Also, for "Auto-Submitted", treat anything other than "no" as
3137 non-personal, in accordance with RFC 3834. (Previously it treated
3138 anything starting "auto-" as non-personal.)
3140 TF/01 The control=submission/name=... option had a problem with syntax
3141 errors if the name included a slash character. The /name= option
3142 now slurps the rest of the string, so it can include any characters
3143 but it must come last in the list of options (after /sender_retain
3146 PH/05 Some modifications to the interface to the fake nameserver for the new
3154 TK/01 Added the "success_on_redirect" address verification option. See
3155 NewStuff for rationale and an example.
3157 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3159 PH/02 Patch to exigrep to allow it to work on syslog lines.
3161 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3162 fread() to skip over the body file's header line, because in Cygwin the
3163 header line is locked and is inaccessible.
3165 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3166 co-exist for some time) to make it clear that it is the Exim ID that is
3167 referenced, not the Message-ID: header line.
3169 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3170 string_format() function, because snprintf() does not exist on all
3173 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3174 ${stat: expansion item.
3176 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3177 protocol synchronization error", to keep the pedants happy.
3179 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3180 well as for IRIX systems, when gcc is being used. See the host.c source
3183 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3185 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3188 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3189 transport and to the smtp transport in LMTP mode.
3191 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3193 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3194 run a filter in a subprocess. This could lead to confusion in subsequent
3195 lookups in the parent process. There should also be a search_tidyup() at
3196 the end of the subprocess.
3198 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3199 only if the host matched helo_try_verify_hosts, which caused the
3200 verification to occur when the EHLO/HELO command was issued. The ACL just
3201 tested the remembered result. Now, if a previous verification attempt has
3202 not happened, "verify = helo" does it there and then.
3204 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3206 TK/03 Fix log output including CR from clamd.
3208 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3209 exist provoked a memory error which could cause a segfault.
3211 PH/15 Installed PCRE 6.2
3213 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3215 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3216 of the problem. Specifically, suggested +O2 rather than +O1 for the
3219 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3221 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3222 record, the port was not correctly logged when the outgoing_port log
3223 selector was set (it logged the transort's default port).
3225 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3226 fallback_hosts, and "hosts" in the smtp transport.
3228 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3229 host errors such as "Connection refused".
3231 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3232 authentication with radiusclient 0.4.9:
3234 - Error returned from rc_read_config was caught wrongly
3235 - Username/password not passed on to radius server due to wrong length.
3237 The presumption is that some radiusclient API changes for 4.51/PH/17
3238 were not taken care of correctly. The code is still untested by me (my
3239 Linux distribution still has 0.3.2 of radiusclient), but it was
3240 contributed by a Radius user.
3242 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3243 expanding the "port" option of the smtp transport.
3245 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3246 while reading a MIME header. Thanks to Tom Hughes for a patch.
3248 PH/24 Include config.h inside local_scan.h so that configuration settings are
3251 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3252 that in an ACL for RCPT (for example), you can examine exactly what was
3255 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3256 commands, but it was not correctly comparing the address with the actual
3257 client host address. Thus, it would show the EHLO address in Received:
3258 header lines when this was not necessary.
3260 PH/27 Added the % operator to ${eval:}.
3262 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3263 it should be ignoring failures (because with -C, for example, it has lost
3264 privilege). It wasn't ignoring creation failures other than "already
3267 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3270 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3271 because an input line was too long, either on its own, or by virtue of
3272 too many continuations, the temporary file was not being removed, and the
3273 return code was incorrect.
3275 PH/31 Missing "BOOL" in function definition in filtertest.c.
3277 PH/32 Applied Sieve patches from the maintainer.
3279 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3281 PH/33 Added "verify = not_blind".
3283 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3284 Local/Makefile (with some defaults set). These are used in built scripts
3285 such as exicyclog, but they have never been used in the exim_install
3286 script (though there are many overriding facilities there). I have
3287 arranged that the exim_install script now takes note of these two
3290 PH/35 Installed configuration files for Dragonfly.
3292 PH/36 When a locally submitted message by a trusted user did not contain a
3293 From: header, and the sender address was obtained from -f or from an SMTP
3294 MAIL command, and the trusted user did not use -F to supply a sender
3295 name, $originator_name was incorrectly used when constructing a From:
3296 header. Furthermore, $originator_name was used for submission mode
3297 messages from external hosts without From: headers in a similar way,
3298 which is clearly wrong.
3300 PH/37 Added control=suppress_local_fixups.
3302 PH/38 When log_selector = +received_sender was set, and the addition of the
3303 sender made the log line's construction buffer exactly full, or one byte
3304 less than full, an overflow happened when the terminating "\n" was
3307 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3308 when the result of a list match is failure because a DNS lookup failed.
3310 PH/40 RM_COMMAND is now used in the building process.
3312 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3313 the "build-* directories that it finds.
3315 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3316 address in a domain literal was a prefix of an interface address.
3318 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3319 when verifying a sender address, unless rewrite_headers is false.
3321 PH/44 Wrote a long comment about why errors_to addresses are verified as
3322 recipients, not senders.
3324 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3325 the ratelimit ACL was added.
3327 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3329 PH/46 Added extra information about PostgreSQL errors to the error string.
3331 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3332 avoiding the need to install special zones in a real server. This is
3333 backwards compatible; if it can't find the fake resolver, it drops back.
3334 Thus, both old and new test suites can be run.
3336 TF/02 Added util/ratelimit.pl
3338 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3339 clock is set back in time.
3341 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3342 Candler <B.Candler@pobox.com>.
3344 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3345 OK for addresses that are the result of redirections.
3347 PH/48 A number of further additions for the benefit of the new test suite,
3348 including a fake gethostbyname() that interfaces to the fake DNS resolver
3351 TF/06 The fix for widen_domains has also been applied to qualify_single and
3352 search_parents which are the other dnslookup options that can cause
3355 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3358 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3360 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3361 service" if it encountered an MX record with an empty target hostname.
3362 The message is now "an MX or SRV record indicated no SMTP service".
3364 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3365 if the DNS of the sending site is misconfigured. This is quite a
3366 common situation. This change restores the behaviour of treating a
3367 helo verification defer as a failure.
3369 PH/51 If self=fail was set on a router, the bounce message did not include the
3370 actual error message.
3376 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3378 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3379 command itself ended in error, the underlying message about the transport
3380 filter timeout was being overwritten with the pipe command error. Now the
3381 underlying error message should be appended to the second error message.
3383 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3385 PH/02 Reduce the amount of output that "make" produces by default. Full output
3386 can still be requested.
3388 PH/03 The warning log line about a condition test deferring for a "warn" verb
3389 was being output only once per connection, rather than after each
3390 occurrence (because it was using the same function as for successful
3391 "warn" verbs). This seems wrong, so I have changed it.
3393 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3394 it should not have, which might have caused a crash in the right
3395 circumstances, but probably never did.
3397 PH/04 Installed a modified version of Tony Finch's patch to make submission
3398 mode fix the return path as well as the Sender: header line, and to
3399 add a /name= option so that you can make the user's friendly name appear
3402 TF/03 Added the control = fakedefer ACL modifier.
3404 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3405 Mark Lowes for thorough testing.
3407 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3409 TK/03 Merged latest SRS patch from Miles Wilton.
3411 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3412 with the definition in sysexits.h (which is #included earlier).
3413 Fortunately, Exim does not actually use EX_OK. The code used to try to
3414 preserve the sysexits.h value, by assumimg that macro definitions were
3415 scanned for macro replacements. I have been disabused of this notion,
3416 so now the code just undefines EX_OK before #including unistd.h.
3418 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3419 in the smtp transport. When a block could not be written in a single
3420 write() function, the timeout was being re-applied to each part-write.
3421 This seems wrong - if the receiver was accepting one byte at a time it
3422 would take for ever. The timeout is now adjusted when this happens. It
3423 doesn't have to be particularly precise.
3425 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3426 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3428 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3429 without a domain if the check to <postmaster@domain> fails.
3431 SC/01 Eximstats: added -xls and the ability to specify output files
3432 (patch written by Frank Heydlauf).
3434 SC/02 Eximstats: use FileHandles for outputing results.
3436 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3438 SC/04 Eximstats: fixed display of large numbers with -nvr option
3440 SC/05 Eximstats: fixed merging of reports with empty tables.
3442 SC/06 Eximstats: added the -include_original_destination flag
3444 SC/07 Eximstats: removed tabs and trailing whitespace.
3446 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3448 TK/06 MBOX spool code: Add real "From " MBOX separator line
3449 so the .eml file is really in mbox format (even though
3450 most programs do not really care). Patch from Alex Miller.
3452 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3453 The latter is generated from $received_to and is only set if the
3454 message has one envelope recipient. SA can use these headers,
3455 obviously out-of-the-box. Patch from Alex Miller.
3457 PH/08 The ${def test on a variable was returning false if the variable's
3458 value was "0", contrary to what the specification has always said!
3459 The result should be true unless the variable is empty.
3461 PH/09 The syntax error of a character other than { following "${if
3462 def:variable_name" (after optional whitespace) was not being diagnosed.
3463 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3464 accidental colon was present, for example, could give incorrect results.
3466 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3467 result is used (not including appendfile, where other changes are about
3470 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3471 This involved changing a lot of size variables from int to off_t. It
3472 should work with maildirs and everything.
3474 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3475 spamd dying while we are connected to it.
3477 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3480 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3481 function for BDB 4.3.
3483 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3485 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3486 the sources for PCRE have been re-arranged and more files are now
3489 PH/15 The code I had for printing potentially long long variables in PH/11
3490 above was not the best (it lost precision). The length of off_t variables
3491 is now inspected at build time, and an appropriate printing format (%ld
3492 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3493 to be "long long int" or "long int". This is needed for the internal
3494 formatting function string_vformat().
3496 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3497 the configuration file to be ":syslog", then the script "guesses" where
3498 the logs files are, rather than using the compiled in default. In our
3499 case the guess is not the same as the compiled default, so the script
3500 suddenly stopped working when I started to use syslog. The patch checks
3501 to see if log_file_path is "". If so, it attempts to read it from exim
3502 with no configuration file to get the compiled in version, before it
3503 falls back to the previous guessing code."
3505 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3506 implementing BATV in an Exim configuration. See NewStuff for the gory
3509 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3510 Makefile that are specific to HP-UX.
3512 PH/18 If the "use_postmaster" option was set for a recipient callout together
3513 with the "random" option, the postmaster address was used as the MAIL
3514 FROM address for the random test, but not for the subsequent recipient
3515 test. It is now used for both.
3517 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3518 patch removes a few documentation additions to RFC 3028, because the
3519 latest draft now contains them. It adds the new en;ascii-case comparator
3520 and a new error check for 8bit text in MIME parts. Comparator and
3521 require names are now matched exactly. I enabled the subaddress
3522 extension, but it is not well tested yet (read: it works for me)."
3524 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3525 rework some of the code of TK/09 above to avoid the hardwired use of
3526 "%lld" and "long long". Replaced the call to snprintf() with a call to
3529 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3530 records point to non-existent hosts", "retry timeout exceeded", and
3531 "retry time not reached for any host after a long failure period".
3533 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3534 experimental DomainKeys support:
3536 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3537 (2) On an error such as an illegally used "control", the wrong name for
3538 the control was given.
3540 These problems did NOT occur unless DomainKeys support was compiled.
3542 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3544 PH/24 Added ${if match_ip condition.
3546 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3547 databases so that it will be absolutely obvious if a crash occurs in the
3548 DB library. This is a regular occurrence (often caused by mis-matched
3551 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3552 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3553 picked up on a user's system that detects such things. There doesn't seem
3554 to be a gcc warning option for this - only an attribute that has to be
3555 put on the function's prototype. It seems that in Fedora Core 4 they have
3556 set this on a number of new functions. No doubt there will be more in due
3559 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3560 specify a transport. However, if an address that was verified by such a
3561 router was the subject of a callout, Exim crashed because it tried to
3562 read the rcpt_include_affixes from the non-existent transport. Now it
3563 just assumes that the setting of that option is false. This bug was
3564 introduced by 4.51/PH/31.
3566 PH/28 Changed -d+all to exclude +memory, because that information is very
3567 rarely of interest, but it makes the output a lot bigger. People tend to
3568 do -d+all out of habit.
3570 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3571 code in os-type was giving problems when libc.so lives in lib64, like on
3574 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3575 aren't the modern standard, and it seems that some systems' include files
3576 don't always have them. Exim was already checking for some of the newer
3577 ones like T_AAAA, and defining it itself. I've added checks for all the
3578 record types that Exim uses.
3580 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3581 not automatically generating a new one, as it is supposed to. This
3582 prevented TLS from working. If the file did exist, but contained invalid
3583 data, a new version was generated, as expected. It was only the case of a
3584 non-existent file that was broken.
3586 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3587 with a change in libdomainkeys > 0.64.
3589 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3590 from DNS. If the selector record carries the flag, it now has
3591 precedence over the domain-wide flag.
3593 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3595 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3596 the use of a port name that isn't defined in /etc/services) occurred, the
3597 message was deferred as in a normal delivery, and thus remained on the
3598 spool, instead of being failed because of the mua_wrapper setting. This
3599 is now fixed, and I tidied up some of the mua_wrapper messages at the
3602 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3603 the messages in a hash of arrays rather than using individual hashes.
3604 This is a bit cleaner and results in dramatic memory savings, albeit
3605 at a slight CPU cost.
3607 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3608 as requested by Marc Sherman.
3610 SC/10 Eximstats: added histograms for user specified patterns as requested
3613 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3615 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3616 fopen() in the content-scanning modules that did not already have it.
3622 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3623 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3625 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3627 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3629 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3630 to test Sieve filters that use "vacation".
3632 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3633 that changes the way the GnuTLS parameters are stored in the cache file.
3634 The new format can be generated externally. For backward compatibility,
3635 if the data in the cache doesn't make sense, Exim assumes it has read an
3636 old-format file, and it generates new data and writes a new file. This
3637 means that you can't go back to an older release without removing the
3640 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3641 work if there are any delivery delays because "one_time" forces the
3642 parent to be marked "delivered", so its unseen clone is never tried
3643 again. For this reason, Exim now forbids the simultaneous setting of
3646 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3647 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3648 entry if you want to know the details. The fix, however, neglected to
3649 consider the case when local delivery batching is involved. The test for
3650 "previously delivered" was not happening when checking to see if an
3651 address could be batched with a previous (undelivered) one; under
3652 certain circumstances this could lead to multiple deliveries to the same
3655 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3656 in its include files, and this causes problems building Exim.
3658 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3659 header_syntax) but Exim was just ignoring anything given after a slash.
3660 In particular, this caused confusion with an attempt to use "verify =
3661 reverse_host_lookup/defer_ok". An error is now given when options are
3662 supplied for verify items that do not have them. (Maybe reverse_host_
3663 lookup should have a defer_ok option, but that's a different point.)
3665 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3666 defined by RFC 821) to 2048, because there were problems with some AUTH
3667 commands, and RFC 1869 says the size should be increased for extended
3668 SMTP commands that take arguments.
3670 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3673 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3674 "unknown" error; now it says that the functionality isn't in the binary.
3676 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3677 an address' error message when a string expansion fails (syntax or
3678 whatever). Otherwise the password may appear in the log. Following change
3679 PH/42 below, there is no longer a chance of it appearing in a bounce
3682 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3684 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3685 of its addresses was ever tried. (Bugzilla bug #2.)
3687 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3688 the result incorrectly in the debug output. (It correctly added a newline
3689 to what was transported.)
3691 TF/01 Added $received_time.
3693 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3694 commented out examples of how to interface to a virus scanner and to
3695 SpamAssassin. Also added commented examples of av_scanner and
3696 spamd_address settings.
3698 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3699 and controls are allowed in which ACLs. There were a couple of minor
3700 errors. Some of the entries in the conditions table (which is a table of
3701 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3702 negation of where the condition IS allowed.
3704 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3706 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3707 header file does not have a version number, so I've had to invent a new
3708 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3709 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3710 radiusclient), but it was contributed by a Radius user.
3712 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3713 files or directories, for OpenSSL.
3715 PH/19 When an Exim process that is running as root has to create an Exim log
3716 file, it does so in a subprocess that runs as exim:exim so as to get the
3717 ownership right at creation (otherwise, other Exim processes might see
3718 the file with the wrong ownership). There was no test for failure of this
3719 fork() call, which would lead to the process getting stuck as it waited
3720 for a non-existent subprocess. Forks do occasionally fail when resources
3721 run out. I reviewed all the other calls to fork(); they all seem to check
3724 PH/20 When checking for unexpected SMTP input at connect time (before writing
3725 the banner), Exim was not dealing correctly with a non-positive return
3726 from the read() function. If the client had disconnected by this time,
3727 the result was a log entry for a synchronization error with an empty
3728 string after "input=" when read() returned zero. If read() returned -1
3729 (an event I could not check), uninitialized data bytes were printed.
3730 There were reports of junk text (parts of files, etc) appearing after
3733 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3735 PH/22 Added support for macro redefinition, and (re)definition in between
3736 driver and ACL definitions.
3738 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3739 forgetting to use the resulting value; it was using the unexpanded value.
3741 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3742 hadn't been configured. The fix is from Juergen Kreileder, who
3743 understands it better than I do:
3745 "Here's what I see happening with three configured cyrus_sasl
3746 authenticators configured (plain, login, cram-md5):
3748 On startup auth_cyrus_sasl_init() gets called for each of these.
3749 This means three calls to sasl_listmech() without a specified mech_list.
3750 => SASL tests which mechs of all available mechs actually work
3751 => three warnings about OTP not working
3752 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3754 With the patch, sasl_listmech() also gets called three times. But now
3755 SASL's mech_list option is set to the server_mech specified in the the
3756 authenticator. Or in other words, the answer from sasl_listmech()
3757 gets limited to just the mech you're testing for (which is different
3759 => the return list contains just 'plain' or 'login', 'cram-md5' or
3760 nothing depending on the value of ob->server_mech.
3762 I've just tested the patch: Authentication still works fine,
3763 unavailable mechs specified in the exim configuration are still
3764 caught, and the auth.log warnings about OTP are gone."
3766 PH/25 When debugging is enabled, the contents of the command line are added
3767 to the debugging output, even when log_selector=+arguments is not
3770 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3771 answer is "GNU", and only if the return is "GNU/something" is the answer
3774 PH/27 $acl_verify_message is now set immediately after the failure of a
3775 verification in an ACL, and so is available in subsequent modifiers. In
3776 particular, the message can be preserved by coding like this:
3778 warn !verify = sender
3779 set acl_m0 = $acl_verify_message
3781 Previously, $acl_verify_message was set only while expanding "message"
3782 and "log_message" when a very denied access.
3784 PH/28 Modified OS/os.c-Linux with
3786 -#ifndef OS_LOAD_AVERAGE
3787 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3789 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3790 nomenclature these days.)
3792 PH/29 Installed patch from the Sieve maintainer that adds the options
3793 sieve_useraddress and sieve_subaddress to the redirect router.
3795 PH/30 In these circumstances:
3796 . Two addresses routed to the same list of hosts;
3797 . First host does not offer TLS;
3798 . First host accepts first address;
3799 . First host gives temporary error to second address;
3800 . Second host offers TLS and a TLS session is established;
3801 . Second host accepts second address.
3802 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3803 and peerdn, if requested) that were in fact used only for the second
3806 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3807 attention to any local part prefix or suffix that was matched by the
3808 router that accepted the address. It now behaves in the same way as it
3809 does for delivery: the affixes are removed from the local part unless
3810 rcpt_include_affixes is set on the transport.
3812 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3813 timeout during the DATA phase of an incoming message.
3815 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3816 applied a patch sanctioned by the Sieve maintainer.
3818 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3819 the uid or gid is negative. A case of a negative gid caused this to be
3820 noticed. The fix allows for either to be negative.
3822 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3823 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3826 PH/36 The change PH/12 above was broken. Fixed it.
3828 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3829 the grounds that routing the same address twice would always produce the
3830 same answer. This might have been true once, but it is certainly no
3831 longer true now. Routing a child address may depend on the previous
3832 routing that produced that child. Some complicated redirection strategies
3833 went wrong when messages had multiple recipients, and made Exim's
3834 behaviour dependent on the order in which the addresses were given.
3836 I have moved the duplicate checking until after the routing is complete.
3837 Exim scans the addresses that are assigned to local and remote
3838 transports, and removes any duplicates. This means that more work will be
3839 done, as duplicates will always all be routed, but duplicates are
3840 presumably rare, so I don't expect this is of any significance.
3842 For deliveries to pipes, files, and autoreplies, the duplicate checking
3843 still happens during the routing process, since they are not going to be
3846 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3847 It corrects a timeout issue with spamd. This is Ian's comment: "The
3848 background is that sometimes spamd either never reads data from a
3849 connection it has accepted, or it never writes response data. The exiscan
3850 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3851 blindly assumes that writes won't block so it may never time out."
3853 PH/39 Allow G after quota size as well as K and M.
3855 PH/40 The value set for $authenticated_id in an authenticator may not contain
3856 binary zeroes or newlines because the value is written to log lines and
3857 to spool files. There was no check on this. Now the value is run through
3858 the string_printing() function so that such characters are converted to
3859 printable escape sequences.
3861 PH/41 $message_linecount is a new variable that contains the total number of
3862 lines in the message. Compare $body_linecount, which is the count for the
3865 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3866 bounce and delay warning messages, except in certain special cases, which
3869 (a) An SMTP error message from a remote host;
3870 (b) A message specified in a :fail: redirection;
3871 (c) A message specified in a "fail" command in a system filter;
3872 (d) A message specified in a FAIL return from the queryprogram router;
3873 (e) A message specified by the cannot_route_message router option.
3875 In these cases only, Exim does include the error details in bounce and
3876 warning messages. There are also a few cases where bland messages such
3877 as "unrouteable address" or "local delivery error" are given.
3879 PH/43 $value is now also set for the "else" part of a ${run expansion.
3881 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3882 being worked on, but at least Exim now implements the latest version to
3885 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3886 process to complete was treated as a delivery failure, a timeout while
3887 writing the message to the pipe was logged, but erroneously treated as a
3888 successful delivery. Such timeouts include transport filter timeouts. For
3889 consistency with the overall process timeout, these timeouts are now
3890 treated as errors, giving rise to delivery failures by default. However,
3891 there is now a new Boolean option for the pipe transport called
3892 timeout_defer, which, if set TRUE, converts the failures into defers for
3893 both kinds of timeout. A transport filter timeout is now identified in
3896 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3897 systems where "make" and "gmake" are different, calling "gmake" at top
3898 level broke things. I've arranged for the value of $(MAKE) to be passed
3899 from the Makefile to this script so that it can call the same version of
3903 A note about Exim versions 4.44 and 4.50
3904 ----------------------------------------
3906 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3907 changes of various kinds. As a consequence, a big documentation update was
3908 needed. This delayed the release for rather longer than seemed good, especially
3909 in the light of a couple of (minor) security issues. Therefore, the changes
3910 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3911 release. So 4.44 and 4.50 are in effect two different branches that both start
3914 I have left the 4.50 change log unchanged; it contains all the changes since
3915 4.43. The change log for 4.44 is below; many of its items are identical to
3916 those for 4.50. This seems to be the most sensible way to preserve the
3917 historical information.
3923 1. Minor wording change to the doc/README.SIEVE file.
3925 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3926 computation of the current number of files was incorrect.
3928 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3929 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3932 4. Give more explanation in the error message when the command for a transport
3933 filter fails to execute.
3935 5. There are several places where Exim runs a non-Exim command in a
3936 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3937 was being done only for the command run by the queryprogram router. It is
3938 now done for all such subprocesses. The other cases are: ${run, transport
3939 filters, and the commands run by the lmtp and pipe transports.
3941 6. Added CONFIGURE_GROUP build-time option.
3943 7. Some older OS have a limit of 256 on the maximum number of file
3944 descriptors. Exim was using setrlimit() to set 1000 as a large value
3945 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3946 systems. I've change it so that if it can't get 1000, it tries for 256.
3948 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3949 was an oversight, and furthermore, ever since the addition of extra
3950 controls (e.g. 4.43/32), the checks on when to allow different forms of
3951 "control" were broken. There should now be diagnostics for all cases when a
3952 control that does not make sense is encountered.
3954 9. Added the /retain_sender option to "control=submission".
3956 10. $recipients is now available in the predata ACL (oversight).
3958 11. Tidy the search cache before the fork to do a delivery from a message
3959 received from the command line. Otherwise the child will trigger a lookup
3960 failure and thereby defer the delivery if it tries to use (for example) a
3961 cached ldap connection that the parent has called unbind on.
3963 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3964 of $address_data from the recipient verification was clobbered by the
3965 sender verification.
3967 13. The value of address_data from a sender verification is now available in
3968 $sender_address_data in subsequent conditions in the ACL statement.
3970 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3972 15. Added a new option "connect=<time>" to callout options, to set a different
3975 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3976 was its contents. (It was OK if the option was not defined at all.)
3978 17. A "Completed" log line is now written for messages that are removed from
3979 the spool by the -Mrm option.
3981 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3982 information about exactly what failed.
3984 19. Added -dd to debug only the daemon process.
3986 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3987 handles timeouts, both on the server side and network timeouts. Renamed the
3988 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3990 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3991 It is now set to "smtps".
3993 22. $host_address is now set to the target address during the checking of
3994 ignore_target_hosts.
3996 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3997 being passed; this would have caused $sender_host_name to have been used if
3998 matching the list had actually called for a host name (not very likely,
3999 since this list is usually IP addresses). A host name is now passed as
4002 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4003 code that specifies a non-restarting handler (typically sigaction() in
4004 modern systems) in an attempt to fix a rare and obscure crash bug.
4006 25. Narrowed the window for a race in the daemon that could cause it to ignore
4007 SIGCHLD signals. This is not a major problem, because they are used only to
4008 wake it up if nothing else does.
4010 26. A malformed maildirsize file could cause Exim to calculate negative values
4011 for the mailbox size or file count. Odd effects could occur as a result.
4012 The maildirsize information is now recalculated if the size or filecount
4015 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4016 support for a long time. Removed HAVE_SYS_VFS_H.
4018 28. Installed the latest version of exipick from John Jetmore.
4020 29. In an address list, if the pattern was not a regular expression, an empty
4021 subject address (from a bounce message) matched only if the pattern was an
4022 empty string. Non-empty patterns were not even tested. This was the wrong
4023 because it is perfectly reasonable to use an empty address as part of a
4024 database query. An empty address is now tested by patterns that are
4025 lookups. However, all the other forms of pattern expect the subject to
4026 contain a local part and a domain, and therefore, for them, an empty
4027 address still always fails if the pattern is not itself empty.
4029 30. Exim went into a mad DNS loop when attempting to do a callout where the
4030 host was specified on an smtp transport, and looking it up yielded more
4031 than one IP address.
4033 31. Re-factored the code for checking spool and log partition space into a
4034 function that finds that data and another that does the check. The former
4035 is then used to implement four new variables: $spool_space, $log_space,
4036 $spool_inodes, and $log_inodes.
4038 32. The RFC2047 encoding function was originally intended for short strings
4039 such as real names; it was not keeping to the 75-character limit for
4040 encoded words that the RFC imposes. It now respects the limit, and
4041 generates multiple encoded words if necessary. To be on the safe side, I
4042 have increased the buffer size for the ${rfc2047: expansion operator from
4045 33. It is now permitted to omit both strings after an "if" condition; if the
4046 condition is true, the result is "true". As before, when the second string
4047 is omitted, a false condition yields an empty string. This makes it less
4048 cumbersome to write custom ACL and router conditions.
4050 34. Failure to deliver a bounce message always caused it to be frozen, even if
4051 there was an errors_to setting on the router. The errors_to setting is now
4054 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4055 canonical form (fully expanded) before being placed in
4056 $sender_host_address.
4058 36. The table in the code that translates DNS record types into text (T_A to
4059 "A" for instance) was missing entries for NS and CNAME. It is just possible
4060 that this could have caused confusion if both these types were looked up
4061 for the same domain, because the text type is used as part of Exim's
4062 per-process caching. But the chance of anyone hitting this buglet seems
4065 37. The dnsdb lookup has been extended in a number of ways.
4067 (1) There is a new type, "zns", which walks up the domain tree until it
4068 finds some nameserver records. It should be used with care.
4070 (2) There is a new type, "mxh", which is like "mx" except that it returns
4071 just the host names, not the priorities.
4073 (3) It is now possible to give a list of domains (or IP addresses) to be
4074 looked up. The behaviour when one of the lookups defers can be
4075 controlled by a keyword.
4077 (4) It is now possible to specify the separator character for use when
4078 multiple records are returned.
4080 38. The dnslists ACL condition has been extended: it is now possible to supply
4081 a list of IP addresses and/or domains to be looked up in a particular DNS
4084 39. Added log_selector=+queue_time_overall.
4086 40. When running the queue in the test harness, wait just a tad after forking a
4087 delivery process, to get repeatability of debugging output.
4089 41. Include certificate and key file names in error message when GnuTLS fails
4090 to set them up, because the GnuTLS error message doesn't include the name
4091 of the failing file when there is a problem reading it.
4093 42. Allow both -bf and -bF in the same test run.
4095 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4097 44. The "Exiscan patch" is now merged into the mainline Exim source.
4099 45. Sometimes the final signoff response after QUIT could fail to get
4100 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4101 < 0 before doing a fflush(). This bug looks as though it goes back to the
4102 introduction of TLS in release 3.20, but "sometimes" must have been rare
4103 because the tests only now provoked it.
4105 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4106 (this can affect the format of dates).
4108 47. exim_tidydb, when checking for the continued existence of a message for
4109 which it has found a message-specific retry record, was not finding
4110 messages that were in split spool directories. Consequently, it was
4111 deleting retry records that should have stayed in existence.
4113 48. Steve fixed some bugs in eximstats.
4115 49. The SPA authentication driver was not abandoning authentication and moving
4116 on to the next authenticator when an expansion was forced to fail,
4117 contradicting the general specification for all authenticators. Instead it
4118 was generating a temporary error. It now behaves as specified.
4120 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4121 (the order specifies the preference for clients). The order is now AES256,
4122 AES128, 3DES, ARCFOUR128.
4124 51. Small patch to Sieve code - explicitly set From: when generating an
4127 52. Exim crashed if a remote delivery caused a very long error message to be
4128 recorded - for instance if somebody sent an entire SpamAssassin report back
4129 as a large number of 550 error lines. This bug was coincidentally fixed by
4130 increasing the size of one of Exim's internal buffers (big_buffer) that
4131 happened as part of the Exiscan merge. However, to be on the safe side, I
4132 have made the code more robust (and fixed the comments that describe what
4135 53. Now that there can be additional text after "Completed" in log lines (if
4136 the queue_time_overall log selector is set), a one-byte patch to exigrep
4137 was needed to allow it to recognize "Completed" as not the last thing in
4140 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4141 patch that reportedly fixes this has been added. I am not expert enough to
4142 create a test for it. This is what the patch creator wrote:
4144 "I found a little strange behaviour of ldap code when working with
4145 Windows 2003 AD Domain, where users was placed in more than one
4146 Organization Units. When I tried to give exim partial DN, the exit code
4147 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4148 But simultaneously result of request was absolutely normal ldap result,
4149 so I produce this patch..."
4151 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4152 REFERENCE, so I have modified the code to exclude the patch when that macro
4155 55. Some experimental protocols are using DNS PTR records for new purposes. The
4156 keys for these records are domain names, not reversed IP addresses. The
4157 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4158 leaves it alone. Component reversal etc. now happens only for IP addresses.
4161 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4163 57. Double the size of the debug message buffer (to 2048) so that more of very
4164 long debug lines gets shown.
4166 58. The exicyclog utility now does better if the number of log files to keep
4167 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4169 59. Two changes related to the smtp_active_hostname option:
4171 (1) $smtp_active_hostname is now available as a variable.
4172 (2) The default for smtp_banner uses $smtp_active_hostname instead
4173 of $primary_hostname.
4175 60. The host_aton() function is supposed to be passed a string that is known
4176 to be a valid IP address. However, in the case of IPv6 addresses, it was
4177 not checking this. This is a hostage to fortune. Exim now panics and dies
4178 if the condition is not met. A case was found where this could be provoked
4179 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4180 components; fortuitously, this particular loophole had already been fixed
4181 by change 4.50/55 above.
4183 If there are any other similar loopholes, the new check in host_aton()
4184 itself should stop them being exploited. The report I received stated that
4185 data on the command line could provoke the exploit when Exim was running as
4186 exim, but did not say which command line option was involved. All I could
4187 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4188 running as the user.
4191 61. There was a buffer overflow vulnerability in the SPA authentication code
4192 (which came originally from the Samba project). I have added a test to the
4193 spa_base64_to_bits() function which I hope fixes it.
4196 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4197 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4199 63. The daemon start-up calls getloadavg() while still root for those OS that
4200 need the first call to be done as root, but it missed one case: when
4201 deliver_queue_load_max is set with deliver_drop_privilege. This is
4202 necessary for the benefit of the queue runner, because there is no re-exec
4203 when deliver_drop_privilege is set.
4205 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4206 This has been fixed.
4208 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4209 was in use, was not putting the data itself into the right store pool;
4210 consequently, it could be overwritten for a subsequent message in the same
4211 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4214 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4216 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4217 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4218 the function were treating the return as a boolean value, which happened to
4219 work because 0=false and not-0=true, but is not correct code.
4221 68. The host_aton() function was not handling scoped IPv6 addresses (those
4222 with, for example, "%eth0" on the end) correctly.
4224 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4225 negated items (that is, ~something) in unsigned ints. Some compilers
4226 apparently mutter when there is no cast.
4228 70. If an address verification called from an ACL failed, and did not produce a
4229 user-specific message (i.e. there was only a "system" message), nothing was
4230 put in $acl_verify_message. In this situation, it now puts the system
4233 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4234 session; change 4.31/43 added the unwanted input to the log line - except
4235 that it did not do this in the start of session case. It now does.
4237 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4238 This is wrong and can cause the other end to generate a synchronization
4239 error if it is another Exim or anything else that does the synchronization
4240 check. A QUIT command is no longer sent after a timeout.
4242 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4243 during host lookups.
4245 74. The defer_ok option of callout verification was not working if it was used
4246 when verifying addresses in header lines, that is, for this case:
4248 verify = header_sender/callout=defer_ok
4250 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4251 those file descriptors could be used for SMTP connections. If anything
4252 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4253 could be sent to the SMTP client, causing chaos. The daemon now opens
4254 stdin, stdout, and stderr to /dev/null when it puts itself into the
4257 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4258 log by default. The user can override this with suitable Perl magic.
4260 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4261 the log message when discard triggers, was not working for the DATA ACL or
4262 for the non-SMTP ACL.
4264 78. Error message wording change in sieve.c.
4266 79. If smtp_accept_max_per_host was set, the number of connections could be
4267 restricted to fewer than expected, because the daemon was trying to set up
4268 a new connection before checking whether the processes handling previous
4269 connections had finished. The check for completed processes is now done
4270 earlier. On busy systems, this bug wouldn't be noticed because something
4271 else would have woken the daemon, and it would have reaped the completed
4274 80. If a message was submitted locally by a user whose login name contained one
4275 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4276 It caused a spool format error. I have fixed the spool reading code. A
4277 related problem was that the "from" clause in the Received: line became
4278 illegal because of the space(s). It is now covered by ${quote_local_part.
4280 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4283 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4285 83. Patch from Sieve maintainer to fix unterminated string problem in
4286 "vacation" handling.
4288 84. Some minor changes to the Linux configuration files to help with other
4289 OS variants using glibc.
4291 85. One more patch for Sieve to update vacation handling to latest spec.
4294 ----------------------------------------------------
4295 See the note above about the 4.44 and 4.50 releases.
4296 ----------------------------------------------------
4302 1. Change 4.43/35 introduced a bug that caused file counts to be
4303 incorrectly computed when quota_filecount was set in an appendfile
4306 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4307 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4310 3. Give more explanation in the error message when the command for a transport
4311 filter fails to execute.
4313 4. There are several places where Exim runs a non-Exim command in a
4314 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4315 was being done only for the command run by the queryprogram router. It is
4316 now done for all such subprocesses. The other cases are: ${run, transport
4317 filters, and the commands run by the lmtp and pipe transports.
4319 5. Some older OS have a limit of 256 on the maximum number of file
4320 descriptors. Exim was using setrlimit() to set 1000 as a large value
4321 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4322 systems. I've change it so that if it can't get 1000, it tries for 256.
4324 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4325 was an oversight, and furthermore, ever since the addition of extra
4326 controls (e.g. 4.43/32), the checks on when to allow different forms of
4327 "control" were broken. There should now be diagnostics for all cases when a
4328 control that does not make sense is encountered.
4330 7. $recipients is now available in the predata ACL (oversight).
4332 8. Tidy the search cache before the fork to do a delivery from a message
4333 received from the command line. Otherwise the child will trigger a lookup
4334 failure and thereby defer the delivery if it tries to use (for example) a
4335 cached ldap connection that the parent has called unbind on.
4337 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4338 of $address_data from the recipient verification was clobbered by the
4339 sender verification.
4341 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4342 was its contents. (It was OK if the option was not defined at all.)
4344 11. A "Completed" log line is now written for messages that are removed from
4345 the spool by the -Mrm option.
4347 12. $host_address is now set to the target address during the checking of
4348 ignore_target_hosts.
4350 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4351 being passed; this would have caused $sender_host_name to have been used if
4352 matching the list had actually called for a host name (not very likely,
4353 since this list is usually IP addresses). A host name is now passed as
4356 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4357 code that specifies a non-restarting handler (typically sigaction() in
4358 modern systems) in an attempt to fix a rare and obscure crash bug.
4360 15. Narrowed the window for a race in the daemon that could cause it to ignore
4361 SIGCHLD signals. This is not a major problem, because they are used only to
4362 wake it up if nothing else does.
4364 16. A malformed maildirsize file could cause Exim to calculate negative values
4365 for the mailbox size or file count. Odd effects could occur as a result.
4366 The maildirsize information is now recalculated if the size or filecount
4369 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4370 support for a long time. Removed HAVE_SYS_VFS_H.
4372 18. Updated exipick to current release from John Jetmore.
4374 19. Allow an empty sender to be matched against a lookup in an address list.
4375 Previously the only cases considered were a regular expression, or an
4378 20. Exim went into a mad DNS lookup loop when doing a callout where the
4379 host was specified on the transport, if the DNS lookup yielded more than
4382 21. The RFC2047 encoding function was originally intended for short strings
4383 such as real names; it was not keeping to the 75-character limit for
4384 encoded words that the RFC imposes. It now respects the limit, and
4385 generates multiple encoded words if necessary. To be on the safe side, I
4386 have increased the buffer size for the ${rfc2047: expansion operator from
4389 22. Failure to deliver a bounce message always caused it to be frozen, even if
4390 there was an errors_to setting on the router. The errors_to setting is now
4393 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4394 canonical form (fully expanded) before being placed in
4395 $sender_host_address.
4397 24. Updated eximstats to version 1.33
4399 25. Include certificate and key file names in error message when GnuTLS fails
4400 to set them up, because the GnuTLS error message doesn't include the name
4401 of the failing file when there is a problem reading it.
4403 26. Expand error message when OpenSSL has problems setting up cert/key files.
4406 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4407 (this can affect the format of dates).
4409 28. exim_tidydb, when checking for the continued existence of a message for
4410 which it has found a message-specific retry record, was not finding
4411 messages that were in split spool directories. Consequently, it was
4412 deleting retry records that should have stayed in existence.
4414 29. eximstats updated to version 1.35
4415 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4416 1.35 - bugfix such that pie charts by volume are generated correctly
4418 30. The SPA authentication driver was not abandoning authentication and moving
4419 on to the next authenticator when an expansion was forced to fail,
4420 contradicting the general specification for all authenticators. Instead it
4421 was generating a temporary error. It now behaves as specified.
4423 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4424 (the order specifies the preference for clients). The order is now AES256,
4425 AES128, 3DES, ARCFOUR128.
4427 31. Small patch to Sieve code - explicitly set From: when generating an
4430 32. Exim crashed if a remote delivery caused a very long error message to be
4431 recorded - for instance if somebody sent an entire SpamAssassin report back
4432 as a large number of 550 error lines. This bug was coincidentally fixed by
4433 increasing the size of one of Exim's internal buffers (big_buffer) that
4434 happened as part of the Exiscan merge. However, to be on the safe side, I
4435 have made the code more robust (and fixed the comments that describe what
4438 33. Some experimental protocols are using DNS PTR records for new purposes. The
4439 keys for these records are domain names, not reversed IP addresses. The
4440 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4441 leaves it alone. Component reversal etc. now happens only for IP addresses.
4444 34. The host_aton() function is supposed to be passed a string that is known
4445 to be a valid IP address. However, in the case of IPv6 addresses, it was
4446 not checking this. This is a hostage to fortune. Exim now panics and dies
4447 if the condition is not met. A case was found where this could be provoked
4448 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4449 components; fortuitously, this particular loophole had already been fixed
4450 by change 4.50/55 or 4.44/33 above.
4452 If there are any other similar loopholes, the new check in host_aton()
4453 itself should stop them being exploited. The report I received stated that
4454 data on the command line could provoke the exploit when Exim was running as
4455 exim, but did not say which command line option was involved. All I could
4456 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4457 running as the user.
4460 35. There was a buffer overflow vulnerability in the SPA authentication code
4461 (which came originally from the Samba project). I have added a test to the
4462 spa_base64_to_bits() function which I hope fixes it.
4465 36. The daemon start-up calls getloadavg() while still root for those OS that
4466 need the first call to be done as root, but it missed one case: when
4467 deliver_queue_load_max is set with deliver_drop_privilege. This is
4468 necessary for the benefit of the queue runner, because there is no re-exec
4469 when deliver_drop_privilege is set.
4471 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4472 was in use, was not putting the data itself into the right store pool;
4473 consequently, it could be overwritten for a subsequent message in the same
4474 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4477 38. Sometimes the final signoff response after QUIT could fail to get
4478 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4479 < 0 before doing a fflush(). This bug looks as though it goes back to the
4480 introduction of TLS in release 3.20, but "sometimes" must have been rare
4481 because the tests only now provoked it.
4487 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4488 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4489 Now it returns an integer. A number of calls were still expecting a T/F
4490 return. Fortuitously, in all cases, the tests worked in OK situations,
4491 which is the norm. However, things would have gone wrong on any write
4492 failures on the smtp file descriptor. This function is used when sending
4493 messages over SMTP and also when doing verify callouts.
4495 2. When Exim is called to do synchronous delivery of a locally submitted
4496 message (the -odf or -odi options), it no longer closes stderr before doing
4499 3. Implemented the mua_wrapper option.
4501 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4503 5. Implemented the functions header_remove(), header_testname(),
4504 header_add_at_position(), and receive_remove_recipient(), and exported them
4507 6. If an ACL "warn" statement specified the addition of headers, Exim already
4508 inserted X-ACL-Warn: at the start if there was no header name. However, it
4509 was not making this test for the second and subsequent header lines if
4510 there were newlines in the string. This meant that an invalid header could
4511 be inserted if Exim was badly configured.
4513 7. Allow an ACL "warn" statement to add header lines at the start or after all
4514 the Received: headers, as well as at the end.
4516 8. Added the rcpt_4xx retry error code.
4518 9. Added postmaster_mailfrom=xxx to callout verification option.
4520 10. Added mailfrom=xxxx to the callout verification option, for verify=
4523 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4524 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4526 12. Inserted some casts to stop certain compilers warning when using pointer
4527 differences as field lengths or precisions in printf-type calls (mostly
4528 affecting debugging statements).
4530 13. Added optional readline() support for -be (dynamically loaded).
4532 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4533 same clock tick as a message's arrival, so that its received time was the
4534 same as the "first fail" time on the retry record, and that message
4535 remained on the queue past the ultimate address timeout, every queue runner
4536 would try a delivery (because it was past the ultimate address timeout) but
4537 after another failure, the ultimate address timeout, which should have then
4538 bounced the address, did not kick in. This was a "< instead of <=" error;
4539 in most cases the first failure would have been in the next clock tick
4540 after the received time, and all would be well.
4542 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4543 being recognized when the domain list was tested by the match_domain
4544 condition in an expansion string.
4546 16. Added the ${str2b64: operator.
4548 17. Exim was always calling setrlimit() to set a large limit for the number of
4549 processes, without checking whether the existing limit was already
4550 adequate. (It did check for the limit on file descriptors.) Furthermore,
4551 errors from getrlimit() and setrlimit() were being ignored. Now they are
4552 logged to the main and panic logs, but Exim does carry on, to try to do its
4553 job under whatever limits there are.
4555 18. Imported PCRE 5.0.
4557 19. Trivial typo in log message " temporarily refused connection" (the leading
4560 20. If the log selector return_path_on_delivery was set and an address was
4561 redirected to /dev/null, the delivery process crashed because it assumed
4562 that a return path would always be set for a "successful" delivery. In this
4563 case, the whole delivery is bypassed as an optimization, and therefore no
4566 21. Internal re-arrangement: the function for sending a challenge and reading
4567 a response while authentication was assuming a zero-terminated challenge
4568 string. It's now changed to take a pointer and a length, to allow for
4569 binary data in such strings.
4571 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4573 23. Exim was not respecting finduser_retries when seeking the login of the
4574 uid under which it was called; it was always trying 10 times. (The default
4575 setting of finduser_retries is zero.) Also, it was sleeping after the final
4576 failure, which is pointless.
4578 24. Implemented tls_on_connect_ports.
4580 25. Implemented acl_smtp_predata.
4582 26. If the domain in control=submission is set empty, Exim assumes that the
4583 authenticated id is a complete email address when it generates From: or
4584 Sender: header lines.
4586 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4587 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4588 chown and chgrp in /bin and hostname in /usr/bin.
4590 28. Exim was keeping the "process log" file open after each use, just as it
4591 does for the main log. This opens the possibility of it remaining open for
4592 long periods when the USR1 signal hits a daemon. Occasional processlog
4593 errors were reported, that could have been caused by this. Anyway, it seems
4594 much more sensible not to leave this file open at all, so that is what now
4597 29. The long-running daemon process does not normally write to the log once it
4598 has entered its main loop, and it closes the log before doing so. This is
4599 so that log files can straightforwardly be renamed and moved. However,
4600 there are a couple of unusual error situations where the daemon does write
4601 log entries, and I had neglected to close the log afterwards.
4603 30. The text of an SMTP error response that was received during a remote
4604 delivery was being truncated at 512 bytes. This is too short for some of
4605 the long messages that one sometimes sees. I've increased the limit to
4608 31. It is now possible to make retry rules that apply only when a message has a
4609 specific sender, in particular, an empty sender.
4611 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4612 it possible to be selective about when SMTP synchronization is enforced.
4614 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4616 32. Implemented hosts_connection_nolog.
4618 33. Added an ACL for QUIT.
4620 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4623 35. Added mailbox_size and mailbox_filecount to appendfile.
4625 36. Added control = no_multiline_responses to ACLs.
4627 37. There was a bug in the logic of the code that waits for the clock to tick
4628 in the case where the clock went backwards by a substantial amount such
4629 that the microsecond fraction of "now" was more than the microsecond
4630 fraction of "then" (but the whole seconds number was less).
4632 38. Added support for the libradius Radius client library this is found on
4633 FreeBSD (previously only the radiusclient library was supported).
4639 1. When certain lookups returned multiple values in the form name=value, the
4640 quoting of the values was not always being done properly. Specifically:
4641 (a) If the value started with a double quote, but contained no whitespace,
4643 (b) If the value contained whitespace other than a space character (i.e.
4644 tabs or newlines or carriage returns) it was not quoted.
4645 This fix has been applied to the mysql and pgsql lookups by writing a
4646 separate quoting function and calling it from the lookup code. The fix
4647 should probably also be applied to nisplus, ibase and oracle lookups, but
4648 since I cannot test any of those, I have not disturbed their existing code.
4650 2. A hit in the callout cache for a specific address caused a log line with no
4651 reason for rejecting RCPT. Now it says "Previous (cached) callout
4652 verification failure".
4654 3. There was an off-by-one bug in the queryprogram router. An over-long
4655 return line was truncated at 256 instead of 255 characters, thereby
4656 overflowing its buffer with the terminating zero. As well as fixing this, I
4657 have increased the buffer size to 1024 (and made a note to document this).
4659 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4660 when Exim is waiting for an SMTP response from a remote server, Exim
4661 restarts its select() call on the socket, thereby resetting its timeout.
4662 This is not a problem when such interrupts are rare. Somebody set up a cron
4663 job to run exiwhat every 2 minutes, which is less than the normal select()
4664 timeout (5 or 10 minutes). This meant that the select() timeout never
4665 kicked in because it was always reset. I have fixed this by comparing the
4666 time when an interrupt arrives with the time at the start of the first call
4667 to select(). If more time than the timeout has elapsed, the interrupt is
4668 treated as a timeout.
4670 5. Some internal re-factoring in preparation for the addition of Sieve
4671 extensions (by MH). In particular, the "personal" test is moved to a
4672 separate function, and given an option for scanning Cc: and Bcc: (which is
4673 not set for Exim filters).
4675 6. When Exim created an email address using the login of the caller as the
4676 local part (e.g. when creating a From: or Sender: header line), it was not
4677 quoting the local part when it contained special characters such as @.
4679 7. Installed new OpenBSD configuration files.
4681 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4682 try to make them clearer.
4684 9. Callout options, other than the timeout value, were being ignored when
4685 verifying sender addresses in header lines. For example, when using
4687 verify = header_sender/callout=no_cache
4689 the cache was (incorrectly) being used.
4691 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4692 only the Cygwin environment.
4694 11. When return_path_on_delivery was set as a log selector, if different remote
4695 addresses in the same message used different return paths and parallel
4696 remote delivery occurred, the wrong values would sometimes be logged.
4697 (Whenever a remote delivery process finished, the return path value from
4698 the most recently started remote delivery process was logged.)
4700 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4701 lines when AUTH and/or TLS are in use. This is the "received protocol"
4702 field. Exim used to use "asmtp" for authenticated SMTP, without any
4703 indication (in the protocol name) for TLS use. Now it follows the RFC and
4704 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4705 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4706 names appear in log lines as well as in Received: header lines.
4708 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4709 extensions, and comparison tests, and to fix some bugs.
4711 14. Changes to the "personal" filter test:
4713 (1) The test was buggy in that it was just doing the equivalent of
4714 "contains" tests on header lines. For example, if a user's address was
4715 anne@some.where, the "personal" test would incorrectly be true for
4717 To: susanne@some.where
4719 This test is now done by extracting each address from the header in turn,
4720 and checking the entire address. Other tests that are part of "personal"
4721 are now done using regular expressions (for example, to check local parts
4722 of addresses in From: header lines).
4724 (2) The list of non-personal local parts in From: addresses has been
4725 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4726 taken from the Sieve specification recommendations.
4728 (3) If the message contains any header line starting with "List-" it is
4729 treated as non-personal.
4731 (4) The test for "circular" in the Subject: header line has been removed
4732 because it now seems ill-conceived.
4734 15. Minor typos in src/EDITME comments corrected.
4736 16. Installed latest exipick from John Jetmore.
4738 17. If headers_add on a router specified a text string that was too long for
4739 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4740 of string_sprintf() is now avoided.
4742 18. $message_body_size was not set (it was always zero) when running the DATA
4743 ACL and the local_scan() function.
4745 19. For the "mail" command in an Exim filter, no default was being set for
4746 the once_repeat time, causing a random time value to be used if "once" was
4747 specified. (If the value happened to be <= 0, no repeat happened.) The
4748 default is now 0s, meaning "never repeat". The "vacation" command was OK
4749 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4750 (I found it when inspecting the code).
4752 20. There is now an overall timeout for performing a callout verification. It
4753 defaults to 4 times the callout timeout, which applies to individual SMTP
4754 commands during the callout. The overall timeout applies when there is more
4755 than one host that can be tried. The timeout is checked before trying the
4756 next host. This prevents very long delays if there are a large number of
4757 hosts and all are timing out (e.g. when the network connections are timing
4758 out). The value of the overall timeout can be changed by specifying an
4759 additional sub-option for "callout", called "maxwait". For example:
4761 verify = sender/callout=5s,maxwait=20s
4763 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4764 to the end before writing, but this should make it even safer).
4766 22. Exim was forgetting that it had advertised PIPELINING for the second and
4767 subsequent messages on an SMTP connection. It was also not resetting its
4768 memory on STARTTLS and an internal HELO.
4770 23. When Exim logs an SMTP synchronization error within a session, it now
4771 records whether PIPELINING has been advertised or not.
4773 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4774 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4775 rather than long int.
4777 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4779 26. Added the never_mail option to autoreply.
4785 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4786 crash if the getsockname() call failed; this can happen if a connection is
4787 closed very soon after it is established. The problem was simply in the
4788 order in which certain operations were done, causing Exim to try to write
4789 to the SMTP stream before it had set up the file descriptor. The bug has
4790 been fixed by making things happen in the correct order.
4796 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4797 before the connection was closed, thus losing the rejection response.
4799 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4800 some early Solaris releases, but causes trouble in current releases where
4801 socklen_t is defined.
4803 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4806 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4809 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4810 that releasing the top of it at the end releases what was used for sub-
4811 expansions (unless the block got too big). However, discard this block if
4812 the first thing is a variable or header, so that we can use its block when
4813 it is dynamic (useful for very large $message_headers, for example).
4815 6. Lookups now cache *every* query, not just the most recent. A new, separate
4816 store pool is used for this. It can be recovered when all lookup caches are
4817 flushed. Lookups now release memory at the end of their result strings.
4818 This has involved some general refactoring of the lookup sources.
4820 7. Some code has been added to the store_xxx() functions to reduce the amount
4821 of flapping under certain conditions.
4823 8. log_incoming_interface used to affect only the <= reception log lines. Now
4824 it causes the local interface and port to be added to several more SMTP log
4825 lines, for example "SMTP connection from", and rejection lines.
4827 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4829 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4831 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4832 could be overwritten at the end of the current message (or the start of a
4833 new message if it was set in a HELO ACL). The value is now preserved for
4834 the duration of the SMTP connection.
4836 12. If a transport had a headers_rewrite setting, and a matching header line
4837 contained an unqualified address, that address was qualified, even if it
4838 did not match any rewriting rules. The underlying bug was that the values
4839 of the flags that permit the existence of unqualified sender and recipient
4840 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4841 non-local messages, and by -bnq for local messages) were not being
4842 preserved with the message after it was received.
4844 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4845 "next input=" as part of the text comprising the host identity instead of
4846 the correct text. The code was using the same buffer for two different
4847 strings. However, depending on which order the printing function evaluated
4848 its arguments, the bug did not always show up. Under Linux, for example, my
4849 test suite worked just fine.
4851 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4852 doesn't work with some older versions of Perl. It has been changed to "my",
4853 which in any case is probably the better facility to use.
4855 15. A really picky compiler found some instances of statements for creating
4856 error messages that either had too many or two few arguments for the format
4859 16. The size of the buffer for calls to the DNS resolver has been increased
4860 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4861 for addresses that have a lot of PTR records. This alleviates a problem; it
4862 does not fully solve it.
4864 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4865 buffer now truncates the list and logs the incident, which is the same
4866 action as happens when Exim is looking up a host name and its aliases.
4867 Previously in this situation something unpredictable would happen;
4868 sometimes it was "internal error: store_reset failed".
4870 18. If a server dropped the connection unexpectedly when an Exim client was
4871 using GnuTLS and trying to read a response, the client delivery process
4872 crashed while trying to generate an error log message.
4874 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4875 string, for example:
4877 warn message = H1: something\nH2: something
4879 the text was added as a single header line from Exim's point of view
4880 though it ended up OK in the delivered message. However, searching for the
4881 second and subsequent header lines using $h_h2: did not work. This has been
4882 fixed. Similarly, if a system filter added multiple headers in this way,
4883 the routers could not see them.
4885 20. Expanded the error message when iplsearch is called with an invalid key to
4886 suggest using net-iplsearch in a host list.
4888 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4889 ACLs are no longer actually imposed (and a message to that effect is
4892 22. If a "gecos" field in a passwd entry contained escaped characters, in
4893 particular, if it contained a \" sequence, Exim got it wrong when building
4894 a From: or a Sender: header from that name. A second bug also caused
4895 incorrect handling when an unquoted " was present following a character
4896 that needed quoting.
4898 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4899 was not being matched caselessly.
4901 24. Arranged for all hyphens in the exim.8 source to be escaped with
4904 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4905 properly. Recipient callout cache records were still being keyed to include
4906 the sender, even when use_sender was set false. This led to far more
4907 callouts that were necessary. The sender is no longer included in the key
4908 when use_sender is false.
4910 26. Added "control = submission" modifier to ACLs.
4912 27. Added the ${base62d: operator to decode base 62 numbers.
4914 28. dnsdb lookups can now access SRV records.
4916 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4917 the configuration file.
4919 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4920 (-v) mode. This makes the output for a verbose queue run more intelligible.
4922 31. Added a use_postmaster feature to recipient callouts.
4924 32. Added the $body_zerocount variable, containing the number of binary zero
4925 bytes in the message body.
4927 33. The time of last modification of the "new" subdirectory is now used as the
4928 "mailbox time last read" when there is a quota error for a maildir
4931 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4933 35. Added +ignore_unknown as a special item in host lists.
4935 36. Code for decoding IPv6 addresses in host lists is now included, even if
4936 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4937 address was recognized as an IP address, but was then not correctly decoded
4938 into binary, causing unexpected and incorrect effects when compared with
4945 1. Very minor rewording of debugging text in manualroute to say "list of
4946 hosts" instead of "hostlist".
4948 2. If verify=header_syntax was set, and a header line with an unqualified
4949 address (no domain) and a large number of spaces between the end of the
4950 name and the colon was received, the reception process suffered a buffer
4951 overflow, and (when I tested it) crashed. This was caused by some obsolete
4952 code that should have been removed. The fix is to remove it!
4954 3. When running in the test harness, delay a bit after writing a bounce
4955 message to get a bit more predictability in the log output.
4957 4. Added a call to search_tidyup() just before forking a reception process. In
4958 theory, someone could use a lookup in the expansion of smtp_accept_max_
4959 per_host which, without the tidyup, could leave open a database connection.
4961 5. Added the variables $recipient_data and $sender_data which get set from a
4962 lookup success in an ACL "recipients" or "senders" condition, or a router
4963 "senders" option, similar to $domain_data and $local_part_data.
4965 6. Moved the writing of debug_print from before to after the "senders" test
4968 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4969 problems for message scanning, either using a data ACL, or using
4970 local_scan() because the Received: header was not generated till after they
4971 were called (in order to set the time as the time of reception completion).
4972 I have revised the way this works. The header is now generated after the
4973 body is received, but before the ACL or local_scan() are called. After they
4974 are run, the timestamp in the header is updated.
4980 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4981 before starting a queue runner without re-exec. This happened only when
4982 deliver_drop_privilege was set or when the Exim user was set to root. The
4983 effect of the bug was that timeouts during subsequent deliveries caused
4984 crashes instead of being properly handled. The handler is now left at its
4985 default (and expected) setting.
4987 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4988 message, again when deliver_drop_privilege is set or Exim is run as root.
4989 The bug described in (1) was not present in this case, but the tidying up
4990 of the other signals was missing. I have made the two cases consistent.
4992 3. The ignore_target_hosts setting on a manualroute router was being ignored
4993 for hosts that were looked up using the /MX notation.
4995 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4998 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4999 operated on the sender address. After changing the $sender_address to <>
5000 for the sender address verify, Exim was re-instated it as the original
5001 (before rewriting) address, but remembering that it had rewritten it, so it
5002 wasn't rewriting it again. This bug also had the effect of breaking the
5003 sender address verification caching when the sender address was rewritten.
5005 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5006 This has been changed so that if the ip literal address matches
5007 ignore_target_hosts, the router declines.
5009 7. Added expansion conditions match_domain, match_address, and match_local_
5010 part (NOT match_host).
5012 8. The placeholder for the Received: header didn't have a length field set.
5014 9. Added code to Exim itself and to exim_lock to test for a specific race
5015 condition that could lead to file corruption when using MBX delivery. The
5016 issue is with the lockfile that is created in /tmp. If this file is removed
5017 after a process has opened it but before that process has acquired a lock,
5018 there is the potential for a second process to recreate the file and also
5019 acquire a lock. This could lead to two Exim processes writing to the file
5020 at the same time. The added code performs the same test as UW imapd; it
5021 checks after acquiring the lock that its file descriptor still refers to
5022 the same named file.
5024 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5025 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5028 11. Added the smtp_active_hostname option. If used, this will typically be made
5029 to depend on the incoming interface address. Because $interface_address is
5030 not set up until the daemon has forked a reception process, error responses
5031 that can happen earlier (such as "too many connections") no longer contain
5034 12. If an expansion in a condition on a "warn" statement fails because a lookup
5035 defers, the "warn" statement is abandoned, and the next ACL statement is
5036 processed. Previously this caused the whole ACL to be aborted.
5038 13. Added the iplsearch lookup type.
5040 14. Added ident_timeout as a log selector.
5042 15. Added tls_certificate_verified as a log selector.
5044 16. Added a global option tls_require_ciphers (compare the smtp transport
5045 option of the same name). This controls incoming TLS connections.
5047 17. I finally figured out how to make tls_require_ciphers do a similar thing
5048 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5049 before starting the TLS session.
5051 18. Tabs are now shown as \t in -bP output.
5053 19. If the log selector return_path_on_delivery was set, Exim crashed when
5054 bouncing a message because it had too many Received: header lines.
5056 20. If two routers both had headers_remove settings, and the first one included
5057 a superfluous trailing colon, the final name in the first list and the
5058 first name in the second list were incorrectly joined into one item (with a
5059 colon in the middle).
5065 1. Added -C and -D options to the exinext utility, mainly to make it easier
5066 to include in the automated testing, but these could be helpful when
5067 multiple configurations are in use.
5069 2. The exinext utility was not formatting the output nicely when there was
5070 an alternate port involved in the retry record key, nor when there was a
5071 message id as well (for retries that were specific to a specific message
5072 and a specific host). It was also confused by IPv6 addresses, because of
5073 the additional colons they contain. I have fixed the IPv4 problem, and
5074 patched it up to do a reasonable job for IPv6.
5076 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5077 delivery, the log line now contains "pipelined" if PIPELINING was used.
5079 4. An SMTP transport process used to panic and die if the bind() call to set
5080 an explicit outgoing interface failed. This has been changed; it is now
5081 treated in the same way as a connect() failure.
5083 5. A reference to $sender_host_name in the part of a conditional expansion
5084 that was being skipped was still causing a DNS lookup. This no longer
5087 6. The def: expansion condition was not recognizing references to header lines
5088 that used bh_ and bheader_.
5090 7. Added the _cache feature to named lists.
5092 8. The code for checking quota_filecount in the appendfile transport was
5093 allowing one more file than it should have been.
5095 9. For compatibility with Sendmail, the command line option
5103 and sets the incoming protocol and host name (for trusted callers). The
5104 host name and its colon can be omitted when only the protocol is to be set.
5105 Note the Exim already has two private options, -pd and -ps, that refer to
5106 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5107 "s", but I don't think that's a major issue.
5109 10. A number of refactoring changes to the code, none of which should affect
5112 (a) The number of logging options was getting close to filling up the
5113 32-bit word that was used as a bit map. I have split them into two classes:
5114 those that are passed in the argument to log_write(), and those that are
5115 only ever tested independently outside of that function. These are now in
5116 separate 32-bit words, so there is plenty of room for expansion again.
5117 There is no change in the user interface or the logging behaviour.
5119 (b) When building, for example, log lines, the code previously used a
5120 macro that called string_cat() twice, in order to add two strings. This is
5121 not really sufficiently general. Furthermore, there was one instance where
5122 it was actually wrong because one of the argument was used twice, and in
5123 one call a function was used. (As it happened, calling the function twice
5124 did not affect the overall behaviour.) The macro has been replaced by a
5125 function that can join an arbitrary number of extra strings onto a growing
5128 (c) The code for expansion conditions now uses a table and a binary chop
5129 instead of a serial search (which was left over from when there were very
5130 few conditions). Also, it now recognizes conditions like "pam" even when
5131 the relevant support is not compiled in: a suitably worded error message is
5132 given if an attempt is made to use such a condition.
5134 11. Added ${time_interval:xxxxx}.
5136 12. A bug was causing one of the ddress fields not to be passed back correctly
5137 from remote delivery subprocesses. The field in question was not being
5138 subsequently used, so this caused to problems in practice.
5140 13. Added new log selectors queue_time and deliver_time.
5142 14. Might have fixed a bug in maildirsizefile handling that threw up
5143 "unexpected character" debug warnings, and recalculated the data
5144 unnecessarily. In any case, I expanded the warning message to give more
5147 15. Added the message "Restricted characters in address" to the statements in
5148 the default ACL that block characters like @ and % in local parts.
5150 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5151 Three changes have been made:
5153 (a) There was a serious bug; a negative response to MAIL caused the whole
5154 recipient domain to be cached as invalid, thereby blocking all messages
5155 to all local parts at the same domain, from all senders. This bug has
5156 been fixed. The domain is no longer cached after a negative response to
5157 MAIL if the sender used is not empty.
5159 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5162 (c) A new callout option, "use_sender" has been added for people who want
5163 the modified behaviour.
5169 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5172 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5173 indeed breaks things for older releases.
5175 3. Added additional logging to the case where there is a problem reading data
5176 from a filter that is running in a subprocess using a pipe, in order to
5177 try to track down a specific problem.
5179 4. Testing facility fudge: when running in the test harness and attempting
5180 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5181 getting "No route to host". Convert this to a timeout.
5183 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5186 6. Some OS don't have socklen_t but use size_t instead. This affects the
5187 fifth argument of getsockopt() amongst other things. This is now
5188 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5189 can be set for individual OS. I have set it for SunOS5, OSF1, and
5190 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5191 some earlier ones do not.
5193 7. Change 4.30/15 was not doing the test caselessly.
5195 8. The standard form for an IPv6 address literal was being rejected by address
5196 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5197 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5198 this, as well as the form without the "IPv6" on the front (but only when
5199 address literals are enabled, of course).
5201 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5203 10. Exim crashed if a message with an empty sender address specified by -f
5204 encountered a router with an errors_to setting. This could be provoked only
5205 by a command such as
5209 where an empty string was supplied; "<>" did not hit this bug.
5211 11. Installed PCRE release 4.5.
5213 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5214 remained set. It is now erased.
5216 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5217 times from message ids (which are base 36 rather than the normal 62).
5219 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5220 were being counted as actual protocol errors, and logged if the log
5221 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5222 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5223 and DATA following a set of rejected RCPTs do not count as protocol errors.
5224 In other words, Exim assumes they were pipelined, though this may not
5225 actually be the case. Of course, in all cases the client gets an
5226 appropriate error code.
5228 15. If a lookup fails in an ACL condition, a message about the failure may
5229 be available; it is used if testing the ACL cannot continue, because most
5230 such messages specify what the cause of the deferral is. However, some
5231 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5232 that caused an old message to be retained and used if a later statement
5233 caused a defer, replacing the real cause of the deferral.
5235 16. If an IP address had so many PTR records that the DNS lookup buffer
5236 was not large enough to hold them, Exim could crash while trying to process
5237 the truncated data. It now detects and logs this case.
5239 17. Further to 4.21/58, another change has been made: if (and only if) the
5240 first line of a message (the first header line) ends with CRLF, a bare LF
5241 in a subsequent header line has a space inserted after it, so as not to
5242 terminate the header.
5244 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5245 unnecessarily, used atoi() instead of strtol(), and didn't check the
5246 termination when getting file sizes from file names by regex.
5248 19. Completely re-implemented the support for maildirsize files, in the light
5249 of a number of problems with the previous contributed implementation
5250 (4.30/29). In particular:
5252 . If the quota is zero, the maildirsize file is maintained, but no quota is
5255 . If the maildir directory does not exist, it is created before any attempt
5256 to write a maildirsize file.
5258 . The quota value in the file is just a cache; if the quota is changed in
5259 the transport, the new value overrides.
5261 . A regular expression is available for excluding directories from the
5264 20. The autoreply transport checks the characters in options that define the
5265 message's headers; it allows continued headers, but it was checking with
5266 isspace() after an embedded newline instead of explicitly looking for a
5269 21. If all the "regular" hosts to which an address was routed had passed their
5270 expiry times, and had not reached their retry times, the address was
5271 bounced, even if fallback hosts were defined. Now Exim should go on to try
5274 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5275 equivalent code in the SMTP transport. Some hosts send humungous responses
5276 to HELO/EHLO, more than 1024 it seems.
5278 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5279 gives compiler warnings in some environments. I've now done it "properly",
5282 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5283 (because of problems with the built-in one) was declared to return uschar *
5284 instead of char *, causing compiler failure.
5286 25. Fixed a file descriptor leak when processing alias/forward files.
5288 26. Fixed a minor format string issue in dbfn.c.
5290 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5292 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5293 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5296 29. When a sender address is verified, it is cached, to save repeating the test
5297 when there is more than one recipient in a message. However, when the
5298 verification involves a callout, it is possible for different callout
5299 options to be set for different recipients. It is too complicated to keep
5300 track of this in the cache, so now Exim always runs a verification when a
5301 callout is required, relying on the callout cache for the optimization.
5302 The overhead is duplication of the address routing, but this should not be
5305 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5306 to be verified with callout=postmaster, and the main callout worked but the
5307 postmaster check failed, the verification correctly failed. However, if a
5308 subsequent RCPT command asked for sender verification *without* the
5309 postmaster check, incorrect caching caused this verification also to fail,
5312 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5313 it was not caching the DNS options (qualify_single, search_parents) that
5314 were used when the lookup failed. A subsequent lookup with different
5315 options therefore always gave the same answer, though there were cases
5316 where it should not have. (Example: a "domains = !$mx_any" option on a
5317 dnslookup router: the "domains" option is always processed without any
5318 widening, but the router might have qualify_single set.) Now Exim uses the
5319 cached value only when the same options are set.
5321 32. Added John Jetmore's "exipick" utility to the distribution.
5323 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5324 than a timeout (e.g. a certificate is required, and is not provided), an
5325 Exim server now closes the connection immediately. Previously it waited for
5326 the client to close - but if the client is SSL, it seems that they each
5327 wait for each other, leading to a delay before one of them times out.
5329 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5330 maintained 0.8.x compatibility because I don't think many are using it, and
5331 it is clearly obsolete.
5333 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5336 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5337 client certificate was expired. A simple patch fixes this, though I don't
5338 understand the full logic of why the verify callback is called multiple
5341 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5342 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5343 which causes problems with some clients (such as the Certicom SSL Plus
5344 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5345 disables the coutermeasure allowing Eudora to connect."
5347 38. Exim was not checking that a write() to a log file succeeded. This could
5348 lead to Bad Things if a log got too big, in particular if it hit a file
5349 size limit. Exim now panics and dies if it cannot write to a log file, just
5350 as it does if it cannot open a log file.
5352 39. Modified OS/Makefile-Linux so that it now contains
5354 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5356 The two -D definitions ensure that Exim is compiled with large file
5357 support, which makes it possible to handle log files that are bigger than
5360 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5361 instance) a domain was checked against a named list that involved a lookup,
5362 causing $domain_data to be set, then another domain was checked against the
5363 same list, then the first domain was re-checked, the value of $domain_data
5364 after the final check could be wrong. In particular, if the second check
5365 failed, it could be set empty. This bug probably also applied to
5368 41. The strip_trailing_dot option was not being applied to the address given
5369 with the -f command-line option.
5371 42. The code for reading a message's header from the spool was incrementing
5372 $received_count, but never initializing it. This meant that the value was
5373 incorrect (doubled) while delivering a message in the same process in which
5374 it was received. In the most common configuration of Exim, this never
5375 happens - a fresh exec is done - but it can happen when
5376 deliver_drop_privilege is set.
5378 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5379 it now includes up to 150 characters of the unexpected data in the log
5382 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5383 and building data strings. The size of both of these buffers was 10 000
5384 bytes - far larger than anybody would *ever* want, thought I. Needless to
5385 say, somebody hit the limit. I have increased the maximum line length to
5386 20 000 and the maximum data length of concatenated lines to 100 000. I have
5387 also fixed two bugs, because there was no checking on these buffers. Tsk,
5388 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5389 buffer is too small.
5391 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5392 lsearch lookups. Now it does.
5394 46. When parsing a route_list item in a manualroute router, a fixed-length
5395 buffer was used for the list of hosts. I made this 1024 bytes long,
5396 thinking that nobody would ever have a list of hosts that long. Wrong.
5397 Somebody had a whole pile of complicated expansion conditions, and the
5398 string was silently truncated, leading to an expansion error. It turns out
5399 that it is easier to change to an unlimited length (owing to other changes
5400 that have happened since this code was originally written) than to build
5401 structure for giving a limitation error. The length of the item that
5402 expands into the list of hosts is now unlimited.
5404 47. The lsearch lookup could not handle data where the length of text line was
5405 more than 4095 characters. Such lines were truncated, leading to shortened
5406 data being returned. It should now handle lines of any length.
5408 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5409 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5412 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5413 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5414 no longer happens in comment lines. A list of replacements is now placed
5415 at the head of all of the source files, except those whose only change is
5416 to replace PERL_COMMAND in the very first #! line.
5418 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5419 messages on the queue, with a bottom-up merge sort, using code contributed
5420 by Michael Haardt. This should make operations like -bp somewhat faster on
5421 large queues. It won't affect queue runners, except when queue_run_in_order
5424 51. Installed eximstats 1.31 in the distribution.
5426 52. Added support for SRV lookups to the dnslookup router.
5428 53. If an ACL referred to $message_body or $message_body_end, the value was not
5429 reset for any messages that followed in the same SMTP session.
5431 54. The store-handling optimization for building very long strings was not
5432 differentiating between the different store pools. I don't think this
5433 actually made any difference in practice, but I've tidied it.
5435 55. While running the routers to verify a sender address, $sender_address
5436 was still set to the sender address. This is wrong, because when routing to
5437 send a bounce to the sender, it would be empty. Therefore, I have changed
5438 it so that, while verifying a sender address, $sender_address is set to <>.
5439 (There is no change to what happens when verifying a recipient address.)
5441 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5442 target A or AAAA records (if not already returned) without resetting the
5443 qualify_single or search_parents options of the DNS resolver. These are
5444 inappropriate in this case because the targets of MX and SRV records must
5445 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5446 target that, when qualified, matched something in the local domain. These
5447 two options are now turned off when doing these lookups.
5449 57. It seems that at least some releases of Reiserfs (which does not have the
5450 concept of a fixed number of inodes) returns zero and not -1 for the
5451 number of available inodes. This interacted badly with check_spool_inodes,
5452 which assumed that -1 was the "no such thing" setting. What I have done is
5453 to check that the total number of inodes is greater than zero before doing
5454 the test of how many are available.
5456 58. When a "warn" ACL statement has a log_message modifier, the message is
5457 remembered, and not repeated. This is to avoid a lot of repetition when a
5458 message has many recipients that cause the same warning to be written.
5459 Howewer, Exim was preserving the list of already written lines for an
5460 entire SMTP session, which doesn't seem right. The memory is now reset if a
5461 new message is started.
5463 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5464 debugging output unless log_rewrite was also set.
5466 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5467 of (int)(handle) when we know that handle contains (void *)(-1).
5469 61. The Exim daemon panic-logs an error return when it closes the incoming
5470 connection. However "connection reset by peer" seems to be common, and
5471 isn't really an error worthy of noting specially, so that particular error
5474 62. When Exim is trying to find all the local interfaces, it used to panic and
5475 die if the ioctl to get the interface flags failed. However, it seems that
5476 on at least one OS (Solaris 9) it is possible to have an interface that is
5477 included in the list of interfaces, but for which you get a failure error
5478 for this call. This happens when the interface is not "plumbed" into a
5479 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5480 failure of the "get flags" call assumes that the interface is down.
5482 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5483 makes life easier for people who are doing arithmetic on fields extracted
5484 from dates, where you often get leading zeros that should not be
5485 interpreted as octal.
5487 64. Added qualify_domain to the redirect router, to override the global
5490 65. If a pathologically long header line contained very many addresses (the
5491 report of this problem mentioned 10 000) and each of them was rewritten,
5492 Exim could use up a very large amount of memory. (It kept on making new
5493 copies of the header line as it rewrote, and never released the old ones.)
5494 At the expense of a bit more processing, the header rewriting function has
5495 been changed so that it no longer eats memory in this way.
5497 66. The generation of the Received: header has been moved from the time that a
5498 message starts to be received, to the time that it finishes. The timestamp
5499 in the Received: header should now be very close to that of the <= log
5500 line. There are two side-effects of this change:
5502 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5503 logged header lines no longer include the local Received: line, because
5504 it has not yet been created. The same applies to a copy of the message
5505 that is returned to a non-SMTP sender when a message is rejected.
5507 (b) When a filter file is tested using -bf, no additional Received: header
5508 is added to the test message. After some thought, I decided that this
5511 This change does not affect the value of $received_for. It is still set
5512 after address rewriting, but before local_scan() is called.
5514 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5516 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5517 gave an unhelpful panic error message, and a defer error. I have managed to
5518 change this behaviour so that it now rejects any supplied certificate,
5519 which seems right, as the list of acceptable certificates is empty.
5521 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5522 gave an unhelpful defer error. I have not managed to make this reject any
5523 supplied certificates, but the error message it gives is "no certificate
5524 supplied", which is not helpful.
5526 70. exigrep's output now also includes lines that are not associated with any
5527 message, but which match the given pattern. Implemented by a patch from
5528 Martin Sluka, which also tidied up the Perl a bit.
5530 71. Recipient callout verification, like sender verification, was using <> in
5531 the MAIL FROM command. This isn't really the right thing, since the actual
5532 sender may affect whether the remote host accepts the recipient or not. I
5533 have changed it to use the actual sender in the callout; this means that
5534 the cache record is now keyed on a recipient/sender pair, not just the
5535 recipient address. There doesn't seem to be a real danger of callout loops,
5536 since a callout by the remote host to check the sender would use <>.
5537 [SEE ABOVE: changed after hitting problems.]
5539 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5540 temporary errors. However, in the case of such a code being given after
5541 the end of a data transmission (i.e. after ".") Exim was failing to write
5542 a retry record for the message. (Yes, there was some broken host that was
5543 actually sending 8xx at this point.)
5545 73. An unknown lookup type in a host list could cause Exim to panic-die when
5546 the list was checked. (An example that provoked this was putting <; in the
5547 middle of a list instead of at the start.) If this happened during a DATA
5548 ACL check, a -D file could be left lying around. This kind of configuration
5549 error no longer causes Exim to die; instead it causes a defer errror. The
5550 incident is still logged to the main and panic logs.
5552 74. Buglet left over from Exim 3 conversion. The message "too many messages
5553 in one connection" was written to the rejectlog but not the mainlog, except
5554 when address rewriting (yes!) was being logged.
5556 75. Added write_rejectlog option.
5558 76. When a system filter was run not as root (that is, when system_filter_user
5559 was set), the values of the $n variables were not being returned to the
5560 main process; thus, they were not subsequently available in the $sn
5563 77. Added +return_path_on_delivery log selector.
5565 78. A connection timeout was being treated differently from recipients deferred
5566 when testing hosts_max_try with a message that was older than the host's
5567 retry timeout. (The host should not be counted, thus allowing all hosts to
5568 be tried at least once before bouncing.) This may have been the cause of an
5569 occasionally reported bug whereby a message would remain on the queue
5570 longer than the retry timeout, but would be bounced if a delivery was
5571 forced. I say "may" because I never totally pinned down the problem;
5572 setting up timeout/retry tests is difficult. See also the next item.
5574 79. The ultimate address timeout was not being applied to errors that involved
5575 a combination of host plus message (for example, a timeout on a MAIL
5576 command). When an address resolved to a number of possible hosts, and they
5577 were not all tried for each delivery (e.g. because of hosts_max_try), a
5578 message could remain on the queue longer than the retry timeout.
5580 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5583 81. Fixed an obscure SMTP outgoing bug which required at least the following
5584 conditions: (a) there was another message waiting for the same server;
5585 (b) the server returned 5xx to all RCPT commands in the first message so
5586 that the message was not completed; (c) the server dropped the connection
5587 or gave a negative response to the RSET that Exim sends to abort the
5588 transaction. The observed case was a dropped connection after DATA that had
5589 been sent in pipelining mode. That is, the server had advertised PIPELINING
5590 but was not implementing it correctly. The effect of the bug was incorrect
5591 behaviour, such as trying another host, and this could lead to a crash.
5597 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5598 and daemon.c were passed as pointers to ints; they should have been
5599 pointers to socklen_t variables (which are typically unsigned ints).
5601 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5604 3. Fixed a really odd bug that affected only the testing scheme; patching a
5605 certain fixed string in the binary changed the value of another string that
5606 happened to be identical to the end of the original first string.
5608 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5609 name", it returns that address as the IP address. On some operating
5610 systems (e.g. Solaris), it also passes back the IP address string as the
5611 "host name". However, on others (e.g. Linux), it passes back an empty
5612 string. Exim wasn't checking for this, and was changing the host name to an
5613 empty string, assuming it had been canonicized.
5615 5. Although rare, it is permitted to have more than one PTR record for a given
5616 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5617 all the names associated with an address, because they do in Solaris.
5618 However, it seems that they do not in Linux for data that comes from the
5619 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5620 I found this out when I moved to a new Linux workstation and tried to run
5621 the Exim test suite.
5623 To get round this problem I have changed the code so that it now does its
5624 own call to the DNS to look up PTR records when searching for a host name.
5625 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5626 addresses that are only in /etc/hosts are still found.
5628 This behaviour is, however, controlled by an option called host_lookup_
5629 order, which defaults to "bydns:byaddr". If people want to use the other
5630 order, or indeed, just use one or the other means of lookup, they can
5631 specify it in this variable.
5633 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5634 some operating systems, this comes back from gethostbyaddr() as an empty
5635 string, and this is what Exim used to test for. However, it seems that in
5636 other systems, "." is yielded. Exim now tests for this case too.
5638 7. The values of check_spool_space and check_log_space are now held internally
5639 as a number of kilobytes instead of an absolute number of bytes. If a
5640 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5641 kilobyte. This means that much larger values can be stored.
5643 8. Exim monitor: an attempt to get the action menu when not actually pointing
5644 at a message produces an empty menu entitled "No message selected". This
5645 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5646 no entries in it ("Shell widget menu has zero width and/or height"). So I
5647 have added a single, blank menu entry in this case.
5649 9. Added ${quote_local_part.
5651 10. MIME decoding is now applied to the contents of Subject: header lines when
5654 11. Now that a reference to $sender_host_address automatically causes a reverse
5655 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5656 host lookup before query-style lookups in lists that might use this
5657 variable. This has therefore been abolished, and the "net-" prefix is no
5658 longer necessary for query-style lookups.
5660 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5661 have been a typo for LFLAGS, so it has been changed.
5663 13. The install script calls Exim with "-C /dev/null" in order to find the
5664 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5665 to be output. Howeve, since Exim outputs its version number before the
5666 error, it didn't break the script. It just looked ugly. I fixed this by
5667 always allowing "-C /dev/null" if the caller is root.
5669 14. Ignore overlarge ACL variable number when reading spool file - insurance
5670 against a later release with more variables having written the file.
5672 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5673 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5674 this, as well as the form without the "IPv6" on the front.
5676 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5677 OS/Makefile-Darwin file.
5679 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5680 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5682 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5683 variable. However, some OS use ints for inodes. I've added cast to long int
5684 to get rid of the compiler warning.
5686 19. I had forgotten to lock out "/../" in configuration file names when
5687 ALT_CONFIG_PREFIX was set.
5689 20. Routers used for verification do not need to specify transports. However,
5690 if such a router generated a host list, and callout was configured, Exim
5691 crashed, because it could not find a port number from the (non-existent)
5692 transport. It now assumes port 25 in this circumstance.
5694 21. Added the -t option to exigrep.
5696 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5697 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5698 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5699 them is set, LOOKUP_LSEARCH is forced.
5701 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5703 24. Added sender and host information to the "rejected by local_scan()" log
5704 line; previously there was no indication of these.
5706 25. Added .include_if_exists.
5708 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5709 Linux. It turns out that not all file systems support this. Apparently some
5710 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5711 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5712 not supported on the file descriptor, is now ignored when Exim is trying to
5713 sync a directory. This applies only to Linux.
5715 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5717 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5718 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5721 29. Added support for maildirsize files from supplied patch (modified a bit).
5723 30. The use of :fail: followed by an empty string could lead Exim to respond to
5724 sender verification failures with (e.g.):
5726 550 Verification failed for <xxx>
5727 550 Sender verify failed
5729 where the first response line was missing the '-' that indicates it is not
5730 the final line of the response.
5732 31. The loop for finding the name of the user that called Exim had a hardwired
5733 limit of 10; it now uses the value of finduser_retries, which is used for
5734 all other user lookups.
5736 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5739 33. Exim was neglecting to zero errno before one call of strtol() when
5740 expanding a string and expecting an integer value. On some systems this
5741 resulted in spurious "integer overflow" errors. Also, it was casting the
5742 result into an int without checking.
5744 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5745 did not work. The code looks as if it has *never* worked, though it appears
5746 to have been documented since at least releast 1.62. I have made it work.
5748 35. The "timeout_DNS" error in retry rules, also documented since at least
5749 1.62, also never worked. As it isn't clear exactly what this means, and
5750 clearly it isn't a major issue, I have abolished the feature by treating it
5751 as "timeout", and writing a warning to the main and panic logs.
5753 36. The display of retry rules for -brt wasn't always showing the error code
5756 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5757 timeout_connect_A, timeout_connect_MX.
5759 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5760 to the empty sender.
5762 39. The daemon was not analysing the content of -oX till after it had closed
5763 stderr and disconnected from the controlling terminal. This meant that any
5764 syntax errors were only noted on the panic log, and the return code from
5765 the command was 0. By re-arranging the code a little, I've made the
5766 decoding happen first, so such errors now appear on stderr, and the return
5767 code is 1. However, the actual setting up of the sockets still happens in
5768 the disconnected process, so errors there are still only recorded on the
5771 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5772 connections (as happens on some IP stacks) was logged at start up time as
5773 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5774 it from "IPv6 and IPv4", which means that two separate sockets are being
5777 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5778 says whether AF_INET or AF_INET6 was passed as an argument.
5780 42. Exiwhat output was messed up when time zones were included in log
5783 43. Exiwhat now gives more information about the daemon's listening ports,
5784 and whether -tls-on-connect was used.
5786 44. The "port" option of the smtp transport is now expanded.
5788 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5789 silently ignored. Now an error message is written to the main and panic
5792 46. There's a new ACL modifier called "logwrite" which writes to a log file
5793 as soon as it is encountered.
5795 47. Added $local_user_uid and $local_user_gid at routing time.
5797 48. Exim crashed when trying to verify a sender address that was being
5800 49. Exim was recognizing only a space character after ".include". It now also
5801 recognizes a tab character.
5803 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5804 extracting the relevant information from the specification. The man page no
5805 longer contains scrambled data for the -d option, and I've added a section
5806 at the front about calling Exim under different names.
5808 51. Added "extra_headers" argument to the "mail" command in filter files.
5810 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5813 53. Installed eximstats 1.29.
5815 54. Added transport_filter_timeout as a generic transport option.
5817 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5818 Cc: header lines. This was required by RFC 822, but it not required by RFC
5821 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5822 incoming messages that did not have them. Now it does so only if the
5823 message originates locally, that is, if there is no associated remote host
5824 address. When Resent- header lines are present, this applies to the Resent-
5825 lines rather than the non-Resent- lines.
5827 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5828 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5830 58. Messages for configuration errors now include the name of the main
5831 configuration file - useful now that there may be more than one file in a
5832 list (.included file names were always shown).
5834 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5835 for those rare installations that do not start the daemon as root or run it
5836 setuid root. I've cut out the call to initgroups() if the daemon is not
5839 60. The Exim user and group can now be bound into the binary as text strings
5840 that are looked up at the start of Exim's processing.
5842 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5844 62. Added $mailstore_basename variable.
5846 63. Installed patch to sieve.c from Michael Haardt.
5848 64. When Exim failed to open the panic log after failing to open the main log,
5849 the original message it was trying to log was written to stderr and debug
5850 output, but if they were not available (the usual case in production), it
5851 was lost. Now it is written to syslog before the two lines that record the
5852 failures to open the logs.
5854 65. Users' Exim filters run in subprocesses under the user's uid. It is
5855 possible for a "deliver" command or an alias in a "personal" command to
5856 provoke an address rewrite. If logging of address rewriting is configured,
5857 this fails because the process is not running as root or exim. There may be
5858 a better way of dealing with this, but for the moment (because 4.30 needs
5859 to be released), I have disabled address rewrite logging when running a
5860 filter in a non-root, non-exim process.
5866 1. The buildconfig auxiliary program wasn't quoting the value set for
5867 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5868 not defined. This bug was masked in 4.22 by the effect that was fixed in
5871 2. Some messages that were rejected after a message id was allocated were
5872 shown as "incomplete" by exigrep. It no longer does this for messages that
5873 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5875 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5876 have allow_domain_literals set, the ID did not get logged in the <= line.
5877 Domain literals are now always recognized in Message-ID: header lines.
5879 4. The first argument for a ${extract expansion item is the key name or field
5880 number. Leading and trailing spaces in this item were not being ignored,
5881 causing some misleading effects.
5883 5. When deliver_drop_privilege was set, single queue runner processes started
5884 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5885 same command in the process it spins off) were not dropping privilege.
5887 6. When the daemon running as "exim" started a queue runner, it always
5888 re-executed Exim in the spun-off process. This is a waste of effort when
5889 deliver_drop_privilege is set. The new process now just calls the
5890 queue-runner function directly.
5896 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5899 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5900 ignored. Though the use of -oP was forcing the writing of a pid file, it
5901 was always written to the default place.
5903 3. If the message "no IP address found for host xxxx" is generated during
5904 incoming verification, it is now followed by identification of the incoming
5905 connection (so you can more easily find what provoked it).
5907 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5909 5. Added some features to "harden" Exim a bit more against certain attacks:
5911 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5912 be put in Local/Makefile. This is like the never_users runtime option,
5913 but it cannot be overridden. The default setting is "root".
5915 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5916 prefix string with which any file named in a -C command line option
5919 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5920 is retained for -C and -D only if the caller of Exim is root. Without
5921 it, the exim user may also use -C and -D and retain privilege.
5923 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5924 command line option is disabled.
5926 6. Macro names set by the -D option must start with an upper case letter, just
5927 like macro names defined in the configuration file.
5929 7. Added "dereference=" facility to LDAP.
5931 8. Two instances of the typo "uknown" in the source files are fixed.
5933 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5934 the Configure-Makefile script screwed up while processing it.
5936 10. Incorporated PCRE 4.4.
5938 11. The SMTP synchronization check was not operating right at the start of an
5939 SMTP session. For example, it could not catch a HELO sent before the client
5940 waited for the greeting. There is now a check for outstanding input at the
5941 point when the greeting is written. Because of the duplex, asynchronous
5942 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5943 way, but not yet received, when the check is performed.
5945 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5946 on TCP/IP sockets, because this apparently causes some broken clients to
5949 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5950 unchanged) from the Cygwin maintainer.
5952 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5953 was supported instead of testing for mbx. Effectively a typo.
5955 15. The spa authenticator server code was not checking that the input it
5956 received was valid base64.
5958 16. The debug output line for the "set" modifier in ACLs was not showing the
5959 name of the variable that was being set.
5961 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5963 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5964 lookup on its own. Something else was needed to trigger the lookup. For
5965 example, a match in host_lookup or the need for a host name in a host list.
5966 Now, if $sender_host_name is referenced and the host name has not yet been
5967 looked up, a lookup is performed. If the lookup fails, the variable remains
5968 empty, and $host_lookup_failed is set to "1".
5970 19. Added "eqi" as a case-independent comparison operator.
5972 20. The saslauthd authentication condition could segfault if neither service
5973 nor realm was specified.
5975 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5976 error message that was logged was misleading, and incoming SMTP
5977 connections were dropped. The message is now more accurate, and temporary
5978 errors are given to SMTP connections.
5980 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5981 (except RSET) with a 421 error, until QUIT is received. However, it was
5982 failing to send a response to QUIT.
5984 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5985 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5986 ACL is now run after the helo_try_verify_hosts code.
5988 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5989 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5990 case-independent, so other case variants are also recognized). Apparently
5991 some systems use these upper case variants.
5993 25. If more than two messages were waiting for the same host, and a transport
5994 filter was specified for the transport, Exim sent two messages over the
5995 same TCP/IP connection, and then failed with "socket operation on non-
5996 socket" when it tried to send the third.
5998 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6000 27. The extern definition of crypt16() in expand.c was not being excluded when
6001 the OS had its own crypt16() function.
6003 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6004 as a preferred synonym for return_size_limit, both as an option and as an
6007 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6009 30. Changed the default configuration ACL to relax the local part checking rule
6010 for addresses that are not in any local domains. For these addresses,
6011 slashes and pipe symbols are allowed within local parts, but the sequence
6012 /../ is explicitly forbidden.
6014 31. SPA server authentication was not clearing the challenge buffer before
6017 32. log_message in a "warn" ACL statement was writing to the reject log as
6018 well as to the main log, which contradicts the documentation and doesn't
6019 seem right (because no rejection is happening). So I have stopped it.
6021 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6022 However, I am unable to do any testing of this.
6024 34. Fixed an infelicity in the appendfile transport. When checking directories
6025 for a mailbox, to see if any needed to be created, it was accidentally
6026 using path names with one or more superfluous leading slashes; tracing
6027 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6029 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6030 contents are added to the log line that is written for every discarded
6031 recipient. (Previously a log_message setting was ignored.)
6033 36. The ${quote: operator now quotes the string if it is empty.
6035 37. The install script runs exim in order to find its version number. If for
6036 some reason other than non-existence or emptiness, which it checks, it
6037 could not run './exim', it was installing it with an empty version number,
6038 i.e. as "exim-". This error state is now caught, and the installation is
6041 38. An argument was missing from the function that creates an error message
6042 when Exim fails to connect to the socket for saslauthd authentication.
6043 This could cause Exim to crash, or give a corrupted message.
6045 39. Added isip, isip4, and isip6 to ${if conditions.
6047 40. The ACL variables $acl_xx are now saved with the message, and can be
6048 accessed later in routers, transports, and filters.
6050 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6051 strings in the file are not string-expanded.
6053 42. If a MAIL command specified a SIZE value that was too large to fit into an
6054 int variable, the check against message_size_limit failed. Such values are
6055 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6056 day this will have to be increased, but I don't think I want to be around
6057 when emails are that large.
6064 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6065 iconv() is not standard in FreeBSD.
6067 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6068 IPv6 enabled. The observed symptom was a segmentation fault on return from
6069 the function os_common_find_running_interfaces() in src/os.c.
6071 3. In the check_special_case() function in daemon.c I had used "errno" as an
6072 argument name, which causes warnings on some systems. This was basically a
6073 typo, since it was named "eno" in the comments!
6075 4. The code that waits for the clock to tick (at a resolution of some fraction
6076 of a second) so as to ensure message-id uniqueness was always waiting for
6077 at least one whole tick, when it could have waited for less. [This is
6078 almost certainly not relevant at current processor speeds, where it is
6079 unlikely to ever wait at all. But we try to future-proof.]
6081 5. The function that sleeps for a time interval that includes fractions of a
6082 second contained a race. It did not block SIGALRM between setting the
6083 timer, and suspending (a couple of lines later). If the interval was short
6084 and the sigsuspend() was delayed until after it had expired, the suspension
6085 never ended. On busy systems this could lead to processes getting stuck for
6088 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6089 process, before it forks any delivery processes. The open lookup caching
6090 mechanism meant that the open file or database connection was passed into
6091 the delivery process. The problem was that delivery processes always tidy
6092 up cached lookup data. This could cause a problem for the next delivery
6093 process started by the queue runner, because the external queue runner
6094 process does not know about the closure. So the next delivery process
6095 still has data in the lookup cache. In the case of a file lookup, there was
6096 no problem because closing a file descriptor in a subprocess doesn't affect
6097 the parent. However, if the lookup was caching a connection to a database,
6098 the connection was closed, and the second delivery process was likely to
6099 see errors such as "PGSQL: query failed: server closed the connection
6100 unexpectedly". The problem has been fixed by closing all cached lookups
6101 in a queue runner before running a delivery process.
6103 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6104 seem to have the "const" qualifier which it has on other OS. I've
6107 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6108 *of the same type* (client or server) with the same public name that an
6109 error should be diagnosed.
6111 9. When Exim looked up a host name for an IP address, but failed to find the
6112 original IP address when looking up the host name (a safety check), it
6113 output the message "<ip address> does not match any IP for NULL", which was
6114 confusing, to say the least. The bug was that the host name should have
6115 appeared instead of "NULL".
6117 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6118 user that is built into the binary, and the -C or -D options is used, root
6119 privilege is dropped before the configuration file is read. In addition,
6120 logging is switched to stderr instead of the normal log files. If the
6121 configuration then re-defines the Exim user, the unprivileged environment
6122 is probably not what is expected, so Exim logs a panic warning message (but
6125 However, if deliver_drop_privilege is set, the unprivileged state may well
6126 be exactly what is intended, so the warning has been cut out in that case,
6127 and Exim is allowed to try to write to its normal log files.
6133 1. smtp_return_error_details was not giving details for temporary sender
6134 or receiver verification errors.
6136 2. Diagnose a configuration error if two authenticators have the same public
6139 3. Exim used not to create the message log file for a message until the first
6140 delivery attempt. This could be confusing when incoming messages were held
6141 for policy or load reasons. The message log file is now created at the time
6142 the message is received, and an initial "Received" line is written to it.
6144 4. The automatically generated man page for command line options had a minor
6145 bug that caused no ill effects; however, a more serious problem was that
6146 the procedure for building the man page automatically didn't always
6147 operate. Consequently, release 4.20 contains an out-of-date version. This
6148 shouldn't happen again.
6150 5. When building Exim with embedded Perl support, the script that builds the
6151 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6152 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6154 6. The freeze_tell option was not being used for messages that were frozen on
6155 arrival, either by an ACL or by local_scan().
6157 7. Added the smtp_incomplete_transaction log selector.
6159 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6160 was accepting AUTH without a new EHLO.
6162 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6163 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6166 10. Logging of TCP/IP connections (when configured) now happens in the main
6167 daemon process instead of the child process, so that the TCP/IP connection
6168 count is more accurate (but it can never be perfect).
6170 11. The use of "drop" in a nested ACL was not being handled correctly in the
6171 outer ACL. Now, if condition failure induced by the nested "drop" causes
6172 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6173 or "require"), the connection is dropped.
6175 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6176 that yield "discard" can now be used with an "accept" or a "discard" verb,
6177 but an error is generated for any others (because I can't see a useful way
6178 to define what should happen).
6180 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6181 are now processed in the same way as lines in the Exim configuration file.
6182 In particular, continuation lines are supported.
6184 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6186 15. Added -ti meaning -t -i.
6188 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6189 domains, and warn by logging if others are found.
6191 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6192 returned from GIFCONF when Exim is trying to find the list of interfaces on
6193 a host. The code in os.c has been modified to copy each ifreq to an aligned
6194 structure in all cases.
6196 Also, in some cases, the returned ifreq's were being copied to a 'struct
6197 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6198 means the last couple of bytes of an IPv6 address could be chopped if the
6199 ifreq contained only a normal sockaddr (14 bytes storage).
6201 18. Named domain lists were not supported in the hosts_treat_as_local option.
6202 An entry such as +xxxx was not recognized, and was treated as a literal
6205 19. Ensure that header lines added by a DATA ACL are included in the reject log
6206 if the ACL subsequently rejects the message.
6208 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6209 MD5 (which is deprecated).
6211 21. When testing a filter file using -bf, Exim was writing a message when it
6212 took the sender from a "From " line in the message, but it was not doing so
6213 when it took $return_path from a Return-Path: header line. It now does.
6215 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6216 with a valid header line field name (a series of printing characters
6217 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6219 23. Changed "disc" in the source to "disk" to conform to the documentation and
6220 the book and for uniformity.
6222 24. Ignore Sendmail's -Ooption=value command line item.
6224 25. When execve() failed while trying to run a command in a pipe transport,
6225 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6226 could be confused with a return value of 69 from the command itself. This
6227 has been changed to 127, the value the shell returns if it is asked to run
6228 a non-existent command. The wording for the related log line suggests a
6229 non-existent command as the problem.
6231 26. If received_header_text expands to an empty string, do not add a Received:
6232 header line to the message. (Well, it adds a token one on the spool, but
6233 marks it "old" so that it doesn't get used or transmitted.)
6235 27. Installed eximstats 1.28 (addition of -nt option).
6237 28. There was no check for failure on the call to getsockname() in the daemon
6238 code. This can fail if there is a shortage of resources on the system, with
6239 ENOMEM, for example. A temporary error is now given on failure.
6241 29. Contrary to the C standard, it seems that in some environments, the
6242 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6243 program. Exim now does this explicitly; it affects the formatting of
6244 timestamps using strftime().
6246 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6247 complaints. I've now initialized all the variables, to avoid this.
6249 32. Header lines added by a system filter were not being "seen" during
6250 transport-time rewrites.
6252 33. The info_callback() function passed to OpenSSL is set up with type void
6253 (*)(SSL *, int, int), as described somewhere. However, when calling the
6254 function (actually a macro) that sets it up, the type void(*)() is
6255 expected. I've put in a cast to prevent warnings from picky compilers.
6257 34. If a DNS black list lookup found a CNAME record, but there were no A
6258 records associated with the domain it pointed at, Exim crashed.
6260 35. If a DNS black list lookup returned more than one A record, Exim ignored
6261 all but the first. It now scans all returned addresses if a particular IP
6262 value is being sought. In this situation, the contents of the
6263 $dnslist_value variable are a list of all the addresses, separated by a
6266 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6267 to accept a host name and all its aliases if the forward lookup for any of
6268 them yielded the IP address of the incoming connection. Now it accepts only
6269 those names whose forward lookup yields the correct IP address. Any other
6270 names are discarded. This closes a loophole whereby a rogue DNS
6271 administrator could create reverse DNS records to break through a
6272 wildcarded host restriction in an ACL.
6274 37. If a user filter or a system filter that ran in a subprocess used any of
6275 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6276 the wrong values were passed to the pipe command ($thisaddress had the
6277 value of $0, $0 had the value of $1, etc). This bug was introduced by
6278 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6280 38. Improved the line breaking for long SMTP error messages from ACLs.
6281 Previously, if there was no break point between 40 and 75 characters, Exim
6282 left the rest of the message alone. Two changes have been made: (a) I've
6283 reduced the minimum length to 35 characters; (b) if it can't find a break
6284 point between 35 and 75 characters, it looks ahead and uses the first one
6285 that it finds. This may give the occasional overlong line, but at least the
6286 remaining text gets split now.
6288 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6289 file descriptors might be low, and that setting 1000 would always raise it.
6290 It turns out that in some environments, the limit is already over 1000 and
6291 that lowering it causes trouble. So now Exim takes care not to decrease it.
6293 40. When delivering a message, the value of $return_path is set to $sender_
6294 address at the start of routing (routers may change the value). By an
6295 oversight, this default was not being set up when an address was tested by
6296 -bt or -bv, which affected the outcome if any router or filter referred to
6299 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6300 log only when "message" or "log_message" are set. However, if one of the
6301 conditions was an address verification, or a call to a nested ACL, the
6302 messages generated by the underlying test were being passed through. This
6303 no longer happens. The underlying message is available in $acl_verify_
6304 message for both "message" and "log_message" expansions, so it can be
6305 passed through if needed.
6307 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6308 new expansion $bh_ to give the encoded byte string without charset
6309 translation. Translation happens only if iconv() is available; HAVE_ICONV
6310 indicates this at build time. HEADERS_CHARSET gives the charset to
6311 translate to; headers_charset can change it in the configuration, and
6312 "headers charset" can change it in an individual filter file.
6314 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6315 that creates RFC 2047 encoded "words" labels them as that charset instead
6316 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6317 expansion operator; (ii) when Exim creates a From: line for a local
6318 message; (iii) when a header line is rewritten to include a "phrase" part.
6320 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6321 buggy, causing it to skip the first lines of messages whose message ID
6322 ended in 'D'. This would not have bitten before Exim release 4.14, because
6323 message IDs were unlikely to end in 'D' before then. The effect was to have
6324 incorrect size information for certain domains.
6326 45. #include "config.h" was missing at the start of the crypt16.c module. This
6327 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6330 46. If there was a timeout during a "random" callout check, Exim treated it as
6331 a failure of the random address, and carried on sending RSET and the real
6332 address. If the delay was just some slowness somewhere, the response to the
6333 original RCPT would be taken as a response to RSET and so on, causing
6334 mayhem of various kinds.
6336 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6337 when I implemented it. It didn't allow for the fact that some option values
6338 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6339 the right test for positive values.
6341 48. Domain names in DNS records are case-independent. Exim always looks them up
6342 in lower case. Some resolvers return domain names in exactly the case they
6343 appear in the zone file, that is, they may contain uppercase letters. Not
6344 all resolvers do this - some return always lower case. Exim was treating a
6345 change of case by a resolver as a change of domain, similar to a widening
6346 of a domain abbreviation. This triggered its re-routing code and so it was
6347 trying to route what was effectively the same domain again. This normally
6348 caused routing to fail (because the router wouldn't handle the domain
6349 twice). Now Exim checks for this case specially, and just changes the
6350 casing of the domain that it ultimately uses when it transmits the message
6353 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6356 50. If a filter generated a file delivery with a non-absolute name (possible if
6357 no home directory exists for the router), the forbid_file option was not
6360 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6361 the existing equality matching.
6363 52. Exim was using ints instead of ino_t variables in some places where it was
6364 dealing with inode numbers.
6366 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6367 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6368 called TMPDIR, and if it finds it is different, it changes its value.
6370 54. The smtp_printf() function is now made available to local_scan() so
6371 additional output lines can be written before returning. There is also an
6372 smtp_fflush() function to enable the detection of a dropped connection.
6373 The variables smtp_input and smtp_batched_input are exported to
6376 55. Changed the default runtime configuration: the message "Unknown user"
6377 has been removed from the ACL, and instead placed on the localuser router,
6378 using the cannot_route_message feature. This means that any verification
6379 failures that generate their own messages won't get overridden. Similarly,
6380 the "Unrouteable address" message that was in the ACL for unverifiable
6381 relay addresses has also been removed.
6383 56. Added hosts_avoid_esmtp to the smtp transport.
6385 57. The exicyclog script was not checking for the esoteric option
6386 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6387 will work only if exicyclog is run under the appropriate euid.
6389 58. Following a discussion on the list, the rules by which Exim recognises line
6390 endings on incoming messages have been changed. The -dropcr and drop_cr
6391 options are now no-ops, retained only for backwards compatibility. The
6392 following line terminators are recognized: LF CRLF CR. However, special
6393 processing applies to CR:
6395 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6396 nor a local message in the state where . is a terminator.
6398 (ii) If a bare CR is encountered in a header line, an extra space is added
6399 after the line terminator so as not to end the header. The reasoning
6400 behind this is that bare CRs in header lines are most likely either
6401 to be mistakes, or people trying to play silly games.
6403 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6404 was being incorrectly given as 18 bytes larger than it should have been.
6405 This is a VOB (very old bug).
6407 60. This may never have affected anything current, but just in case it has:
6408 When the local host is found other than at the start of a list of hosts,
6409 the local host, those with the same MX, and any that follow, are discarded.
6410 When the list in question was part of a longer list of hosts, the following
6411 hosts (not currently being processed) were also being discarded. This no
6412 longer happens. I'm not sure if this situation could ever has previously
6415 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6418 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6419 header. This is something that is recommended in a new Internet Draft, and
6420 is something that is documented as being done by Sendmail. There are two
6421 possible values. For messages generated by the autoreply transport, Exim
6424 Auto-Submitted: auto-replied
6426 whereas for all other generated messages (e.g. bounces) it adds
6428 Auto-Submitted: auto-generated
6430 63. The "personal" condition in filters now includes a test for the
6431 Auto-Submitted: header. If it contains the string "auto-" the message it
6432 not considered personal.
6434 64. Added rcpt_include_affixes as a generic transport option.
6436 65. Added queue_only_override (default true).
6438 66. Added the syslog_duplication option.
6440 67. If what should have been the first header line of a message consisted of
6441 a space followed by a colon, Exim was mis-interpreting it as a header line.
6442 It isn't of course - it is syntactically invalid and should therefore be
6443 treated as the start of the message body. The misbehaviour could have
6444 caused a number of strange effects, including loss of data in subsequent
6445 header lines, and spool format errors.
6447 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6448 client host had authenticated. This control can now be exercised by an ACL
6449 for more flexibility.
6451 69. By default, callouts do not happen when testing with -bh. There is now a
6452 variant, -bhc, which does actually run the callout code, including
6453 consulting and updating the callout cache.
6455 70. Added support for saslauthd authentication, courtesy of Alexander
6458 71. If statvfs() failed on the spool or log directories while checking their
6459 size for availability, Exim confusingly gave the error "space shortage".
6460 Furthermore, in debugging mode it crashed with a floating point exception.
6461 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6462 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6463 serious problem, Exim now writes to the main and panic logs when this
6464 happens, with details of the failure. It then refuses to accept the
6465 incoming message, giving the message "spool directory problem" or "log
6466 directory problem" with a 421 code for SMTP messages.
6468 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6469 0, 1, and 2 exist, because some OS complain for execs without them (see
6470 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6471 descriptors. However, the code omitted to check that the open succeeded,
6472 causing mysterious errors if for some reason the permissions on /dev/null
6473 got screwed. Now Exim writes a message to the main and panic logs, and
6474 bombs out if it can't open /dev/null.
6476 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6477 interact so that it is all more flexible. It is supposed to remain
6478 backwards compatible. Also added extra_local_interfaces.
6480 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6481 to bomb out with an assertion failure - to the client this appears as a
6482 connection drop. This problem occurs in the part of the code that was taken
6483 from the Samba project. Fortunately, the assertion is in a very simple
6484 function, so I have fixed this by reproducing the function inline in the
6485 one place where it is called, and arranging for authentication to fail
6486 instead of killing the process with assert().
6488 75. The SPA client code was not working when the server requested OEM rather
6489 than Unicode encoding.
6491 76. Added code to make require_files with a specific uid setting more usable in
6492 the case where statting the file as root fails - usually a non-root-mounted
6493 NFS file system. When this happens and the failure is EACCES, Exim now
6494 forks a subprocess and does the per-uid checking as the relevant uid.
6496 77. Added process_log_path.
6498 78. If log_file_path was not explicitly set, a setting of check_log_space or
6499 check_log_inodes was ignored.
6501 79. If a space check for the spool or log partitions fails, the incident is now
6502 logged. Of course, in the latter case the data may get lost...
6504 80. Added the %p formatting code to string_format() so that it can be used to
6505 print addresses in debug_print(). Adjusted all the address printing in the
6506 debugging in store.c to use %p rather than %d.
6508 81. There was a concern that a line of code in smtp_in.c could overflow a
6509 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6510 initially expressed, the concern was not well-founded, because trailing
6511 spaces are removed early. However, if the trailing spaces were followed by
6512 a NULL, they did not get removed, so the overflow was possible. Two fixes
6515 (a) I re-wrote the offending code in a cleaner fashion.
6516 (b) If an incoming SMTP command contains a NULL character, it is rejected
6519 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6520 runs initgroups(), so that if the Exim user is in any additional groups,
6521 they will be used during message reception.
6527 The change log for 4.20 and earlier releases has been archived.