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.
16 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
17 and 3.4.4 - once the server is enabled to respond to an OCSP request
18 it does even when not requested, resulting in a stapling non-aware
19 client dropping the TLS connection.
21 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
22 support variable-length bit vectors. No functional change.
24 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
25 The I= interface field on outgoing lines is now after the H= remote
26 host field, same as incoming lines. There is a separate
27 outgoing_interface log selector which allows you to disable the
30 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
31 If not running log_selector +smtp_connection the mainlog would be held
32 open indefinitely after a "too many connections" event, including to a
33 deleted file after a log rotate. Leave the per net connection logging
34 leaving it open for efficiency as that will be quickly detected by the
35 check on the next write.
37 HS/01 Bug 1671: Fix post transport crash.
38 Processing the wait-<transport> messages could crash the delivery
39 process if the message IDs didn't exist for some reason. When
40 using 'split_spool_directory=yes' the construction of the spool
41 file name failed already, exposing the same netto behaviour.
43 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
44 mime_regex ACL conditions.
46 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
47 to DSN fail messages (bounces): remote IP, remote greeting, remote response
48 to HELO, local diagnostic string.
50 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
51 log line to debug. Even when configured with a tls authenticator many
52 client connections are expected to not authenticate in this way, so
53 an authenticate fail is not an error.
55 HS/02 Add the Exim version string to the process info. This way exiwhat
56 gives some more detail about the running daemon.
58 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
59 matter for fast-change records such as DNSBLs.
61 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
62 retry record, even if constant. There may be multiple transports with
63 different interface settings and the retry behaviour needs to be kept
66 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
68 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
70 JH/10 Bug 840: fix log_defer_output option of pipe transport
72 JH/11 Bug 830: use same host for all RCPTS of a message, even under
73 hosts_randomize. This matters a lot when combined with mua_wrapper.
75 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
76 ${quote_pgsql:<string>} operator.
78 JH/13 Bug 1708: avoid misaligned access in cached lookup.
80 JH/14 Change header file name for freeradius-client. Relevant if compiling
81 with Radius support; from the Gentoo tree and checked under Fedora.
83 JH/15 Bug 1712: Introduce $prdr_requested flag variable
85 JH/16 Bug 1714: Permit an empty string as expansion result for transport
86 option transport_filter, meaning no filtering.
88 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
90 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
91 defaults to "*" (all hosts). The variable is now available when not built
92 with TLS, default unset, mainly to enable keeping the testuite sane.
93 If a server certificate is not supplied (via tls_certificate) an error is
94 logged, and clients will find TLS connections fail on startup. Presumably
95 they will retry in-clear.
96 Packagers of Exim are strongly encouraged to create a server certificate
99 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
100 with the $config_file variable.
102 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
103 in transport context, after the attempt, and per-recipient. The latter type
104 is per host attempted. The event data is the error message, and the errno
105 information encodes the lookup type (A vs. MX) used for the (first) host,
106 and the trailing two digits of the smtp 4xx reponse.
108 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
109 to write to mainlog (or rejectlog, paniclog) in the window between file
110 creation and permissions/ownership being changed. Particularly affects
111 installations where exicyclog is run as root, rather than exim user;
112 result is that the running daemon panics and dies.
114 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
116 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
117 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
118 "pri" and "weight". Note that the previous implicit priority given by the
119 list order is no longer honoured.
121 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
124 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
125 by defining SUPPORT_SOCKS.
127 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
128 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
129 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
130 variables are renamed to proxy_{local,external}_{address,port}.
132 JH/27 Move Internationalisation support from Experimental to mainline, enabled
133 for a build by defining SUPPORT_I18N
135 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
136 of the query string, and make ${quote_redis:} do that quoting.
138 JH/29 Move Events support from Experimental to mainline, enabled by default
139 and removable for a build by defining DISABLE_EVENT.
141 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
143 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
144 cached by the daemon.
146 JH/32 Move Redis support from Experimental to mainline, enabled for a build
147 by defining LOOKUP_REDIS. The libhiredis library is required.
149 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
150 keys are given for lookup.
152 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
153 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
154 only supported when built with TLS support. The PolarSSL SHA routines
155 are still used when the TLS library is too old for convenient support.
157 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
158 openssl_options), for security. OpenSSL forces this from version 1.1.0
159 server-side so match that on older versions.
161 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
162 allocation for $value could be released as the expansion processing
163 concluded, but leaving the global pointer active for it.
165 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
166 and to use the domains and local_parts ACL conditions.
168 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
169 incorrectly not doubled on cutthrough transmission, hence seen as a
170 body-termination at the receiving system - resulting in truncated mails.
171 Commonly the sender saw a TCP-level error, and retransmitted the message
172 via the normal store-and-forward channel. This could result in duplicates
173 received - but deduplicating mailstores were liable to retain only the
174 initial truncated version.
176 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
178 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
180 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
181 we're in there, support oversigning also; bug 1309.
183 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
185 HS/04 Add support for keep_environment and add_environment options.
187 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
188 either intentional arithmetic overflow during PRNG, or testing config-
191 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
192 delivery resulted in actual delivery. Cancel cutthrough before DATA
195 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
196 rejects a recipient: pass the reject to the originator.
198 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
199 Many were false-positives and ignorable, but it's worth fixing the
202 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
203 for the new environment-manipulation done at startup. Move the routines
204 from being local to tls.c to being global via the os.c file.
206 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
207 an extract embedded as result-arg for a map, the first arg for extract
208 is unavailable so we cannot tell if this is a numbered or keyed
209 extraction. Accept either.
215 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
218 JH/02 The smtp transport option "multi_domain" is now expanded.
220 JH/03 The smtp transport now requests PRDR by default, if the server offers
223 JH/04 Certificate name checking on server certificates, when exim is a client,
224 is now done by default. The transport option tls_verify_cert_hostnames
225 can be used to disable this per-host. The build option
226 EXPERIMENTAL_CERTNAMES is withdrawn.
228 JH/05 The value of the tls_verify_certificates smtp transport and main options
229 default to the word "system" to access the system default CA bundle.
230 For GnuTLS, only version 3.0.20 or later.
232 JH/06 Verification of the server certificate for a TLS connection is now tried
233 (but not required) by default. The verification status is now logged by
234 default, for both outbound TLS and client-certificate supplying inbound
237 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
240 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
241 Status Notification (bounce) messages are now MIME format per RFC 3464.
242 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
243 under the control of the dsn_advertise_hosts option, and routers may
244 have a dsn_lasthop option.
246 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
247 default, modifiable by a malware= option. The list separator for
248 the options can now be changed in the usual way. Bug 68.
250 JH/10 The smtp_receive_timeout main option is now expanded before use.
252 JH/11 The incoming_interface log option now also enables logging of the
253 local interface on delivery outgoing connections.
255 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
256 if the interface and destination host and port all match.
258 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
261 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
262 Patch from Andrew Lewis.
264 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
265 now supports optional time-restrictions, weighting, and priority
266 modifiers per server. Patch originally by <rommer@active.by>.
268 JH/16 The spamd_address main option now supports a mixed list of local
269 and remote servers. Remote servers can be IPv6 addresses, and
270 specify a port-range.
272 JH/17 Bug 68: The spamd_address main option now supports an optional
273 timeout value per server.
275 JH/18 Bug 1581: Router and transport options headers_add/remove can
276 now have the list separator specified.
278 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
281 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
284 JH/21 Support for the A6 type of dns record is withdrawn.
286 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
287 rather than the verbs used.
289 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
290 from 255 to 1024 chars.
292 JH/24 Verification callouts now attempt to use TLS by default.
294 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
295 are generic router options now. The defaults didn't change.
297 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
298 Original patch from Alexander Shikoff, worked over by JH.
300 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
301 files if ZCAT_COMMAND is not executable.
303 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
305 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
307 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
308 Normally benign, it bites when the pair was led to by a CNAME;
309 modern usage is to not canoicalize the domain to a CNAME target
310 (and we were inconsistent anyway for A-only vs AAAA+A).
312 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
314 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
315 when evaluating $sender_host_dnssec.
317 JH/31 Check the HELO verification lookup for DNSSEC, adding new
318 $sender_helo_dnssec variable.
320 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
322 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
324 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
326 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
327 documented as working, but never had. Support all but $spam_report.
329 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
330 added for tls authenticator.
335 TL/01 When running the test suite, the README says that variables such as
336 no_msglog_check are global and can be placed anywhere in a specific
337 test's script, however it was observed that placement needed to be near
338 the beginning for it to behave that way. Changed the runtest perl
339 script to read through the entire script once to detect and set these
340 variables, reset to the beginning of the script, and then run through
341 the script parsing/test process like normal.
343 TL/02 The BSD's have an arc4random API. One of the functions to induce
344 adding randomness was arc4random_stir(), but it has been removed in
345 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
346 function when detected.
348 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
349 cause callback expansion.
351 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
352 syntax errors in an expansion can be treated as a string instead of
353 logging or causing an error, due to the internal use of bool_lax
354 instead of bool when processing it.
356 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
357 server certificates when making smtp deliveries.
359 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
361 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
363 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
365 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
366 Merged patch from Sebastian Wiedenroth.
368 JH/05 Fix results-pipe from transport process. Several recipients, combined
369 with certificate use, exposed issues where response data items split
370 over buffer boundaries were not parsed properly. This eventually
371 resulted in duplicates being sent. This issue only became common enough
372 to notice due to the introduction of conection certificate information,
373 the item size being so much larger. Found and fixed by Wolfgang Breyha.
375 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
376 size buffer was used, resulting in syntax errors when an expansion
379 JH/07 Add support for directories of certificates when compiled with a GnuTLS
380 version 3.3.6 or later.
382 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
383 is EXPERIMENTAL_EVENT, the main-configuration and transport options
384 both become "event_action", the variables become $event_name, $event_data
385 and $event_defer_errno. There is a new variable $verify_mode, usable in
386 routers, transports and related events. The tls:cert event is now also
387 raised for inbound connections, if the main configuration event_action
390 TL/06 In test suite, disable OCSP for old versions of openssl which contained
391 early OCSP support, but no stapling (appears to be less than 1.0.0).
393 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
394 server certificate names available under the smtp transport option
395 "tls_verify_cert_hostname" now do not permit multi-component wildcard
398 JH/10 Time-related extraction expansions from certificates now use the main
399 option "timezone" setting for output formatting, and are consistent
400 between OpenSSL and GnuTLS compilations. Bug 1541.
402 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
403 encoded parameter in the incoming message. Bug 1558.
405 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
406 include certificate info, eximon was claiming there were spoolfile
409 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
411 JH/14 Log delivery-related information more consistently, using the sequence
412 "H=<name> [<ip>]" wherever possible.
414 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
415 are problematic for Debian distribution, omit them from the release
418 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
420 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
423 JH/17 Fix a null-indirection in certextract expansions when a nondefault
424 output list separator was used.
429 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
430 checkers that were complaining about end of non-void function with no
433 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
434 This was a regression intruduced in 4.83 by another bugfix.
436 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
438 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
439 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
445 TF/01 Correctly close the server side of TLS when forking for delivery.
447 When a message was received over SMTP with TLS, Exim failed to clear up
448 the incoming connection properly after forking off the child process to
449 deliver the message. In some situations the subsequent outgoing
450 delivery connection happened to have the same fd number as the incoming
451 connection previously had. Exim would try to use TLS and fail, logging
452 a "Bad file descriptor" error.
454 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
455 utilities have not been installed.
457 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
458 temporary space as the ACL may create new global variables.
460 TL/01 LDAP support uses per connection or global context settings, depending
461 upon the detected version of the libraries at build time.
463 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
464 to extract and use the src ip:port in logging and expansions as if it
465 were a direct connection from the outside internet. PPv2 support was
466 updated based on HAProxy spec change in May 2014.
468 JH/02 Add ${listextract {number}{list}{success}{fail}}.
470 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
471 Properly escape header and check for NULL return.
473 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
476 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
478 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
479 characters in header names, implemented as a verify condition.
480 Contributed by Michael Fischer v. Mollard.
482 TL/05 Rename SPF condition results err_perm and err_temp to standardized
483 results permerror and temperror. Previous values are deprecated but
484 still accepted. In a future release, err_perm and err_temp will be
485 completely removed, which will be a backward incompatibility if the
486 ACL tests for either of these two old results. Patch contributed by
487 user bes-internal on the mailing list.
489 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
491 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
492 selectors, in both main and reject logs.
494 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
497 JH/07 Add malware type "sock" for talking to simple daemon.
499 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
501 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
502 routers/transports under cutthrough routing.
504 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
505 numbers. Touch up "bool" conditional to keep the same definition.
507 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
509 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
511 JH/12 Expand items in router/transport headers_add or headers_remove lists
512 individually rather than the list as a whole. Bug 1452.
514 Required for reasonable handling of multiple headers_ options when
515 they may be empty; requires that headers_remove items with embedded
516 colons must have them doubled (or the list-separator changed).
518 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
519 view the policy declared in the DMARC record. Currently, $dmarc_status
520 is a combined value of both the record presence and the result of the
523 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
525 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
526 dnslookup router and the smtp transport (applying to the forward
529 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
530 of ldap servers used for a specific lookup. Patch provided by Heiko
533 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
534 New variable $lookup_dnssec_authenticated for observability.
536 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
537 Patch submitted by Lars Timman.
539 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
541 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
542 Requires trusted mode and valid format message id, aborts otherwise.
543 Patch contributed by Heiko Schlichting.
545 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
546 certextract with support for various fields. Bug 1358.
548 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
549 is requested by default, modifiable by smtp transport option
552 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
553 operate on certificate variables to give certificate fingerprints
554 Also new ${sha256:cert_variable}.
556 JH/23 The PRDR feature is moved from being Experimental into the mainline.
558 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
559 Christian Aistleitner.
561 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
563 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
564 file. Patch from Wolfgang Breyha.
566 JH/25 Expand the coverage of the delivery $host and $host_address to
567 client authenticators run in verify callout. Bug 1476.
569 JH/26 Port service names are now accepted for tls_on_connect_ports, to
570 align with daemon_smtp_ports. Bug 72.
572 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
573 support and error reporting did not work properly.
575 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
576 and is readable. Patch from Andrew Colin Kissa.
578 TL/14 Enhance documentation of ${run expansion and how it parses the
579 commandline after expansion, particularly in the case when an
580 unquoted variable expansion results in an empty value.
582 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
584 PP/02 Fix internal collision of T_APL on systems which support RFC3123
585 by renaming away from it. Addresses GH issue 15, reported by
588 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
590 TL/15 SECURITY: prevent double expansion in math comparison functions
591 (can expand unsanitized data). Not remotely exploitable.
598 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
600 PP/02 Make -n do something, by making it not do something.
601 When combined with -bP, the name of an option is not output.
603 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
606 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
607 $sender_host_name and config options to manage this, and basic check
610 PP/05 DSCP support for outbound connections and control modifier for inbound.
612 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
613 (Only plugin which currently uses this is kerberos4, which nobody should
614 be using, but we should make it available and other future plugins might
615 conceivably use it, even though it would break NAT; stuff *should* be
616 using channel bindings instead).
618 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
619 name; added for Sendmail compatibility; requires admin caller.
620 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
621 just ignore it); requires trusted caller.
622 Also parse but ignore: -Ac -Am -X<logfile>
625 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
627 TL/02 Add +smtp_confirmation as a default logging option.
629 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
630 Patch by Magnus Holmgren from 2007-02-20.
632 TL/04 Bugzilla 1281 - Spec typo.
633 Bugzilla 1283 - Spec typo.
634 Bugzilla 1290 - Spec grammar fixes.
636 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
638 TL/06 Add Experimental DMARC support using libopendmarc libraries.
640 TL/07 Fix an out of order global option causing a segfault. Reported to dev
641 mailing list by by Dmitry Isaikin.
643 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
645 JH/02 Support "G" suffix to numbers in ${if comparisons.
647 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
649 NM/01 Bugzilla 1197 - Spec typo
650 Bugzilla 1196 - Spec examples corrections
652 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
654 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
655 gnutls_enable_pkcs11, but renamed to more accurately indicate its
658 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
659 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
661 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
662 "acl {{name}{arg}...}", and optional args on acl condition
665 JH/05 Permit multiple router/transport headers_add/remove lines.
667 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
669 JH/07 Avoid using a waiting database for a single-message-only transport.
670 Performance patch from Paul Fisher. Bugzilla 1262.
672 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
675 JH/09 Add $headers_added variable, with content from use of ACL modifier
676 add_header (but not yet added to the message). Bugzilla 199.
678 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
679 Pulled from Bugzilla 817 by Wolfgang Breyha.
681 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
683 (nb: this is the same fix as in Exim 4.80.1)
685 JH/11 Add A= logging on delivery lines, and a client_set_id option on
688 JH/12 Add optional authenticated_sender logging to A= and a log_selector
691 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
693 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
694 advertise SMTP AUTH mechanism to us, instead of a generic
695 protocol violation error. Also, make Exim more robust to bad
696 data from the Dovecot auth socket.
698 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
700 When a queue runner is handling a message, Exim first routes the
701 recipient addresses, during which it prunes them based on the retry
702 hints database. After that it attempts to deliver the message to
703 any remaining recipients. It then updates the hints database using
706 So if a recipient address works intermittently, it can get repeatedly
707 deferred at routing time. The retry hints record remains fresh so the
708 address never reaches the final cutoff time.
710 This is a fairly common occurrence when a user is bumping up against
711 their storage quota. Exim had some logic in its local delivery code
712 to deal with this. However it did not apply to per-recipient defers
713 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
715 This change adds a proper retry rule check during routing so that the
716 final cutoff time is checked against the message's age. We only do
717 this check if there is an address retry record and there is not a
718 domain retry record; this implies that previous attempts to handle
719 the address had the retry_use_local_parts option turned on. We use
720 this as an approximation for the destination being like a local
721 delivery, as in LMTP.
723 I suspect this new check makes the old local delivery cutoff check
724 redundant, but I have not verified this so I left the code in place.
726 TF/02 Correct gecos expansion when From: is a prefix of the username.
728 Test 0254 submits a message to Exim with the header
732 When I ran the test suite under the user fanf2, Exim expanded
733 the header to contain my full name, whereas it should have added
734 a Resent-Sender: header. It erroneously treats any prefix of the
735 username as equal to the username.
737 This change corrects that bug.
739 GF/01 DCC debug and logging tidyup
740 Error conditions log to paniclog rather than rejectlog.
741 Debug lines prefixed by "DCC: " to remove any ambiguity.
743 TF/03 Avoid unnecessary rebuilds of lookup-related code.
745 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
746 Bug spotted by Jeremy Harris; was flawed since initial commit.
747 Would have resulted in OCSP responses post-SNI triggering an Exim
748 NULL dereference and crash.
750 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
752 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
753 Bug detection, analysis and fix by Samuel Thibault.
754 Bugzilla 1331, Debian bug #698092.
756 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
758 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
759 Server implementation by Todd Lyons, client by JH.
760 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
761 config variable "prdr_enable" controls whether the server
762 advertises the facility. If the client requests PRDR a new
763 acl_data_smtp_prdr ACL is called once for each recipient, after
764 the body content is received and before the acl_smtp_data ACL.
765 The client is controlled by bolth of: a hosts_try_prdr option
766 on the smtp transport, and the server advertisement.
767 Default client logging of deliveries and rejections involving
768 PRDR are flagged with the string "PRDR".
770 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
771 fclose(). Diagnosis by Todd Lyons.
773 PP/17 Update configure.default to handle IPv6 localhost better.
774 Patch by Alain Williams (plus minor tweaks).
777 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
778 This is now consistent with GnuTLS, and is now documented: the
779 previous undocumented portable approach to treating the option as
780 unset was to force an expansion failure. That still works, and
781 an empty string is now equivalent.
783 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
784 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
785 not performing validation itself.
787 PP/20 Added force_command boolean option to pipe transport.
788 Patch from Nick Koston, of cPanel Inc.
790 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
793 TF/04 Added udpsend ACL modifer and hexquote expansion operator
795 PP/21 Fix eximon continuous updating with timestamped log-files.
796 Broken in a format-string cleanup in 4.80, missed when I repaired the
797 other false fix of the same issue.
798 Report and fix from Heiko Schlichting.
801 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
802 Report from Prashanth Katuri.
804 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
805 It's SecureTransport, so affects any MacOS clients which use the
806 system-integrated TLS libraries, including email clients.
808 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
809 using a MIME ACL for non-SMTP local injection.
810 Report and assistance in diagnosis by Warren Baker.
812 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
814 JH/16 Fix comparisons for 64b. Bugzilla 1385.
816 TL/09 Add expansion variable $authenticated_fail_id to keep track of
817 last id that failed so it may be referenced in subsequent ACL's.
819 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
822 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
823 ldap library initialization, allowing self-signed CA's to be
824 used. Also properly sets require_cert option later in code by
825 using NULL (global ldap config) instead of ldap handle (per
826 session). Bug diagnosis and testing by alxgomz.
828 TL/12 Enhanced documentation in the ratelimit.pl script provided in
829 the src/util/ subdirectory.
831 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
832 renamed to Transport Post Delivery Action by Jeremy Harris, as
835 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
836 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
837 redis_servers = needs to be configured which will be used by the redis
838 lookup. Patch from Warren Baker, of The Packet Hub.
840 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
842 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
843 hostname or reverse DNS when processing a host list. Used suggestions
844 from multiple comments on this bug.
846 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
848 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
849 Missed a few lines, added it to make the runtest require no keyboard
852 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
853 contains upper case chars. Make router use caseful_local_part.
855 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
856 support when GnuTLS has been built with p11-kit.
862 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
864 This, or similar/improved, will also be change PP/11 of 4.82.
870 PP/01 Handle short writes when writing local log-files.
871 In practice, only affects FreeBSD (8 onwards).
872 Bugzilla 1053, with thanks to Dmitry Isaikin.
874 NM/01 Bugzilla 949 - Documentation tweak
876 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
879 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
881 PP/02 Implemented gsasl authenticator.
883 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
885 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
886 `pkg-config foo` for cflags/libs.
888 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
889 with rest of GSASL and with heimdal_gssapi.
891 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
892 `pkg-config foo` for cflags/libs for the TLS implementation.
894 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
895 properties get this fed in as external SSF. A number of robustness
896 and debugging improvements to the cyrus_sasl authenticator.
898 PP/08 cyrus_sasl server now expands the server_realm option.
900 PP/09 Bugzilla 1214 - Log authentication information in reject log.
901 Patch by Jeremy Harris.
903 PP/10 Added dbmjz lookup type.
905 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
907 PP/12 MAIL args handles TAB as well as SP, for better interop with
908 non-compliant senders.
909 Analysis and variant patch by Todd Lyons.
911 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
912 Bug report from Lars Müller <lars@samba.org> (via SUSE),
913 Patch from Dirk Mueller <dmueller@suse.com>
915 PP/13 tls_peerdn now print-escaped for spool files.
916 Observed some $tls_peerdn in wild which contained \n, which resulted
917 in spool file corruption.
919 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
920 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
921 or write after TLS renegotiation, which otherwise led to messages
924 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
925 as a tracking header (ie: a signed header comes before the signature).
926 Patch from Wolfgang Breyha.
928 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
929 comma-sep list; embedded commas doubled.
931 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
933 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
935 Report and patch from Dmitry Banschikov.
937 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
938 Removed SSL_clear() after SSL_new() which led to protocol negotiation
939 failures. We appear to now support TLS1.1+ with Exim.
941 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
942 lets Exim select keys and certificates based upon TLS SNI from client.
943 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
944 before an outbound SMTP session. New log_selector, +tls_sni.
946 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
947 NULL dereference. Report and patch from Alun Jones.
949 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
950 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
951 Not seeing resolver debug output on NetBSD, but suspect this is a
952 resolver implementation change.
954 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
955 Left warnings. Added "eximon gdb" invocation mode.
957 PP/21 Defaulting "accept_8bitmime" to true, not false.
959 PP/22 Added -bw for inetd wait mode support.
961 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
962 locate the relevant includes and libraries. Made this the default.
964 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
965 Bugzilla 1246, report and most of solution from Tomasz Kusy.
967 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
968 This may cause build issues on older platforms.
970 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
971 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
972 gnutls_require_mac & gnutls_require_protocols (no longer supported).
973 Added SNI support via GnuTLS too.
974 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
976 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
978 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
979 Applied second patch from Janne, implementing suggestion to default
980 multiple-strings-in-record handling to match SPF spec.
982 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
984 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
985 read-only, out of scope).
986 Patch from Wolfgang Breyha, report from Stuart Northfield.
988 PP/29 Fix three issues highlighted by clang analyser static analysis.
989 Only crash-plausible issue would require the Cambridge-specific
990 iplookup router and a misconfiguration.
991 Report from Marcin Mirosław.
993 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
995 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
996 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
997 As part of this, removing so much warning spew let me fix some minor
998 real issues in debug logging.
1000 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1001 assignment on my part. Fixed.
1003 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1004 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1005 Janne Snabb (who went above and beyond: thank you).
1007 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1008 string otherwise requires a connection and a bunch more work and it's
1009 relatively easy to get wrong. Should also expose TLS library linkage
1012 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1013 64-bit ${eval} (JH/03).
1015 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1016 GNU libc to support some of the 64-bit stuff, should not lead to
1017 conflicts. Defined before os.h is pulled in, so if a given platform
1018 needs to override this, it can.
1020 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1021 protection layer was required, which is not implemented.
1022 Bugzilla 1254, patch from Wolfgang Breyha.
1024 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1025 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1026 tls_dhparam take prime identifiers. Also unbreak combination of
1027 OpenSSL+DH_params+TLSSNI.
1029 PP/39 Disable SSLv2 by default in OpenSSL support.
1035 PP/01 Solaris build fix for Oracle's LDAP libraries.
1036 Bugzilla 1109, patch from Stephen Usher.
1038 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1040 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1043 TF/02 Fix a couple more cases where we did not log the error message
1044 when unlink() failed. See also change 4.74-TF/03.
1046 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1047 lock up or crash if it happened to be inside a call to libc when it
1048 got a SIGUSR1 from exiwhat.
1050 The SIGUSR1 handler appends the current process status to the process
1051 log which is later printed by exiwhat. It used to use the general
1052 purpose logging code to do this, but several functions it calls are
1053 not safe for signals.
1055 The new output code in the SIGUSR1 handler is specific to the process
1056 log, and simple enough that it's easy to inspect for signal safety.
1057 Removing some special cases also simplifies the general logging code.
1058 Removing the spurious timestamps from the process log simplifies
1061 TF/04 Improved ratelimit ACL condition.
1063 The /noupdate option has been deprecated in favour of /readonly which
1064 has clearer semantics. The /leaky, /strict, and /readonly update modes
1065 are mutually exclusive. The update mode is no longer included in the
1066 database key; it just determines when the database is updated. (This
1067 means that when you upgrde Exim will forget old rate measurements.)
1069 Exim now checks that the per_* options are used with an update mode that
1070 makes sense for the current ACL. For example, when Exim is processing a
1071 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1072 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1073 must specify per_mail/readonly. If you omit the update mode it defaults to
1074 /leaky where that makes sense (as before) or /readonly where required.
1076 The /noupdate option is now undocumented but still supported for
1077 backwards compatibility. It is equivalent to /readonly except that in
1078 ACLs where /readonly is required you may specify /leaky/noupdate or
1079 /strict/noupdate which are treated the same as /readonly.
1081 A useful new feature is the /count= option. This is a generalization
1082 of the per_byte option, so that you can measure the throughput of other
1083 aggregate values. For example, the per_byte option is now equivalent
1084 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1086 The per_rcpt option has been generalized using the /count= mechanism
1087 (though it's more complicated than the per_byte equivalence). When it is
1088 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1089 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1090 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1091 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1092 Note that using per_rcpt with a non-readonly update mode in more than
1093 one ACL will cause the recipients to be double-counted. (The per_mail
1094 and per_byte options don't have this problem.)
1096 The handling of very low rates has changed slightly. If the computed rate
1097 is less than the event's count (usually one) then this event is the first
1098 after a long gap. In this case the rate is set to the same as this event's
1099 count, so that the first message of a spam run is counted properly.
1101 The major new feature is a mechanism for counting the rate of unique
1102 events. The new per_addr option counts the number of different
1103 recipients that someone has sent messages to in the last time period. It
1104 behaves like per_rcpt if all the recipient addresses are different, but
1105 duplicate recipient addresses do not increase the measured rate. Like
1106 the /count= option this is a general mechanism, so the per_addr option
1107 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1108 example, measure the rate that a client uses different sender addresses
1109 with the options per_mail/unique=$sender_address. There are further
1110 details in the main documentation.
1112 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1114 TF/06 Removed a few PCRE remnants.
1116 TF/07 Automatically extract Exim's version number from tags in the git
1117 repository when doing development or release builds.
1119 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1120 Bugzilla 879. Patch from Paul Fisher.
1122 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1123 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1126 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1128 PP/05 Variable $av_failed, true if the AV scanner deferred.
1129 Bugzilla 1078. Patch from John Horne.
1131 PP/06 Stop make process more reliably on build failure.
1132 Bugzilla 1087. Patch from Heiko Schlittermann.
1134 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1135 Bugzilla 1089. Patch from Heiko Schlittermann.
1137 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1138 Bugzilla 1131. Patch from Holger Weiß.
1140 PP/09 Handle IPv6 addresses with SPF.
1141 Bugzilla 860. Patch from Wolfgang Breyha.
1143 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1145 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1148 PP/11 match_* no longer expand right-hand-side by default.
1149 New compile-time build option, EXPAND_LISTMATCH_RHS.
1150 New expansion conditions, "inlist", "inlisti".
1152 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1154 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1156 PP/14 fix log_write() format string regression from TF/03.
1157 Bugzilla 1152. Patch from Dmitry Isaikin.
1163 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1165 PP/02 Harmonised TLS library version reporting; only show if debugging.
1166 Layout now matches that introduced for other libraries in 4.74 PP/03.
1168 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1170 PP/04 New "dns_use_edns0" global option.
1172 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1175 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1176 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1178 TK/01 Updated PolarSSL code to 0.14.2.
1179 Bugzilla 1097. Patch from Andreas Metzler.
1181 PP/07 Catch divide-by-zero in ${eval:...}.
1182 Fixes bugzilla 1102.
1184 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1187 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1188 format-string attack -- SECURITY: remote arbitrary code execution.
1190 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1191 time unintentionally subject to list matching rules, letting the header
1192 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1193 arbitrary string expansion). This allowed for information disclosure.
1195 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1196 INT_MIN/-1 -- value coerced to INT_MAX.
1202 NM/01 Workround for PCRE version dependency in version reporting
1205 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1206 This fixes portability to compilers other than gcc, notably
1207 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1209 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1210 makefiles for portability to HP-UX and POSIX correctness.
1212 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1213 Also via indented variable definition in the Makefile.
1214 (Debugging by Oliver Heesakkers).
1216 PP/02 Restore caching of spamd results with expanded spamd_address.
1217 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1219 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1220 Improves build reliability. Fix from: Frank Elsner
1222 NM/02 Fix wide character breakage in the rfc2047 coding
1223 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1225 NM/03 Allow underscore in dnslist lookups
1226 Fixes bug 1026. Patch from Graeme Fowler
1228 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1229 Code patches from Adam Ciarcinski of NetBSD.
1231 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1234 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1235 is logged, to avoid truncation. Patch from John Horne.
1237 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1238 Patch from Jakob Hirsch.
1240 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1241 SQL string expansion failure details.
1242 Patch from Andrey Oktyabrski.
1244 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1245 Patch from Simon Arlott.
1247 PP/09 New lookups functionality failed to compile on old gcc which rejects
1248 extern declarations in function scope.
1249 Patch from Oliver Fleischmann
1251 PP/10 Use sig_atomic_t for flags set from signal handlers.
1252 Check getgroups() return and improve debugging.
1253 Fixed developed for diagnosis in bug 927 (which turned out to be
1256 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1257 Patch from Mark Zealey.
1259 PP/12 Bugzilla 1056: Improved spamd server selection.
1260 Patch from Mark Zealey.
1262 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1263 Based on patch from Heiko Schlittermann.
1265 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1266 Patch from Uwe Doering, sign-off by Michael Haardt.
1268 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1269 variable declaration deep within a block. Bug and patch from
1272 PP/15 lookups-Makefile IRIX compatibilty coercion.
1274 PP/16 Make DISABLE_DKIM build knob functional.
1276 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1277 Patch by Simon Arlott
1279 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1280 variable argument macros. Our copy now differs from upstream.
1286 TF/01 Failure to get a lock on a hints database can have serious
1287 consequences so log it to the panic log.
1289 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1290 controlled using the smtp_confirmation log selector.
1292 TF/03 Include the error message when we fail to unlink a spool file.
1294 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1295 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1296 for maintaining out-of-tree patches for some time.
1298 PP/01 Bugzilla 139: Documentation and portability issues.
1299 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1300 Handle per-OS dynamic-module compilation flags.
1302 PP/02 Let /dev/null have normal permissions.
1303 The 4.73 fixes were a little too stringent and complained about the
1304 permissions on /dev/null. Exempt it from some checks.
1305 Reported by Andreas M. Kirchwitz.
1307 PP/03 Report version information for many libraries, including
1308 Exim version information for dynamically loaded libraries. Created
1309 version.h, now support a version extension string for distributors
1310 who patch heavily. Dynamic module ABI change.
1312 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1313 privilege escalation vulnerability whereby the Exim run-time user
1314 can cause root to append content of the attacker's choosing to
1317 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1320 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1321 If dropping privileges for untrusted macros, we disabled normal logging
1322 on the basis that it would fail; for the Exim run-time user, this is not
1323 the case, and it resulted in successful deliveries going unlogged.
1324 Fixed. Reported by Andreas Metzler.
1330 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1331 only prepend for the Resent-* case. Fixes regression introduced in
1332 Exim 4.70 by NM/22 for Bugzilla 607.
1334 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1335 increasing numbers of administrators be bitten by this.
1337 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1339 PP/03 Bugzilla 994: added openssl_options main configuration option.
1341 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1343 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1345 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1347 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1348 without a peer certificate, leading to a segfault because of an
1349 assumption that peers always have certificates. Be a little more
1350 paranoid. Problem reported by Martin Tscholak.
1352 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1353 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1354 NB: ClamAV planning to remove STREAM in "middle of 2010".
1355 CL also introduces -bmalware, various -d+acl logging additions and
1356 more caution in buffer sizes.
1358 PP/09 Implemented reverse_ip expansion operator.
1360 PP/10 Bugzilla 937: provide a "debug" ACL control.
1362 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1364 PP/12 Bugzilla 973: Implement --version.
1366 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1368 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1370 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1372 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1373 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1374 ignore trailing whitespace.
1376 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1378 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1381 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1382 Notification from Dr Andrew Aitchison.
1384 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1385 ExtendedDetectionInfo response format.
1386 Notification from John Horne.
1388 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1391 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1392 XSL and documented dependency on system catalogs, with examples of how
1395 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1398 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1399 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1400 configuration file which is writeable by the Exim user or group.
1402 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1403 of configuration files to cover files specified with the -C option if
1404 they are going to be used with root privileges, not just the default
1407 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1408 option (effectively making it always true).
1410 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1411 files to be used while preserving root privileges.
1413 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1414 that rogue child processes cannot use them.
1416 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1417 run-time user, instead of root.
1419 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1420 Exim run-time user without dropping privileges.
1422 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1423 result string, instead of calling string_vformat() twice with the same
1426 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1427 for other users. Others should always drop root privileges if they use
1428 -C on the command line, even for a whitelisted configure file.
1430 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1432 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1438 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1439 $data_path, and $header_path variables; fixed documentation bugs and
1442 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1443 exipick to access non-standard spools, including the "frozen" queue
1446 NM/01 Bugzilla 965: Support mysql stored procedures.
1447 Patch from Alain Williams
1449 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1451 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1452 Patch from Andreas Metzler
1454 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1455 Patch from Kirill Miazine
1457 NM/05 Bugzilla 671: Added umask to procmail example.
1459 JJ/03 installed exipick 20100323.0, fixing doc bug
1461 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1462 directory. Notification and patch from Dan Rosenberg.
1464 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1466 TK/02 Improve log output when DKIM signing operation fails.
1468 MH/01 Treat the transport option dkim_domain as a colon separated
1469 list, not as a single string, and sign the message with each element,
1470 omitting multiple occurences of the same signer.
1472 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1473 Bugzilla 985, 986. Patch by Simon Arlott
1475 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1476 Patch by Simon Arlott
1478 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1479 MBX locking. Notification from Dan Rosenberg.
1485 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1487 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1489 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1491 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1493 NM/04 Bugzilla 915: Flag broken perl installation during build.
1499 TK/01 Added patch by Johannes Berg that expands the main option
1500 "spamd_address" if it starts with a dollar sign.
1502 TK/02 Write list of recipients to X-Envelope-Sender header when building
1503 the mbox-format spool file for content scanning (suggested by Jakob
1506 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1507 (http://www.dcc-servers.net/) support via dccifd. Activated by
1508 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1510 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1511 by Mark Daniel Reidel <mr@df.eu>.
1513 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1514 When building exim an external PCRE library is now needed -
1515 PCRE is a system library on the majority of modern systems.
1516 See entry on PCRE_LIBS in EDITME file.
1518 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1519 conversation. Added nologin parameter to request.
1520 Patch contributed by Kirill Miazine.
1522 TF/01 Do not log submission mode rewrites if they do not change the address.
1524 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1526 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1527 log files in place. Contributed by Roberto Lima.
1529 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1531 TF/03 Bugzilla 615: When checking the local_parts router precondition
1532 after a local_part_suffix or local_part_prefix option, Exim now
1533 does not use the address's named list lookup cache, since this
1534 contains cached lookups for the whole local part.
1536 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1537 Robert Millan. Documentation is in experimental-spec.txt.
1539 TF/04 Bugzilla 668: Fix parallel build (make -j).
1541 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1543 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1544 Patch provided by Jan Srzednicki.
1546 TF/05 Leading white space used to be stripped from $spam_report which
1547 wrecked the formatting. Now it is preserved.
1549 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1550 that they are available at delivery time.
1552 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1554 TF/08 TLS error reporting now respects the incoming_interface and
1555 incoming_port log selectors.
1557 TF/09 Produce a more useful error message if an SMTP transport's hosts
1558 setting expands to an empty string.
1560 NM/06 Bugzilla 744: EXPN did not work under TLS.
1561 Patch provided by Phil Pennock.
1563 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1564 Patch provided by Richard Godbee.
1566 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1567 acl_smtp_notquit, added index entry.
1569 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1570 Patch provided by Eugene Bujak.
1572 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1573 accept(). Patch provided by Maxim Dounin.
1575 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1576 Patch provided by Phil Pennock.
1578 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1580 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1581 Patch provided by Brad "anomie" Jorsch.
1583 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1584 Patch provided by Dean Brooks.
1586 TK/05 Add native DKIM support (does not depend on external libraries).
1588 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1589 Patch provided by Graeme Fowler.
1591 NM/16 Bugzilla 851: Documentation example syntax fix.
1593 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1595 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1598 NM/19 Bugzilla 745: TLS version reporting.
1599 Patch provided by Phil Pennock.
1601 NM/20 Bugzilla 167: bool: condition support.
1602 Patch provided by Phil Pennock.
1604 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1605 clients. Patch provided by Phil Pennock.
1607 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1608 Patch provided by Brad "anomie" Jorsch.
1610 NM/23 Bugzilla 687: Fix misparses in eximstats.
1611 Patch provided by Heiko Schlittermann.
1613 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1614 Patch provided by Heiko Schlittermann.
1616 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1617 plus update to original patch.
1619 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1621 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1622 Patch provided by David Brownlee.
1624 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1626 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1628 NM/30 Bugzilla 888: TLS documentation bugfixes.
1630 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1632 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1633 Unlike the original bugzilla I have changed all shell scripts in src tree.
1635 NM/33 Bugzilla 898: Transport filter timeout fix.
1636 Patch by Todd Rinaldo.
1638 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1639 Patch by Serge Demonchaux.
1641 NM/35 Bugzilla 39: Base64 decode bug fixes.
1642 Patch by Jakob Hirsch.
1644 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1646 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1648 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1650 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1656 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1657 ALT-N's libdkim that I have put here:
1658 http://duncanthrax.net/exim-experimental/
1660 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1661 were called 'true' and it seems that C99 defines that as a reserved
1662 keyword to be used with 'bool' variable types. That means you could
1663 not include C99-style headers which use bools without triggering
1664 build errors in sieve.c.
1666 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1667 as mailq or other aliases. Changed the --help handling significantly
1668 to do whats expected. exim_usage() emits usage/help information.
1670 SC/01 Added the -bylocaldomain option to eximstats.
1672 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1674 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1676 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1682 PH/01 Another patch from the Sieve maintainer.
1684 PH/02 When an IPv6 address is converted to a string for single-key lookup
1685 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1686 dots are used instead of colons so that keys in lsearch files need not
1687 contain colons. This was done some time before quoting was made available
1688 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1689 (notated using the quote facility) so as to distinguish them from IPv4
1690 keys. This meant that lookups for IP addresses in host lists did not work
1691 for iplsearch lookups.
1693 This has been fixed by arranging for IPv6 addresses to be expressed with
1694 colons if the lookup type is iplsearch. This is not incompatible, because
1695 previously such lookups could never work.
1697 The situation is now rather anomolous, since one *can* have colons in
1698 ordinary lsearch keys. However, making the change in all cases is
1699 incompatible and would probably break a number of configurations.
1701 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1704 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1705 conversion specification without a maximum field width, thereby enabling
1706 a rogue spamd server to cause a buffer overflow. While nobody in their
1707 right mind would setup Exim to query an untrusted spamd server, an
1708 attacker that gains access to a server running spamd could potentially
1709 exploit this vulnerability to run arbitrary code as the Exim user.
1711 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1712 $primary_hostname instead of what libspf2 thinks the hosts name is.
1714 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1715 a directory entry by the name of the lookup key. Previously, if a
1716 symlink pointed to a non-existing file or a file in a directory that
1717 Exim lacked permissions to read, a lookup for a key matching that
1718 symlink would fail. Now it is enough that a matching directory entry
1719 exists, symlink or not. (Bugzilla 503.)
1721 PH/03 The body_linecount and body_zerocount variables are now exported in the
1724 PH/04 Added the $dnslist_matched variable.
1726 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1727 This means they are set thereafter only if the connection becomes
1730 PH/06 Added the client_condition to authenticators so that some can be skipped
1731 by clients under certain conditions.
1733 PH/07 The error message for a badly-placed control=no_multiline_responses left
1734 "_responses" off the end of the name.
1736 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1738 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1739 (without spaces) instead of just copying the configuration text.
1741 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1743 PH/11 Added $max_received_linelength.
1745 PH/12 Added +ignore_defer and +include_defer to host lists.
1747 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1748 way in which PCRE > 7.0 is built.
1750 PH/14 Implemented queue_only_load_latch.
1752 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1753 MAIL command. The effect was to mangle the value on 64-bit systems.
1755 PH/16 Another patch from the Sieve maintainer.
1757 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1759 PH/18 If a system quota error occurred while trying to create the file for
1760 a maildir delivery, the message "Mailbox is full" was not appended to the
1761 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1762 only to a quota excession during the actual writing of the file.
1764 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1765 characters?) which causes problems in log lines. The DN values are now
1766 passed through string_printing() before being added to log lines.
1768 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1769 and InterBase are left for another time.)
1771 PH/21 Added message_body_newlines option.
1773 PH/22 Guard against possible overflow in moan_check_errorcopy().
1775 PH/23 POSIX allows open() to be a macro; guard against that.
1777 PH/24 If the recipient of an error message contained an @ in the local part
1778 (suitably quoted, of course), incorrect values were put in $domain and
1779 $local_part during the evaluation of errors_copy.
1785 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1786 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1789 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1790 issue a MAIL command.
1792 PH/02 In an ACL statement such as
1794 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1796 if a client was not listed at all, or was listed with a value other than
1797 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1798 the condition was not true (as it should be), so access was not denied.
1799 The bug was that the ! inversion was incorrectly passed on to the second
1800 item. This has been fixed.
1802 PH/03 Added additional dnslists conditions == and =& which are different from
1803 = and & when the dns lookup returns more than one IP address.
1805 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1806 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1808 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1809 FSYNC, which compiles an option called disable_fsync that allows for
1810 bypassing fsync(). The documentation is heavily laced with warnings.
1812 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1814 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1815 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1816 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1817 including adding "make clean"; (3) Added -fPIC when compiling the test
1818 dynamically loaded module, to get rid of a warning.
1820 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1821 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1822 = 0s. The bug is otherwise harmless.
1824 PH/07 There was a bug in the dovecot authenticator such that the value of
1825 $auth1 could be overwritten, and so not correctly preserved, after a
1826 successful authentication. This usually meant that the value preserved by
1827 the server_setid option was incorrect.
1829 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1831 PH/09 Installed PCRE release 7.0.
1833 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1834 run for batched SMTP input. It is now run at the start of every message
1835 in the batch. While fixing this I discovered that the process information
1836 (output by running exiwhat) was not always getting set for -bs and -bS
1837 input. This is fixed, and it now also says "batched" for BSMTP.
1839 PH/11 Added control=no_pipelining.
1841 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1842 patch, slightly modified), and move the expansion of helo_data till after
1843 the connection is made in the smtp transport (so it can use these
1846 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1848 PH/14 Added log_selector = +pid.
1850 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1852 PH/16 Add ${if forany and ${if forall.
1854 PH/17 Added dsn_from option to vary the From: line in DSNs.
1856 PH/18 Flush SMTP output before performing a callout, unless control =
1857 no_callout_flush is set.
1859 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1860 was true (the default) a successful delivery failed to delete the retry
1861 item, thus causing premature timeout of the address. The bug is now
1864 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1866 PH/21 Long custom messages for fakedefer and fakereject are now split up
1867 into multiline reponses in the same way that messages for "deny" and
1868 other ACL rejections are.
1870 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1871 with slight modification.
1873 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1874 draft, changing the syntax and factoring some duplicate code".
1876 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1877 for deliveries of the second and subsequent messages over the same SMTP
1880 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1881 ${reduce, with only minor "tidies".
1883 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1885 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1886 expansion side effects.
1888 PH/27 When a message times out after an over-quota error from an Exim-imposed
1889 quota, the bounce message says "mailbox is full". This message was not
1890 being given when it was a system quota that was exceeded. It now should
1893 MH/03 Made $recipients available in local_scan(). local_scan() already has
1894 better access to the recipient list through recipients_list[], but
1895 $recipients can be useful in postmaster-provided expansion strings.
1897 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1898 in the case of a MAIL command with additional options following the
1899 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1900 were accidentally chopped off.
1902 PH/29 SMTP synchronization checks are implemented when a command is read -
1903 there is a check that no more input is waiting when there shouldn't be
1904 any. However, for some commands, a delay in an ACL can mean that it is
1905 some time before the response is written. In this time, more input might
1906 arrive, invalidly. So now there are extra checks after an ACL has run for
1907 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1908 pipelining has not been advertised.
1910 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1912 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1913 returned in the SMTP response when smtp_return_error_details was false.
1914 This has been fixed.
1916 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1917 instead of the C I/O that was originally supplied, because problems were
1918 reported on Solaris.
1920 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1921 Exim which did not show up earlier: it was assuming that a call to
1922 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1923 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1924 was a macro that became an assignment, so it seemed to work. This has
1925 changed to a proper function call with a void return, hence the compile
1926 error. Exim's code has been fixed.
1928 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1931 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1933 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1935 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1936 because it turns out to be too big or there is a timeout) there is no
1937 "Completed" line in the log. When some messages of this type were
1938 selected by exigrep, they were listed as "not completed". Others were
1939 picked up by some special patterns. I have improved the selection
1940 criteria to be more general.
1942 PH/38 The host_find_failed option in the manualroute router can now be set
1943 to "ignore", to completely ignore a host whose IP address cannot be
1944 found. If all hosts are ignored, the behaviour is controlled by the new
1945 host_all_ignored option.
1947 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1948 homing or because of multiple MX records with /mx) generated more than
1949 one IP address, and the following item turned out to be the local host,
1950 all the secondary addresses of the first item were incorrectly removed
1951 from the list, along with the local host and any following hosts (which
1952 is what is supposed to happen).
1954 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1955 whoever called Exim into the -H file. In the case of the daemon it was
1956 behaving confusingly. When first started, it used values for whoever
1957 started the daemon, but after a SIGHUP it used the Exim user (because it
1958 calls itself on a restart). I have changed the code so that it now always
1961 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1962 message are rejected with the same error (e.g. no authentication or bad
1963 sender address), and a DATA command is nevertheless sent (as can happen
1964 with PIPELINING or a stupid MUA), the error message that was given to the
1965 RCPT commands is included in the rejection of the DATA command. This is
1966 intended to be helpful for MUAs that show only the final error to their
1969 PH/42 Another patch from the Sieve maintainer.
1971 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1972 Eximstats - Fixed some broken HTML links and added missing column headers
1974 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1975 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1977 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1979 PH/43 Yet another patch from the Sieve maintainer.
1981 PH/44 I found a way to check for a TCP/IP connection going away before sending
1982 the response to the final '.' that terminates a message, but only in the
1983 case where the client has not sent further data following the '.'
1984 (unfortunately, this is allowed). However, in many cases there won't be
1985 any further data because there won't be any more messages to send. A call
1986 to select() can be used: if it shows that the input is "ready", there is
1987 either input waiting, or the socket has been closed. An attempt to read
1988 the next input character can distinguish the two cases. Previously, Exim
1989 would have sent an OK response which the client would never have see.
1990 This could lead to message repetition. This fix should cure that, at
1991 least in a lot of common cases.
1993 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1994 advertised in response to EHLO.
2000 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2001 fixed by 4.65/MH/01 (is this a record?) are fixed:
2003 (i) An empty string was always treated as zero by the numeric comparison
2004 operators. This behaviour has been restored.
2006 (ii) It is documented that the numeric comparison operators always treat
2007 their arguments as decimal numbers. This was broken in that numbers
2008 starting with 0 were being interpreted as octal.
2010 While fixing these problems I realized that there was another issue that
2011 hadn't been noticed. Values of message_size_limit (both the global option
2012 and the transport option) were treated as octal if they started with 0.
2013 The documentation was vague. These values are now always treated as
2014 decimal, and I will make that clear in the documentation.
2020 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2021 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2024 MH/01 Don't check that the operands of numeric comparison operators are
2025 integers when their expansion is in "skipping" mode (fixes bug
2026 introduced by 4.64-PH/07).
2028 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2029 child addresses, Exim now panics and dies. Previously, because the count
2030 is held in a short int, deliveries were likely to be lost. As such a
2031 large number of recipients for a single message is ridiculous
2032 (performance will be very, very poor), I have chosen to impose a limit
2033 rather than extend the field.
2039 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2040 leftover -K file (the existence of which was triggered by #402).
2041 While we were at it, introduced process PID as part of the -K
2042 filename. This should rule out race conditions when creating
2045 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2046 processing considerably. Previous code took too long for large mails,
2047 triggering a timeout which in turn triggers #401.
2049 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2050 in the DK code in transports.c. sendfile() is not really portable,
2051 hence the _LINUX specificness.
2053 TF/01 In the add_headers option to the mail command in an Exim filter,
2054 there was a bug that Exim would claim a syntax error in any
2055 header after the first one which had an odd number of characters
2058 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2059 callout verification, Exim cached a "reject" for the entire domain. This
2060 is correct for most verifications, but it is not correct for a recipient
2061 verification with use_sender or use_postmaster set, because in that case
2062 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2063 case of MAIL FROM:<> rejection from other early rejections (e.g.
2064 rejection of HELO). When verifying a recipient using a non-null MAIL
2065 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2066 Whatever the result of the callout, the value of the domain cache is
2067 left unchanged (for any other kind of callout, getting as far as trying
2068 RCPT means that the domain itself is ok).
2070 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2073 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2074 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2075 session, the daemon ignores ECONNECT errors and logs others; it now
2076 ignores EPIPE as well.
2078 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2079 (quoted-printable decoding).
2081 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2082 later the small subsequent patch to fix an introduced bug.
2084 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2086 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2088 PH/08 An error is now given if message_size_limit is specified negative.
2090 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2091 to be given (somewhat) arbitrary names.
2093 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2096 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2097 miscellaneous code fixes
2099 PH/10 Added the log_reject_target ACL modifier to specify where to log
2102 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2103 hostname. This is wrong, because it relates to the incoming message (and
2104 probably the interface on which it is arriving) and not to the outgoing
2105 callout (which could be using a different interface). This has been
2106 changed to use the value of the helo_data option from the smtp transport
2107 instead - this is what is used when a message is actually being sent. If
2108 there is no remote transport (possible with a router that sets up host
2109 addresses), $smtp_active_hostname is used.
2111 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2112 tweaks were necessary in order to get it to work (see also 21 below):
2113 (a) The code assumed that strncpy() returns a negative number on buffer
2114 overflow, which isn't the case. Replaced with Exim's string_format()
2116 (b) There were several signed/unsigned issues. I just did the minimum
2117 hacking in of casts. There is scope for a larger refactoring.
2118 (c) The code used strcasecmp() which is not a standard C function.
2119 Replaced with Exim's strcmpic() function.
2120 (d) The code set only $1; it now sets $auth1 as well.
2121 (e) A simple test gave the error "authentication client didn't specify
2122 service in request". It would seem that Dovecot has changed its
2123 interface. Fortunately there's a specification; I followed it and
2124 changed what the client sends and it appears to be working now.
2126 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2129 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2130 address A is aliased to B and C, where B exists and C does not. Without
2131 -v the output is "A verified" because verification stops after a
2132 successful redirection if more than one address is generated. However,
2133 with -v the child addresses are also verified. Exim was outputting "A
2134 failed to verify" and then showing the successful verification for C,
2135 with its parentage. It now outputs "B failed to verify", showing B's
2136 parentage before showing the successful verification of C.
2138 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2139 look up a TXT record in a specific list after matching in a combined
2142 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2143 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2144 they consult the DNS. I had assumed they would set it the way they
2145 wanted; and indeed my experiments on Linux seem to show that in some
2146 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2147 To be on the safe side, however, I have now made the interface to
2148 host_find_byname() similar to host_find_bydns(), with an argument
2149 containing the DNS resolver options. The host_find_byname() function now
2150 sets these options at its start, just as host_find_bydns() does. The smtp
2151 transport options dns_qualify_single and dns_search_parents are passed to
2152 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2153 of host_find_byname() use the default settings of RES_DEFNAMES
2154 (qualify_single) but not RES_DNSRCH (search_parents).
2156 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2157 spool_read_header() do less string testing, by means of a preliminary
2158 switch on the second character of optional "-foo" lines. (This is
2159 overdue, caused by the large number of possibilities that now exist.
2160 Originally there were few.) While I was there, I also converted the
2161 str(n)cmp tests so they don't re-test the leading "-" and the first
2162 character, in the hope this might squeeze out yet more improvement.
2164 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2165 flag allowing group syntax was set by the header_syntax check but not
2166 turned off, possible causing trouble later; (2) The flag was not being
2167 set at all for the header_verify test, causing "group"-style headers to
2168 be rejected. I have now set it in this case, and also caused header_
2169 verify to ignore an empty address taken from a group. While doing this, I
2170 came across some other cases where the code for allowing group syntax
2171 while scanning a header line wasn't quite right (mostly, not resetting
2172 the flag correctly in the right place). These bugs could have caused
2173 trouble for malformed header lines. I hope it is now all correct.
2175 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2176 with the "reply" argument non-NULL. The code, however (which originally
2177 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2178 but it didn't always do it. This confused somebody who was copying the
2179 code for some other use. I have removed all the tests.
2181 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2182 feature that was used to support insecure browsers during the U.S. crypto
2183 embargo. It requires special client support, and Exim is probably the
2184 only MTA that supported it -- and would never use it because real RSA is
2185 always available. This code has been removed, because it had the bad
2186 effect of slowing Exim down by computing (never used) parameters for the
2187 RSA_EXPORT functionality.
2189 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2190 authenticator to fail if there's a tab character in the incoming data
2191 (there should never be unless someone is messing about, as it's supposed
2192 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2193 if the connection is using TLS or if the remote IP is the same as the
2194 local IP, and the "valid-client-cert option" if a client certificate has
2197 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2198 authenticators. This can be used for authorization after authentication
2199 succeeds. (In the case of plaintext, it servers for both authentication
2202 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2203 if any retry times were supplied.
2205 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2206 connection, where there is no client IP address to check. In this
2207 situation, the verify now always succeeds.
2209 PH/25 Applied John Jetmore's -Mset patch.
2211 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2213 PH/27 In a string expansion for a processed (not raw) header when multiple
2214 headers of the same name were present, leading whitespace was being
2215 removed from all of them, but trailing whitespace was being removed only
2216 from the last one. Now trailing whitespace is removed from each header
2217 before concatenation. Completely empty headers in a concatenation (as
2218 before) are ignored.
2220 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2221 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2223 PH/29 [Removed. This was a change that I later backed out, and forgot to
2224 correct the ChangeLog entry (that I had efficiently created) before
2225 committing the later change.]
2227 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2228 address errors (4xx response to RCPT) over the same connection as other
2229 messages routed to the same hosts. Such deliveries are always "forced",
2230 so retry times are not inspected. This resulted in far too many retries
2231 for the affected addresses. The effect occurred only when there were more
2232 hosts than the hosts_max_try setting in the smtp transport when it had
2233 the 4xx errors. Those hosts that it had tried were not added to the list
2234 of hosts for which the message was waiting, so if all were tried, there
2235 was no problem. Two fixes have been applied:
2237 (i) If there are any address or message errors in an SMTP delivery, none
2238 of the hosts (tried or untried) are now added to the list of hosts
2239 for which the message is waiting, so the message should not be a
2240 candidate for sending over the same connection that was used for a
2241 successful delivery of some other message. This seems entirely
2242 reasonable: after all the message is NOT "waiting for some host".
2243 This is so "obvious" that I'm not sure why it wasn't done
2244 previously. Hope I haven't missed anything, but it can't do any
2245 harm, as the worst effect is to miss an optimization.
2247 (ii) If, despite (i), such a delivery is accidentally attempted, the
2248 routing retry time is respected, so at least it doesn't keep
2249 hammering the server.
2251 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2252 in ${readsocket because some servers need this prod.
2254 PH/32 Added some extra debug output when updating a wait-xxx database.
2256 PH/33 The hint "could be header name not terminated by colon", which has been
2257 given for certain expansion errors for a long time, was not being given
2258 for the ${if def:h_colon_omitted{... case.
2260 PH/34 The spec says: "With one important exception, whenever a domain list is
2261 being scanned, $domain contains the subject domain." There was at least
2262 one case where this was not true.
2264 PH/35 The error "getsockname() failed: connection reset by peer" was being
2265 written to the panic log as well as the main log, but it isn't really
2266 panic-worthy as it just means the connection died rather early on. I have
2267 removed the panic log writing for the ECONNRESET error when getsockname()
2270 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2271 runs only) independently of the message's sender address. This meant
2272 that, if the 4xx error was in fact related to the sender, a different
2273 message to the same recipient with a different sender could confuse
2274 things. In particualar, this can happen when sending to a greylisting
2275 server, but other circumstances could also provoke similar problems.
2276 I have changed the default so that the retry time for these errors is now
2277 based a combination of the sender and recipient addresses. This change
2278 can be overridden by setting address_retry_include_sender=false in the
2281 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2282 remote server are returned as part of bounce messages. This was not
2283 happening for LMTP over a pipe (the lmtp transport), but now it is the
2284 same for both kinds of LMTP.
2286 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2287 in header lines that were in fact CNAMEs. This is no longer the case.
2289 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2290 and queue runs started by the daemon processed all messages. This has
2291 been fixed so that -R and -S can now usefully be given with -q<time>.
2293 PH/40 Import PCRE release 6.7 (fixes some bugs).
2295 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2297 PH/42 Give an error if -q is specified more than once.
2299 PH/43 Renamed the variables $interface_address and $interface_port as
2300 $received_ip_address and $received_port, to make it clear that these
2301 values apply to message reception, and not to the outgoing interface when
2302 a message is delivered. (The old names remain recognized, of course.)
2304 PH/44 There was no timeout on the connect() call when using a Unix domain
2305 socket in the ${readsocket expansion. There now is.
2307 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2308 be meaningful with "accept".
2310 SC/01 Eximstats V1.43
2311 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2313 SC/02 Eximstats V1.44
2314 Use a glob alias rather than an array ref in the generated
2315 parser. This improves both readability and performance.
2317 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2318 Collect SpamAssassin and rejection statistics.
2319 Don't display local sender or destination tables unless
2320 there is data to show.
2321 Added average volumes into the top table text output.
2323 SC/04 Eximstats V1.46
2324 Collect data on the number of addresses (recipients)
2325 as well as the number of messages.
2327 SC/05 Eximstats V1.47
2328 Added 'Message too big' to the list of mail rejection
2329 reasons (thanks to Marco Gaiarin).
2331 SC/06 Eximstats V1.48
2332 Mainlog lines which have GMT offsets and are too short to
2333 have a flag are now skipped.
2335 SC/07 Eximstats V1.49 (Alain Williams)
2336 Added the -emptyok flag.
2338 SC/08 Eximstats V1.50
2339 Fixes for obtaining the IP address from reject messages.
2341 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2342 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2343 whitesspace changes from 4.64-PH/27
2345 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2348 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2349 are found, allow negative numbers in numeric criteria)
2351 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2353 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2354 to match changes made in 4.64-PH/43
2356 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2358 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2359 "domain" is actually an IP address, and force a failure. This locks out
2360 those revolvers/nameservers that support "A-for-A" lookups, in
2361 contravention of the specifications.
2363 PH/48 When a host name was looked up from an IP address, and the subsequent
2364 forward lookup of the name timed out, the host name was left in
2365 $sender_host_name, contrary to the specification.
2367 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2368 restricted to single-key lookups, Exim was not diagnosing an error if
2369 * or *@ was used with a query-style lookup.
2371 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2373 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2374 long ago, but noone interested enough thought of it. Let's just say that
2375 the "1.1" means that there are some new functions that weren't there at
2376 some point in the past.
2378 PH/51 Error processing for expansion failure of helo_data from an smtp
2379 transport during callout processing was broken.
2381 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2382 tested/used via the -bh/-bhc/-bs options.
2384 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2385 bug, fixed in subsequent PCRE releases).
2387 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2388 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2390 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2396 SC/01 Use a glob alias rather than an array ref in eximstats generated
2397 parser. This improves both readability and performance.
2399 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2400 Don't display local sender or destination tables in eximstats unless
2401 there is data to show.
2402 Added average volumes into the eximstats top table text output.
2404 SC/03 Collect data on the number of addresses (recipients) as well
2405 as the number of messages in eximstats.
2407 TF/01 Correct an error in the documentation for the redirect router. Exim
2408 does (usually) call initgroups() when daemonizing.
2410 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2411 with consistent privilege compared to when running as a daemon.
2413 TF/03 Note in the spec that $authenticated_id is not set for local
2414 submissions from trusted users.
2416 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2417 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2419 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2420 by adding some example configuration directives to the default
2421 configuration file. A little bit of work is required to uncomment the
2422 directives and define how usernames and passwords are checked, but
2423 there is now a framework to start from.
2425 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2426 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2427 without this. I don't know how relevant this is to other LDAP libraries.
2429 PH/02 Add the verb name to the "unknown ACL verb" error.
2431 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2433 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2435 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2436 directory not expanded when it should be if an expanded home directory
2437 was set for the address (which is overridden by the transport).
2439 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2442 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2443 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2444 because it is too late at that time, and has no effect.
2446 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2447 security issue with \' (bugzilla #107). I could not use the
2448 PQescapeStringConn() function, because it needs a PGconn value as one of
2451 PH/08 When testing addresses using -bt, indicate those final addresses that
2452 are duplicates that would not cause an additional delivery. At least one
2453 person was confused, thinking that -bt output corresponded to deliveries.
2454 (Suppressing duplicates isn't a good idea as you lose the information
2455 about possibly different redirections that led to the duplicates.)
2457 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2458 systems where poll() doesn't work, in particular OS X.
2460 PH/10 Added more information to debugging output for retry time not reached.
2462 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2463 operations in malware.c.
2465 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2468 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2469 syslog_duplication set false, log lines that would normally be written
2470 both the the main log and to the reject log were not written to syslog at
2473 PH/14 In the default configuration, change the use of "message" in ACL warn
2474 statements to "add_header".
2476 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2477 not followed by a command (e.g. "seen endif").
2479 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2480 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2483 PH/17 Added extra conditions to the default value of delay_warning_condition
2487 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2488 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2489 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2492 The Auto-Submitted: and various List- headers are standardised, whereas I
2493 don't think Precedence: ever was.
2495 PH/18 Refactored debugging code in route_finduser() to show more information,
2496 in particular, the error code if getpwnam() issues one.
2498 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2499 This is apparently needed in addition to the PH/07 change above to avoid
2500 any possible encoding problems.
2502 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2503 but not after initializing Perl.
2505 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2506 output them only if debugging. By default they are written stderr,
2507 apparently, which is not desirable.
2509 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2512 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2515 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2517 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2518 authenticated or an ident call has been made. Suppress the default
2519 values for $authenticated_id and $authenticated_sender (but permit -oMai
2520 and -oMas) when testing with -bh.
2522 PH/24 Re-jigged the order of the tests in the default configuration so that the
2523 tests for valid domains and recipients precede the DNS black list and CSA
2524 tests, on the grounds that those ones are more expensive.
2526 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2527 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2528 This bug exists in every version of Exim that I still have, right back to
2531 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2532 However, an attempt to turn on case-sensitivity in a regex key by
2533 including (?-i) didn't work because the subject string was already
2534 lowercased, and the effects were non-intuitive. It turns out that a
2535 one-line patch can be used to allow (?-i) to work as expected.
2541 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2542 other effects) broke the use of negated acl sub-conditions.
2544 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2547 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2548 "Deny" causes Exim to reject the incoming connection with a 554 error.
2549 Unfortunately, if there is a major crisis, such as a disk failure,
2550 tcp-wrappers gives "deny", whereas what one would like would be some
2551 kind of temporary error. A kludge has been added to help with this.
2552 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2553 554 error is used if errno is still zero or contains ENOENT (which occurs
2554 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2557 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2559 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2560 errors. Otherwise a message that provokes a temporary error (when other
2561 messages do not) can cause a whole host to time out.
2563 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2564 addresses were routed directly to files or pipes from a redirect router.
2565 File deliveries just didn't batch; pipe deliveries might have suffered
2568 PH/06 A failure to get a lock for a hints database would erroneously always say
2569 "Failed to get write lock", even when it was really a read lock.
2571 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2572 of 0600. This has been changed to use the value of the lockfile_mode
2573 option (which defaults to 0600).
2575 PH/08 Applied small patch from the Sieve maintainer.
2577 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2578 folder from quota calculations, a direct delivery into this folder messed
2579 up the contents of the maildirsize file. This was because the regex was
2580 used only to exclude .Trash (or whatever) when the size of the mailbox
2581 was calculated. There was no check that a delivery was happening into an
2582 excluded directory. This bug has been fixed by ignoring all quota
2583 processing for deliveries into excluded directories.
2585 PH/10 Added the maildirfolder_create_regex option to appendfile.
2591 PH/01 The code for finding all the local interface addresses on a FreeBSD
2592 system running IPv6 was broken. This may well have applied to all BSD
2593 systems, as well as to others that have similar system calls. The broken
2594 code found IPv4 interfaces correctly, but gave incorrect values for the
2595 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2596 that it would not match correctly against @[] and not recognize the IPv6
2599 PH/02 The ipliteral router was not recognizing addresses of the form user@
2600 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2602 PH/03 Added disable_ipv6.
2604 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2605 decoded form, because it is most often used to construct To: headers
2606 lines in autoreplies, and the decoded form may well be syntactically
2607 invalid. However, $reply_address has leading white space removed, and all
2608 newlines turned into spaces so that the autoreply transport does not
2611 PH/05 If group was specified without a user on a router, and no group or user
2612 was specified on a transport, the group from the router was ignored.
2614 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2615 for visible compile-time settings that can be used to change these
2616 numbers, for those that want even more. Backwards compatibility with old
2617 spool files has been maintained. However, going back to a previous Exim
2618 release will lost any variables that are in spool files.
2620 PH/07 Two small changes when running in the test harness: increase delay when
2621 passing a TCP/IP connection to a new process, in case the original
2622 process has to generate a bounce, and remove special handling of
2623 127.0.0.2 (sic), which is no longer necessary.
2625 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2626 be the same on different OS.
2628 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2631 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2632 whether --show-vars was specified or not
2634 JJ/02 exipick: Added support for new ACL variable spool format introduced
2637 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2638 syntactically invalid From: or Reply-to: line, and a filter used this to
2639 generate an autoreply, and therefore failed to obtain an address for the
2640 autoreply, Exim could try to deliver to a non-existent relative file
2641 name, causing unrelated and misleading errors. What now happens is that
2642 it logs this as a hard delivery error, but does not attempt to create a
2645 PH/11 The exinext utility has a -C option for testing purposes, but although
2646 the given file was scanned by exinext itself; it wasn't being passed on
2647 when Exim was called.
2649 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2650 an end-of-file indication when reading a command response.
2652 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2653 compiled. In many other places in Exim, IPv6 addresses are always
2654 recognized, so I have changed this. It also means that IPv4 domain
2655 literals of the form [IPV4:n.n.n.n] are now always recognized.
2657 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2658 used if the router is not running as root, for example, when verifying at
2659 ACL time, or when using -bh. The debugging output from this situation was
2660 non-existent - all you got was a failure to exec. I have made two
2663 (a) Failures to set uid/gid, the current directory, or a process leader
2664 in a subprocess such as that created by queryprogram now generate
2665 suitable debugging ouput when -d is set.
2667 (b) The queryprogram router detects when it is not running as root,
2668 outputs suitable debugging information if -d is set, and then runs
2669 the subprocess without attempting to change uid/gid.
2671 PH/15 Minor change to Makefile for building test_host (undocumented testing
2674 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2675 additional section of a DNS packet that returns MX or SRV records.
2676 Instead, it always explicitly searches for A/AAAA records. This avoids
2677 major problems that occur when a DNS server includes only records of one
2678 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2679 fixed another bug: if SRV records were looked up and the corresponding
2680 address records were *not* found in the additional section, the port
2681 values from the SRV records were lost.
2683 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2684 using the correct key (the original address) when searching the retry
2685 rules in order to find which one to use for generating the retry hint.
2687 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2688 adding the default one. Similarly, if it contains a Reply-To: header, the
2689 errors_reply_to option, if set, is not used.
2691 PH/19 When calculating a retry time, Exim used to measure the "time since
2692 failure" by looking at the "first failed" field in the retry record. Now
2693 it does not use this if it is later than than the arrival time of the
2694 message. Instead it uses the arrival time. This makes for better
2695 behaviour in cases where some deliveries succeed, thus re-setting the
2696 "first failed" field. An example is a quota failure for a huge message
2697 when small messages continue to be delivered. Without this change, the
2698 "time since failure" will always be short, possible causing more frequent
2699 delivery attempts for the huge message than are intended.
2700 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2702 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2703 $1, $2, $3) because the numerical variables can be reset during some
2704 expansion items (e.g. "match"), thereby losing the authentication data.
2706 PH/21 Make -bV show the size of off_t variables so that the test suite can
2707 decide whether to run tests for quotas > 2G.
2709 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2710 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2711 filecount value is greater than 2G or if a quota value is greater than 2G
2712 on a system where the size of off_t is not greater than 4, a panic error
2715 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2716 never match. The debug and -bh output now contains an explicit error
2717 message indicating a malformed IPv4 address or mask.
2719 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2720 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2721 PH/23 above applies.
2723 PH/25 Do not write to syslog when running in the test harness. The only
2724 occasion when this arises is a failure to open the main or panic logs
2725 (for which there is an explicit test).
2727 PH/26 Added the /no_tell option to "control=freeze".
2729 PH/27 If a host name lookup failed very early in a connection, for example, if
2730 the IP address matched host_lookup and the reverse lookup yielded a name
2731 that did not have a forward lookup, an error message of the form "no IP
2732 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2733 could be logged. Now it outputs the IP address instead of "NULL".
2735 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2736 allows the sender and the authenticated sender to be set when
2737 submitting a message from within Exim. Since child_open_exim() is
2738 documented for local_scan(), the new function should be too.
2740 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2741 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2742 results in an empty string is now treated as unset.
2744 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2746 PH/31 Added sender_verify_fail logging option.
2748 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2749 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2750 tidied the source and removed it altogether.
2752 PH/33 When a queue run was abandoned because the load average was too high, a
2753 log line was always written; now it is written only if the queue_run log
2754 selector is set. In addition, the log line for abandonment now contains
2755 information about the queue run such as the pid. This is always present
2756 in "start" and "stop" lines but was omitted from the "abandon" line.
2758 PH/34 Omit spaces between a header name and the colon in the error message that
2759 is given when verify = headers_syntax fails (if there are lots of them,
2760 the message gets confusing).
2762 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2763 names, as there are now some PTR records that contain slashes. This check
2764 is only to protect against broken name servers that fall over on strange
2765 characters, so the fact that it applies to all lookups doesn't matter.
2767 PH/36 Now that the new test suite is complete, we can remove some of the
2768 special code in Exim that was needed for the old test suite. For example,
2769 sorting DNS records because real resolvers return them in an arbitrary
2770 order. The new test suite's fake resolver always returns records in the
2773 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2774 bounces) except when queue_only is set, to avoid logging races between
2775 the different processes.
2777 PH/38 Panic-die if .include specifies a non-absolute path.
2779 PH/39 A tweak to the "H" retry rule from its user.
2781 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2782 a label. They prevented compilation on older perls.
2784 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2785 a warning to be raised on newish perls.
2787 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2788 on queue. Changes to match documented behaviour of showing count of
2789 messages matching specified criteria.
2791 PH/40 Changed the default ident timeout from 30s to 5s.
2793 PH/41 Added support for the use of login_cap features, on those BSD systems
2794 that have them, for controlling the resources used by pipe deliveries.
2796 PH/42 The content-scanning code uses fopen() to create files in which to put
2797 message data. Previously it was not paying any attention to the mode of
2798 the files. Exim runs with umask(0) because the rest of the code creates
2799 files with open(), and sets the required mode explicitly. Thus, these
2800 files were ending up world-writeable. This was not a big issue, because,
2801 being within the spool directory, they were not world-accessible. I have
2802 created a function called modefopen, which takes an additional mode
2803 argument. It sets umask(777), creates the file, chmods it to the required
2804 mode, then resets the umask. All the relevant calls to fopen() in the
2805 content scanning code have been changed to use this function.
2807 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2808 to 24 hours. This avoids potential overflow problems when processing G
2809 and H retry rules. I suspect nobody ever tinkers with this value.
2811 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2813 PH/45 When the plaintext authenticator is running as a client, the server's
2814 challenges are checked to ensure they are valid base64 strings. By
2815 default, the authentication attempt is cancelled if an invalid string is
2816 received. Setting client_ignore_invalid_base64 true ignores these errors.
2817 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2818 they are received. Thus, the responses can be made to depend on the
2819 challenges. If an invalid string is ignored, an empty string is placed in
2822 PH/46 Messages that are created by the autoreply transport now contains a
2823 References: header, in accordance with RFCs 2822 and 3834.
2825 PH/47 Added authenticated_sender_force to the smtp transport.
2827 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2829 PH/49 Installed latest patch from the Sieve maintainer.
2831 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2832 was also set, the appendfile transport was unnecessarily scanning a
2833 directory of message files (e.g. for maildir delivery) to find the count
2834 of files (along with the size), even though it did not need this
2835 information. It now does the scan only if it needs to find either the
2836 size of the count of files.
2838 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2840 PH/52 Two bugs concerned with error handling when the smtp transport is
2843 (i) Exim was not creating retry information for temporary errors given
2844 for individual recipients after the DATA command when the smtp transport
2845 was used in LMTP mode. This meant that they could be retried too
2846 frequently, and not timed out correctly.
2848 (ii) Exim was setting the flag that allows error details to be returned
2849 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2850 recipients that were returned after the DATA command.
2852 PH/53 This is related to PH/52, but is more general: for any failing address,
2853 when detailed error information was permitted to be returned to the
2854 sender, but the error was temporary, then after the final timeout, only
2855 "retry timeout exceeded" was returned. Now it returns the full error as
2856 well as "retry timeout exceeded".
2858 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2859 do this, and (what is worse) MTAs that accept it.
2861 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2862 will now be deprecated.
2864 PH/56 New os.c-cygwin from the Cygwin maintainer.
2866 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2867 formats (previously only available in exim formats via -bpr, -bpru,
2868 and -bpra. Now also available in native and exiqgrep formats)
2870 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2871 with very large, slow to parse queues
2873 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2875 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2877 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2878 responses to authentication challenges, though it was showing the
2879 challenges; (ii) I've removed the CR characters from the debug output for
2882 PH/58 Allow for the insertion of a newline as well as a space when a string
2883 is turned into more than one encoded-word during RFC 2047 encoding. The
2884 Sieve code now uses this.
2886 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2887 data_4xx, lost_connection, tls_required.
2889 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2890 message was being sent as an SMTP response.
2892 PH/61 Add -l and -k options to exicyclog.
2894 PH/62 When verifying, if an address was redirected to one new address, so that
2895 verification continued, and the new address failed or deferred after
2896 having set something in $address_data, the value of $address_data was not
2897 passed back to the ACL. This was different to the case when no
2898 redirection occurred. The value is now passed back in both cases.
2900 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2901 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2902 use login_cap.h, so on its own it isn't the distinguishing feature. The
2903 new name refers directly to the setclassresources() function.
2905 PH/65 Added configuration files for NetBSD3.
2907 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2909 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2910 is preferred over IPv4.
2912 PH/68 The bounce_return_message and bounce_return_body options were not being
2913 honoured for bounces generated during the reception of non-SMTP messages.
2914 In particular, this applied to messages rejected by the ACL. This bug has
2915 been fixed. However, if bounce_return_message is true and bounce_return_
2916 body is false, the headers that are returned for a non-SMTP message
2917 include only those that have been read before the error was detected.
2918 (In the case of an ACL rejection, they have all been read.)
2920 PH/69 The HTML version of the specification is now built in a directory called
2921 spec_html instead of spec.html, because the latter looks like a path with
2922 a MIME-type, and this confuses some software.
2924 PH/70 Catch two compiler warnings in sieve.c.
2926 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2927 function verify_get_ident() calls ip_connect() to connect a socket, but
2928 if the "connect()" function timed out, ip_connect() used to close the
2929 socket. However, verify_get_ident() also closes the socket later, and in
2930 between Exim writes to the log, which may get opened at this point. When
2931 the socket was closed in ip_connect(), the log could get the same file
2932 descriptor number as the socket. This naturally causes chaos. The fix is
2933 not to close the socket in ip_connect(); the socket should be closed by
2934 the function that creates it. There was only one place in the code where
2935 this was missing, in the iplookup router, which I don't think anybody now
2936 uses, but I've fixed it anyway.
2938 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2939 well as to direct DNS lookups. Otherwise the handling of names in host
2940 lists is inconsistent and therefore confusing.
2946 PH/01 Two changes to the default runtime configuration:
2948 (1) Move the checks for relay_from_hosts and authenticated clients from
2949 after to before the (commented out) DNS black list checks.
2951 (2) Add control=submission to the relay_from_hosts and authenticated
2952 clients checks, on the grounds that messages accepted by these
2953 statements are most likely to be submissions.
2955 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2957 (1) Generate an error if the third argument for the ${prvs expansion is
2960 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2963 (3) Reset the variables that are obtained from the first argument of
2964 ${prvscheck and used in the second argument before leaving the code,
2965 because their memory is reclaimed, so using them afterwards may do
2968 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2969 one (it's much easier than Tom thought :-).
2971 (5) Because of (4), we can now allow for the use of $prvscheck_result
2972 inside the third argument.
2974 PH/03 For some reason, the default setting of PATH when running a command from
2975 a pipe transport was just "/usr/bin". I have changed it to
2978 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2979 anything to be listed in the output from -bV.
2981 PH/05 When a filter generated an autoreply, the entire To: header line was
2982 quoted in the delivery log line, like this:
2984 => >A.N.Other <ano@some.domain> <original@ddress> ...
2986 This has been changed so that it extracts the operative address. There
2987 may be more than one such address. If so, they are comma-separated, like
2990 => >ano@some.domain,ona@other.domain <original@ddress> ...
2992 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2993 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2994 not being looked up in the rDNS to get a host name, Exim was showing the
2995 IP address twice in Received: lines, even though the IP addresses were
2996 identical. For example:
2998 Received: from [1.2.3.4] (helo=[1.2.3.4])
3000 However, if the real host name was known, it was omitting the HELO data
3001 if it matched the actual IP address. This has been tidied up so that it
3002 doesn't show the same IP address twice.
3004 PH/07 When both +timestamp and +memory debugging was on, the value given by
3005 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3006 called by the debug printing, thereby overwriting the timestamp buffer.
3007 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3009 PH/08 When the original message was included in an autoreply transport, it
3010 always said "this is a copy of the message, including all the headers",
3011 even if body_only or headers_only was set. It now gives an appropriate
3014 PH/09 Applied a patch from the Sieve maintainer which:
3016 o fixes some comments
3017 o adds the (disabled) notify extension core
3018 o adds some debug output for the result of if/elsif tests
3019 o points to the current vacation draft in the documentation
3020 and documents the missing references header update
3024 o fixes a bug in processing the envelope test (when testing
3025 multiple envelope elements, the last element determinted the
3028 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3029 Electronic Mail") by including:
3031 Auto-submitted: auto-generated
3033 in the messages that it generates (bounce messages and others, such as
3034 warnings). In the case of bounce messages for non-SMTP mesages, there was
3035 also a typo: it was using "Auto_submitted" (underscore instead of
3036 hyphen). Since every message generated by Exim is necessarily in response
3037 to another message, thes have all been changed to:
3039 Auto-Submitted: auto-replied
3041 in accordance with these statements in the RFC:
3043 The auto-replied keyword:
3045 - SHOULD be used on messages sent in direct response to another
3046 message by an automatic process,
3048 - MUST NOT be used on manually-generated messages,
3050 - MAY be used on Delivery Status Notifications (DSNs) and Message
3051 Disposition Notifications (MDNs),
3053 - MUST NOT be used on messages generated by automatic or periodic
3054 processes, except for messages which are automatic responses to
3057 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3058 to the default Received: header definition.
3060 PH/12 Added log selector acl_warn_skipped (default on).
3062 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3063 variables because (a) they are in the wrong storage pool and (b) even if
3064 they were copied, it wouldn't work properly because of the caching.
3066 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3067 checking when decoding. Apparently there are clients that generate
3068 overlong encoded strings. Why am I not surprised?
3070 PH/15 If the first argument of "${if match_address" was not empty, but did not
3071 contain an "@" character, Exim crashed. Now it writes a panic log message
3072 and treats the condition as false.
3074 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3076 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3077 extension "envelope-auth". The code is finished and in agreement with
3078 other implementations, but there is no documentation so far and in fact,
3079 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3080 not changing the active code.
3082 Print executed "if" and "elsif" statements when debugging is used. This
3083 helps a great deal to understand what a filter does.
3085 Document more things not specified clearly in RFC3028. I had all this
3086 sorted out, when out of a sudden new issues came to my mind. Oops."
3088 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3091 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3092 canonical form (as documented). However, after a host name lookup from
3093 the IP address, check_host() was doing a simple string comparison with
3094 addresses acquired from the DNS when checking that the found name did
3095 have the original IP as one of its addresses. Since any found IPv6
3096 addresses are likely to be in abbreviated form, the comparison could
3097 fail. Luckily, there already exists a function for doing the comparison
3098 by converting both addresses to binary, so now that is used instead of
3099 the text comparison.
3101 PH/20 There was another similar case to PH/19, when a complete host name was
3102 given in a host list; looking up its IP address could give an abbreviated
3103 form, whereas the current host's name might or might not be abbreviated.
3104 The same fix has been applied.
3110 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3111 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3114 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3115 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3117 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3119 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3120 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3121 Also, for "Auto-Submitted", treat anything other than "no" as
3122 non-personal, in accordance with RFC 3834. (Previously it treated
3123 anything starting "auto-" as non-personal.)
3125 TF/01 The control=submission/name=... option had a problem with syntax
3126 errors if the name included a slash character. The /name= option
3127 now slurps the rest of the string, so it can include any characters
3128 but it must come last in the list of options (after /sender_retain
3131 PH/05 Some modifications to the interface to the fake nameserver for the new
3139 TK/01 Added the "success_on_redirect" address verification option. See
3140 NewStuff for rationale and an example.
3142 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3144 PH/02 Patch to exigrep to allow it to work on syslog lines.
3146 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3147 fread() to skip over the body file's header line, because in Cygwin the
3148 header line is locked and is inaccessible.
3150 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3151 co-exist for some time) to make it clear that it is the Exim ID that is
3152 referenced, not the Message-ID: header line.
3154 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3155 string_format() function, because snprintf() does not exist on all
3158 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3159 ${stat: expansion item.
3161 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3162 protocol synchronization error", to keep the pedants happy.
3164 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3165 well as for IRIX systems, when gcc is being used. See the host.c source
3168 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3170 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3173 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3174 transport and to the smtp transport in LMTP mode.
3176 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3178 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3179 run a filter in a subprocess. This could lead to confusion in subsequent
3180 lookups in the parent process. There should also be a search_tidyup() at
3181 the end of the subprocess.
3183 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3184 only if the host matched helo_try_verify_hosts, which caused the
3185 verification to occur when the EHLO/HELO command was issued. The ACL just
3186 tested the remembered result. Now, if a previous verification attempt has
3187 not happened, "verify = helo" does it there and then.
3189 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3191 TK/03 Fix log output including CR from clamd.
3193 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3194 exist provoked a memory error which could cause a segfault.
3196 PH/15 Installed PCRE 6.2
3198 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3200 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3201 of the problem. Specifically, suggested +O2 rather than +O1 for the
3204 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3206 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3207 record, the port was not correctly logged when the outgoing_port log
3208 selector was set (it logged the transort's default port).
3210 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3211 fallback_hosts, and "hosts" in the smtp transport.
3213 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3214 host errors such as "Connection refused".
3216 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3217 authentication with radiusclient 0.4.9:
3219 - Error returned from rc_read_config was caught wrongly
3220 - Username/password not passed on to radius server due to wrong length.
3222 The presumption is that some radiusclient API changes for 4.51/PH/17
3223 were not taken care of correctly. The code is still untested by me (my
3224 Linux distribution still has 0.3.2 of radiusclient), but it was
3225 contributed by a Radius user.
3227 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3228 expanding the "port" option of the smtp transport.
3230 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3231 while reading a MIME header. Thanks to Tom Hughes for a patch.
3233 PH/24 Include config.h inside local_scan.h so that configuration settings are
3236 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3237 that in an ACL for RCPT (for example), you can examine exactly what was
3240 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3241 commands, but it was not correctly comparing the address with the actual
3242 client host address. Thus, it would show the EHLO address in Received:
3243 header lines when this was not necessary.
3245 PH/27 Added the % operator to ${eval:}.
3247 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3248 it should be ignoring failures (because with -C, for example, it has lost
3249 privilege). It wasn't ignoring creation failures other than "already
3252 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3255 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3256 because an input line was too long, either on its own, or by virtue of
3257 too many continuations, the temporary file was not being removed, and the
3258 return code was incorrect.
3260 PH/31 Missing "BOOL" in function definition in filtertest.c.
3262 PH/32 Applied Sieve patches from the maintainer.
3264 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3266 PH/33 Added "verify = not_blind".
3268 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3269 Local/Makefile (with some defaults set). These are used in built scripts
3270 such as exicyclog, but they have never been used in the exim_install
3271 script (though there are many overriding facilities there). I have
3272 arranged that the exim_install script now takes note of these two
3275 PH/35 Installed configuration files for Dragonfly.
3277 PH/36 When a locally submitted message by a trusted user did not contain a
3278 From: header, and the sender address was obtained from -f or from an SMTP
3279 MAIL command, and the trusted user did not use -F to supply a sender
3280 name, $originator_name was incorrectly used when constructing a From:
3281 header. Furthermore, $originator_name was used for submission mode
3282 messages from external hosts without From: headers in a similar way,
3283 which is clearly wrong.
3285 PH/37 Added control=suppress_local_fixups.
3287 PH/38 When log_selector = +received_sender was set, and the addition of the
3288 sender made the log line's construction buffer exactly full, or one byte
3289 less than full, an overflow happened when the terminating "\n" was
3292 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3293 when the result of a list match is failure because a DNS lookup failed.
3295 PH/40 RM_COMMAND is now used in the building process.
3297 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3298 the "build-* directories that it finds.
3300 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3301 address in a domain literal was a prefix of an interface address.
3303 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3304 when verifying a sender address, unless rewrite_headers is false.
3306 PH/44 Wrote a long comment about why errors_to addresses are verified as
3307 recipients, not senders.
3309 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3310 the ratelimit ACL was added.
3312 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3314 PH/46 Added extra information about PostgreSQL errors to the error string.
3316 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3317 avoiding the need to install special zones in a real server. This is
3318 backwards compatible; if it can't find the fake resolver, it drops back.
3319 Thus, both old and new test suites can be run.
3321 TF/02 Added util/ratelimit.pl
3323 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3324 clock is set back in time.
3326 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3327 Candler <B.Candler@pobox.com>.
3329 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3330 OK for addresses that are the result of redirections.
3332 PH/48 A number of further additions for the benefit of the new test suite,
3333 including a fake gethostbyname() that interfaces to the fake DNS resolver
3336 TF/06 The fix for widen_domains has also been applied to qualify_single and
3337 search_parents which are the other dnslookup options that can cause
3340 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3343 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3345 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3346 service" if it encountered an MX record with an empty target hostname.
3347 The message is now "an MX or SRV record indicated no SMTP service".
3349 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3350 if the DNS of the sending site is misconfigured. This is quite a
3351 common situation. This change restores the behaviour of treating a
3352 helo verification defer as a failure.
3354 PH/51 If self=fail was set on a router, the bounce message did not include the
3355 actual error message.
3361 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3363 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3364 command itself ended in error, the underlying message about the transport
3365 filter timeout was being overwritten with the pipe command error. Now the
3366 underlying error message should be appended to the second error message.
3368 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3370 PH/02 Reduce the amount of output that "make" produces by default. Full output
3371 can still be requested.
3373 PH/03 The warning log line about a condition test deferring for a "warn" verb
3374 was being output only once per connection, rather than after each
3375 occurrence (because it was using the same function as for successful
3376 "warn" verbs). This seems wrong, so I have changed it.
3378 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3379 it should not have, which might have caused a crash in the right
3380 circumstances, but probably never did.
3382 PH/04 Installed a modified version of Tony Finch's patch to make submission
3383 mode fix the return path as well as the Sender: header line, and to
3384 add a /name= option so that you can make the user's friendly name appear
3387 TF/03 Added the control = fakedefer ACL modifier.
3389 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3390 Mark Lowes for thorough testing.
3392 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3394 TK/03 Merged latest SRS patch from Miles Wilton.
3396 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3397 with the definition in sysexits.h (which is #included earlier).
3398 Fortunately, Exim does not actually use EX_OK. The code used to try to
3399 preserve the sysexits.h value, by assumimg that macro definitions were
3400 scanned for macro replacements. I have been disabused of this notion,
3401 so now the code just undefines EX_OK before #including unistd.h.
3403 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3404 in the smtp transport. When a block could not be written in a single
3405 write() function, the timeout was being re-applied to each part-write.
3406 This seems wrong - if the receiver was accepting one byte at a time it
3407 would take for ever. The timeout is now adjusted when this happens. It
3408 doesn't have to be particularly precise.
3410 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3411 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3413 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3414 without a domain if the check to <postmaster@domain> fails.
3416 SC/01 Eximstats: added -xls and the ability to specify output files
3417 (patch written by Frank Heydlauf).
3419 SC/02 Eximstats: use FileHandles for outputing results.
3421 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3423 SC/04 Eximstats: fixed display of large numbers with -nvr option
3425 SC/05 Eximstats: fixed merging of reports with empty tables.
3427 SC/06 Eximstats: added the -include_original_destination flag
3429 SC/07 Eximstats: removed tabs and trailing whitespace.
3431 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3433 TK/06 MBOX spool code: Add real "From " MBOX separator line
3434 so the .eml file is really in mbox format (even though
3435 most programs do not really care). Patch from Alex Miller.
3437 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3438 The latter is generated from $received_to and is only set if the
3439 message has one envelope recipient. SA can use these headers,
3440 obviously out-of-the-box. Patch from Alex Miller.
3442 PH/08 The ${def test on a variable was returning false if the variable's
3443 value was "0", contrary to what the specification has always said!
3444 The result should be true unless the variable is empty.
3446 PH/09 The syntax error of a character other than { following "${if
3447 def:variable_name" (after optional whitespace) was not being diagnosed.
3448 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3449 accidental colon was present, for example, could give incorrect results.
3451 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3452 result is used (not including appendfile, where other changes are about
3455 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3456 This involved changing a lot of size variables from int to off_t. It
3457 should work with maildirs and everything.
3459 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3460 spamd dying while we are connected to it.
3462 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3465 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3466 function for BDB 4.3.
3468 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3470 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3471 the sources for PCRE have been re-arranged and more files are now
3474 PH/15 The code I had for printing potentially long long variables in PH/11
3475 above was not the best (it lost precision). The length of off_t variables
3476 is now inspected at build time, and an appropriate printing format (%ld
3477 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3478 to be "long long int" or "long int". This is needed for the internal
3479 formatting function string_vformat().
3481 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3482 the configuration file to be ":syslog", then the script "guesses" where
3483 the logs files are, rather than using the compiled in default. In our
3484 case the guess is not the same as the compiled default, so the script
3485 suddenly stopped working when I started to use syslog. The patch checks
3486 to see if log_file_path is "". If so, it attempts to read it from exim
3487 with no configuration file to get the compiled in version, before it
3488 falls back to the previous guessing code."
3490 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3491 implementing BATV in an Exim configuration. See NewStuff for the gory
3494 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3495 Makefile that are specific to HP-UX.
3497 PH/18 If the "use_postmaster" option was set for a recipient callout together
3498 with the "random" option, the postmaster address was used as the MAIL
3499 FROM address for the random test, but not for the subsequent recipient
3500 test. It is now used for both.
3502 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3503 patch removes a few documentation additions to RFC 3028, because the
3504 latest draft now contains them. It adds the new en;ascii-case comparator
3505 and a new error check for 8bit text in MIME parts. Comparator and
3506 require names are now matched exactly. I enabled the subaddress
3507 extension, but it is not well tested yet (read: it works for me)."
3509 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3510 rework some of the code of TK/09 above to avoid the hardwired use of
3511 "%lld" and "long long". Replaced the call to snprintf() with a call to
3514 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3515 records point to non-existent hosts", "retry timeout exceeded", and
3516 "retry time not reached for any host after a long failure period".
3518 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3519 experimental DomainKeys support:
3521 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3522 (2) On an error such as an illegally used "control", the wrong name for
3523 the control was given.
3525 These problems did NOT occur unless DomainKeys support was compiled.
3527 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3529 PH/24 Added ${if match_ip condition.
3531 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3532 databases so that it will be absolutely obvious if a crash occurs in the
3533 DB library. This is a regular occurrence (often caused by mis-matched
3536 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3537 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3538 picked up on a user's system that detects such things. There doesn't seem
3539 to be a gcc warning option for this - only an attribute that has to be
3540 put on the function's prototype. It seems that in Fedora Core 4 they have
3541 set this on a number of new functions. No doubt there will be more in due
3544 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3545 specify a transport. However, if an address that was verified by such a
3546 router was the subject of a callout, Exim crashed because it tried to
3547 read the rcpt_include_affixes from the non-existent transport. Now it
3548 just assumes that the setting of that option is false. This bug was
3549 introduced by 4.51/PH/31.
3551 PH/28 Changed -d+all to exclude +memory, because that information is very
3552 rarely of interest, but it makes the output a lot bigger. People tend to
3553 do -d+all out of habit.
3555 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3556 code in os-type was giving problems when libc.so lives in lib64, like on
3559 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3560 aren't the modern standard, and it seems that some systems' include files
3561 don't always have them. Exim was already checking for some of the newer
3562 ones like T_AAAA, and defining it itself. I've added checks for all the
3563 record types that Exim uses.
3565 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3566 not automatically generating a new one, as it is supposed to. This
3567 prevented TLS from working. If the file did exist, but contained invalid
3568 data, a new version was generated, as expected. It was only the case of a
3569 non-existent file that was broken.
3571 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3572 with a change in libdomainkeys > 0.64.
3574 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3575 from DNS. If the selector record carries the flag, it now has
3576 precedence over the domain-wide flag.
3578 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3580 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3581 the use of a port name that isn't defined in /etc/services) occurred, the
3582 message was deferred as in a normal delivery, and thus remained on the
3583 spool, instead of being failed because of the mua_wrapper setting. This
3584 is now fixed, and I tidied up some of the mua_wrapper messages at the
3587 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3588 the messages in a hash of arrays rather than using individual hashes.
3589 This is a bit cleaner and results in dramatic memory savings, albeit
3590 at a slight CPU cost.
3592 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3593 as requested by Marc Sherman.
3595 SC/10 Eximstats: added histograms for user specified patterns as requested
3598 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3600 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3601 fopen() in the content-scanning modules that did not already have it.
3607 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3608 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3610 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3612 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3614 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3615 to test Sieve filters that use "vacation".
3617 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3618 that changes the way the GnuTLS parameters are stored in the cache file.
3619 The new format can be generated externally. For backward compatibility,
3620 if the data in the cache doesn't make sense, Exim assumes it has read an
3621 old-format file, and it generates new data and writes a new file. This
3622 means that you can't go back to an older release without removing the
3625 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3626 work if there are any delivery delays because "one_time" forces the
3627 parent to be marked "delivered", so its unseen clone is never tried
3628 again. For this reason, Exim now forbids the simultaneous setting of
3631 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3632 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3633 entry if you want to know the details. The fix, however, neglected to
3634 consider the case when local delivery batching is involved. The test for
3635 "previously delivered" was not happening when checking to see if an
3636 address could be batched with a previous (undelivered) one; under
3637 certain circumstances this could lead to multiple deliveries to the same
3640 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3641 in its include files, and this causes problems building Exim.
3643 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3644 header_syntax) but Exim was just ignoring anything given after a slash.
3645 In particular, this caused confusion with an attempt to use "verify =
3646 reverse_host_lookup/defer_ok". An error is now given when options are
3647 supplied for verify items that do not have them. (Maybe reverse_host_
3648 lookup should have a defer_ok option, but that's a different point.)
3650 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3651 defined by RFC 821) to 2048, because there were problems with some AUTH
3652 commands, and RFC 1869 says the size should be increased for extended
3653 SMTP commands that take arguments.
3655 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3658 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3659 "unknown" error; now it says that the functionality isn't in the binary.
3661 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3662 an address' error message when a string expansion fails (syntax or
3663 whatever). Otherwise the password may appear in the log. Following change
3664 PH/42 below, there is no longer a chance of it appearing in a bounce
3667 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3669 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3670 of its addresses was ever tried. (Bugzilla bug #2.)
3672 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3673 the result incorrectly in the debug output. (It correctly added a newline
3674 to what was transported.)
3676 TF/01 Added $received_time.
3678 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3679 commented out examples of how to interface to a virus scanner and to
3680 SpamAssassin. Also added commented examples of av_scanner and
3681 spamd_address settings.
3683 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3684 and controls are allowed in which ACLs. There were a couple of minor
3685 errors. Some of the entries in the conditions table (which is a table of
3686 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3687 negation of where the condition IS allowed.
3689 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3691 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3692 header file does not have a version number, so I've had to invent a new
3693 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3694 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3695 radiusclient), but it was contributed by a Radius user.
3697 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3698 files or directories, for OpenSSL.
3700 PH/19 When an Exim process that is running as root has to create an Exim log
3701 file, it does so in a subprocess that runs as exim:exim so as to get the
3702 ownership right at creation (otherwise, other Exim processes might see
3703 the file with the wrong ownership). There was no test for failure of this
3704 fork() call, which would lead to the process getting stuck as it waited
3705 for a non-existent subprocess. Forks do occasionally fail when resources
3706 run out. I reviewed all the other calls to fork(); they all seem to check
3709 PH/20 When checking for unexpected SMTP input at connect time (before writing
3710 the banner), Exim was not dealing correctly with a non-positive return
3711 from the read() function. If the client had disconnected by this time,
3712 the result was a log entry for a synchronization error with an empty
3713 string after "input=" when read() returned zero. If read() returned -1
3714 (an event I could not check), uninitialized data bytes were printed.
3715 There were reports of junk text (parts of files, etc) appearing after
3718 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3720 PH/22 Added support for macro redefinition, and (re)definition in between
3721 driver and ACL definitions.
3723 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3724 forgetting to use the resulting value; it was using the unexpanded value.
3726 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3727 hadn't been configured. The fix is from Juergen Kreileder, who
3728 understands it better than I do:
3730 "Here's what I see happening with three configured cyrus_sasl
3731 authenticators configured (plain, login, cram-md5):
3733 On startup auth_cyrus_sasl_init() gets called for each of these.
3734 This means three calls to sasl_listmech() without a specified mech_list.
3735 => SASL tests which mechs of all available mechs actually work
3736 => three warnings about OTP not working
3737 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3739 With the patch, sasl_listmech() also gets called three times. But now
3740 SASL's mech_list option is set to the server_mech specified in the the
3741 authenticator. Or in other words, the answer from sasl_listmech()
3742 gets limited to just the mech you're testing for (which is different
3744 => the return list contains just 'plain' or 'login', 'cram-md5' or
3745 nothing depending on the value of ob->server_mech.
3747 I've just tested the patch: Authentication still works fine,
3748 unavailable mechs specified in the exim configuration are still
3749 caught, and the auth.log warnings about OTP are gone."
3751 PH/25 When debugging is enabled, the contents of the command line are added
3752 to the debugging output, even when log_selector=+arguments is not
3755 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3756 answer is "GNU", and only if the return is "GNU/something" is the answer
3759 PH/27 $acl_verify_message is now set immediately after the failure of a
3760 verification in an ACL, and so is available in subsequent modifiers. In
3761 particular, the message can be preserved by coding like this:
3763 warn !verify = sender
3764 set acl_m0 = $acl_verify_message
3766 Previously, $acl_verify_message was set only while expanding "message"
3767 and "log_message" when a very denied access.
3769 PH/28 Modified OS/os.c-Linux with
3771 -#ifndef OS_LOAD_AVERAGE
3772 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3774 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3775 nomenclature these days.)
3777 PH/29 Installed patch from the Sieve maintainer that adds the options
3778 sieve_useraddress and sieve_subaddress to the redirect router.
3780 PH/30 In these circumstances:
3781 . Two addresses routed to the same list of hosts;
3782 . First host does not offer TLS;
3783 . First host accepts first address;
3784 . First host gives temporary error to second address;
3785 . Second host offers TLS and a TLS session is established;
3786 . Second host accepts second address.
3787 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3788 and peerdn, if requested) that were in fact used only for the second
3791 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3792 attention to any local part prefix or suffix that was matched by the
3793 router that accepted the address. It now behaves in the same way as it
3794 does for delivery: the affixes are removed from the local part unless
3795 rcpt_include_affixes is set on the transport.
3797 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3798 timeout during the DATA phase of an incoming message.
3800 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3801 applied a patch sanctioned by the Sieve maintainer.
3803 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3804 the uid or gid is negative. A case of a negative gid caused this to be
3805 noticed. The fix allows for either to be negative.
3807 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3808 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3811 PH/36 The change PH/12 above was broken. Fixed it.
3813 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3814 the grounds that routing the same address twice would always produce the
3815 same answer. This might have been true once, but it is certainly no
3816 longer true now. Routing a child address may depend on the previous
3817 routing that produced that child. Some complicated redirection strategies
3818 went wrong when messages had multiple recipients, and made Exim's
3819 behaviour dependent on the order in which the addresses were given.
3821 I have moved the duplicate checking until after the routing is complete.
3822 Exim scans the addresses that are assigned to local and remote
3823 transports, and removes any duplicates. This means that more work will be
3824 done, as duplicates will always all be routed, but duplicates are
3825 presumably rare, so I don't expect this is of any significance.
3827 For deliveries to pipes, files, and autoreplies, the duplicate checking
3828 still happens during the routing process, since they are not going to be
3831 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3832 It corrects a timeout issue with spamd. This is Ian's comment: "The
3833 background is that sometimes spamd either never reads data from a
3834 connection it has accepted, or it never writes response data. The exiscan
3835 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3836 blindly assumes that writes won't block so it may never time out."
3838 PH/39 Allow G after quota size as well as K and M.
3840 PH/40 The value set for $authenticated_id in an authenticator may not contain
3841 binary zeroes or newlines because the value is written to log lines and
3842 to spool files. There was no check on this. Now the value is run through
3843 the string_printing() function so that such characters are converted to
3844 printable escape sequences.
3846 PH/41 $message_linecount is a new variable that contains the total number of
3847 lines in the message. Compare $body_linecount, which is the count for the
3850 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3851 bounce and delay warning messages, except in certain special cases, which
3854 (a) An SMTP error message from a remote host;
3855 (b) A message specified in a :fail: redirection;
3856 (c) A message specified in a "fail" command in a system filter;
3857 (d) A message specified in a FAIL return from the queryprogram router;
3858 (e) A message specified by the cannot_route_message router option.
3860 In these cases only, Exim does include the error details in bounce and
3861 warning messages. There are also a few cases where bland messages such
3862 as "unrouteable address" or "local delivery error" are given.
3864 PH/43 $value is now also set for the "else" part of a ${run expansion.
3866 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3867 being worked on, but at least Exim now implements the latest version to
3870 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3871 process to complete was treated as a delivery failure, a timeout while
3872 writing the message to the pipe was logged, but erroneously treated as a
3873 successful delivery. Such timeouts include transport filter timeouts. For
3874 consistency with the overall process timeout, these timeouts are now
3875 treated as errors, giving rise to delivery failures by default. However,
3876 there is now a new Boolean option for the pipe transport called
3877 timeout_defer, which, if set TRUE, converts the failures into defers for
3878 both kinds of timeout. A transport filter timeout is now identified in
3881 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3882 systems where "make" and "gmake" are different, calling "gmake" at top
3883 level broke things. I've arranged for the value of $(MAKE) to be passed
3884 from the Makefile to this script so that it can call the same version of
3888 A note about Exim versions 4.44 and 4.50
3889 ----------------------------------------
3891 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3892 changes of various kinds. As a consequence, a big documentation update was
3893 needed. This delayed the release for rather longer than seemed good, especially
3894 in the light of a couple of (minor) security issues. Therefore, the changes
3895 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3896 release. So 4.44 and 4.50 are in effect two different branches that both start
3899 I have left the 4.50 change log unchanged; it contains all the changes since
3900 4.43. The change log for 4.44 is below; many of its items are identical to
3901 those for 4.50. This seems to be the most sensible way to preserve the
3902 historical information.
3908 1. Minor wording change to the doc/README.SIEVE file.
3910 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3911 computation of the current number of files was incorrect.
3913 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3914 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3917 4. Give more explanation in the error message when the command for a transport
3918 filter fails to execute.
3920 5. There are several places where Exim runs a non-Exim command in a
3921 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3922 was being done only for the command run by the queryprogram router. It is
3923 now done for all such subprocesses. The other cases are: ${run, transport
3924 filters, and the commands run by the lmtp and pipe transports.
3926 6. Added CONFIGURE_GROUP build-time option.
3928 7. Some older OS have a limit of 256 on the maximum number of file
3929 descriptors. Exim was using setrlimit() to set 1000 as a large value
3930 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3931 systems. I've change it so that if it can't get 1000, it tries for 256.
3933 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3934 was an oversight, and furthermore, ever since the addition of extra
3935 controls (e.g. 4.43/32), the checks on when to allow different forms of
3936 "control" were broken. There should now be diagnostics for all cases when a
3937 control that does not make sense is encountered.
3939 9. Added the /retain_sender option to "control=submission".
3941 10. $recipients is now available in the predata ACL (oversight).
3943 11. Tidy the search cache before the fork to do a delivery from a message
3944 received from the command line. Otherwise the child will trigger a lookup
3945 failure and thereby defer the delivery if it tries to use (for example) a
3946 cached ldap connection that the parent has called unbind on.
3948 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3949 of $address_data from the recipient verification was clobbered by the
3950 sender verification.
3952 13. The value of address_data from a sender verification is now available in
3953 $sender_address_data in subsequent conditions in the ACL statement.
3955 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3957 15. Added a new option "connect=<time>" to callout options, to set a different
3960 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3961 was its contents. (It was OK if the option was not defined at all.)
3963 17. A "Completed" log line is now written for messages that are removed from
3964 the spool by the -Mrm option.
3966 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3967 information about exactly what failed.
3969 19. Added -dd to debug only the daemon process.
3971 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3972 handles timeouts, both on the server side and network timeouts. Renamed the
3973 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3975 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3976 It is now set to "smtps".
3978 22. $host_address is now set to the target address during the checking of
3979 ignore_target_hosts.
3981 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3982 being passed; this would have caused $sender_host_name to have been used if
3983 matching the list had actually called for a host name (not very likely,
3984 since this list is usually IP addresses). A host name is now passed as
3987 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3988 code that specifies a non-restarting handler (typically sigaction() in
3989 modern systems) in an attempt to fix a rare and obscure crash bug.
3991 25. Narrowed the window for a race in the daemon that could cause it to ignore
3992 SIGCHLD signals. This is not a major problem, because they are used only to
3993 wake it up if nothing else does.
3995 26. A malformed maildirsize file could cause Exim to calculate negative values
3996 for the mailbox size or file count. Odd effects could occur as a result.
3997 The maildirsize information is now recalculated if the size or filecount
4000 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4001 support for a long time. Removed HAVE_SYS_VFS_H.
4003 28. Installed the latest version of exipick from John Jetmore.
4005 29. In an address list, if the pattern was not a regular expression, an empty
4006 subject address (from a bounce message) matched only if the pattern was an
4007 empty string. Non-empty patterns were not even tested. This was the wrong
4008 because it is perfectly reasonable to use an empty address as part of a
4009 database query. An empty address is now tested by patterns that are
4010 lookups. However, all the other forms of pattern expect the subject to
4011 contain a local part and a domain, and therefore, for them, an empty
4012 address still always fails if the pattern is not itself empty.
4014 30. Exim went into a mad DNS loop when attempting to do a callout where the
4015 host was specified on an smtp transport, and looking it up yielded more
4016 than one IP address.
4018 31. Re-factored the code for checking spool and log partition space into a
4019 function that finds that data and another that does the check. The former
4020 is then used to implement four new variables: $spool_space, $log_space,
4021 $spool_inodes, and $log_inodes.
4023 32. The RFC2047 encoding function was originally intended for short strings
4024 such as real names; it was not keeping to the 75-character limit for
4025 encoded words that the RFC imposes. It now respects the limit, and
4026 generates multiple encoded words if necessary. To be on the safe side, I
4027 have increased the buffer size for the ${rfc2047: expansion operator from
4030 33. It is now permitted to omit both strings after an "if" condition; if the
4031 condition is true, the result is "true". As before, when the second string
4032 is omitted, a false condition yields an empty string. This makes it less
4033 cumbersome to write custom ACL and router conditions.
4035 34. Failure to deliver a bounce message always caused it to be frozen, even if
4036 there was an errors_to setting on the router. The errors_to setting is now
4039 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4040 canonical form (fully expanded) before being placed in
4041 $sender_host_address.
4043 36. The table in the code that translates DNS record types into text (T_A to
4044 "A" for instance) was missing entries for NS and CNAME. It is just possible
4045 that this could have caused confusion if both these types were looked up
4046 for the same domain, because the text type is used as part of Exim's
4047 per-process caching. But the chance of anyone hitting this buglet seems
4050 37. The dnsdb lookup has been extended in a number of ways.
4052 (1) There is a new type, "zns", which walks up the domain tree until it
4053 finds some nameserver records. It should be used with care.
4055 (2) There is a new type, "mxh", which is like "mx" except that it returns
4056 just the host names, not the priorities.
4058 (3) It is now possible to give a list of domains (or IP addresses) to be
4059 looked up. The behaviour when one of the lookups defers can be
4060 controlled by a keyword.
4062 (4) It is now possible to specify the separator character for use when
4063 multiple records are returned.
4065 38. The dnslists ACL condition has been extended: it is now possible to supply
4066 a list of IP addresses and/or domains to be looked up in a particular DNS
4069 39. Added log_selector=+queue_time_overall.
4071 40. When running the queue in the test harness, wait just a tad after forking a
4072 delivery process, to get repeatability of debugging output.
4074 41. Include certificate and key file names in error message when GnuTLS fails
4075 to set them up, because the GnuTLS error message doesn't include the name
4076 of the failing file when there is a problem reading it.
4078 42. Allow both -bf and -bF in the same test run.
4080 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4082 44. The "Exiscan patch" is now merged into the mainline Exim source.
4084 45. Sometimes the final signoff response after QUIT could fail to get
4085 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4086 < 0 before doing a fflush(). This bug looks as though it goes back to the
4087 introduction of TLS in release 3.20, but "sometimes" must have been rare
4088 because the tests only now provoked it.
4090 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4091 (this can affect the format of dates).
4093 47. exim_tidydb, when checking for the continued existence of a message for
4094 which it has found a message-specific retry record, was not finding
4095 messages that were in split spool directories. Consequently, it was
4096 deleting retry records that should have stayed in existence.
4098 48. Steve fixed some bugs in eximstats.
4100 49. The SPA authentication driver was not abandoning authentication and moving
4101 on to the next authenticator when an expansion was forced to fail,
4102 contradicting the general specification for all authenticators. Instead it
4103 was generating a temporary error. It now behaves as specified.
4105 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4106 (the order specifies the preference for clients). The order is now AES256,
4107 AES128, 3DES, ARCFOUR128.
4109 51. Small patch to Sieve code - explicitly set From: when generating an
4112 52. Exim crashed if a remote delivery caused a very long error message to be
4113 recorded - for instance if somebody sent an entire SpamAssassin report back
4114 as a large number of 550 error lines. This bug was coincidentally fixed by
4115 increasing the size of one of Exim's internal buffers (big_buffer) that
4116 happened as part of the Exiscan merge. However, to be on the safe side, I
4117 have made the code more robust (and fixed the comments that describe what
4120 53. Now that there can be additional text after "Completed" in log lines (if
4121 the queue_time_overall log selector is set), a one-byte patch to exigrep
4122 was needed to allow it to recognize "Completed" as not the last thing in
4125 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4126 patch that reportedly fixes this has been added. I am not expert enough to
4127 create a test for it. This is what the patch creator wrote:
4129 "I found a little strange behaviour of ldap code when working with
4130 Windows 2003 AD Domain, where users was placed in more than one
4131 Organization Units. When I tried to give exim partial DN, the exit code
4132 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4133 But simultaneously result of request was absolutely normal ldap result,
4134 so I produce this patch..."
4136 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4137 REFERENCE, so I have modified the code to exclude the patch when that macro
4140 55. Some experimental protocols are using DNS PTR records for new purposes. The
4141 keys for these records are domain names, not reversed IP addresses. The
4142 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4143 leaves it alone. Component reversal etc. now happens only for IP addresses.
4146 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4148 57. Double the size of the debug message buffer (to 2048) so that more of very
4149 long debug lines gets shown.
4151 58. The exicyclog utility now does better if the number of log files to keep
4152 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4154 59. Two changes related to the smtp_active_hostname option:
4156 (1) $smtp_active_hostname is now available as a variable.
4157 (2) The default for smtp_banner uses $smtp_active_hostname instead
4158 of $primary_hostname.
4160 60. The host_aton() function is supposed to be passed a string that is known
4161 to be a valid IP address. However, in the case of IPv6 addresses, it was
4162 not checking this. This is a hostage to fortune. Exim now panics and dies
4163 if the condition is not met. A case was found where this could be provoked
4164 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4165 components; fortuitously, this particular loophole had already been fixed
4166 by change 4.50/55 above.
4168 If there are any other similar loopholes, the new check in host_aton()
4169 itself should stop them being exploited. The report I received stated that
4170 data on the command line could provoke the exploit when Exim was running as
4171 exim, but did not say which command line option was involved. All I could
4172 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4173 running as the user.
4176 61. There was a buffer overflow vulnerability in the SPA authentication code
4177 (which came originally from the Samba project). I have added a test to the
4178 spa_base64_to_bits() function which I hope fixes it.
4181 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4182 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4184 63. The daemon start-up calls getloadavg() while still root for those OS that
4185 need the first call to be done as root, but it missed one case: when
4186 deliver_queue_load_max is set with deliver_drop_privilege. This is
4187 necessary for the benefit of the queue runner, because there is no re-exec
4188 when deliver_drop_privilege is set.
4190 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4191 This has been fixed.
4193 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4194 was in use, was not putting the data itself into the right store pool;
4195 consequently, it could be overwritten for a subsequent message in the same
4196 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4199 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4201 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4202 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4203 the function were treating the return as a boolean value, which happened to
4204 work because 0=false and not-0=true, but is not correct code.
4206 68. The host_aton() function was not handling scoped IPv6 addresses (those
4207 with, for example, "%eth0" on the end) correctly.
4209 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4210 negated items (that is, ~something) in unsigned ints. Some compilers
4211 apparently mutter when there is no cast.
4213 70. If an address verification called from an ACL failed, and did not produce a
4214 user-specific message (i.e. there was only a "system" message), nothing was
4215 put in $acl_verify_message. In this situation, it now puts the system
4218 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4219 session; change 4.31/43 added the unwanted input to the log line - except
4220 that it did not do this in the start of session case. It now does.
4222 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4223 This is wrong and can cause the other end to generate a synchronization
4224 error if it is another Exim or anything else that does the synchronization
4225 check. A QUIT command is no longer sent after a timeout.
4227 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4228 during host lookups.
4230 74. The defer_ok option of callout verification was not working if it was used
4231 when verifying addresses in header lines, that is, for this case:
4233 verify = header_sender/callout=defer_ok
4235 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4236 those file descriptors could be used for SMTP connections. If anything
4237 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4238 could be sent to the SMTP client, causing chaos. The daemon now opens
4239 stdin, stdout, and stderr to /dev/null when it puts itself into the
4242 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4243 log by default. The user can override this with suitable Perl magic.
4245 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4246 the log message when discard triggers, was not working for the DATA ACL or
4247 for the non-SMTP ACL.
4249 78. Error message wording change in sieve.c.
4251 79. If smtp_accept_max_per_host was set, the number of connections could be
4252 restricted to fewer than expected, because the daemon was trying to set up
4253 a new connection before checking whether the processes handling previous
4254 connections had finished. The check for completed processes is now done
4255 earlier. On busy systems, this bug wouldn't be noticed because something
4256 else would have woken the daemon, and it would have reaped the completed
4259 80. If a message was submitted locally by a user whose login name contained one
4260 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4261 It caused a spool format error. I have fixed the spool reading code. A
4262 related problem was that the "from" clause in the Received: line became
4263 illegal because of the space(s). It is now covered by ${quote_local_part.
4265 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4268 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4270 83. Patch from Sieve maintainer to fix unterminated string problem in
4271 "vacation" handling.
4273 84. Some minor changes to the Linux configuration files to help with other
4274 OS variants using glibc.
4276 85. One more patch for Sieve to update vacation handling to latest spec.
4279 ----------------------------------------------------
4280 See the note above about the 4.44 and 4.50 releases.
4281 ----------------------------------------------------
4287 1. Change 4.43/35 introduced a bug that caused file counts to be
4288 incorrectly computed when quota_filecount was set in an appendfile
4291 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4292 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4295 3. Give more explanation in the error message when the command for a transport
4296 filter fails to execute.
4298 4. There are several places where Exim runs a non-Exim command in a
4299 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4300 was being done only for the command run by the queryprogram router. It is
4301 now done for all such subprocesses. The other cases are: ${run, transport
4302 filters, and the commands run by the lmtp and pipe transports.
4304 5. Some older OS have a limit of 256 on the maximum number of file
4305 descriptors. Exim was using setrlimit() to set 1000 as a large value
4306 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4307 systems. I've change it so that if it can't get 1000, it tries for 256.
4309 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4310 was an oversight, and furthermore, ever since the addition of extra
4311 controls (e.g. 4.43/32), the checks on when to allow different forms of
4312 "control" were broken. There should now be diagnostics for all cases when a
4313 control that does not make sense is encountered.
4315 7. $recipients is now available in the predata ACL (oversight).
4317 8. Tidy the search cache before the fork to do a delivery from a message
4318 received from the command line. Otherwise the child will trigger a lookup
4319 failure and thereby defer the delivery if it tries to use (for example) a
4320 cached ldap connection that the parent has called unbind on.
4322 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4323 of $address_data from the recipient verification was clobbered by the
4324 sender verification.
4326 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4327 was its contents. (It was OK if the option was not defined at all.)
4329 11. A "Completed" log line is now written for messages that are removed from
4330 the spool by the -Mrm option.
4332 12. $host_address is now set to the target address during the checking of
4333 ignore_target_hosts.
4335 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4336 being passed; this would have caused $sender_host_name to have been used if
4337 matching the list had actually called for a host name (not very likely,
4338 since this list is usually IP addresses). A host name is now passed as
4341 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4342 code that specifies a non-restarting handler (typically sigaction() in
4343 modern systems) in an attempt to fix a rare and obscure crash bug.
4345 15. Narrowed the window for a race in the daemon that could cause it to ignore
4346 SIGCHLD signals. This is not a major problem, because they are used only to
4347 wake it up if nothing else does.
4349 16. A malformed maildirsize file could cause Exim to calculate negative values
4350 for the mailbox size or file count. Odd effects could occur as a result.
4351 The maildirsize information is now recalculated if the size or filecount
4354 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4355 support for a long time. Removed HAVE_SYS_VFS_H.
4357 18. Updated exipick to current release from John Jetmore.
4359 19. Allow an empty sender to be matched against a lookup in an address list.
4360 Previously the only cases considered were a regular expression, or an
4363 20. Exim went into a mad DNS lookup loop when doing a callout where the
4364 host was specified on the transport, if the DNS lookup yielded more than
4367 21. The RFC2047 encoding function was originally intended for short strings
4368 such as real names; it was not keeping to the 75-character limit for
4369 encoded words that the RFC imposes. It now respects the limit, and
4370 generates multiple encoded words if necessary. To be on the safe side, I
4371 have increased the buffer size for the ${rfc2047: expansion operator from
4374 22. Failure to deliver a bounce message always caused it to be frozen, even if
4375 there was an errors_to setting on the router. The errors_to setting is now
4378 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4379 canonical form (fully expanded) before being placed in
4380 $sender_host_address.
4382 24. Updated eximstats to version 1.33
4384 25. Include certificate and key file names in error message when GnuTLS fails
4385 to set them up, because the GnuTLS error message doesn't include the name
4386 of the failing file when there is a problem reading it.
4388 26. Expand error message when OpenSSL has problems setting up cert/key files.
4391 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4392 (this can affect the format of dates).
4394 28. exim_tidydb, when checking for the continued existence of a message for
4395 which it has found a message-specific retry record, was not finding
4396 messages that were in split spool directories. Consequently, it was
4397 deleting retry records that should have stayed in existence.
4399 29. eximstats updated to version 1.35
4400 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4401 1.35 - bugfix such that pie charts by volume are generated correctly
4403 30. The SPA authentication driver was not abandoning authentication and moving
4404 on to the next authenticator when an expansion was forced to fail,
4405 contradicting the general specification for all authenticators. Instead it
4406 was generating a temporary error. It now behaves as specified.
4408 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4409 (the order specifies the preference for clients). The order is now AES256,
4410 AES128, 3DES, ARCFOUR128.
4412 31. Small patch to Sieve code - explicitly set From: when generating an
4415 32. Exim crashed if a remote delivery caused a very long error message to be
4416 recorded - for instance if somebody sent an entire SpamAssassin report back
4417 as a large number of 550 error lines. This bug was coincidentally fixed by
4418 increasing the size of one of Exim's internal buffers (big_buffer) that
4419 happened as part of the Exiscan merge. However, to be on the safe side, I
4420 have made the code more robust (and fixed the comments that describe what
4423 33. Some experimental protocols are using DNS PTR records for new purposes. The
4424 keys for these records are domain names, not reversed IP addresses. The
4425 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4426 leaves it alone. Component reversal etc. now happens only for IP addresses.
4429 34. The host_aton() function is supposed to be passed a string that is known
4430 to be a valid IP address. However, in the case of IPv6 addresses, it was
4431 not checking this. This is a hostage to fortune. Exim now panics and dies
4432 if the condition is not met. A case was found where this could be provoked
4433 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4434 components; fortuitously, this particular loophole had already been fixed
4435 by change 4.50/55 or 4.44/33 above.
4437 If there are any other similar loopholes, the new check in host_aton()
4438 itself should stop them being exploited. The report I received stated that
4439 data on the command line could provoke the exploit when Exim was running as
4440 exim, but did not say which command line option was involved. All I could
4441 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4442 running as the user.
4445 35. There was a buffer overflow vulnerability in the SPA authentication code
4446 (which came originally from the Samba project). I have added a test to the
4447 spa_base64_to_bits() function which I hope fixes it.
4450 36. The daemon start-up calls getloadavg() while still root for those OS that
4451 need the first call to be done as root, but it missed one case: when
4452 deliver_queue_load_max is set with deliver_drop_privilege. This is
4453 necessary for the benefit of the queue runner, because there is no re-exec
4454 when deliver_drop_privilege is set.
4456 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4457 was in use, was not putting the data itself into the right store pool;
4458 consequently, it could be overwritten for a subsequent message in the same
4459 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4462 38. Sometimes the final signoff response after QUIT could fail to get
4463 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4464 < 0 before doing a fflush(). This bug looks as though it goes back to the
4465 introduction of TLS in release 3.20, but "sometimes" must have been rare
4466 because the tests only now provoked it.
4472 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4473 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4474 Now it returns an integer. A number of calls were still expecting a T/F
4475 return. Fortuitously, in all cases, the tests worked in OK situations,
4476 which is the norm. However, things would have gone wrong on any write
4477 failures on the smtp file descriptor. This function is used when sending
4478 messages over SMTP and also when doing verify callouts.
4480 2. When Exim is called to do synchronous delivery of a locally submitted
4481 message (the -odf or -odi options), it no longer closes stderr before doing
4484 3. Implemented the mua_wrapper option.
4486 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4488 5. Implemented the functions header_remove(), header_testname(),
4489 header_add_at_position(), and receive_remove_recipient(), and exported them
4492 6. If an ACL "warn" statement specified the addition of headers, Exim already
4493 inserted X-ACL-Warn: at the start if there was no header name. However, it
4494 was not making this test for the second and subsequent header lines if
4495 there were newlines in the string. This meant that an invalid header could
4496 be inserted if Exim was badly configured.
4498 7. Allow an ACL "warn" statement to add header lines at the start or after all
4499 the Received: headers, as well as at the end.
4501 8. Added the rcpt_4xx retry error code.
4503 9. Added postmaster_mailfrom=xxx to callout verification option.
4505 10. Added mailfrom=xxxx to the callout verification option, for verify=
4508 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4509 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4511 12. Inserted some casts to stop certain compilers warning when using pointer
4512 differences as field lengths or precisions in printf-type calls (mostly
4513 affecting debugging statements).
4515 13. Added optional readline() support for -be (dynamically loaded).
4517 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4518 same clock tick as a message's arrival, so that its received time was the
4519 same as the "first fail" time on the retry record, and that message
4520 remained on the queue past the ultimate address timeout, every queue runner
4521 would try a delivery (because it was past the ultimate address timeout) but
4522 after another failure, the ultimate address timeout, which should have then
4523 bounced the address, did not kick in. This was a "< instead of <=" error;
4524 in most cases the first failure would have been in the next clock tick
4525 after the received time, and all would be well.
4527 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4528 being recognized when the domain list was tested by the match_domain
4529 condition in an expansion string.
4531 16. Added the ${str2b64: operator.
4533 17. Exim was always calling setrlimit() to set a large limit for the number of
4534 processes, without checking whether the existing limit was already
4535 adequate. (It did check for the limit on file descriptors.) Furthermore,
4536 errors from getrlimit() and setrlimit() were being ignored. Now they are
4537 logged to the main and panic logs, but Exim does carry on, to try to do its
4538 job under whatever limits there are.
4540 18. Imported PCRE 5.0.
4542 19. Trivial typo in log message " temporarily refused connection" (the leading
4545 20. If the log selector return_path_on_delivery was set and an address was
4546 redirected to /dev/null, the delivery process crashed because it assumed
4547 that a return path would always be set for a "successful" delivery. In this
4548 case, the whole delivery is bypassed as an optimization, and therefore no
4551 21. Internal re-arrangement: the function for sending a challenge and reading
4552 a response while authentication was assuming a zero-terminated challenge
4553 string. It's now changed to take a pointer and a length, to allow for
4554 binary data in such strings.
4556 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4558 23. Exim was not respecting finduser_retries when seeking the login of the
4559 uid under which it was called; it was always trying 10 times. (The default
4560 setting of finduser_retries is zero.) Also, it was sleeping after the final
4561 failure, which is pointless.
4563 24. Implemented tls_on_connect_ports.
4565 25. Implemented acl_smtp_predata.
4567 26. If the domain in control=submission is set empty, Exim assumes that the
4568 authenticated id is a complete email address when it generates From: or
4569 Sender: header lines.
4571 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4572 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4573 chown and chgrp in /bin and hostname in /usr/bin.
4575 28. Exim was keeping the "process log" file open after each use, just as it
4576 does for the main log. This opens the possibility of it remaining open for
4577 long periods when the USR1 signal hits a daemon. Occasional processlog
4578 errors were reported, that could have been caused by this. Anyway, it seems
4579 much more sensible not to leave this file open at all, so that is what now
4582 29. The long-running daemon process does not normally write to the log once it
4583 has entered its main loop, and it closes the log before doing so. This is
4584 so that log files can straightforwardly be renamed and moved. However,
4585 there are a couple of unusual error situations where the daemon does write
4586 log entries, and I had neglected to close the log afterwards.
4588 30. The text of an SMTP error response that was received during a remote
4589 delivery was being truncated at 512 bytes. This is too short for some of
4590 the long messages that one sometimes sees. I've increased the limit to
4593 31. It is now possible to make retry rules that apply only when a message has a
4594 specific sender, in particular, an empty sender.
4596 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4597 it possible to be selective about when SMTP synchronization is enforced.
4599 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4601 32. Implemented hosts_connection_nolog.
4603 33. Added an ACL for QUIT.
4605 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4608 35. Added mailbox_size and mailbox_filecount to appendfile.
4610 36. Added control = no_multiline_responses to ACLs.
4612 37. There was a bug in the logic of the code that waits for the clock to tick
4613 in the case where the clock went backwards by a substantial amount such
4614 that the microsecond fraction of "now" was more than the microsecond
4615 fraction of "then" (but the whole seconds number was less).
4617 38. Added support for the libradius Radius client library this is found on
4618 FreeBSD (previously only the radiusclient library was supported).
4624 1. When certain lookups returned multiple values in the form name=value, the
4625 quoting of the values was not always being done properly. Specifically:
4626 (a) If the value started with a double quote, but contained no whitespace,
4628 (b) If the value contained whitespace other than a space character (i.e.
4629 tabs or newlines or carriage returns) it was not quoted.
4630 This fix has been applied to the mysql and pgsql lookups by writing a
4631 separate quoting function and calling it from the lookup code. The fix
4632 should probably also be applied to nisplus, ibase and oracle lookups, but
4633 since I cannot test any of those, I have not disturbed their existing code.
4635 2. A hit in the callout cache for a specific address caused a log line with no
4636 reason for rejecting RCPT. Now it says "Previous (cached) callout
4637 verification failure".
4639 3. There was an off-by-one bug in the queryprogram router. An over-long
4640 return line was truncated at 256 instead of 255 characters, thereby
4641 overflowing its buffer with the terminating zero. As well as fixing this, I
4642 have increased the buffer size to 1024 (and made a note to document this).
4644 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4645 when Exim is waiting for an SMTP response from a remote server, Exim
4646 restarts its select() call on the socket, thereby resetting its timeout.
4647 This is not a problem when such interrupts are rare. Somebody set up a cron
4648 job to run exiwhat every 2 minutes, which is less than the normal select()
4649 timeout (5 or 10 minutes). This meant that the select() timeout never
4650 kicked in because it was always reset. I have fixed this by comparing the
4651 time when an interrupt arrives with the time at the start of the first call
4652 to select(). If more time than the timeout has elapsed, the interrupt is
4653 treated as a timeout.
4655 5. Some internal re-factoring in preparation for the addition of Sieve
4656 extensions (by MH). In particular, the "personal" test is moved to a
4657 separate function, and given an option for scanning Cc: and Bcc: (which is
4658 not set for Exim filters).
4660 6. When Exim created an email address using the login of the caller as the
4661 local part (e.g. when creating a From: or Sender: header line), it was not
4662 quoting the local part when it contained special characters such as @.
4664 7. Installed new OpenBSD configuration files.
4666 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4667 try to make them clearer.
4669 9. Callout options, other than the timeout value, were being ignored when
4670 verifying sender addresses in header lines. For example, when using
4672 verify = header_sender/callout=no_cache
4674 the cache was (incorrectly) being used.
4676 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4677 only the Cygwin environment.
4679 11. When return_path_on_delivery was set as a log selector, if different remote
4680 addresses in the same message used different return paths and parallel
4681 remote delivery occurred, the wrong values would sometimes be logged.
4682 (Whenever a remote delivery process finished, the return path value from
4683 the most recently started remote delivery process was logged.)
4685 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4686 lines when AUTH and/or TLS are in use. This is the "received protocol"
4687 field. Exim used to use "asmtp" for authenticated SMTP, without any
4688 indication (in the protocol name) for TLS use. Now it follows the RFC and
4689 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4690 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4691 names appear in log lines as well as in Received: header lines.
4693 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4694 extensions, and comparison tests, and to fix some bugs.
4696 14. Changes to the "personal" filter test:
4698 (1) The test was buggy in that it was just doing the equivalent of
4699 "contains" tests on header lines. For example, if a user's address was
4700 anne@some.where, the "personal" test would incorrectly be true for
4702 To: susanne@some.where
4704 This test is now done by extracting each address from the header in turn,
4705 and checking the entire address. Other tests that are part of "personal"
4706 are now done using regular expressions (for example, to check local parts
4707 of addresses in From: header lines).
4709 (2) The list of non-personal local parts in From: addresses has been
4710 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4711 taken from the Sieve specification recommendations.
4713 (3) If the message contains any header line starting with "List-" it is
4714 treated as non-personal.
4716 (4) The test for "circular" in the Subject: header line has been removed
4717 because it now seems ill-conceived.
4719 15. Minor typos in src/EDITME comments corrected.
4721 16. Installed latest exipick from John Jetmore.
4723 17. If headers_add on a router specified a text string that was too long for
4724 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4725 of string_sprintf() is now avoided.
4727 18. $message_body_size was not set (it was always zero) when running the DATA
4728 ACL and the local_scan() function.
4730 19. For the "mail" command in an Exim filter, no default was being set for
4731 the once_repeat time, causing a random time value to be used if "once" was
4732 specified. (If the value happened to be <= 0, no repeat happened.) The
4733 default is now 0s, meaning "never repeat". The "vacation" command was OK
4734 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4735 (I found it when inspecting the code).
4737 20. There is now an overall timeout for performing a callout verification. It
4738 defaults to 4 times the callout timeout, which applies to individual SMTP
4739 commands during the callout. The overall timeout applies when there is more
4740 than one host that can be tried. The timeout is checked before trying the
4741 next host. This prevents very long delays if there are a large number of
4742 hosts and all are timing out (e.g. when the network connections are timing
4743 out). The value of the overall timeout can be changed by specifying an
4744 additional sub-option for "callout", called "maxwait". For example:
4746 verify = sender/callout=5s,maxwait=20s
4748 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4749 to the end before writing, but this should make it even safer).
4751 22. Exim was forgetting that it had advertised PIPELINING for the second and
4752 subsequent messages on an SMTP connection. It was also not resetting its
4753 memory on STARTTLS and an internal HELO.
4755 23. When Exim logs an SMTP synchronization error within a session, it now
4756 records whether PIPELINING has been advertised or not.
4758 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4759 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4760 rather than long int.
4762 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4764 26. Added the never_mail option to autoreply.
4770 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4771 crash if the getsockname() call failed; this can happen if a connection is
4772 closed very soon after it is established. The problem was simply in the
4773 order in which certain operations were done, causing Exim to try to write
4774 to the SMTP stream before it had set up the file descriptor. The bug has
4775 been fixed by making things happen in the correct order.
4781 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4782 before the connection was closed, thus losing the rejection response.
4784 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4785 some early Solaris releases, but causes trouble in current releases where
4786 socklen_t is defined.
4788 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4791 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4794 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4795 that releasing the top of it at the end releases what was used for sub-
4796 expansions (unless the block got too big). However, discard this block if
4797 the first thing is a variable or header, so that we can use its block when
4798 it is dynamic (useful for very large $message_headers, for example).
4800 6. Lookups now cache *every* query, not just the most recent. A new, separate
4801 store pool is used for this. It can be recovered when all lookup caches are
4802 flushed. Lookups now release memory at the end of their result strings.
4803 This has involved some general refactoring of the lookup sources.
4805 7. Some code has been added to the store_xxx() functions to reduce the amount
4806 of flapping under certain conditions.
4808 8. log_incoming_interface used to affect only the <= reception log lines. Now
4809 it causes the local interface and port to be added to several more SMTP log
4810 lines, for example "SMTP connection from", and rejection lines.
4812 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4814 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4816 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4817 could be overwritten at the end of the current message (or the start of a
4818 new message if it was set in a HELO ACL). The value is now preserved for
4819 the duration of the SMTP connection.
4821 12. If a transport had a headers_rewrite setting, and a matching header line
4822 contained an unqualified address, that address was qualified, even if it
4823 did not match any rewriting rules. The underlying bug was that the values
4824 of the flags that permit the existence of unqualified sender and recipient
4825 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4826 non-local messages, and by -bnq for local messages) were not being
4827 preserved with the message after it was received.
4829 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4830 "next input=" as part of the text comprising the host identity instead of
4831 the correct text. The code was using the same buffer for two different
4832 strings. However, depending on which order the printing function evaluated
4833 its arguments, the bug did not always show up. Under Linux, for example, my
4834 test suite worked just fine.
4836 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4837 doesn't work with some older versions of Perl. It has been changed to "my",
4838 which in any case is probably the better facility to use.
4840 15. A really picky compiler found some instances of statements for creating
4841 error messages that either had too many or two few arguments for the format
4844 16. The size of the buffer for calls to the DNS resolver has been increased
4845 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4846 for addresses that have a lot of PTR records. This alleviates a problem; it
4847 does not fully solve it.
4849 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4850 buffer now truncates the list and logs the incident, which is the same
4851 action as happens when Exim is looking up a host name and its aliases.
4852 Previously in this situation something unpredictable would happen;
4853 sometimes it was "internal error: store_reset failed".
4855 18. If a server dropped the connection unexpectedly when an Exim client was
4856 using GnuTLS and trying to read a response, the client delivery process
4857 crashed while trying to generate an error log message.
4859 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4860 string, for example:
4862 warn message = H1: something\nH2: something
4864 the text was added as a single header line from Exim's point of view
4865 though it ended up OK in the delivered message. However, searching for the
4866 second and subsequent header lines using $h_h2: did not work. This has been
4867 fixed. Similarly, if a system filter added multiple headers in this way,
4868 the routers could not see them.
4870 20. Expanded the error message when iplsearch is called with an invalid key to
4871 suggest using net-iplsearch in a host list.
4873 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4874 ACLs are no longer actually imposed (and a message to that effect is
4877 22. If a "gecos" field in a passwd entry contained escaped characters, in
4878 particular, if it contained a \" sequence, Exim got it wrong when building
4879 a From: or a Sender: header from that name. A second bug also caused
4880 incorrect handling when an unquoted " was present following a character
4881 that needed quoting.
4883 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4884 was not being matched caselessly.
4886 24. Arranged for all hyphens in the exim.8 source to be escaped with
4889 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4890 properly. Recipient callout cache records were still being keyed to include
4891 the sender, even when use_sender was set false. This led to far more
4892 callouts that were necessary. The sender is no longer included in the key
4893 when use_sender is false.
4895 26. Added "control = submission" modifier to ACLs.
4897 27. Added the ${base62d: operator to decode base 62 numbers.
4899 28. dnsdb lookups can now access SRV records.
4901 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4902 the configuration file.
4904 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4905 (-v) mode. This makes the output for a verbose queue run more intelligible.
4907 31. Added a use_postmaster feature to recipient callouts.
4909 32. Added the $body_zerocount variable, containing the number of binary zero
4910 bytes in the message body.
4912 33. The time of last modification of the "new" subdirectory is now used as the
4913 "mailbox time last read" when there is a quota error for a maildir
4916 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4918 35. Added +ignore_unknown as a special item in host lists.
4920 36. Code for decoding IPv6 addresses in host lists is now included, even if
4921 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4922 address was recognized as an IP address, but was then not correctly decoded
4923 into binary, causing unexpected and incorrect effects when compared with
4930 1. Very minor rewording of debugging text in manualroute to say "list of
4931 hosts" instead of "hostlist".
4933 2. If verify=header_syntax was set, and a header line with an unqualified
4934 address (no domain) and a large number of spaces between the end of the
4935 name and the colon was received, the reception process suffered a buffer
4936 overflow, and (when I tested it) crashed. This was caused by some obsolete
4937 code that should have been removed. The fix is to remove it!
4939 3. When running in the test harness, delay a bit after writing a bounce
4940 message to get a bit more predictability in the log output.
4942 4. Added a call to search_tidyup() just before forking a reception process. In
4943 theory, someone could use a lookup in the expansion of smtp_accept_max_
4944 per_host which, without the tidyup, could leave open a database connection.
4946 5. Added the variables $recipient_data and $sender_data which get set from a
4947 lookup success in an ACL "recipients" or "senders" condition, or a router
4948 "senders" option, similar to $domain_data and $local_part_data.
4950 6. Moved the writing of debug_print from before to after the "senders" test
4953 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4954 problems for message scanning, either using a data ACL, or using
4955 local_scan() because the Received: header was not generated till after they
4956 were called (in order to set the time as the time of reception completion).
4957 I have revised the way this works. The header is now generated after the
4958 body is received, but before the ACL or local_scan() are called. After they
4959 are run, the timestamp in the header is updated.
4965 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4966 before starting a queue runner without re-exec. This happened only when
4967 deliver_drop_privilege was set or when the Exim user was set to root. The
4968 effect of the bug was that timeouts during subsequent deliveries caused
4969 crashes instead of being properly handled. The handler is now left at its
4970 default (and expected) setting.
4972 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4973 message, again when deliver_drop_privilege is set or Exim is run as root.
4974 The bug described in (1) was not present in this case, but the tidying up
4975 of the other signals was missing. I have made the two cases consistent.
4977 3. The ignore_target_hosts setting on a manualroute router was being ignored
4978 for hosts that were looked up using the /MX notation.
4980 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4983 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4984 operated on the sender address. After changing the $sender_address to <>
4985 for the sender address verify, Exim was re-instated it as the original
4986 (before rewriting) address, but remembering that it had rewritten it, so it
4987 wasn't rewriting it again. This bug also had the effect of breaking the
4988 sender address verification caching when the sender address was rewritten.
4990 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4991 This has been changed so that if the ip literal address matches
4992 ignore_target_hosts, the router declines.
4994 7. Added expansion conditions match_domain, match_address, and match_local_
4995 part (NOT match_host).
4997 8. The placeholder for the Received: header didn't have a length field set.
4999 9. Added code to Exim itself and to exim_lock to test for a specific race
5000 condition that could lead to file corruption when using MBX delivery. The
5001 issue is with the lockfile that is created in /tmp. If this file is removed
5002 after a process has opened it but before that process has acquired a lock,
5003 there is the potential for a second process to recreate the file and also
5004 acquire a lock. This could lead to two Exim processes writing to the file
5005 at the same time. The added code performs the same test as UW imapd; it
5006 checks after acquiring the lock that its file descriptor still refers to
5007 the same named file.
5009 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5010 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5013 11. Added the smtp_active_hostname option. If used, this will typically be made
5014 to depend on the incoming interface address. Because $interface_address is
5015 not set up until the daemon has forked a reception process, error responses
5016 that can happen earlier (such as "too many connections") no longer contain
5019 12. If an expansion in a condition on a "warn" statement fails because a lookup
5020 defers, the "warn" statement is abandoned, and the next ACL statement is
5021 processed. Previously this caused the whole ACL to be aborted.
5023 13. Added the iplsearch lookup type.
5025 14. Added ident_timeout as a log selector.
5027 15. Added tls_certificate_verified as a log selector.
5029 16. Added a global option tls_require_ciphers (compare the smtp transport
5030 option of the same name). This controls incoming TLS connections.
5032 17. I finally figured out how to make tls_require_ciphers do a similar thing
5033 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5034 before starting the TLS session.
5036 18. Tabs are now shown as \t in -bP output.
5038 19. If the log selector return_path_on_delivery was set, Exim crashed when
5039 bouncing a message because it had too many Received: header lines.
5041 20. If two routers both had headers_remove settings, and the first one included
5042 a superfluous trailing colon, the final name in the first list and the
5043 first name in the second list were incorrectly joined into one item (with a
5044 colon in the middle).
5050 1. Added -C and -D options to the exinext utility, mainly to make it easier
5051 to include in the automated testing, but these could be helpful when
5052 multiple configurations are in use.
5054 2. The exinext utility was not formatting the output nicely when there was
5055 an alternate port involved in the retry record key, nor when there was a
5056 message id as well (for retries that were specific to a specific message
5057 and a specific host). It was also confused by IPv6 addresses, because of
5058 the additional colons they contain. I have fixed the IPv4 problem, and
5059 patched it up to do a reasonable job for IPv6.
5061 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5062 delivery, the log line now contains "pipelined" if PIPELINING was used.
5064 4. An SMTP transport process used to panic and die if the bind() call to set
5065 an explicit outgoing interface failed. This has been changed; it is now
5066 treated in the same way as a connect() failure.
5068 5. A reference to $sender_host_name in the part of a conditional expansion
5069 that was being skipped was still causing a DNS lookup. This no longer
5072 6. The def: expansion condition was not recognizing references to header lines
5073 that used bh_ and bheader_.
5075 7. Added the _cache feature to named lists.
5077 8. The code for checking quota_filecount in the appendfile transport was
5078 allowing one more file than it should have been.
5080 9. For compatibility with Sendmail, the command line option
5088 and sets the incoming protocol and host name (for trusted callers). The
5089 host name and its colon can be omitted when only the protocol is to be set.
5090 Note the Exim already has two private options, -pd and -ps, that refer to
5091 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5092 "s", but I don't think that's a major issue.
5094 10. A number of refactoring changes to the code, none of which should affect
5097 (a) The number of logging options was getting close to filling up the
5098 32-bit word that was used as a bit map. I have split them into two classes:
5099 those that are passed in the argument to log_write(), and those that are
5100 only ever tested independently outside of that function. These are now in
5101 separate 32-bit words, so there is plenty of room for expansion again.
5102 There is no change in the user interface or the logging behaviour.
5104 (b) When building, for example, log lines, the code previously used a
5105 macro that called string_cat() twice, in order to add two strings. This is
5106 not really sufficiently general. Furthermore, there was one instance where
5107 it was actually wrong because one of the argument was used twice, and in
5108 one call a function was used. (As it happened, calling the function twice
5109 did not affect the overall behaviour.) The macro has been replaced by a
5110 function that can join an arbitrary number of extra strings onto a growing
5113 (c) The code for expansion conditions now uses a table and a binary chop
5114 instead of a serial search (which was left over from when there were very
5115 few conditions). Also, it now recognizes conditions like "pam" even when
5116 the relevant support is not compiled in: a suitably worded error message is
5117 given if an attempt is made to use such a condition.
5119 11. Added ${time_interval:xxxxx}.
5121 12. A bug was causing one of the ddress fields not to be passed back correctly
5122 from remote delivery subprocesses. The field in question was not being
5123 subsequently used, so this caused to problems in practice.
5125 13. Added new log selectors queue_time and deliver_time.
5127 14. Might have fixed a bug in maildirsizefile handling that threw up
5128 "unexpected character" debug warnings, and recalculated the data
5129 unnecessarily. In any case, I expanded the warning message to give more
5132 15. Added the message "Restricted characters in address" to the statements in
5133 the default ACL that block characters like @ and % in local parts.
5135 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5136 Three changes have been made:
5138 (a) There was a serious bug; a negative response to MAIL caused the whole
5139 recipient domain to be cached as invalid, thereby blocking all messages
5140 to all local parts at the same domain, from all senders. This bug has
5141 been fixed. The domain is no longer cached after a negative response to
5142 MAIL if the sender used is not empty.
5144 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5147 (c) A new callout option, "use_sender" has been added for people who want
5148 the modified behaviour.
5154 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5157 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5158 indeed breaks things for older releases.
5160 3. Added additional logging to the case where there is a problem reading data
5161 from a filter that is running in a subprocess using a pipe, in order to
5162 try to track down a specific problem.
5164 4. Testing facility fudge: when running in the test harness and attempting
5165 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5166 getting "No route to host". Convert this to a timeout.
5168 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5171 6. Some OS don't have socklen_t but use size_t instead. This affects the
5172 fifth argument of getsockopt() amongst other things. This is now
5173 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5174 can be set for individual OS. I have set it for SunOS5, OSF1, and
5175 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5176 some earlier ones do not.
5178 7. Change 4.30/15 was not doing the test caselessly.
5180 8. The standard form for an IPv6 address literal was being rejected by address
5181 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5182 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5183 this, as well as the form without the "IPv6" on the front (but only when
5184 address literals are enabled, of course).
5186 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5188 10. Exim crashed if a message with an empty sender address specified by -f
5189 encountered a router with an errors_to setting. This could be provoked only
5190 by a command such as
5194 where an empty string was supplied; "<>" did not hit this bug.
5196 11. Installed PCRE release 4.5.
5198 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5199 remained set. It is now erased.
5201 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5202 times from message ids (which are base 36 rather than the normal 62).
5204 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5205 were being counted as actual protocol errors, and logged if the log
5206 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5207 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5208 and DATA following a set of rejected RCPTs do not count as protocol errors.
5209 In other words, Exim assumes they were pipelined, though this may not
5210 actually be the case. Of course, in all cases the client gets an
5211 appropriate error code.
5213 15. If a lookup fails in an ACL condition, a message about the failure may
5214 be available; it is used if testing the ACL cannot continue, because most
5215 such messages specify what the cause of the deferral is. However, some
5216 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5217 that caused an old message to be retained and used if a later statement
5218 caused a defer, replacing the real cause of the deferral.
5220 16. If an IP address had so many PTR records that the DNS lookup buffer
5221 was not large enough to hold them, Exim could crash while trying to process
5222 the truncated data. It now detects and logs this case.
5224 17. Further to 4.21/58, another change has been made: if (and only if) the
5225 first line of a message (the first header line) ends with CRLF, a bare LF
5226 in a subsequent header line has a space inserted after it, so as not to
5227 terminate the header.
5229 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5230 unnecessarily, used atoi() instead of strtol(), and didn't check the
5231 termination when getting file sizes from file names by regex.
5233 19. Completely re-implemented the support for maildirsize files, in the light
5234 of a number of problems with the previous contributed implementation
5235 (4.30/29). In particular:
5237 . If the quota is zero, the maildirsize file is maintained, but no quota is
5240 . If the maildir directory does not exist, it is created before any attempt
5241 to write a maildirsize file.
5243 . The quota value in the file is just a cache; if the quota is changed in
5244 the transport, the new value overrides.
5246 . A regular expression is available for excluding directories from the
5249 20. The autoreply transport checks the characters in options that define the
5250 message's headers; it allows continued headers, but it was checking with
5251 isspace() after an embedded newline instead of explicitly looking for a
5254 21. If all the "regular" hosts to which an address was routed had passed their
5255 expiry times, and had not reached their retry times, the address was
5256 bounced, even if fallback hosts were defined. Now Exim should go on to try
5259 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5260 equivalent code in the SMTP transport. Some hosts send humungous responses
5261 to HELO/EHLO, more than 1024 it seems.
5263 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5264 gives compiler warnings in some environments. I've now done it "properly",
5267 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5268 (because of problems with the built-in one) was declared to return uschar *
5269 instead of char *, causing compiler failure.
5271 25. Fixed a file descriptor leak when processing alias/forward files.
5273 26. Fixed a minor format string issue in dbfn.c.
5275 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5277 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5278 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5281 29. When a sender address is verified, it is cached, to save repeating the test
5282 when there is more than one recipient in a message. However, when the
5283 verification involves a callout, it is possible for different callout
5284 options to be set for different recipients. It is too complicated to keep
5285 track of this in the cache, so now Exim always runs a verification when a
5286 callout is required, relying on the callout cache for the optimization.
5287 The overhead is duplication of the address routing, but this should not be
5290 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5291 to be verified with callout=postmaster, and the main callout worked but the
5292 postmaster check failed, the verification correctly failed. However, if a
5293 subsequent RCPT command asked for sender verification *without* the
5294 postmaster check, incorrect caching caused this verification also to fail,
5297 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5298 it was not caching the DNS options (qualify_single, search_parents) that
5299 were used when the lookup failed. A subsequent lookup with different
5300 options therefore always gave the same answer, though there were cases
5301 where it should not have. (Example: a "domains = !$mx_any" option on a
5302 dnslookup router: the "domains" option is always processed without any
5303 widening, but the router might have qualify_single set.) Now Exim uses the
5304 cached value only when the same options are set.
5306 32. Added John Jetmore's "exipick" utility to the distribution.
5308 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5309 than a timeout (e.g. a certificate is required, and is not provided), an
5310 Exim server now closes the connection immediately. Previously it waited for
5311 the client to close - but if the client is SSL, it seems that they each
5312 wait for each other, leading to a delay before one of them times out.
5314 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5315 maintained 0.8.x compatibility because I don't think many are using it, and
5316 it is clearly obsolete.
5318 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5321 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5322 client certificate was expired. A simple patch fixes this, though I don't
5323 understand the full logic of why the verify callback is called multiple
5326 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5327 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5328 which causes problems with some clients (such as the Certicom SSL Plus
5329 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5330 disables the coutermeasure allowing Eudora to connect."
5332 38. Exim was not checking that a write() to a log file succeeded. This could
5333 lead to Bad Things if a log got too big, in particular if it hit a file
5334 size limit. Exim now panics and dies if it cannot write to a log file, just
5335 as it does if it cannot open a log file.
5337 39. Modified OS/Makefile-Linux so that it now contains
5339 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5341 The two -D definitions ensure that Exim is compiled with large file
5342 support, which makes it possible to handle log files that are bigger than
5345 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5346 instance) a domain was checked against a named list that involved a lookup,
5347 causing $domain_data to be set, then another domain was checked against the
5348 same list, then the first domain was re-checked, the value of $domain_data
5349 after the final check could be wrong. In particular, if the second check
5350 failed, it could be set empty. This bug probably also applied to
5353 41. The strip_trailing_dot option was not being applied to the address given
5354 with the -f command-line option.
5356 42. The code for reading a message's header from the spool was incrementing
5357 $received_count, but never initializing it. This meant that the value was
5358 incorrect (doubled) while delivering a message in the same process in which
5359 it was received. In the most common configuration of Exim, this never
5360 happens - a fresh exec is done - but it can happen when
5361 deliver_drop_privilege is set.
5363 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5364 it now includes up to 150 characters of the unexpected data in the log
5367 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5368 and building data strings. The size of both of these buffers was 10 000
5369 bytes - far larger than anybody would *ever* want, thought I. Needless to
5370 say, somebody hit the limit. I have increased the maximum line length to
5371 20 000 and the maximum data length of concatenated lines to 100 000. I have
5372 also fixed two bugs, because there was no checking on these buffers. Tsk,
5373 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5374 buffer is too small.
5376 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5377 lsearch lookups. Now it does.
5379 46. When parsing a route_list item in a manualroute router, a fixed-length
5380 buffer was used for the list of hosts. I made this 1024 bytes long,
5381 thinking that nobody would ever have a list of hosts that long. Wrong.
5382 Somebody had a whole pile of complicated expansion conditions, and the
5383 string was silently truncated, leading to an expansion error. It turns out
5384 that it is easier to change to an unlimited length (owing to other changes
5385 that have happened since this code was originally written) than to build
5386 structure for giving a limitation error. The length of the item that
5387 expands into the list of hosts is now unlimited.
5389 47. The lsearch lookup could not handle data where the length of text line was
5390 more than 4095 characters. Such lines were truncated, leading to shortened
5391 data being returned. It should now handle lines of any length.
5393 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5394 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5397 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5398 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5399 no longer happens in comment lines. A list of replacements is now placed
5400 at the head of all of the source files, except those whose only change is
5401 to replace PERL_COMMAND in the very first #! line.
5403 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5404 messages on the queue, with a bottom-up merge sort, using code contributed
5405 by Michael Haardt. This should make operations like -bp somewhat faster on
5406 large queues. It won't affect queue runners, except when queue_run_in_order
5409 51. Installed eximstats 1.31 in the distribution.
5411 52. Added support for SRV lookups to the dnslookup router.
5413 53. If an ACL referred to $message_body or $message_body_end, the value was not
5414 reset for any messages that followed in the same SMTP session.
5416 54. The store-handling optimization for building very long strings was not
5417 differentiating between the different store pools. I don't think this
5418 actually made any difference in practice, but I've tidied it.
5420 55. While running the routers to verify a sender address, $sender_address
5421 was still set to the sender address. This is wrong, because when routing to
5422 send a bounce to the sender, it would be empty. Therefore, I have changed
5423 it so that, while verifying a sender address, $sender_address is set to <>.
5424 (There is no change to what happens when verifying a recipient address.)
5426 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5427 target A or AAAA records (if not already returned) without resetting the
5428 qualify_single or search_parents options of the DNS resolver. These are
5429 inappropriate in this case because the targets of MX and SRV records must
5430 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5431 target that, when qualified, matched something in the local domain. These
5432 two options are now turned off when doing these lookups.
5434 57. It seems that at least some releases of Reiserfs (which does not have the
5435 concept of a fixed number of inodes) returns zero and not -1 for the
5436 number of available inodes. This interacted badly with check_spool_inodes,
5437 which assumed that -1 was the "no such thing" setting. What I have done is
5438 to check that the total number of inodes is greater than zero before doing
5439 the test of how many are available.
5441 58. When a "warn" ACL statement has a log_message modifier, the message is
5442 remembered, and not repeated. This is to avoid a lot of repetition when a
5443 message has many recipients that cause the same warning to be written.
5444 Howewer, Exim was preserving the list of already written lines for an
5445 entire SMTP session, which doesn't seem right. The memory is now reset if a
5446 new message is started.
5448 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5449 debugging output unless log_rewrite was also set.
5451 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5452 of (int)(handle) when we know that handle contains (void *)(-1).
5454 61. The Exim daemon panic-logs an error return when it closes the incoming
5455 connection. However "connection reset by peer" seems to be common, and
5456 isn't really an error worthy of noting specially, so that particular error
5459 62. When Exim is trying to find all the local interfaces, it used to panic and
5460 die if the ioctl to get the interface flags failed. However, it seems that
5461 on at least one OS (Solaris 9) it is possible to have an interface that is
5462 included in the list of interfaces, but for which you get a failure error
5463 for this call. This happens when the interface is not "plumbed" into a
5464 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5465 failure of the "get flags" call assumes that the interface is down.
5467 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5468 makes life easier for people who are doing arithmetic on fields extracted
5469 from dates, where you often get leading zeros that should not be
5470 interpreted as octal.
5472 64. Added qualify_domain to the redirect router, to override the global
5475 65. If a pathologically long header line contained very many addresses (the
5476 report of this problem mentioned 10 000) and each of them was rewritten,
5477 Exim could use up a very large amount of memory. (It kept on making new
5478 copies of the header line as it rewrote, and never released the old ones.)
5479 At the expense of a bit more processing, the header rewriting function has
5480 been changed so that it no longer eats memory in this way.
5482 66. The generation of the Received: header has been moved from the time that a
5483 message starts to be received, to the time that it finishes. The timestamp
5484 in the Received: header should now be very close to that of the <= log
5485 line. There are two side-effects of this change:
5487 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5488 logged header lines no longer include the local Received: line, because
5489 it has not yet been created. The same applies to a copy of the message
5490 that is returned to a non-SMTP sender when a message is rejected.
5492 (b) When a filter file is tested using -bf, no additional Received: header
5493 is added to the test message. After some thought, I decided that this
5496 This change does not affect the value of $received_for. It is still set
5497 after address rewriting, but before local_scan() is called.
5499 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5501 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5502 gave an unhelpful panic error message, and a defer error. I have managed to
5503 change this behaviour so that it now rejects any supplied certificate,
5504 which seems right, as the list of acceptable certificates is empty.
5506 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5507 gave an unhelpful defer error. I have not managed to make this reject any
5508 supplied certificates, but the error message it gives is "no certificate
5509 supplied", which is not helpful.
5511 70. exigrep's output now also includes lines that are not associated with any
5512 message, but which match the given pattern. Implemented by a patch from
5513 Martin Sluka, which also tidied up the Perl a bit.
5515 71. Recipient callout verification, like sender verification, was using <> in
5516 the MAIL FROM command. This isn't really the right thing, since the actual
5517 sender may affect whether the remote host accepts the recipient or not. I
5518 have changed it to use the actual sender in the callout; this means that
5519 the cache record is now keyed on a recipient/sender pair, not just the
5520 recipient address. There doesn't seem to be a real danger of callout loops,
5521 since a callout by the remote host to check the sender would use <>.
5522 [SEE ABOVE: changed after hitting problems.]
5524 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5525 temporary errors. However, in the case of such a code being given after
5526 the end of a data transmission (i.e. after ".") Exim was failing to write
5527 a retry record for the message. (Yes, there was some broken host that was
5528 actually sending 8xx at this point.)
5530 73. An unknown lookup type in a host list could cause Exim to panic-die when
5531 the list was checked. (An example that provoked this was putting <; in the
5532 middle of a list instead of at the start.) If this happened during a DATA
5533 ACL check, a -D file could be left lying around. This kind of configuration
5534 error no longer causes Exim to die; instead it causes a defer errror. The
5535 incident is still logged to the main and panic logs.
5537 74. Buglet left over from Exim 3 conversion. The message "too many messages
5538 in one connection" was written to the rejectlog but not the mainlog, except
5539 when address rewriting (yes!) was being logged.
5541 75. Added write_rejectlog option.
5543 76. When a system filter was run not as root (that is, when system_filter_user
5544 was set), the values of the $n variables were not being returned to the
5545 main process; thus, they were not subsequently available in the $sn
5548 77. Added +return_path_on_delivery log selector.
5550 78. A connection timeout was being treated differently from recipients deferred
5551 when testing hosts_max_try with a message that was older than the host's
5552 retry timeout. (The host should not be counted, thus allowing all hosts to
5553 be tried at least once before bouncing.) This may have been the cause of an
5554 occasionally reported bug whereby a message would remain on the queue
5555 longer than the retry timeout, but would be bounced if a delivery was
5556 forced. I say "may" because I never totally pinned down the problem;
5557 setting up timeout/retry tests is difficult. See also the next item.
5559 79. The ultimate address timeout was not being applied to errors that involved
5560 a combination of host plus message (for example, a timeout on a MAIL
5561 command). When an address resolved to a number of possible hosts, and they
5562 were not all tried for each delivery (e.g. because of hosts_max_try), a
5563 message could remain on the queue longer than the retry timeout.
5565 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5568 81. Fixed an obscure SMTP outgoing bug which required at least the following
5569 conditions: (a) there was another message waiting for the same server;
5570 (b) the server returned 5xx to all RCPT commands in the first message so
5571 that the message was not completed; (c) the server dropped the connection
5572 or gave a negative response to the RSET that Exim sends to abort the
5573 transaction. The observed case was a dropped connection after DATA that had
5574 been sent in pipelining mode. That is, the server had advertised PIPELINING
5575 but was not implementing it correctly. The effect of the bug was incorrect
5576 behaviour, such as trying another host, and this could lead to a crash.
5582 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5583 and daemon.c were passed as pointers to ints; they should have been
5584 pointers to socklen_t variables (which are typically unsigned ints).
5586 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5589 3. Fixed a really odd bug that affected only the testing scheme; patching a
5590 certain fixed string in the binary changed the value of another string that
5591 happened to be identical to the end of the original first string.
5593 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5594 name", it returns that address as the IP address. On some operating
5595 systems (e.g. Solaris), it also passes back the IP address string as the
5596 "host name". However, on others (e.g. Linux), it passes back an empty
5597 string. Exim wasn't checking for this, and was changing the host name to an
5598 empty string, assuming it had been canonicized.
5600 5. Although rare, it is permitted to have more than one PTR record for a given
5601 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5602 all the names associated with an address, because they do in Solaris.
5603 However, it seems that they do not in Linux for data that comes from the
5604 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5605 I found this out when I moved to a new Linux workstation and tried to run
5606 the Exim test suite.
5608 To get round this problem I have changed the code so that it now does its
5609 own call to the DNS to look up PTR records when searching for a host name.
5610 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5611 addresses that are only in /etc/hosts are still found.
5613 This behaviour is, however, controlled by an option called host_lookup_
5614 order, which defaults to "bydns:byaddr". If people want to use the other
5615 order, or indeed, just use one or the other means of lookup, they can
5616 specify it in this variable.
5618 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5619 some operating systems, this comes back from gethostbyaddr() as an empty
5620 string, and this is what Exim used to test for. However, it seems that in
5621 other systems, "." is yielded. Exim now tests for this case too.
5623 7. The values of check_spool_space and check_log_space are now held internally
5624 as a number of kilobytes instead of an absolute number of bytes. If a
5625 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5626 kilobyte. This means that much larger values can be stored.
5628 8. Exim monitor: an attempt to get the action menu when not actually pointing
5629 at a message produces an empty menu entitled "No message selected". This
5630 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5631 no entries in it ("Shell widget menu has zero width and/or height"). So I
5632 have added a single, blank menu entry in this case.
5634 9. Added ${quote_local_part.
5636 10. MIME decoding is now applied to the contents of Subject: header lines when
5639 11. Now that a reference to $sender_host_address automatically causes a reverse
5640 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5641 host lookup before query-style lookups in lists that might use this
5642 variable. This has therefore been abolished, and the "net-" prefix is no
5643 longer necessary for query-style lookups.
5645 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5646 have been a typo for LFLAGS, so it has been changed.
5648 13. The install script calls Exim with "-C /dev/null" in order to find the
5649 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5650 to be output. Howeve, since Exim outputs its version number before the
5651 error, it didn't break the script. It just looked ugly. I fixed this by
5652 always allowing "-C /dev/null" if the caller is root.
5654 14. Ignore overlarge ACL variable number when reading spool file - insurance
5655 against a later release with more variables having written the file.
5657 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5658 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5659 this, as well as the form without the "IPv6" on the front.
5661 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5662 OS/Makefile-Darwin file.
5664 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5665 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5667 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5668 variable. However, some OS use ints for inodes. I've added cast to long int
5669 to get rid of the compiler warning.
5671 19. I had forgotten to lock out "/../" in configuration file names when
5672 ALT_CONFIG_PREFIX was set.
5674 20. Routers used for verification do not need to specify transports. However,
5675 if such a router generated a host list, and callout was configured, Exim
5676 crashed, because it could not find a port number from the (non-existent)
5677 transport. It now assumes port 25 in this circumstance.
5679 21. Added the -t option to exigrep.
5681 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5682 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5683 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5684 them is set, LOOKUP_LSEARCH is forced.
5686 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5688 24. Added sender and host information to the "rejected by local_scan()" log
5689 line; previously there was no indication of these.
5691 25. Added .include_if_exists.
5693 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5694 Linux. It turns out that not all file systems support this. Apparently some
5695 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5696 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5697 not supported on the file descriptor, is now ignored when Exim is trying to
5698 sync a directory. This applies only to Linux.
5700 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5702 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5703 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5706 29. Added support for maildirsize files from supplied patch (modified a bit).
5708 30. The use of :fail: followed by an empty string could lead Exim to respond to
5709 sender verification failures with (e.g.):
5711 550 Verification failed for <xxx>
5712 550 Sender verify failed
5714 where the first response line was missing the '-' that indicates it is not
5715 the final line of the response.
5717 31. The loop for finding the name of the user that called Exim had a hardwired
5718 limit of 10; it now uses the value of finduser_retries, which is used for
5719 all other user lookups.
5721 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5724 33. Exim was neglecting to zero errno before one call of strtol() when
5725 expanding a string and expecting an integer value. On some systems this
5726 resulted in spurious "integer overflow" errors. Also, it was casting the
5727 result into an int without checking.
5729 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5730 did not work. The code looks as if it has *never* worked, though it appears
5731 to have been documented since at least releast 1.62. I have made it work.
5733 35. The "timeout_DNS" error in retry rules, also documented since at least
5734 1.62, also never worked. As it isn't clear exactly what this means, and
5735 clearly it isn't a major issue, I have abolished the feature by treating it
5736 as "timeout", and writing a warning to the main and panic logs.
5738 36. The display of retry rules for -brt wasn't always showing the error code
5741 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5742 timeout_connect_A, timeout_connect_MX.
5744 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5745 to the empty sender.
5747 39. The daemon was not analysing the content of -oX till after it had closed
5748 stderr and disconnected from the controlling terminal. This meant that any
5749 syntax errors were only noted on the panic log, and the return code from
5750 the command was 0. By re-arranging the code a little, I've made the
5751 decoding happen first, so such errors now appear on stderr, and the return
5752 code is 1. However, the actual setting up of the sockets still happens in
5753 the disconnected process, so errors there are still only recorded on the
5756 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5757 connections (as happens on some IP stacks) was logged at start up time as
5758 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5759 it from "IPv6 and IPv4", which means that two separate sockets are being
5762 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5763 says whether AF_INET or AF_INET6 was passed as an argument.
5765 42. Exiwhat output was messed up when time zones were included in log
5768 43. Exiwhat now gives more information about the daemon's listening ports,
5769 and whether -tls-on-connect was used.
5771 44. The "port" option of the smtp transport is now expanded.
5773 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5774 silently ignored. Now an error message is written to the main and panic
5777 46. There's a new ACL modifier called "logwrite" which writes to a log file
5778 as soon as it is encountered.
5780 47. Added $local_user_uid and $local_user_gid at routing time.
5782 48. Exim crashed when trying to verify a sender address that was being
5785 49. Exim was recognizing only a space character after ".include". It now also
5786 recognizes a tab character.
5788 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5789 extracting the relevant information from the specification. The man page no
5790 longer contains scrambled data for the -d option, and I've added a section
5791 at the front about calling Exim under different names.
5793 51. Added "extra_headers" argument to the "mail" command in filter files.
5795 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5798 53. Installed eximstats 1.29.
5800 54. Added transport_filter_timeout as a generic transport option.
5802 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5803 Cc: header lines. This was required by RFC 822, but it not required by RFC
5806 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5807 incoming messages that did not have them. Now it does so only if the
5808 message originates locally, that is, if there is no associated remote host
5809 address. When Resent- header lines are present, this applies to the Resent-
5810 lines rather than the non-Resent- lines.
5812 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5813 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5815 58. Messages for configuration errors now include the name of the main
5816 configuration file - useful now that there may be more than one file in a
5817 list (.included file names were always shown).
5819 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5820 for those rare installations that do not start the daemon as root or run it
5821 setuid root. I've cut out the call to initgroups() if the daemon is not
5824 60. The Exim user and group can now be bound into the binary as text strings
5825 that are looked up at the start of Exim's processing.
5827 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5829 62. Added $mailstore_basename variable.
5831 63. Installed patch to sieve.c from Michael Haardt.
5833 64. When Exim failed to open the panic log after failing to open the main log,
5834 the original message it was trying to log was written to stderr and debug
5835 output, but if they were not available (the usual case in production), it
5836 was lost. Now it is written to syslog before the two lines that record the
5837 failures to open the logs.
5839 65. Users' Exim filters run in subprocesses under the user's uid. It is
5840 possible for a "deliver" command or an alias in a "personal" command to
5841 provoke an address rewrite. If logging of address rewriting is configured,
5842 this fails because the process is not running as root or exim. There may be
5843 a better way of dealing with this, but for the moment (because 4.30 needs
5844 to be released), I have disabled address rewrite logging when running a
5845 filter in a non-root, non-exim process.
5851 1. The buildconfig auxiliary program wasn't quoting the value set for
5852 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5853 not defined. This bug was masked in 4.22 by the effect that was fixed in
5856 2. Some messages that were rejected after a message id was allocated were
5857 shown as "incomplete" by exigrep. It no longer does this for messages that
5858 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5860 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5861 have allow_domain_literals set, the ID did not get logged in the <= line.
5862 Domain literals are now always recognized in Message-ID: header lines.
5864 4. The first argument for a ${extract expansion item is the key name or field
5865 number. Leading and trailing spaces in this item were not being ignored,
5866 causing some misleading effects.
5868 5. When deliver_drop_privilege was set, single queue runner processes started
5869 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5870 same command in the process it spins off) were not dropping privilege.
5872 6. When the daemon running as "exim" started a queue runner, it always
5873 re-executed Exim in the spun-off process. This is a waste of effort when
5874 deliver_drop_privilege is set. The new process now just calls the
5875 queue-runner function directly.
5881 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5884 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5885 ignored. Though the use of -oP was forcing the writing of a pid file, it
5886 was always written to the default place.
5888 3. If the message "no IP address found for host xxxx" is generated during
5889 incoming verification, it is now followed by identification of the incoming
5890 connection (so you can more easily find what provoked it).
5892 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5894 5. Added some features to "harden" Exim a bit more against certain attacks:
5896 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5897 be put in Local/Makefile. This is like the never_users runtime option,
5898 but it cannot be overridden. The default setting is "root".
5900 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5901 prefix string with which any file named in a -C command line option
5904 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5905 is retained for -C and -D only if the caller of Exim is root. Without
5906 it, the exim user may also use -C and -D and retain privilege.
5908 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5909 command line option is disabled.
5911 6. Macro names set by the -D option must start with an upper case letter, just
5912 like macro names defined in the configuration file.
5914 7. Added "dereference=" facility to LDAP.
5916 8. Two instances of the typo "uknown" in the source files are fixed.
5918 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5919 the Configure-Makefile script screwed up while processing it.
5921 10. Incorporated PCRE 4.4.
5923 11. The SMTP synchronization check was not operating right at the start of an
5924 SMTP session. For example, it could not catch a HELO sent before the client
5925 waited for the greeting. There is now a check for outstanding input at the
5926 point when the greeting is written. Because of the duplex, asynchronous
5927 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5928 way, but not yet received, when the check is performed.
5930 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5931 on TCP/IP sockets, because this apparently causes some broken clients to
5934 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5935 unchanged) from the Cygwin maintainer.
5937 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5938 was supported instead of testing for mbx. Effectively a typo.
5940 15. The spa authenticator server code was not checking that the input it
5941 received was valid base64.
5943 16. The debug output line for the "set" modifier in ACLs was not showing the
5944 name of the variable that was being set.
5946 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5948 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5949 lookup on its own. Something else was needed to trigger the lookup. For
5950 example, a match in host_lookup or the need for a host name in a host list.
5951 Now, if $sender_host_name is referenced and the host name has not yet been
5952 looked up, a lookup is performed. If the lookup fails, the variable remains
5953 empty, and $host_lookup_failed is set to "1".
5955 19. Added "eqi" as a case-independent comparison operator.
5957 20. The saslauthd authentication condition could segfault if neither service
5958 nor realm was specified.
5960 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5961 error message that was logged was misleading, and incoming SMTP
5962 connections were dropped. The message is now more accurate, and temporary
5963 errors are given to SMTP connections.
5965 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5966 (except RSET) with a 421 error, until QUIT is received. However, it was
5967 failing to send a response to QUIT.
5969 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5970 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5971 ACL is now run after the helo_try_verify_hosts code.
5973 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5974 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5975 case-independent, so other case variants are also recognized). Apparently
5976 some systems use these upper case variants.
5978 25. If more than two messages were waiting for the same host, and a transport
5979 filter was specified for the transport, Exim sent two messages over the
5980 same TCP/IP connection, and then failed with "socket operation on non-
5981 socket" when it tried to send the third.
5983 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5985 27. The extern definition of crypt16() in expand.c was not being excluded when
5986 the OS had its own crypt16() function.
5988 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5989 as a preferred synonym for return_size_limit, both as an option and as an
5992 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5994 30. Changed the default configuration ACL to relax the local part checking rule
5995 for addresses that are not in any local domains. For these addresses,
5996 slashes and pipe symbols are allowed within local parts, but the sequence
5997 /../ is explicitly forbidden.
5999 31. SPA server authentication was not clearing the challenge buffer before
6002 32. log_message in a "warn" ACL statement was writing to the reject log as
6003 well as to the main log, which contradicts the documentation and doesn't
6004 seem right (because no rejection is happening). So I have stopped it.
6006 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6007 However, I am unable to do any testing of this.
6009 34. Fixed an infelicity in the appendfile transport. When checking directories
6010 for a mailbox, to see if any needed to be created, it was accidentally
6011 using path names with one or more superfluous leading slashes; tracing
6012 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6014 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6015 contents are added to the log line that is written for every discarded
6016 recipient. (Previously a log_message setting was ignored.)
6018 36. The ${quote: operator now quotes the string if it is empty.
6020 37. The install script runs exim in order to find its version number. If for
6021 some reason other than non-existence or emptiness, which it checks, it
6022 could not run './exim', it was installing it with an empty version number,
6023 i.e. as "exim-". This error state is now caught, and the installation is
6026 38. An argument was missing from the function that creates an error message
6027 when Exim fails to connect to the socket for saslauthd authentication.
6028 This could cause Exim to crash, or give a corrupted message.
6030 39. Added isip, isip4, and isip6 to ${if conditions.
6032 40. The ACL variables $acl_xx are now saved with the message, and can be
6033 accessed later in routers, transports, and filters.
6035 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6036 strings in the file are not string-expanded.
6038 42. If a MAIL command specified a SIZE value that was too large to fit into an
6039 int variable, the check against message_size_limit failed. Such values are
6040 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6041 day this will have to be increased, but I don't think I want to be around
6042 when emails are that large.
6049 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6050 iconv() is not standard in FreeBSD.
6052 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6053 IPv6 enabled. The observed symptom was a segmentation fault on return from
6054 the function os_common_find_running_interfaces() in src/os.c.
6056 3. In the check_special_case() function in daemon.c I had used "errno" as an
6057 argument name, which causes warnings on some systems. This was basically a
6058 typo, since it was named "eno" in the comments!
6060 4. The code that waits for the clock to tick (at a resolution of some fraction
6061 of a second) so as to ensure message-id uniqueness was always waiting for
6062 at least one whole tick, when it could have waited for less. [This is
6063 almost certainly not relevant at current processor speeds, where it is
6064 unlikely to ever wait at all. But we try to future-proof.]
6066 5. The function that sleeps for a time interval that includes fractions of a
6067 second contained a race. It did not block SIGALRM between setting the
6068 timer, and suspending (a couple of lines later). If the interval was short
6069 and the sigsuspend() was delayed until after it had expired, the suspension
6070 never ended. On busy systems this could lead to processes getting stuck for
6073 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6074 process, before it forks any delivery processes. The open lookup caching
6075 mechanism meant that the open file or database connection was passed into
6076 the delivery process. The problem was that delivery processes always tidy
6077 up cached lookup data. This could cause a problem for the next delivery
6078 process started by the queue runner, because the external queue runner
6079 process does not know about the closure. So the next delivery process
6080 still has data in the lookup cache. In the case of a file lookup, there was
6081 no problem because closing a file descriptor in a subprocess doesn't affect
6082 the parent. However, if the lookup was caching a connection to a database,
6083 the connection was closed, and the second delivery process was likely to
6084 see errors such as "PGSQL: query failed: server closed the connection
6085 unexpectedly". The problem has been fixed by closing all cached lookups
6086 in a queue runner before running a delivery process.
6088 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6089 seem to have the "const" qualifier which it has on other OS. I've
6092 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6093 *of the same type* (client or server) with the same public name that an
6094 error should be diagnosed.
6096 9. When Exim looked up a host name for an IP address, but failed to find the
6097 original IP address when looking up the host name (a safety check), it
6098 output the message "<ip address> does not match any IP for NULL", which was
6099 confusing, to say the least. The bug was that the host name should have
6100 appeared instead of "NULL".
6102 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6103 user that is built into the binary, and the -C or -D options is used, root
6104 privilege is dropped before the configuration file is read. In addition,
6105 logging is switched to stderr instead of the normal log files. If the
6106 configuration then re-defines the Exim user, the unprivileged environment
6107 is probably not what is expected, so Exim logs a panic warning message (but
6110 However, if deliver_drop_privilege is set, the unprivileged state may well
6111 be exactly what is intended, so the warning has been cut out in that case,
6112 and Exim is allowed to try to write to its normal log files.
6118 1. smtp_return_error_details was not giving details for temporary sender
6119 or receiver verification errors.
6121 2. Diagnose a configuration error if two authenticators have the same public
6124 3. Exim used not to create the message log file for a message until the first
6125 delivery attempt. This could be confusing when incoming messages were held
6126 for policy or load reasons. The message log file is now created at the time
6127 the message is received, and an initial "Received" line is written to it.
6129 4. The automatically generated man page for command line options had a minor
6130 bug that caused no ill effects; however, a more serious problem was that
6131 the procedure for building the man page automatically didn't always
6132 operate. Consequently, release 4.20 contains an out-of-date version. This
6133 shouldn't happen again.
6135 5. When building Exim with embedded Perl support, the script that builds the
6136 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6137 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6139 6. The freeze_tell option was not being used for messages that were frozen on
6140 arrival, either by an ACL or by local_scan().
6142 7. Added the smtp_incomplete_transaction log selector.
6144 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6145 was accepting AUTH without a new EHLO.
6147 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6148 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6151 10. Logging of TCP/IP connections (when configured) now happens in the main
6152 daemon process instead of the child process, so that the TCP/IP connection
6153 count is more accurate (but it can never be perfect).
6155 11. The use of "drop" in a nested ACL was not being handled correctly in the
6156 outer ACL. Now, if condition failure induced by the nested "drop" causes
6157 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6158 or "require"), the connection is dropped.
6160 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6161 that yield "discard" can now be used with an "accept" or a "discard" verb,
6162 but an error is generated for any others (because I can't see a useful way
6163 to define what should happen).
6165 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6166 are now processed in the same way as lines in the Exim configuration file.
6167 In particular, continuation lines are supported.
6169 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6171 15. Added -ti meaning -t -i.
6173 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6174 domains, and warn by logging if others are found.
6176 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6177 returned from GIFCONF when Exim is trying to find the list of interfaces on
6178 a host. The code in os.c has been modified to copy each ifreq to an aligned
6179 structure in all cases.
6181 Also, in some cases, the returned ifreq's were being copied to a 'struct
6182 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6183 means the last couple of bytes of an IPv6 address could be chopped if the
6184 ifreq contained only a normal sockaddr (14 bytes storage).
6186 18. Named domain lists were not supported in the hosts_treat_as_local option.
6187 An entry such as +xxxx was not recognized, and was treated as a literal
6190 19. Ensure that header lines added by a DATA ACL are included in the reject log
6191 if the ACL subsequently rejects the message.
6193 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6194 MD5 (which is deprecated).
6196 21. When testing a filter file using -bf, Exim was writing a message when it
6197 took the sender from a "From " line in the message, but it was not doing so
6198 when it took $return_path from a Return-Path: header line. It now does.
6200 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6201 with a valid header line field name (a series of printing characters
6202 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6204 23. Changed "disc" in the source to "disk" to conform to the documentation and
6205 the book and for uniformity.
6207 24. Ignore Sendmail's -Ooption=value command line item.
6209 25. When execve() failed while trying to run a command in a pipe transport,
6210 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6211 could be confused with a return value of 69 from the command itself. This
6212 has been changed to 127, the value the shell returns if it is asked to run
6213 a non-existent command. The wording for the related log line suggests a
6214 non-existent command as the problem.
6216 26. If received_header_text expands to an empty string, do not add a Received:
6217 header line to the message. (Well, it adds a token one on the spool, but
6218 marks it "old" so that it doesn't get used or transmitted.)
6220 27. Installed eximstats 1.28 (addition of -nt option).
6222 28. There was no check for failure on the call to getsockname() in the daemon
6223 code. This can fail if there is a shortage of resources on the system, with
6224 ENOMEM, for example. A temporary error is now given on failure.
6226 29. Contrary to the C standard, it seems that in some environments, the
6227 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6228 program. Exim now does this explicitly; it affects the formatting of
6229 timestamps using strftime().
6231 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6232 complaints. I've now initialized all the variables, to avoid this.
6234 32. Header lines added by a system filter were not being "seen" during
6235 transport-time rewrites.
6237 33. The info_callback() function passed to OpenSSL is set up with type void
6238 (*)(SSL *, int, int), as described somewhere. However, when calling the
6239 function (actually a macro) that sets it up, the type void(*)() is
6240 expected. I've put in a cast to prevent warnings from picky compilers.
6242 34. If a DNS black list lookup found a CNAME record, but there were no A
6243 records associated with the domain it pointed at, Exim crashed.
6245 35. If a DNS black list lookup returned more than one A record, Exim ignored
6246 all but the first. It now scans all returned addresses if a particular IP
6247 value is being sought. In this situation, the contents of the
6248 $dnslist_value variable are a list of all the addresses, separated by a
6251 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6252 to accept a host name and all its aliases if the forward lookup for any of
6253 them yielded the IP address of the incoming connection. Now it accepts only
6254 those names whose forward lookup yields the correct IP address. Any other
6255 names are discarded. This closes a loophole whereby a rogue DNS
6256 administrator could create reverse DNS records to break through a
6257 wildcarded host restriction in an ACL.
6259 37. If a user filter or a system filter that ran in a subprocess used any of
6260 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6261 the wrong values were passed to the pipe command ($thisaddress had the
6262 value of $0, $0 had the value of $1, etc). This bug was introduced by
6263 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6265 38. Improved the line breaking for long SMTP error messages from ACLs.
6266 Previously, if there was no break point between 40 and 75 characters, Exim
6267 left the rest of the message alone. Two changes have been made: (a) I've
6268 reduced the minimum length to 35 characters; (b) if it can't find a break
6269 point between 35 and 75 characters, it looks ahead and uses the first one
6270 that it finds. This may give the occasional overlong line, but at least the
6271 remaining text gets split now.
6273 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6274 file descriptors might be low, and that setting 1000 would always raise it.
6275 It turns out that in some environments, the limit is already over 1000 and
6276 that lowering it causes trouble. So now Exim takes care not to decrease it.
6278 40. When delivering a message, the value of $return_path is set to $sender_
6279 address at the start of routing (routers may change the value). By an
6280 oversight, this default was not being set up when an address was tested by
6281 -bt or -bv, which affected the outcome if any router or filter referred to
6284 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6285 log only when "message" or "log_message" are set. However, if one of the
6286 conditions was an address verification, or a call to a nested ACL, the
6287 messages generated by the underlying test were being passed through. This
6288 no longer happens. The underlying message is available in $acl_verify_
6289 message for both "message" and "log_message" expansions, so it can be
6290 passed through if needed.
6292 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6293 new expansion $bh_ to give the encoded byte string without charset
6294 translation. Translation happens only if iconv() is available; HAVE_ICONV
6295 indicates this at build time. HEADERS_CHARSET gives the charset to
6296 translate to; headers_charset can change it in the configuration, and
6297 "headers charset" can change it in an individual filter file.
6299 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6300 that creates RFC 2047 encoded "words" labels them as that charset instead
6301 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6302 expansion operator; (ii) when Exim creates a From: line for a local
6303 message; (iii) when a header line is rewritten to include a "phrase" part.
6305 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6306 buggy, causing it to skip the first lines of messages whose message ID
6307 ended in 'D'. This would not have bitten before Exim release 4.14, because
6308 message IDs were unlikely to end in 'D' before then. The effect was to have
6309 incorrect size information for certain domains.
6311 45. #include "config.h" was missing at the start of the crypt16.c module. This
6312 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6315 46. If there was a timeout during a "random" callout check, Exim treated it as
6316 a failure of the random address, and carried on sending RSET and the real
6317 address. If the delay was just some slowness somewhere, the response to the
6318 original RCPT would be taken as a response to RSET and so on, causing
6319 mayhem of various kinds.
6321 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6322 when I implemented it. It didn't allow for the fact that some option values
6323 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6324 the right test for positive values.
6326 48. Domain names in DNS records are case-independent. Exim always looks them up
6327 in lower case. Some resolvers return domain names in exactly the case they
6328 appear in the zone file, that is, they may contain uppercase letters. Not
6329 all resolvers do this - some return always lower case. Exim was treating a
6330 change of case by a resolver as a change of domain, similar to a widening
6331 of a domain abbreviation. This triggered its re-routing code and so it was
6332 trying to route what was effectively the same domain again. This normally
6333 caused routing to fail (because the router wouldn't handle the domain
6334 twice). Now Exim checks for this case specially, and just changes the
6335 casing of the domain that it ultimately uses when it transmits the message
6338 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6341 50. If a filter generated a file delivery with a non-absolute name (possible if
6342 no home directory exists for the router), the forbid_file option was not
6345 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6346 the existing equality matching.
6348 52. Exim was using ints instead of ino_t variables in some places where it was
6349 dealing with inode numbers.
6351 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6352 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6353 called TMPDIR, and if it finds it is different, it changes its value.
6355 54. The smtp_printf() function is now made available to local_scan() so
6356 additional output lines can be written before returning. There is also an
6357 smtp_fflush() function to enable the detection of a dropped connection.
6358 The variables smtp_input and smtp_batched_input are exported to
6361 55. Changed the default runtime configuration: the message "Unknown user"
6362 has been removed from the ACL, and instead placed on the localuser router,
6363 using the cannot_route_message feature. This means that any verification
6364 failures that generate their own messages won't get overridden. Similarly,
6365 the "Unrouteable address" message that was in the ACL for unverifiable
6366 relay addresses has also been removed.
6368 56. Added hosts_avoid_esmtp to the smtp transport.
6370 57. The exicyclog script was not checking for the esoteric option
6371 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6372 will work only if exicyclog is run under the appropriate euid.
6374 58. Following a discussion on the list, the rules by which Exim recognises line
6375 endings on incoming messages have been changed. The -dropcr and drop_cr
6376 options are now no-ops, retained only for backwards compatibility. The
6377 following line terminators are recognized: LF CRLF CR. However, special
6378 processing applies to CR:
6380 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6381 nor a local message in the state where . is a terminator.
6383 (ii) If a bare CR is encountered in a header line, an extra space is added
6384 after the line terminator so as not to end the header. The reasoning
6385 behind this is that bare CRs in header lines are most likely either
6386 to be mistakes, or people trying to play silly games.
6388 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6389 was being incorrectly given as 18 bytes larger than it should have been.
6390 This is a VOB (very old bug).
6392 60. This may never have affected anything current, but just in case it has:
6393 When the local host is found other than at the start of a list of hosts,
6394 the local host, those with the same MX, and any that follow, are discarded.
6395 When the list in question was part of a longer list of hosts, the following
6396 hosts (not currently being processed) were also being discarded. This no
6397 longer happens. I'm not sure if this situation could ever has previously
6400 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6403 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6404 header. This is something that is recommended in a new Internet Draft, and
6405 is something that is documented as being done by Sendmail. There are two
6406 possible values. For messages generated by the autoreply transport, Exim
6409 Auto-Submitted: auto-replied
6411 whereas for all other generated messages (e.g. bounces) it adds
6413 Auto-Submitted: auto-generated
6415 63. The "personal" condition in filters now includes a test for the
6416 Auto-Submitted: header. If it contains the string "auto-" the message it
6417 not considered personal.
6419 64. Added rcpt_include_affixes as a generic transport option.
6421 65. Added queue_only_override (default true).
6423 66. Added the syslog_duplication option.
6425 67. If what should have been the first header line of a message consisted of
6426 a space followed by a colon, Exim was mis-interpreting it as a header line.
6427 It isn't of course - it is syntactically invalid and should therefore be
6428 treated as the start of the message body. The misbehaviour could have
6429 caused a number of strange effects, including loss of data in subsequent
6430 header lines, and spool format errors.
6432 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6433 client host had authenticated. This control can now be exercised by an ACL
6434 for more flexibility.
6436 69. By default, callouts do not happen when testing with -bh. There is now a
6437 variant, -bhc, which does actually run the callout code, including
6438 consulting and updating the callout cache.
6440 70. Added support for saslauthd authentication, courtesy of Alexander
6443 71. If statvfs() failed on the spool or log directories while checking their
6444 size for availability, Exim confusingly gave the error "space shortage".
6445 Furthermore, in debugging mode it crashed with a floating point exception.
6446 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6447 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6448 serious problem, Exim now writes to the main and panic logs when this
6449 happens, with details of the failure. It then refuses to accept the
6450 incoming message, giving the message "spool directory problem" or "log
6451 directory problem" with a 421 code for SMTP messages.
6453 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6454 0, 1, and 2 exist, because some OS complain for execs without them (see
6455 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6456 descriptors. However, the code omitted to check that the open succeeded,
6457 causing mysterious errors if for some reason the permissions on /dev/null
6458 got screwed. Now Exim writes a message to the main and panic logs, and
6459 bombs out if it can't open /dev/null.
6461 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6462 interact so that it is all more flexible. It is supposed to remain
6463 backwards compatible. Also added extra_local_interfaces.
6465 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6466 to bomb out with an assertion failure - to the client this appears as a
6467 connection drop. This problem occurs in the part of the code that was taken
6468 from the Samba project. Fortunately, the assertion is in a very simple
6469 function, so I have fixed this by reproducing the function inline in the
6470 one place where it is called, and arranging for authentication to fail
6471 instead of killing the process with assert().
6473 75. The SPA client code was not working when the server requested OEM rather
6474 than Unicode encoding.
6476 76. Added code to make require_files with a specific uid setting more usable in
6477 the case where statting the file as root fails - usually a non-root-mounted
6478 NFS file system. When this happens and the failure is EACCES, Exim now
6479 forks a subprocess and does the per-uid checking as the relevant uid.
6481 77. Added process_log_path.
6483 78. If log_file_path was not explicitly set, a setting of check_log_space or
6484 check_log_inodes was ignored.
6486 79. If a space check for the spool or log partitions fails, the incident is now
6487 logged. Of course, in the latter case the data may get lost...
6489 80. Added the %p formatting code to string_format() so that it can be used to
6490 print addresses in debug_print(). Adjusted all the address printing in the
6491 debugging in store.c to use %p rather than %d.
6493 81. There was a concern that a line of code in smtp_in.c could overflow a
6494 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6495 initially expressed, the concern was not well-founded, because trailing
6496 spaces are removed early. However, if the trailing spaces were followed by
6497 a NULL, they did not get removed, so the overflow was possible. Two fixes
6500 (a) I re-wrote the offending code in a cleaner fashion.
6501 (b) If an incoming SMTP command contains a NULL character, it is rejected
6504 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6505 runs initgroups(), so that if the Exim user is in any additional groups,
6506 they will be used during message reception.
6512 The change log for 4.20 and earlier releases has been archived.