1 Change log file for Exim from version 4.21
2 -------------------------------------------
6 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
7 supports it and a size is available (ie. the sending peer gave us one).
9 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
10 years of being deprecated). The replacements are the ACLs
11 acl_smtp_mime and acl_not_smtp_mime.
13 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
14 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
15 or even an in-clear connection were permitted. Now, if the host lookup
16 was dnssec and dane was requested then the host is only used if the
17 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
18 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
19 if one fails this test.
20 This means that a poorly-configured remote DNS will make it incommunicado;
21 but it protects against a DNS-interception attack on it.
23 JH/04 Bug 1810: make continued-use of an open smtp transport connection
24 non-noisy when a race steals the message being considered.
26 JH/05 If main configuration option tls_certificate is unset, generate a
27 selfsigned certificate for inbound TLS connections.
29 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
30 in rewrites and routers.
32 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
33 and logged a warning sing 4.83; now they are a configuration file error.
35 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
36 (lacking @domain). Apply the same qualification processing as RCPT.
41 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
42 and 3.4.4 - once the server is enabled to respond to an OCSP request
43 it does even when not requested, resulting in a stapling non-aware
44 client dropping the TLS connection.
46 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
47 support variable-length bit vectors. No functional change.
49 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
50 The I= interface field on outgoing lines is now after the H= remote
51 host field, same as incoming lines. There is a separate
52 outgoing_interface log selector which allows you to disable the
55 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
56 If not running log_selector +smtp_connection the mainlog would be held
57 open indefinitely after a "too many connections" event, including to a
58 deleted file after a log rotate. Leave the per net connection logging
59 leaving it open for efficiency as that will be quickly detected by the
60 check on the next write.
62 HS/01 Bug 1671: Fix post transport crash.
63 Processing the wait-<transport> messages could crash the delivery
64 process if the message IDs didn't exist for some reason. When
65 using 'split_spool_directory=yes' the construction of the spool
66 file name failed already, exposing the same netto behaviour.
68 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
69 mime_regex ACL conditions.
71 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
72 to DSN fail messages (bounces): remote IP, remote greeting, remote response
73 to HELO, local diagnostic string.
75 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
76 log line to debug. Even when configured with a tls authenticator many
77 client connections are expected to not authenticate in this way, so
78 an authenticate fail is not an error.
80 HS/02 Add the Exim version string to the process info. This way exiwhat
81 gives some more detail about the running daemon.
83 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
84 matter for fast-change records such as DNSBLs.
86 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
87 retry record, even if constant. There may be multiple transports with
88 different interface settings and the retry behaviour needs to be kept
91 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
93 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
95 JH/10 Bug 840: fix log_defer_output option of pipe transport
97 JH/11 Bug 830: use same host for all RCPTS of a message, even under
98 hosts_randomize. This matters a lot when combined with mua_wrapper.
100 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
101 ${quote_pgsql:<string>} operator.
103 JH/13 Bug 1708: avoid misaligned access in cached lookup.
105 JH/14 Change header file name for freeradius-client. Relevant if compiling
106 with Radius support; from the Gentoo tree and checked under Fedora.
108 JH/15 Bug 1712: Introduce $prdr_requested flag variable
110 JH/16 Bug 1714: Permit an empty string as expansion result for transport
111 option transport_filter, meaning no filtering.
113 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
115 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
116 defaults to "*" (all hosts). The variable is now available when not built
117 with TLS, default unset, mainly to enable keeping the testuite sane.
118 If a server certificate is not supplied (via tls_certificate) an error is
119 logged, and clients will find TLS connections fail on startup. Presumably
120 they will retry in-clear.
121 Packagers of Exim are strongly encouraged to create a server certificate
122 at installation time.
124 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
125 with the $config_file variable.
127 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
128 in transport context, after the attempt, and per-recipient. The latter type
129 is per host attempted. The event data is the error message, and the errno
130 information encodes the lookup type (A vs. MX) used for the (first) host,
131 and the trailing two digits of the smtp 4xx reponse.
133 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
134 to write to mainlog (or rejectlog, paniclog) in the window between file
135 creation and permissions/ownership being changed. Particularly affects
136 installations where exicyclog is run as root, rather than exim user;
137 result is that the running daemon panics and dies.
139 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
141 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
142 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
143 "pri" and "weight". Note that the previous implicit priority given by the
144 list order is no longer honoured.
146 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
149 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
150 by defining SUPPORT_SOCKS.
152 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
153 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
154 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
155 variables are renamed to proxy_{local,external}_{address,port}.
157 JH/27 Move Internationalisation support from Experimental to mainline, enabled
158 for a build by defining SUPPORT_I18N
160 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
161 of the query string, and make ${quote_redis:} do that quoting.
163 JH/29 Move Events support from Experimental to mainline, enabled by default
164 and removable for a build by defining DISABLE_EVENT.
166 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
168 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
169 cached by the daemon.
171 JH/32 Move Redis support from Experimental to mainline, enabled for a build
172 by defining LOOKUP_REDIS. The libhiredis library is required.
174 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
175 keys are given for lookup.
177 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
178 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
179 only supported when built with TLS support. The PolarSSL SHA routines
180 are still used when the TLS library is too old for convenient support.
182 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
183 openssl_options), for security. OpenSSL forces this from version 1.1.0
184 server-side so match that on older versions.
186 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
187 allocation for $value could be released as the expansion processing
188 concluded, but leaving the global pointer active for it.
190 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
191 and to use the domains and local_parts ACL conditions.
193 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
194 incorrectly not doubled on cutthrough transmission, hence seen as a
195 body-termination at the receiving system - resulting in truncated mails.
196 Commonly the sender saw a TCP-level error, and retransmitted the message
197 via the normal store-and-forward channel. This could result in duplicates
198 received - but deduplicating mailstores were liable to retain only the
199 initial truncated version.
201 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
203 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
205 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
206 we're in there, support oversigning also; bug 1309.
208 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
210 HS/04 Add support for keep_environment and add_environment options.
212 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
213 either intentional arithmetic overflow during PRNG, or testing config-
216 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
217 delivery resulted in actual delivery. Cancel cutthrough before DATA
220 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
221 rejects a recipient: pass the reject to the originator.
223 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
224 Many were false-positives and ignorable, but it's worth fixing the
227 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
228 for the new environment-manipulation done at startup. Move the routines
229 from being local to tls.c to being global via the os.c file.
231 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
232 an extract embedded as result-arg for a map, the first arg for extract
233 is unavailable so we cannot tell if this is a numbered or keyed
234 extraction. Accept either.
240 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
243 JH/02 The smtp transport option "multi_domain" is now expanded.
245 JH/03 The smtp transport now requests PRDR by default, if the server offers
248 JH/04 Certificate name checking on server certificates, when exim is a client,
249 is now done by default. The transport option tls_verify_cert_hostnames
250 can be used to disable this per-host. The build option
251 EXPERIMENTAL_CERTNAMES is withdrawn.
253 JH/05 The value of the tls_verify_certificates smtp transport and main options
254 default to the word "system" to access the system default CA bundle.
255 For GnuTLS, only version 3.0.20 or later.
257 JH/06 Verification of the server certificate for a TLS connection is now tried
258 (but not required) by default. The verification status is now logged by
259 default, for both outbound TLS and client-certificate supplying inbound
262 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
265 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
266 Status Notification (bounce) messages are now MIME format per RFC 3464.
267 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
268 under the control of the dsn_advertise_hosts option, and routers may
269 have a dsn_lasthop option.
271 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
272 default, modifiable by a malware= option. The list separator for
273 the options can now be changed in the usual way. Bug 68.
275 JH/10 The smtp_receive_timeout main option is now expanded before use.
277 JH/11 The incoming_interface log option now also enables logging of the
278 local interface on delivery outgoing connections.
280 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
281 if the interface and destination host and port all match.
283 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
286 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
287 Patch from Andrew Lewis.
289 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
290 now supports optional time-restrictions, weighting, and priority
291 modifiers per server. Patch originally by <rommer@active.by>.
293 JH/16 The spamd_address main option now supports a mixed list of local
294 and remote servers. Remote servers can be IPv6 addresses, and
295 specify a port-range.
297 JH/17 Bug 68: The spamd_address main option now supports an optional
298 timeout value per server.
300 JH/18 Bug 1581: Router and transport options headers_add/remove can
301 now have the list separator specified.
303 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
306 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
309 JH/21 Support for the A6 type of dns record is withdrawn.
311 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
312 rather than the verbs used.
314 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
315 from 255 to 1024 chars.
317 JH/24 Verification callouts now attempt to use TLS by default.
319 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
320 are generic router options now. The defaults didn't change.
322 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
323 Original patch from Alexander Shikoff, worked over by JH.
325 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
326 files if ZCAT_COMMAND is not executable.
328 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
330 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
332 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
333 Normally benign, it bites when the pair was led to by a CNAME;
334 modern usage is to not canoicalize the domain to a CNAME target
335 (and we were inconsistent anyway for A-only vs AAAA+A).
337 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
339 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
340 when evaluating $sender_host_dnssec.
342 JH/31 Check the HELO verification lookup for DNSSEC, adding new
343 $sender_helo_dnssec variable.
345 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
347 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
349 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
351 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
352 documented as working, but never had. Support all but $spam_report.
354 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
355 added for tls authenticator.
357 HS/03 Add perl_taintmode main config option
362 TL/01 When running the test suite, the README says that variables such as
363 no_msglog_check are global and can be placed anywhere in a specific
364 test's script, however it was observed that placement needed to be near
365 the beginning for it to behave that way. Changed the runtest perl
366 script to read through the entire script once to detect and set these
367 variables, reset to the beginning of the script, and then run through
368 the script parsing/test process like normal.
370 TL/02 The BSD's have an arc4random API. One of the functions to induce
371 adding randomness was arc4random_stir(), but it has been removed in
372 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
373 function when detected.
375 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
376 cause callback expansion.
378 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
379 syntax errors in an expansion can be treated as a string instead of
380 logging or causing an error, due to the internal use of bool_lax
381 instead of bool when processing it.
383 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
384 server certificates when making smtp deliveries.
386 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
388 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
390 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
392 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
393 Merged patch from Sebastian Wiedenroth.
395 JH/05 Fix results-pipe from transport process. Several recipients, combined
396 with certificate use, exposed issues where response data items split
397 over buffer boundaries were not parsed properly. This eventually
398 resulted in duplicates being sent. This issue only became common enough
399 to notice due to the introduction of conection certificate information,
400 the item size being so much larger. Found and fixed by Wolfgang Breyha.
402 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
403 size buffer was used, resulting in syntax errors when an expansion
406 JH/07 Add support for directories of certificates when compiled with a GnuTLS
407 version 3.3.6 or later.
409 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
410 is EXPERIMENTAL_EVENT, the main-configuration and transport options
411 both become "event_action", the variables become $event_name, $event_data
412 and $event_defer_errno. There is a new variable $verify_mode, usable in
413 routers, transports and related events. The tls:cert event is now also
414 raised for inbound connections, if the main configuration event_action
417 TL/06 In test suite, disable OCSP for old versions of openssl which contained
418 early OCSP support, but no stapling (appears to be less than 1.0.0).
420 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
421 server certificate names available under the smtp transport option
422 "tls_verify_cert_hostname" now do not permit multi-component wildcard
425 JH/10 Time-related extraction expansions from certificates now use the main
426 option "timezone" setting for output formatting, and are consistent
427 between OpenSSL and GnuTLS compilations. Bug 1541.
429 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
430 encoded parameter in the incoming message. Bug 1558.
432 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
433 include certificate info, eximon was claiming there were spoolfile
436 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
438 JH/14 Log delivery-related information more consistently, using the sequence
439 "H=<name> [<ip>]" wherever possible.
441 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
442 are problematic for Debian distribution, omit them from the release
445 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
447 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
450 JH/17 Fix a null-indirection in certextract expansions when a nondefault
451 output list separator was used.
456 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
457 checkers that were complaining about end of non-void function with no
460 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
461 This was a regression intruduced in 4.83 by another bugfix.
463 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
465 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
466 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
472 TF/01 Correctly close the server side of TLS when forking for delivery.
474 When a message was received over SMTP with TLS, Exim failed to clear up
475 the incoming connection properly after forking off the child process to
476 deliver the message. In some situations the subsequent outgoing
477 delivery connection happened to have the same fd number as the incoming
478 connection previously had. Exim would try to use TLS and fail, logging
479 a "Bad file descriptor" error.
481 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
482 utilities have not been installed.
484 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
485 temporary space as the ACL may create new global variables.
487 TL/01 LDAP support uses per connection or global context settings, depending
488 upon the detected version of the libraries at build time.
490 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
491 to extract and use the src ip:port in logging and expansions as if it
492 were a direct connection from the outside internet. PPv2 support was
493 updated based on HAProxy spec change in May 2014.
495 JH/02 Add ${listextract {number}{list}{success}{fail}}.
497 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
498 Properly escape header and check for NULL return.
500 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
503 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
505 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
506 characters in header names, implemented as a verify condition.
507 Contributed by Michael Fischer v. Mollard.
509 TL/05 Rename SPF condition results err_perm and err_temp to standardized
510 results permerror and temperror. Previous values are deprecated but
511 still accepted. In a future release, err_perm and err_temp will be
512 completely removed, which will be a backward incompatibility if the
513 ACL tests for either of these two old results. Patch contributed by
514 user bes-internal on the mailing list.
516 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
518 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
519 selectors, in both main and reject logs.
521 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
524 JH/07 Add malware type "sock" for talking to simple daemon.
526 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
528 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
529 routers/transports under cutthrough routing.
531 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
532 numbers. Touch up "bool" conditional to keep the same definition.
534 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
536 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
538 JH/12 Expand items in router/transport headers_add or headers_remove lists
539 individually rather than the list as a whole. Bug 1452.
541 Required for reasonable handling of multiple headers_ options when
542 they may be empty; requires that headers_remove items with embedded
543 colons must have them doubled (or the list-separator changed).
545 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
546 view the policy declared in the DMARC record. Currently, $dmarc_status
547 is a combined value of both the record presence and the result of the
550 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
552 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
553 dnslookup router and the smtp transport (applying to the forward
556 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
557 of ldap servers used for a specific lookup. Patch provided by Heiko
560 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
561 New variable $lookup_dnssec_authenticated for observability.
563 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
564 Patch submitted by Lars Timman.
566 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
568 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
569 Requires trusted mode and valid format message id, aborts otherwise.
570 Patch contributed by Heiko Schlichting.
572 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
573 certextract with support for various fields. Bug 1358.
575 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
576 is requested by default, modifiable by smtp transport option
579 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
580 operate on certificate variables to give certificate fingerprints
581 Also new ${sha256:cert_variable}.
583 JH/23 The PRDR feature is moved from being Experimental into the mainline.
585 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
586 Christian Aistleitner.
588 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
590 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
591 file. Patch from Wolfgang Breyha.
593 JH/25 Expand the coverage of the delivery $host and $host_address to
594 client authenticators run in verify callout. Bug 1476.
596 JH/26 Port service names are now accepted for tls_on_connect_ports, to
597 align with daemon_smtp_ports. Bug 72.
599 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
600 support and error reporting did not work properly.
602 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
603 and is readable. Patch from Andrew Colin Kissa.
605 TL/14 Enhance documentation of ${run expansion and how it parses the
606 commandline after expansion, particularly in the case when an
607 unquoted variable expansion results in an empty value.
609 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
611 PP/02 Fix internal collision of T_APL on systems which support RFC3123
612 by renaming away from it. Addresses GH issue 15, reported by
615 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
617 TL/15 SECURITY: prevent double expansion in math comparison functions
618 (can expand unsanitized data). Not remotely exploitable.
625 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
627 PP/02 Make -n do something, by making it not do something.
628 When combined with -bP, the name of an option is not output.
630 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
633 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
634 $sender_host_name and config options to manage this, and basic check
637 PP/05 DSCP support for outbound connections and control modifier for inbound.
639 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
640 (Only plugin which currently uses this is kerberos4, which nobody should
641 be using, but we should make it available and other future plugins might
642 conceivably use it, even though it would break NAT; stuff *should* be
643 using channel bindings instead).
645 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
646 name; added for Sendmail compatibility; requires admin caller.
647 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
648 just ignore it); requires trusted caller.
649 Also parse but ignore: -Ac -Am -X<logfile>
652 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
654 TL/02 Add +smtp_confirmation as a default logging option.
656 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
657 Patch by Magnus Holmgren from 2007-02-20.
659 TL/04 Bugzilla 1281 - Spec typo.
660 Bugzilla 1283 - Spec typo.
661 Bugzilla 1290 - Spec grammar fixes.
663 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
665 TL/06 Add Experimental DMARC support using libopendmarc libraries.
667 TL/07 Fix an out of order global option causing a segfault. Reported to dev
668 mailing list by by Dmitry Isaikin.
670 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
672 JH/02 Support "G" suffix to numbers in ${if comparisons.
674 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
676 NM/01 Bugzilla 1197 - Spec typo
677 Bugzilla 1196 - Spec examples corrections
679 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
681 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
682 gnutls_enable_pkcs11, but renamed to more accurately indicate its
685 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
686 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
688 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
689 "acl {{name}{arg}...}", and optional args on acl condition
692 JH/05 Permit multiple router/transport headers_add/remove lines.
694 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
696 JH/07 Avoid using a waiting database for a single-message-only transport.
697 Performance patch from Paul Fisher. Bugzilla 1262.
699 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
702 JH/09 Add $headers_added variable, with content from use of ACL modifier
703 add_header (but not yet added to the message). Bugzilla 199.
705 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
706 Pulled from Bugzilla 817 by Wolfgang Breyha.
708 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
710 (nb: this is the same fix as in Exim 4.80.1)
712 JH/11 Add A= logging on delivery lines, and a client_set_id option on
715 JH/12 Add optional authenticated_sender logging to A= and a log_selector
718 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
720 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
721 advertise SMTP AUTH mechanism to us, instead of a generic
722 protocol violation error. Also, make Exim more robust to bad
723 data from the Dovecot auth socket.
725 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
727 When a queue runner is handling a message, Exim first routes the
728 recipient addresses, during which it prunes them based on the retry
729 hints database. After that it attempts to deliver the message to
730 any remaining recipients. It then updates the hints database using
733 So if a recipient address works intermittently, it can get repeatedly
734 deferred at routing time. The retry hints record remains fresh so the
735 address never reaches the final cutoff time.
737 This is a fairly common occurrence when a user is bumping up against
738 their storage quota. Exim had some logic in its local delivery code
739 to deal with this. However it did not apply to per-recipient defers
740 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
742 This change adds a proper retry rule check during routing so that the
743 final cutoff time is checked against the message's age. We only do
744 this check if there is an address retry record and there is not a
745 domain retry record; this implies that previous attempts to handle
746 the address had the retry_use_local_parts option turned on. We use
747 this as an approximation for the destination being like a local
748 delivery, as in LMTP.
750 I suspect this new check makes the old local delivery cutoff check
751 redundant, but I have not verified this so I left the code in place.
753 TF/02 Correct gecos expansion when From: is a prefix of the username.
755 Test 0254 submits a message to Exim with the header
759 When I ran the test suite under the user fanf2, Exim expanded
760 the header to contain my full name, whereas it should have added
761 a Resent-Sender: header. It erroneously treats any prefix of the
762 username as equal to the username.
764 This change corrects that bug.
766 GF/01 DCC debug and logging tidyup
767 Error conditions log to paniclog rather than rejectlog.
768 Debug lines prefixed by "DCC: " to remove any ambiguity.
770 TF/03 Avoid unnecessary rebuilds of lookup-related code.
772 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
773 Bug spotted by Jeremy Harris; was flawed since initial commit.
774 Would have resulted in OCSP responses post-SNI triggering an Exim
775 NULL dereference and crash.
777 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
779 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
780 Bug detection, analysis and fix by Samuel Thibault.
781 Bugzilla 1331, Debian bug #698092.
783 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
785 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
786 Server implementation by Todd Lyons, client by JH.
787 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
788 config variable "prdr_enable" controls whether the server
789 advertises the facility. If the client requests PRDR a new
790 acl_data_smtp_prdr ACL is called once for each recipient, after
791 the body content is received and before the acl_smtp_data ACL.
792 The client is controlled by bolth of: a hosts_try_prdr option
793 on the smtp transport, and the server advertisement.
794 Default client logging of deliveries and rejections involving
795 PRDR are flagged with the string "PRDR".
797 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
798 fclose(). Diagnosis by Todd Lyons.
800 PP/17 Update configure.default to handle IPv6 localhost better.
801 Patch by Alain Williams (plus minor tweaks).
804 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
805 This is now consistent with GnuTLS, and is now documented: the
806 previous undocumented portable approach to treating the option as
807 unset was to force an expansion failure. That still works, and
808 an empty string is now equivalent.
810 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
811 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
812 not performing validation itself.
814 PP/20 Added force_command boolean option to pipe transport.
815 Patch from Nick Koston, of cPanel Inc.
817 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
820 TF/04 Added udpsend ACL modifer and hexquote expansion operator
822 PP/21 Fix eximon continuous updating with timestamped log-files.
823 Broken in a format-string cleanup in 4.80, missed when I repaired the
824 other false fix of the same issue.
825 Report and fix from Heiko Schlichting.
828 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
829 Report from Prashanth Katuri.
831 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
832 It's SecureTransport, so affects any MacOS clients which use the
833 system-integrated TLS libraries, including email clients.
835 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
836 using a MIME ACL for non-SMTP local injection.
837 Report and assistance in diagnosis by Warren Baker.
839 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
841 JH/16 Fix comparisons for 64b. Bugzilla 1385.
843 TL/09 Add expansion variable $authenticated_fail_id to keep track of
844 last id that failed so it may be referenced in subsequent ACL's.
846 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
849 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
850 ldap library initialization, allowing self-signed CA's to be
851 used. Also properly sets require_cert option later in code by
852 using NULL (global ldap config) instead of ldap handle (per
853 session). Bug diagnosis and testing by alxgomz.
855 TL/12 Enhanced documentation in the ratelimit.pl script provided in
856 the src/util/ subdirectory.
858 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
859 renamed to Transport Post Delivery Action by Jeremy Harris, as
862 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
863 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
864 redis_servers = needs to be configured which will be used by the redis
865 lookup. Patch from Warren Baker, of The Packet Hub.
867 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
869 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
870 hostname or reverse DNS when processing a host list. Used suggestions
871 from multiple comments on this bug.
873 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
875 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
876 Missed a few lines, added it to make the runtest require no keyboard
879 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
880 contains upper case chars. Make router use caseful_local_part.
882 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
883 support when GnuTLS has been built with p11-kit.
889 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
891 This, or similar/improved, will also be change PP/11 of 4.82.
897 PP/01 Handle short writes when writing local log-files.
898 In practice, only affects FreeBSD (8 onwards).
899 Bugzilla 1053, with thanks to Dmitry Isaikin.
901 NM/01 Bugzilla 949 - Documentation tweak
903 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
906 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
908 PP/02 Implemented gsasl authenticator.
910 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
912 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
913 `pkg-config foo` for cflags/libs.
915 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
916 with rest of GSASL and with heimdal_gssapi.
918 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
919 `pkg-config foo` for cflags/libs for the TLS implementation.
921 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
922 properties get this fed in as external SSF. A number of robustness
923 and debugging improvements to the cyrus_sasl authenticator.
925 PP/08 cyrus_sasl server now expands the server_realm option.
927 PP/09 Bugzilla 1214 - Log authentication information in reject log.
928 Patch by Jeremy Harris.
930 PP/10 Added dbmjz lookup type.
932 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
934 PP/12 MAIL args handles TAB as well as SP, for better interop with
935 non-compliant senders.
936 Analysis and variant patch by Todd Lyons.
938 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
939 Bug report from Lars Müller <lars@samba.org> (via SUSE),
940 Patch from Dirk Mueller <dmueller@suse.com>
942 PP/13 tls_peerdn now print-escaped for spool files.
943 Observed some $tls_peerdn in wild which contained \n, which resulted
944 in spool file corruption.
946 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
947 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
948 or write after TLS renegotiation, which otherwise led to messages
951 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
952 as a tracking header (ie: a signed header comes before the signature).
953 Patch from Wolfgang Breyha.
955 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
956 comma-sep list; embedded commas doubled.
958 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
960 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
962 Report and patch from Dmitry Banschikov.
964 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
965 Removed SSL_clear() after SSL_new() which led to protocol negotiation
966 failures. We appear to now support TLS1.1+ with Exim.
968 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
969 lets Exim select keys and certificates based upon TLS SNI from client.
970 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
971 before an outbound SMTP session. New log_selector, +tls_sni.
973 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
974 NULL dereference. Report and patch from Alun Jones.
976 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
977 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
978 Not seeing resolver debug output on NetBSD, but suspect this is a
979 resolver implementation change.
981 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
982 Left warnings. Added "eximon gdb" invocation mode.
984 PP/21 Defaulting "accept_8bitmime" to true, not false.
986 PP/22 Added -bw for inetd wait mode support.
988 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
989 locate the relevant includes and libraries. Made this the default.
991 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
992 Bugzilla 1246, report and most of solution from Tomasz Kusy.
994 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
995 This may cause build issues on older platforms.
997 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
998 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
999 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1000 Added SNI support via GnuTLS too.
1001 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1003 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1005 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1006 Applied second patch from Janne, implementing suggestion to default
1007 multiple-strings-in-record handling to match SPF spec.
1009 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1011 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1012 read-only, out of scope).
1013 Patch from Wolfgang Breyha, report from Stuart Northfield.
1015 PP/29 Fix three issues highlighted by clang analyser static analysis.
1016 Only crash-plausible issue would require the Cambridge-specific
1017 iplookup router and a misconfiguration.
1018 Report from Marcin Mirosław.
1020 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1022 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1023 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1024 As part of this, removing so much warning spew let me fix some minor
1025 real issues in debug logging.
1027 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1028 assignment on my part. Fixed.
1030 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1031 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1032 Janne Snabb (who went above and beyond: thank you).
1034 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1035 string otherwise requires a connection and a bunch more work and it's
1036 relatively easy to get wrong. Should also expose TLS library linkage
1039 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1040 64-bit ${eval} (JH/03).
1042 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1043 GNU libc to support some of the 64-bit stuff, should not lead to
1044 conflicts. Defined before os.h is pulled in, so if a given platform
1045 needs to override this, it can.
1047 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1048 protection layer was required, which is not implemented.
1049 Bugzilla 1254, patch from Wolfgang Breyha.
1051 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1052 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1053 tls_dhparam take prime identifiers. Also unbreak combination of
1054 OpenSSL+DH_params+TLSSNI.
1056 PP/39 Disable SSLv2 by default in OpenSSL support.
1062 PP/01 Solaris build fix for Oracle's LDAP libraries.
1063 Bugzilla 1109, patch from Stephen Usher.
1065 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1067 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1070 TF/02 Fix a couple more cases where we did not log the error message
1071 when unlink() failed. See also change 4.74-TF/03.
1073 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1074 lock up or crash if it happened to be inside a call to libc when it
1075 got a SIGUSR1 from exiwhat.
1077 The SIGUSR1 handler appends the current process status to the process
1078 log which is later printed by exiwhat. It used to use the general
1079 purpose logging code to do this, but several functions it calls are
1080 not safe for signals.
1082 The new output code in the SIGUSR1 handler is specific to the process
1083 log, and simple enough that it's easy to inspect for signal safety.
1084 Removing some special cases also simplifies the general logging code.
1085 Removing the spurious timestamps from the process log simplifies
1088 TF/04 Improved ratelimit ACL condition.
1090 The /noupdate option has been deprecated in favour of /readonly which
1091 has clearer semantics. The /leaky, /strict, and /readonly update modes
1092 are mutually exclusive. The update mode is no longer included in the
1093 database key; it just determines when the database is updated. (This
1094 means that when you upgrde Exim will forget old rate measurements.)
1096 Exim now checks that the per_* options are used with an update mode that
1097 makes sense for the current ACL. For example, when Exim is processing a
1098 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1099 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1100 must specify per_mail/readonly. If you omit the update mode it defaults to
1101 /leaky where that makes sense (as before) or /readonly where required.
1103 The /noupdate option is now undocumented but still supported for
1104 backwards compatibility. It is equivalent to /readonly except that in
1105 ACLs where /readonly is required you may specify /leaky/noupdate or
1106 /strict/noupdate which are treated the same as /readonly.
1108 A useful new feature is the /count= option. This is a generalization
1109 of the per_byte option, so that you can measure the throughput of other
1110 aggregate values. For example, the per_byte option is now equivalent
1111 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1113 The per_rcpt option has been generalized using the /count= mechanism
1114 (though it's more complicated than the per_byte equivalence). When it is
1115 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1116 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1117 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1118 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1119 Note that using per_rcpt with a non-readonly update mode in more than
1120 one ACL will cause the recipients to be double-counted. (The per_mail
1121 and per_byte options don't have this problem.)
1123 The handling of very low rates has changed slightly. If the computed rate
1124 is less than the event's count (usually one) then this event is the first
1125 after a long gap. In this case the rate is set to the same as this event's
1126 count, so that the first message of a spam run is counted properly.
1128 The major new feature is a mechanism for counting the rate of unique
1129 events. The new per_addr option counts the number of different
1130 recipients that someone has sent messages to in the last time period. It
1131 behaves like per_rcpt if all the recipient addresses are different, but
1132 duplicate recipient addresses do not increase the measured rate. Like
1133 the /count= option this is a general mechanism, so the per_addr option
1134 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1135 example, measure the rate that a client uses different sender addresses
1136 with the options per_mail/unique=$sender_address. There are further
1137 details in the main documentation.
1139 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1141 TF/06 Removed a few PCRE remnants.
1143 TF/07 Automatically extract Exim's version number from tags in the git
1144 repository when doing development or release builds.
1146 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1147 Bugzilla 879. Patch from Paul Fisher.
1149 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1150 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1153 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1155 PP/05 Variable $av_failed, true if the AV scanner deferred.
1156 Bugzilla 1078. Patch from John Horne.
1158 PP/06 Stop make process more reliably on build failure.
1159 Bugzilla 1087. Patch from Heiko Schlittermann.
1161 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1162 Bugzilla 1089. Patch from Heiko Schlittermann.
1164 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1165 Bugzilla 1131. Patch from Holger Weiß.
1167 PP/09 Handle IPv6 addresses with SPF.
1168 Bugzilla 860. Patch from Wolfgang Breyha.
1170 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1172 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1175 PP/11 match_* no longer expand right-hand-side by default.
1176 New compile-time build option, EXPAND_LISTMATCH_RHS.
1177 New expansion conditions, "inlist", "inlisti".
1179 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1181 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1183 PP/14 fix log_write() format string regression from TF/03.
1184 Bugzilla 1152. Patch from Dmitry Isaikin.
1190 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1192 PP/02 Harmonised TLS library version reporting; only show if debugging.
1193 Layout now matches that introduced for other libraries in 4.74 PP/03.
1195 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1197 PP/04 New "dns_use_edns0" global option.
1199 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1202 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1203 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1205 TK/01 Updated PolarSSL code to 0.14.2.
1206 Bugzilla 1097. Patch from Andreas Metzler.
1208 PP/07 Catch divide-by-zero in ${eval:...}.
1209 Fixes bugzilla 1102.
1211 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1214 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1215 format-string attack -- SECURITY: remote arbitrary code execution.
1217 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1218 time unintentionally subject to list matching rules, letting the header
1219 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1220 arbitrary string expansion). This allowed for information disclosure.
1222 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1223 INT_MIN/-1 -- value coerced to INT_MAX.
1229 NM/01 Workround for PCRE version dependency in version reporting
1232 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1233 This fixes portability to compilers other than gcc, notably
1234 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1236 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1237 makefiles for portability to HP-UX and POSIX correctness.
1239 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1240 Also via indented variable definition in the Makefile.
1241 (Debugging by Oliver Heesakkers).
1243 PP/02 Restore caching of spamd results with expanded spamd_address.
1244 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1246 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1247 Improves build reliability. Fix from: Frank Elsner
1249 NM/02 Fix wide character breakage in the rfc2047 coding
1250 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1252 NM/03 Allow underscore in dnslist lookups
1253 Fixes bug 1026. Patch from Graeme Fowler
1255 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1256 Code patches from Adam Ciarcinski of NetBSD.
1258 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1261 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1262 is logged, to avoid truncation. Patch from John Horne.
1264 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1265 Patch from Jakob Hirsch.
1267 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1268 SQL string expansion failure details.
1269 Patch from Andrey Oktyabrski.
1271 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1272 Patch from Simon Arlott.
1274 PP/09 New lookups functionality failed to compile on old gcc which rejects
1275 extern declarations in function scope.
1276 Patch from Oliver Fleischmann
1278 PP/10 Use sig_atomic_t for flags set from signal handlers.
1279 Check getgroups() return and improve debugging.
1280 Fixed developed for diagnosis in bug 927 (which turned out to be
1283 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1284 Patch from Mark Zealey.
1286 PP/12 Bugzilla 1056: Improved spamd server selection.
1287 Patch from Mark Zealey.
1289 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1290 Based on patch from Heiko Schlittermann.
1292 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1293 Patch from Uwe Doering, sign-off by Michael Haardt.
1295 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1296 variable declaration deep within a block. Bug and patch from
1299 PP/15 lookups-Makefile IRIX compatibilty coercion.
1301 PP/16 Make DISABLE_DKIM build knob functional.
1303 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1304 Patch by Simon Arlott
1306 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1307 variable argument macros. Our copy now differs from upstream.
1313 TF/01 Failure to get a lock on a hints database can have serious
1314 consequences so log it to the panic log.
1316 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1317 controlled using the smtp_confirmation log selector.
1319 TF/03 Include the error message when we fail to unlink a spool file.
1321 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1322 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1323 for maintaining out-of-tree patches for some time.
1325 PP/01 Bugzilla 139: Documentation and portability issues.
1326 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1327 Handle per-OS dynamic-module compilation flags.
1329 PP/02 Let /dev/null have normal permissions.
1330 The 4.73 fixes were a little too stringent and complained about the
1331 permissions on /dev/null. Exempt it from some checks.
1332 Reported by Andreas M. Kirchwitz.
1334 PP/03 Report version information for many libraries, including
1335 Exim version information for dynamically loaded libraries. Created
1336 version.h, now support a version extension string for distributors
1337 who patch heavily. Dynamic module ABI change.
1339 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1340 privilege escalation vulnerability whereby the Exim run-time user
1341 can cause root to append content of the attacker's choosing to
1344 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1347 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1348 If dropping privileges for untrusted macros, we disabled normal logging
1349 on the basis that it would fail; for the Exim run-time user, this is not
1350 the case, and it resulted in successful deliveries going unlogged.
1351 Fixed. Reported by Andreas Metzler.
1357 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1358 only prepend for the Resent-* case. Fixes regression introduced in
1359 Exim 4.70 by NM/22 for Bugzilla 607.
1361 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1362 increasing numbers of administrators be bitten by this.
1364 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1366 PP/03 Bugzilla 994: added openssl_options main configuration option.
1368 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1370 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1372 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1374 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1375 without a peer certificate, leading to a segfault because of an
1376 assumption that peers always have certificates. Be a little more
1377 paranoid. Problem reported by Martin Tscholak.
1379 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1380 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1381 NB: ClamAV planning to remove STREAM in "middle of 2010".
1382 CL also introduces -bmalware, various -d+acl logging additions and
1383 more caution in buffer sizes.
1385 PP/09 Implemented reverse_ip expansion operator.
1387 PP/10 Bugzilla 937: provide a "debug" ACL control.
1389 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1391 PP/12 Bugzilla 973: Implement --version.
1393 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1395 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1397 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1399 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1400 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1401 ignore trailing whitespace.
1403 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1405 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1408 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1409 Notification from Dr Andrew Aitchison.
1411 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1412 ExtendedDetectionInfo response format.
1413 Notification from John Horne.
1415 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1418 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1419 XSL and documented dependency on system catalogs, with examples of how
1422 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1425 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1426 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1427 configuration file which is writeable by the Exim user or group.
1429 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1430 of configuration files to cover files specified with the -C option if
1431 they are going to be used with root privileges, not just the default
1434 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1435 option (effectively making it always true).
1437 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1438 files to be used while preserving root privileges.
1440 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1441 that rogue child processes cannot use them.
1443 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1444 run-time user, instead of root.
1446 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1447 Exim run-time user without dropping privileges.
1449 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1450 result string, instead of calling string_vformat() twice with the same
1453 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1454 for other users. Others should always drop root privileges if they use
1455 -C on the command line, even for a whitelisted configure file.
1457 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1459 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1465 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1466 $data_path, and $header_path variables; fixed documentation bugs and
1469 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1470 exipick to access non-standard spools, including the "frozen" queue
1473 NM/01 Bugzilla 965: Support mysql stored procedures.
1474 Patch from Alain Williams
1476 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1478 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1479 Patch from Andreas Metzler
1481 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1482 Patch from Kirill Miazine
1484 NM/05 Bugzilla 671: Added umask to procmail example.
1486 JJ/03 installed exipick 20100323.0, fixing doc bug
1488 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1489 directory. Notification and patch from Dan Rosenberg.
1491 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1493 TK/02 Improve log output when DKIM signing operation fails.
1495 MH/01 Treat the transport option dkim_domain as a colon separated
1496 list, not as a single string, and sign the message with each element,
1497 omitting multiple occurences of the same signer.
1499 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1500 Bugzilla 985, 986. Patch by Simon Arlott
1502 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1503 Patch by Simon Arlott
1505 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1506 MBX locking. Notification from Dan Rosenberg.
1512 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1514 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1516 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1518 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1520 NM/04 Bugzilla 915: Flag broken perl installation during build.
1526 TK/01 Added patch by Johannes Berg that expands the main option
1527 "spamd_address" if it starts with a dollar sign.
1529 TK/02 Write list of recipients to X-Envelope-Sender header when building
1530 the mbox-format spool file for content scanning (suggested by Jakob
1533 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1534 (http://www.dcc-servers.net/) support via dccifd. Activated by
1535 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1537 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1538 by Mark Daniel Reidel <mr@df.eu>.
1540 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1541 When building exim an external PCRE library is now needed -
1542 PCRE is a system library on the majority of modern systems.
1543 See entry on PCRE_LIBS in EDITME file.
1545 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1546 conversation. Added nologin parameter to request.
1547 Patch contributed by Kirill Miazine.
1549 TF/01 Do not log submission mode rewrites if they do not change the address.
1551 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1553 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1554 log files in place. Contributed by Roberto Lima.
1556 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1558 TF/03 Bugzilla 615: When checking the local_parts router precondition
1559 after a local_part_suffix or local_part_prefix option, Exim now
1560 does not use the address's named list lookup cache, since this
1561 contains cached lookups for the whole local part.
1563 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1564 Robert Millan. Documentation is in experimental-spec.txt.
1566 TF/04 Bugzilla 668: Fix parallel build (make -j).
1568 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1570 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1571 Patch provided by Jan Srzednicki.
1573 TF/05 Leading white space used to be stripped from $spam_report which
1574 wrecked the formatting. Now it is preserved.
1576 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1577 that they are available at delivery time.
1579 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1581 TF/08 TLS error reporting now respects the incoming_interface and
1582 incoming_port log selectors.
1584 TF/09 Produce a more useful error message if an SMTP transport's hosts
1585 setting expands to an empty string.
1587 NM/06 Bugzilla 744: EXPN did not work under TLS.
1588 Patch provided by Phil Pennock.
1590 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1591 Patch provided by Richard Godbee.
1593 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1594 acl_smtp_notquit, added index entry.
1596 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1597 Patch provided by Eugene Bujak.
1599 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1600 accept(). Patch provided by Maxim Dounin.
1602 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1603 Patch provided by Phil Pennock.
1605 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1607 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1608 Patch provided by Brad "anomie" Jorsch.
1610 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1611 Patch provided by Dean Brooks.
1613 TK/05 Add native DKIM support (does not depend on external libraries).
1615 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1616 Patch provided by Graeme Fowler.
1618 NM/16 Bugzilla 851: Documentation example syntax fix.
1620 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1622 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1625 NM/19 Bugzilla 745: TLS version reporting.
1626 Patch provided by Phil Pennock.
1628 NM/20 Bugzilla 167: bool: condition support.
1629 Patch provided by Phil Pennock.
1631 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1632 clients. Patch provided by Phil Pennock.
1634 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1635 Patch provided by Brad "anomie" Jorsch.
1637 NM/23 Bugzilla 687: Fix misparses in eximstats.
1638 Patch provided by Heiko Schlittermann.
1640 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1641 Patch provided by Heiko Schlittermann.
1643 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1644 plus update to original patch.
1646 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1648 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1649 Patch provided by David Brownlee.
1651 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1653 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1655 NM/30 Bugzilla 888: TLS documentation bugfixes.
1657 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1659 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1660 Unlike the original bugzilla I have changed all shell scripts in src tree.
1662 NM/33 Bugzilla 898: Transport filter timeout fix.
1663 Patch by Todd Rinaldo.
1665 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1666 Patch by Serge Demonchaux.
1668 NM/35 Bugzilla 39: Base64 decode bug fixes.
1669 Patch by Jakob Hirsch.
1671 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1673 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1675 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1677 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1683 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1684 ALT-N's libdkim that I have put here:
1685 http://duncanthrax.net/exim-experimental/
1687 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1688 were called 'true' and it seems that C99 defines that as a reserved
1689 keyword to be used with 'bool' variable types. That means you could
1690 not include C99-style headers which use bools without triggering
1691 build errors in sieve.c.
1693 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1694 as mailq or other aliases. Changed the --help handling significantly
1695 to do whats expected. exim_usage() emits usage/help information.
1697 SC/01 Added the -bylocaldomain option to eximstats.
1699 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1701 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1703 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1709 PH/01 Another patch from the Sieve maintainer.
1711 PH/02 When an IPv6 address is converted to a string for single-key lookup
1712 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1713 dots are used instead of colons so that keys in lsearch files need not
1714 contain colons. This was done some time before quoting was made available
1715 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1716 (notated using the quote facility) so as to distinguish them from IPv4
1717 keys. This meant that lookups for IP addresses in host lists did not work
1718 for iplsearch lookups.
1720 This has been fixed by arranging for IPv6 addresses to be expressed with
1721 colons if the lookup type is iplsearch. This is not incompatible, because
1722 previously such lookups could never work.
1724 The situation is now rather anomolous, since one *can* have colons in
1725 ordinary lsearch keys. However, making the change in all cases is
1726 incompatible and would probably break a number of configurations.
1728 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1731 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1732 conversion specification without a maximum field width, thereby enabling
1733 a rogue spamd server to cause a buffer overflow. While nobody in their
1734 right mind would setup Exim to query an untrusted spamd server, an
1735 attacker that gains access to a server running spamd could potentially
1736 exploit this vulnerability to run arbitrary code as the Exim user.
1738 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1739 $primary_hostname instead of what libspf2 thinks the hosts name is.
1741 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1742 a directory entry by the name of the lookup key. Previously, if a
1743 symlink pointed to a non-existing file or a file in a directory that
1744 Exim lacked permissions to read, a lookup for a key matching that
1745 symlink would fail. Now it is enough that a matching directory entry
1746 exists, symlink or not. (Bugzilla 503.)
1748 PH/03 The body_linecount and body_zerocount variables are now exported in the
1751 PH/04 Added the $dnslist_matched variable.
1753 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1754 This means they are set thereafter only if the connection becomes
1757 PH/06 Added the client_condition to authenticators so that some can be skipped
1758 by clients under certain conditions.
1760 PH/07 The error message for a badly-placed control=no_multiline_responses left
1761 "_responses" off the end of the name.
1763 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1765 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1766 (without spaces) instead of just copying the configuration text.
1768 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1770 PH/11 Added $max_received_linelength.
1772 PH/12 Added +ignore_defer and +include_defer to host lists.
1774 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1775 way in which PCRE > 7.0 is built.
1777 PH/14 Implemented queue_only_load_latch.
1779 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1780 MAIL command. The effect was to mangle the value on 64-bit systems.
1782 PH/16 Another patch from the Sieve maintainer.
1784 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1786 PH/18 If a system quota error occurred while trying to create the file for
1787 a maildir delivery, the message "Mailbox is full" was not appended to the
1788 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1789 only to a quota excession during the actual writing of the file.
1791 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1792 characters?) which causes problems in log lines. The DN values are now
1793 passed through string_printing() before being added to log lines.
1795 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1796 and InterBase are left for another time.)
1798 PH/21 Added message_body_newlines option.
1800 PH/22 Guard against possible overflow in moan_check_errorcopy().
1802 PH/23 POSIX allows open() to be a macro; guard against that.
1804 PH/24 If the recipient of an error message contained an @ in the local part
1805 (suitably quoted, of course), incorrect values were put in $domain and
1806 $local_part during the evaluation of errors_copy.
1812 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1813 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1816 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1817 issue a MAIL command.
1819 PH/02 In an ACL statement such as
1821 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1823 if a client was not listed at all, or was listed with a value other than
1824 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1825 the condition was not true (as it should be), so access was not denied.
1826 The bug was that the ! inversion was incorrectly passed on to the second
1827 item. This has been fixed.
1829 PH/03 Added additional dnslists conditions == and =& which are different from
1830 = and & when the dns lookup returns more than one IP address.
1832 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1833 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1835 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1836 FSYNC, which compiles an option called disable_fsync that allows for
1837 bypassing fsync(). The documentation is heavily laced with warnings.
1839 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1841 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1842 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1843 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1844 including adding "make clean"; (3) Added -fPIC when compiling the test
1845 dynamically loaded module, to get rid of a warning.
1847 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1848 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1849 = 0s. The bug is otherwise harmless.
1851 PH/07 There was a bug in the dovecot authenticator such that the value of
1852 $auth1 could be overwritten, and so not correctly preserved, after a
1853 successful authentication. This usually meant that the value preserved by
1854 the server_setid option was incorrect.
1856 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1858 PH/09 Installed PCRE release 7.0.
1860 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1861 run for batched SMTP input. It is now run at the start of every message
1862 in the batch. While fixing this I discovered that the process information
1863 (output by running exiwhat) was not always getting set for -bs and -bS
1864 input. This is fixed, and it now also says "batched" for BSMTP.
1866 PH/11 Added control=no_pipelining.
1868 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1869 patch, slightly modified), and move the expansion of helo_data till after
1870 the connection is made in the smtp transport (so it can use these
1873 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1875 PH/14 Added log_selector = +pid.
1877 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1879 PH/16 Add ${if forany and ${if forall.
1881 PH/17 Added dsn_from option to vary the From: line in DSNs.
1883 PH/18 Flush SMTP output before performing a callout, unless control =
1884 no_callout_flush is set.
1886 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1887 was true (the default) a successful delivery failed to delete the retry
1888 item, thus causing premature timeout of the address. The bug is now
1891 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1893 PH/21 Long custom messages for fakedefer and fakereject are now split up
1894 into multiline reponses in the same way that messages for "deny" and
1895 other ACL rejections are.
1897 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1898 with slight modification.
1900 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1901 draft, changing the syntax and factoring some duplicate code".
1903 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1904 for deliveries of the second and subsequent messages over the same SMTP
1907 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1908 ${reduce, with only minor "tidies".
1910 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1912 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1913 expansion side effects.
1915 PH/27 When a message times out after an over-quota error from an Exim-imposed
1916 quota, the bounce message says "mailbox is full". This message was not
1917 being given when it was a system quota that was exceeded. It now should
1920 MH/03 Made $recipients available in local_scan(). local_scan() already has
1921 better access to the recipient list through recipients_list[], but
1922 $recipients can be useful in postmaster-provided expansion strings.
1924 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1925 in the case of a MAIL command with additional options following the
1926 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1927 were accidentally chopped off.
1929 PH/29 SMTP synchronization checks are implemented when a command is read -
1930 there is a check that no more input is waiting when there shouldn't be
1931 any. However, for some commands, a delay in an ACL can mean that it is
1932 some time before the response is written. In this time, more input might
1933 arrive, invalidly. So now there are extra checks after an ACL has run for
1934 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1935 pipelining has not been advertised.
1937 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1939 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1940 returned in the SMTP response when smtp_return_error_details was false.
1941 This has been fixed.
1943 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1944 instead of the C I/O that was originally supplied, because problems were
1945 reported on Solaris.
1947 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1948 Exim which did not show up earlier: it was assuming that a call to
1949 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1950 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1951 was a macro that became an assignment, so it seemed to work. This has
1952 changed to a proper function call with a void return, hence the compile
1953 error. Exim's code has been fixed.
1955 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1958 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1960 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1962 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1963 because it turns out to be too big or there is a timeout) there is no
1964 "Completed" line in the log. When some messages of this type were
1965 selected by exigrep, they were listed as "not completed". Others were
1966 picked up by some special patterns. I have improved the selection
1967 criteria to be more general.
1969 PH/38 The host_find_failed option in the manualroute router can now be set
1970 to "ignore", to completely ignore a host whose IP address cannot be
1971 found. If all hosts are ignored, the behaviour is controlled by the new
1972 host_all_ignored option.
1974 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1975 homing or because of multiple MX records with /mx) generated more than
1976 one IP address, and the following item turned out to be the local host,
1977 all the secondary addresses of the first item were incorrectly removed
1978 from the list, along with the local host and any following hosts (which
1979 is what is supposed to happen).
1981 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1982 whoever called Exim into the -H file. In the case of the daemon it was
1983 behaving confusingly. When first started, it used values for whoever
1984 started the daemon, but after a SIGHUP it used the Exim user (because it
1985 calls itself on a restart). I have changed the code so that it now always
1988 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1989 message are rejected with the same error (e.g. no authentication or bad
1990 sender address), and a DATA command is nevertheless sent (as can happen
1991 with PIPELINING or a stupid MUA), the error message that was given to the
1992 RCPT commands is included in the rejection of the DATA command. This is
1993 intended to be helpful for MUAs that show only the final error to their
1996 PH/42 Another patch from the Sieve maintainer.
1998 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1999 Eximstats - Fixed some broken HTML links and added missing column headers
2001 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2002 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2004 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2006 PH/43 Yet another patch from the Sieve maintainer.
2008 PH/44 I found a way to check for a TCP/IP connection going away before sending
2009 the response to the final '.' that terminates a message, but only in the
2010 case where the client has not sent further data following the '.'
2011 (unfortunately, this is allowed). However, in many cases there won't be
2012 any further data because there won't be any more messages to send. A call
2013 to select() can be used: if it shows that the input is "ready", there is
2014 either input waiting, or the socket has been closed. An attempt to read
2015 the next input character can distinguish the two cases. Previously, Exim
2016 would have sent an OK response which the client would never have see.
2017 This could lead to message repetition. This fix should cure that, at
2018 least in a lot of common cases.
2020 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2021 advertised in response to EHLO.
2027 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2028 fixed by 4.65/MH/01 (is this a record?) are fixed:
2030 (i) An empty string was always treated as zero by the numeric comparison
2031 operators. This behaviour has been restored.
2033 (ii) It is documented that the numeric comparison operators always treat
2034 their arguments as decimal numbers. This was broken in that numbers
2035 starting with 0 were being interpreted as octal.
2037 While fixing these problems I realized that there was another issue that
2038 hadn't been noticed. Values of message_size_limit (both the global option
2039 and the transport option) were treated as octal if they started with 0.
2040 The documentation was vague. These values are now always treated as
2041 decimal, and I will make that clear in the documentation.
2047 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2048 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2051 MH/01 Don't check that the operands of numeric comparison operators are
2052 integers when their expansion is in "skipping" mode (fixes bug
2053 introduced by 4.64-PH/07).
2055 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2056 child addresses, Exim now panics and dies. Previously, because the count
2057 is held in a short int, deliveries were likely to be lost. As such a
2058 large number of recipients for a single message is ridiculous
2059 (performance will be very, very poor), I have chosen to impose a limit
2060 rather than extend the field.
2066 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2067 leftover -K file (the existence of which was triggered by #402).
2068 While we were at it, introduced process PID as part of the -K
2069 filename. This should rule out race conditions when creating
2072 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2073 processing considerably. Previous code took too long for large mails,
2074 triggering a timeout which in turn triggers #401.
2076 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2077 in the DK code in transports.c. sendfile() is not really portable,
2078 hence the _LINUX specificness.
2080 TF/01 In the add_headers option to the mail command in an Exim filter,
2081 there was a bug that Exim would claim a syntax error in any
2082 header after the first one which had an odd number of characters
2085 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2086 callout verification, Exim cached a "reject" for the entire domain. This
2087 is correct for most verifications, but it is not correct for a recipient
2088 verification with use_sender or use_postmaster set, because in that case
2089 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2090 case of MAIL FROM:<> rejection from other early rejections (e.g.
2091 rejection of HELO). When verifying a recipient using a non-null MAIL
2092 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2093 Whatever the result of the callout, the value of the domain cache is
2094 left unchanged (for any other kind of callout, getting as far as trying
2095 RCPT means that the domain itself is ok).
2097 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2100 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2101 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2102 session, the daemon ignores ECONNECT errors and logs others; it now
2103 ignores EPIPE as well.
2105 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2106 (quoted-printable decoding).
2108 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2109 later the small subsequent patch to fix an introduced bug.
2111 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2113 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2115 PH/08 An error is now given if message_size_limit is specified negative.
2117 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2118 to be given (somewhat) arbitrary names.
2120 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2123 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2124 miscellaneous code fixes
2126 PH/10 Added the log_reject_target ACL modifier to specify where to log
2129 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2130 hostname. This is wrong, because it relates to the incoming message (and
2131 probably the interface on which it is arriving) and not to the outgoing
2132 callout (which could be using a different interface). This has been
2133 changed to use the value of the helo_data option from the smtp transport
2134 instead - this is what is used when a message is actually being sent. If
2135 there is no remote transport (possible with a router that sets up host
2136 addresses), $smtp_active_hostname is used.
2138 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2139 tweaks were necessary in order to get it to work (see also 21 below):
2140 (a) The code assumed that strncpy() returns a negative number on buffer
2141 overflow, which isn't the case. Replaced with Exim's string_format()
2143 (b) There were several signed/unsigned issues. I just did the minimum
2144 hacking in of casts. There is scope for a larger refactoring.
2145 (c) The code used strcasecmp() which is not a standard C function.
2146 Replaced with Exim's strcmpic() function.
2147 (d) The code set only $1; it now sets $auth1 as well.
2148 (e) A simple test gave the error "authentication client didn't specify
2149 service in request". It would seem that Dovecot has changed its
2150 interface. Fortunately there's a specification; I followed it and
2151 changed what the client sends and it appears to be working now.
2153 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2156 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2157 address A is aliased to B and C, where B exists and C does not. Without
2158 -v the output is "A verified" because verification stops after a
2159 successful redirection if more than one address is generated. However,
2160 with -v the child addresses are also verified. Exim was outputting "A
2161 failed to verify" and then showing the successful verification for C,
2162 with its parentage. It now outputs "B failed to verify", showing B's
2163 parentage before showing the successful verification of C.
2165 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2166 look up a TXT record in a specific list after matching in a combined
2169 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2170 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2171 they consult the DNS. I had assumed they would set it the way they
2172 wanted; and indeed my experiments on Linux seem to show that in some
2173 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2174 To be on the safe side, however, I have now made the interface to
2175 host_find_byname() similar to host_find_bydns(), with an argument
2176 containing the DNS resolver options. The host_find_byname() function now
2177 sets these options at its start, just as host_find_bydns() does. The smtp
2178 transport options dns_qualify_single and dns_search_parents are passed to
2179 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2180 of host_find_byname() use the default settings of RES_DEFNAMES
2181 (qualify_single) but not RES_DNSRCH (search_parents).
2183 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2184 spool_read_header() do less string testing, by means of a preliminary
2185 switch on the second character of optional "-foo" lines. (This is
2186 overdue, caused by the large number of possibilities that now exist.
2187 Originally there were few.) While I was there, I also converted the
2188 str(n)cmp tests so they don't re-test the leading "-" and the first
2189 character, in the hope this might squeeze out yet more improvement.
2191 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2192 flag allowing group syntax was set by the header_syntax check but not
2193 turned off, possible causing trouble later; (2) The flag was not being
2194 set at all for the header_verify test, causing "group"-style headers to
2195 be rejected. I have now set it in this case, and also caused header_
2196 verify to ignore an empty address taken from a group. While doing this, I
2197 came across some other cases where the code for allowing group syntax
2198 while scanning a header line wasn't quite right (mostly, not resetting
2199 the flag correctly in the right place). These bugs could have caused
2200 trouble for malformed header lines. I hope it is now all correct.
2202 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2203 with the "reply" argument non-NULL. The code, however (which originally
2204 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2205 but it didn't always do it. This confused somebody who was copying the
2206 code for some other use. I have removed all the tests.
2208 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2209 feature that was used to support insecure browsers during the U.S. crypto
2210 embargo. It requires special client support, and Exim is probably the
2211 only MTA that supported it -- and would never use it because real RSA is
2212 always available. This code has been removed, because it had the bad
2213 effect of slowing Exim down by computing (never used) parameters for the
2214 RSA_EXPORT functionality.
2216 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2217 authenticator to fail if there's a tab character in the incoming data
2218 (there should never be unless someone is messing about, as it's supposed
2219 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2220 if the connection is using TLS or if the remote IP is the same as the
2221 local IP, and the "valid-client-cert option" if a client certificate has
2224 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2225 authenticators. This can be used for authorization after authentication
2226 succeeds. (In the case of plaintext, it servers for both authentication
2229 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2230 if any retry times were supplied.
2232 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2233 connection, where there is no client IP address to check. In this
2234 situation, the verify now always succeeds.
2236 PH/25 Applied John Jetmore's -Mset patch.
2238 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2240 PH/27 In a string expansion for a processed (not raw) header when multiple
2241 headers of the same name were present, leading whitespace was being
2242 removed from all of them, but trailing whitespace was being removed only
2243 from the last one. Now trailing whitespace is removed from each header
2244 before concatenation. Completely empty headers in a concatenation (as
2245 before) are ignored.
2247 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2248 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2250 PH/29 [Removed. This was a change that I later backed out, and forgot to
2251 correct the ChangeLog entry (that I had efficiently created) before
2252 committing the later change.]
2254 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2255 address errors (4xx response to RCPT) over the same connection as other
2256 messages routed to the same hosts. Such deliveries are always "forced",
2257 so retry times are not inspected. This resulted in far too many retries
2258 for the affected addresses. The effect occurred only when there were more
2259 hosts than the hosts_max_try setting in the smtp transport when it had
2260 the 4xx errors. Those hosts that it had tried were not added to the list
2261 of hosts for which the message was waiting, so if all were tried, there
2262 was no problem. Two fixes have been applied:
2264 (i) If there are any address or message errors in an SMTP delivery, none
2265 of the hosts (tried or untried) are now added to the list of hosts
2266 for which the message is waiting, so the message should not be a
2267 candidate for sending over the same connection that was used for a
2268 successful delivery of some other message. This seems entirely
2269 reasonable: after all the message is NOT "waiting for some host".
2270 This is so "obvious" that I'm not sure why it wasn't done
2271 previously. Hope I haven't missed anything, but it can't do any
2272 harm, as the worst effect is to miss an optimization.
2274 (ii) If, despite (i), such a delivery is accidentally attempted, the
2275 routing retry time is respected, so at least it doesn't keep
2276 hammering the server.
2278 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2279 in ${readsocket because some servers need this prod.
2281 PH/32 Added some extra debug output when updating a wait-xxx database.
2283 PH/33 The hint "could be header name not terminated by colon", which has been
2284 given for certain expansion errors for a long time, was not being given
2285 for the ${if def:h_colon_omitted{... case.
2287 PH/34 The spec says: "With one important exception, whenever a domain list is
2288 being scanned, $domain contains the subject domain." There was at least
2289 one case where this was not true.
2291 PH/35 The error "getsockname() failed: connection reset by peer" was being
2292 written to the panic log as well as the main log, but it isn't really
2293 panic-worthy as it just means the connection died rather early on. I have
2294 removed the panic log writing for the ECONNRESET error when getsockname()
2297 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2298 runs only) independently of the message's sender address. This meant
2299 that, if the 4xx error was in fact related to the sender, a different
2300 message to the same recipient with a different sender could confuse
2301 things. In particualar, this can happen when sending to a greylisting
2302 server, but other circumstances could also provoke similar problems.
2303 I have changed the default so that the retry time for these errors is now
2304 based a combination of the sender and recipient addresses. This change
2305 can be overridden by setting address_retry_include_sender=false in the
2308 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2309 remote server are returned as part of bounce messages. This was not
2310 happening for LMTP over a pipe (the lmtp transport), but now it is the
2311 same for both kinds of LMTP.
2313 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2314 in header lines that were in fact CNAMEs. This is no longer the case.
2316 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2317 and queue runs started by the daemon processed all messages. This has
2318 been fixed so that -R and -S can now usefully be given with -q<time>.
2320 PH/40 Import PCRE release 6.7 (fixes some bugs).
2322 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2324 PH/42 Give an error if -q is specified more than once.
2326 PH/43 Renamed the variables $interface_address and $interface_port as
2327 $received_ip_address and $received_port, to make it clear that these
2328 values apply to message reception, and not to the outgoing interface when
2329 a message is delivered. (The old names remain recognized, of course.)
2331 PH/44 There was no timeout on the connect() call when using a Unix domain
2332 socket in the ${readsocket expansion. There now is.
2334 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2335 be meaningful with "accept".
2337 SC/01 Eximstats V1.43
2338 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2340 SC/02 Eximstats V1.44
2341 Use a glob alias rather than an array ref in the generated
2342 parser. This improves both readability and performance.
2344 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2345 Collect SpamAssassin and rejection statistics.
2346 Don't display local sender or destination tables unless
2347 there is data to show.
2348 Added average volumes into the top table text output.
2350 SC/04 Eximstats V1.46
2351 Collect data on the number of addresses (recipients)
2352 as well as the number of messages.
2354 SC/05 Eximstats V1.47
2355 Added 'Message too big' to the list of mail rejection
2356 reasons (thanks to Marco Gaiarin).
2358 SC/06 Eximstats V1.48
2359 Mainlog lines which have GMT offsets and are too short to
2360 have a flag are now skipped.
2362 SC/07 Eximstats V1.49 (Alain Williams)
2363 Added the -emptyok flag.
2365 SC/08 Eximstats V1.50
2366 Fixes for obtaining the IP address from reject messages.
2368 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2369 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2370 whitesspace changes from 4.64-PH/27
2372 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2375 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2376 are found, allow negative numbers in numeric criteria)
2378 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2380 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2381 to match changes made in 4.64-PH/43
2383 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2385 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2386 "domain" is actually an IP address, and force a failure. This locks out
2387 those revolvers/nameservers that support "A-for-A" lookups, in
2388 contravention of the specifications.
2390 PH/48 When a host name was looked up from an IP address, and the subsequent
2391 forward lookup of the name timed out, the host name was left in
2392 $sender_host_name, contrary to the specification.
2394 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2395 restricted to single-key lookups, Exim was not diagnosing an error if
2396 * or *@ was used with a query-style lookup.
2398 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2400 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2401 long ago, but noone interested enough thought of it. Let's just say that
2402 the "1.1" means that there are some new functions that weren't there at
2403 some point in the past.
2405 PH/51 Error processing for expansion failure of helo_data from an smtp
2406 transport during callout processing was broken.
2408 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2409 tested/used via the -bh/-bhc/-bs options.
2411 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2412 bug, fixed in subsequent PCRE releases).
2414 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2415 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2417 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2423 SC/01 Use a glob alias rather than an array ref in eximstats generated
2424 parser. This improves both readability and performance.
2426 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2427 Don't display local sender or destination tables in eximstats unless
2428 there is data to show.
2429 Added average volumes into the eximstats top table text output.
2431 SC/03 Collect data on the number of addresses (recipients) as well
2432 as the number of messages in eximstats.
2434 TF/01 Correct an error in the documentation for the redirect router. Exim
2435 does (usually) call initgroups() when daemonizing.
2437 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2438 with consistent privilege compared to when running as a daemon.
2440 TF/03 Note in the spec that $authenticated_id is not set for local
2441 submissions from trusted users.
2443 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2444 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2446 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2447 by adding some example configuration directives to the default
2448 configuration file. A little bit of work is required to uncomment the
2449 directives and define how usernames and passwords are checked, but
2450 there is now a framework to start from.
2452 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2453 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2454 without this. I don't know how relevant this is to other LDAP libraries.
2456 PH/02 Add the verb name to the "unknown ACL verb" error.
2458 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2460 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2462 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2463 directory not expanded when it should be if an expanded home directory
2464 was set for the address (which is overridden by the transport).
2466 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2469 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2470 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2471 because it is too late at that time, and has no effect.
2473 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2474 security issue with \' (bugzilla #107). I could not use the
2475 PQescapeStringConn() function, because it needs a PGconn value as one of
2478 PH/08 When testing addresses using -bt, indicate those final addresses that
2479 are duplicates that would not cause an additional delivery. At least one
2480 person was confused, thinking that -bt output corresponded to deliveries.
2481 (Suppressing duplicates isn't a good idea as you lose the information
2482 about possibly different redirections that led to the duplicates.)
2484 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2485 systems where poll() doesn't work, in particular OS X.
2487 PH/10 Added more information to debugging output for retry time not reached.
2489 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2490 operations in malware.c.
2492 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2495 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2496 syslog_duplication set false, log lines that would normally be written
2497 both the the main log and to the reject log were not written to syslog at
2500 PH/14 In the default configuration, change the use of "message" in ACL warn
2501 statements to "add_header".
2503 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2504 not followed by a command (e.g. "seen endif").
2506 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2507 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2510 PH/17 Added extra conditions to the default value of delay_warning_condition
2514 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2515 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2516 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2519 The Auto-Submitted: and various List- headers are standardised, whereas I
2520 don't think Precedence: ever was.
2522 PH/18 Refactored debugging code in route_finduser() to show more information,
2523 in particular, the error code if getpwnam() issues one.
2525 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2526 This is apparently needed in addition to the PH/07 change above to avoid
2527 any possible encoding problems.
2529 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2530 but not after initializing Perl.
2532 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2533 output them only if debugging. By default they are written stderr,
2534 apparently, which is not desirable.
2536 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2539 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2542 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2544 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2545 authenticated or an ident call has been made. Suppress the default
2546 values for $authenticated_id and $authenticated_sender (but permit -oMai
2547 and -oMas) when testing with -bh.
2549 PH/24 Re-jigged the order of the tests in the default configuration so that the
2550 tests for valid domains and recipients precede the DNS black list and CSA
2551 tests, on the grounds that those ones are more expensive.
2553 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2554 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2555 This bug exists in every version of Exim that I still have, right back to
2558 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2559 However, an attempt to turn on case-sensitivity in a regex key by
2560 including (?-i) didn't work because the subject string was already
2561 lowercased, and the effects were non-intuitive. It turns out that a
2562 one-line patch can be used to allow (?-i) to work as expected.
2568 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2569 other effects) broke the use of negated acl sub-conditions.
2571 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2574 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2575 "Deny" causes Exim to reject the incoming connection with a 554 error.
2576 Unfortunately, if there is a major crisis, such as a disk failure,
2577 tcp-wrappers gives "deny", whereas what one would like would be some
2578 kind of temporary error. A kludge has been added to help with this.
2579 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2580 554 error is used if errno is still zero or contains ENOENT (which occurs
2581 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2584 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2586 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2587 errors. Otherwise a message that provokes a temporary error (when other
2588 messages do not) can cause a whole host to time out.
2590 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2591 addresses were routed directly to files or pipes from a redirect router.
2592 File deliveries just didn't batch; pipe deliveries might have suffered
2595 PH/06 A failure to get a lock for a hints database would erroneously always say
2596 "Failed to get write lock", even when it was really a read lock.
2598 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2599 of 0600. This has been changed to use the value of the lockfile_mode
2600 option (which defaults to 0600).
2602 PH/08 Applied small patch from the Sieve maintainer.
2604 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2605 folder from quota calculations, a direct delivery into this folder messed
2606 up the contents of the maildirsize file. This was because the regex was
2607 used only to exclude .Trash (or whatever) when the size of the mailbox
2608 was calculated. There was no check that a delivery was happening into an
2609 excluded directory. This bug has been fixed by ignoring all quota
2610 processing for deliveries into excluded directories.
2612 PH/10 Added the maildirfolder_create_regex option to appendfile.
2618 PH/01 The code for finding all the local interface addresses on a FreeBSD
2619 system running IPv6 was broken. This may well have applied to all BSD
2620 systems, as well as to others that have similar system calls. The broken
2621 code found IPv4 interfaces correctly, but gave incorrect values for the
2622 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2623 that it would not match correctly against @[] and not recognize the IPv6
2626 PH/02 The ipliteral router was not recognizing addresses of the form user@
2627 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2629 PH/03 Added disable_ipv6.
2631 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2632 decoded form, because it is most often used to construct To: headers
2633 lines in autoreplies, and the decoded form may well be syntactically
2634 invalid. However, $reply_address has leading white space removed, and all
2635 newlines turned into spaces so that the autoreply transport does not
2638 PH/05 If group was specified without a user on a router, and no group or user
2639 was specified on a transport, the group from the router was ignored.
2641 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2642 for visible compile-time settings that can be used to change these
2643 numbers, for those that want even more. Backwards compatibility with old
2644 spool files has been maintained. However, going back to a previous Exim
2645 release will lost any variables that are in spool files.
2647 PH/07 Two small changes when running in the test harness: increase delay when
2648 passing a TCP/IP connection to a new process, in case the original
2649 process has to generate a bounce, and remove special handling of
2650 127.0.0.2 (sic), which is no longer necessary.
2652 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2653 be the same on different OS.
2655 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2658 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2659 whether --show-vars was specified or not
2661 JJ/02 exipick: Added support for new ACL variable spool format introduced
2664 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2665 syntactically invalid From: or Reply-to: line, and a filter used this to
2666 generate an autoreply, and therefore failed to obtain an address for the
2667 autoreply, Exim could try to deliver to a non-existent relative file
2668 name, causing unrelated and misleading errors. What now happens is that
2669 it logs this as a hard delivery error, but does not attempt to create a
2672 PH/11 The exinext utility has a -C option for testing purposes, but although
2673 the given file was scanned by exinext itself; it wasn't being passed on
2674 when Exim was called.
2676 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2677 an end-of-file indication when reading a command response.
2679 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2680 compiled. In many other places in Exim, IPv6 addresses are always
2681 recognized, so I have changed this. It also means that IPv4 domain
2682 literals of the form [IPV4:n.n.n.n] are now always recognized.
2684 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2685 used if the router is not running as root, for example, when verifying at
2686 ACL time, or when using -bh. The debugging output from this situation was
2687 non-existent - all you got was a failure to exec. I have made two
2690 (a) Failures to set uid/gid, the current directory, or a process leader
2691 in a subprocess such as that created by queryprogram now generate
2692 suitable debugging ouput when -d is set.
2694 (b) The queryprogram router detects when it is not running as root,
2695 outputs suitable debugging information if -d is set, and then runs
2696 the subprocess without attempting to change uid/gid.
2698 PH/15 Minor change to Makefile for building test_host (undocumented testing
2701 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2702 additional section of a DNS packet that returns MX or SRV records.
2703 Instead, it always explicitly searches for A/AAAA records. This avoids
2704 major problems that occur when a DNS server includes only records of one
2705 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2706 fixed another bug: if SRV records were looked up and the corresponding
2707 address records were *not* found in the additional section, the port
2708 values from the SRV records were lost.
2710 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2711 using the correct key (the original address) when searching the retry
2712 rules in order to find which one to use for generating the retry hint.
2714 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2715 adding the default one. Similarly, if it contains a Reply-To: header, the
2716 errors_reply_to option, if set, is not used.
2718 PH/19 When calculating a retry time, Exim used to measure the "time since
2719 failure" by looking at the "first failed" field in the retry record. Now
2720 it does not use this if it is later than than the arrival time of the
2721 message. Instead it uses the arrival time. This makes for better
2722 behaviour in cases where some deliveries succeed, thus re-setting the
2723 "first failed" field. An example is a quota failure for a huge message
2724 when small messages continue to be delivered. Without this change, the
2725 "time since failure" will always be short, possible causing more frequent
2726 delivery attempts for the huge message than are intended.
2727 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2729 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2730 $1, $2, $3) because the numerical variables can be reset during some
2731 expansion items (e.g. "match"), thereby losing the authentication data.
2733 PH/21 Make -bV show the size of off_t variables so that the test suite can
2734 decide whether to run tests for quotas > 2G.
2736 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2737 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2738 filecount value is greater than 2G or if a quota value is greater than 2G
2739 on a system where the size of off_t is not greater than 4, a panic error
2742 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2743 never match. The debug and -bh output now contains an explicit error
2744 message indicating a malformed IPv4 address or mask.
2746 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2747 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2748 PH/23 above applies.
2750 PH/25 Do not write to syslog when running in the test harness. The only
2751 occasion when this arises is a failure to open the main or panic logs
2752 (for which there is an explicit test).
2754 PH/26 Added the /no_tell option to "control=freeze".
2756 PH/27 If a host name lookup failed very early in a connection, for example, if
2757 the IP address matched host_lookup and the reverse lookup yielded a name
2758 that did not have a forward lookup, an error message of the form "no IP
2759 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2760 could be logged. Now it outputs the IP address instead of "NULL".
2762 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2763 allows the sender and the authenticated sender to be set when
2764 submitting a message from within Exim. Since child_open_exim() is
2765 documented for local_scan(), the new function should be too.
2767 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2768 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2769 results in an empty string is now treated as unset.
2771 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2773 PH/31 Added sender_verify_fail logging option.
2775 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2776 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2777 tidied the source and removed it altogether.
2779 PH/33 When a queue run was abandoned because the load average was too high, a
2780 log line was always written; now it is written only if the queue_run log
2781 selector is set. In addition, the log line for abandonment now contains
2782 information about the queue run such as the pid. This is always present
2783 in "start" and "stop" lines but was omitted from the "abandon" line.
2785 PH/34 Omit spaces between a header name and the colon in the error message that
2786 is given when verify = headers_syntax fails (if there are lots of them,
2787 the message gets confusing).
2789 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2790 names, as there are now some PTR records that contain slashes. This check
2791 is only to protect against broken name servers that fall over on strange
2792 characters, so the fact that it applies to all lookups doesn't matter.
2794 PH/36 Now that the new test suite is complete, we can remove some of the
2795 special code in Exim that was needed for the old test suite. For example,
2796 sorting DNS records because real resolvers return them in an arbitrary
2797 order. The new test suite's fake resolver always returns records in the
2800 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2801 bounces) except when queue_only is set, to avoid logging races between
2802 the different processes.
2804 PH/38 Panic-die if .include specifies a non-absolute path.
2806 PH/39 A tweak to the "H" retry rule from its user.
2808 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2809 a label. They prevented compilation on older perls.
2811 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2812 a warning to be raised on newish perls.
2814 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2815 on queue. Changes to match documented behaviour of showing count of
2816 messages matching specified criteria.
2818 PH/40 Changed the default ident timeout from 30s to 5s.
2820 PH/41 Added support for the use of login_cap features, on those BSD systems
2821 that have them, for controlling the resources used by pipe deliveries.
2823 PH/42 The content-scanning code uses fopen() to create files in which to put
2824 message data. Previously it was not paying any attention to the mode of
2825 the files. Exim runs with umask(0) because the rest of the code creates
2826 files with open(), and sets the required mode explicitly. Thus, these
2827 files were ending up world-writeable. This was not a big issue, because,
2828 being within the spool directory, they were not world-accessible. I have
2829 created a function called modefopen, which takes an additional mode
2830 argument. It sets umask(777), creates the file, chmods it to the required
2831 mode, then resets the umask. All the relevant calls to fopen() in the
2832 content scanning code have been changed to use this function.
2834 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2835 to 24 hours. This avoids potential overflow problems when processing G
2836 and H retry rules. I suspect nobody ever tinkers with this value.
2838 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2840 PH/45 When the plaintext authenticator is running as a client, the server's
2841 challenges are checked to ensure they are valid base64 strings. By
2842 default, the authentication attempt is cancelled if an invalid string is
2843 received. Setting client_ignore_invalid_base64 true ignores these errors.
2844 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2845 they are received. Thus, the responses can be made to depend on the
2846 challenges. If an invalid string is ignored, an empty string is placed in
2849 PH/46 Messages that are created by the autoreply transport now contains a
2850 References: header, in accordance with RFCs 2822 and 3834.
2852 PH/47 Added authenticated_sender_force to the smtp transport.
2854 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2856 PH/49 Installed latest patch from the Sieve maintainer.
2858 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2859 was also set, the appendfile transport was unnecessarily scanning a
2860 directory of message files (e.g. for maildir delivery) to find the count
2861 of files (along with the size), even though it did not need this
2862 information. It now does the scan only if it needs to find either the
2863 size of the count of files.
2865 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2867 PH/52 Two bugs concerned with error handling when the smtp transport is
2870 (i) Exim was not creating retry information for temporary errors given
2871 for individual recipients after the DATA command when the smtp transport
2872 was used in LMTP mode. This meant that they could be retried too
2873 frequently, and not timed out correctly.
2875 (ii) Exim was setting the flag that allows error details to be returned
2876 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2877 recipients that were returned after the DATA command.
2879 PH/53 This is related to PH/52, but is more general: for any failing address,
2880 when detailed error information was permitted to be returned to the
2881 sender, but the error was temporary, then after the final timeout, only
2882 "retry timeout exceeded" was returned. Now it returns the full error as
2883 well as "retry timeout exceeded".
2885 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2886 do this, and (what is worse) MTAs that accept it.
2888 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2889 will now be deprecated.
2891 PH/56 New os.c-cygwin from the Cygwin maintainer.
2893 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2894 formats (previously only available in exim formats via -bpr, -bpru,
2895 and -bpra. Now also available in native and exiqgrep formats)
2897 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2898 with very large, slow to parse queues
2900 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2902 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2904 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2905 responses to authentication challenges, though it was showing the
2906 challenges; (ii) I've removed the CR characters from the debug output for
2909 PH/58 Allow for the insertion of a newline as well as a space when a string
2910 is turned into more than one encoded-word during RFC 2047 encoding. The
2911 Sieve code now uses this.
2913 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2914 data_4xx, lost_connection, tls_required.
2916 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2917 message was being sent as an SMTP response.
2919 PH/61 Add -l and -k options to exicyclog.
2921 PH/62 When verifying, if an address was redirected to one new address, so that
2922 verification continued, and the new address failed or deferred after
2923 having set something in $address_data, the value of $address_data was not
2924 passed back to the ACL. This was different to the case when no
2925 redirection occurred. The value is now passed back in both cases.
2927 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2928 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2929 use login_cap.h, so on its own it isn't the distinguishing feature. The
2930 new name refers directly to the setclassresources() function.
2932 PH/65 Added configuration files for NetBSD3.
2934 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2936 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2937 is preferred over IPv4.
2939 PH/68 The bounce_return_message and bounce_return_body options were not being
2940 honoured for bounces generated during the reception of non-SMTP messages.
2941 In particular, this applied to messages rejected by the ACL. This bug has
2942 been fixed. However, if bounce_return_message is true and bounce_return_
2943 body is false, the headers that are returned for a non-SMTP message
2944 include only those that have been read before the error was detected.
2945 (In the case of an ACL rejection, they have all been read.)
2947 PH/69 The HTML version of the specification is now built in a directory called
2948 spec_html instead of spec.html, because the latter looks like a path with
2949 a MIME-type, and this confuses some software.
2951 PH/70 Catch two compiler warnings in sieve.c.
2953 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2954 function verify_get_ident() calls ip_connect() to connect a socket, but
2955 if the "connect()" function timed out, ip_connect() used to close the
2956 socket. However, verify_get_ident() also closes the socket later, and in
2957 between Exim writes to the log, which may get opened at this point. When
2958 the socket was closed in ip_connect(), the log could get the same file
2959 descriptor number as the socket. This naturally causes chaos. The fix is
2960 not to close the socket in ip_connect(); the socket should be closed by
2961 the function that creates it. There was only one place in the code where
2962 this was missing, in the iplookup router, which I don't think anybody now
2963 uses, but I've fixed it anyway.
2965 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2966 well as to direct DNS lookups. Otherwise the handling of names in host
2967 lists is inconsistent and therefore confusing.
2973 PH/01 Two changes to the default runtime configuration:
2975 (1) Move the checks for relay_from_hosts and authenticated clients from
2976 after to before the (commented out) DNS black list checks.
2978 (2) Add control=submission to the relay_from_hosts and authenticated
2979 clients checks, on the grounds that messages accepted by these
2980 statements are most likely to be submissions.
2982 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2984 (1) Generate an error if the third argument for the ${prvs expansion is
2987 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2990 (3) Reset the variables that are obtained from the first argument of
2991 ${prvscheck and used in the second argument before leaving the code,
2992 because their memory is reclaimed, so using them afterwards may do
2995 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2996 one (it's much easier than Tom thought :-).
2998 (5) Because of (4), we can now allow for the use of $prvscheck_result
2999 inside the third argument.
3001 PH/03 For some reason, the default setting of PATH when running a command from
3002 a pipe transport was just "/usr/bin". I have changed it to
3005 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3006 anything to be listed in the output from -bV.
3008 PH/05 When a filter generated an autoreply, the entire To: header line was
3009 quoted in the delivery log line, like this:
3011 => >A.N.Other <ano@some.domain> <original@ddress> ...
3013 This has been changed so that it extracts the operative address. There
3014 may be more than one such address. If so, they are comma-separated, like
3017 => >ano@some.domain,ona@other.domain <original@ddress> ...
3019 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3020 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3021 not being looked up in the rDNS to get a host name, Exim was showing the
3022 IP address twice in Received: lines, even though the IP addresses were
3023 identical. For example:
3025 Received: from [1.2.3.4] (helo=[1.2.3.4])
3027 However, if the real host name was known, it was omitting the HELO data
3028 if it matched the actual IP address. This has been tidied up so that it
3029 doesn't show the same IP address twice.
3031 PH/07 When both +timestamp and +memory debugging was on, the value given by
3032 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3033 called by the debug printing, thereby overwriting the timestamp buffer.
3034 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3036 PH/08 When the original message was included in an autoreply transport, it
3037 always said "this is a copy of the message, including all the headers",
3038 even if body_only or headers_only was set. It now gives an appropriate
3041 PH/09 Applied a patch from the Sieve maintainer which:
3043 o fixes some comments
3044 o adds the (disabled) notify extension core
3045 o adds some debug output for the result of if/elsif tests
3046 o points to the current vacation draft in the documentation
3047 and documents the missing references header update
3051 o fixes a bug in processing the envelope test (when testing
3052 multiple envelope elements, the last element determinted the
3055 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3056 Electronic Mail") by including:
3058 Auto-submitted: auto-generated
3060 in the messages that it generates (bounce messages and others, such as
3061 warnings). In the case of bounce messages for non-SMTP mesages, there was
3062 also a typo: it was using "Auto_submitted" (underscore instead of
3063 hyphen). Since every message generated by Exim is necessarily in response
3064 to another message, thes have all been changed to:
3066 Auto-Submitted: auto-replied
3068 in accordance with these statements in the RFC:
3070 The auto-replied keyword:
3072 - SHOULD be used on messages sent in direct response to another
3073 message by an automatic process,
3075 - MUST NOT be used on manually-generated messages,
3077 - MAY be used on Delivery Status Notifications (DSNs) and Message
3078 Disposition Notifications (MDNs),
3080 - MUST NOT be used on messages generated by automatic or periodic
3081 processes, except for messages which are automatic responses to
3084 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3085 to the default Received: header definition.
3087 PH/12 Added log selector acl_warn_skipped (default on).
3089 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3090 variables because (a) they are in the wrong storage pool and (b) even if
3091 they were copied, it wouldn't work properly because of the caching.
3093 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3094 checking when decoding. Apparently there are clients that generate
3095 overlong encoded strings. Why am I not surprised?
3097 PH/15 If the first argument of "${if match_address" was not empty, but did not
3098 contain an "@" character, Exim crashed. Now it writes a panic log message
3099 and treats the condition as false.
3101 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3103 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3104 extension "envelope-auth". The code is finished and in agreement with
3105 other implementations, but there is no documentation so far and in fact,
3106 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3107 not changing the active code.
3109 Print executed "if" and "elsif" statements when debugging is used. This
3110 helps a great deal to understand what a filter does.
3112 Document more things not specified clearly in RFC3028. I had all this
3113 sorted out, when out of a sudden new issues came to my mind. Oops."
3115 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3118 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3119 canonical form (as documented). However, after a host name lookup from
3120 the IP address, check_host() was doing a simple string comparison with
3121 addresses acquired from the DNS when checking that the found name did
3122 have the original IP as one of its addresses. Since any found IPv6
3123 addresses are likely to be in abbreviated form, the comparison could
3124 fail. Luckily, there already exists a function for doing the comparison
3125 by converting both addresses to binary, so now that is used instead of
3126 the text comparison.
3128 PH/20 There was another similar case to PH/19, when a complete host name was
3129 given in a host list; looking up its IP address could give an abbreviated
3130 form, whereas the current host's name might or might not be abbreviated.
3131 The same fix has been applied.
3137 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3138 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3141 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3142 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3144 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3146 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3147 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3148 Also, for "Auto-Submitted", treat anything other than "no" as
3149 non-personal, in accordance with RFC 3834. (Previously it treated
3150 anything starting "auto-" as non-personal.)
3152 TF/01 The control=submission/name=... option had a problem with syntax
3153 errors if the name included a slash character. The /name= option
3154 now slurps the rest of the string, so it can include any characters
3155 but it must come last in the list of options (after /sender_retain
3158 PH/05 Some modifications to the interface to the fake nameserver for the new
3166 TK/01 Added the "success_on_redirect" address verification option. See
3167 NewStuff for rationale and an example.
3169 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3171 PH/02 Patch to exigrep to allow it to work on syslog lines.
3173 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3174 fread() to skip over the body file's header line, because in Cygwin the
3175 header line is locked and is inaccessible.
3177 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3178 co-exist for some time) to make it clear that it is the Exim ID that is
3179 referenced, not the Message-ID: header line.
3181 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3182 string_format() function, because snprintf() does not exist on all
3185 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3186 ${stat: expansion item.
3188 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3189 protocol synchronization error", to keep the pedants happy.
3191 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3192 well as for IRIX systems, when gcc is being used. See the host.c source
3195 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3197 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3200 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3201 transport and to the smtp transport in LMTP mode.
3203 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3205 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3206 run a filter in a subprocess. This could lead to confusion in subsequent
3207 lookups in the parent process. There should also be a search_tidyup() at
3208 the end of the subprocess.
3210 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3211 only if the host matched helo_try_verify_hosts, which caused the
3212 verification to occur when the EHLO/HELO command was issued. The ACL just
3213 tested the remembered result. Now, if a previous verification attempt has
3214 not happened, "verify = helo" does it there and then.
3216 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3218 TK/03 Fix log output including CR from clamd.
3220 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3221 exist provoked a memory error which could cause a segfault.
3223 PH/15 Installed PCRE 6.2
3225 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3227 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3228 of the problem. Specifically, suggested +O2 rather than +O1 for the
3231 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3233 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3234 record, the port was not correctly logged when the outgoing_port log
3235 selector was set (it logged the transort's default port).
3237 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3238 fallback_hosts, and "hosts" in the smtp transport.
3240 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3241 host errors such as "Connection refused".
3243 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3244 authentication with radiusclient 0.4.9:
3246 - Error returned from rc_read_config was caught wrongly
3247 - Username/password not passed on to radius server due to wrong length.
3249 The presumption is that some radiusclient API changes for 4.51/PH/17
3250 were not taken care of correctly. The code is still untested by me (my
3251 Linux distribution still has 0.3.2 of radiusclient), but it was
3252 contributed by a Radius user.
3254 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3255 expanding the "port" option of the smtp transport.
3257 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3258 while reading a MIME header. Thanks to Tom Hughes for a patch.
3260 PH/24 Include config.h inside local_scan.h so that configuration settings are
3263 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3264 that in an ACL for RCPT (for example), you can examine exactly what was
3267 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3268 commands, but it was not correctly comparing the address with the actual
3269 client host address. Thus, it would show the EHLO address in Received:
3270 header lines when this was not necessary.
3272 PH/27 Added the % operator to ${eval:}.
3274 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3275 it should be ignoring failures (because with -C, for example, it has lost
3276 privilege). It wasn't ignoring creation failures other than "already
3279 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3282 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3283 because an input line was too long, either on its own, or by virtue of
3284 too many continuations, the temporary file was not being removed, and the
3285 return code was incorrect.
3287 PH/31 Missing "BOOL" in function definition in filtertest.c.
3289 PH/32 Applied Sieve patches from the maintainer.
3291 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3293 PH/33 Added "verify = not_blind".
3295 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3296 Local/Makefile (with some defaults set). These are used in built scripts
3297 such as exicyclog, but they have never been used in the exim_install
3298 script (though there are many overriding facilities there). I have
3299 arranged that the exim_install script now takes note of these two
3302 PH/35 Installed configuration files for Dragonfly.
3304 PH/36 When a locally submitted message by a trusted user did not contain a
3305 From: header, and the sender address was obtained from -f or from an SMTP
3306 MAIL command, and the trusted user did not use -F to supply a sender
3307 name, $originator_name was incorrectly used when constructing a From:
3308 header. Furthermore, $originator_name was used for submission mode
3309 messages from external hosts without From: headers in a similar way,
3310 which is clearly wrong.
3312 PH/37 Added control=suppress_local_fixups.
3314 PH/38 When log_selector = +received_sender was set, and the addition of the
3315 sender made the log line's construction buffer exactly full, or one byte
3316 less than full, an overflow happened when the terminating "\n" was
3319 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3320 when the result of a list match is failure because a DNS lookup failed.
3322 PH/40 RM_COMMAND is now used in the building process.
3324 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3325 the "build-* directories that it finds.
3327 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3328 address in a domain literal was a prefix of an interface address.
3330 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3331 when verifying a sender address, unless rewrite_headers is false.
3333 PH/44 Wrote a long comment about why errors_to addresses are verified as
3334 recipients, not senders.
3336 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3337 the ratelimit ACL was added.
3339 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3341 PH/46 Added extra information about PostgreSQL errors to the error string.
3343 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3344 avoiding the need to install special zones in a real server. This is
3345 backwards compatible; if it can't find the fake resolver, it drops back.
3346 Thus, both old and new test suites can be run.
3348 TF/02 Added util/ratelimit.pl
3350 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3351 clock is set back in time.
3353 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3354 Candler <B.Candler@pobox.com>.
3356 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3357 OK for addresses that are the result of redirections.
3359 PH/48 A number of further additions for the benefit of the new test suite,
3360 including a fake gethostbyname() that interfaces to the fake DNS resolver
3363 TF/06 The fix for widen_domains has also been applied to qualify_single and
3364 search_parents which are the other dnslookup options that can cause
3367 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3370 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3372 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3373 service" if it encountered an MX record with an empty target hostname.
3374 The message is now "an MX or SRV record indicated no SMTP service".
3376 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3377 if the DNS of the sending site is misconfigured. This is quite a
3378 common situation. This change restores the behaviour of treating a
3379 helo verification defer as a failure.
3381 PH/51 If self=fail was set on a router, the bounce message did not include the
3382 actual error message.
3388 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3390 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3391 command itself ended in error, the underlying message about the transport
3392 filter timeout was being overwritten with the pipe command error. Now the
3393 underlying error message should be appended to the second error message.
3395 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3397 PH/02 Reduce the amount of output that "make" produces by default. Full output
3398 can still be requested.
3400 PH/03 The warning log line about a condition test deferring for a "warn" verb
3401 was being output only once per connection, rather than after each
3402 occurrence (because it was using the same function as for successful
3403 "warn" verbs). This seems wrong, so I have changed it.
3405 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3406 it should not have, which might have caused a crash in the right
3407 circumstances, but probably never did.
3409 PH/04 Installed a modified version of Tony Finch's patch to make submission
3410 mode fix the return path as well as the Sender: header line, and to
3411 add a /name= option so that you can make the user's friendly name appear
3414 TF/03 Added the control = fakedefer ACL modifier.
3416 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3417 Mark Lowes for thorough testing.
3419 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3421 TK/03 Merged latest SRS patch from Miles Wilton.
3423 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3424 with the definition in sysexits.h (which is #included earlier).
3425 Fortunately, Exim does not actually use EX_OK. The code used to try to
3426 preserve the sysexits.h value, by assumimg that macro definitions were
3427 scanned for macro replacements. I have been disabused of this notion,
3428 so now the code just undefines EX_OK before #including unistd.h.
3430 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3431 in the smtp transport. When a block could not be written in a single
3432 write() function, the timeout was being re-applied to each part-write.
3433 This seems wrong - if the receiver was accepting one byte at a time it
3434 would take for ever. The timeout is now adjusted when this happens. It
3435 doesn't have to be particularly precise.
3437 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3438 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3440 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3441 without a domain if the check to <postmaster@domain> fails.
3443 SC/01 Eximstats: added -xls and the ability to specify output files
3444 (patch written by Frank Heydlauf).
3446 SC/02 Eximstats: use FileHandles for outputing results.
3448 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3450 SC/04 Eximstats: fixed display of large numbers with -nvr option
3452 SC/05 Eximstats: fixed merging of reports with empty tables.
3454 SC/06 Eximstats: added the -include_original_destination flag
3456 SC/07 Eximstats: removed tabs and trailing whitespace.
3458 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3460 TK/06 MBOX spool code: Add real "From " MBOX separator line
3461 so the .eml file is really in mbox format (even though
3462 most programs do not really care). Patch from Alex Miller.
3464 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3465 The latter is generated from $received_to and is only set if the
3466 message has one envelope recipient. SA can use these headers,
3467 obviously out-of-the-box. Patch from Alex Miller.
3469 PH/08 The ${def test on a variable was returning false if the variable's
3470 value was "0", contrary to what the specification has always said!
3471 The result should be true unless the variable is empty.
3473 PH/09 The syntax error of a character other than { following "${if
3474 def:variable_name" (after optional whitespace) was not being diagnosed.
3475 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3476 accidental colon was present, for example, could give incorrect results.
3478 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3479 result is used (not including appendfile, where other changes are about
3482 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3483 This involved changing a lot of size variables from int to off_t. It
3484 should work with maildirs and everything.
3486 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3487 spamd dying while we are connected to it.
3489 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3492 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3493 function for BDB 4.3.
3495 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3497 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3498 the sources for PCRE have been re-arranged and more files are now
3501 PH/15 The code I had for printing potentially long long variables in PH/11
3502 above was not the best (it lost precision). The length of off_t variables
3503 is now inspected at build time, and an appropriate printing format (%ld
3504 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3505 to be "long long int" or "long int". This is needed for the internal
3506 formatting function string_vformat().
3508 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3509 the configuration file to be ":syslog", then the script "guesses" where
3510 the logs files are, rather than using the compiled in default. In our
3511 case the guess is not the same as the compiled default, so the script
3512 suddenly stopped working when I started to use syslog. The patch checks
3513 to see if log_file_path is "". If so, it attempts to read it from exim
3514 with no configuration file to get the compiled in version, before it
3515 falls back to the previous guessing code."
3517 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3518 implementing BATV in an Exim configuration. See NewStuff for the gory
3521 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3522 Makefile that are specific to HP-UX.
3524 PH/18 If the "use_postmaster" option was set for a recipient callout together
3525 with the "random" option, the postmaster address was used as the MAIL
3526 FROM address for the random test, but not for the subsequent recipient
3527 test. It is now used for both.
3529 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3530 patch removes a few documentation additions to RFC 3028, because the
3531 latest draft now contains them. It adds the new en;ascii-case comparator
3532 and a new error check for 8bit text in MIME parts. Comparator and
3533 require names are now matched exactly. I enabled the subaddress
3534 extension, but it is not well tested yet (read: it works for me)."
3536 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3537 rework some of the code of TK/09 above to avoid the hardwired use of
3538 "%lld" and "long long". Replaced the call to snprintf() with a call to
3541 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3542 records point to non-existent hosts", "retry timeout exceeded", and
3543 "retry time not reached for any host after a long failure period".
3545 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3546 experimental DomainKeys support:
3548 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3549 (2) On an error such as an illegally used "control", the wrong name for
3550 the control was given.
3552 These problems did NOT occur unless DomainKeys support was compiled.
3554 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3556 PH/24 Added ${if match_ip condition.
3558 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3559 databases so that it will be absolutely obvious if a crash occurs in the
3560 DB library. This is a regular occurrence (often caused by mis-matched
3563 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3564 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3565 picked up on a user's system that detects such things. There doesn't seem
3566 to be a gcc warning option for this - only an attribute that has to be
3567 put on the function's prototype. It seems that in Fedora Core 4 they have
3568 set this on a number of new functions. No doubt there will be more in due
3571 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3572 specify a transport. However, if an address that was verified by such a
3573 router was the subject of a callout, Exim crashed because it tried to
3574 read the rcpt_include_affixes from the non-existent transport. Now it
3575 just assumes that the setting of that option is false. This bug was
3576 introduced by 4.51/PH/31.
3578 PH/28 Changed -d+all to exclude +memory, because that information is very
3579 rarely of interest, but it makes the output a lot bigger. People tend to
3580 do -d+all out of habit.
3582 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3583 code in os-type was giving problems when libc.so lives in lib64, like on
3586 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3587 aren't the modern standard, and it seems that some systems' include files
3588 don't always have them. Exim was already checking for some of the newer
3589 ones like T_AAAA, and defining it itself. I've added checks for all the
3590 record types that Exim uses.
3592 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3593 not automatically generating a new one, as it is supposed to. This
3594 prevented TLS from working. If the file did exist, but contained invalid
3595 data, a new version was generated, as expected. It was only the case of a
3596 non-existent file that was broken.
3598 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3599 with a change in libdomainkeys > 0.64.
3601 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3602 from DNS. If the selector record carries the flag, it now has
3603 precedence over the domain-wide flag.
3605 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3607 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3608 the use of a port name that isn't defined in /etc/services) occurred, the
3609 message was deferred as in a normal delivery, and thus remained on the
3610 spool, instead of being failed because of the mua_wrapper setting. This
3611 is now fixed, and I tidied up some of the mua_wrapper messages at the
3614 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3615 the messages in a hash of arrays rather than using individual hashes.
3616 This is a bit cleaner and results in dramatic memory savings, albeit
3617 at a slight CPU cost.
3619 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3620 as requested by Marc Sherman.
3622 SC/10 Eximstats: added histograms for user specified patterns as requested
3625 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3627 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3628 fopen() in the content-scanning modules that did not already have it.
3634 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3635 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3637 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3639 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3641 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3642 to test Sieve filters that use "vacation".
3644 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3645 that changes the way the GnuTLS parameters are stored in the cache file.
3646 The new format can be generated externally. For backward compatibility,
3647 if the data in the cache doesn't make sense, Exim assumes it has read an
3648 old-format file, and it generates new data and writes a new file. This
3649 means that you can't go back to an older release without removing the
3652 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3653 work if there are any delivery delays because "one_time" forces the
3654 parent to be marked "delivered", so its unseen clone is never tried
3655 again. For this reason, Exim now forbids the simultaneous setting of
3658 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3659 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3660 entry if you want to know the details. The fix, however, neglected to
3661 consider the case when local delivery batching is involved. The test for
3662 "previously delivered" was not happening when checking to see if an
3663 address could be batched with a previous (undelivered) one; under
3664 certain circumstances this could lead to multiple deliveries to the same
3667 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3668 in its include files, and this causes problems building Exim.
3670 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3671 header_syntax) but Exim was just ignoring anything given after a slash.
3672 In particular, this caused confusion with an attempt to use "verify =
3673 reverse_host_lookup/defer_ok". An error is now given when options are
3674 supplied for verify items that do not have them. (Maybe reverse_host_
3675 lookup should have a defer_ok option, but that's a different point.)
3677 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3678 defined by RFC 821) to 2048, because there were problems with some AUTH
3679 commands, and RFC 1869 says the size should be increased for extended
3680 SMTP commands that take arguments.
3682 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3685 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3686 "unknown" error; now it says that the functionality isn't in the binary.
3688 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3689 an address' error message when a string expansion fails (syntax or
3690 whatever). Otherwise the password may appear in the log. Following change
3691 PH/42 below, there is no longer a chance of it appearing in a bounce
3694 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3696 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3697 of its addresses was ever tried. (Bugzilla bug #2.)
3699 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3700 the result incorrectly in the debug output. (It correctly added a newline
3701 to what was transported.)
3703 TF/01 Added $received_time.
3705 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3706 commented out examples of how to interface to a virus scanner and to
3707 SpamAssassin. Also added commented examples of av_scanner and
3708 spamd_address settings.
3710 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3711 and controls are allowed in which ACLs. There were a couple of minor
3712 errors. Some of the entries in the conditions table (which is a table of
3713 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3714 negation of where the condition IS allowed.
3716 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3718 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3719 header file does not have a version number, so I've had to invent a new
3720 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3721 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3722 radiusclient), but it was contributed by a Radius user.
3724 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3725 files or directories, for OpenSSL.
3727 PH/19 When an Exim process that is running as root has to create an Exim log
3728 file, it does so in a subprocess that runs as exim:exim so as to get the
3729 ownership right at creation (otherwise, other Exim processes might see
3730 the file with the wrong ownership). There was no test for failure of this
3731 fork() call, which would lead to the process getting stuck as it waited
3732 for a non-existent subprocess. Forks do occasionally fail when resources
3733 run out. I reviewed all the other calls to fork(); they all seem to check
3736 PH/20 When checking for unexpected SMTP input at connect time (before writing
3737 the banner), Exim was not dealing correctly with a non-positive return
3738 from the read() function. If the client had disconnected by this time,
3739 the result was a log entry for a synchronization error with an empty
3740 string after "input=" when read() returned zero. If read() returned -1
3741 (an event I could not check), uninitialized data bytes were printed.
3742 There were reports of junk text (parts of files, etc) appearing after
3745 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3747 PH/22 Added support for macro redefinition, and (re)definition in between
3748 driver and ACL definitions.
3750 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3751 forgetting to use the resulting value; it was using the unexpanded value.
3753 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3754 hadn't been configured. The fix is from Juergen Kreileder, who
3755 understands it better than I do:
3757 "Here's what I see happening with three configured cyrus_sasl
3758 authenticators configured (plain, login, cram-md5):
3760 On startup auth_cyrus_sasl_init() gets called for each of these.
3761 This means three calls to sasl_listmech() without a specified mech_list.
3762 => SASL tests which mechs of all available mechs actually work
3763 => three warnings about OTP not working
3764 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3766 With the patch, sasl_listmech() also gets called three times. But now
3767 SASL's mech_list option is set to the server_mech specified in the the
3768 authenticator. Or in other words, the answer from sasl_listmech()
3769 gets limited to just the mech you're testing for (which is different
3771 => the return list contains just 'plain' or 'login', 'cram-md5' or
3772 nothing depending on the value of ob->server_mech.
3774 I've just tested the patch: Authentication still works fine,
3775 unavailable mechs specified in the exim configuration are still
3776 caught, and the auth.log warnings about OTP are gone."
3778 PH/25 When debugging is enabled, the contents of the command line are added
3779 to the debugging output, even when log_selector=+arguments is not
3782 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3783 answer is "GNU", and only if the return is "GNU/something" is the answer
3786 PH/27 $acl_verify_message is now set immediately after the failure of a
3787 verification in an ACL, and so is available in subsequent modifiers. In
3788 particular, the message can be preserved by coding like this:
3790 warn !verify = sender
3791 set acl_m0 = $acl_verify_message
3793 Previously, $acl_verify_message was set only while expanding "message"
3794 and "log_message" when a very denied access.
3796 PH/28 Modified OS/os.c-Linux with
3798 -#ifndef OS_LOAD_AVERAGE
3799 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3801 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3802 nomenclature these days.)
3804 PH/29 Installed patch from the Sieve maintainer that adds the options
3805 sieve_useraddress and sieve_subaddress to the redirect router.
3807 PH/30 In these circumstances:
3808 . Two addresses routed to the same list of hosts;
3809 . First host does not offer TLS;
3810 . First host accepts first address;
3811 . First host gives temporary error to second address;
3812 . Second host offers TLS and a TLS session is established;
3813 . Second host accepts second address.
3814 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3815 and peerdn, if requested) that were in fact used only for the second
3818 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3819 attention to any local part prefix or suffix that was matched by the
3820 router that accepted the address. It now behaves in the same way as it
3821 does for delivery: the affixes are removed from the local part unless
3822 rcpt_include_affixes is set on the transport.
3824 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3825 timeout during the DATA phase of an incoming message.
3827 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3828 applied a patch sanctioned by the Sieve maintainer.
3830 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3831 the uid or gid is negative. A case of a negative gid caused this to be
3832 noticed. The fix allows for either to be negative.
3834 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3835 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3838 PH/36 The change PH/12 above was broken. Fixed it.
3840 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3841 the grounds that routing the same address twice would always produce the
3842 same answer. This might have been true once, but it is certainly no
3843 longer true now. Routing a child address may depend on the previous
3844 routing that produced that child. Some complicated redirection strategies
3845 went wrong when messages had multiple recipients, and made Exim's
3846 behaviour dependent on the order in which the addresses were given.
3848 I have moved the duplicate checking until after the routing is complete.
3849 Exim scans the addresses that are assigned to local and remote
3850 transports, and removes any duplicates. This means that more work will be
3851 done, as duplicates will always all be routed, but duplicates are
3852 presumably rare, so I don't expect this is of any significance.
3854 For deliveries to pipes, files, and autoreplies, the duplicate checking
3855 still happens during the routing process, since they are not going to be
3858 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3859 It corrects a timeout issue with spamd. This is Ian's comment: "The
3860 background is that sometimes spamd either never reads data from a
3861 connection it has accepted, or it never writes response data. The exiscan
3862 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3863 blindly assumes that writes won't block so it may never time out."
3865 PH/39 Allow G after quota size as well as K and M.
3867 PH/40 The value set for $authenticated_id in an authenticator may not contain
3868 binary zeroes or newlines because the value is written to log lines and
3869 to spool files. There was no check on this. Now the value is run through
3870 the string_printing() function so that such characters are converted to
3871 printable escape sequences.
3873 PH/41 $message_linecount is a new variable that contains the total number of
3874 lines in the message. Compare $body_linecount, which is the count for the
3877 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3878 bounce and delay warning messages, except in certain special cases, which
3881 (a) An SMTP error message from a remote host;
3882 (b) A message specified in a :fail: redirection;
3883 (c) A message specified in a "fail" command in a system filter;
3884 (d) A message specified in a FAIL return from the queryprogram router;
3885 (e) A message specified by the cannot_route_message router option.
3887 In these cases only, Exim does include the error details in bounce and
3888 warning messages. There are also a few cases where bland messages such
3889 as "unrouteable address" or "local delivery error" are given.
3891 PH/43 $value is now also set for the "else" part of a ${run expansion.
3893 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3894 being worked on, but at least Exim now implements the latest version to
3897 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3898 process to complete was treated as a delivery failure, a timeout while
3899 writing the message to the pipe was logged, but erroneously treated as a
3900 successful delivery. Such timeouts include transport filter timeouts. For
3901 consistency with the overall process timeout, these timeouts are now
3902 treated as errors, giving rise to delivery failures by default. However,
3903 there is now a new Boolean option for the pipe transport called
3904 timeout_defer, which, if set TRUE, converts the failures into defers for
3905 both kinds of timeout. A transport filter timeout is now identified in
3908 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3909 systems where "make" and "gmake" are different, calling "gmake" at top
3910 level broke things. I've arranged for the value of $(MAKE) to be passed
3911 from the Makefile to this script so that it can call the same version of
3915 A note about Exim versions 4.44 and 4.50
3916 ----------------------------------------
3918 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3919 changes of various kinds. As a consequence, a big documentation update was
3920 needed. This delayed the release for rather longer than seemed good, especially
3921 in the light of a couple of (minor) security issues. Therefore, the changes
3922 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3923 release. So 4.44 and 4.50 are in effect two different branches that both start
3926 I have left the 4.50 change log unchanged; it contains all the changes since
3927 4.43. The change log for 4.44 is below; many of its items are identical to
3928 those for 4.50. This seems to be the most sensible way to preserve the
3929 historical information.
3935 1. Minor wording change to the doc/README.SIEVE file.
3937 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3938 computation of the current number of files was incorrect.
3940 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3941 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3944 4. Give more explanation in the error message when the command for a transport
3945 filter fails to execute.
3947 5. There are several places where Exim runs a non-Exim command in a
3948 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3949 was being done only for the command run by the queryprogram router. It is
3950 now done for all such subprocesses. The other cases are: ${run, transport
3951 filters, and the commands run by the lmtp and pipe transports.
3953 6. Added CONFIGURE_GROUP build-time option.
3955 7. Some older OS have a limit of 256 on the maximum number of file
3956 descriptors. Exim was using setrlimit() to set 1000 as a large value
3957 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3958 systems. I've change it so that if it can't get 1000, it tries for 256.
3960 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3961 was an oversight, and furthermore, ever since the addition of extra
3962 controls (e.g. 4.43/32), the checks on when to allow different forms of
3963 "control" were broken. There should now be diagnostics for all cases when a
3964 control that does not make sense is encountered.
3966 9. Added the /retain_sender option to "control=submission".
3968 10. $recipients is now available in the predata ACL (oversight).
3970 11. Tidy the search cache before the fork to do a delivery from a message
3971 received from the command line. Otherwise the child will trigger a lookup
3972 failure and thereby defer the delivery if it tries to use (for example) a
3973 cached ldap connection that the parent has called unbind on.
3975 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3976 of $address_data from the recipient verification was clobbered by the
3977 sender verification.
3979 13. The value of address_data from a sender verification is now available in
3980 $sender_address_data in subsequent conditions in the ACL statement.
3982 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3984 15. Added a new option "connect=<time>" to callout options, to set a different
3987 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3988 was its contents. (It was OK if the option was not defined at all.)
3990 17. A "Completed" log line is now written for messages that are removed from
3991 the spool by the -Mrm option.
3993 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3994 information about exactly what failed.
3996 19. Added -dd to debug only the daemon process.
3998 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3999 handles timeouts, both on the server side and network timeouts. Renamed the
4000 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4002 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4003 It is now set to "smtps".
4005 22. $host_address is now set to the target address during the checking of
4006 ignore_target_hosts.
4008 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4009 being passed; this would have caused $sender_host_name to have been used if
4010 matching the list had actually called for a host name (not very likely,
4011 since this list is usually IP addresses). A host name is now passed as
4014 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4015 code that specifies a non-restarting handler (typically sigaction() in
4016 modern systems) in an attempt to fix a rare and obscure crash bug.
4018 25. Narrowed the window for a race in the daemon that could cause it to ignore
4019 SIGCHLD signals. This is not a major problem, because they are used only to
4020 wake it up if nothing else does.
4022 26. A malformed maildirsize file could cause Exim to calculate negative values
4023 for the mailbox size or file count. Odd effects could occur as a result.
4024 The maildirsize information is now recalculated if the size or filecount
4027 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4028 support for a long time. Removed HAVE_SYS_VFS_H.
4030 28. Installed the latest version of exipick from John Jetmore.
4032 29. In an address list, if the pattern was not a regular expression, an empty
4033 subject address (from a bounce message) matched only if the pattern was an
4034 empty string. Non-empty patterns were not even tested. This was the wrong
4035 because it is perfectly reasonable to use an empty address as part of a
4036 database query. An empty address is now tested by patterns that are
4037 lookups. However, all the other forms of pattern expect the subject to
4038 contain a local part and a domain, and therefore, for them, an empty
4039 address still always fails if the pattern is not itself empty.
4041 30. Exim went into a mad DNS loop when attempting to do a callout where the
4042 host was specified on an smtp transport, and looking it up yielded more
4043 than one IP address.
4045 31. Re-factored the code for checking spool and log partition space into a
4046 function that finds that data and another that does the check. The former
4047 is then used to implement four new variables: $spool_space, $log_space,
4048 $spool_inodes, and $log_inodes.
4050 32. The RFC2047 encoding function was originally intended for short strings
4051 such as real names; it was not keeping to the 75-character limit for
4052 encoded words that the RFC imposes. It now respects the limit, and
4053 generates multiple encoded words if necessary. To be on the safe side, I
4054 have increased the buffer size for the ${rfc2047: expansion operator from
4057 33. It is now permitted to omit both strings after an "if" condition; if the
4058 condition is true, the result is "true". As before, when the second string
4059 is omitted, a false condition yields an empty string. This makes it less
4060 cumbersome to write custom ACL and router conditions.
4062 34. Failure to deliver a bounce message always caused it to be frozen, even if
4063 there was an errors_to setting on the router. The errors_to setting is now
4066 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4067 canonical form (fully expanded) before being placed in
4068 $sender_host_address.
4070 36. The table in the code that translates DNS record types into text (T_A to
4071 "A" for instance) was missing entries for NS and CNAME. It is just possible
4072 that this could have caused confusion if both these types were looked up
4073 for the same domain, because the text type is used as part of Exim's
4074 per-process caching. But the chance of anyone hitting this buglet seems
4077 37. The dnsdb lookup has been extended in a number of ways.
4079 (1) There is a new type, "zns", which walks up the domain tree until it
4080 finds some nameserver records. It should be used with care.
4082 (2) There is a new type, "mxh", which is like "mx" except that it returns
4083 just the host names, not the priorities.
4085 (3) It is now possible to give a list of domains (or IP addresses) to be
4086 looked up. The behaviour when one of the lookups defers can be
4087 controlled by a keyword.
4089 (4) It is now possible to specify the separator character for use when
4090 multiple records are returned.
4092 38. The dnslists ACL condition has been extended: it is now possible to supply
4093 a list of IP addresses and/or domains to be looked up in a particular DNS
4096 39. Added log_selector=+queue_time_overall.
4098 40. When running the queue in the test harness, wait just a tad after forking a
4099 delivery process, to get repeatability of debugging output.
4101 41. Include certificate and key file names in error message when GnuTLS fails
4102 to set them up, because the GnuTLS error message doesn't include the name
4103 of the failing file when there is a problem reading it.
4105 42. Allow both -bf and -bF in the same test run.
4107 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4109 44. The "Exiscan patch" is now merged into the mainline Exim source.
4111 45. Sometimes the final signoff response after QUIT could fail to get
4112 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4113 < 0 before doing a fflush(). This bug looks as though it goes back to the
4114 introduction of TLS in release 3.20, but "sometimes" must have been rare
4115 because the tests only now provoked it.
4117 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4118 (this can affect the format of dates).
4120 47. exim_tidydb, when checking for the continued existence of a message for
4121 which it has found a message-specific retry record, was not finding
4122 messages that were in split spool directories. Consequently, it was
4123 deleting retry records that should have stayed in existence.
4125 48. Steve fixed some bugs in eximstats.
4127 49. The SPA authentication driver was not abandoning authentication and moving
4128 on to the next authenticator when an expansion was forced to fail,
4129 contradicting the general specification for all authenticators. Instead it
4130 was generating a temporary error. It now behaves as specified.
4132 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4133 (the order specifies the preference for clients). The order is now AES256,
4134 AES128, 3DES, ARCFOUR128.
4136 51. Small patch to Sieve code - explicitly set From: when generating an
4139 52. Exim crashed if a remote delivery caused a very long error message to be
4140 recorded - for instance if somebody sent an entire SpamAssassin report back
4141 as a large number of 550 error lines. This bug was coincidentally fixed by
4142 increasing the size of one of Exim's internal buffers (big_buffer) that
4143 happened as part of the Exiscan merge. However, to be on the safe side, I
4144 have made the code more robust (and fixed the comments that describe what
4147 53. Now that there can be additional text after "Completed" in log lines (if
4148 the queue_time_overall log selector is set), a one-byte patch to exigrep
4149 was needed to allow it to recognize "Completed" as not the last thing in
4152 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4153 patch that reportedly fixes this has been added. I am not expert enough to
4154 create a test for it. This is what the patch creator wrote:
4156 "I found a little strange behaviour of ldap code when working with
4157 Windows 2003 AD Domain, where users was placed in more than one
4158 Organization Units. When I tried to give exim partial DN, the exit code
4159 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4160 But simultaneously result of request was absolutely normal ldap result,
4161 so I produce this patch..."
4163 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4164 REFERENCE, so I have modified the code to exclude the patch when that macro
4167 55. Some experimental protocols are using DNS PTR records for new purposes. The
4168 keys for these records are domain names, not reversed IP addresses. The
4169 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4170 leaves it alone. Component reversal etc. now happens only for IP addresses.
4173 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4175 57. Double the size of the debug message buffer (to 2048) so that more of very
4176 long debug lines gets shown.
4178 58. The exicyclog utility now does better if the number of log files to keep
4179 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4181 59. Two changes related to the smtp_active_hostname option:
4183 (1) $smtp_active_hostname is now available as a variable.
4184 (2) The default for smtp_banner uses $smtp_active_hostname instead
4185 of $primary_hostname.
4187 60. The host_aton() function is supposed to be passed a string that is known
4188 to be a valid IP address. However, in the case of IPv6 addresses, it was
4189 not checking this. This is a hostage to fortune. Exim now panics and dies
4190 if the condition is not met. A case was found where this could be provoked
4191 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4192 components; fortuitously, this particular loophole had already been fixed
4193 by change 4.50/55 above.
4195 If there are any other similar loopholes, the new check in host_aton()
4196 itself should stop them being exploited. The report I received stated that
4197 data on the command line could provoke the exploit when Exim was running as
4198 exim, but did not say which command line option was involved. All I could
4199 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4200 running as the user.
4203 61. There was a buffer overflow vulnerability in the SPA authentication code
4204 (which came originally from the Samba project). I have added a test to the
4205 spa_base64_to_bits() function which I hope fixes it.
4208 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4209 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4211 63. The daemon start-up calls getloadavg() while still root for those OS that
4212 need the first call to be done as root, but it missed one case: when
4213 deliver_queue_load_max is set with deliver_drop_privilege. This is
4214 necessary for the benefit of the queue runner, because there is no re-exec
4215 when deliver_drop_privilege is set.
4217 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4218 This has been fixed.
4220 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4221 was in use, was not putting the data itself into the right store pool;
4222 consequently, it could be overwritten for a subsequent message in the same
4223 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4226 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4228 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4229 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4230 the function were treating the return as a boolean value, which happened to
4231 work because 0=false and not-0=true, but is not correct code.
4233 68. The host_aton() function was not handling scoped IPv6 addresses (those
4234 with, for example, "%eth0" on the end) correctly.
4236 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4237 negated items (that is, ~something) in unsigned ints. Some compilers
4238 apparently mutter when there is no cast.
4240 70. If an address verification called from an ACL failed, and did not produce a
4241 user-specific message (i.e. there was only a "system" message), nothing was
4242 put in $acl_verify_message. In this situation, it now puts the system
4245 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4246 session; change 4.31/43 added the unwanted input to the log line - except
4247 that it did not do this in the start of session case. It now does.
4249 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4250 This is wrong and can cause the other end to generate a synchronization
4251 error if it is another Exim or anything else that does the synchronization
4252 check. A QUIT command is no longer sent after a timeout.
4254 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4255 during host lookups.
4257 74. The defer_ok option of callout verification was not working if it was used
4258 when verifying addresses in header lines, that is, for this case:
4260 verify = header_sender/callout=defer_ok
4262 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4263 those file descriptors could be used for SMTP connections. If anything
4264 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4265 could be sent to the SMTP client, causing chaos. The daemon now opens
4266 stdin, stdout, and stderr to /dev/null when it puts itself into the
4269 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4270 log by default. The user can override this with suitable Perl magic.
4272 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4273 the log message when discard triggers, was not working for the DATA ACL or
4274 for the non-SMTP ACL.
4276 78. Error message wording change in sieve.c.
4278 79. If smtp_accept_max_per_host was set, the number of connections could be
4279 restricted to fewer than expected, because the daemon was trying to set up
4280 a new connection before checking whether the processes handling previous
4281 connections had finished. The check for completed processes is now done
4282 earlier. On busy systems, this bug wouldn't be noticed because something
4283 else would have woken the daemon, and it would have reaped the completed
4286 80. If a message was submitted locally by a user whose login name contained one
4287 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4288 It caused a spool format error. I have fixed the spool reading code. A
4289 related problem was that the "from" clause in the Received: line became
4290 illegal because of the space(s). It is now covered by ${quote_local_part.
4292 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4295 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4297 83. Patch from Sieve maintainer to fix unterminated string problem in
4298 "vacation" handling.
4300 84. Some minor changes to the Linux configuration files to help with other
4301 OS variants using glibc.
4303 85. One more patch for Sieve to update vacation handling to latest spec.
4306 ----------------------------------------------------
4307 See the note above about the 4.44 and 4.50 releases.
4308 ----------------------------------------------------
4314 1. Change 4.43/35 introduced a bug that caused file counts to be
4315 incorrectly computed when quota_filecount was set in an appendfile
4318 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4319 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4322 3. Give more explanation in the error message when the command for a transport
4323 filter fails to execute.
4325 4. There are several places where Exim runs a non-Exim command in a
4326 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4327 was being done only for the command run by the queryprogram router. It is
4328 now done for all such subprocesses. The other cases are: ${run, transport
4329 filters, and the commands run by the lmtp and pipe transports.
4331 5. Some older OS have a limit of 256 on the maximum number of file
4332 descriptors. Exim was using setrlimit() to set 1000 as a large value
4333 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4334 systems. I've change it so that if it can't get 1000, it tries for 256.
4336 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4337 was an oversight, and furthermore, ever since the addition of extra
4338 controls (e.g. 4.43/32), the checks on when to allow different forms of
4339 "control" were broken. There should now be diagnostics for all cases when a
4340 control that does not make sense is encountered.
4342 7. $recipients is now available in the predata ACL (oversight).
4344 8. Tidy the search cache before the fork to do a delivery from a message
4345 received from the command line. Otherwise the child will trigger a lookup
4346 failure and thereby defer the delivery if it tries to use (for example) a
4347 cached ldap connection that the parent has called unbind on.
4349 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4350 of $address_data from the recipient verification was clobbered by the
4351 sender verification.
4353 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4354 was its contents. (It was OK if the option was not defined at all.)
4356 11. A "Completed" log line is now written for messages that are removed from
4357 the spool by the -Mrm option.
4359 12. $host_address is now set to the target address during the checking of
4360 ignore_target_hosts.
4362 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4363 being passed; this would have caused $sender_host_name to have been used if
4364 matching the list had actually called for a host name (not very likely,
4365 since this list is usually IP addresses). A host name is now passed as
4368 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4369 code that specifies a non-restarting handler (typically sigaction() in
4370 modern systems) in an attempt to fix a rare and obscure crash bug.
4372 15. Narrowed the window for a race in the daemon that could cause it to ignore
4373 SIGCHLD signals. This is not a major problem, because they are used only to
4374 wake it up if nothing else does.
4376 16. A malformed maildirsize file could cause Exim to calculate negative values
4377 for the mailbox size or file count. Odd effects could occur as a result.
4378 The maildirsize information is now recalculated if the size or filecount
4381 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4382 support for a long time. Removed HAVE_SYS_VFS_H.
4384 18. Updated exipick to current release from John Jetmore.
4386 19. Allow an empty sender to be matched against a lookup in an address list.
4387 Previously the only cases considered were a regular expression, or an
4390 20. Exim went into a mad DNS lookup loop when doing a callout where the
4391 host was specified on the transport, if the DNS lookup yielded more than
4394 21. The RFC2047 encoding function was originally intended for short strings
4395 such as real names; it was not keeping to the 75-character limit for
4396 encoded words that the RFC imposes. It now respects the limit, and
4397 generates multiple encoded words if necessary. To be on the safe side, I
4398 have increased the buffer size for the ${rfc2047: expansion operator from
4401 22. Failure to deliver a bounce message always caused it to be frozen, even if
4402 there was an errors_to setting on the router. The errors_to setting is now
4405 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4406 canonical form (fully expanded) before being placed in
4407 $sender_host_address.
4409 24. Updated eximstats to version 1.33
4411 25. Include certificate and key file names in error message when GnuTLS fails
4412 to set them up, because the GnuTLS error message doesn't include the name
4413 of the failing file when there is a problem reading it.
4415 26. Expand error message when OpenSSL has problems setting up cert/key files.
4418 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4419 (this can affect the format of dates).
4421 28. exim_tidydb, when checking for the continued existence of a message for
4422 which it has found a message-specific retry record, was not finding
4423 messages that were in split spool directories. Consequently, it was
4424 deleting retry records that should have stayed in existence.
4426 29. eximstats updated to version 1.35
4427 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4428 1.35 - bugfix such that pie charts by volume are generated correctly
4430 30. The SPA authentication driver was not abandoning authentication and moving
4431 on to the next authenticator when an expansion was forced to fail,
4432 contradicting the general specification for all authenticators. Instead it
4433 was generating a temporary error. It now behaves as specified.
4435 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4436 (the order specifies the preference for clients). The order is now AES256,
4437 AES128, 3DES, ARCFOUR128.
4439 31. Small patch to Sieve code - explicitly set From: when generating an
4442 32. Exim crashed if a remote delivery caused a very long error message to be
4443 recorded - for instance if somebody sent an entire SpamAssassin report back
4444 as a large number of 550 error lines. This bug was coincidentally fixed by
4445 increasing the size of one of Exim's internal buffers (big_buffer) that
4446 happened as part of the Exiscan merge. However, to be on the safe side, I
4447 have made the code more robust (and fixed the comments that describe what
4450 33. Some experimental protocols are using DNS PTR records for new purposes. The
4451 keys for these records are domain names, not reversed IP addresses. The
4452 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4453 leaves it alone. Component reversal etc. now happens only for IP addresses.
4456 34. The host_aton() function is supposed to be passed a string that is known
4457 to be a valid IP address. However, in the case of IPv6 addresses, it was
4458 not checking this. This is a hostage to fortune. Exim now panics and dies
4459 if the condition is not met. A case was found where this could be provoked
4460 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4461 components; fortuitously, this particular loophole had already been fixed
4462 by change 4.50/55 or 4.44/33 above.
4464 If there are any other similar loopholes, the new check in host_aton()
4465 itself should stop them being exploited. The report I received stated that
4466 data on the command line could provoke the exploit when Exim was running as
4467 exim, but did not say which command line option was involved. All I could
4468 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4469 running as the user.
4472 35. There was a buffer overflow vulnerability in the SPA authentication code
4473 (which came originally from the Samba project). I have added a test to the
4474 spa_base64_to_bits() function which I hope fixes it.
4477 36. The daemon start-up calls getloadavg() while still root for those OS that
4478 need the first call to be done as root, but it missed one case: when
4479 deliver_queue_load_max is set with deliver_drop_privilege. This is
4480 necessary for the benefit of the queue runner, because there is no re-exec
4481 when deliver_drop_privilege is set.
4483 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4484 was in use, was not putting the data itself into the right store pool;
4485 consequently, it could be overwritten for a subsequent message in the same
4486 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4489 38. Sometimes the final signoff response after QUIT could fail to get
4490 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4491 < 0 before doing a fflush(). This bug looks as though it goes back to the
4492 introduction of TLS in release 3.20, but "sometimes" must have been rare
4493 because the tests only now provoked it.
4499 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4500 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4501 Now it returns an integer. A number of calls were still expecting a T/F
4502 return. Fortuitously, in all cases, the tests worked in OK situations,
4503 which is the norm. However, things would have gone wrong on any write
4504 failures on the smtp file descriptor. This function is used when sending
4505 messages over SMTP and also when doing verify callouts.
4507 2. When Exim is called to do synchronous delivery of a locally submitted
4508 message (the -odf or -odi options), it no longer closes stderr before doing
4511 3. Implemented the mua_wrapper option.
4513 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4515 5. Implemented the functions header_remove(), header_testname(),
4516 header_add_at_position(), and receive_remove_recipient(), and exported them
4519 6. If an ACL "warn" statement specified the addition of headers, Exim already
4520 inserted X-ACL-Warn: at the start if there was no header name. However, it
4521 was not making this test for the second and subsequent header lines if
4522 there were newlines in the string. This meant that an invalid header could
4523 be inserted if Exim was badly configured.
4525 7. Allow an ACL "warn" statement to add header lines at the start or after all
4526 the Received: headers, as well as at the end.
4528 8. Added the rcpt_4xx retry error code.
4530 9. Added postmaster_mailfrom=xxx to callout verification option.
4532 10. Added mailfrom=xxxx to the callout verification option, for verify=
4535 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4536 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4538 12. Inserted some casts to stop certain compilers warning when using pointer
4539 differences as field lengths or precisions in printf-type calls (mostly
4540 affecting debugging statements).
4542 13. Added optional readline() support for -be (dynamically loaded).
4544 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4545 same clock tick as a message's arrival, so that its received time was the
4546 same as the "first fail" time on the retry record, and that message
4547 remained on the queue past the ultimate address timeout, every queue runner
4548 would try a delivery (because it was past the ultimate address timeout) but
4549 after another failure, the ultimate address timeout, which should have then
4550 bounced the address, did not kick in. This was a "< instead of <=" error;
4551 in most cases the first failure would have been in the next clock tick
4552 after the received time, and all would be well.
4554 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4555 being recognized when the domain list was tested by the match_domain
4556 condition in an expansion string.
4558 16. Added the ${str2b64: operator.
4560 17. Exim was always calling setrlimit() to set a large limit for the number of
4561 processes, without checking whether the existing limit was already
4562 adequate. (It did check for the limit on file descriptors.) Furthermore,
4563 errors from getrlimit() and setrlimit() were being ignored. Now they are
4564 logged to the main and panic logs, but Exim does carry on, to try to do its
4565 job under whatever limits there are.
4567 18. Imported PCRE 5.0.
4569 19. Trivial typo in log message " temporarily refused connection" (the leading
4572 20. If the log selector return_path_on_delivery was set and an address was
4573 redirected to /dev/null, the delivery process crashed because it assumed
4574 that a return path would always be set for a "successful" delivery. In this
4575 case, the whole delivery is bypassed as an optimization, and therefore no
4578 21. Internal re-arrangement: the function for sending a challenge and reading
4579 a response while authentication was assuming a zero-terminated challenge
4580 string. It's now changed to take a pointer and a length, to allow for
4581 binary data in such strings.
4583 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4585 23. Exim was not respecting finduser_retries when seeking the login of the
4586 uid under which it was called; it was always trying 10 times. (The default
4587 setting of finduser_retries is zero.) Also, it was sleeping after the final
4588 failure, which is pointless.
4590 24. Implemented tls_on_connect_ports.
4592 25. Implemented acl_smtp_predata.
4594 26. If the domain in control=submission is set empty, Exim assumes that the
4595 authenticated id is a complete email address when it generates From: or
4596 Sender: header lines.
4598 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4599 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4600 chown and chgrp in /bin and hostname in /usr/bin.
4602 28. Exim was keeping the "process log" file open after each use, just as it
4603 does for the main log. This opens the possibility of it remaining open for
4604 long periods when the USR1 signal hits a daemon. Occasional processlog
4605 errors were reported, that could have been caused by this. Anyway, it seems
4606 much more sensible not to leave this file open at all, so that is what now
4609 29. The long-running daemon process does not normally write to the log once it
4610 has entered its main loop, and it closes the log before doing so. This is
4611 so that log files can straightforwardly be renamed and moved. However,
4612 there are a couple of unusual error situations where the daemon does write
4613 log entries, and I had neglected to close the log afterwards.
4615 30. The text of an SMTP error response that was received during a remote
4616 delivery was being truncated at 512 bytes. This is too short for some of
4617 the long messages that one sometimes sees. I've increased the limit to
4620 31. It is now possible to make retry rules that apply only when a message has a
4621 specific sender, in particular, an empty sender.
4623 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4624 it possible to be selective about when SMTP synchronization is enforced.
4626 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4628 32. Implemented hosts_connection_nolog.
4630 33. Added an ACL for QUIT.
4632 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4635 35. Added mailbox_size and mailbox_filecount to appendfile.
4637 36. Added control = no_multiline_responses to ACLs.
4639 37. There was a bug in the logic of the code that waits for the clock to tick
4640 in the case where the clock went backwards by a substantial amount such
4641 that the microsecond fraction of "now" was more than the microsecond
4642 fraction of "then" (but the whole seconds number was less).
4644 38. Added support for the libradius Radius client library this is found on
4645 FreeBSD (previously only the radiusclient library was supported).
4651 1. When certain lookups returned multiple values in the form name=value, the
4652 quoting of the values was not always being done properly. Specifically:
4653 (a) If the value started with a double quote, but contained no whitespace,
4655 (b) If the value contained whitespace other than a space character (i.e.
4656 tabs or newlines or carriage returns) it was not quoted.
4657 This fix has been applied to the mysql and pgsql lookups by writing a
4658 separate quoting function and calling it from the lookup code. The fix
4659 should probably also be applied to nisplus, ibase and oracle lookups, but
4660 since I cannot test any of those, I have not disturbed their existing code.
4662 2. A hit in the callout cache for a specific address caused a log line with no
4663 reason for rejecting RCPT. Now it says "Previous (cached) callout
4664 verification failure".
4666 3. There was an off-by-one bug in the queryprogram router. An over-long
4667 return line was truncated at 256 instead of 255 characters, thereby
4668 overflowing its buffer with the terminating zero. As well as fixing this, I
4669 have increased the buffer size to 1024 (and made a note to document this).
4671 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4672 when Exim is waiting for an SMTP response from a remote server, Exim
4673 restarts its select() call on the socket, thereby resetting its timeout.
4674 This is not a problem when such interrupts are rare. Somebody set up a cron
4675 job to run exiwhat every 2 minutes, which is less than the normal select()
4676 timeout (5 or 10 minutes). This meant that the select() timeout never
4677 kicked in because it was always reset. I have fixed this by comparing the
4678 time when an interrupt arrives with the time at the start of the first call
4679 to select(). If more time than the timeout has elapsed, the interrupt is
4680 treated as a timeout.
4682 5. Some internal re-factoring in preparation for the addition of Sieve
4683 extensions (by MH). In particular, the "personal" test is moved to a
4684 separate function, and given an option for scanning Cc: and Bcc: (which is
4685 not set for Exim filters).
4687 6. When Exim created an email address using the login of the caller as the
4688 local part (e.g. when creating a From: or Sender: header line), it was not
4689 quoting the local part when it contained special characters such as @.
4691 7. Installed new OpenBSD configuration files.
4693 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4694 try to make them clearer.
4696 9. Callout options, other than the timeout value, were being ignored when
4697 verifying sender addresses in header lines. For example, when using
4699 verify = header_sender/callout=no_cache
4701 the cache was (incorrectly) being used.
4703 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4704 only the Cygwin environment.
4706 11. When return_path_on_delivery was set as a log selector, if different remote
4707 addresses in the same message used different return paths and parallel
4708 remote delivery occurred, the wrong values would sometimes be logged.
4709 (Whenever a remote delivery process finished, the return path value from
4710 the most recently started remote delivery process was logged.)
4712 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4713 lines when AUTH and/or TLS are in use. This is the "received protocol"
4714 field. Exim used to use "asmtp" for authenticated SMTP, without any
4715 indication (in the protocol name) for TLS use. Now it follows the RFC and
4716 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4717 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4718 names appear in log lines as well as in Received: header lines.
4720 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4721 extensions, and comparison tests, and to fix some bugs.
4723 14. Changes to the "personal" filter test:
4725 (1) The test was buggy in that it was just doing the equivalent of
4726 "contains" tests on header lines. For example, if a user's address was
4727 anne@some.where, the "personal" test would incorrectly be true for
4729 To: susanne@some.where
4731 This test is now done by extracting each address from the header in turn,
4732 and checking the entire address. Other tests that are part of "personal"
4733 are now done using regular expressions (for example, to check local parts
4734 of addresses in From: header lines).
4736 (2) The list of non-personal local parts in From: addresses has been
4737 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4738 taken from the Sieve specification recommendations.
4740 (3) If the message contains any header line starting with "List-" it is
4741 treated as non-personal.
4743 (4) The test for "circular" in the Subject: header line has been removed
4744 because it now seems ill-conceived.
4746 15. Minor typos in src/EDITME comments corrected.
4748 16. Installed latest exipick from John Jetmore.
4750 17. If headers_add on a router specified a text string that was too long for
4751 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4752 of string_sprintf() is now avoided.
4754 18. $message_body_size was not set (it was always zero) when running the DATA
4755 ACL and the local_scan() function.
4757 19. For the "mail" command in an Exim filter, no default was being set for
4758 the once_repeat time, causing a random time value to be used if "once" was
4759 specified. (If the value happened to be <= 0, no repeat happened.) The
4760 default is now 0s, meaning "never repeat". The "vacation" command was OK
4761 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4762 (I found it when inspecting the code).
4764 20. There is now an overall timeout for performing a callout verification. It
4765 defaults to 4 times the callout timeout, which applies to individual SMTP
4766 commands during the callout. The overall timeout applies when there is more
4767 than one host that can be tried. The timeout is checked before trying the
4768 next host. This prevents very long delays if there are a large number of
4769 hosts and all are timing out (e.g. when the network connections are timing
4770 out). The value of the overall timeout can be changed by specifying an
4771 additional sub-option for "callout", called "maxwait". For example:
4773 verify = sender/callout=5s,maxwait=20s
4775 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4776 to the end before writing, but this should make it even safer).
4778 22. Exim was forgetting that it had advertised PIPELINING for the second and
4779 subsequent messages on an SMTP connection. It was also not resetting its
4780 memory on STARTTLS and an internal HELO.
4782 23. When Exim logs an SMTP synchronization error within a session, it now
4783 records whether PIPELINING has been advertised or not.
4785 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4786 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4787 rather than long int.
4789 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4791 26. Added the never_mail option to autoreply.
4797 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4798 crash if the getsockname() call failed; this can happen if a connection is
4799 closed very soon after it is established. The problem was simply in the
4800 order in which certain operations were done, causing Exim to try to write
4801 to the SMTP stream before it had set up the file descriptor. The bug has
4802 been fixed by making things happen in the correct order.
4808 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4809 before the connection was closed, thus losing the rejection response.
4811 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4812 some early Solaris releases, but causes trouble in current releases where
4813 socklen_t is defined.
4815 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4818 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4821 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4822 that releasing the top of it at the end releases what was used for sub-
4823 expansions (unless the block got too big). However, discard this block if
4824 the first thing is a variable or header, so that we can use its block when
4825 it is dynamic (useful for very large $message_headers, for example).
4827 6. Lookups now cache *every* query, not just the most recent. A new, separate
4828 store pool is used for this. It can be recovered when all lookup caches are
4829 flushed. Lookups now release memory at the end of their result strings.
4830 This has involved some general refactoring of the lookup sources.
4832 7. Some code has been added to the store_xxx() functions to reduce the amount
4833 of flapping under certain conditions.
4835 8. log_incoming_interface used to affect only the <= reception log lines. Now
4836 it causes the local interface and port to be added to several more SMTP log
4837 lines, for example "SMTP connection from", and rejection lines.
4839 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4841 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4843 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4844 could be overwritten at the end of the current message (or the start of a
4845 new message if it was set in a HELO ACL). The value is now preserved for
4846 the duration of the SMTP connection.
4848 12. If a transport had a headers_rewrite setting, and a matching header line
4849 contained an unqualified address, that address was qualified, even if it
4850 did not match any rewriting rules. The underlying bug was that the values
4851 of the flags that permit the existence of unqualified sender and recipient
4852 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4853 non-local messages, and by -bnq for local messages) were not being
4854 preserved with the message after it was received.
4856 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4857 "next input=" as part of the text comprising the host identity instead of
4858 the correct text. The code was using the same buffer for two different
4859 strings. However, depending on which order the printing function evaluated
4860 its arguments, the bug did not always show up. Under Linux, for example, my
4861 test suite worked just fine.
4863 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4864 doesn't work with some older versions of Perl. It has been changed to "my",
4865 which in any case is probably the better facility to use.
4867 15. A really picky compiler found some instances of statements for creating
4868 error messages that either had too many or two few arguments for the format
4871 16. The size of the buffer for calls to the DNS resolver has been increased
4872 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4873 for addresses that have a lot of PTR records. This alleviates a problem; it
4874 does not fully solve it.
4876 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4877 buffer now truncates the list and logs the incident, which is the same
4878 action as happens when Exim is looking up a host name and its aliases.
4879 Previously in this situation something unpredictable would happen;
4880 sometimes it was "internal error: store_reset failed".
4882 18. If a server dropped the connection unexpectedly when an Exim client was
4883 using GnuTLS and trying to read a response, the client delivery process
4884 crashed while trying to generate an error log message.
4886 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4887 string, for example:
4889 warn message = H1: something\nH2: something
4891 the text was added as a single header line from Exim's point of view
4892 though it ended up OK in the delivered message. However, searching for the
4893 second and subsequent header lines using $h_h2: did not work. This has been
4894 fixed. Similarly, if a system filter added multiple headers in this way,
4895 the routers could not see them.
4897 20. Expanded the error message when iplsearch is called with an invalid key to
4898 suggest using net-iplsearch in a host list.
4900 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4901 ACLs are no longer actually imposed (and a message to that effect is
4904 22. If a "gecos" field in a passwd entry contained escaped characters, in
4905 particular, if it contained a \" sequence, Exim got it wrong when building
4906 a From: or a Sender: header from that name. A second bug also caused
4907 incorrect handling when an unquoted " was present following a character
4908 that needed quoting.
4910 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4911 was not being matched caselessly.
4913 24. Arranged for all hyphens in the exim.8 source to be escaped with
4916 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4917 properly. Recipient callout cache records were still being keyed to include
4918 the sender, even when use_sender was set false. This led to far more
4919 callouts that were necessary. The sender is no longer included in the key
4920 when use_sender is false.
4922 26. Added "control = submission" modifier to ACLs.
4924 27. Added the ${base62d: operator to decode base 62 numbers.
4926 28. dnsdb lookups can now access SRV records.
4928 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4929 the configuration file.
4931 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4932 (-v) mode. This makes the output for a verbose queue run more intelligible.
4934 31. Added a use_postmaster feature to recipient callouts.
4936 32. Added the $body_zerocount variable, containing the number of binary zero
4937 bytes in the message body.
4939 33. The time of last modification of the "new" subdirectory is now used as the
4940 "mailbox time last read" when there is a quota error for a maildir
4943 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4945 35. Added +ignore_unknown as a special item in host lists.
4947 36. Code for decoding IPv6 addresses in host lists is now included, even if
4948 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4949 address was recognized as an IP address, but was then not correctly decoded
4950 into binary, causing unexpected and incorrect effects when compared with
4957 1. Very minor rewording of debugging text in manualroute to say "list of
4958 hosts" instead of "hostlist".
4960 2. If verify=header_syntax was set, and a header line with an unqualified
4961 address (no domain) and a large number of spaces between the end of the
4962 name and the colon was received, the reception process suffered a buffer
4963 overflow, and (when I tested it) crashed. This was caused by some obsolete
4964 code that should have been removed. The fix is to remove it!
4966 3. When running in the test harness, delay a bit after writing a bounce
4967 message to get a bit more predictability in the log output.
4969 4. Added a call to search_tidyup() just before forking a reception process. In
4970 theory, someone could use a lookup in the expansion of smtp_accept_max_
4971 per_host which, without the tidyup, could leave open a database connection.
4973 5. Added the variables $recipient_data and $sender_data which get set from a
4974 lookup success in an ACL "recipients" or "senders" condition, or a router
4975 "senders" option, similar to $domain_data and $local_part_data.
4977 6. Moved the writing of debug_print from before to after the "senders" test
4980 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4981 problems for message scanning, either using a data ACL, or using
4982 local_scan() because the Received: header was not generated till after they
4983 were called (in order to set the time as the time of reception completion).
4984 I have revised the way this works. The header is now generated after the
4985 body is received, but before the ACL or local_scan() are called. After they
4986 are run, the timestamp in the header is updated.
4992 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4993 before starting a queue runner without re-exec. This happened only when
4994 deliver_drop_privilege was set or when the Exim user was set to root. The
4995 effect of the bug was that timeouts during subsequent deliveries caused
4996 crashes instead of being properly handled. The handler is now left at its
4997 default (and expected) setting.
4999 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5000 message, again when deliver_drop_privilege is set or Exim is run as root.
5001 The bug described in (1) was not present in this case, but the tidying up
5002 of the other signals was missing. I have made the two cases consistent.
5004 3. The ignore_target_hosts setting on a manualroute router was being ignored
5005 for hosts that were looked up using the /MX notation.
5007 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5010 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5011 operated on the sender address. After changing the $sender_address to <>
5012 for the sender address verify, Exim was re-instated it as the original
5013 (before rewriting) address, but remembering that it had rewritten it, so it
5014 wasn't rewriting it again. This bug also had the effect of breaking the
5015 sender address verification caching when the sender address was rewritten.
5017 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5018 This has been changed so that if the ip literal address matches
5019 ignore_target_hosts, the router declines.
5021 7. Added expansion conditions match_domain, match_address, and match_local_
5022 part (NOT match_host).
5024 8. The placeholder for the Received: header didn't have a length field set.
5026 9. Added code to Exim itself and to exim_lock to test for a specific race
5027 condition that could lead to file corruption when using MBX delivery. The
5028 issue is with the lockfile that is created in /tmp. If this file is removed
5029 after a process has opened it but before that process has acquired a lock,
5030 there is the potential for a second process to recreate the file and also
5031 acquire a lock. This could lead to two Exim processes writing to the file
5032 at the same time. The added code performs the same test as UW imapd; it
5033 checks after acquiring the lock that its file descriptor still refers to
5034 the same named file.
5036 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5037 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5040 11. Added the smtp_active_hostname option. If used, this will typically be made
5041 to depend on the incoming interface address. Because $interface_address is
5042 not set up until the daemon has forked a reception process, error responses
5043 that can happen earlier (such as "too many connections") no longer contain
5046 12. If an expansion in a condition on a "warn" statement fails because a lookup
5047 defers, the "warn" statement is abandoned, and the next ACL statement is
5048 processed. Previously this caused the whole ACL to be aborted.
5050 13. Added the iplsearch lookup type.
5052 14. Added ident_timeout as a log selector.
5054 15. Added tls_certificate_verified as a log selector.
5056 16. Added a global option tls_require_ciphers (compare the smtp transport
5057 option of the same name). This controls incoming TLS connections.
5059 17. I finally figured out how to make tls_require_ciphers do a similar thing
5060 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5061 before starting the TLS session.
5063 18. Tabs are now shown as \t in -bP output.
5065 19. If the log selector return_path_on_delivery was set, Exim crashed when
5066 bouncing a message because it had too many Received: header lines.
5068 20. If two routers both had headers_remove settings, and the first one included
5069 a superfluous trailing colon, the final name in the first list and the
5070 first name in the second list were incorrectly joined into one item (with a
5071 colon in the middle).
5077 1. Added -C and -D options to the exinext utility, mainly to make it easier
5078 to include in the automated testing, but these could be helpful when
5079 multiple configurations are in use.
5081 2. The exinext utility was not formatting the output nicely when there was
5082 an alternate port involved in the retry record key, nor when there was a
5083 message id as well (for retries that were specific to a specific message
5084 and a specific host). It was also confused by IPv6 addresses, because of
5085 the additional colons they contain. I have fixed the IPv4 problem, and
5086 patched it up to do a reasonable job for IPv6.
5088 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5089 delivery, the log line now contains "pipelined" if PIPELINING was used.
5091 4. An SMTP transport process used to panic and die if the bind() call to set
5092 an explicit outgoing interface failed. This has been changed; it is now
5093 treated in the same way as a connect() failure.
5095 5. A reference to $sender_host_name in the part of a conditional expansion
5096 that was being skipped was still causing a DNS lookup. This no longer
5099 6. The def: expansion condition was not recognizing references to header lines
5100 that used bh_ and bheader_.
5102 7. Added the _cache feature to named lists.
5104 8. The code for checking quota_filecount in the appendfile transport was
5105 allowing one more file than it should have been.
5107 9. For compatibility with Sendmail, the command line option
5115 and sets the incoming protocol and host name (for trusted callers). The
5116 host name and its colon can be omitted when only the protocol is to be set.
5117 Note the Exim already has two private options, -pd and -ps, that refer to
5118 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5119 "s", but I don't think that's a major issue.
5121 10. A number of refactoring changes to the code, none of which should affect
5124 (a) The number of logging options was getting close to filling up the
5125 32-bit word that was used as a bit map. I have split them into two classes:
5126 those that are passed in the argument to log_write(), and those that are
5127 only ever tested independently outside of that function. These are now in
5128 separate 32-bit words, so there is plenty of room for expansion again.
5129 There is no change in the user interface or the logging behaviour.
5131 (b) When building, for example, log lines, the code previously used a
5132 macro that called string_cat() twice, in order to add two strings. This is
5133 not really sufficiently general. Furthermore, there was one instance where
5134 it was actually wrong because one of the argument was used twice, and in
5135 one call a function was used. (As it happened, calling the function twice
5136 did not affect the overall behaviour.) The macro has been replaced by a
5137 function that can join an arbitrary number of extra strings onto a growing
5140 (c) The code for expansion conditions now uses a table and a binary chop
5141 instead of a serial search (which was left over from when there were very
5142 few conditions). Also, it now recognizes conditions like "pam" even when
5143 the relevant support is not compiled in: a suitably worded error message is
5144 given if an attempt is made to use such a condition.
5146 11. Added ${time_interval:xxxxx}.
5148 12. A bug was causing one of the ddress fields not to be passed back correctly
5149 from remote delivery subprocesses. The field in question was not being
5150 subsequently used, so this caused to problems in practice.
5152 13. Added new log selectors queue_time and deliver_time.
5154 14. Might have fixed a bug in maildirsizefile handling that threw up
5155 "unexpected character" debug warnings, and recalculated the data
5156 unnecessarily. In any case, I expanded the warning message to give more
5159 15. Added the message "Restricted characters in address" to the statements in
5160 the default ACL that block characters like @ and % in local parts.
5162 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5163 Three changes have been made:
5165 (a) There was a serious bug; a negative response to MAIL caused the whole
5166 recipient domain to be cached as invalid, thereby blocking all messages
5167 to all local parts at the same domain, from all senders. This bug has
5168 been fixed. The domain is no longer cached after a negative response to
5169 MAIL if the sender used is not empty.
5171 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5174 (c) A new callout option, "use_sender" has been added for people who want
5175 the modified behaviour.
5181 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5184 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5185 indeed breaks things for older releases.
5187 3. Added additional logging to the case where there is a problem reading data
5188 from a filter that is running in a subprocess using a pipe, in order to
5189 try to track down a specific problem.
5191 4. Testing facility fudge: when running in the test harness and attempting
5192 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5193 getting "No route to host". Convert this to a timeout.
5195 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5198 6. Some OS don't have socklen_t but use size_t instead. This affects the
5199 fifth argument of getsockopt() amongst other things. This is now
5200 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5201 can be set for individual OS. I have set it for SunOS5, OSF1, and
5202 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5203 some earlier ones do not.
5205 7. Change 4.30/15 was not doing the test caselessly.
5207 8. The standard form for an IPv6 address literal was being rejected by address
5208 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5209 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5210 this, as well as the form without the "IPv6" on the front (but only when
5211 address literals are enabled, of course).
5213 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5215 10. Exim crashed if a message with an empty sender address specified by -f
5216 encountered a router with an errors_to setting. This could be provoked only
5217 by a command such as
5221 where an empty string was supplied; "<>" did not hit this bug.
5223 11. Installed PCRE release 4.5.
5225 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5226 remained set. It is now erased.
5228 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5229 times from message ids (which are base 36 rather than the normal 62).
5231 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5232 were being counted as actual protocol errors, and logged if the log
5233 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5234 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5235 and DATA following a set of rejected RCPTs do not count as protocol errors.
5236 In other words, Exim assumes they were pipelined, though this may not
5237 actually be the case. Of course, in all cases the client gets an
5238 appropriate error code.
5240 15. If a lookup fails in an ACL condition, a message about the failure may
5241 be available; it is used if testing the ACL cannot continue, because most
5242 such messages specify what the cause of the deferral is. However, some
5243 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5244 that caused an old message to be retained and used if a later statement
5245 caused a defer, replacing the real cause of the deferral.
5247 16. If an IP address had so many PTR records that the DNS lookup buffer
5248 was not large enough to hold them, Exim could crash while trying to process
5249 the truncated data. It now detects and logs this case.
5251 17. Further to 4.21/58, another change has been made: if (and only if) the
5252 first line of a message (the first header line) ends with CRLF, a bare LF
5253 in a subsequent header line has a space inserted after it, so as not to
5254 terminate the header.
5256 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5257 unnecessarily, used atoi() instead of strtol(), and didn't check the
5258 termination when getting file sizes from file names by regex.
5260 19. Completely re-implemented the support for maildirsize files, in the light
5261 of a number of problems with the previous contributed implementation
5262 (4.30/29). In particular:
5264 . If the quota is zero, the maildirsize file is maintained, but no quota is
5267 . If the maildir directory does not exist, it is created before any attempt
5268 to write a maildirsize file.
5270 . The quota value in the file is just a cache; if the quota is changed in
5271 the transport, the new value overrides.
5273 . A regular expression is available for excluding directories from the
5276 20. The autoreply transport checks the characters in options that define the
5277 message's headers; it allows continued headers, but it was checking with
5278 isspace() after an embedded newline instead of explicitly looking for a
5281 21. If all the "regular" hosts to which an address was routed had passed their
5282 expiry times, and had not reached their retry times, the address was
5283 bounced, even if fallback hosts were defined. Now Exim should go on to try
5286 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5287 equivalent code in the SMTP transport. Some hosts send humungous responses
5288 to HELO/EHLO, more than 1024 it seems.
5290 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5291 gives compiler warnings in some environments. I've now done it "properly",
5294 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5295 (because of problems with the built-in one) was declared to return uschar *
5296 instead of char *, causing compiler failure.
5298 25. Fixed a file descriptor leak when processing alias/forward files.
5300 26. Fixed a minor format string issue in dbfn.c.
5302 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5304 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5305 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5308 29. When a sender address is verified, it is cached, to save repeating the test
5309 when there is more than one recipient in a message. However, when the
5310 verification involves a callout, it is possible for different callout
5311 options to be set for different recipients. It is too complicated to keep
5312 track of this in the cache, so now Exim always runs a verification when a
5313 callout is required, relying on the callout cache for the optimization.
5314 The overhead is duplication of the address routing, but this should not be
5317 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5318 to be verified with callout=postmaster, and the main callout worked but the
5319 postmaster check failed, the verification correctly failed. However, if a
5320 subsequent RCPT command asked for sender verification *without* the
5321 postmaster check, incorrect caching caused this verification also to fail,
5324 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5325 it was not caching the DNS options (qualify_single, search_parents) that
5326 were used when the lookup failed. A subsequent lookup with different
5327 options therefore always gave the same answer, though there were cases
5328 where it should not have. (Example: a "domains = !$mx_any" option on a
5329 dnslookup router: the "domains" option is always processed without any
5330 widening, but the router might have qualify_single set.) Now Exim uses the
5331 cached value only when the same options are set.
5333 32. Added John Jetmore's "exipick" utility to the distribution.
5335 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5336 than a timeout (e.g. a certificate is required, and is not provided), an
5337 Exim server now closes the connection immediately. Previously it waited for
5338 the client to close - but if the client is SSL, it seems that they each
5339 wait for each other, leading to a delay before one of them times out.
5341 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5342 maintained 0.8.x compatibility because I don't think many are using it, and
5343 it is clearly obsolete.
5345 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5348 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5349 client certificate was expired. A simple patch fixes this, though I don't
5350 understand the full logic of why the verify callback is called multiple
5353 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5354 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5355 which causes problems with some clients (such as the Certicom SSL Plus
5356 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5357 disables the coutermeasure allowing Eudora to connect."
5359 38. Exim was not checking that a write() to a log file succeeded. This could
5360 lead to Bad Things if a log got too big, in particular if it hit a file
5361 size limit. Exim now panics and dies if it cannot write to a log file, just
5362 as it does if it cannot open a log file.
5364 39. Modified OS/Makefile-Linux so that it now contains
5366 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5368 The two -D definitions ensure that Exim is compiled with large file
5369 support, which makes it possible to handle log files that are bigger than
5372 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5373 instance) a domain was checked against a named list that involved a lookup,
5374 causing $domain_data to be set, then another domain was checked against the
5375 same list, then the first domain was re-checked, the value of $domain_data
5376 after the final check could be wrong. In particular, if the second check
5377 failed, it could be set empty. This bug probably also applied to
5380 41. The strip_trailing_dot option was not being applied to the address given
5381 with the -f command-line option.
5383 42. The code for reading a message's header from the spool was incrementing
5384 $received_count, but never initializing it. This meant that the value was
5385 incorrect (doubled) while delivering a message in the same process in which
5386 it was received. In the most common configuration of Exim, this never
5387 happens - a fresh exec is done - but it can happen when
5388 deliver_drop_privilege is set.
5390 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5391 it now includes up to 150 characters of the unexpected data in the log
5394 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5395 and building data strings. The size of both of these buffers was 10 000
5396 bytes - far larger than anybody would *ever* want, thought I. Needless to
5397 say, somebody hit the limit. I have increased the maximum line length to
5398 20 000 and the maximum data length of concatenated lines to 100 000. I have
5399 also fixed two bugs, because there was no checking on these buffers. Tsk,
5400 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5401 buffer is too small.
5403 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5404 lsearch lookups. Now it does.
5406 46. When parsing a route_list item in a manualroute router, a fixed-length
5407 buffer was used for the list of hosts. I made this 1024 bytes long,
5408 thinking that nobody would ever have a list of hosts that long. Wrong.
5409 Somebody had a whole pile of complicated expansion conditions, and the
5410 string was silently truncated, leading to an expansion error. It turns out
5411 that it is easier to change to an unlimited length (owing to other changes
5412 that have happened since this code was originally written) than to build
5413 structure for giving a limitation error. The length of the item that
5414 expands into the list of hosts is now unlimited.
5416 47. The lsearch lookup could not handle data where the length of text line was
5417 more than 4095 characters. Such lines were truncated, leading to shortened
5418 data being returned. It should now handle lines of any length.
5420 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5421 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5424 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5425 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5426 no longer happens in comment lines. A list of replacements is now placed
5427 at the head of all of the source files, except those whose only change is
5428 to replace PERL_COMMAND in the very first #! line.
5430 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5431 messages on the queue, with a bottom-up merge sort, using code contributed
5432 by Michael Haardt. This should make operations like -bp somewhat faster on
5433 large queues. It won't affect queue runners, except when queue_run_in_order
5436 51. Installed eximstats 1.31 in the distribution.
5438 52. Added support for SRV lookups to the dnslookup router.
5440 53. If an ACL referred to $message_body or $message_body_end, the value was not
5441 reset for any messages that followed in the same SMTP session.
5443 54. The store-handling optimization for building very long strings was not
5444 differentiating between the different store pools. I don't think this
5445 actually made any difference in practice, but I've tidied it.
5447 55. While running the routers to verify a sender address, $sender_address
5448 was still set to the sender address. This is wrong, because when routing to
5449 send a bounce to the sender, it would be empty. Therefore, I have changed
5450 it so that, while verifying a sender address, $sender_address is set to <>.
5451 (There is no change to what happens when verifying a recipient address.)
5453 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5454 target A or AAAA records (if not already returned) without resetting the
5455 qualify_single or search_parents options of the DNS resolver. These are
5456 inappropriate in this case because the targets of MX and SRV records must
5457 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5458 target that, when qualified, matched something in the local domain. These
5459 two options are now turned off when doing these lookups.
5461 57. It seems that at least some releases of Reiserfs (which does not have the
5462 concept of a fixed number of inodes) returns zero and not -1 for the
5463 number of available inodes. This interacted badly with check_spool_inodes,
5464 which assumed that -1 was the "no such thing" setting. What I have done is
5465 to check that the total number of inodes is greater than zero before doing
5466 the test of how many are available.
5468 58. When a "warn" ACL statement has a log_message modifier, the message is
5469 remembered, and not repeated. This is to avoid a lot of repetition when a
5470 message has many recipients that cause the same warning to be written.
5471 Howewer, Exim was preserving the list of already written lines for an
5472 entire SMTP session, which doesn't seem right. The memory is now reset if a
5473 new message is started.
5475 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5476 debugging output unless log_rewrite was also set.
5478 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5479 of (int)(handle) when we know that handle contains (void *)(-1).
5481 61. The Exim daemon panic-logs an error return when it closes the incoming
5482 connection. However "connection reset by peer" seems to be common, and
5483 isn't really an error worthy of noting specially, so that particular error
5486 62. When Exim is trying to find all the local interfaces, it used to panic and
5487 die if the ioctl to get the interface flags failed. However, it seems that
5488 on at least one OS (Solaris 9) it is possible to have an interface that is
5489 included in the list of interfaces, but for which you get a failure error
5490 for this call. This happens when the interface is not "plumbed" into a
5491 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5492 failure of the "get flags" call assumes that the interface is down.
5494 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5495 makes life easier for people who are doing arithmetic on fields extracted
5496 from dates, where you often get leading zeros that should not be
5497 interpreted as octal.
5499 64. Added qualify_domain to the redirect router, to override the global
5502 65. If a pathologically long header line contained very many addresses (the
5503 report of this problem mentioned 10 000) and each of them was rewritten,
5504 Exim could use up a very large amount of memory. (It kept on making new
5505 copies of the header line as it rewrote, and never released the old ones.)
5506 At the expense of a bit more processing, the header rewriting function has
5507 been changed so that it no longer eats memory in this way.
5509 66. The generation of the Received: header has been moved from the time that a
5510 message starts to be received, to the time that it finishes. The timestamp
5511 in the Received: header should now be very close to that of the <= log
5512 line. There are two side-effects of this change:
5514 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5515 logged header lines no longer include the local Received: line, because
5516 it has not yet been created. The same applies to a copy of the message
5517 that is returned to a non-SMTP sender when a message is rejected.
5519 (b) When a filter file is tested using -bf, no additional Received: header
5520 is added to the test message. After some thought, I decided that this
5523 This change does not affect the value of $received_for. It is still set
5524 after address rewriting, but before local_scan() is called.
5526 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5528 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5529 gave an unhelpful panic error message, and a defer error. I have managed to
5530 change this behaviour so that it now rejects any supplied certificate,
5531 which seems right, as the list of acceptable certificates is empty.
5533 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5534 gave an unhelpful defer error. I have not managed to make this reject any
5535 supplied certificates, but the error message it gives is "no certificate
5536 supplied", which is not helpful.
5538 70. exigrep's output now also includes lines that are not associated with any
5539 message, but which match the given pattern. Implemented by a patch from
5540 Martin Sluka, which also tidied up the Perl a bit.
5542 71. Recipient callout verification, like sender verification, was using <> in
5543 the MAIL FROM command. This isn't really the right thing, since the actual
5544 sender may affect whether the remote host accepts the recipient or not. I
5545 have changed it to use the actual sender in the callout; this means that
5546 the cache record is now keyed on a recipient/sender pair, not just the
5547 recipient address. There doesn't seem to be a real danger of callout loops,
5548 since a callout by the remote host to check the sender would use <>.
5549 [SEE ABOVE: changed after hitting problems.]
5551 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5552 temporary errors. However, in the case of such a code being given after
5553 the end of a data transmission (i.e. after ".") Exim was failing to write
5554 a retry record for the message. (Yes, there was some broken host that was
5555 actually sending 8xx at this point.)
5557 73. An unknown lookup type in a host list could cause Exim to panic-die when
5558 the list was checked. (An example that provoked this was putting <; in the
5559 middle of a list instead of at the start.) If this happened during a DATA
5560 ACL check, a -D file could be left lying around. This kind of configuration
5561 error no longer causes Exim to die; instead it causes a defer errror. The
5562 incident is still logged to the main and panic logs.
5564 74. Buglet left over from Exim 3 conversion. The message "too many messages
5565 in one connection" was written to the rejectlog but not the mainlog, except
5566 when address rewriting (yes!) was being logged.
5568 75. Added write_rejectlog option.
5570 76. When a system filter was run not as root (that is, when system_filter_user
5571 was set), the values of the $n variables were not being returned to the
5572 main process; thus, they were not subsequently available in the $sn
5575 77. Added +return_path_on_delivery log selector.
5577 78. A connection timeout was being treated differently from recipients deferred
5578 when testing hosts_max_try with a message that was older than the host's
5579 retry timeout. (The host should not be counted, thus allowing all hosts to
5580 be tried at least once before bouncing.) This may have been the cause of an
5581 occasionally reported bug whereby a message would remain on the queue
5582 longer than the retry timeout, but would be bounced if a delivery was
5583 forced. I say "may" because I never totally pinned down the problem;
5584 setting up timeout/retry tests is difficult. See also the next item.
5586 79. The ultimate address timeout was not being applied to errors that involved
5587 a combination of host plus message (for example, a timeout on a MAIL
5588 command). When an address resolved to a number of possible hosts, and they
5589 were not all tried for each delivery (e.g. because of hosts_max_try), a
5590 message could remain on the queue longer than the retry timeout.
5592 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5595 81. Fixed an obscure SMTP outgoing bug which required at least the following
5596 conditions: (a) there was another message waiting for the same server;
5597 (b) the server returned 5xx to all RCPT commands in the first message so
5598 that the message was not completed; (c) the server dropped the connection
5599 or gave a negative response to the RSET that Exim sends to abort the
5600 transaction. The observed case was a dropped connection after DATA that had
5601 been sent in pipelining mode. That is, the server had advertised PIPELINING
5602 but was not implementing it correctly. The effect of the bug was incorrect
5603 behaviour, such as trying another host, and this could lead to a crash.
5609 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5610 and daemon.c were passed as pointers to ints; they should have been
5611 pointers to socklen_t variables (which are typically unsigned ints).
5613 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5616 3. Fixed a really odd bug that affected only the testing scheme; patching a
5617 certain fixed string in the binary changed the value of another string that
5618 happened to be identical to the end of the original first string.
5620 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5621 name", it returns that address as the IP address. On some operating
5622 systems (e.g. Solaris), it also passes back the IP address string as the
5623 "host name". However, on others (e.g. Linux), it passes back an empty
5624 string. Exim wasn't checking for this, and was changing the host name to an
5625 empty string, assuming it had been canonicized.
5627 5. Although rare, it is permitted to have more than one PTR record for a given
5628 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5629 all the names associated with an address, because they do in Solaris.
5630 However, it seems that they do not in Linux for data that comes from the
5631 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5632 I found this out when I moved to a new Linux workstation and tried to run
5633 the Exim test suite.
5635 To get round this problem I have changed the code so that it now does its
5636 own call to the DNS to look up PTR records when searching for a host name.
5637 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5638 addresses that are only in /etc/hosts are still found.
5640 This behaviour is, however, controlled by an option called host_lookup_
5641 order, which defaults to "bydns:byaddr". If people want to use the other
5642 order, or indeed, just use one or the other means of lookup, they can
5643 specify it in this variable.
5645 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5646 some operating systems, this comes back from gethostbyaddr() as an empty
5647 string, and this is what Exim used to test for. However, it seems that in
5648 other systems, "." is yielded. Exim now tests for this case too.
5650 7. The values of check_spool_space and check_log_space are now held internally
5651 as a number of kilobytes instead of an absolute number of bytes. If a
5652 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5653 kilobyte. This means that much larger values can be stored.
5655 8. Exim monitor: an attempt to get the action menu when not actually pointing
5656 at a message produces an empty menu entitled "No message selected". This
5657 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5658 no entries in it ("Shell widget menu has zero width and/or height"). So I
5659 have added a single, blank menu entry in this case.
5661 9. Added ${quote_local_part.
5663 10. MIME decoding is now applied to the contents of Subject: header lines when
5666 11. Now that a reference to $sender_host_address automatically causes a reverse
5667 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5668 host lookup before query-style lookups in lists that might use this
5669 variable. This has therefore been abolished, and the "net-" prefix is no
5670 longer necessary for query-style lookups.
5672 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5673 have been a typo for LFLAGS, so it has been changed.
5675 13. The install script calls Exim with "-C /dev/null" in order to find the
5676 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5677 to be output. Howeve, since Exim outputs its version number before the
5678 error, it didn't break the script. It just looked ugly. I fixed this by
5679 always allowing "-C /dev/null" if the caller is root.
5681 14. Ignore overlarge ACL variable number when reading spool file - insurance
5682 against a later release with more variables having written the file.
5684 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5685 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5686 this, as well as the form without the "IPv6" on the front.
5688 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5689 OS/Makefile-Darwin file.
5691 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5692 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5694 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5695 variable. However, some OS use ints for inodes. I've added cast to long int
5696 to get rid of the compiler warning.
5698 19. I had forgotten to lock out "/../" in configuration file names when
5699 ALT_CONFIG_PREFIX was set.
5701 20. Routers used for verification do not need to specify transports. However,
5702 if such a router generated a host list, and callout was configured, Exim
5703 crashed, because it could not find a port number from the (non-existent)
5704 transport. It now assumes port 25 in this circumstance.
5706 21. Added the -t option to exigrep.
5708 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5709 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5710 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5711 them is set, LOOKUP_LSEARCH is forced.
5713 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5715 24. Added sender and host information to the "rejected by local_scan()" log
5716 line; previously there was no indication of these.
5718 25. Added .include_if_exists.
5720 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5721 Linux. It turns out that not all file systems support this. Apparently some
5722 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5723 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5724 not supported on the file descriptor, is now ignored when Exim is trying to
5725 sync a directory. This applies only to Linux.
5727 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5729 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5730 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5733 29. Added support for maildirsize files from supplied patch (modified a bit).
5735 30. The use of :fail: followed by an empty string could lead Exim to respond to
5736 sender verification failures with (e.g.):
5738 550 Verification failed for <xxx>
5739 550 Sender verify failed
5741 where the first response line was missing the '-' that indicates it is not
5742 the final line of the response.
5744 31. The loop for finding the name of the user that called Exim had a hardwired
5745 limit of 10; it now uses the value of finduser_retries, which is used for
5746 all other user lookups.
5748 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5751 33. Exim was neglecting to zero errno before one call of strtol() when
5752 expanding a string and expecting an integer value. On some systems this
5753 resulted in spurious "integer overflow" errors. Also, it was casting the
5754 result into an int without checking.
5756 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5757 did not work. The code looks as if it has *never* worked, though it appears
5758 to have been documented since at least releast 1.62. I have made it work.
5760 35. The "timeout_DNS" error in retry rules, also documented since at least
5761 1.62, also never worked. As it isn't clear exactly what this means, and
5762 clearly it isn't a major issue, I have abolished the feature by treating it
5763 as "timeout", and writing a warning to the main and panic logs.
5765 36. The display of retry rules for -brt wasn't always showing the error code
5768 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5769 timeout_connect_A, timeout_connect_MX.
5771 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5772 to the empty sender.
5774 39. The daemon was not analysing the content of -oX till after it had closed
5775 stderr and disconnected from the controlling terminal. This meant that any
5776 syntax errors were only noted on the panic log, and the return code from
5777 the command was 0. By re-arranging the code a little, I've made the
5778 decoding happen first, so such errors now appear on stderr, and the return
5779 code is 1. However, the actual setting up of the sockets still happens in
5780 the disconnected process, so errors there are still only recorded on the
5783 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5784 connections (as happens on some IP stacks) was logged at start up time as
5785 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5786 it from "IPv6 and IPv4", which means that two separate sockets are being
5789 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5790 says whether AF_INET or AF_INET6 was passed as an argument.
5792 42. Exiwhat output was messed up when time zones were included in log
5795 43. Exiwhat now gives more information about the daemon's listening ports,
5796 and whether -tls-on-connect was used.
5798 44. The "port" option of the smtp transport is now expanded.
5800 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5801 silently ignored. Now an error message is written to the main and panic
5804 46. There's a new ACL modifier called "logwrite" which writes to a log file
5805 as soon as it is encountered.
5807 47. Added $local_user_uid and $local_user_gid at routing time.
5809 48. Exim crashed when trying to verify a sender address that was being
5812 49. Exim was recognizing only a space character after ".include". It now also
5813 recognizes a tab character.
5815 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5816 extracting the relevant information from the specification. The man page no
5817 longer contains scrambled data for the -d option, and I've added a section
5818 at the front about calling Exim under different names.
5820 51. Added "extra_headers" argument to the "mail" command in filter files.
5822 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5825 53. Installed eximstats 1.29.
5827 54. Added transport_filter_timeout as a generic transport option.
5829 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5830 Cc: header lines. This was required by RFC 822, but it not required by RFC
5833 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5834 incoming messages that did not have them. Now it does so only if the
5835 message originates locally, that is, if there is no associated remote host
5836 address. When Resent- header lines are present, this applies to the Resent-
5837 lines rather than the non-Resent- lines.
5839 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5840 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5842 58. Messages for configuration errors now include the name of the main
5843 configuration file - useful now that there may be more than one file in a
5844 list (.included file names were always shown).
5846 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5847 for those rare installations that do not start the daemon as root or run it
5848 setuid root. I've cut out the call to initgroups() if the daemon is not
5851 60. The Exim user and group can now be bound into the binary as text strings
5852 that are looked up at the start of Exim's processing.
5854 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5856 62. Added $mailstore_basename variable.
5858 63. Installed patch to sieve.c from Michael Haardt.
5860 64. When Exim failed to open the panic log after failing to open the main log,
5861 the original message it was trying to log was written to stderr and debug
5862 output, but if they were not available (the usual case in production), it
5863 was lost. Now it is written to syslog before the two lines that record the
5864 failures to open the logs.
5866 65. Users' Exim filters run in subprocesses under the user's uid. It is
5867 possible for a "deliver" command or an alias in a "personal" command to
5868 provoke an address rewrite. If logging of address rewriting is configured,
5869 this fails because the process is not running as root or exim. There may be
5870 a better way of dealing with this, but for the moment (because 4.30 needs
5871 to be released), I have disabled address rewrite logging when running a
5872 filter in a non-root, non-exim process.
5878 1. The buildconfig auxiliary program wasn't quoting the value set for
5879 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5880 not defined. This bug was masked in 4.22 by the effect that was fixed in
5883 2. Some messages that were rejected after a message id was allocated were
5884 shown as "incomplete" by exigrep. It no longer does this for messages that
5885 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5887 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5888 have allow_domain_literals set, the ID did not get logged in the <= line.
5889 Domain literals are now always recognized in Message-ID: header lines.
5891 4. The first argument for a ${extract expansion item is the key name or field
5892 number. Leading and trailing spaces in this item were not being ignored,
5893 causing some misleading effects.
5895 5. When deliver_drop_privilege was set, single queue runner processes started
5896 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5897 same command in the process it spins off) were not dropping privilege.
5899 6. When the daemon running as "exim" started a queue runner, it always
5900 re-executed Exim in the spun-off process. This is a waste of effort when
5901 deliver_drop_privilege is set. The new process now just calls the
5902 queue-runner function directly.
5908 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5911 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5912 ignored. Though the use of -oP was forcing the writing of a pid file, it
5913 was always written to the default place.
5915 3. If the message "no IP address found for host xxxx" is generated during
5916 incoming verification, it is now followed by identification of the incoming
5917 connection (so you can more easily find what provoked it).
5919 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5921 5. Added some features to "harden" Exim a bit more against certain attacks:
5923 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5924 be put in Local/Makefile. This is like the never_users runtime option,
5925 but it cannot be overridden. The default setting is "root".
5927 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5928 prefix string with which any file named in a -C command line option
5931 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5932 is retained for -C and -D only if the caller of Exim is root. Without
5933 it, the exim user may also use -C and -D and retain privilege.
5935 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5936 command line option is disabled.
5938 6. Macro names set by the -D option must start with an upper case letter, just
5939 like macro names defined in the configuration file.
5941 7. Added "dereference=" facility to LDAP.
5943 8. Two instances of the typo "uknown" in the source files are fixed.
5945 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5946 the Configure-Makefile script screwed up while processing it.
5948 10. Incorporated PCRE 4.4.
5950 11. The SMTP synchronization check was not operating right at the start of an
5951 SMTP session. For example, it could not catch a HELO sent before the client
5952 waited for the greeting. There is now a check for outstanding input at the
5953 point when the greeting is written. Because of the duplex, asynchronous
5954 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5955 way, but not yet received, when the check is performed.
5957 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5958 on TCP/IP sockets, because this apparently causes some broken clients to
5961 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5962 unchanged) from the Cygwin maintainer.
5964 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5965 was supported instead of testing for mbx. Effectively a typo.
5967 15. The spa authenticator server code was not checking that the input it
5968 received was valid base64.
5970 16. The debug output line for the "set" modifier in ACLs was not showing the
5971 name of the variable that was being set.
5973 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5975 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5976 lookup on its own. Something else was needed to trigger the lookup. For
5977 example, a match in host_lookup or the need for a host name in a host list.
5978 Now, if $sender_host_name is referenced and the host name has not yet been
5979 looked up, a lookup is performed. If the lookup fails, the variable remains
5980 empty, and $host_lookup_failed is set to "1".
5982 19. Added "eqi" as a case-independent comparison operator.
5984 20. The saslauthd authentication condition could segfault if neither service
5985 nor realm was specified.
5987 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5988 error message that was logged was misleading, and incoming SMTP
5989 connections were dropped. The message is now more accurate, and temporary
5990 errors are given to SMTP connections.
5992 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5993 (except RSET) with a 421 error, until QUIT is received. However, it was
5994 failing to send a response to QUIT.
5996 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5997 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5998 ACL is now run after the helo_try_verify_hosts code.
6000 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6001 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6002 case-independent, so other case variants are also recognized). Apparently
6003 some systems use these upper case variants.
6005 25. If more than two messages were waiting for the same host, and a transport
6006 filter was specified for the transport, Exim sent two messages over the
6007 same TCP/IP connection, and then failed with "socket operation on non-
6008 socket" when it tried to send the third.
6010 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6012 27. The extern definition of crypt16() in expand.c was not being excluded when
6013 the OS had its own crypt16() function.
6015 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6016 as a preferred synonym for return_size_limit, both as an option and as an
6019 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6021 30. Changed the default configuration ACL to relax the local part checking rule
6022 for addresses that are not in any local domains. For these addresses,
6023 slashes and pipe symbols are allowed within local parts, but the sequence
6024 /../ is explicitly forbidden.
6026 31. SPA server authentication was not clearing the challenge buffer before
6029 32. log_message in a "warn" ACL statement was writing to the reject log as
6030 well as to the main log, which contradicts the documentation and doesn't
6031 seem right (because no rejection is happening). So I have stopped it.
6033 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6034 However, I am unable to do any testing of this.
6036 34. Fixed an infelicity in the appendfile transport. When checking directories
6037 for a mailbox, to see if any needed to be created, it was accidentally
6038 using path names with one or more superfluous leading slashes; tracing
6039 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6041 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6042 contents are added to the log line that is written for every discarded
6043 recipient. (Previously a log_message setting was ignored.)
6045 36. The ${quote: operator now quotes the string if it is empty.
6047 37. The install script runs exim in order to find its version number. If for
6048 some reason other than non-existence or emptiness, which it checks, it
6049 could not run './exim', it was installing it with an empty version number,
6050 i.e. as "exim-". This error state is now caught, and the installation is
6053 38. An argument was missing from the function that creates an error message
6054 when Exim fails to connect to the socket for saslauthd authentication.
6055 This could cause Exim to crash, or give a corrupted message.
6057 39. Added isip, isip4, and isip6 to ${if conditions.
6059 40. The ACL variables $acl_xx are now saved with the message, and can be
6060 accessed later in routers, transports, and filters.
6062 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6063 strings in the file are not string-expanded.
6065 42. If a MAIL command specified a SIZE value that was too large to fit into an
6066 int variable, the check against message_size_limit failed. Such values are
6067 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6068 day this will have to be increased, but I don't think I want to be around
6069 when emails are that large.
6076 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6077 iconv() is not standard in FreeBSD.
6079 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6080 IPv6 enabled. The observed symptom was a segmentation fault on return from
6081 the function os_common_find_running_interfaces() in src/os.c.
6083 3. In the check_special_case() function in daemon.c I had used "errno" as an
6084 argument name, which causes warnings on some systems. This was basically a
6085 typo, since it was named "eno" in the comments!
6087 4. The code that waits for the clock to tick (at a resolution of some fraction
6088 of a second) so as to ensure message-id uniqueness was always waiting for
6089 at least one whole tick, when it could have waited for less. [This is
6090 almost certainly not relevant at current processor speeds, where it is
6091 unlikely to ever wait at all. But we try to future-proof.]
6093 5. The function that sleeps for a time interval that includes fractions of a
6094 second contained a race. It did not block SIGALRM between setting the
6095 timer, and suspending (a couple of lines later). If the interval was short
6096 and the sigsuspend() was delayed until after it had expired, the suspension
6097 never ended. On busy systems this could lead to processes getting stuck for
6100 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6101 process, before it forks any delivery processes. The open lookup caching
6102 mechanism meant that the open file or database connection was passed into
6103 the delivery process. The problem was that delivery processes always tidy
6104 up cached lookup data. This could cause a problem for the next delivery
6105 process started by the queue runner, because the external queue runner
6106 process does not know about the closure. So the next delivery process
6107 still has data in the lookup cache. In the case of a file lookup, there was
6108 no problem because closing a file descriptor in a subprocess doesn't affect
6109 the parent. However, if the lookup was caching a connection to a database,
6110 the connection was closed, and the second delivery process was likely to
6111 see errors such as "PGSQL: query failed: server closed the connection
6112 unexpectedly". The problem has been fixed by closing all cached lookups
6113 in a queue runner before running a delivery process.
6115 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6116 seem to have the "const" qualifier which it has on other OS. I've
6119 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6120 *of the same type* (client or server) with the same public name that an
6121 error should be diagnosed.
6123 9. When Exim looked up a host name for an IP address, but failed to find the
6124 original IP address when looking up the host name (a safety check), it
6125 output the message "<ip address> does not match any IP for NULL", which was
6126 confusing, to say the least. The bug was that the host name should have
6127 appeared instead of "NULL".
6129 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6130 user that is built into the binary, and the -C or -D options is used, root
6131 privilege is dropped before the configuration file is read. In addition,
6132 logging is switched to stderr instead of the normal log files. If the
6133 configuration then re-defines the Exim user, the unprivileged environment
6134 is probably not what is expected, so Exim logs a panic warning message (but
6137 However, if deliver_drop_privilege is set, the unprivileged state may well
6138 be exactly what is intended, so the warning has been cut out in that case,
6139 and Exim is allowed to try to write to its normal log files.
6145 1. smtp_return_error_details was not giving details for temporary sender
6146 or receiver verification errors.
6148 2. Diagnose a configuration error if two authenticators have the same public
6151 3. Exim used not to create the message log file for a message until the first
6152 delivery attempt. This could be confusing when incoming messages were held
6153 for policy or load reasons. The message log file is now created at the time
6154 the message is received, and an initial "Received" line is written to it.
6156 4. The automatically generated man page for command line options had a minor
6157 bug that caused no ill effects; however, a more serious problem was that
6158 the procedure for building the man page automatically didn't always
6159 operate. Consequently, release 4.20 contains an out-of-date version. This
6160 shouldn't happen again.
6162 5. When building Exim with embedded Perl support, the script that builds the
6163 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6164 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6166 6. The freeze_tell option was not being used for messages that were frozen on
6167 arrival, either by an ACL or by local_scan().
6169 7. Added the smtp_incomplete_transaction log selector.
6171 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6172 was accepting AUTH without a new EHLO.
6174 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6175 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6178 10. Logging of TCP/IP connections (when configured) now happens in the main
6179 daemon process instead of the child process, so that the TCP/IP connection
6180 count is more accurate (but it can never be perfect).
6182 11. The use of "drop" in a nested ACL was not being handled correctly in the
6183 outer ACL. Now, if condition failure induced by the nested "drop" causes
6184 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6185 or "require"), the connection is dropped.
6187 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6188 that yield "discard" can now be used with an "accept" or a "discard" verb,
6189 but an error is generated for any others (because I can't see a useful way
6190 to define what should happen).
6192 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6193 are now processed in the same way as lines in the Exim configuration file.
6194 In particular, continuation lines are supported.
6196 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6198 15. Added -ti meaning -t -i.
6200 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6201 domains, and warn by logging if others are found.
6203 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6204 returned from GIFCONF when Exim is trying to find the list of interfaces on
6205 a host. The code in os.c has been modified to copy each ifreq to an aligned
6206 structure in all cases.
6208 Also, in some cases, the returned ifreq's were being copied to a 'struct
6209 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6210 means the last couple of bytes of an IPv6 address could be chopped if the
6211 ifreq contained only a normal sockaddr (14 bytes storage).
6213 18. Named domain lists were not supported in the hosts_treat_as_local option.
6214 An entry such as +xxxx was not recognized, and was treated as a literal
6217 19. Ensure that header lines added by a DATA ACL are included in the reject log
6218 if the ACL subsequently rejects the message.
6220 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6221 MD5 (which is deprecated).
6223 21. When testing a filter file using -bf, Exim was writing a message when it
6224 took the sender from a "From " line in the message, but it was not doing so
6225 when it took $return_path from a Return-Path: header line. It now does.
6227 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6228 with a valid header line field name (a series of printing characters
6229 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6231 23. Changed "disc" in the source to "disk" to conform to the documentation and
6232 the book and for uniformity.
6234 24. Ignore Sendmail's -Ooption=value command line item.
6236 25. When execve() failed while trying to run a command in a pipe transport,
6237 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6238 could be confused with a return value of 69 from the command itself. This
6239 has been changed to 127, the value the shell returns if it is asked to run
6240 a non-existent command. The wording for the related log line suggests a
6241 non-existent command as the problem.
6243 26. If received_header_text expands to an empty string, do not add a Received:
6244 header line to the message. (Well, it adds a token one on the spool, but
6245 marks it "old" so that it doesn't get used or transmitted.)
6247 27. Installed eximstats 1.28 (addition of -nt option).
6249 28. There was no check for failure on the call to getsockname() in the daemon
6250 code. This can fail if there is a shortage of resources on the system, with
6251 ENOMEM, for example. A temporary error is now given on failure.
6253 29. Contrary to the C standard, it seems that in some environments, the
6254 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6255 program. Exim now does this explicitly; it affects the formatting of
6256 timestamps using strftime().
6258 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6259 complaints. I've now initialized all the variables, to avoid this.
6261 32. Header lines added by a system filter were not being "seen" during
6262 transport-time rewrites.
6264 33. The info_callback() function passed to OpenSSL is set up with type void
6265 (*)(SSL *, int, int), as described somewhere. However, when calling the
6266 function (actually a macro) that sets it up, the type void(*)() is
6267 expected. I've put in a cast to prevent warnings from picky compilers.
6269 34. If a DNS black list lookup found a CNAME record, but there were no A
6270 records associated with the domain it pointed at, Exim crashed.
6272 35. If a DNS black list lookup returned more than one A record, Exim ignored
6273 all but the first. It now scans all returned addresses if a particular IP
6274 value is being sought. In this situation, the contents of the
6275 $dnslist_value variable are a list of all the addresses, separated by a
6278 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6279 to accept a host name and all its aliases if the forward lookup for any of
6280 them yielded the IP address of the incoming connection. Now it accepts only
6281 those names whose forward lookup yields the correct IP address. Any other
6282 names are discarded. This closes a loophole whereby a rogue DNS
6283 administrator could create reverse DNS records to break through a
6284 wildcarded host restriction in an ACL.
6286 37. If a user filter or a system filter that ran in a subprocess used any of
6287 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6288 the wrong values were passed to the pipe command ($thisaddress had the
6289 value of $0, $0 had the value of $1, etc). This bug was introduced by
6290 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6292 38. Improved the line breaking for long SMTP error messages from ACLs.
6293 Previously, if there was no break point between 40 and 75 characters, Exim
6294 left the rest of the message alone. Two changes have been made: (a) I've
6295 reduced the minimum length to 35 characters; (b) if it can't find a break
6296 point between 35 and 75 characters, it looks ahead and uses the first one
6297 that it finds. This may give the occasional overlong line, but at least the
6298 remaining text gets split now.
6300 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6301 file descriptors might be low, and that setting 1000 would always raise it.
6302 It turns out that in some environments, the limit is already over 1000 and
6303 that lowering it causes trouble. So now Exim takes care not to decrease it.
6305 40. When delivering a message, the value of $return_path is set to $sender_
6306 address at the start of routing (routers may change the value). By an
6307 oversight, this default was not being set up when an address was tested by
6308 -bt or -bv, which affected the outcome if any router or filter referred to
6311 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6312 log only when "message" or "log_message" are set. However, if one of the
6313 conditions was an address verification, or a call to a nested ACL, the
6314 messages generated by the underlying test were being passed through. This
6315 no longer happens. The underlying message is available in $acl_verify_
6316 message for both "message" and "log_message" expansions, so it can be
6317 passed through if needed.
6319 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6320 new expansion $bh_ to give the encoded byte string without charset
6321 translation. Translation happens only if iconv() is available; HAVE_ICONV
6322 indicates this at build time. HEADERS_CHARSET gives the charset to
6323 translate to; headers_charset can change it in the configuration, and
6324 "headers charset" can change it in an individual filter file.
6326 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6327 that creates RFC 2047 encoded "words" labels them as that charset instead
6328 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6329 expansion operator; (ii) when Exim creates a From: line for a local
6330 message; (iii) when a header line is rewritten to include a "phrase" part.
6332 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6333 buggy, causing it to skip the first lines of messages whose message ID
6334 ended in 'D'. This would not have bitten before Exim release 4.14, because
6335 message IDs were unlikely to end in 'D' before then. The effect was to have
6336 incorrect size information for certain domains.
6338 45. #include "config.h" was missing at the start of the crypt16.c module. This
6339 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6342 46. If there was a timeout during a "random" callout check, Exim treated it as
6343 a failure of the random address, and carried on sending RSET and the real
6344 address. If the delay was just some slowness somewhere, the response to the
6345 original RCPT would be taken as a response to RSET and so on, causing
6346 mayhem of various kinds.
6348 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6349 when I implemented it. It didn't allow for the fact that some option values
6350 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6351 the right test for positive values.
6353 48. Domain names in DNS records are case-independent. Exim always looks them up
6354 in lower case. Some resolvers return domain names in exactly the case they
6355 appear in the zone file, that is, they may contain uppercase letters. Not
6356 all resolvers do this - some return always lower case. Exim was treating a
6357 change of case by a resolver as a change of domain, similar to a widening
6358 of a domain abbreviation. This triggered its re-routing code and so it was
6359 trying to route what was effectively the same domain again. This normally
6360 caused routing to fail (because the router wouldn't handle the domain
6361 twice). Now Exim checks for this case specially, and just changes the
6362 casing of the domain that it ultimately uses when it transmits the message
6365 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6368 50. If a filter generated a file delivery with a non-absolute name (possible if
6369 no home directory exists for the router), the forbid_file option was not
6372 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6373 the existing equality matching.
6375 52. Exim was using ints instead of ino_t variables in some places where it was
6376 dealing with inode numbers.
6378 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6379 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6380 called TMPDIR, and if it finds it is different, it changes its value.
6382 54. The smtp_printf() function is now made available to local_scan() so
6383 additional output lines can be written before returning. There is also an
6384 smtp_fflush() function to enable the detection of a dropped connection.
6385 The variables smtp_input and smtp_batched_input are exported to
6388 55. Changed the default runtime configuration: the message "Unknown user"
6389 has been removed from the ACL, and instead placed on the localuser router,
6390 using the cannot_route_message feature. This means that any verification
6391 failures that generate their own messages won't get overridden. Similarly,
6392 the "Unrouteable address" message that was in the ACL for unverifiable
6393 relay addresses has also been removed.
6395 56. Added hosts_avoid_esmtp to the smtp transport.
6397 57. The exicyclog script was not checking for the esoteric option
6398 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6399 will work only if exicyclog is run under the appropriate euid.
6401 58. Following a discussion on the list, the rules by which Exim recognises line
6402 endings on incoming messages have been changed. The -dropcr and drop_cr
6403 options are now no-ops, retained only for backwards compatibility. The
6404 following line terminators are recognized: LF CRLF CR. However, special
6405 processing applies to CR:
6407 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6408 nor a local message in the state where . is a terminator.
6410 (ii) If a bare CR is encountered in a header line, an extra space is added
6411 after the line terminator so as not to end the header. The reasoning
6412 behind this is that bare CRs in header lines are most likely either
6413 to be mistakes, or people trying to play silly games.
6415 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6416 was being incorrectly given as 18 bytes larger than it should have been.
6417 This is a VOB (very old bug).
6419 60. This may never have affected anything current, but just in case it has:
6420 When the local host is found other than at the start of a list of hosts,
6421 the local host, those with the same MX, and any that follow, are discarded.
6422 When the list in question was part of a longer list of hosts, the following
6423 hosts (not currently being processed) were also being discarded. This no
6424 longer happens. I'm not sure if this situation could ever has previously
6427 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6430 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6431 header. This is something that is recommended in a new Internet Draft, and
6432 is something that is documented as being done by Sendmail. There are two
6433 possible values. For messages generated by the autoreply transport, Exim
6436 Auto-Submitted: auto-replied
6438 whereas for all other generated messages (e.g. bounces) it adds
6440 Auto-Submitted: auto-generated
6442 63. The "personal" condition in filters now includes a test for the
6443 Auto-Submitted: header. If it contains the string "auto-" the message it
6444 not considered personal.
6446 64. Added rcpt_include_affixes as a generic transport option.
6448 65. Added queue_only_override (default true).
6450 66. Added the syslog_duplication option.
6452 67. If what should have been the first header line of a message consisted of
6453 a space followed by a colon, Exim was mis-interpreting it as a header line.
6454 It isn't of course - it is syntactically invalid and should therefore be
6455 treated as the start of the message body. The misbehaviour could have
6456 caused a number of strange effects, including loss of data in subsequent
6457 header lines, and spool format errors.
6459 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6460 client host had authenticated. This control can now be exercised by an ACL
6461 for more flexibility.
6463 69. By default, callouts do not happen when testing with -bh. There is now a
6464 variant, -bhc, which does actually run the callout code, including
6465 consulting and updating the callout cache.
6467 70. Added support for saslauthd authentication, courtesy of Alexander
6470 71. If statvfs() failed on the spool or log directories while checking their
6471 size for availability, Exim confusingly gave the error "space shortage".
6472 Furthermore, in debugging mode it crashed with a floating point exception.
6473 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6474 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6475 serious problem, Exim now writes to the main and panic logs when this
6476 happens, with details of the failure. It then refuses to accept the
6477 incoming message, giving the message "spool directory problem" or "log
6478 directory problem" with a 421 code for SMTP messages.
6480 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6481 0, 1, and 2 exist, because some OS complain for execs without them (see
6482 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6483 descriptors. However, the code omitted to check that the open succeeded,
6484 causing mysterious errors if for some reason the permissions on /dev/null
6485 got screwed. Now Exim writes a message to the main and panic logs, and
6486 bombs out if it can't open /dev/null.
6488 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6489 interact so that it is all more flexible. It is supposed to remain
6490 backwards compatible. Also added extra_local_interfaces.
6492 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6493 to bomb out with an assertion failure - to the client this appears as a
6494 connection drop. This problem occurs in the part of the code that was taken
6495 from the Samba project. Fortunately, the assertion is in a very simple
6496 function, so I have fixed this by reproducing the function inline in the
6497 one place where it is called, and arranging for authentication to fail
6498 instead of killing the process with assert().
6500 75. The SPA client code was not working when the server requested OEM rather
6501 than Unicode encoding.
6503 76. Added code to make require_files with a specific uid setting more usable in
6504 the case where statting the file as root fails - usually a non-root-mounted
6505 NFS file system. When this happens and the failure is EACCES, Exim now
6506 forks a subprocess and does the per-uid checking as the relevant uid.
6508 77. Added process_log_path.
6510 78. If log_file_path was not explicitly set, a setting of check_log_space or
6511 check_log_inodes was ignored.
6513 79. If a space check for the spool or log partitions fails, the incident is now
6514 logged. Of course, in the latter case the data may get lost...
6516 80. Added the %p formatting code to string_format() so that it can be used to
6517 print addresses in debug_print(). Adjusted all the address printing in the
6518 debugging in store.c to use %p rather than %d.
6520 81. There was a concern that a line of code in smtp_in.c could overflow a
6521 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6522 initially expressed, the concern was not well-founded, because trailing
6523 spaces are removed early. However, if the trailing spaces were followed by
6524 a NULL, they did not get removed, so the overflow was possible. Two fixes
6527 (a) I re-wrote the offending code in a cleaner fashion.
6528 (b) If an incoming SMTP command contains a NULL character, it is rejected
6531 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6532 runs initgroups(), so that if the Exim user is in any additional groups,
6533 they will be used during message reception.
6539 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob