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.
35 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
36 and 3.4.4 - once the server is enabled to respond to an OCSP request
37 it does even when not requested, resulting in a stapling non-aware
38 client dropping the TLS connection.
40 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
41 support variable-length bit vectors. No functional change.
43 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
44 The I= interface field on outgoing lines is now after the H= remote
45 host field, same as incoming lines. There is a separate
46 outgoing_interface log selector which allows you to disable the
49 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
50 If not running log_selector +smtp_connection the mainlog would be held
51 open indefinitely after a "too many connections" event, including to a
52 deleted file after a log rotate. Leave the per net connection logging
53 leaving it open for efficiency as that will be quickly detected by the
54 check on the next write.
56 HS/01 Bug 1671: Fix post transport crash.
57 Processing the wait-<transport> messages could crash the delivery
58 process if the message IDs didn't exist for some reason. When
59 using 'split_spool_directory=yes' the construction of the spool
60 file name failed already, exposing the same netto behaviour.
62 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
63 mime_regex ACL conditions.
65 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
66 to DSN fail messages (bounces): remote IP, remote greeting, remote response
67 to HELO, local diagnostic string.
69 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
70 log line to debug. Even when configured with a tls authenticator many
71 client connections are expected to not authenticate in this way, so
72 an authenticate fail is not an error.
74 HS/02 Add the Exim version string to the process info. This way exiwhat
75 gives some more detail about the running daemon.
77 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
78 matter for fast-change records such as DNSBLs.
80 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
81 retry record, even if constant. There may be multiple transports with
82 different interface settings and the retry behaviour needs to be kept
85 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
87 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
89 JH/10 Bug 840: fix log_defer_output option of pipe transport
91 JH/11 Bug 830: use same host for all RCPTS of a message, even under
92 hosts_randomize. This matters a lot when combined with mua_wrapper.
94 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
95 ${quote_pgsql:<string>} operator.
97 JH/13 Bug 1708: avoid misaligned access in cached lookup.
99 JH/14 Change header file name for freeradius-client. Relevant if compiling
100 with Radius support; from the Gentoo tree and checked under Fedora.
102 JH/15 Bug 1712: Introduce $prdr_requested flag variable
104 JH/16 Bug 1714: Permit an empty string as expansion result for transport
105 option transport_filter, meaning no filtering.
107 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
109 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
110 defaults to "*" (all hosts). The variable is now available when not built
111 with TLS, default unset, mainly to enable keeping the testuite sane.
112 If a server certificate is not supplied (via tls_certificate) an error is
113 logged, and clients will find TLS connections fail on startup. Presumably
114 they will retry in-clear.
115 Packagers of Exim are strongly encouraged to create a server certificate
116 at installation time.
118 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
119 with the $config_file variable.
121 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
122 in transport context, after the attempt, and per-recipient. The latter type
123 is per host attempted. The event data is the error message, and the errno
124 information encodes the lookup type (A vs. MX) used for the (first) host,
125 and the trailing two digits of the smtp 4xx reponse.
127 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
128 to write to mainlog (or rejectlog, paniclog) in the window between file
129 creation and permissions/ownership being changed. Particularly affects
130 installations where exicyclog is run as root, rather than exim user;
131 result is that the running daemon panics and dies.
133 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
135 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
136 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
137 "pri" and "weight". Note that the previous implicit priority given by the
138 list order is no longer honoured.
140 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
143 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
144 by defining SUPPORT_SOCKS.
146 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
147 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
148 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
149 variables are renamed to proxy_{local,external}_{address,port}.
151 JH/27 Move Internationalisation support from Experimental to mainline, enabled
152 for a build by defining SUPPORT_I18N
154 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
155 of the query string, and make ${quote_redis:} do that quoting.
157 JH/29 Move Events support from Experimental to mainline, enabled by default
158 and removable for a build by defining DISABLE_EVENT.
160 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
162 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
163 cached by the daemon.
165 JH/32 Move Redis support from Experimental to mainline, enabled for a build
166 by defining LOOKUP_REDIS. The libhiredis library is required.
168 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
169 keys are given for lookup.
171 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
172 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
173 only supported when built with TLS support. The PolarSSL SHA routines
174 are still used when the TLS library is too old for convenient support.
176 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
177 openssl_options), for security. OpenSSL forces this from version 1.1.0
178 server-side so match that on older versions.
180 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
181 allocation for $value could be released as the expansion processing
182 concluded, but leaving the global pointer active for it.
184 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
185 and to use the domains and local_parts ACL conditions.
187 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
188 incorrectly not doubled on cutthrough transmission, hence seen as a
189 body-termination at the receiving system - resulting in truncated mails.
190 Commonly the sender saw a TCP-level error, and retransmitted the message
191 via the normal store-and-forward channel. This could result in duplicates
192 received - but deduplicating mailstores were liable to retain only the
193 initial truncated version.
195 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
197 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
199 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
200 we're in there, support oversigning also; bug 1309.
202 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
204 HS/04 Add support for keep_environment and add_environment options.
206 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
207 either intentional arithmetic overflow during PRNG, or testing config-
210 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
211 delivery resulted in actual delivery. Cancel cutthrough before DATA
214 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
215 rejects a recipient: pass the reject to the originator.
217 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
218 Many were false-positives and ignorable, but it's worth fixing the
221 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
222 for the new environment-manipulation done at startup. Move the routines
223 from being local to tls.c to being global via the os.c file.
225 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
226 an extract embedded as result-arg for a map, the first arg for extract
227 is unavailable so we cannot tell if this is a numbered or keyed
228 extraction. Accept either.
234 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
237 JH/02 The smtp transport option "multi_domain" is now expanded.
239 JH/03 The smtp transport now requests PRDR by default, if the server offers
242 JH/04 Certificate name checking on server certificates, when exim is a client,
243 is now done by default. The transport option tls_verify_cert_hostnames
244 can be used to disable this per-host. The build option
245 EXPERIMENTAL_CERTNAMES is withdrawn.
247 JH/05 The value of the tls_verify_certificates smtp transport and main options
248 default to the word "system" to access the system default CA bundle.
249 For GnuTLS, only version 3.0.20 or later.
251 JH/06 Verification of the server certificate for a TLS connection is now tried
252 (but not required) by default. The verification status is now logged by
253 default, for both outbound TLS and client-certificate supplying inbound
256 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
259 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
260 Status Notification (bounce) messages are now MIME format per RFC 3464.
261 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
262 under the control of the dsn_advertise_hosts option, and routers may
263 have a dsn_lasthop option.
265 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
266 default, modifiable by a malware= option. The list separator for
267 the options can now be changed in the usual way. Bug 68.
269 JH/10 The smtp_receive_timeout main option is now expanded before use.
271 JH/11 The incoming_interface log option now also enables logging of the
272 local interface on delivery outgoing connections.
274 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
275 if the interface and destination host and port all match.
277 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
280 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
281 Patch from Andrew Lewis.
283 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
284 now supports optional time-restrictions, weighting, and priority
285 modifiers per server. Patch originally by <rommer@active.by>.
287 JH/16 The spamd_address main option now supports a mixed list of local
288 and remote servers. Remote servers can be IPv6 addresses, and
289 specify a port-range.
291 JH/17 Bug 68: The spamd_address main option now supports an optional
292 timeout value per server.
294 JH/18 Bug 1581: Router and transport options headers_add/remove can
295 now have the list separator specified.
297 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
300 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
303 JH/21 Support for the A6 type of dns record is withdrawn.
305 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
306 rather than the verbs used.
308 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
309 from 255 to 1024 chars.
311 JH/24 Verification callouts now attempt to use TLS by default.
313 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
314 are generic router options now. The defaults didn't change.
316 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
317 Original patch from Alexander Shikoff, worked over by JH.
319 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
320 files if ZCAT_COMMAND is not executable.
322 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
324 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
326 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
327 Normally benign, it bites when the pair was led to by a CNAME;
328 modern usage is to not canoicalize the domain to a CNAME target
329 (and we were inconsistent anyway for A-only vs AAAA+A).
331 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
333 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
334 when evaluating $sender_host_dnssec.
336 JH/31 Check the HELO verification lookup for DNSSEC, adding new
337 $sender_helo_dnssec variable.
339 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
341 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
343 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
345 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
346 documented as working, but never had. Support all but $spam_report.
348 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
349 added for tls authenticator.
351 HS/03 Add perl_taintmode main config option
356 TL/01 When running the test suite, the README says that variables such as
357 no_msglog_check are global and can be placed anywhere in a specific
358 test's script, however it was observed that placement needed to be near
359 the beginning for it to behave that way. Changed the runtest perl
360 script to read through the entire script once to detect and set these
361 variables, reset to the beginning of the script, and then run through
362 the script parsing/test process like normal.
364 TL/02 The BSD's have an arc4random API. One of the functions to induce
365 adding randomness was arc4random_stir(), but it has been removed in
366 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
367 function when detected.
369 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
370 cause callback expansion.
372 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
373 syntax errors in an expansion can be treated as a string instead of
374 logging or causing an error, due to the internal use of bool_lax
375 instead of bool when processing it.
377 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
378 server certificates when making smtp deliveries.
380 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
382 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
384 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
386 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
387 Merged patch from Sebastian Wiedenroth.
389 JH/05 Fix results-pipe from transport process. Several recipients, combined
390 with certificate use, exposed issues where response data items split
391 over buffer boundaries were not parsed properly. This eventually
392 resulted in duplicates being sent. This issue only became common enough
393 to notice due to the introduction of conection certificate information,
394 the item size being so much larger. Found and fixed by Wolfgang Breyha.
396 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
397 size buffer was used, resulting in syntax errors when an expansion
400 JH/07 Add support for directories of certificates when compiled with a GnuTLS
401 version 3.3.6 or later.
403 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
404 is EXPERIMENTAL_EVENT, the main-configuration and transport options
405 both become "event_action", the variables become $event_name, $event_data
406 and $event_defer_errno. There is a new variable $verify_mode, usable in
407 routers, transports and related events. The tls:cert event is now also
408 raised for inbound connections, if the main configuration event_action
411 TL/06 In test suite, disable OCSP for old versions of openssl which contained
412 early OCSP support, but no stapling (appears to be less than 1.0.0).
414 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
415 server certificate names available under the smtp transport option
416 "tls_verify_cert_hostname" now do not permit multi-component wildcard
419 JH/10 Time-related extraction expansions from certificates now use the main
420 option "timezone" setting for output formatting, and are consistent
421 between OpenSSL and GnuTLS compilations. Bug 1541.
423 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
424 encoded parameter in the incoming message. Bug 1558.
426 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
427 include certificate info, eximon was claiming there were spoolfile
430 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
432 JH/14 Log delivery-related information more consistently, using the sequence
433 "H=<name> [<ip>]" wherever possible.
435 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
436 are problematic for Debian distribution, omit them from the release
439 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
441 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
444 JH/17 Fix a null-indirection in certextract expansions when a nondefault
445 output list separator was used.
450 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
451 checkers that were complaining about end of non-void function with no
454 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
455 This was a regression intruduced in 4.83 by another bugfix.
457 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
459 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
460 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
466 TF/01 Correctly close the server side of TLS when forking for delivery.
468 When a message was received over SMTP with TLS, Exim failed to clear up
469 the incoming connection properly after forking off the child process to
470 deliver the message. In some situations the subsequent outgoing
471 delivery connection happened to have the same fd number as the incoming
472 connection previously had. Exim would try to use TLS and fail, logging
473 a "Bad file descriptor" error.
475 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
476 utilities have not been installed.
478 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
479 temporary space as the ACL may create new global variables.
481 TL/01 LDAP support uses per connection or global context settings, depending
482 upon the detected version of the libraries at build time.
484 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
485 to extract and use the src ip:port in logging and expansions as if it
486 were a direct connection from the outside internet. PPv2 support was
487 updated based on HAProxy spec change in May 2014.
489 JH/02 Add ${listextract {number}{list}{success}{fail}}.
491 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
492 Properly escape header and check for NULL return.
494 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
497 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
499 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
500 characters in header names, implemented as a verify condition.
501 Contributed by Michael Fischer v. Mollard.
503 TL/05 Rename SPF condition results err_perm and err_temp to standardized
504 results permerror and temperror. Previous values are deprecated but
505 still accepted. In a future release, err_perm and err_temp will be
506 completely removed, which will be a backward incompatibility if the
507 ACL tests for either of these two old results. Patch contributed by
508 user bes-internal on the mailing list.
510 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
512 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
513 selectors, in both main and reject logs.
515 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
518 JH/07 Add malware type "sock" for talking to simple daemon.
520 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
522 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
523 routers/transports under cutthrough routing.
525 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
526 numbers. Touch up "bool" conditional to keep the same definition.
528 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
530 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
532 JH/12 Expand items in router/transport headers_add or headers_remove lists
533 individually rather than the list as a whole. Bug 1452.
535 Required for reasonable handling of multiple headers_ options when
536 they may be empty; requires that headers_remove items with embedded
537 colons must have them doubled (or the list-separator changed).
539 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
540 view the policy declared in the DMARC record. Currently, $dmarc_status
541 is a combined value of both the record presence and the result of the
544 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
546 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
547 dnslookup router and the smtp transport (applying to the forward
550 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
551 of ldap servers used for a specific lookup. Patch provided by Heiko
554 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
555 New variable $lookup_dnssec_authenticated for observability.
557 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
558 Patch submitted by Lars Timman.
560 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
562 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
563 Requires trusted mode and valid format message id, aborts otherwise.
564 Patch contributed by Heiko Schlichting.
566 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
567 certextract with support for various fields. Bug 1358.
569 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
570 is requested by default, modifiable by smtp transport option
573 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
574 operate on certificate variables to give certificate fingerprints
575 Also new ${sha256:cert_variable}.
577 JH/23 The PRDR feature is moved from being Experimental into the mainline.
579 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
580 Christian Aistleitner.
582 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
584 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
585 file. Patch from Wolfgang Breyha.
587 JH/25 Expand the coverage of the delivery $host and $host_address to
588 client authenticators run in verify callout. Bug 1476.
590 JH/26 Port service names are now accepted for tls_on_connect_ports, to
591 align with daemon_smtp_ports. Bug 72.
593 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
594 support and error reporting did not work properly.
596 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
597 and is readable. Patch from Andrew Colin Kissa.
599 TL/14 Enhance documentation of ${run expansion and how it parses the
600 commandline after expansion, particularly in the case when an
601 unquoted variable expansion results in an empty value.
603 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
605 PP/02 Fix internal collision of T_APL on systems which support RFC3123
606 by renaming away from it. Addresses GH issue 15, reported by
609 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
611 TL/15 SECURITY: prevent double expansion in math comparison functions
612 (can expand unsanitized data). Not remotely exploitable.
619 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
621 PP/02 Make -n do something, by making it not do something.
622 When combined with -bP, the name of an option is not output.
624 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
627 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
628 $sender_host_name and config options to manage this, and basic check
631 PP/05 DSCP support for outbound connections and control modifier for inbound.
633 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
634 (Only plugin which currently uses this is kerberos4, which nobody should
635 be using, but we should make it available and other future plugins might
636 conceivably use it, even though it would break NAT; stuff *should* be
637 using channel bindings instead).
639 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
640 name; added for Sendmail compatibility; requires admin caller.
641 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
642 just ignore it); requires trusted caller.
643 Also parse but ignore: -Ac -Am -X<logfile>
646 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
648 TL/02 Add +smtp_confirmation as a default logging option.
650 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
651 Patch by Magnus Holmgren from 2007-02-20.
653 TL/04 Bugzilla 1281 - Spec typo.
654 Bugzilla 1283 - Spec typo.
655 Bugzilla 1290 - Spec grammar fixes.
657 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
659 TL/06 Add Experimental DMARC support using libopendmarc libraries.
661 TL/07 Fix an out of order global option causing a segfault. Reported to dev
662 mailing list by by Dmitry Isaikin.
664 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
666 JH/02 Support "G" suffix to numbers in ${if comparisons.
668 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
670 NM/01 Bugzilla 1197 - Spec typo
671 Bugzilla 1196 - Spec examples corrections
673 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
675 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
676 gnutls_enable_pkcs11, but renamed to more accurately indicate its
679 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
680 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
682 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
683 "acl {{name}{arg}...}", and optional args on acl condition
686 JH/05 Permit multiple router/transport headers_add/remove lines.
688 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
690 JH/07 Avoid using a waiting database for a single-message-only transport.
691 Performance patch from Paul Fisher. Bugzilla 1262.
693 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
696 JH/09 Add $headers_added variable, with content from use of ACL modifier
697 add_header (but not yet added to the message). Bugzilla 199.
699 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
700 Pulled from Bugzilla 817 by Wolfgang Breyha.
702 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
704 (nb: this is the same fix as in Exim 4.80.1)
706 JH/11 Add A= logging on delivery lines, and a client_set_id option on
709 JH/12 Add optional authenticated_sender logging to A= and a log_selector
712 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
714 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
715 advertise SMTP AUTH mechanism to us, instead of a generic
716 protocol violation error. Also, make Exim more robust to bad
717 data from the Dovecot auth socket.
719 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
721 When a queue runner is handling a message, Exim first routes the
722 recipient addresses, during which it prunes them based on the retry
723 hints database. After that it attempts to deliver the message to
724 any remaining recipients. It then updates the hints database using
727 So if a recipient address works intermittently, it can get repeatedly
728 deferred at routing time. The retry hints record remains fresh so the
729 address never reaches the final cutoff time.
731 This is a fairly common occurrence when a user is bumping up against
732 their storage quota. Exim had some logic in its local delivery code
733 to deal with this. However it did not apply to per-recipient defers
734 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
736 This change adds a proper retry rule check during routing so that the
737 final cutoff time is checked against the message's age. We only do
738 this check if there is an address retry record and there is not a
739 domain retry record; this implies that previous attempts to handle
740 the address had the retry_use_local_parts option turned on. We use
741 this as an approximation for the destination being like a local
742 delivery, as in LMTP.
744 I suspect this new check makes the old local delivery cutoff check
745 redundant, but I have not verified this so I left the code in place.
747 TF/02 Correct gecos expansion when From: is a prefix of the username.
749 Test 0254 submits a message to Exim with the header
753 When I ran the test suite under the user fanf2, Exim expanded
754 the header to contain my full name, whereas it should have added
755 a Resent-Sender: header. It erroneously treats any prefix of the
756 username as equal to the username.
758 This change corrects that bug.
760 GF/01 DCC debug and logging tidyup
761 Error conditions log to paniclog rather than rejectlog.
762 Debug lines prefixed by "DCC: " to remove any ambiguity.
764 TF/03 Avoid unnecessary rebuilds of lookup-related code.
766 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
767 Bug spotted by Jeremy Harris; was flawed since initial commit.
768 Would have resulted in OCSP responses post-SNI triggering an Exim
769 NULL dereference and crash.
771 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
773 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
774 Bug detection, analysis and fix by Samuel Thibault.
775 Bugzilla 1331, Debian bug #698092.
777 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
779 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
780 Server implementation by Todd Lyons, client by JH.
781 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
782 config variable "prdr_enable" controls whether the server
783 advertises the facility. If the client requests PRDR a new
784 acl_data_smtp_prdr ACL is called once for each recipient, after
785 the body content is received and before the acl_smtp_data ACL.
786 The client is controlled by bolth of: a hosts_try_prdr option
787 on the smtp transport, and the server advertisement.
788 Default client logging of deliveries and rejections involving
789 PRDR are flagged with the string "PRDR".
791 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
792 fclose(). Diagnosis by Todd Lyons.
794 PP/17 Update configure.default to handle IPv6 localhost better.
795 Patch by Alain Williams (plus minor tweaks).
798 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
799 This is now consistent with GnuTLS, and is now documented: the
800 previous undocumented portable approach to treating the option as
801 unset was to force an expansion failure. That still works, and
802 an empty string is now equivalent.
804 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
805 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
806 not performing validation itself.
808 PP/20 Added force_command boolean option to pipe transport.
809 Patch from Nick Koston, of cPanel Inc.
811 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
814 TF/04 Added udpsend ACL modifer and hexquote expansion operator
816 PP/21 Fix eximon continuous updating with timestamped log-files.
817 Broken in a format-string cleanup in 4.80, missed when I repaired the
818 other false fix of the same issue.
819 Report and fix from Heiko Schlichting.
822 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
823 Report from Prashanth Katuri.
825 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
826 It's SecureTransport, so affects any MacOS clients which use the
827 system-integrated TLS libraries, including email clients.
829 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
830 using a MIME ACL for non-SMTP local injection.
831 Report and assistance in diagnosis by Warren Baker.
833 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
835 JH/16 Fix comparisons for 64b. Bugzilla 1385.
837 TL/09 Add expansion variable $authenticated_fail_id to keep track of
838 last id that failed so it may be referenced in subsequent ACL's.
840 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
843 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
844 ldap library initialization, allowing self-signed CA's to be
845 used. Also properly sets require_cert option later in code by
846 using NULL (global ldap config) instead of ldap handle (per
847 session). Bug diagnosis and testing by alxgomz.
849 TL/12 Enhanced documentation in the ratelimit.pl script provided in
850 the src/util/ subdirectory.
852 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
853 renamed to Transport Post Delivery Action by Jeremy Harris, as
856 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
857 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
858 redis_servers = needs to be configured which will be used by the redis
859 lookup. Patch from Warren Baker, of The Packet Hub.
861 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
863 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
864 hostname or reverse DNS when processing a host list. Used suggestions
865 from multiple comments on this bug.
867 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
869 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
870 Missed a few lines, added it to make the runtest require no keyboard
873 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
874 contains upper case chars. Make router use caseful_local_part.
876 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
877 support when GnuTLS has been built with p11-kit.
883 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
885 This, or similar/improved, will also be change PP/11 of 4.82.
891 PP/01 Handle short writes when writing local log-files.
892 In practice, only affects FreeBSD (8 onwards).
893 Bugzilla 1053, with thanks to Dmitry Isaikin.
895 NM/01 Bugzilla 949 - Documentation tweak
897 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
900 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
902 PP/02 Implemented gsasl authenticator.
904 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
906 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
907 `pkg-config foo` for cflags/libs.
909 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
910 with rest of GSASL and with heimdal_gssapi.
912 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
913 `pkg-config foo` for cflags/libs for the TLS implementation.
915 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
916 properties get this fed in as external SSF. A number of robustness
917 and debugging improvements to the cyrus_sasl authenticator.
919 PP/08 cyrus_sasl server now expands the server_realm option.
921 PP/09 Bugzilla 1214 - Log authentication information in reject log.
922 Patch by Jeremy Harris.
924 PP/10 Added dbmjz lookup type.
926 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
928 PP/12 MAIL args handles TAB as well as SP, for better interop with
929 non-compliant senders.
930 Analysis and variant patch by Todd Lyons.
932 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
933 Bug report from Lars Müller <lars@samba.org> (via SUSE),
934 Patch from Dirk Mueller <dmueller@suse.com>
936 PP/13 tls_peerdn now print-escaped for spool files.
937 Observed some $tls_peerdn in wild which contained \n, which resulted
938 in spool file corruption.
940 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
941 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
942 or write after TLS renegotiation, which otherwise led to messages
945 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
946 as a tracking header (ie: a signed header comes before the signature).
947 Patch from Wolfgang Breyha.
949 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
950 comma-sep list; embedded commas doubled.
952 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
954 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
956 Report and patch from Dmitry Banschikov.
958 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
959 Removed SSL_clear() after SSL_new() which led to protocol negotiation
960 failures. We appear to now support TLS1.1+ with Exim.
962 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
963 lets Exim select keys and certificates based upon TLS SNI from client.
964 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
965 before an outbound SMTP session. New log_selector, +tls_sni.
967 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
968 NULL dereference. Report and patch from Alun Jones.
970 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
971 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
972 Not seeing resolver debug output on NetBSD, but suspect this is a
973 resolver implementation change.
975 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
976 Left warnings. Added "eximon gdb" invocation mode.
978 PP/21 Defaulting "accept_8bitmime" to true, not false.
980 PP/22 Added -bw for inetd wait mode support.
982 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
983 locate the relevant includes and libraries. Made this the default.
985 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
986 Bugzilla 1246, report and most of solution from Tomasz Kusy.
988 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
989 This may cause build issues on older platforms.
991 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
992 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
993 gnutls_require_mac & gnutls_require_protocols (no longer supported).
994 Added SNI support via GnuTLS too.
995 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
997 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
999 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1000 Applied second patch from Janne, implementing suggestion to default
1001 multiple-strings-in-record handling to match SPF spec.
1003 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1005 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1006 read-only, out of scope).
1007 Patch from Wolfgang Breyha, report from Stuart Northfield.
1009 PP/29 Fix three issues highlighted by clang analyser static analysis.
1010 Only crash-plausible issue would require the Cambridge-specific
1011 iplookup router and a misconfiguration.
1012 Report from Marcin Mirosław.
1014 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1016 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1017 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1018 As part of this, removing so much warning spew let me fix some minor
1019 real issues in debug logging.
1021 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1022 assignment on my part. Fixed.
1024 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1025 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1026 Janne Snabb (who went above and beyond: thank you).
1028 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1029 string otherwise requires a connection and a bunch more work and it's
1030 relatively easy to get wrong. Should also expose TLS library linkage
1033 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1034 64-bit ${eval} (JH/03).
1036 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1037 GNU libc to support some of the 64-bit stuff, should not lead to
1038 conflicts. Defined before os.h is pulled in, so if a given platform
1039 needs to override this, it can.
1041 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1042 protection layer was required, which is not implemented.
1043 Bugzilla 1254, patch from Wolfgang Breyha.
1045 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1046 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1047 tls_dhparam take prime identifiers. Also unbreak combination of
1048 OpenSSL+DH_params+TLSSNI.
1050 PP/39 Disable SSLv2 by default in OpenSSL support.
1056 PP/01 Solaris build fix for Oracle's LDAP libraries.
1057 Bugzilla 1109, patch from Stephen Usher.
1059 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1061 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1064 TF/02 Fix a couple more cases where we did not log the error message
1065 when unlink() failed. See also change 4.74-TF/03.
1067 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1068 lock up or crash if it happened to be inside a call to libc when it
1069 got a SIGUSR1 from exiwhat.
1071 The SIGUSR1 handler appends the current process status to the process
1072 log which is later printed by exiwhat. It used to use the general
1073 purpose logging code to do this, but several functions it calls are
1074 not safe for signals.
1076 The new output code in the SIGUSR1 handler is specific to the process
1077 log, and simple enough that it's easy to inspect for signal safety.
1078 Removing some special cases also simplifies the general logging code.
1079 Removing the spurious timestamps from the process log simplifies
1082 TF/04 Improved ratelimit ACL condition.
1084 The /noupdate option has been deprecated in favour of /readonly which
1085 has clearer semantics. The /leaky, /strict, and /readonly update modes
1086 are mutually exclusive. The update mode is no longer included in the
1087 database key; it just determines when the database is updated. (This
1088 means that when you upgrde Exim will forget old rate measurements.)
1090 Exim now checks that the per_* options are used with an update mode that
1091 makes sense for the current ACL. For example, when Exim is processing a
1092 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1093 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1094 must specify per_mail/readonly. If you omit the update mode it defaults to
1095 /leaky where that makes sense (as before) or /readonly where required.
1097 The /noupdate option is now undocumented but still supported for
1098 backwards compatibility. It is equivalent to /readonly except that in
1099 ACLs where /readonly is required you may specify /leaky/noupdate or
1100 /strict/noupdate which are treated the same as /readonly.
1102 A useful new feature is the /count= option. This is a generalization
1103 of the per_byte option, so that you can measure the throughput of other
1104 aggregate values. For example, the per_byte option is now equivalent
1105 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1107 The per_rcpt option has been generalized using the /count= mechanism
1108 (though it's more complicated than the per_byte equivalence). When it is
1109 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1110 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1111 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1112 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1113 Note that using per_rcpt with a non-readonly update mode in more than
1114 one ACL will cause the recipients to be double-counted. (The per_mail
1115 and per_byte options don't have this problem.)
1117 The handling of very low rates has changed slightly. If the computed rate
1118 is less than the event's count (usually one) then this event is the first
1119 after a long gap. In this case the rate is set to the same as this event's
1120 count, so that the first message of a spam run is counted properly.
1122 The major new feature is a mechanism for counting the rate of unique
1123 events. The new per_addr option counts the number of different
1124 recipients that someone has sent messages to in the last time period. It
1125 behaves like per_rcpt if all the recipient addresses are different, but
1126 duplicate recipient addresses do not increase the measured rate. Like
1127 the /count= option this is a general mechanism, so the per_addr option
1128 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1129 example, measure the rate that a client uses different sender addresses
1130 with the options per_mail/unique=$sender_address. There are further
1131 details in the main documentation.
1133 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1135 TF/06 Removed a few PCRE remnants.
1137 TF/07 Automatically extract Exim's version number from tags in the git
1138 repository when doing development or release builds.
1140 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1141 Bugzilla 879. Patch from Paul Fisher.
1143 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1144 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1147 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1149 PP/05 Variable $av_failed, true if the AV scanner deferred.
1150 Bugzilla 1078. Patch from John Horne.
1152 PP/06 Stop make process more reliably on build failure.
1153 Bugzilla 1087. Patch from Heiko Schlittermann.
1155 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1156 Bugzilla 1089. Patch from Heiko Schlittermann.
1158 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1159 Bugzilla 1131. Patch from Holger Weiß.
1161 PP/09 Handle IPv6 addresses with SPF.
1162 Bugzilla 860. Patch from Wolfgang Breyha.
1164 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1166 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1169 PP/11 match_* no longer expand right-hand-side by default.
1170 New compile-time build option, EXPAND_LISTMATCH_RHS.
1171 New expansion conditions, "inlist", "inlisti".
1173 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1175 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1177 PP/14 fix log_write() format string regression from TF/03.
1178 Bugzilla 1152. Patch from Dmitry Isaikin.
1184 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1186 PP/02 Harmonised TLS library version reporting; only show if debugging.
1187 Layout now matches that introduced for other libraries in 4.74 PP/03.
1189 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1191 PP/04 New "dns_use_edns0" global option.
1193 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1196 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1197 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1199 TK/01 Updated PolarSSL code to 0.14.2.
1200 Bugzilla 1097. Patch from Andreas Metzler.
1202 PP/07 Catch divide-by-zero in ${eval:...}.
1203 Fixes bugzilla 1102.
1205 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1208 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1209 format-string attack -- SECURITY: remote arbitrary code execution.
1211 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1212 time unintentionally subject to list matching rules, letting the header
1213 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1214 arbitrary string expansion). This allowed for information disclosure.
1216 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1217 INT_MIN/-1 -- value coerced to INT_MAX.
1223 NM/01 Workround for PCRE version dependency in version reporting
1226 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1227 This fixes portability to compilers other than gcc, notably
1228 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1230 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1231 makefiles for portability to HP-UX and POSIX correctness.
1233 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1234 Also via indented variable definition in the Makefile.
1235 (Debugging by Oliver Heesakkers).
1237 PP/02 Restore caching of spamd results with expanded spamd_address.
1238 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1240 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1241 Improves build reliability. Fix from: Frank Elsner
1243 NM/02 Fix wide character breakage in the rfc2047 coding
1244 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1246 NM/03 Allow underscore in dnslist lookups
1247 Fixes bug 1026. Patch from Graeme Fowler
1249 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1250 Code patches from Adam Ciarcinski of NetBSD.
1252 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1255 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1256 is logged, to avoid truncation. Patch from John Horne.
1258 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1259 Patch from Jakob Hirsch.
1261 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1262 SQL string expansion failure details.
1263 Patch from Andrey Oktyabrski.
1265 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1266 Patch from Simon Arlott.
1268 PP/09 New lookups functionality failed to compile on old gcc which rejects
1269 extern declarations in function scope.
1270 Patch from Oliver Fleischmann
1272 PP/10 Use sig_atomic_t for flags set from signal handlers.
1273 Check getgroups() return and improve debugging.
1274 Fixed developed for diagnosis in bug 927 (which turned out to be
1277 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1278 Patch from Mark Zealey.
1280 PP/12 Bugzilla 1056: Improved spamd server selection.
1281 Patch from Mark Zealey.
1283 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1284 Based on patch from Heiko Schlittermann.
1286 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1287 Patch from Uwe Doering, sign-off by Michael Haardt.
1289 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1290 variable declaration deep within a block. Bug and patch from
1293 PP/15 lookups-Makefile IRIX compatibilty coercion.
1295 PP/16 Make DISABLE_DKIM build knob functional.
1297 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1298 Patch by Simon Arlott
1300 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1301 variable argument macros. Our copy now differs from upstream.
1307 TF/01 Failure to get a lock on a hints database can have serious
1308 consequences so log it to the panic log.
1310 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1311 controlled using the smtp_confirmation log selector.
1313 TF/03 Include the error message when we fail to unlink a spool file.
1315 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1316 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1317 for maintaining out-of-tree patches for some time.
1319 PP/01 Bugzilla 139: Documentation and portability issues.
1320 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1321 Handle per-OS dynamic-module compilation flags.
1323 PP/02 Let /dev/null have normal permissions.
1324 The 4.73 fixes were a little too stringent and complained about the
1325 permissions on /dev/null. Exempt it from some checks.
1326 Reported by Andreas M. Kirchwitz.
1328 PP/03 Report version information for many libraries, including
1329 Exim version information for dynamically loaded libraries. Created
1330 version.h, now support a version extension string for distributors
1331 who patch heavily. Dynamic module ABI change.
1333 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1334 privilege escalation vulnerability whereby the Exim run-time user
1335 can cause root to append content of the attacker's choosing to
1338 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1341 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1342 If dropping privileges for untrusted macros, we disabled normal logging
1343 on the basis that it would fail; for the Exim run-time user, this is not
1344 the case, and it resulted in successful deliveries going unlogged.
1345 Fixed. Reported by Andreas Metzler.
1351 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1352 only prepend for the Resent-* case. Fixes regression introduced in
1353 Exim 4.70 by NM/22 for Bugzilla 607.
1355 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1356 increasing numbers of administrators be bitten by this.
1358 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1360 PP/03 Bugzilla 994: added openssl_options main configuration option.
1362 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1364 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1366 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1368 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1369 without a peer certificate, leading to a segfault because of an
1370 assumption that peers always have certificates. Be a little more
1371 paranoid. Problem reported by Martin Tscholak.
1373 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1374 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1375 NB: ClamAV planning to remove STREAM in "middle of 2010".
1376 CL also introduces -bmalware, various -d+acl logging additions and
1377 more caution in buffer sizes.
1379 PP/09 Implemented reverse_ip expansion operator.
1381 PP/10 Bugzilla 937: provide a "debug" ACL control.
1383 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1385 PP/12 Bugzilla 973: Implement --version.
1387 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1389 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1391 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1393 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1394 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1395 ignore trailing whitespace.
1397 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1399 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1402 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1403 Notification from Dr Andrew Aitchison.
1405 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1406 ExtendedDetectionInfo response format.
1407 Notification from John Horne.
1409 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1412 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1413 XSL and documented dependency on system catalogs, with examples of how
1416 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1419 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1420 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1421 configuration file which is writeable by the Exim user or group.
1423 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1424 of configuration files to cover files specified with the -C option if
1425 they are going to be used with root privileges, not just the default
1428 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1429 option (effectively making it always true).
1431 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1432 files to be used while preserving root privileges.
1434 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1435 that rogue child processes cannot use them.
1437 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1438 run-time user, instead of root.
1440 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1441 Exim run-time user without dropping privileges.
1443 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1444 result string, instead of calling string_vformat() twice with the same
1447 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1448 for other users. Others should always drop root privileges if they use
1449 -C on the command line, even for a whitelisted configure file.
1451 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1453 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1459 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1460 $data_path, and $header_path variables; fixed documentation bugs and
1463 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1464 exipick to access non-standard spools, including the "frozen" queue
1467 NM/01 Bugzilla 965: Support mysql stored procedures.
1468 Patch from Alain Williams
1470 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1472 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1473 Patch from Andreas Metzler
1475 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1476 Patch from Kirill Miazine
1478 NM/05 Bugzilla 671: Added umask to procmail example.
1480 JJ/03 installed exipick 20100323.0, fixing doc bug
1482 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1483 directory. Notification and patch from Dan Rosenberg.
1485 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1487 TK/02 Improve log output when DKIM signing operation fails.
1489 MH/01 Treat the transport option dkim_domain as a colon separated
1490 list, not as a single string, and sign the message with each element,
1491 omitting multiple occurences of the same signer.
1493 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1494 Bugzilla 985, 986. Patch by Simon Arlott
1496 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1497 Patch by Simon Arlott
1499 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1500 MBX locking. Notification from Dan Rosenberg.
1506 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1508 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1510 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1512 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1514 NM/04 Bugzilla 915: Flag broken perl installation during build.
1520 TK/01 Added patch by Johannes Berg that expands the main option
1521 "spamd_address" if it starts with a dollar sign.
1523 TK/02 Write list of recipients to X-Envelope-Sender header when building
1524 the mbox-format spool file for content scanning (suggested by Jakob
1527 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1528 (http://www.dcc-servers.net/) support via dccifd. Activated by
1529 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1531 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1532 by Mark Daniel Reidel <mr@df.eu>.
1534 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1535 When building exim an external PCRE library is now needed -
1536 PCRE is a system library on the majority of modern systems.
1537 See entry on PCRE_LIBS in EDITME file.
1539 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1540 conversation. Added nologin parameter to request.
1541 Patch contributed by Kirill Miazine.
1543 TF/01 Do not log submission mode rewrites if they do not change the address.
1545 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1547 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1548 log files in place. Contributed by Roberto Lima.
1550 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1552 TF/03 Bugzilla 615: When checking the local_parts router precondition
1553 after a local_part_suffix or local_part_prefix option, Exim now
1554 does not use the address's named list lookup cache, since this
1555 contains cached lookups for the whole local part.
1557 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1558 Robert Millan. Documentation is in experimental-spec.txt.
1560 TF/04 Bugzilla 668: Fix parallel build (make -j).
1562 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1564 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1565 Patch provided by Jan Srzednicki.
1567 TF/05 Leading white space used to be stripped from $spam_report which
1568 wrecked the formatting. Now it is preserved.
1570 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1571 that they are available at delivery time.
1573 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1575 TF/08 TLS error reporting now respects the incoming_interface and
1576 incoming_port log selectors.
1578 TF/09 Produce a more useful error message if an SMTP transport's hosts
1579 setting expands to an empty string.
1581 NM/06 Bugzilla 744: EXPN did not work under TLS.
1582 Patch provided by Phil Pennock.
1584 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1585 Patch provided by Richard Godbee.
1587 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1588 acl_smtp_notquit, added index entry.
1590 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1591 Patch provided by Eugene Bujak.
1593 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1594 accept(). Patch provided by Maxim Dounin.
1596 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1597 Patch provided by Phil Pennock.
1599 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1601 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1602 Patch provided by Brad "anomie" Jorsch.
1604 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1605 Patch provided by Dean Brooks.
1607 TK/05 Add native DKIM support (does not depend on external libraries).
1609 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1610 Patch provided by Graeme Fowler.
1612 NM/16 Bugzilla 851: Documentation example syntax fix.
1614 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1616 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1619 NM/19 Bugzilla 745: TLS version reporting.
1620 Patch provided by Phil Pennock.
1622 NM/20 Bugzilla 167: bool: condition support.
1623 Patch provided by Phil Pennock.
1625 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1626 clients. Patch provided by Phil Pennock.
1628 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1629 Patch provided by Brad "anomie" Jorsch.
1631 NM/23 Bugzilla 687: Fix misparses in eximstats.
1632 Patch provided by Heiko Schlittermann.
1634 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1635 Patch provided by Heiko Schlittermann.
1637 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1638 plus update to original patch.
1640 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1642 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1643 Patch provided by David Brownlee.
1645 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1647 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1649 NM/30 Bugzilla 888: TLS documentation bugfixes.
1651 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1653 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1654 Unlike the original bugzilla I have changed all shell scripts in src tree.
1656 NM/33 Bugzilla 898: Transport filter timeout fix.
1657 Patch by Todd Rinaldo.
1659 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1660 Patch by Serge Demonchaux.
1662 NM/35 Bugzilla 39: Base64 decode bug fixes.
1663 Patch by Jakob Hirsch.
1665 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1667 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1669 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1671 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1677 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1678 ALT-N's libdkim that I have put here:
1679 http://duncanthrax.net/exim-experimental/
1681 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1682 were called 'true' and it seems that C99 defines that as a reserved
1683 keyword to be used with 'bool' variable types. That means you could
1684 not include C99-style headers which use bools without triggering
1685 build errors in sieve.c.
1687 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1688 as mailq or other aliases. Changed the --help handling significantly
1689 to do whats expected. exim_usage() emits usage/help information.
1691 SC/01 Added the -bylocaldomain option to eximstats.
1693 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1695 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1697 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1703 PH/01 Another patch from the Sieve maintainer.
1705 PH/02 When an IPv6 address is converted to a string for single-key lookup
1706 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1707 dots are used instead of colons so that keys in lsearch files need not
1708 contain colons. This was done some time before quoting was made available
1709 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1710 (notated using the quote facility) so as to distinguish them from IPv4
1711 keys. This meant that lookups for IP addresses in host lists did not work
1712 for iplsearch lookups.
1714 This has been fixed by arranging for IPv6 addresses to be expressed with
1715 colons if the lookup type is iplsearch. This is not incompatible, because
1716 previously such lookups could never work.
1718 The situation is now rather anomolous, since one *can* have colons in
1719 ordinary lsearch keys. However, making the change in all cases is
1720 incompatible and would probably break a number of configurations.
1722 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1725 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1726 conversion specification without a maximum field width, thereby enabling
1727 a rogue spamd server to cause a buffer overflow. While nobody in their
1728 right mind would setup Exim to query an untrusted spamd server, an
1729 attacker that gains access to a server running spamd could potentially
1730 exploit this vulnerability to run arbitrary code as the Exim user.
1732 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1733 $primary_hostname instead of what libspf2 thinks the hosts name is.
1735 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1736 a directory entry by the name of the lookup key. Previously, if a
1737 symlink pointed to a non-existing file or a file in a directory that
1738 Exim lacked permissions to read, a lookup for a key matching that
1739 symlink would fail. Now it is enough that a matching directory entry
1740 exists, symlink or not. (Bugzilla 503.)
1742 PH/03 The body_linecount and body_zerocount variables are now exported in the
1745 PH/04 Added the $dnslist_matched variable.
1747 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1748 This means they are set thereafter only if the connection becomes
1751 PH/06 Added the client_condition to authenticators so that some can be skipped
1752 by clients under certain conditions.
1754 PH/07 The error message for a badly-placed control=no_multiline_responses left
1755 "_responses" off the end of the name.
1757 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1759 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1760 (without spaces) instead of just copying the configuration text.
1762 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1764 PH/11 Added $max_received_linelength.
1766 PH/12 Added +ignore_defer and +include_defer to host lists.
1768 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1769 way in which PCRE > 7.0 is built.
1771 PH/14 Implemented queue_only_load_latch.
1773 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1774 MAIL command. The effect was to mangle the value on 64-bit systems.
1776 PH/16 Another patch from the Sieve maintainer.
1778 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1780 PH/18 If a system quota error occurred while trying to create the file for
1781 a maildir delivery, the message "Mailbox is full" was not appended to the
1782 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1783 only to a quota excession during the actual writing of the file.
1785 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1786 characters?) which causes problems in log lines. The DN values are now
1787 passed through string_printing() before being added to log lines.
1789 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1790 and InterBase are left for another time.)
1792 PH/21 Added message_body_newlines option.
1794 PH/22 Guard against possible overflow in moan_check_errorcopy().
1796 PH/23 POSIX allows open() to be a macro; guard against that.
1798 PH/24 If the recipient of an error message contained an @ in the local part
1799 (suitably quoted, of course), incorrect values were put in $domain and
1800 $local_part during the evaluation of errors_copy.
1806 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1807 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1810 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1811 issue a MAIL command.
1813 PH/02 In an ACL statement such as
1815 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1817 if a client was not listed at all, or was listed with a value other than
1818 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1819 the condition was not true (as it should be), so access was not denied.
1820 The bug was that the ! inversion was incorrectly passed on to the second
1821 item. This has been fixed.
1823 PH/03 Added additional dnslists conditions == and =& which are different from
1824 = and & when the dns lookup returns more than one IP address.
1826 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1827 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1829 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1830 FSYNC, which compiles an option called disable_fsync that allows for
1831 bypassing fsync(). The documentation is heavily laced with warnings.
1833 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1835 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1836 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1837 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1838 including adding "make clean"; (3) Added -fPIC when compiling the test
1839 dynamically loaded module, to get rid of a warning.
1841 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1842 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1843 = 0s. The bug is otherwise harmless.
1845 PH/07 There was a bug in the dovecot authenticator such that the value of
1846 $auth1 could be overwritten, and so not correctly preserved, after a
1847 successful authentication. This usually meant that the value preserved by
1848 the server_setid option was incorrect.
1850 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1852 PH/09 Installed PCRE release 7.0.
1854 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1855 run for batched SMTP input. It is now run at the start of every message
1856 in the batch. While fixing this I discovered that the process information
1857 (output by running exiwhat) was not always getting set for -bs and -bS
1858 input. This is fixed, and it now also says "batched" for BSMTP.
1860 PH/11 Added control=no_pipelining.
1862 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1863 patch, slightly modified), and move the expansion of helo_data till after
1864 the connection is made in the smtp transport (so it can use these
1867 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1869 PH/14 Added log_selector = +pid.
1871 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1873 PH/16 Add ${if forany and ${if forall.
1875 PH/17 Added dsn_from option to vary the From: line in DSNs.
1877 PH/18 Flush SMTP output before performing a callout, unless control =
1878 no_callout_flush is set.
1880 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1881 was true (the default) a successful delivery failed to delete the retry
1882 item, thus causing premature timeout of the address. The bug is now
1885 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1887 PH/21 Long custom messages for fakedefer and fakereject are now split up
1888 into multiline reponses in the same way that messages for "deny" and
1889 other ACL rejections are.
1891 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1892 with slight modification.
1894 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1895 draft, changing the syntax and factoring some duplicate code".
1897 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1898 for deliveries of the second and subsequent messages over the same SMTP
1901 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1902 ${reduce, with only minor "tidies".
1904 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1906 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1907 expansion side effects.
1909 PH/27 When a message times out after an over-quota error from an Exim-imposed
1910 quota, the bounce message says "mailbox is full". This message was not
1911 being given when it was a system quota that was exceeded. It now should
1914 MH/03 Made $recipients available in local_scan(). local_scan() already has
1915 better access to the recipient list through recipients_list[], but
1916 $recipients can be useful in postmaster-provided expansion strings.
1918 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1919 in the case of a MAIL command with additional options following the
1920 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1921 were accidentally chopped off.
1923 PH/29 SMTP synchronization checks are implemented when a command is read -
1924 there is a check that no more input is waiting when there shouldn't be
1925 any. However, for some commands, a delay in an ACL can mean that it is
1926 some time before the response is written. In this time, more input might
1927 arrive, invalidly. So now there are extra checks after an ACL has run for
1928 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1929 pipelining has not been advertised.
1931 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1933 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1934 returned in the SMTP response when smtp_return_error_details was false.
1935 This has been fixed.
1937 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1938 instead of the C I/O that was originally supplied, because problems were
1939 reported on Solaris.
1941 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1942 Exim which did not show up earlier: it was assuming that a call to
1943 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1944 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1945 was a macro that became an assignment, so it seemed to work. This has
1946 changed to a proper function call with a void return, hence the compile
1947 error. Exim's code has been fixed.
1949 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1952 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1954 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1956 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1957 because it turns out to be too big or there is a timeout) there is no
1958 "Completed" line in the log. When some messages of this type were
1959 selected by exigrep, they were listed as "not completed". Others were
1960 picked up by some special patterns. I have improved the selection
1961 criteria to be more general.
1963 PH/38 The host_find_failed option in the manualroute router can now be set
1964 to "ignore", to completely ignore a host whose IP address cannot be
1965 found. If all hosts are ignored, the behaviour is controlled by the new
1966 host_all_ignored option.
1968 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1969 homing or because of multiple MX records with /mx) generated more than
1970 one IP address, and the following item turned out to be the local host,
1971 all the secondary addresses of the first item were incorrectly removed
1972 from the list, along with the local host and any following hosts (which
1973 is what is supposed to happen).
1975 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1976 whoever called Exim into the -H file. In the case of the daemon it was
1977 behaving confusingly. When first started, it used values for whoever
1978 started the daemon, but after a SIGHUP it used the Exim user (because it
1979 calls itself on a restart). I have changed the code so that it now always
1982 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1983 message are rejected with the same error (e.g. no authentication or bad
1984 sender address), and a DATA command is nevertheless sent (as can happen
1985 with PIPELINING or a stupid MUA), the error message that was given to the
1986 RCPT commands is included in the rejection of the DATA command. This is
1987 intended to be helpful for MUAs that show only the final error to their
1990 PH/42 Another patch from the Sieve maintainer.
1992 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1993 Eximstats - Fixed some broken HTML links and added missing column headers
1995 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1996 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1998 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2000 PH/43 Yet another patch from the Sieve maintainer.
2002 PH/44 I found a way to check for a TCP/IP connection going away before sending
2003 the response to the final '.' that terminates a message, but only in the
2004 case where the client has not sent further data following the '.'
2005 (unfortunately, this is allowed). However, in many cases there won't be
2006 any further data because there won't be any more messages to send. A call
2007 to select() can be used: if it shows that the input is "ready", there is
2008 either input waiting, or the socket has been closed. An attempt to read
2009 the next input character can distinguish the two cases. Previously, Exim
2010 would have sent an OK response which the client would never have see.
2011 This could lead to message repetition. This fix should cure that, at
2012 least in a lot of common cases.
2014 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2015 advertised in response to EHLO.
2021 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2022 fixed by 4.65/MH/01 (is this a record?) are fixed:
2024 (i) An empty string was always treated as zero by the numeric comparison
2025 operators. This behaviour has been restored.
2027 (ii) It is documented that the numeric comparison operators always treat
2028 their arguments as decimal numbers. This was broken in that numbers
2029 starting with 0 were being interpreted as octal.
2031 While fixing these problems I realized that there was another issue that
2032 hadn't been noticed. Values of message_size_limit (both the global option
2033 and the transport option) were treated as octal if they started with 0.
2034 The documentation was vague. These values are now always treated as
2035 decimal, and I will make that clear in the documentation.
2041 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2042 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2045 MH/01 Don't check that the operands of numeric comparison operators are
2046 integers when their expansion is in "skipping" mode (fixes bug
2047 introduced by 4.64-PH/07).
2049 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2050 child addresses, Exim now panics and dies. Previously, because the count
2051 is held in a short int, deliveries were likely to be lost. As such a
2052 large number of recipients for a single message is ridiculous
2053 (performance will be very, very poor), I have chosen to impose a limit
2054 rather than extend the field.
2060 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2061 leftover -K file (the existence of which was triggered by #402).
2062 While we were at it, introduced process PID as part of the -K
2063 filename. This should rule out race conditions when creating
2066 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2067 processing considerably. Previous code took too long for large mails,
2068 triggering a timeout which in turn triggers #401.
2070 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2071 in the DK code in transports.c. sendfile() is not really portable,
2072 hence the _LINUX specificness.
2074 TF/01 In the add_headers option to the mail command in an Exim filter,
2075 there was a bug that Exim would claim a syntax error in any
2076 header after the first one which had an odd number of characters
2079 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2080 callout verification, Exim cached a "reject" for the entire domain. This
2081 is correct for most verifications, but it is not correct for a recipient
2082 verification with use_sender or use_postmaster set, because in that case
2083 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2084 case of MAIL FROM:<> rejection from other early rejections (e.g.
2085 rejection of HELO). When verifying a recipient using a non-null MAIL
2086 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2087 Whatever the result of the callout, the value of the domain cache is
2088 left unchanged (for any other kind of callout, getting as far as trying
2089 RCPT means that the domain itself is ok).
2091 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2094 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2095 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2096 session, the daemon ignores ECONNECT errors and logs others; it now
2097 ignores EPIPE as well.
2099 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2100 (quoted-printable decoding).
2102 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2103 later the small subsequent patch to fix an introduced bug.
2105 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2107 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2109 PH/08 An error is now given if message_size_limit is specified negative.
2111 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2112 to be given (somewhat) arbitrary names.
2114 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2117 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2118 miscellaneous code fixes
2120 PH/10 Added the log_reject_target ACL modifier to specify where to log
2123 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2124 hostname. This is wrong, because it relates to the incoming message (and
2125 probably the interface on which it is arriving) and not to the outgoing
2126 callout (which could be using a different interface). This has been
2127 changed to use the value of the helo_data option from the smtp transport
2128 instead - this is what is used when a message is actually being sent. If
2129 there is no remote transport (possible with a router that sets up host
2130 addresses), $smtp_active_hostname is used.
2132 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2133 tweaks were necessary in order to get it to work (see also 21 below):
2134 (a) The code assumed that strncpy() returns a negative number on buffer
2135 overflow, which isn't the case. Replaced with Exim's string_format()
2137 (b) There were several signed/unsigned issues. I just did the minimum
2138 hacking in of casts. There is scope for a larger refactoring.
2139 (c) The code used strcasecmp() which is not a standard C function.
2140 Replaced with Exim's strcmpic() function.
2141 (d) The code set only $1; it now sets $auth1 as well.
2142 (e) A simple test gave the error "authentication client didn't specify
2143 service in request". It would seem that Dovecot has changed its
2144 interface. Fortunately there's a specification; I followed it and
2145 changed what the client sends and it appears to be working now.
2147 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2150 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2151 address A is aliased to B and C, where B exists and C does not. Without
2152 -v the output is "A verified" because verification stops after a
2153 successful redirection if more than one address is generated. However,
2154 with -v the child addresses are also verified. Exim was outputting "A
2155 failed to verify" and then showing the successful verification for C,
2156 with its parentage. It now outputs "B failed to verify", showing B's
2157 parentage before showing the successful verification of C.
2159 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2160 look up a TXT record in a specific list after matching in a combined
2163 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2164 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2165 they consult the DNS. I had assumed they would set it the way they
2166 wanted; and indeed my experiments on Linux seem to show that in some
2167 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2168 To be on the safe side, however, I have now made the interface to
2169 host_find_byname() similar to host_find_bydns(), with an argument
2170 containing the DNS resolver options. The host_find_byname() function now
2171 sets these options at its start, just as host_find_bydns() does. The smtp
2172 transport options dns_qualify_single and dns_search_parents are passed to
2173 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2174 of host_find_byname() use the default settings of RES_DEFNAMES
2175 (qualify_single) but not RES_DNSRCH (search_parents).
2177 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2178 spool_read_header() do less string testing, by means of a preliminary
2179 switch on the second character of optional "-foo" lines. (This is
2180 overdue, caused by the large number of possibilities that now exist.
2181 Originally there were few.) While I was there, I also converted the
2182 str(n)cmp tests so they don't re-test the leading "-" and the first
2183 character, in the hope this might squeeze out yet more improvement.
2185 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2186 flag allowing group syntax was set by the header_syntax check but not
2187 turned off, possible causing trouble later; (2) The flag was not being
2188 set at all for the header_verify test, causing "group"-style headers to
2189 be rejected. I have now set it in this case, and also caused header_
2190 verify to ignore an empty address taken from a group. While doing this, I
2191 came across some other cases where the code for allowing group syntax
2192 while scanning a header line wasn't quite right (mostly, not resetting
2193 the flag correctly in the right place). These bugs could have caused
2194 trouble for malformed header lines. I hope it is now all correct.
2196 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2197 with the "reply" argument non-NULL. The code, however (which originally
2198 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2199 but it didn't always do it. This confused somebody who was copying the
2200 code for some other use. I have removed all the tests.
2202 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2203 feature that was used to support insecure browsers during the U.S. crypto
2204 embargo. It requires special client support, and Exim is probably the
2205 only MTA that supported it -- and would never use it because real RSA is
2206 always available. This code has been removed, because it had the bad
2207 effect of slowing Exim down by computing (never used) parameters for the
2208 RSA_EXPORT functionality.
2210 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2211 authenticator to fail if there's a tab character in the incoming data
2212 (there should never be unless someone is messing about, as it's supposed
2213 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2214 if the connection is using TLS or if the remote IP is the same as the
2215 local IP, and the "valid-client-cert option" if a client certificate has
2218 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2219 authenticators. This can be used for authorization after authentication
2220 succeeds. (In the case of plaintext, it servers for both authentication
2223 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2224 if any retry times were supplied.
2226 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2227 connection, where there is no client IP address to check. In this
2228 situation, the verify now always succeeds.
2230 PH/25 Applied John Jetmore's -Mset patch.
2232 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2234 PH/27 In a string expansion for a processed (not raw) header when multiple
2235 headers of the same name were present, leading whitespace was being
2236 removed from all of them, but trailing whitespace was being removed only
2237 from the last one. Now trailing whitespace is removed from each header
2238 before concatenation. Completely empty headers in a concatenation (as
2239 before) are ignored.
2241 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2242 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2244 PH/29 [Removed. This was a change that I later backed out, and forgot to
2245 correct the ChangeLog entry (that I had efficiently created) before
2246 committing the later change.]
2248 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2249 address errors (4xx response to RCPT) over the same connection as other
2250 messages routed to the same hosts. Such deliveries are always "forced",
2251 so retry times are not inspected. This resulted in far too many retries
2252 for the affected addresses. The effect occurred only when there were more
2253 hosts than the hosts_max_try setting in the smtp transport when it had
2254 the 4xx errors. Those hosts that it had tried were not added to the list
2255 of hosts for which the message was waiting, so if all were tried, there
2256 was no problem. Two fixes have been applied:
2258 (i) If there are any address or message errors in an SMTP delivery, none
2259 of the hosts (tried or untried) are now added to the list of hosts
2260 for which the message is waiting, so the message should not be a
2261 candidate for sending over the same connection that was used for a
2262 successful delivery of some other message. This seems entirely
2263 reasonable: after all the message is NOT "waiting for some host".
2264 This is so "obvious" that I'm not sure why it wasn't done
2265 previously. Hope I haven't missed anything, but it can't do any
2266 harm, as the worst effect is to miss an optimization.
2268 (ii) If, despite (i), such a delivery is accidentally attempted, the
2269 routing retry time is respected, so at least it doesn't keep
2270 hammering the server.
2272 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2273 in ${readsocket because some servers need this prod.
2275 PH/32 Added some extra debug output when updating a wait-xxx database.
2277 PH/33 The hint "could be header name not terminated by colon", which has been
2278 given for certain expansion errors for a long time, was not being given
2279 for the ${if def:h_colon_omitted{... case.
2281 PH/34 The spec says: "With one important exception, whenever a domain list is
2282 being scanned, $domain contains the subject domain." There was at least
2283 one case where this was not true.
2285 PH/35 The error "getsockname() failed: connection reset by peer" was being
2286 written to the panic log as well as the main log, but it isn't really
2287 panic-worthy as it just means the connection died rather early on. I have
2288 removed the panic log writing for the ECONNRESET error when getsockname()
2291 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2292 runs only) independently of the message's sender address. This meant
2293 that, if the 4xx error was in fact related to the sender, a different
2294 message to the same recipient with a different sender could confuse
2295 things. In particualar, this can happen when sending to a greylisting
2296 server, but other circumstances could also provoke similar problems.
2297 I have changed the default so that the retry time for these errors is now
2298 based a combination of the sender and recipient addresses. This change
2299 can be overridden by setting address_retry_include_sender=false in the
2302 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2303 remote server are returned as part of bounce messages. This was not
2304 happening for LMTP over a pipe (the lmtp transport), but now it is the
2305 same for both kinds of LMTP.
2307 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2308 in header lines that were in fact CNAMEs. This is no longer the case.
2310 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2311 and queue runs started by the daemon processed all messages. This has
2312 been fixed so that -R and -S can now usefully be given with -q<time>.
2314 PH/40 Import PCRE release 6.7 (fixes some bugs).
2316 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2318 PH/42 Give an error if -q is specified more than once.
2320 PH/43 Renamed the variables $interface_address and $interface_port as
2321 $received_ip_address and $received_port, to make it clear that these
2322 values apply to message reception, and not to the outgoing interface when
2323 a message is delivered. (The old names remain recognized, of course.)
2325 PH/44 There was no timeout on the connect() call when using a Unix domain
2326 socket in the ${readsocket expansion. There now is.
2328 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2329 be meaningful with "accept".
2331 SC/01 Eximstats V1.43
2332 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2334 SC/02 Eximstats V1.44
2335 Use a glob alias rather than an array ref in the generated
2336 parser. This improves both readability and performance.
2338 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2339 Collect SpamAssassin and rejection statistics.
2340 Don't display local sender or destination tables unless
2341 there is data to show.
2342 Added average volumes into the top table text output.
2344 SC/04 Eximstats V1.46
2345 Collect data on the number of addresses (recipients)
2346 as well as the number of messages.
2348 SC/05 Eximstats V1.47
2349 Added 'Message too big' to the list of mail rejection
2350 reasons (thanks to Marco Gaiarin).
2352 SC/06 Eximstats V1.48
2353 Mainlog lines which have GMT offsets and are too short to
2354 have a flag are now skipped.
2356 SC/07 Eximstats V1.49 (Alain Williams)
2357 Added the -emptyok flag.
2359 SC/08 Eximstats V1.50
2360 Fixes for obtaining the IP address from reject messages.
2362 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2363 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2364 whitesspace changes from 4.64-PH/27
2366 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2369 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2370 are found, allow negative numbers in numeric criteria)
2372 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2374 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2375 to match changes made in 4.64-PH/43
2377 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2379 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2380 "domain" is actually an IP address, and force a failure. This locks out
2381 those revolvers/nameservers that support "A-for-A" lookups, in
2382 contravention of the specifications.
2384 PH/48 When a host name was looked up from an IP address, and the subsequent
2385 forward lookup of the name timed out, the host name was left in
2386 $sender_host_name, contrary to the specification.
2388 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2389 restricted to single-key lookups, Exim was not diagnosing an error if
2390 * or *@ was used with a query-style lookup.
2392 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2394 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2395 long ago, but noone interested enough thought of it. Let's just say that
2396 the "1.1" means that there are some new functions that weren't there at
2397 some point in the past.
2399 PH/51 Error processing for expansion failure of helo_data from an smtp
2400 transport during callout processing was broken.
2402 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2403 tested/used via the -bh/-bhc/-bs options.
2405 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2406 bug, fixed in subsequent PCRE releases).
2408 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2409 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2411 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2417 SC/01 Use a glob alias rather than an array ref in eximstats generated
2418 parser. This improves both readability and performance.
2420 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2421 Don't display local sender or destination tables in eximstats unless
2422 there is data to show.
2423 Added average volumes into the eximstats top table text output.
2425 SC/03 Collect data on the number of addresses (recipients) as well
2426 as the number of messages in eximstats.
2428 TF/01 Correct an error in the documentation for the redirect router. Exim
2429 does (usually) call initgroups() when daemonizing.
2431 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2432 with consistent privilege compared to when running as a daemon.
2434 TF/03 Note in the spec that $authenticated_id is not set for local
2435 submissions from trusted users.
2437 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2438 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2440 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2441 by adding some example configuration directives to the default
2442 configuration file. A little bit of work is required to uncomment the
2443 directives and define how usernames and passwords are checked, but
2444 there is now a framework to start from.
2446 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2447 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2448 without this. I don't know how relevant this is to other LDAP libraries.
2450 PH/02 Add the verb name to the "unknown ACL verb" error.
2452 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2454 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2456 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2457 directory not expanded when it should be if an expanded home directory
2458 was set for the address (which is overridden by the transport).
2460 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2463 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2464 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2465 because it is too late at that time, and has no effect.
2467 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2468 security issue with \' (bugzilla #107). I could not use the
2469 PQescapeStringConn() function, because it needs a PGconn value as one of
2472 PH/08 When testing addresses using -bt, indicate those final addresses that
2473 are duplicates that would not cause an additional delivery. At least one
2474 person was confused, thinking that -bt output corresponded to deliveries.
2475 (Suppressing duplicates isn't a good idea as you lose the information
2476 about possibly different redirections that led to the duplicates.)
2478 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2479 systems where poll() doesn't work, in particular OS X.
2481 PH/10 Added more information to debugging output for retry time not reached.
2483 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2484 operations in malware.c.
2486 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2489 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2490 syslog_duplication set false, log lines that would normally be written
2491 both the the main log and to the reject log were not written to syslog at
2494 PH/14 In the default configuration, change the use of "message" in ACL warn
2495 statements to "add_header".
2497 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2498 not followed by a command (e.g. "seen endif").
2500 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2501 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2504 PH/17 Added extra conditions to the default value of delay_warning_condition
2508 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2509 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2510 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2513 The Auto-Submitted: and various List- headers are standardised, whereas I
2514 don't think Precedence: ever was.
2516 PH/18 Refactored debugging code in route_finduser() to show more information,
2517 in particular, the error code if getpwnam() issues one.
2519 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2520 This is apparently needed in addition to the PH/07 change above to avoid
2521 any possible encoding problems.
2523 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2524 but not after initializing Perl.
2526 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2527 output them only if debugging. By default they are written stderr,
2528 apparently, which is not desirable.
2530 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2533 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2536 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2538 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2539 authenticated or an ident call has been made. Suppress the default
2540 values for $authenticated_id and $authenticated_sender (but permit -oMai
2541 and -oMas) when testing with -bh.
2543 PH/24 Re-jigged the order of the tests in the default configuration so that the
2544 tests for valid domains and recipients precede the DNS black list and CSA
2545 tests, on the grounds that those ones are more expensive.
2547 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2548 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2549 This bug exists in every version of Exim that I still have, right back to
2552 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2553 However, an attempt to turn on case-sensitivity in a regex key by
2554 including (?-i) didn't work because the subject string was already
2555 lowercased, and the effects were non-intuitive. It turns out that a
2556 one-line patch can be used to allow (?-i) to work as expected.
2562 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2563 other effects) broke the use of negated acl sub-conditions.
2565 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2568 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2569 "Deny" causes Exim to reject the incoming connection with a 554 error.
2570 Unfortunately, if there is a major crisis, such as a disk failure,
2571 tcp-wrappers gives "deny", whereas what one would like would be some
2572 kind of temporary error. A kludge has been added to help with this.
2573 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2574 554 error is used if errno is still zero or contains ENOENT (which occurs
2575 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2578 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2580 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2581 errors. Otherwise a message that provokes a temporary error (when other
2582 messages do not) can cause a whole host to time out.
2584 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2585 addresses were routed directly to files or pipes from a redirect router.
2586 File deliveries just didn't batch; pipe deliveries might have suffered
2589 PH/06 A failure to get a lock for a hints database would erroneously always say
2590 "Failed to get write lock", even when it was really a read lock.
2592 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2593 of 0600. This has been changed to use the value of the lockfile_mode
2594 option (which defaults to 0600).
2596 PH/08 Applied small patch from the Sieve maintainer.
2598 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2599 folder from quota calculations, a direct delivery into this folder messed
2600 up the contents of the maildirsize file. This was because the regex was
2601 used only to exclude .Trash (or whatever) when the size of the mailbox
2602 was calculated. There was no check that a delivery was happening into an
2603 excluded directory. This bug has been fixed by ignoring all quota
2604 processing for deliveries into excluded directories.
2606 PH/10 Added the maildirfolder_create_regex option to appendfile.
2612 PH/01 The code for finding all the local interface addresses on a FreeBSD
2613 system running IPv6 was broken. This may well have applied to all BSD
2614 systems, as well as to others that have similar system calls. The broken
2615 code found IPv4 interfaces correctly, but gave incorrect values for the
2616 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2617 that it would not match correctly against @[] and not recognize the IPv6
2620 PH/02 The ipliteral router was not recognizing addresses of the form user@
2621 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2623 PH/03 Added disable_ipv6.
2625 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2626 decoded form, because it is most often used to construct To: headers
2627 lines in autoreplies, and the decoded form may well be syntactically
2628 invalid. However, $reply_address has leading white space removed, and all
2629 newlines turned into spaces so that the autoreply transport does not
2632 PH/05 If group was specified without a user on a router, and no group or user
2633 was specified on a transport, the group from the router was ignored.
2635 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2636 for visible compile-time settings that can be used to change these
2637 numbers, for those that want even more. Backwards compatibility with old
2638 spool files has been maintained. However, going back to a previous Exim
2639 release will lost any variables that are in spool files.
2641 PH/07 Two small changes when running in the test harness: increase delay when
2642 passing a TCP/IP connection to a new process, in case the original
2643 process has to generate a bounce, and remove special handling of
2644 127.0.0.2 (sic), which is no longer necessary.
2646 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2647 be the same on different OS.
2649 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2652 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2653 whether --show-vars was specified or not
2655 JJ/02 exipick: Added support for new ACL variable spool format introduced
2658 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2659 syntactically invalid From: or Reply-to: line, and a filter used this to
2660 generate an autoreply, and therefore failed to obtain an address for the
2661 autoreply, Exim could try to deliver to a non-existent relative file
2662 name, causing unrelated and misleading errors. What now happens is that
2663 it logs this as a hard delivery error, but does not attempt to create a
2666 PH/11 The exinext utility has a -C option for testing purposes, but although
2667 the given file was scanned by exinext itself; it wasn't being passed on
2668 when Exim was called.
2670 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2671 an end-of-file indication when reading a command response.
2673 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2674 compiled. In many other places in Exim, IPv6 addresses are always
2675 recognized, so I have changed this. It also means that IPv4 domain
2676 literals of the form [IPV4:n.n.n.n] are now always recognized.
2678 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2679 used if the router is not running as root, for example, when verifying at
2680 ACL time, or when using -bh. The debugging output from this situation was
2681 non-existent - all you got was a failure to exec. I have made two
2684 (a) Failures to set uid/gid, the current directory, or a process leader
2685 in a subprocess such as that created by queryprogram now generate
2686 suitable debugging ouput when -d is set.
2688 (b) The queryprogram router detects when it is not running as root,
2689 outputs suitable debugging information if -d is set, and then runs
2690 the subprocess without attempting to change uid/gid.
2692 PH/15 Minor change to Makefile for building test_host (undocumented testing
2695 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2696 additional section of a DNS packet that returns MX or SRV records.
2697 Instead, it always explicitly searches for A/AAAA records. This avoids
2698 major problems that occur when a DNS server includes only records of one
2699 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2700 fixed another bug: if SRV records were looked up and the corresponding
2701 address records were *not* found in the additional section, the port
2702 values from the SRV records were lost.
2704 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2705 using the correct key (the original address) when searching the retry
2706 rules in order to find which one to use for generating the retry hint.
2708 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2709 adding the default one. Similarly, if it contains a Reply-To: header, the
2710 errors_reply_to option, if set, is not used.
2712 PH/19 When calculating a retry time, Exim used to measure the "time since
2713 failure" by looking at the "first failed" field in the retry record. Now
2714 it does not use this if it is later than than the arrival time of the
2715 message. Instead it uses the arrival time. This makes for better
2716 behaviour in cases where some deliveries succeed, thus re-setting the
2717 "first failed" field. An example is a quota failure for a huge message
2718 when small messages continue to be delivered. Without this change, the
2719 "time since failure" will always be short, possible causing more frequent
2720 delivery attempts for the huge message than are intended.
2721 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2723 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2724 $1, $2, $3) because the numerical variables can be reset during some
2725 expansion items (e.g. "match"), thereby losing the authentication data.
2727 PH/21 Make -bV show the size of off_t variables so that the test suite can
2728 decide whether to run tests for quotas > 2G.
2730 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2731 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2732 filecount value is greater than 2G or if a quota value is greater than 2G
2733 on a system where the size of off_t is not greater than 4, a panic error
2736 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2737 never match. The debug and -bh output now contains an explicit error
2738 message indicating a malformed IPv4 address or mask.
2740 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2741 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2742 PH/23 above applies.
2744 PH/25 Do not write to syslog when running in the test harness. The only
2745 occasion when this arises is a failure to open the main or panic logs
2746 (for which there is an explicit test).
2748 PH/26 Added the /no_tell option to "control=freeze".
2750 PH/27 If a host name lookup failed very early in a connection, for example, if
2751 the IP address matched host_lookup and the reverse lookup yielded a name
2752 that did not have a forward lookup, an error message of the form "no IP
2753 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2754 could be logged. Now it outputs the IP address instead of "NULL".
2756 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2757 allows the sender and the authenticated sender to be set when
2758 submitting a message from within Exim. Since child_open_exim() is
2759 documented for local_scan(), the new function should be too.
2761 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2762 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2763 results in an empty string is now treated as unset.
2765 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2767 PH/31 Added sender_verify_fail logging option.
2769 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2770 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2771 tidied the source and removed it altogether.
2773 PH/33 When a queue run was abandoned because the load average was too high, a
2774 log line was always written; now it is written only if the queue_run log
2775 selector is set. In addition, the log line for abandonment now contains
2776 information about the queue run such as the pid. This is always present
2777 in "start" and "stop" lines but was omitted from the "abandon" line.
2779 PH/34 Omit spaces between a header name and the colon in the error message that
2780 is given when verify = headers_syntax fails (if there are lots of them,
2781 the message gets confusing).
2783 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2784 names, as there are now some PTR records that contain slashes. This check
2785 is only to protect against broken name servers that fall over on strange
2786 characters, so the fact that it applies to all lookups doesn't matter.
2788 PH/36 Now that the new test suite is complete, we can remove some of the
2789 special code in Exim that was needed for the old test suite. For example,
2790 sorting DNS records because real resolvers return them in an arbitrary
2791 order. The new test suite's fake resolver always returns records in the
2794 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2795 bounces) except when queue_only is set, to avoid logging races between
2796 the different processes.
2798 PH/38 Panic-die if .include specifies a non-absolute path.
2800 PH/39 A tweak to the "H" retry rule from its user.
2802 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2803 a label. They prevented compilation on older perls.
2805 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2806 a warning to be raised on newish perls.
2808 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2809 on queue. Changes to match documented behaviour of showing count of
2810 messages matching specified criteria.
2812 PH/40 Changed the default ident timeout from 30s to 5s.
2814 PH/41 Added support for the use of login_cap features, on those BSD systems
2815 that have them, for controlling the resources used by pipe deliveries.
2817 PH/42 The content-scanning code uses fopen() to create files in which to put
2818 message data. Previously it was not paying any attention to the mode of
2819 the files. Exim runs with umask(0) because the rest of the code creates
2820 files with open(), and sets the required mode explicitly. Thus, these
2821 files were ending up world-writeable. This was not a big issue, because,
2822 being within the spool directory, they were not world-accessible. I have
2823 created a function called modefopen, which takes an additional mode
2824 argument. It sets umask(777), creates the file, chmods it to the required
2825 mode, then resets the umask. All the relevant calls to fopen() in the
2826 content scanning code have been changed to use this function.
2828 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2829 to 24 hours. This avoids potential overflow problems when processing G
2830 and H retry rules. I suspect nobody ever tinkers with this value.
2832 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2834 PH/45 When the plaintext authenticator is running as a client, the server's
2835 challenges are checked to ensure they are valid base64 strings. By
2836 default, the authentication attempt is cancelled if an invalid string is
2837 received. Setting client_ignore_invalid_base64 true ignores these errors.
2838 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2839 they are received. Thus, the responses can be made to depend on the
2840 challenges. If an invalid string is ignored, an empty string is placed in
2843 PH/46 Messages that are created by the autoreply transport now contains a
2844 References: header, in accordance with RFCs 2822 and 3834.
2846 PH/47 Added authenticated_sender_force to the smtp transport.
2848 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2850 PH/49 Installed latest patch from the Sieve maintainer.
2852 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2853 was also set, the appendfile transport was unnecessarily scanning a
2854 directory of message files (e.g. for maildir delivery) to find the count
2855 of files (along with the size), even though it did not need this
2856 information. It now does the scan only if it needs to find either the
2857 size of the count of files.
2859 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2861 PH/52 Two bugs concerned with error handling when the smtp transport is
2864 (i) Exim was not creating retry information for temporary errors given
2865 for individual recipients after the DATA command when the smtp transport
2866 was used in LMTP mode. This meant that they could be retried too
2867 frequently, and not timed out correctly.
2869 (ii) Exim was setting the flag that allows error details to be returned
2870 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2871 recipients that were returned after the DATA command.
2873 PH/53 This is related to PH/52, but is more general: for any failing address,
2874 when detailed error information was permitted to be returned to the
2875 sender, but the error was temporary, then after the final timeout, only
2876 "retry timeout exceeded" was returned. Now it returns the full error as
2877 well as "retry timeout exceeded".
2879 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2880 do this, and (what is worse) MTAs that accept it.
2882 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2883 will now be deprecated.
2885 PH/56 New os.c-cygwin from the Cygwin maintainer.
2887 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2888 formats (previously only available in exim formats via -bpr, -bpru,
2889 and -bpra. Now also available in native and exiqgrep formats)
2891 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2892 with very large, slow to parse queues
2894 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2896 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2898 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2899 responses to authentication challenges, though it was showing the
2900 challenges; (ii) I've removed the CR characters from the debug output for
2903 PH/58 Allow for the insertion of a newline as well as a space when a string
2904 is turned into more than one encoded-word during RFC 2047 encoding. The
2905 Sieve code now uses this.
2907 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2908 data_4xx, lost_connection, tls_required.
2910 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2911 message was being sent as an SMTP response.
2913 PH/61 Add -l and -k options to exicyclog.
2915 PH/62 When verifying, if an address was redirected to one new address, so that
2916 verification continued, and the new address failed or deferred after
2917 having set something in $address_data, the value of $address_data was not
2918 passed back to the ACL. This was different to the case when no
2919 redirection occurred. The value is now passed back in both cases.
2921 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2922 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2923 use login_cap.h, so on its own it isn't the distinguishing feature. The
2924 new name refers directly to the setclassresources() function.
2926 PH/65 Added configuration files for NetBSD3.
2928 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2930 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2931 is preferred over IPv4.
2933 PH/68 The bounce_return_message and bounce_return_body options were not being
2934 honoured for bounces generated during the reception of non-SMTP messages.
2935 In particular, this applied to messages rejected by the ACL. This bug has
2936 been fixed. However, if bounce_return_message is true and bounce_return_
2937 body is false, the headers that are returned for a non-SMTP message
2938 include only those that have been read before the error was detected.
2939 (In the case of an ACL rejection, they have all been read.)
2941 PH/69 The HTML version of the specification is now built in a directory called
2942 spec_html instead of spec.html, because the latter looks like a path with
2943 a MIME-type, and this confuses some software.
2945 PH/70 Catch two compiler warnings in sieve.c.
2947 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2948 function verify_get_ident() calls ip_connect() to connect a socket, but
2949 if the "connect()" function timed out, ip_connect() used to close the
2950 socket. However, verify_get_ident() also closes the socket later, and in
2951 between Exim writes to the log, which may get opened at this point. When
2952 the socket was closed in ip_connect(), the log could get the same file
2953 descriptor number as the socket. This naturally causes chaos. The fix is
2954 not to close the socket in ip_connect(); the socket should be closed by
2955 the function that creates it. There was only one place in the code where
2956 this was missing, in the iplookup router, which I don't think anybody now
2957 uses, but I've fixed it anyway.
2959 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2960 well as to direct DNS lookups. Otherwise the handling of names in host
2961 lists is inconsistent and therefore confusing.
2967 PH/01 Two changes to the default runtime configuration:
2969 (1) Move the checks for relay_from_hosts and authenticated clients from
2970 after to before the (commented out) DNS black list checks.
2972 (2) Add control=submission to the relay_from_hosts and authenticated
2973 clients checks, on the grounds that messages accepted by these
2974 statements are most likely to be submissions.
2976 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2978 (1) Generate an error if the third argument for the ${prvs expansion is
2981 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2984 (3) Reset the variables that are obtained from the first argument of
2985 ${prvscheck and used in the second argument before leaving the code,
2986 because their memory is reclaimed, so using them afterwards may do
2989 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2990 one (it's much easier than Tom thought :-).
2992 (5) Because of (4), we can now allow for the use of $prvscheck_result
2993 inside the third argument.
2995 PH/03 For some reason, the default setting of PATH when running a command from
2996 a pipe transport was just "/usr/bin". I have changed it to
2999 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3000 anything to be listed in the output from -bV.
3002 PH/05 When a filter generated an autoreply, the entire To: header line was
3003 quoted in the delivery log line, like this:
3005 => >A.N.Other <ano@some.domain> <original@ddress> ...
3007 This has been changed so that it extracts the operative address. There
3008 may be more than one such address. If so, they are comma-separated, like
3011 => >ano@some.domain,ona@other.domain <original@ddress> ...
3013 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3014 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3015 not being looked up in the rDNS to get a host name, Exim was showing the
3016 IP address twice in Received: lines, even though the IP addresses were
3017 identical. For example:
3019 Received: from [1.2.3.4] (helo=[1.2.3.4])
3021 However, if the real host name was known, it was omitting the HELO data
3022 if it matched the actual IP address. This has been tidied up so that it
3023 doesn't show the same IP address twice.
3025 PH/07 When both +timestamp and +memory debugging was on, the value given by
3026 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3027 called by the debug printing, thereby overwriting the timestamp buffer.
3028 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3030 PH/08 When the original message was included in an autoreply transport, it
3031 always said "this is a copy of the message, including all the headers",
3032 even if body_only or headers_only was set. It now gives an appropriate
3035 PH/09 Applied a patch from the Sieve maintainer which:
3037 o fixes some comments
3038 o adds the (disabled) notify extension core
3039 o adds some debug output for the result of if/elsif tests
3040 o points to the current vacation draft in the documentation
3041 and documents the missing references header update
3045 o fixes a bug in processing the envelope test (when testing
3046 multiple envelope elements, the last element determinted the
3049 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3050 Electronic Mail") by including:
3052 Auto-submitted: auto-generated
3054 in the messages that it generates (bounce messages and others, such as
3055 warnings). In the case of bounce messages for non-SMTP mesages, there was
3056 also a typo: it was using "Auto_submitted" (underscore instead of
3057 hyphen). Since every message generated by Exim is necessarily in response
3058 to another message, thes have all been changed to:
3060 Auto-Submitted: auto-replied
3062 in accordance with these statements in the RFC:
3064 The auto-replied keyword:
3066 - SHOULD be used on messages sent in direct response to another
3067 message by an automatic process,
3069 - MUST NOT be used on manually-generated messages,
3071 - MAY be used on Delivery Status Notifications (DSNs) and Message
3072 Disposition Notifications (MDNs),
3074 - MUST NOT be used on messages generated by automatic or periodic
3075 processes, except for messages which are automatic responses to
3078 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3079 to the default Received: header definition.
3081 PH/12 Added log selector acl_warn_skipped (default on).
3083 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3084 variables because (a) they are in the wrong storage pool and (b) even if
3085 they were copied, it wouldn't work properly because of the caching.
3087 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3088 checking when decoding. Apparently there are clients that generate
3089 overlong encoded strings. Why am I not surprised?
3091 PH/15 If the first argument of "${if match_address" was not empty, but did not
3092 contain an "@" character, Exim crashed. Now it writes a panic log message
3093 and treats the condition as false.
3095 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3097 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3098 extension "envelope-auth". The code is finished and in agreement with
3099 other implementations, but there is no documentation so far and in fact,
3100 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3101 not changing the active code.
3103 Print executed "if" and "elsif" statements when debugging is used. This
3104 helps a great deal to understand what a filter does.
3106 Document more things not specified clearly in RFC3028. I had all this
3107 sorted out, when out of a sudden new issues came to my mind. Oops."
3109 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3112 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3113 canonical form (as documented). However, after a host name lookup from
3114 the IP address, check_host() was doing a simple string comparison with
3115 addresses acquired from the DNS when checking that the found name did
3116 have the original IP as one of its addresses. Since any found IPv6
3117 addresses are likely to be in abbreviated form, the comparison could
3118 fail. Luckily, there already exists a function for doing the comparison
3119 by converting both addresses to binary, so now that is used instead of
3120 the text comparison.
3122 PH/20 There was another similar case to PH/19, when a complete host name was
3123 given in a host list; looking up its IP address could give an abbreviated
3124 form, whereas the current host's name might or might not be abbreviated.
3125 The same fix has been applied.
3131 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3132 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3135 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3136 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3138 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3140 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3141 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3142 Also, for "Auto-Submitted", treat anything other than "no" as
3143 non-personal, in accordance with RFC 3834. (Previously it treated
3144 anything starting "auto-" as non-personal.)
3146 TF/01 The control=submission/name=... option had a problem with syntax
3147 errors if the name included a slash character. The /name= option
3148 now slurps the rest of the string, so it can include any characters
3149 but it must come last in the list of options (after /sender_retain
3152 PH/05 Some modifications to the interface to the fake nameserver for the new
3160 TK/01 Added the "success_on_redirect" address verification option. See
3161 NewStuff for rationale and an example.
3163 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3165 PH/02 Patch to exigrep to allow it to work on syslog lines.
3167 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3168 fread() to skip over the body file's header line, because in Cygwin the
3169 header line is locked and is inaccessible.
3171 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3172 co-exist for some time) to make it clear that it is the Exim ID that is
3173 referenced, not the Message-ID: header line.
3175 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3176 string_format() function, because snprintf() does not exist on all
3179 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3180 ${stat: expansion item.
3182 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3183 protocol synchronization error", to keep the pedants happy.
3185 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3186 well as for IRIX systems, when gcc is being used. See the host.c source
3189 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3191 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3194 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3195 transport and to the smtp transport in LMTP mode.
3197 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3199 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3200 run a filter in a subprocess. This could lead to confusion in subsequent
3201 lookups in the parent process. There should also be a search_tidyup() at
3202 the end of the subprocess.
3204 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3205 only if the host matched helo_try_verify_hosts, which caused the
3206 verification to occur when the EHLO/HELO command was issued. The ACL just
3207 tested the remembered result. Now, if a previous verification attempt has
3208 not happened, "verify = helo" does it there and then.
3210 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3212 TK/03 Fix log output including CR from clamd.
3214 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3215 exist provoked a memory error which could cause a segfault.
3217 PH/15 Installed PCRE 6.2
3219 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3221 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3222 of the problem. Specifically, suggested +O2 rather than +O1 for the
3225 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3227 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3228 record, the port was not correctly logged when the outgoing_port log
3229 selector was set (it logged the transort's default port).
3231 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3232 fallback_hosts, and "hosts" in the smtp transport.
3234 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3235 host errors such as "Connection refused".
3237 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3238 authentication with radiusclient 0.4.9:
3240 - Error returned from rc_read_config was caught wrongly
3241 - Username/password not passed on to radius server due to wrong length.
3243 The presumption is that some radiusclient API changes for 4.51/PH/17
3244 were not taken care of correctly. The code is still untested by me (my
3245 Linux distribution still has 0.3.2 of radiusclient), but it was
3246 contributed by a Radius user.
3248 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3249 expanding the "port" option of the smtp transport.
3251 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3252 while reading a MIME header. Thanks to Tom Hughes for a patch.
3254 PH/24 Include config.h inside local_scan.h so that configuration settings are
3257 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3258 that in an ACL for RCPT (for example), you can examine exactly what was
3261 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3262 commands, but it was not correctly comparing the address with the actual
3263 client host address. Thus, it would show the EHLO address in Received:
3264 header lines when this was not necessary.
3266 PH/27 Added the % operator to ${eval:}.
3268 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3269 it should be ignoring failures (because with -C, for example, it has lost
3270 privilege). It wasn't ignoring creation failures other than "already
3273 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3276 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3277 because an input line was too long, either on its own, or by virtue of
3278 too many continuations, the temporary file was not being removed, and the
3279 return code was incorrect.
3281 PH/31 Missing "BOOL" in function definition in filtertest.c.
3283 PH/32 Applied Sieve patches from the maintainer.
3285 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3287 PH/33 Added "verify = not_blind".
3289 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3290 Local/Makefile (with some defaults set). These are used in built scripts
3291 such as exicyclog, but they have never been used in the exim_install
3292 script (though there are many overriding facilities there). I have
3293 arranged that the exim_install script now takes note of these two
3296 PH/35 Installed configuration files for Dragonfly.
3298 PH/36 When a locally submitted message by a trusted user did not contain a
3299 From: header, and the sender address was obtained from -f or from an SMTP
3300 MAIL command, and the trusted user did not use -F to supply a sender
3301 name, $originator_name was incorrectly used when constructing a From:
3302 header. Furthermore, $originator_name was used for submission mode
3303 messages from external hosts without From: headers in a similar way,
3304 which is clearly wrong.
3306 PH/37 Added control=suppress_local_fixups.
3308 PH/38 When log_selector = +received_sender was set, and the addition of the
3309 sender made the log line's construction buffer exactly full, or one byte
3310 less than full, an overflow happened when the terminating "\n" was
3313 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3314 when the result of a list match is failure because a DNS lookup failed.
3316 PH/40 RM_COMMAND is now used in the building process.
3318 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3319 the "build-* directories that it finds.
3321 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3322 address in a domain literal was a prefix of an interface address.
3324 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3325 when verifying a sender address, unless rewrite_headers is false.
3327 PH/44 Wrote a long comment about why errors_to addresses are verified as
3328 recipients, not senders.
3330 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3331 the ratelimit ACL was added.
3333 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3335 PH/46 Added extra information about PostgreSQL errors to the error string.
3337 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3338 avoiding the need to install special zones in a real server. This is
3339 backwards compatible; if it can't find the fake resolver, it drops back.
3340 Thus, both old and new test suites can be run.
3342 TF/02 Added util/ratelimit.pl
3344 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3345 clock is set back in time.
3347 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3348 Candler <B.Candler@pobox.com>.
3350 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3351 OK for addresses that are the result of redirections.
3353 PH/48 A number of further additions for the benefit of the new test suite,
3354 including a fake gethostbyname() that interfaces to the fake DNS resolver
3357 TF/06 The fix for widen_domains has also been applied to qualify_single and
3358 search_parents which are the other dnslookup options that can cause
3361 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3364 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3366 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3367 service" if it encountered an MX record with an empty target hostname.
3368 The message is now "an MX or SRV record indicated no SMTP service".
3370 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3371 if the DNS of the sending site is misconfigured. This is quite a
3372 common situation. This change restores the behaviour of treating a
3373 helo verification defer as a failure.
3375 PH/51 If self=fail was set on a router, the bounce message did not include the
3376 actual error message.
3382 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3384 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3385 command itself ended in error, the underlying message about the transport
3386 filter timeout was being overwritten with the pipe command error. Now the
3387 underlying error message should be appended to the second error message.
3389 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3391 PH/02 Reduce the amount of output that "make" produces by default. Full output
3392 can still be requested.
3394 PH/03 The warning log line about a condition test deferring for a "warn" verb
3395 was being output only once per connection, rather than after each
3396 occurrence (because it was using the same function as for successful
3397 "warn" verbs). This seems wrong, so I have changed it.
3399 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3400 it should not have, which might have caused a crash in the right
3401 circumstances, but probably never did.
3403 PH/04 Installed a modified version of Tony Finch's patch to make submission
3404 mode fix the return path as well as the Sender: header line, and to
3405 add a /name= option so that you can make the user's friendly name appear
3408 TF/03 Added the control = fakedefer ACL modifier.
3410 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3411 Mark Lowes for thorough testing.
3413 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3415 TK/03 Merged latest SRS patch from Miles Wilton.
3417 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3418 with the definition in sysexits.h (which is #included earlier).
3419 Fortunately, Exim does not actually use EX_OK. The code used to try to
3420 preserve the sysexits.h value, by assumimg that macro definitions were
3421 scanned for macro replacements. I have been disabused of this notion,
3422 so now the code just undefines EX_OK before #including unistd.h.
3424 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3425 in the smtp transport. When a block could not be written in a single
3426 write() function, the timeout was being re-applied to each part-write.
3427 This seems wrong - if the receiver was accepting one byte at a time it
3428 would take for ever. The timeout is now adjusted when this happens. It
3429 doesn't have to be particularly precise.
3431 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3432 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3434 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3435 without a domain if the check to <postmaster@domain> fails.
3437 SC/01 Eximstats: added -xls and the ability to specify output files
3438 (patch written by Frank Heydlauf).
3440 SC/02 Eximstats: use FileHandles for outputing results.
3442 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3444 SC/04 Eximstats: fixed display of large numbers with -nvr option
3446 SC/05 Eximstats: fixed merging of reports with empty tables.
3448 SC/06 Eximstats: added the -include_original_destination flag
3450 SC/07 Eximstats: removed tabs and trailing whitespace.
3452 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3454 TK/06 MBOX spool code: Add real "From " MBOX separator line
3455 so the .eml file is really in mbox format (even though
3456 most programs do not really care). Patch from Alex Miller.
3458 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3459 The latter is generated from $received_to and is only set if the
3460 message has one envelope recipient. SA can use these headers,
3461 obviously out-of-the-box. Patch from Alex Miller.
3463 PH/08 The ${def test on a variable was returning false if the variable's
3464 value was "0", contrary to what the specification has always said!
3465 The result should be true unless the variable is empty.
3467 PH/09 The syntax error of a character other than { following "${if
3468 def:variable_name" (after optional whitespace) was not being diagnosed.
3469 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3470 accidental colon was present, for example, could give incorrect results.
3472 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3473 result is used (not including appendfile, where other changes are about
3476 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3477 This involved changing a lot of size variables from int to off_t. It
3478 should work with maildirs and everything.
3480 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3481 spamd dying while we are connected to it.
3483 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3486 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3487 function for BDB 4.3.
3489 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3491 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3492 the sources for PCRE have been re-arranged and more files are now
3495 PH/15 The code I had for printing potentially long long variables in PH/11
3496 above was not the best (it lost precision). The length of off_t variables
3497 is now inspected at build time, and an appropriate printing format (%ld
3498 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3499 to be "long long int" or "long int". This is needed for the internal
3500 formatting function string_vformat().
3502 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3503 the configuration file to be ":syslog", then the script "guesses" where
3504 the logs files are, rather than using the compiled in default. In our
3505 case the guess is not the same as the compiled default, so the script
3506 suddenly stopped working when I started to use syslog. The patch checks
3507 to see if log_file_path is "". If so, it attempts to read it from exim
3508 with no configuration file to get the compiled in version, before it
3509 falls back to the previous guessing code."
3511 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3512 implementing BATV in an Exim configuration. See NewStuff for the gory
3515 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3516 Makefile that are specific to HP-UX.
3518 PH/18 If the "use_postmaster" option was set for a recipient callout together
3519 with the "random" option, the postmaster address was used as the MAIL
3520 FROM address for the random test, but not for the subsequent recipient
3521 test. It is now used for both.
3523 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3524 patch removes a few documentation additions to RFC 3028, because the
3525 latest draft now contains them. It adds the new en;ascii-case comparator
3526 and a new error check for 8bit text in MIME parts. Comparator and
3527 require names are now matched exactly. I enabled the subaddress
3528 extension, but it is not well tested yet (read: it works for me)."
3530 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3531 rework some of the code of TK/09 above to avoid the hardwired use of
3532 "%lld" and "long long". Replaced the call to snprintf() with a call to
3535 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3536 records point to non-existent hosts", "retry timeout exceeded", and
3537 "retry time not reached for any host after a long failure period".
3539 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3540 experimental DomainKeys support:
3542 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3543 (2) On an error such as an illegally used "control", the wrong name for
3544 the control was given.
3546 These problems did NOT occur unless DomainKeys support was compiled.
3548 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3550 PH/24 Added ${if match_ip condition.
3552 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3553 databases so that it will be absolutely obvious if a crash occurs in the
3554 DB library. This is a regular occurrence (often caused by mis-matched
3557 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3558 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3559 picked up on a user's system that detects such things. There doesn't seem
3560 to be a gcc warning option for this - only an attribute that has to be
3561 put on the function's prototype. It seems that in Fedora Core 4 they have
3562 set this on a number of new functions. No doubt there will be more in due
3565 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3566 specify a transport. However, if an address that was verified by such a
3567 router was the subject of a callout, Exim crashed because it tried to
3568 read the rcpt_include_affixes from the non-existent transport. Now it
3569 just assumes that the setting of that option is false. This bug was
3570 introduced by 4.51/PH/31.
3572 PH/28 Changed -d+all to exclude +memory, because that information is very
3573 rarely of interest, but it makes the output a lot bigger. People tend to
3574 do -d+all out of habit.
3576 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3577 code in os-type was giving problems when libc.so lives in lib64, like on
3580 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3581 aren't the modern standard, and it seems that some systems' include files
3582 don't always have them. Exim was already checking for some of the newer
3583 ones like T_AAAA, and defining it itself. I've added checks for all the
3584 record types that Exim uses.
3586 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3587 not automatically generating a new one, as it is supposed to. This
3588 prevented TLS from working. If the file did exist, but contained invalid
3589 data, a new version was generated, as expected. It was only the case of a
3590 non-existent file that was broken.
3592 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3593 with a change in libdomainkeys > 0.64.
3595 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3596 from DNS. If the selector record carries the flag, it now has
3597 precedence over the domain-wide flag.
3599 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3601 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3602 the use of a port name that isn't defined in /etc/services) occurred, the
3603 message was deferred as in a normal delivery, and thus remained on the
3604 spool, instead of being failed because of the mua_wrapper setting. This
3605 is now fixed, and I tidied up some of the mua_wrapper messages at the
3608 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3609 the messages in a hash of arrays rather than using individual hashes.
3610 This is a bit cleaner and results in dramatic memory savings, albeit
3611 at a slight CPU cost.
3613 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3614 as requested by Marc Sherman.
3616 SC/10 Eximstats: added histograms for user specified patterns as requested
3619 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3621 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3622 fopen() in the content-scanning modules that did not already have it.
3628 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3629 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3631 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3633 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3635 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3636 to test Sieve filters that use "vacation".
3638 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3639 that changes the way the GnuTLS parameters are stored in the cache file.
3640 The new format can be generated externally. For backward compatibility,
3641 if the data in the cache doesn't make sense, Exim assumes it has read an
3642 old-format file, and it generates new data and writes a new file. This
3643 means that you can't go back to an older release without removing the
3646 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3647 work if there are any delivery delays because "one_time" forces the
3648 parent to be marked "delivered", so its unseen clone is never tried
3649 again. For this reason, Exim now forbids the simultaneous setting of
3652 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3653 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3654 entry if you want to know the details. The fix, however, neglected to
3655 consider the case when local delivery batching is involved. The test for
3656 "previously delivered" was not happening when checking to see if an
3657 address could be batched with a previous (undelivered) one; under
3658 certain circumstances this could lead to multiple deliveries to the same
3661 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3662 in its include files, and this causes problems building Exim.
3664 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3665 header_syntax) but Exim was just ignoring anything given after a slash.
3666 In particular, this caused confusion with an attempt to use "verify =
3667 reverse_host_lookup/defer_ok". An error is now given when options are
3668 supplied for verify items that do not have them. (Maybe reverse_host_
3669 lookup should have a defer_ok option, but that's a different point.)
3671 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3672 defined by RFC 821) to 2048, because there were problems with some AUTH
3673 commands, and RFC 1869 says the size should be increased for extended
3674 SMTP commands that take arguments.
3676 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3679 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3680 "unknown" error; now it says that the functionality isn't in the binary.
3682 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3683 an address' error message when a string expansion fails (syntax or
3684 whatever). Otherwise the password may appear in the log. Following change
3685 PH/42 below, there is no longer a chance of it appearing in a bounce
3688 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3690 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3691 of its addresses was ever tried. (Bugzilla bug #2.)
3693 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3694 the result incorrectly in the debug output. (It correctly added a newline
3695 to what was transported.)
3697 TF/01 Added $received_time.
3699 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3700 commented out examples of how to interface to a virus scanner and to
3701 SpamAssassin. Also added commented examples of av_scanner and
3702 spamd_address settings.
3704 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3705 and controls are allowed in which ACLs. There were a couple of minor
3706 errors. Some of the entries in the conditions table (which is a table of
3707 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3708 negation of where the condition IS allowed.
3710 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3712 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3713 header file does not have a version number, so I've had to invent a new
3714 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3715 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3716 radiusclient), but it was contributed by a Radius user.
3718 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3719 files or directories, for OpenSSL.
3721 PH/19 When an Exim process that is running as root has to create an Exim log
3722 file, it does so in a subprocess that runs as exim:exim so as to get the
3723 ownership right at creation (otherwise, other Exim processes might see
3724 the file with the wrong ownership). There was no test for failure of this
3725 fork() call, which would lead to the process getting stuck as it waited
3726 for a non-existent subprocess. Forks do occasionally fail when resources
3727 run out. I reviewed all the other calls to fork(); they all seem to check
3730 PH/20 When checking for unexpected SMTP input at connect time (before writing
3731 the banner), Exim was not dealing correctly with a non-positive return
3732 from the read() function. If the client had disconnected by this time,
3733 the result was a log entry for a synchronization error with an empty
3734 string after "input=" when read() returned zero. If read() returned -1
3735 (an event I could not check), uninitialized data bytes were printed.
3736 There were reports of junk text (parts of files, etc) appearing after
3739 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3741 PH/22 Added support for macro redefinition, and (re)definition in between
3742 driver and ACL definitions.
3744 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3745 forgetting to use the resulting value; it was using the unexpanded value.
3747 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3748 hadn't been configured. The fix is from Juergen Kreileder, who
3749 understands it better than I do:
3751 "Here's what I see happening with three configured cyrus_sasl
3752 authenticators configured (plain, login, cram-md5):
3754 On startup auth_cyrus_sasl_init() gets called for each of these.
3755 This means three calls to sasl_listmech() without a specified mech_list.
3756 => SASL tests which mechs of all available mechs actually work
3757 => three warnings about OTP not working
3758 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3760 With the patch, sasl_listmech() also gets called three times. But now
3761 SASL's mech_list option is set to the server_mech specified in the the
3762 authenticator. Or in other words, the answer from sasl_listmech()
3763 gets limited to just the mech you're testing for (which is different
3765 => the return list contains just 'plain' or 'login', 'cram-md5' or
3766 nothing depending on the value of ob->server_mech.
3768 I've just tested the patch: Authentication still works fine,
3769 unavailable mechs specified in the exim configuration are still
3770 caught, and the auth.log warnings about OTP are gone."
3772 PH/25 When debugging is enabled, the contents of the command line are added
3773 to the debugging output, even when log_selector=+arguments is not
3776 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3777 answer is "GNU", and only if the return is "GNU/something" is the answer
3780 PH/27 $acl_verify_message is now set immediately after the failure of a
3781 verification in an ACL, and so is available in subsequent modifiers. In
3782 particular, the message can be preserved by coding like this:
3784 warn !verify = sender
3785 set acl_m0 = $acl_verify_message
3787 Previously, $acl_verify_message was set only while expanding "message"
3788 and "log_message" when a very denied access.
3790 PH/28 Modified OS/os.c-Linux with
3792 -#ifndef OS_LOAD_AVERAGE
3793 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3795 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3796 nomenclature these days.)
3798 PH/29 Installed patch from the Sieve maintainer that adds the options
3799 sieve_useraddress and sieve_subaddress to the redirect router.
3801 PH/30 In these circumstances:
3802 . Two addresses routed to the same list of hosts;
3803 . First host does not offer TLS;
3804 . First host accepts first address;
3805 . First host gives temporary error to second address;
3806 . Second host offers TLS and a TLS session is established;
3807 . Second host accepts second address.
3808 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3809 and peerdn, if requested) that were in fact used only for the second
3812 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3813 attention to any local part prefix or suffix that was matched by the
3814 router that accepted the address. It now behaves in the same way as it
3815 does for delivery: the affixes are removed from the local part unless
3816 rcpt_include_affixes is set on the transport.
3818 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3819 timeout during the DATA phase of an incoming message.
3821 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3822 applied a patch sanctioned by the Sieve maintainer.
3824 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3825 the uid or gid is negative. A case of a negative gid caused this to be
3826 noticed. The fix allows for either to be negative.
3828 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3829 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3832 PH/36 The change PH/12 above was broken. Fixed it.
3834 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3835 the grounds that routing the same address twice would always produce the
3836 same answer. This might have been true once, but it is certainly no
3837 longer true now. Routing a child address may depend on the previous
3838 routing that produced that child. Some complicated redirection strategies
3839 went wrong when messages had multiple recipients, and made Exim's
3840 behaviour dependent on the order in which the addresses were given.
3842 I have moved the duplicate checking until after the routing is complete.
3843 Exim scans the addresses that are assigned to local and remote
3844 transports, and removes any duplicates. This means that more work will be
3845 done, as duplicates will always all be routed, but duplicates are
3846 presumably rare, so I don't expect this is of any significance.
3848 For deliveries to pipes, files, and autoreplies, the duplicate checking
3849 still happens during the routing process, since they are not going to be
3852 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3853 It corrects a timeout issue with spamd. This is Ian's comment: "The
3854 background is that sometimes spamd either never reads data from a
3855 connection it has accepted, or it never writes response data. The exiscan
3856 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3857 blindly assumes that writes won't block so it may never time out."
3859 PH/39 Allow G after quota size as well as K and M.
3861 PH/40 The value set for $authenticated_id in an authenticator may not contain
3862 binary zeroes or newlines because the value is written to log lines and
3863 to spool files. There was no check on this. Now the value is run through
3864 the string_printing() function so that such characters are converted to
3865 printable escape sequences.
3867 PH/41 $message_linecount is a new variable that contains the total number of
3868 lines in the message. Compare $body_linecount, which is the count for the
3871 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3872 bounce and delay warning messages, except in certain special cases, which
3875 (a) An SMTP error message from a remote host;
3876 (b) A message specified in a :fail: redirection;
3877 (c) A message specified in a "fail" command in a system filter;
3878 (d) A message specified in a FAIL return from the queryprogram router;
3879 (e) A message specified by the cannot_route_message router option.
3881 In these cases only, Exim does include the error details in bounce and
3882 warning messages. There are also a few cases where bland messages such
3883 as "unrouteable address" or "local delivery error" are given.
3885 PH/43 $value is now also set for the "else" part of a ${run expansion.
3887 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3888 being worked on, but at least Exim now implements the latest version to
3891 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3892 process to complete was treated as a delivery failure, a timeout while
3893 writing the message to the pipe was logged, but erroneously treated as a
3894 successful delivery. Such timeouts include transport filter timeouts. For
3895 consistency with the overall process timeout, these timeouts are now
3896 treated as errors, giving rise to delivery failures by default. However,
3897 there is now a new Boolean option for the pipe transport called
3898 timeout_defer, which, if set TRUE, converts the failures into defers for
3899 both kinds of timeout. A transport filter timeout is now identified in
3902 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3903 systems where "make" and "gmake" are different, calling "gmake" at top
3904 level broke things. I've arranged for the value of $(MAKE) to be passed
3905 from the Makefile to this script so that it can call the same version of
3909 A note about Exim versions 4.44 and 4.50
3910 ----------------------------------------
3912 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3913 changes of various kinds. As a consequence, a big documentation update was
3914 needed. This delayed the release for rather longer than seemed good, especially
3915 in the light of a couple of (minor) security issues. Therefore, the changes
3916 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3917 release. So 4.44 and 4.50 are in effect two different branches that both start
3920 I have left the 4.50 change log unchanged; it contains all the changes since
3921 4.43. The change log for 4.44 is below; many of its items are identical to
3922 those for 4.50. This seems to be the most sensible way to preserve the
3923 historical information.
3929 1. Minor wording change to the doc/README.SIEVE file.
3931 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3932 computation of the current number of files was incorrect.
3934 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3935 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3938 4. Give more explanation in the error message when the command for a transport
3939 filter fails to execute.
3941 5. There are several places where Exim runs a non-Exim command in a
3942 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3943 was being done only for the command run by the queryprogram router. It is
3944 now done for all such subprocesses. The other cases are: ${run, transport
3945 filters, and the commands run by the lmtp and pipe transports.
3947 6. Added CONFIGURE_GROUP build-time option.
3949 7. Some older OS have a limit of 256 on the maximum number of file
3950 descriptors. Exim was using setrlimit() to set 1000 as a large value
3951 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3952 systems. I've change it so that if it can't get 1000, it tries for 256.
3954 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3955 was an oversight, and furthermore, ever since the addition of extra
3956 controls (e.g. 4.43/32), the checks on when to allow different forms of
3957 "control" were broken. There should now be diagnostics for all cases when a
3958 control that does not make sense is encountered.
3960 9. Added the /retain_sender option to "control=submission".
3962 10. $recipients is now available in the predata ACL (oversight).
3964 11. Tidy the search cache before the fork to do a delivery from a message
3965 received from the command line. Otherwise the child will trigger a lookup
3966 failure and thereby defer the delivery if it tries to use (for example) a
3967 cached ldap connection that the parent has called unbind on.
3969 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3970 of $address_data from the recipient verification was clobbered by the
3971 sender verification.
3973 13. The value of address_data from a sender verification is now available in
3974 $sender_address_data in subsequent conditions in the ACL statement.
3976 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3978 15. Added a new option "connect=<time>" to callout options, to set a different
3981 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3982 was its contents. (It was OK if the option was not defined at all.)
3984 17. A "Completed" log line is now written for messages that are removed from
3985 the spool by the -Mrm option.
3987 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3988 information about exactly what failed.
3990 19. Added -dd to debug only the daemon process.
3992 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3993 handles timeouts, both on the server side and network timeouts. Renamed the
3994 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3996 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3997 It is now set to "smtps".
3999 22. $host_address is now set to the target address during the checking of
4000 ignore_target_hosts.
4002 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4003 being passed; this would have caused $sender_host_name to have been used if
4004 matching the list had actually called for a host name (not very likely,
4005 since this list is usually IP addresses). A host name is now passed as
4008 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4009 code that specifies a non-restarting handler (typically sigaction() in
4010 modern systems) in an attempt to fix a rare and obscure crash bug.
4012 25. Narrowed the window for a race in the daemon that could cause it to ignore
4013 SIGCHLD signals. This is not a major problem, because they are used only to
4014 wake it up if nothing else does.
4016 26. A malformed maildirsize file could cause Exim to calculate negative values
4017 for the mailbox size or file count. Odd effects could occur as a result.
4018 The maildirsize information is now recalculated if the size or filecount
4021 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4022 support for a long time. Removed HAVE_SYS_VFS_H.
4024 28. Installed the latest version of exipick from John Jetmore.
4026 29. In an address list, if the pattern was not a regular expression, an empty
4027 subject address (from a bounce message) matched only if the pattern was an
4028 empty string. Non-empty patterns were not even tested. This was the wrong
4029 because it is perfectly reasonable to use an empty address as part of a
4030 database query. An empty address is now tested by patterns that are
4031 lookups. However, all the other forms of pattern expect the subject to
4032 contain a local part and a domain, and therefore, for them, an empty
4033 address still always fails if the pattern is not itself empty.
4035 30. Exim went into a mad DNS loop when attempting to do a callout where the
4036 host was specified on an smtp transport, and looking it up yielded more
4037 than one IP address.
4039 31. Re-factored the code for checking spool and log partition space into a
4040 function that finds that data and another that does the check. The former
4041 is then used to implement four new variables: $spool_space, $log_space,
4042 $spool_inodes, and $log_inodes.
4044 32. The RFC2047 encoding function was originally intended for short strings
4045 such as real names; it was not keeping to the 75-character limit for
4046 encoded words that the RFC imposes. It now respects the limit, and
4047 generates multiple encoded words if necessary. To be on the safe side, I
4048 have increased the buffer size for the ${rfc2047: expansion operator from
4051 33. It is now permitted to omit both strings after an "if" condition; if the
4052 condition is true, the result is "true". As before, when the second string
4053 is omitted, a false condition yields an empty string. This makes it less
4054 cumbersome to write custom ACL and router conditions.
4056 34. Failure to deliver a bounce message always caused it to be frozen, even if
4057 there was an errors_to setting on the router. The errors_to setting is now
4060 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4061 canonical form (fully expanded) before being placed in
4062 $sender_host_address.
4064 36. The table in the code that translates DNS record types into text (T_A to
4065 "A" for instance) was missing entries for NS and CNAME. It is just possible
4066 that this could have caused confusion if both these types were looked up
4067 for the same domain, because the text type is used as part of Exim's
4068 per-process caching. But the chance of anyone hitting this buglet seems
4071 37. The dnsdb lookup has been extended in a number of ways.
4073 (1) There is a new type, "zns", which walks up the domain tree until it
4074 finds some nameserver records. It should be used with care.
4076 (2) There is a new type, "mxh", which is like "mx" except that it returns
4077 just the host names, not the priorities.
4079 (3) It is now possible to give a list of domains (or IP addresses) to be
4080 looked up. The behaviour when one of the lookups defers can be
4081 controlled by a keyword.
4083 (4) It is now possible to specify the separator character for use when
4084 multiple records are returned.
4086 38. The dnslists ACL condition has been extended: it is now possible to supply
4087 a list of IP addresses and/or domains to be looked up in a particular DNS
4090 39. Added log_selector=+queue_time_overall.
4092 40. When running the queue in the test harness, wait just a tad after forking a
4093 delivery process, to get repeatability of debugging output.
4095 41. Include certificate and key file names in error message when GnuTLS fails
4096 to set them up, because the GnuTLS error message doesn't include the name
4097 of the failing file when there is a problem reading it.
4099 42. Allow both -bf and -bF in the same test run.
4101 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4103 44. The "Exiscan patch" is now merged into the mainline Exim source.
4105 45. Sometimes the final signoff response after QUIT could fail to get
4106 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4107 < 0 before doing a fflush(). This bug looks as though it goes back to the
4108 introduction of TLS in release 3.20, but "sometimes" must have been rare
4109 because the tests only now provoked it.
4111 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4112 (this can affect the format of dates).
4114 47. exim_tidydb, when checking for the continued existence of a message for
4115 which it has found a message-specific retry record, was not finding
4116 messages that were in split spool directories. Consequently, it was
4117 deleting retry records that should have stayed in existence.
4119 48. Steve fixed some bugs in eximstats.
4121 49. The SPA authentication driver was not abandoning authentication and moving
4122 on to the next authenticator when an expansion was forced to fail,
4123 contradicting the general specification for all authenticators. Instead it
4124 was generating a temporary error. It now behaves as specified.
4126 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4127 (the order specifies the preference for clients). The order is now AES256,
4128 AES128, 3DES, ARCFOUR128.
4130 51. Small patch to Sieve code - explicitly set From: when generating an
4133 52. Exim crashed if a remote delivery caused a very long error message to be
4134 recorded - for instance if somebody sent an entire SpamAssassin report back
4135 as a large number of 550 error lines. This bug was coincidentally fixed by
4136 increasing the size of one of Exim's internal buffers (big_buffer) that
4137 happened as part of the Exiscan merge. However, to be on the safe side, I
4138 have made the code more robust (and fixed the comments that describe what
4141 53. Now that there can be additional text after "Completed" in log lines (if
4142 the queue_time_overall log selector is set), a one-byte patch to exigrep
4143 was needed to allow it to recognize "Completed" as not the last thing in
4146 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4147 patch that reportedly fixes this has been added. I am not expert enough to
4148 create a test for it. This is what the patch creator wrote:
4150 "I found a little strange behaviour of ldap code when working with
4151 Windows 2003 AD Domain, where users was placed in more than one
4152 Organization Units. When I tried to give exim partial DN, the exit code
4153 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4154 But simultaneously result of request was absolutely normal ldap result,
4155 so I produce this patch..."
4157 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4158 REFERENCE, so I have modified the code to exclude the patch when that macro
4161 55. Some experimental protocols are using DNS PTR records for new purposes. The
4162 keys for these records are domain names, not reversed IP addresses. The
4163 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4164 leaves it alone. Component reversal etc. now happens only for IP addresses.
4167 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4169 57. Double the size of the debug message buffer (to 2048) so that more of very
4170 long debug lines gets shown.
4172 58. The exicyclog utility now does better if the number of log files to keep
4173 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4175 59. Two changes related to the smtp_active_hostname option:
4177 (1) $smtp_active_hostname is now available as a variable.
4178 (2) The default for smtp_banner uses $smtp_active_hostname instead
4179 of $primary_hostname.
4181 60. The host_aton() function is supposed to be passed a string that is known
4182 to be a valid IP address. However, in the case of IPv6 addresses, it was
4183 not checking this. This is a hostage to fortune. Exim now panics and dies
4184 if the condition is not met. A case was found where this could be provoked
4185 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4186 components; fortuitously, this particular loophole had already been fixed
4187 by change 4.50/55 above.
4189 If there are any other similar loopholes, the new check in host_aton()
4190 itself should stop them being exploited. The report I received stated that
4191 data on the command line could provoke the exploit when Exim was running as
4192 exim, but did not say which command line option was involved. All I could
4193 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4194 running as the user.
4197 61. There was a buffer overflow vulnerability in the SPA authentication code
4198 (which came originally from the Samba project). I have added a test to the
4199 spa_base64_to_bits() function which I hope fixes it.
4202 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4203 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4205 63. The daemon start-up calls getloadavg() while still root for those OS that
4206 need the first call to be done as root, but it missed one case: when
4207 deliver_queue_load_max is set with deliver_drop_privilege. This is
4208 necessary for the benefit of the queue runner, because there is no re-exec
4209 when deliver_drop_privilege is set.
4211 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4212 This has been fixed.
4214 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4215 was in use, was not putting the data itself into the right store pool;
4216 consequently, it could be overwritten for a subsequent message in the same
4217 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4220 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4222 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4223 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4224 the function were treating the return as a boolean value, which happened to
4225 work because 0=false and not-0=true, but is not correct code.
4227 68. The host_aton() function was not handling scoped IPv6 addresses (those
4228 with, for example, "%eth0" on the end) correctly.
4230 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4231 negated items (that is, ~something) in unsigned ints. Some compilers
4232 apparently mutter when there is no cast.
4234 70. If an address verification called from an ACL failed, and did not produce a
4235 user-specific message (i.e. there was only a "system" message), nothing was
4236 put in $acl_verify_message. In this situation, it now puts the system
4239 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4240 session; change 4.31/43 added the unwanted input to the log line - except
4241 that it did not do this in the start of session case. It now does.
4243 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4244 This is wrong and can cause the other end to generate a synchronization
4245 error if it is another Exim or anything else that does the synchronization
4246 check. A QUIT command is no longer sent after a timeout.
4248 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4249 during host lookups.
4251 74. The defer_ok option of callout verification was not working if it was used
4252 when verifying addresses in header lines, that is, for this case:
4254 verify = header_sender/callout=defer_ok
4256 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4257 those file descriptors could be used for SMTP connections. If anything
4258 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4259 could be sent to the SMTP client, causing chaos. The daemon now opens
4260 stdin, stdout, and stderr to /dev/null when it puts itself into the
4263 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4264 log by default. The user can override this with suitable Perl magic.
4266 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4267 the log message when discard triggers, was not working for the DATA ACL or
4268 for the non-SMTP ACL.
4270 78. Error message wording change in sieve.c.
4272 79. If smtp_accept_max_per_host was set, the number of connections could be
4273 restricted to fewer than expected, because the daemon was trying to set up
4274 a new connection before checking whether the processes handling previous
4275 connections had finished. The check for completed processes is now done
4276 earlier. On busy systems, this bug wouldn't be noticed because something
4277 else would have woken the daemon, and it would have reaped the completed
4280 80. If a message was submitted locally by a user whose login name contained one
4281 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4282 It caused a spool format error. I have fixed the spool reading code. A
4283 related problem was that the "from" clause in the Received: line became
4284 illegal because of the space(s). It is now covered by ${quote_local_part.
4286 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4289 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4291 83. Patch from Sieve maintainer to fix unterminated string problem in
4292 "vacation" handling.
4294 84. Some minor changes to the Linux configuration files to help with other
4295 OS variants using glibc.
4297 85. One more patch for Sieve to update vacation handling to latest spec.
4300 ----------------------------------------------------
4301 See the note above about the 4.44 and 4.50 releases.
4302 ----------------------------------------------------
4308 1. Change 4.43/35 introduced a bug that caused file counts to be
4309 incorrectly computed when quota_filecount was set in an appendfile
4312 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4313 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4316 3. Give more explanation in the error message when the command for a transport
4317 filter fails to execute.
4319 4. There are several places where Exim runs a non-Exim command in a
4320 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4321 was being done only for the command run by the queryprogram router. It is
4322 now done for all such subprocesses. The other cases are: ${run, transport
4323 filters, and the commands run by the lmtp and pipe transports.
4325 5. Some older OS have a limit of 256 on the maximum number of file
4326 descriptors. Exim was using setrlimit() to set 1000 as a large value
4327 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4328 systems. I've change it so that if it can't get 1000, it tries for 256.
4330 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4331 was an oversight, and furthermore, ever since the addition of extra
4332 controls (e.g. 4.43/32), the checks on when to allow different forms of
4333 "control" were broken. There should now be diagnostics for all cases when a
4334 control that does not make sense is encountered.
4336 7. $recipients is now available in the predata ACL (oversight).
4338 8. Tidy the search cache before the fork to do a delivery from a message
4339 received from the command line. Otherwise the child will trigger a lookup
4340 failure and thereby defer the delivery if it tries to use (for example) a
4341 cached ldap connection that the parent has called unbind on.
4343 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4344 of $address_data from the recipient verification was clobbered by the
4345 sender verification.
4347 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4348 was its contents. (It was OK if the option was not defined at all.)
4350 11. A "Completed" log line is now written for messages that are removed from
4351 the spool by the -Mrm option.
4353 12. $host_address is now set to the target address during the checking of
4354 ignore_target_hosts.
4356 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4357 being passed; this would have caused $sender_host_name to have been used if
4358 matching the list had actually called for a host name (not very likely,
4359 since this list is usually IP addresses). A host name is now passed as
4362 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4363 code that specifies a non-restarting handler (typically sigaction() in
4364 modern systems) in an attempt to fix a rare and obscure crash bug.
4366 15. Narrowed the window for a race in the daemon that could cause it to ignore
4367 SIGCHLD signals. This is not a major problem, because they are used only to
4368 wake it up if nothing else does.
4370 16. A malformed maildirsize file could cause Exim to calculate negative values
4371 for the mailbox size or file count. Odd effects could occur as a result.
4372 The maildirsize information is now recalculated if the size or filecount
4375 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4376 support for a long time. Removed HAVE_SYS_VFS_H.
4378 18. Updated exipick to current release from John Jetmore.
4380 19. Allow an empty sender to be matched against a lookup in an address list.
4381 Previously the only cases considered were a regular expression, or an
4384 20. Exim went into a mad DNS lookup loop when doing a callout where the
4385 host was specified on the transport, if the DNS lookup yielded more than
4388 21. The RFC2047 encoding function was originally intended for short strings
4389 such as real names; it was not keeping to the 75-character limit for
4390 encoded words that the RFC imposes. It now respects the limit, and
4391 generates multiple encoded words if necessary. To be on the safe side, I
4392 have increased the buffer size for the ${rfc2047: expansion operator from
4395 22. Failure to deliver a bounce message always caused it to be frozen, even if
4396 there was an errors_to setting on the router. The errors_to setting is now
4399 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4400 canonical form (fully expanded) before being placed in
4401 $sender_host_address.
4403 24. Updated eximstats to version 1.33
4405 25. Include certificate and key file names in error message when GnuTLS fails
4406 to set them up, because the GnuTLS error message doesn't include the name
4407 of the failing file when there is a problem reading it.
4409 26. Expand error message when OpenSSL has problems setting up cert/key files.
4412 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4413 (this can affect the format of dates).
4415 28. exim_tidydb, when checking for the continued existence of a message for
4416 which it has found a message-specific retry record, was not finding
4417 messages that were in split spool directories. Consequently, it was
4418 deleting retry records that should have stayed in existence.
4420 29. eximstats updated to version 1.35
4421 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4422 1.35 - bugfix such that pie charts by volume are generated correctly
4424 30. The SPA authentication driver was not abandoning authentication and moving
4425 on to the next authenticator when an expansion was forced to fail,
4426 contradicting the general specification for all authenticators. Instead it
4427 was generating a temporary error. It now behaves as specified.
4429 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4430 (the order specifies the preference for clients). The order is now AES256,
4431 AES128, 3DES, ARCFOUR128.
4433 31. Small patch to Sieve code - explicitly set From: when generating an
4436 32. Exim crashed if a remote delivery caused a very long error message to be
4437 recorded - for instance if somebody sent an entire SpamAssassin report back
4438 as a large number of 550 error lines. This bug was coincidentally fixed by
4439 increasing the size of one of Exim's internal buffers (big_buffer) that
4440 happened as part of the Exiscan merge. However, to be on the safe side, I
4441 have made the code more robust (and fixed the comments that describe what
4444 33. Some experimental protocols are using DNS PTR records for new purposes. The
4445 keys for these records are domain names, not reversed IP addresses. The
4446 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4447 leaves it alone. Component reversal etc. now happens only for IP addresses.
4450 34. The host_aton() function is supposed to be passed a string that is known
4451 to be a valid IP address. However, in the case of IPv6 addresses, it was
4452 not checking this. This is a hostage to fortune. Exim now panics and dies
4453 if the condition is not met. A case was found where this could be provoked
4454 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4455 components; fortuitously, this particular loophole had already been fixed
4456 by change 4.50/55 or 4.44/33 above.
4458 If there are any other similar loopholes, the new check in host_aton()
4459 itself should stop them being exploited. The report I received stated that
4460 data on the command line could provoke the exploit when Exim was running as
4461 exim, but did not say which command line option was involved. All I could
4462 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4463 running as the user.
4466 35. There was a buffer overflow vulnerability in the SPA authentication code
4467 (which came originally from the Samba project). I have added a test to the
4468 spa_base64_to_bits() function which I hope fixes it.
4471 36. The daemon start-up calls getloadavg() while still root for those OS that
4472 need the first call to be done as root, but it missed one case: when
4473 deliver_queue_load_max is set with deliver_drop_privilege. This is
4474 necessary for the benefit of the queue runner, because there is no re-exec
4475 when deliver_drop_privilege is set.
4477 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4478 was in use, was not putting the data itself into the right store pool;
4479 consequently, it could be overwritten for a subsequent message in the same
4480 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4483 38. Sometimes the final signoff response after QUIT could fail to get
4484 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4485 < 0 before doing a fflush(). This bug looks as though it goes back to the
4486 introduction of TLS in release 3.20, but "sometimes" must have been rare
4487 because the tests only now provoked it.
4493 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4494 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4495 Now it returns an integer. A number of calls were still expecting a T/F
4496 return. Fortuitously, in all cases, the tests worked in OK situations,
4497 which is the norm. However, things would have gone wrong on any write
4498 failures on the smtp file descriptor. This function is used when sending
4499 messages over SMTP and also when doing verify callouts.
4501 2. When Exim is called to do synchronous delivery of a locally submitted
4502 message (the -odf or -odi options), it no longer closes stderr before doing
4505 3. Implemented the mua_wrapper option.
4507 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4509 5. Implemented the functions header_remove(), header_testname(),
4510 header_add_at_position(), and receive_remove_recipient(), and exported them
4513 6. If an ACL "warn" statement specified the addition of headers, Exim already
4514 inserted X-ACL-Warn: at the start if there was no header name. However, it
4515 was not making this test for the second and subsequent header lines if
4516 there were newlines in the string. This meant that an invalid header could
4517 be inserted if Exim was badly configured.
4519 7. Allow an ACL "warn" statement to add header lines at the start or after all
4520 the Received: headers, as well as at the end.
4522 8. Added the rcpt_4xx retry error code.
4524 9. Added postmaster_mailfrom=xxx to callout verification option.
4526 10. Added mailfrom=xxxx to the callout verification option, for verify=
4529 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4530 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4532 12. Inserted some casts to stop certain compilers warning when using pointer
4533 differences as field lengths or precisions in printf-type calls (mostly
4534 affecting debugging statements).
4536 13. Added optional readline() support for -be (dynamically loaded).
4538 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4539 same clock tick as a message's arrival, so that its received time was the
4540 same as the "first fail" time on the retry record, and that message
4541 remained on the queue past the ultimate address timeout, every queue runner
4542 would try a delivery (because it was past the ultimate address timeout) but
4543 after another failure, the ultimate address timeout, which should have then
4544 bounced the address, did not kick in. This was a "< instead of <=" error;
4545 in most cases the first failure would have been in the next clock tick
4546 after the received time, and all would be well.
4548 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4549 being recognized when the domain list was tested by the match_domain
4550 condition in an expansion string.
4552 16. Added the ${str2b64: operator.
4554 17. Exim was always calling setrlimit() to set a large limit for the number of
4555 processes, without checking whether the existing limit was already
4556 adequate. (It did check for the limit on file descriptors.) Furthermore,
4557 errors from getrlimit() and setrlimit() were being ignored. Now they are
4558 logged to the main and panic logs, but Exim does carry on, to try to do its
4559 job under whatever limits there are.
4561 18. Imported PCRE 5.0.
4563 19. Trivial typo in log message " temporarily refused connection" (the leading
4566 20. If the log selector return_path_on_delivery was set and an address was
4567 redirected to /dev/null, the delivery process crashed because it assumed
4568 that a return path would always be set for a "successful" delivery. In this
4569 case, the whole delivery is bypassed as an optimization, and therefore no
4572 21. Internal re-arrangement: the function for sending a challenge and reading
4573 a response while authentication was assuming a zero-terminated challenge
4574 string. It's now changed to take a pointer and a length, to allow for
4575 binary data in such strings.
4577 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4579 23. Exim was not respecting finduser_retries when seeking the login of the
4580 uid under which it was called; it was always trying 10 times. (The default
4581 setting of finduser_retries is zero.) Also, it was sleeping after the final
4582 failure, which is pointless.
4584 24. Implemented tls_on_connect_ports.
4586 25. Implemented acl_smtp_predata.
4588 26. If the domain in control=submission is set empty, Exim assumes that the
4589 authenticated id is a complete email address when it generates From: or
4590 Sender: header lines.
4592 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4593 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4594 chown and chgrp in /bin and hostname in /usr/bin.
4596 28. Exim was keeping the "process log" file open after each use, just as it
4597 does for the main log. This opens the possibility of it remaining open for
4598 long periods when the USR1 signal hits a daemon. Occasional processlog
4599 errors were reported, that could have been caused by this. Anyway, it seems
4600 much more sensible not to leave this file open at all, so that is what now
4603 29. The long-running daemon process does not normally write to the log once it
4604 has entered its main loop, and it closes the log before doing so. This is
4605 so that log files can straightforwardly be renamed and moved. However,
4606 there are a couple of unusual error situations where the daemon does write
4607 log entries, and I had neglected to close the log afterwards.
4609 30. The text of an SMTP error response that was received during a remote
4610 delivery was being truncated at 512 bytes. This is too short for some of
4611 the long messages that one sometimes sees. I've increased the limit to
4614 31. It is now possible to make retry rules that apply only when a message has a
4615 specific sender, in particular, an empty sender.
4617 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4618 it possible to be selective about when SMTP synchronization is enforced.
4620 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4622 32. Implemented hosts_connection_nolog.
4624 33. Added an ACL for QUIT.
4626 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4629 35. Added mailbox_size and mailbox_filecount to appendfile.
4631 36. Added control = no_multiline_responses to ACLs.
4633 37. There was a bug in the logic of the code that waits for the clock to tick
4634 in the case where the clock went backwards by a substantial amount such
4635 that the microsecond fraction of "now" was more than the microsecond
4636 fraction of "then" (but the whole seconds number was less).
4638 38. Added support for the libradius Radius client library this is found on
4639 FreeBSD (previously only the radiusclient library was supported).
4645 1. When certain lookups returned multiple values in the form name=value, the
4646 quoting of the values was not always being done properly. Specifically:
4647 (a) If the value started with a double quote, but contained no whitespace,
4649 (b) If the value contained whitespace other than a space character (i.e.
4650 tabs or newlines or carriage returns) it was not quoted.
4651 This fix has been applied to the mysql and pgsql lookups by writing a
4652 separate quoting function and calling it from the lookup code. The fix
4653 should probably also be applied to nisplus, ibase and oracle lookups, but
4654 since I cannot test any of those, I have not disturbed their existing code.
4656 2. A hit in the callout cache for a specific address caused a log line with no
4657 reason for rejecting RCPT. Now it says "Previous (cached) callout
4658 verification failure".
4660 3. There was an off-by-one bug in the queryprogram router. An over-long
4661 return line was truncated at 256 instead of 255 characters, thereby
4662 overflowing its buffer with the terminating zero. As well as fixing this, I
4663 have increased the buffer size to 1024 (and made a note to document this).
4665 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4666 when Exim is waiting for an SMTP response from a remote server, Exim
4667 restarts its select() call on the socket, thereby resetting its timeout.
4668 This is not a problem when such interrupts are rare. Somebody set up a cron
4669 job to run exiwhat every 2 minutes, which is less than the normal select()
4670 timeout (5 or 10 minutes). This meant that the select() timeout never
4671 kicked in because it was always reset. I have fixed this by comparing the
4672 time when an interrupt arrives with the time at the start of the first call
4673 to select(). If more time than the timeout has elapsed, the interrupt is
4674 treated as a timeout.
4676 5. Some internal re-factoring in preparation for the addition of Sieve
4677 extensions (by MH). In particular, the "personal" test is moved to a
4678 separate function, and given an option for scanning Cc: and Bcc: (which is
4679 not set for Exim filters).
4681 6. When Exim created an email address using the login of the caller as the
4682 local part (e.g. when creating a From: or Sender: header line), it was not
4683 quoting the local part when it contained special characters such as @.
4685 7. Installed new OpenBSD configuration files.
4687 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4688 try to make them clearer.
4690 9. Callout options, other than the timeout value, were being ignored when
4691 verifying sender addresses in header lines. For example, when using
4693 verify = header_sender/callout=no_cache
4695 the cache was (incorrectly) being used.
4697 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4698 only the Cygwin environment.
4700 11. When return_path_on_delivery was set as a log selector, if different remote
4701 addresses in the same message used different return paths and parallel
4702 remote delivery occurred, the wrong values would sometimes be logged.
4703 (Whenever a remote delivery process finished, the return path value from
4704 the most recently started remote delivery process was logged.)
4706 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4707 lines when AUTH and/or TLS are in use. This is the "received protocol"
4708 field. Exim used to use "asmtp" for authenticated SMTP, without any
4709 indication (in the protocol name) for TLS use. Now it follows the RFC and
4710 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4711 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4712 names appear in log lines as well as in Received: header lines.
4714 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4715 extensions, and comparison tests, and to fix some bugs.
4717 14. Changes to the "personal" filter test:
4719 (1) The test was buggy in that it was just doing the equivalent of
4720 "contains" tests on header lines. For example, if a user's address was
4721 anne@some.where, the "personal" test would incorrectly be true for
4723 To: susanne@some.where
4725 This test is now done by extracting each address from the header in turn,
4726 and checking the entire address. Other tests that are part of "personal"
4727 are now done using regular expressions (for example, to check local parts
4728 of addresses in From: header lines).
4730 (2) The list of non-personal local parts in From: addresses has been
4731 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4732 taken from the Sieve specification recommendations.
4734 (3) If the message contains any header line starting with "List-" it is
4735 treated as non-personal.
4737 (4) The test for "circular" in the Subject: header line has been removed
4738 because it now seems ill-conceived.
4740 15. Minor typos in src/EDITME comments corrected.
4742 16. Installed latest exipick from John Jetmore.
4744 17. If headers_add on a router specified a text string that was too long for
4745 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4746 of string_sprintf() is now avoided.
4748 18. $message_body_size was not set (it was always zero) when running the DATA
4749 ACL and the local_scan() function.
4751 19. For the "mail" command in an Exim filter, no default was being set for
4752 the once_repeat time, causing a random time value to be used if "once" was
4753 specified. (If the value happened to be <= 0, no repeat happened.) The
4754 default is now 0s, meaning "never repeat". The "vacation" command was OK
4755 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4756 (I found it when inspecting the code).
4758 20. There is now an overall timeout for performing a callout verification. It
4759 defaults to 4 times the callout timeout, which applies to individual SMTP
4760 commands during the callout. The overall timeout applies when there is more
4761 than one host that can be tried. The timeout is checked before trying the
4762 next host. This prevents very long delays if there are a large number of
4763 hosts and all are timing out (e.g. when the network connections are timing
4764 out). The value of the overall timeout can be changed by specifying an
4765 additional sub-option for "callout", called "maxwait". For example:
4767 verify = sender/callout=5s,maxwait=20s
4769 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4770 to the end before writing, but this should make it even safer).
4772 22. Exim was forgetting that it had advertised PIPELINING for the second and
4773 subsequent messages on an SMTP connection. It was also not resetting its
4774 memory on STARTTLS and an internal HELO.
4776 23. When Exim logs an SMTP synchronization error within a session, it now
4777 records whether PIPELINING has been advertised or not.
4779 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4780 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4781 rather than long int.
4783 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4785 26. Added the never_mail option to autoreply.
4791 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4792 crash if the getsockname() call failed; this can happen if a connection is
4793 closed very soon after it is established. The problem was simply in the
4794 order in which certain operations were done, causing Exim to try to write
4795 to the SMTP stream before it had set up the file descriptor. The bug has
4796 been fixed by making things happen in the correct order.
4802 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4803 before the connection was closed, thus losing the rejection response.
4805 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4806 some early Solaris releases, but causes trouble in current releases where
4807 socklen_t is defined.
4809 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4812 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4815 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4816 that releasing the top of it at the end releases what was used for sub-
4817 expansions (unless the block got too big). However, discard this block if
4818 the first thing is a variable or header, so that we can use its block when
4819 it is dynamic (useful for very large $message_headers, for example).
4821 6. Lookups now cache *every* query, not just the most recent. A new, separate
4822 store pool is used for this. It can be recovered when all lookup caches are
4823 flushed. Lookups now release memory at the end of their result strings.
4824 This has involved some general refactoring of the lookup sources.
4826 7. Some code has been added to the store_xxx() functions to reduce the amount
4827 of flapping under certain conditions.
4829 8. log_incoming_interface used to affect only the <= reception log lines. Now
4830 it causes the local interface and port to be added to several more SMTP log
4831 lines, for example "SMTP connection from", and rejection lines.
4833 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4835 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4837 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4838 could be overwritten at the end of the current message (or the start of a
4839 new message if it was set in a HELO ACL). The value is now preserved for
4840 the duration of the SMTP connection.
4842 12. If a transport had a headers_rewrite setting, and a matching header line
4843 contained an unqualified address, that address was qualified, even if it
4844 did not match any rewriting rules. The underlying bug was that the values
4845 of the flags that permit the existence of unqualified sender and recipient
4846 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4847 non-local messages, and by -bnq for local messages) were not being
4848 preserved with the message after it was received.
4850 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4851 "next input=" as part of the text comprising the host identity instead of
4852 the correct text. The code was using the same buffer for two different
4853 strings. However, depending on which order the printing function evaluated
4854 its arguments, the bug did not always show up. Under Linux, for example, my
4855 test suite worked just fine.
4857 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4858 doesn't work with some older versions of Perl. It has been changed to "my",
4859 which in any case is probably the better facility to use.
4861 15. A really picky compiler found some instances of statements for creating
4862 error messages that either had too many or two few arguments for the format
4865 16. The size of the buffer for calls to the DNS resolver has been increased
4866 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4867 for addresses that have a lot of PTR records. This alleviates a problem; it
4868 does not fully solve it.
4870 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4871 buffer now truncates the list and logs the incident, which is the same
4872 action as happens when Exim is looking up a host name and its aliases.
4873 Previously in this situation something unpredictable would happen;
4874 sometimes it was "internal error: store_reset failed".
4876 18. If a server dropped the connection unexpectedly when an Exim client was
4877 using GnuTLS and trying to read a response, the client delivery process
4878 crashed while trying to generate an error log message.
4880 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4881 string, for example:
4883 warn message = H1: something\nH2: something
4885 the text was added as a single header line from Exim's point of view
4886 though it ended up OK in the delivered message. However, searching for the
4887 second and subsequent header lines using $h_h2: did not work. This has been
4888 fixed. Similarly, if a system filter added multiple headers in this way,
4889 the routers could not see them.
4891 20. Expanded the error message when iplsearch is called with an invalid key to
4892 suggest using net-iplsearch in a host list.
4894 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4895 ACLs are no longer actually imposed (and a message to that effect is
4898 22. If a "gecos" field in a passwd entry contained escaped characters, in
4899 particular, if it contained a \" sequence, Exim got it wrong when building
4900 a From: or a Sender: header from that name. A second bug also caused
4901 incorrect handling when an unquoted " was present following a character
4902 that needed quoting.
4904 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4905 was not being matched caselessly.
4907 24. Arranged for all hyphens in the exim.8 source to be escaped with
4910 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4911 properly. Recipient callout cache records were still being keyed to include
4912 the sender, even when use_sender was set false. This led to far more
4913 callouts that were necessary. The sender is no longer included in the key
4914 when use_sender is false.
4916 26. Added "control = submission" modifier to ACLs.
4918 27. Added the ${base62d: operator to decode base 62 numbers.
4920 28. dnsdb lookups can now access SRV records.
4922 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4923 the configuration file.
4925 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4926 (-v) mode. This makes the output for a verbose queue run more intelligible.
4928 31. Added a use_postmaster feature to recipient callouts.
4930 32. Added the $body_zerocount variable, containing the number of binary zero
4931 bytes in the message body.
4933 33. The time of last modification of the "new" subdirectory is now used as the
4934 "mailbox time last read" when there is a quota error for a maildir
4937 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4939 35. Added +ignore_unknown as a special item in host lists.
4941 36. Code for decoding IPv6 addresses in host lists is now included, even if
4942 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4943 address was recognized as an IP address, but was then not correctly decoded
4944 into binary, causing unexpected and incorrect effects when compared with
4951 1. Very minor rewording of debugging text in manualroute to say "list of
4952 hosts" instead of "hostlist".
4954 2. If verify=header_syntax was set, and a header line with an unqualified
4955 address (no domain) and a large number of spaces between the end of the
4956 name and the colon was received, the reception process suffered a buffer
4957 overflow, and (when I tested it) crashed. This was caused by some obsolete
4958 code that should have been removed. The fix is to remove it!
4960 3. When running in the test harness, delay a bit after writing a bounce
4961 message to get a bit more predictability in the log output.
4963 4. Added a call to search_tidyup() just before forking a reception process. In
4964 theory, someone could use a lookup in the expansion of smtp_accept_max_
4965 per_host which, without the tidyup, could leave open a database connection.
4967 5. Added the variables $recipient_data and $sender_data which get set from a
4968 lookup success in an ACL "recipients" or "senders" condition, or a router
4969 "senders" option, similar to $domain_data and $local_part_data.
4971 6. Moved the writing of debug_print from before to after the "senders" test
4974 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4975 problems for message scanning, either using a data ACL, or using
4976 local_scan() because the Received: header was not generated till after they
4977 were called (in order to set the time as the time of reception completion).
4978 I have revised the way this works. The header is now generated after the
4979 body is received, but before the ACL or local_scan() are called. After they
4980 are run, the timestamp in the header is updated.
4986 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4987 before starting a queue runner without re-exec. This happened only when
4988 deliver_drop_privilege was set or when the Exim user was set to root. The
4989 effect of the bug was that timeouts during subsequent deliveries caused
4990 crashes instead of being properly handled. The handler is now left at its
4991 default (and expected) setting.
4993 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4994 message, again when deliver_drop_privilege is set or Exim is run as root.
4995 The bug described in (1) was not present in this case, but the tidying up
4996 of the other signals was missing. I have made the two cases consistent.
4998 3. The ignore_target_hosts setting on a manualroute router was being ignored
4999 for hosts that were looked up using the /MX notation.
5001 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5004 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5005 operated on the sender address. After changing the $sender_address to <>
5006 for the sender address verify, Exim was re-instated it as the original
5007 (before rewriting) address, but remembering that it had rewritten it, so it
5008 wasn't rewriting it again. This bug also had the effect of breaking the
5009 sender address verification caching when the sender address was rewritten.
5011 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5012 This has been changed so that if the ip literal address matches
5013 ignore_target_hosts, the router declines.
5015 7. Added expansion conditions match_domain, match_address, and match_local_
5016 part (NOT match_host).
5018 8. The placeholder for the Received: header didn't have a length field set.
5020 9. Added code to Exim itself and to exim_lock to test for a specific race
5021 condition that could lead to file corruption when using MBX delivery. The
5022 issue is with the lockfile that is created in /tmp. If this file is removed
5023 after a process has opened it but before that process has acquired a lock,
5024 there is the potential for a second process to recreate the file and also
5025 acquire a lock. This could lead to two Exim processes writing to the file
5026 at the same time. The added code performs the same test as UW imapd; it
5027 checks after acquiring the lock that its file descriptor still refers to
5028 the same named file.
5030 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5031 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5034 11. Added the smtp_active_hostname option. If used, this will typically be made
5035 to depend on the incoming interface address. Because $interface_address is
5036 not set up until the daemon has forked a reception process, error responses
5037 that can happen earlier (such as "too many connections") no longer contain
5040 12. If an expansion in a condition on a "warn" statement fails because a lookup
5041 defers, the "warn" statement is abandoned, and the next ACL statement is
5042 processed. Previously this caused the whole ACL to be aborted.
5044 13. Added the iplsearch lookup type.
5046 14. Added ident_timeout as a log selector.
5048 15. Added tls_certificate_verified as a log selector.
5050 16. Added a global option tls_require_ciphers (compare the smtp transport
5051 option of the same name). This controls incoming TLS connections.
5053 17. I finally figured out how to make tls_require_ciphers do a similar thing
5054 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5055 before starting the TLS session.
5057 18. Tabs are now shown as \t in -bP output.
5059 19. If the log selector return_path_on_delivery was set, Exim crashed when
5060 bouncing a message because it had too many Received: header lines.
5062 20. If two routers both had headers_remove settings, and the first one included
5063 a superfluous trailing colon, the final name in the first list and the
5064 first name in the second list were incorrectly joined into one item (with a
5065 colon in the middle).
5071 1. Added -C and -D options to the exinext utility, mainly to make it easier
5072 to include in the automated testing, but these could be helpful when
5073 multiple configurations are in use.
5075 2. The exinext utility was not formatting the output nicely when there was
5076 an alternate port involved in the retry record key, nor when there was a
5077 message id as well (for retries that were specific to a specific message
5078 and a specific host). It was also confused by IPv6 addresses, because of
5079 the additional colons they contain. I have fixed the IPv4 problem, and
5080 patched it up to do a reasonable job for IPv6.
5082 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5083 delivery, the log line now contains "pipelined" if PIPELINING was used.
5085 4. An SMTP transport process used to panic and die if the bind() call to set
5086 an explicit outgoing interface failed. This has been changed; it is now
5087 treated in the same way as a connect() failure.
5089 5. A reference to $sender_host_name in the part of a conditional expansion
5090 that was being skipped was still causing a DNS lookup. This no longer
5093 6. The def: expansion condition was not recognizing references to header lines
5094 that used bh_ and bheader_.
5096 7. Added the _cache feature to named lists.
5098 8. The code for checking quota_filecount in the appendfile transport was
5099 allowing one more file than it should have been.
5101 9. For compatibility with Sendmail, the command line option
5109 and sets the incoming protocol and host name (for trusted callers). The
5110 host name and its colon can be omitted when only the protocol is to be set.
5111 Note the Exim already has two private options, -pd and -ps, that refer to
5112 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5113 "s", but I don't think that's a major issue.
5115 10. A number of refactoring changes to the code, none of which should affect
5118 (a) The number of logging options was getting close to filling up the
5119 32-bit word that was used as a bit map. I have split them into two classes:
5120 those that are passed in the argument to log_write(), and those that are
5121 only ever tested independently outside of that function. These are now in
5122 separate 32-bit words, so there is plenty of room for expansion again.
5123 There is no change in the user interface or the logging behaviour.
5125 (b) When building, for example, log lines, the code previously used a
5126 macro that called string_cat() twice, in order to add two strings. This is
5127 not really sufficiently general. Furthermore, there was one instance where
5128 it was actually wrong because one of the argument was used twice, and in
5129 one call a function was used. (As it happened, calling the function twice
5130 did not affect the overall behaviour.) The macro has been replaced by a
5131 function that can join an arbitrary number of extra strings onto a growing
5134 (c) The code for expansion conditions now uses a table and a binary chop
5135 instead of a serial search (which was left over from when there were very
5136 few conditions). Also, it now recognizes conditions like "pam" even when
5137 the relevant support is not compiled in: a suitably worded error message is
5138 given if an attempt is made to use such a condition.
5140 11. Added ${time_interval:xxxxx}.
5142 12. A bug was causing one of the ddress fields not to be passed back correctly
5143 from remote delivery subprocesses. The field in question was not being
5144 subsequently used, so this caused to problems in practice.
5146 13. Added new log selectors queue_time and deliver_time.
5148 14. Might have fixed a bug in maildirsizefile handling that threw up
5149 "unexpected character" debug warnings, and recalculated the data
5150 unnecessarily. In any case, I expanded the warning message to give more
5153 15. Added the message "Restricted characters in address" to the statements in
5154 the default ACL that block characters like @ and % in local parts.
5156 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5157 Three changes have been made:
5159 (a) There was a serious bug; a negative response to MAIL caused the whole
5160 recipient domain to be cached as invalid, thereby blocking all messages
5161 to all local parts at the same domain, from all senders. This bug has
5162 been fixed. The domain is no longer cached after a negative response to
5163 MAIL if the sender used is not empty.
5165 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5168 (c) A new callout option, "use_sender" has been added for people who want
5169 the modified behaviour.
5175 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5178 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5179 indeed breaks things for older releases.
5181 3. Added additional logging to the case where there is a problem reading data
5182 from a filter that is running in a subprocess using a pipe, in order to
5183 try to track down a specific problem.
5185 4. Testing facility fudge: when running in the test harness and attempting
5186 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5187 getting "No route to host". Convert this to a timeout.
5189 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5192 6. Some OS don't have socklen_t but use size_t instead. This affects the
5193 fifth argument of getsockopt() amongst other things. This is now
5194 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5195 can be set for individual OS. I have set it for SunOS5, OSF1, and
5196 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5197 some earlier ones do not.
5199 7. Change 4.30/15 was not doing the test caselessly.
5201 8. The standard form for an IPv6 address literal was being rejected by address
5202 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5203 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5204 this, as well as the form without the "IPv6" on the front (but only when
5205 address literals are enabled, of course).
5207 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5209 10. Exim crashed if a message with an empty sender address specified by -f
5210 encountered a router with an errors_to setting. This could be provoked only
5211 by a command such as
5215 where an empty string was supplied; "<>" did not hit this bug.
5217 11. Installed PCRE release 4.5.
5219 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5220 remained set. It is now erased.
5222 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5223 times from message ids (which are base 36 rather than the normal 62).
5225 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5226 were being counted as actual protocol errors, and logged if the log
5227 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5228 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5229 and DATA following a set of rejected RCPTs do not count as protocol errors.
5230 In other words, Exim assumes they were pipelined, though this may not
5231 actually be the case. Of course, in all cases the client gets an
5232 appropriate error code.
5234 15. If a lookup fails in an ACL condition, a message about the failure may
5235 be available; it is used if testing the ACL cannot continue, because most
5236 such messages specify what the cause of the deferral is. However, some
5237 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5238 that caused an old message to be retained and used if a later statement
5239 caused a defer, replacing the real cause of the deferral.
5241 16. If an IP address had so many PTR records that the DNS lookup buffer
5242 was not large enough to hold them, Exim could crash while trying to process
5243 the truncated data. It now detects and logs this case.
5245 17. Further to 4.21/58, another change has been made: if (and only if) the
5246 first line of a message (the first header line) ends with CRLF, a bare LF
5247 in a subsequent header line has a space inserted after it, so as not to
5248 terminate the header.
5250 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5251 unnecessarily, used atoi() instead of strtol(), and didn't check the
5252 termination when getting file sizes from file names by regex.
5254 19. Completely re-implemented the support for maildirsize files, in the light
5255 of a number of problems with the previous contributed implementation
5256 (4.30/29). In particular:
5258 . If the quota is zero, the maildirsize file is maintained, but no quota is
5261 . If the maildir directory does not exist, it is created before any attempt
5262 to write a maildirsize file.
5264 . The quota value in the file is just a cache; if the quota is changed in
5265 the transport, the new value overrides.
5267 . A regular expression is available for excluding directories from the
5270 20. The autoreply transport checks the characters in options that define the
5271 message's headers; it allows continued headers, but it was checking with
5272 isspace() after an embedded newline instead of explicitly looking for a
5275 21. If all the "regular" hosts to which an address was routed had passed their
5276 expiry times, and had not reached their retry times, the address was
5277 bounced, even if fallback hosts were defined. Now Exim should go on to try
5280 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5281 equivalent code in the SMTP transport. Some hosts send humungous responses
5282 to HELO/EHLO, more than 1024 it seems.
5284 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5285 gives compiler warnings in some environments. I've now done it "properly",
5288 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5289 (because of problems with the built-in one) was declared to return uschar *
5290 instead of char *, causing compiler failure.
5292 25. Fixed a file descriptor leak when processing alias/forward files.
5294 26. Fixed a minor format string issue in dbfn.c.
5296 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5298 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5299 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5302 29. When a sender address is verified, it is cached, to save repeating the test
5303 when there is more than one recipient in a message. However, when the
5304 verification involves a callout, it is possible for different callout
5305 options to be set for different recipients. It is too complicated to keep
5306 track of this in the cache, so now Exim always runs a verification when a
5307 callout is required, relying on the callout cache for the optimization.
5308 The overhead is duplication of the address routing, but this should not be
5311 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5312 to be verified with callout=postmaster, and the main callout worked but the
5313 postmaster check failed, the verification correctly failed. However, if a
5314 subsequent RCPT command asked for sender verification *without* the
5315 postmaster check, incorrect caching caused this verification also to fail,
5318 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5319 it was not caching the DNS options (qualify_single, search_parents) that
5320 were used when the lookup failed. A subsequent lookup with different
5321 options therefore always gave the same answer, though there were cases
5322 where it should not have. (Example: a "domains = !$mx_any" option on a
5323 dnslookup router: the "domains" option is always processed without any
5324 widening, but the router might have qualify_single set.) Now Exim uses the
5325 cached value only when the same options are set.
5327 32. Added John Jetmore's "exipick" utility to the distribution.
5329 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5330 than a timeout (e.g. a certificate is required, and is not provided), an
5331 Exim server now closes the connection immediately. Previously it waited for
5332 the client to close - but if the client is SSL, it seems that they each
5333 wait for each other, leading to a delay before one of them times out.
5335 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5336 maintained 0.8.x compatibility because I don't think many are using it, and
5337 it is clearly obsolete.
5339 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5342 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5343 client certificate was expired. A simple patch fixes this, though I don't
5344 understand the full logic of why the verify callback is called multiple
5347 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5348 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5349 which causes problems with some clients (such as the Certicom SSL Plus
5350 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5351 disables the coutermeasure allowing Eudora to connect."
5353 38. Exim was not checking that a write() to a log file succeeded. This could
5354 lead to Bad Things if a log got too big, in particular if it hit a file
5355 size limit. Exim now panics and dies if it cannot write to a log file, just
5356 as it does if it cannot open a log file.
5358 39. Modified OS/Makefile-Linux so that it now contains
5360 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5362 The two -D definitions ensure that Exim is compiled with large file
5363 support, which makes it possible to handle log files that are bigger than
5366 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5367 instance) a domain was checked against a named list that involved a lookup,
5368 causing $domain_data to be set, then another domain was checked against the
5369 same list, then the first domain was re-checked, the value of $domain_data
5370 after the final check could be wrong. In particular, if the second check
5371 failed, it could be set empty. This bug probably also applied to
5374 41. The strip_trailing_dot option was not being applied to the address given
5375 with the -f command-line option.
5377 42. The code for reading a message's header from the spool was incrementing
5378 $received_count, but never initializing it. This meant that the value was
5379 incorrect (doubled) while delivering a message in the same process in which
5380 it was received. In the most common configuration of Exim, this never
5381 happens - a fresh exec is done - but it can happen when
5382 deliver_drop_privilege is set.
5384 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5385 it now includes up to 150 characters of the unexpected data in the log
5388 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5389 and building data strings. The size of both of these buffers was 10 000
5390 bytes - far larger than anybody would *ever* want, thought I. Needless to
5391 say, somebody hit the limit. I have increased the maximum line length to
5392 20 000 and the maximum data length of concatenated lines to 100 000. I have
5393 also fixed two bugs, because there was no checking on these buffers. Tsk,
5394 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5395 buffer is too small.
5397 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5398 lsearch lookups. Now it does.
5400 46. When parsing a route_list item in a manualroute router, a fixed-length
5401 buffer was used for the list of hosts. I made this 1024 bytes long,
5402 thinking that nobody would ever have a list of hosts that long. Wrong.
5403 Somebody had a whole pile of complicated expansion conditions, and the
5404 string was silently truncated, leading to an expansion error. It turns out
5405 that it is easier to change to an unlimited length (owing to other changes
5406 that have happened since this code was originally written) than to build
5407 structure for giving a limitation error. The length of the item that
5408 expands into the list of hosts is now unlimited.
5410 47. The lsearch lookup could not handle data where the length of text line was
5411 more than 4095 characters. Such lines were truncated, leading to shortened
5412 data being returned. It should now handle lines of any length.
5414 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5415 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5418 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5419 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5420 no longer happens in comment lines. A list of replacements is now placed
5421 at the head of all of the source files, except those whose only change is
5422 to replace PERL_COMMAND in the very first #! line.
5424 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5425 messages on the queue, with a bottom-up merge sort, using code contributed
5426 by Michael Haardt. This should make operations like -bp somewhat faster on
5427 large queues. It won't affect queue runners, except when queue_run_in_order
5430 51. Installed eximstats 1.31 in the distribution.
5432 52. Added support for SRV lookups to the dnslookup router.
5434 53. If an ACL referred to $message_body or $message_body_end, the value was not
5435 reset for any messages that followed in the same SMTP session.
5437 54. The store-handling optimization for building very long strings was not
5438 differentiating between the different store pools. I don't think this
5439 actually made any difference in practice, but I've tidied it.
5441 55. While running the routers to verify a sender address, $sender_address
5442 was still set to the sender address. This is wrong, because when routing to
5443 send a bounce to the sender, it would be empty. Therefore, I have changed
5444 it so that, while verifying a sender address, $sender_address is set to <>.
5445 (There is no change to what happens when verifying a recipient address.)
5447 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5448 target A or AAAA records (if not already returned) without resetting the
5449 qualify_single or search_parents options of the DNS resolver. These are
5450 inappropriate in this case because the targets of MX and SRV records must
5451 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5452 target that, when qualified, matched something in the local domain. These
5453 two options are now turned off when doing these lookups.
5455 57. It seems that at least some releases of Reiserfs (which does not have the
5456 concept of a fixed number of inodes) returns zero and not -1 for the
5457 number of available inodes. This interacted badly with check_spool_inodes,
5458 which assumed that -1 was the "no such thing" setting. What I have done is
5459 to check that the total number of inodes is greater than zero before doing
5460 the test of how many are available.
5462 58. When a "warn" ACL statement has a log_message modifier, the message is
5463 remembered, and not repeated. This is to avoid a lot of repetition when a
5464 message has many recipients that cause the same warning to be written.
5465 Howewer, Exim was preserving the list of already written lines for an
5466 entire SMTP session, which doesn't seem right. The memory is now reset if a
5467 new message is started.
5469 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5470 debugging output unless log_rewrite was also set.
5472 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5473 of (int)(handle) when we know that handle contains (void *)(-1).
5475 61. The Exim daemon panic-logs an error return when it closes the incoming
5476 connection. However "connection reset by peer" seems to be common, and
5477 isn't really an error worthy of noting specially, so that particular error
5480 62. When Exim is trying to find all the local interfaces, it used to panic and
5481 die if the ioctl to get the interface flags failed. However, it seems that
5482 on at least one OS (Solaris 9) it is possible to have an interface that is
5483 included in the list of interfaces, but for which you get a failure error
5484 for this call. This happens when the interface is not "plumbed" into a
5485 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5486 failure of the "get flags" call assumes that the interface is down.
5488 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5489 makes life easier for people who are doing arithmetic on fields extracted
5490 from dates, where you often get leading zeros that should not be
5491 interpreted as octal.
5493 64. Added qualify_domain to the redirect router, to override the global
5496 65. If a pathologically long header line contained very many addresses (the
5497 report of this problem mentioned 10 000) and each of them was rewritten,
5498 Exim could use up a very large amount of memory. (It kept on making new
5499 copies of the header line as it rewrote, and never released the old ones.)
5500 At the expense of a bit more processing, the header rewriting function has
5501 been changed so that it no longer eats memory in this way.
5503 66. The generation of the Received: header has been moved from the time that a
5504 message starts to be received, to the time that it finishes. The timestamp
5505 in the Received: header should now be very close to that of the <= log
5506 line. There are two side-effects of this change:
5508 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5509 logged header lines no longer include the local Received: line, because
5510 it has not yet been created. The same applies to a copy of the message
5511 that is returned to a non-SMTP sender when a message is rejected.
5513 (b) When a filter file is tested using -bf, no additional Received: header
5514 is added to the test message. After some thought, I decided that this
5517 This change does not affect the value of $received_for. It is still set
5518 after address rewriting, but before local_scan() is called.
5520 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5522 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5523 gave an unhelpful panic error message, and a defer error. I have managed to
5524 change this behaviour so that it now rejects any supplied certificate,
5525 which seems right, as the list of acceptable certificates is empty.
5527 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5528 gave an unhelpful defer error. I have not managed to make this reject any
5529 supplied certificates, but the error message it gives is "no certificate
5530 supplied", which is not helpful.
5532 70. exigrep's output now also includes lines that are not associated with any
5533 message, but which match the given pattern. Implemented by a patch from
5534 Martin Sluka, which also tidied up the Perl a bit.
5536 71. Recipient callout verification, like sender verification, was using <> in
5537 the MAIL FROM command. This isn't really the right thing, since the actual
5538 sender may affect whether the remote host accepts the recipient or not. I
5539 have changed it to use the actual sender in the callout; this means that
5540 the cache record is now keyed on a recipient/sender pair, not just the
5541 recipient address. There doesn't seem to be a real danger of callout loops,
5542 since a callout by the remote host to check the sender would use <>.
5543 [SEE ABOVE: changed after hitting problems.]
5545 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5546 temporary errors. However, in the case of such a code being given after
5547 the end of a data transmission (i.e. after ".") Exim was failing to write
5548 a retry record for the message. (Yes, there was some broken host that was
5549 actually sending 8xx at this point.)
5551 73. An unknown lookup type in a host list could cause Exim to panic-die when
5552 the list was checked. (An example that provoked this was putting <; in the
5553 middle of a list instead of at the start.) If this happened during a DATA
5554 ACL check, a -D file could be left lying around. This kind of configuration
5555 error no longer causes Exim to die; instead it causes a defer errror. The
5556 incident is still logged to the main and panic logs.
5558 74. Buglet left over from Exim 3 conversion. The message "too many messages
5559 in one connection" was written to the rejectlog but not the mainlog, except
5560 when address rewriting (yes!) was being logged.
5562 75. Added write_rejectlog option.
5564 76. When a system filter was run not as root (that is, when system_filter_user
5565 was set), the values of the $n variables were not being returned to the
5566 main process; thus, they were not subsequently available in the $sn
5569 77. Added +return_path_on_delivery log selector.
5571 78. A connection timeout was being treated differently from recipients deferred
5572 when testing hosts_max_try with a message that was older than the host's
5573 retry timeout. (The host should not be counted, thus allowing all hosts to
5574 be tried at least once before bouncing.) This may have been the cause of an
5575 occasionally reported bug whereby a message would remain on the queue
5576 longer than the retry timeout, but would be bounced if a delivery was
5577 forced. I say "may" because I never totally pinned down the problem;
5578 setting up timeout/retry tests is difficult. See also the next item.
5580 79. The ultimate address timeout was not being applied to errors that involved
5581 a combination of host plus message (for example, a timeout on a MAIL
5582 command). When an address resolved to a number of possible hosts, and they
5583 were not all tried for each delivery (e.g. because of hosts_max_try), a
5584 message could remain on the queue longer than the retry timeout.
5586 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5589 81. Fixed an obscure SMTP outgoing bug which required at least the following
5590 conditions: (a) there was another message waiting for the same server;
5591 (b) the server returned 5xx to all RCPT commands in the first message so
5592 that the message was not completed; (c) the server dropped the connection
5593 or gave a negative response to the RSET that Exim sends to abort the
5594 transaction. The observed case was a dropped connection after DATA that had
5595 been sent in pipelining mode. That is, the server had advertised PIPELINING
5596 but was not implementing it correctly. The effect of the bug was incorrect
5597 behaviour, such as trying another host, and this could lead to a crash.
5603 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5604 and daemon.c were passed as pointers to ints; they should have been
5605 pointers to socklen_t variables (which are typically unsigned ints).
5607 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5610 3. Fixed a really odd bug that affected only the testing scheme; patching a
5611 certain fixed string in the binary changed the value of another string that
5612 happened to be identical to the end of the original first string.
5614 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5615 name", it returns that address as the IP address. On some operating
5616 systems (e.g. Solaris), it also passes back the IP address string as the
5617 "host name". However, on others (e.g. Linux), it passes back an empty
5618 string. Exim wasn't checking for this, and was changing the host name to an
5619 empty string, assuming it had been canonicized.
5621 5. Although rare, it is permitted to have more than one PTR record for a given
5622 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5623 all the names associated with an address, because they do in Solaris.
5624 However, it seems that they do not in Linux for data that comes from the
5625 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5626 I found this out when I moved to a new Linux workstation and tried to run
5627 the Exim test suite.
5629 To get round this problem I have changed the code so that it now does its
5630 own call to the DNS to look up PTR records when searching for a host name.
5631 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5632 addresses that are only in /etc/hosts are still found.
5634 This behaviour is, however, controlled by an option called host_lookup_
5635 order, which defaults to "bydns:byaddr". If people want to use the other
5636 order, or indeed, just use one or the other means of lookup, they can
5637 specify it in this variable.
5639 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5640 some operating systems, this comes back from gethostbyaddr() as an empty
5641 string, and this is what Exim used to test for. However, it seems that in
5642 other systems, "." is yielded. Exim now tests for this case too.
5644 7. The values of check_spool_space and check_log_space are now held internally
5645 as a number of kilobytes instead of an absolute number of bytes. If a
5646 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5647 kilobyte. This means that much larger values can be stored.
5649 8. Exim monitor: an attempt to get the action menu when not actually pointing
5650 at a message produces an empty menu entitled "No message selected". This
5651 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5652 no entries in it ("Shell widget menu has zero width and/or height"). So I
5653 have added a single, blank menu entry in this case.
5655 9. Added ${quote_local_part.
5657 10. MIME decoding is now applied to the contents of Subject: header lines when
5660 11. Now that a reference to $sender_host_address automatically causes a reverse
5661 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5662 host lookup before query-style lookups in lists that might use this
5663 variable. This has therefore been abolished, and the "net-" prefix is no
5664 longer necessary for query-style lookups.
5666 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5667 have been a typo for LFLAGS, so it has been changed.
5669 13. The install script calls Exim with "-C /dev/null" in order to find the
5670 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5671 to be output. Howeve, since Exim outputs its version number before the
5672 error, it didn't break the script. It just looked ugly. I fixed this by
5673 always allowing "-C /dev/null" if the caller is root.
5675 14. Ignore overlarge ACL variable number when reading spool file - insurance
5676 against a later release with more variables having written the file.
5678 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5679 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5680 this, as well as the form without the "IPv6" on the front.
5682 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5683 OS/Makefile-Darwin file.
5685 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5686 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5688 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5689 variable. However, some OS use ints for inodes. I've added cast to long int
5690 to get rid of the compiler warning.
5692 19. I had forgotten to lock out "/../" in configuration file names when
5693 ALT_CONFIG_PREFIX was set.
5695 20. Routers used for verification do not need to specify transports. However,
5696 if such a router generated a host list, and callout was configured, Exim
5697 crashed, because it could not find a port number from the (non-existent)
5698 transport. It now assumes port 25 in this circumstance.
5700 21. Added the -t option to exigrep.
5702 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5703 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5704 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5705 them is set, LOOKUP_LSEARCH is forced.
5707 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5709 24. Added sender and host information to the "rejected by local_scan()" log
5710 line; previously there was no indication of these.
5712 25. Added .include_if_exists.
5714 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5715 Linux. It turns out that not all file systems support this. Apparently some
5716 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5717 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5718 not supported on the file descriptor, is now ignored when Exim is trying to
5719 sync a directory. This applies only to Linux.
5721 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5723 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5724 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5727 29. Added support for maildirsize files from supplied patch (modified a bit).
5729 30. The use of :fail: followed by an empty string could lead Exim to respond to
5730 sender verification failures with (e.g.):
5732 550 Verification failed for <xxx>
5733 550 Sender verify failed
5735 where the first response line was missing the '-' that indicates it is not
5736 the final line of the response.
5738 31. The loop for finding the name of the user that called Exim had a hardwired
5739 limit of 10; it now uses the value of finduser_retries, which is used for
5740 all other user lookups.
5742 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5745 33. Exim was neglecting to zero errno before one call of strtol() when
5746 expanding a string and expecting an integer value. On some systems this
5747 resulted in spurious "integer overflow" errors. Also, it was casting the
5748 result into an int without checking.
5750 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5751 did not work. The code looks as if it has *never* worked, though it appears
5752 to have been documented since at least releast 1.62. I have made it work.
5754 35. The "timeout_DNS" error in retry rules, also documented since at least
5755 1.62, also never worked. As it isn't clear exactly what this means, and
5756 clearly it isn't a major issue, I have abolished the feature by treating it
5757 as "timeout", and writing a warning to the main and panic logs.
5759 36. The display of retry rules for -brt wasn't always showing the error code
5762 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5763 timeout_connect_A, timeout_connect_MX.
5765 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5766 to the empty sender.
5768 39. The daemon was not analysing the content of -oX till after it had closed
5769 stderr and disconnected from the controlling terminal. This meant that any
5770 syntax errors were only noted on the panic log, and the return code from
5771 the command was 0. By re-arranging the code a little, I've made the
5772 decoding happen first, so such errors now appear on stderr, and the return
5773 code is 1. However, the actual setting up of the sockets still happens in
5774 the disconnected process, so errors there are still only recorded on the
5777 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5778 connections (as happens on some IP stacks) was logged at start up time as
5779 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5780 it from "IPv6 and IPv4", which means that two separate sockets are being
5783 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5784 says whether AF_INET or AF_INET6 was passed as an argument.
5786 42. Exiwhat output was messed up when time zones were included in log
5789 43. Exiwhat now gives more information about the daemon's listening ports,
5790 and whether -tls-on-connect was used.
5792 44. The "port" option of the smtp transport is now expanded.
5794 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5795 silently ignored. Now an error message is written to the main and panic
5798 46. There's a new ACL modifier called "logwrite" which writes to a log file
5799 as soon as it is encountered.
5801 47. Added $local_user_uid and $local_user_gid at routing time.
5803 48. Exim crashed when trying to verify a sender address that was being
5806 49. Exim was recognizing only a space character after ".include". It now also
5807 recognizes a tab character.
5809 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5810 extracting the relevant information from the specification. The man page no
5811 longer contains scrambled data for the -d option, and I've added a section
5812 at the front about calling Exim under different names.
5814 51. Added "extra_headers" argument to the "mail" command in filter files.
5816 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5819 53. Installed eximstats 1.29.
5821 54. Added transport_filter_timeout as a generic transport option.
5823 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5824 Cc: header lines. This was required by RFC 822, but it not required by RFC
5827 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5828 incoming messages that did not have them. Now it does so only if the
5829 message originates locally, that is, if there is no associated remote host
5830 address. When Resent- header lines are present, this applies to the Resent-
5831 lines rather than the non-Resent- lines.
5833 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5834 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5836 58. Messages for configuration errors now include the name of the main
5837 configuration file - useful now that there may be more than one file in a
5838 list (.included file names were always shown).
5840 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5841 for those rare installations that do not start the daemon as root or run it
5842 setuid root. I've cut out the call to initgroups() if the daemon is not
5845 60. The Exim user and group can now be bound into the binary as text strings
5846 that are looked up at the start of Exim's processing.
5848 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5850 62. Added $mailstore_basename variable.
5852 63. Installed patch to sieve.c from Michael Haardt.
5854 64. When Exim failed to open the panic log after failing to open the main log,
5855 the original message it was trying to log was written to stderr and debug
5856 output, but if they were not available (the usual case in production), it
5857 was lost. Now it is written to syslog before the two lines that record the
5858 failures to open the logs.
5860 65. Users' Exim filters run in subprocesses under the user's uid. It is
5861 possible for a "deliver" command or an alias in a "personal" command to
5862 provoke an address rewrite. If logging of address rewriting is configured,
5863 this fails because the process is not running as root or exim. There may be
5864 a better way of dealing with this, but for the moment (because 4.30 needs
5865 to be released), I have disabled address rewrite logging when running a
5866 filter in a non-root, non-exim process.
5872 1. The buildconfig auxiliary program wasn't quoting the value set for
5873 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5874 not defined. This bug was masked in 4.22 by the effect that was fixed in
5877 2. Some messages that were rejected after a message id was allocated were
5878 shown as "incomplete" by exigrep. It no longer does this for messages that
5879 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5881 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5882 have allow_domain_literals set, the ID did not get logged in the <= line.
5883 Domain literals are now always recognized in Message-ID: header lines.
5885 4. The first argument for a ${extract expansion item is the key name or field
5886 number. Leading and trailing spaces in this item were not being ignored,
5887 causing some misleading effects.
5889 5. When deliver_drop_privilege was set, single queue runner processes started
5890 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5891 same command in the process it spins off) were not dropping privilege.
5893 6. When the daemon running as "exim" started a queue runner, it always
5894 re-executed Exim in the spun-off process. This is a waste of effort when
5895 deliver_drop_privilege is set. The new process now just calls the
5896 queue-runner function directly.
5902 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5905 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5906 ignored. Though the use of -oP was forcing the writing of a pid file, it
5907 was always written to the default place.
5909 3. If the message "no IP address found for host xxxx" is generated during
5910 incoming verification, it is now followed by identification of the incoming
5911 connection (so you can more easily find what provoked it).
5913 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5915 5. Added some features to "harden" Exim a bit more against certain attacks:
5917 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5918 be put in Local/Makefile. This is like the never_users runtime option,
5919 but it cannot be overridden. The default setting is "root".
5921 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5922 prefix string with which any file named in a -C command line option
5925 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5926 is retained for -C and -D only if the caller of Exim is root. Without
5927 it, the exim user may also use -C and -D and retain privilege.
5929 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5930 command line option is disabled.
5932 6. Macro names set by the -D option must start with an upper case letter, just
5933 like macro names defined in the configuration file.
5935 7. Added "dereference=" facility to LDAP.
5937 8. Two instances of the typo "uknown" in the source files are fixed.
5939 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5940 the Configure-Makefile script screwed up while processing it.
5942 10. Incorporated PCRE 4.4.
5944 11. The SMTP synchronization check was not operating right at the start of an
5945 SMTP session. For example, it could not catch a HELO sent before the client
5946 waited for the greeting. There is now a check for outstanding input at the
5947 point when the greeting is written. Because of the duplex, asynchronous
5948 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5949 way, but not yet received, when the check is performed.
5951 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5952 on TCP/IP sockets, because this apparently causes some broken clients to
5955 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5956 unchanged) from the Cygwin maintainer.
5958 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5959 was supported instead of testing for mbx. Effectively a typo.
5961 15. The spa authenticator server code was not checking that the input it
5962 received was valid base64.
5964 16. The debug output line for the "set" modifier in ACLs was not showing the
5965 name of the variable that was being set.
5967 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5969 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5970 lookup on its own. Something else was needed to trigger the lookup. For
5971 example, a match in host_lookup or the need for a host name in a host list.
5972 Now, if $sender_host_name is referenced and the host name has not yet been
5973 looked up, a lookup is performed. If the lookup fails, the variable remains
5974 empty, and $host_lookup_failed is set to "1".
5976 19. Added "eqi" as a case-independent comparison operator.
5978 20. The saslauthd authentication condition could segfault if neither service
5979 nor realm was specified.
5981 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5982 error message that was logged was misleading, and incoming SMTP
5983 connections were dropped. The message is now more accurate, and temporary
5984 errors are given to SMTP connections.
5986 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5987 (except RSET) with a 421 error, until QUIT is received. However, it was
5988 failing to send a response to QUIT.
5990 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5991 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5992 ACL is now run after the helo_try_verify_hosts code.
5994 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5995 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5996 case-independent, so other case variants are also recognized). Apparently
5997 some systems use these upper case variants.
5999 25. If more than two messages were waiting for the same host, and a transport
6000 filter was specified for the transport, Exim sent two messages over the
6001 same TCP/IP connection, and then failed with "socket operation on non-
6002 socket" when it tried to send the third.
6004 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6006 27. The extern definition of crypt16() in expand.c was not being excluded when
6007 the OS had its own crypt16() function.
6009 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6010 as a preferred synonym for return_size_limit, both as an option and as an
6013 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6015 30. Changed the default configuration ACL to relax the local part checking rule
6016 for addresses that are not in any local domains. For these addresses,
6017 slashes and pipe symbols are allowed within local parts, but the sequence
6018 /../ is explicitly forbidden.
6020 31. SPA server authentication was not clearing the challenge buffer before
6023 32. log_message in a "warn" ACL statement was writing to the reject log as
6024 well as to the main log, which contradicts the documentation and doesn't
6025 seem right (because no rejection is happening). So I have stopped it.
6027 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6028 However, I am unable to do any testing of this.
6030 34. Fixed an infelicity in the appendfile transport. When checking directories
6031 for a mailbox, to see if any needed to be created, it was accidentally
6032 using path names with one or more superfluous leading slashes; tracing
6033 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6035 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6036 contents are added to the log line that is written for every discarded
6037 recipient. (Previously a log_message setting was ignored.)
6039 36. The ${quote: operator now quotes the string if it is empty.
6041 37. The install script runs exim in order to find its version number. If for
6042 some reason other than non-existence or emptiness, which it checks, it
6043 could not run './exim', it was installing it with an empty version number,
6044 i.e. as "exim-". This error state is now caught, and the installation is
6047 38. An argument was missing from the function that creates an error message
6048 when Exim fails to connect to the socket for saslauthd authentication.
6049 This could cause Exim to crash, or give a corrupted message.
6051 39. Added isip, isip4, and isip6 to ${if conditions.
6053 40. The ACL variables $acl_xx are now saved with the message, and can be
6054 accessed later in routers, transports, and filters.
6056 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6057 strings in the file are not string-expanded.
6059 42. If a MAIL command specified a SIZE value that was too large to fit into an
6060 int variable, the check against message_size_limit failed. Such values are
6061 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6062 day this will have to be increased, but I don't think I want to be around
6063 when emails are that large.
6070 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6071 iconv() is not standard in FreeBSD.
6073 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6074 IPv6 enabled. The observed symptom was a segmentation fault on return from
6075 the function os_common_find_running_interfaces() in src/os.c.
6077 3. In the check_special_case() function in daemon.c I had used "errno" as an
6078 argument name, which causes warnings on some systems. This was basically a
6079 typo, since it was named "eno" in the comments!
6081 4. The code that waits for the clock to tick (at a resolution of some fraction
6082 of a second) so as to ensure message-id uniqueness was always waiting for
6083 at least one whole tick, when it could have waited for less. [This is
6084 almost certainly not relevant at current processor speeds, where it is
6085 unlikely to ever wait at all. But we try to future-proof.]
6087 5. The function that sleeps for a time interval that includes fractions of a
6088 second contained a race. It did not block SIGALRM between setting the
6089 timer, and suspending (a couple of lines later). If the interval was short
6090 and the sigsuspend() was delayed until after it had expired, the suspension
6091 never ended. On busy systems this could lead to processes getting stuck for
6094 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6095 process, before it forks any delivery processes. The open lookup caching
6096 mechanism meant that the open file or database connection was passed into
6097 the delivery process. The problem was that delivery processes always tidy
6098 up cached lookup data. This could cause a problem for the next delivery
6099 process started by the queue runner, because the external queue runner
6100 process does not know about the closure. So the next delivery process
6101 still has data in the lookup cache. In the case of a file lookup, there was
6102 no problem because closing a file descriptor in a subprocess doesn't affect
6103 the parent. However, if the lookup was caching a connection to a database,
6104 the connection was closed, and the second delivery process was likely to
6105 see errors such as "PGSQL: query failed: server closed the connection
6106 unexpectedly". The problem has been fixed by closing all cached lookups
6107 in a queue runner before running a delivery process.
6109 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6110 seem to have the "const" qualifier which it has on other OS. I've
6113 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6114 *of the same type* (client or server) with the same public name that an
6115 error should be diagnosed.
6117 9. When Exim looked up a host name for an IP address, but failed to find the
6118 original IP address when looking up the host name (a safety check), it
6119 output the message "<ip address> does not match any IP for NULL", which was
6120 confusing, to say the least. The bug was that the host name should have
6121 appeared instead of "NULL".
6123 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6124 user that is built into the binary, and the -C or -D options is used, root
6125 privilege is dropped before the configuration file is read. In addition,
6126 logging is switched to stderr instead of the normal log files. If the
6127 configuration then re-defines the Exim user, the unprivileged environment
6128 is probably not what is expected, so Exim logs a panic warning message (but
6131 However, if deliver_drop_privilege is set, the unprivileged state may well
6132 be exactly what is intended, so the warning has been cut out in that case,
6133 and Exim is allowed to try to write to its normal log files.
6139 1. smtp_return_error_details was not giving details for temporary sender
6140 or receiver verification errors.
6142 2. Diagnose a configuration error if two authenticators have the same public
6145 3. Exim used not to create the message log file for a message until the first
6146 delivery attempt. This could be confusing when incoming messages were held
6147 for policy or load reasons. The message log file is now created at the time
6148 the message is received, and an initial "Received" line is written to it.
6150 4. The automatically generated man page for command line options had a minor
6151 bug that caused no ill effects; however, a more serious problem was that
6152 the procedure for building the man page automatically didn't always
6153 operate. Consequently, release 4.20 contains an out-of-date version. This
6154 shouldn't happen again.
6156 5. When building Exim with embedded Perl support, the script that builds the
6157 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6158 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6160 6. The freeze_tell option was not being used for messages that were frozen on
6161 arrival, either by an ACL or by local_scan().
6163 7. Added the smtp_incomplete_transaction log selector.
6165 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6166 was accepting AUTH without a new EHLO.
6168 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6169 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6172 10. Logging of TCP/IP connections (when configured) now happens in the main
6173 daemon process instead of the child process, so that the TCP/IP connection
6174 count is more accurate (but it can never be perfect).
6176 11. The use of "drop" in a nested ACL was not being handled correctly in the
6177 outer ACL. Now, if condition failure induced by the nested "drop" causes
6178 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6179 or "require"), the connection is dropped.
6181 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6182 that yield "discard" can now be used with an "accept" or a "discard" verb,
6183 but an error is generated for any others (because I can't see a useful way
6184 to define what should happen).
6186 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6187 are now processed in the same way as lines in the Exim configuration file.
6188 In particular, continuation lines are supported.
6190 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6192 15. Added -ti meaning -t -i.
6194 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6195 domains, and warn by logging if others are found.
6197 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6198 returned from GIFCONF when Exim is trying to find the list of interfaces on
6199 a host. The code in os.c has been modified to copy each ifreq to an aligned
6200 structure in all cases.
6202 Also, in some cases, the returned ifreq's were being copied to a 'struct
6203 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6204 means the last couple of bytes of an IPv6 address could be chopped if the
6205 ifreq contained only a normal sockaddr (14 bytes storage).
6207 18. Named domain lists were not supported in the hosts_treat_as_local option.
6208 An entry such as +xxxx was not recognized, and was treated as a literal
6211 19. Ensure that header lines added by a DATA ACL are included in the reject log
6212 if the ACL subsequently rejects the message.
6214 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6215 MD5 (which is deprecated).
6217 21. When testing a filter file using -bf, Exim was writing a message when it
6218 took the sender from a "From " line in the message, but it was not doing so
6219 when it took $return_path from a Return-Path: header line. It now does.
6221 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6222 with a valid header line field name (a series of printing characters
6223 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6225 23. Changed "disc" in the source to "disk" to conform to the documentation and
6226 the book and for uniformity.
6228 24. Ignore Sendmail's -Ooption=value command line item.
6230 25. When execve() failed while trying to run a command in a pipe transport,
6231 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6232 could be confused with a return value of 69 from the command itself. This
6233 has been changed to 127, the value the shell returns if it is asked to run
6234 a non-existent command. The wording for the related log line suggests a
6235 non-existent command as the problem.
6237 26. If received_header_text expands to an empty string, do not add a Received:
6238 header line to the message. (Well, it adds a token one on the spool, but
6239 marks it "old" so that it doesn't get used or transmitted.)
6241 27. Installed eximstats 1.28 (addition of -nt option).
6243 28. There was no check for failure on the call to getsockname() in the daemon
6244 code. This can fail if there is a shortage of resources on the system, with
6245 ENOMEM, for example. A temporary error is now given on failure.
6247 29. Contrary to the C standard, it seems that in some environments, the
6248 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6249 program. Exim now does this explicitly; it affects the formatting of
6250 timestamps using strftime().
6252 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6253 complaints. I've now initialized all the variables, to avoid this.
6255 32. Header lines added by a system filter were not being "seen" during
6256 transport-time rewrites.
6258 33. The info_callback() function passed to OpenSSL is set up with type void
6259 (*)(SSL *, int, int), as described somewhere. However, when calling the
6260 function (actually a macro) that sets it up, the type void(*)() is
6261 expected. I've put in a cast to prevent warnings from picky compilers.
6263 34. If a DNS black list lookup found a CNAME record, but there were no A
6264 records associated with the domain it pointed at, Exim crashed.
6266 35. If a DNS black list lookup returned more than one A record, Exim ignored
6267 all but the first. It now scans all returned addresses if a particular IP
6268 value is being sought. In this situation, the contents of the
6269 $dnslist_value variable are a list of all the addresses, separated by a
6272 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6273 to accept a host name and all its aliases if the forward lookup for any of
6274 them yielded the IP address of the incoming connection. Now it accepts only
6275 those names whose forward lookup yields the correct IP address. Any other
6276 names are discarded. This closes a loophole whereby a rogue DNS
6277 administrator could create reverse DNS records to break through a
6278 wildcarded host restriction in an ACL.
6280 37. If a user filter or a system filter that ran in a subprocess used any of
6281 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6282 the wrong values were passed to the pipe command ($thisaddress had the
6283 value of $0, $0 had the value of $1, etc). This bug was introduced by
6284 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6286 38. Improved the line breaking for long SMTP error messages from ACLs.
6287 Previously, if there was no break point between 40 and 75 characters, Exim
6288 left the rest of the message alone. Two changes have been made: (a) I've
6289 reduced the minimum length to 35 characters; (b) if it can't find a break
6290 point between 35 and 75 characters, it looks ahead and uses the first one
6291 that it finds. This may give the occasional overlong line, but at least the
6292 remaining text gets split now.
6294 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6295 file descriptors might be low, and that setting 1000 would always raise it.
6296 It turns out that in some environments, the limit is already over 1000 and
6297 that lowering it causes trouble. So now Exim takes care not to decrease it.
6299 40. When delivering a message, the value of $return_path is set to $sender_
6300 address at the start of routing (routers may change the value). By an
6301 oversight, this default was not being set up when an address was tested by
6302 -bt or -bv, which affected the outcome if any router or filter referred to
6305 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6306 log only when "message" or "log_message" are set. However, if one of the
6307 conditions was an address verification, or a call to a nested ACL, the
6308 messages generated by the underlying test were being passed through. This
6309 no longer happens. The underlying message is available in $acl_verify_
6310 message for both "message" and "log_message" expansions, so it can be
6311 passed through if needed.
6313 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6314 new expansion $bh_ to give the encoded byte string without charset
6315 translation. Translation happens only if iconv() is available; HAVE_ICONV
6316 indicates this at build time. HEADERS_CHARSET gives the charset to
6317 translate to; headers_charset can change it in the configuration, and
6318 "headers charset" can change it in an individual filter file.
6320 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6321 that creates RFC 2047 encoded "words" labels them as that charset instead
6322 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6323 expansion operator; (ii) when Exim creates a From: line for a local
6324 message; (iii) when a header line is rewritten to include a "phrase" part.
6326 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6327 buggy, causing it to skip the first lines of messages whose message ID
6328 ended in 'D'. This would not have bitten before Exim release 4.14, because
6329 message IDs were unlikely to end in 'D' before then. The effect was to have
6330 incorrect size information for certain domains.
6332 45. #include "config.h" was missing at the start of the crypt16.c module. This
6333 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6336 46. If there was a timeout during a "random" callout check, Exim treated it as
6337 a failure of the random address, and carried on sending RSET and the real
6338 address. If the delay was just some slowness somewhere, the response to the
6339 original RCPT would be taken as a response to RSET and so on, causing
6340 mayhem of various kinds.
6342 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6343 when I implemented it. It didn't allow for the fact that some option values
6344 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6345 the right test for positive values.
6347 48. Domain names in DNS records are case-independent. Exim always looks them up
6348 in lower case. Some resolvers return domain names in exactly the case they
6349 appear in the zone file, that is, they may contain uppercase letters. Not
6350 all resolvers do this - some return always lower case. Exim was treating a
6351 change of case by a resolver as a change of domain, similar to a widening
6352 of a domain abbreviation. This triggered its re-routing code and so it was
6353 trying to route what was effectively the same domain again. This normally
6354 caused routing to fail (because the router wouldn't handle the domain
6355 twice). Now Exim checks for this case specially, and just changes the
6356 casing of the domain that it ultimately uses when it transmits the message
6359 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6362 50. If a filter generated a file delivery with a non-absolute name (possible if
6363 no home directory exists for the router), the forbid_file option was not
6366 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6367 the existing equality matching.
6369 52. Exim was using ints instead of ino_t variables in some places where it was
6370 dealing with inode numbers.
6372 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6373 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6374 called TMPDIR, and if it finds it is different, it changes its value.
6376 54. The smtp_printf() function is now made available to local_scan() so
6377 additional output lines can be written before returning. There is also an
6378 smtp_fflush() function to enable the detection of a dropped connection.
6379 The variables smtp_input and smtp_batched_input are exported to
6382 55. Changed the default runtime configuration: the message "Unknown user"
6383 has been removed from the ACL, and instead placed on the localuser router,
6384 using the cannot_route_message feature. This means that any verification
6385 failures that generate their own messages won't get overridden. Similarly,
6386 the "Unrouteable address" message that was in the ACL for unverifiable
6387 relay addresses has also been removed.
6389 56. Added hosts_avoid_esmtp to the smtp transport.
6391 57. The exicyclog script was not checking for the esoteric option
6392 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6393 will work only if exicyclog is run under the appropriate euid.
6395 58. Following a discussion on the list, the rules by which Exim recognises line
6396 endings on incoming messages have been changed. The -dropcr and drop_cr
6397 options are now no-ops, retained only for backwards compatibility. The
6398 following line terminators are recognized: LF CRLF CR. However, special
6399 processing applies to CR:
6401 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6402 nor a local message in the state where . is a terminator.
6404 (ii) If a bare CR is encountered in a header line, an extra space is added
6405 after the line terminator so as not to end the header. The reasoning
6406 behind this is that bare CRs in header lines are most likely either
6407 to be mistakes, or people trying to play silly games.
6409 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6410 was being incorrectly given as 18 bytes larger than it should have been.
6411 This is a VOB (very old bug).
6413 60. This may never have affected anything current, but just in case it has:
6414 When the local host is found other than at the start of a list of hosts,
6415 the local host, those with the same MX, and any that follow, are discarded.
6416 When the list in question was part of a longer list of hosts, the following
6417 hosts (not currently being processed) were also being discarded. This no
6418 longer happens. I'm not sure if this situation could ever has previously
6421 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6424 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6425 header. This is something that is recommended in a new Internet Draft, and
6426 is something that is documented as being done by Sendmail. There are two
6427 possible values. For messages generated by the autoreply transport, Exim
6430 Auto-Submitted: auto-replied
6432 whereas for all other generated messages (e.g. bounces) it adds
6434 Auto-Submitted: auto-generated
6436 63. The "personal" condition in filters now includes a test for the
6437 Auto-Submitted: header. If it contains the string "auto-" the message it
6438 not considered personal.
6440 64. Added rcpt_include_affixes as a generic transport option.
6442 65. Added queue_only_override (default true).
6444 66. Added the syslog_duplication option.
6446 67. If what should have been the first header line of a message consisted of
6447 a space followed by a colon, Exim was mis-interpreting it as a header line.
6448 It isn't of course - it is syntactically invalid and should therefore be
6449 treated as the start of the message body. The misbehaviour could have
6450 caused a number of strange effects, including loss of data in subsequent
6451 header lines, and spool format errors.
6453 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6454 client host had authenticated. This control can now be exercised by an ACL
6455 for more flexibility.
6457 69. By default, callouts do not happen when testing with -bh. There is now a
6458 variant, -bhc, which does actually run the callout code, including
6459 consulting and updating the callout cache.
6461 70. Added support for saslauthd authentication, courtesy of Alexander
6464 71. If statvfs() failed on the spool or log directories while checking their
6465 size for availability, Exim confusingly gave the error "space shortage".
6466 Furthermore, in debugging mode it crashed with a floating point exception.
6467 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6468 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6469 serious problem, Exim now writes to the main and panic logs when this
6470 happens, with details of the failure. It then refuses to accept the
6471 incoming message, giving the message "spool directory problem" or "log
6472 directory problem" with a 421 code for SMTP messages.
6474 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6475 0, 1, and 2 exist, because some OS complain for execs without them (see
6476 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6477 descriptors. However, the code omitted to check that the open succeeded,
6478 causing mysterious errors if for some reason the permissions on /dev/null
6479 got screwed. Now Exim writes a message to the main and panic logs, and
6480 bombs out if it can't open /dev/null.
6482 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6483 interact so that it is all more flexible. It is supposed to remain
6484 backwards compatible. Also added extra_local_interfaces.
6486 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6487 to bomb out with an assertion failure - to the client this appears as a
6488 connection drop. This problem occurs in the part of the code that was taken
6489 from the Samba project. Fortunately, the assertion is in a very simple
6490 function, so I have fixed this by reproducing the function inline in the
6491 one place where it is called, and arranging for authentication to fail
6492 instead of killing the process with assert().
6494 75. The SPA client code was not working when the server requested OEM rather
6495 than Unicode encoding.
6497 76. Added code to make require_files with a specific uid setting more usable in
6498 the case where statting the file as root fails - usually a non-root-mounted
6499 NFS file system. When this happens and the failure is EACCES, Exim now
6500 forks a subprocess and does the per-uid checking as the relevant uid.
6502 77. Added process_log_path.
6504 78. If log_file_path was not explicitly set, a setting of check_log_space or
6505 check_log_inodes was ignored.
6507 79. If a space check for the spool or log partitions fails, the incident is now
6508 logged. Of course, in the latter case the data may get lost...
6510 80. Added the %p formatting code to string_format() so that it can be used to
6511 print addresses in debug_print(). Adjusted all the address printing in the
6512 debugging in store.c to use %p rather than %d.
6514 81. There was a concern that a line of code in smtp_in.c could overflow a
6515 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6516 initially expressed, the concern was not well-founded, because trailing
6517 spaces are removed early. However, if the trailing spaces were followed by
6518 a NULL, they did not get removed, so the overflow was possible. Two fixes
6521 (a) I re-wrote the offending code in a cleaner fashion.
6522 (b) If an incoming SMTP command contains a NULL character, it is rejected
6525 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6526 runs initgroups(), so that if the Exim user is in any additional groups,
6527 they will be used during message reception.
6533 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob