1 Change log file for Exim from version 4.21
2 -------------------------------------------
6 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
7 supports it and a size is available (ie. the sending peer gave us one).
9 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
10 years of being deprecated). The replacements are the ACLs
11 acl_smtp_mime and acl_not_smtp_mime.
13 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
14 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
15 or even an in-clear connection were permitted. Now, if the host lookup
16 was dnssec and dane was requested then the host is only used if the
17 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
18 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
19 if one fails this test.
20 This means that a poorly-configured remote DNS will make it incommunicado;
21 but it protects against a DNS-interception attack on it.
23 JH/04 Bug 1810: make continued-use of an open smtp transport connection
24 non-noisy when a race steals the message being considered.
26 JH/05 If main configuration option tls_certificate is unset, generate a
27 selfsigned certificate for inbound TLS connections.
29 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
30 in rewrites and routers.
32 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
33 and logged a warning sing 4.83; now they are a configuration file error.
35 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
36 (lacking @domain). Apply the same qualification processing as RCPT.
38 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
40 JH/10 Support ${sha256:} applied to a string (as well as the previous
43 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
44 a cutthrough deliver is pending, as we always want to make a connection.
45 This also avoids re-routing the message when later placing the cutthrough
46 connection after a verify cache hit.
47 Do not update it with the verify result either.
49 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
50 when routing results in more than one destination address.
52 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
53 signing (which inhibits the cutthrough capability). Previously only
54 the presence of an option was tested; now an expansion evaluating as
55 empty is permissible (obviously it should depend only on data available
56 when the cutthrough connection is made).
58 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
59 the relevant preceding SMTP command did not note the pipelining mode.
61 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
62 Previously they were not counted.
64 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
65 as one having no matching records. Previously we deferred the message
66 that needed the lookup.
68 JH/17 Fakereject: previously logged as a norml message arrival "<="; now
69 distinguished as "(=".
71 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
72 for missing MX records. Previously it only worked for missing A records.
77 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
78 and 3.4.4 - once the server is enabled to respond to an OCSP request
79 it does even when not requested, resulting in a stapling non-aware
80 client dropping the TLS connection.
82 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
83 support variable-length bit vectors. No functional change.
85 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
86 The I= interface field on outgoing lines is now after the H= remote
87 host field, same as incoming lines. There is a separate
88 outgoing_interface log selector which allows you to disable the
91 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
92 If not running log_selector +smtp_connection the mainlog would be held
93 open indefinitely after a "too many connections" event, including to a
94 deleted file after a log rotate. Leave the per net connection logging
95 leaving it open for efficiency as that will be quickly detected by the
96 check on the next write.
98 HS/01 Bug 1671: Fix post transport crash.
99 Processing the wait-<transport> messages could crash the delivery
100 process if the message IDs didn't exist for some reason. When
101 using 'split_spool_directory=yes' the construction of the spool
102 file name failed already, exposing the same netto behaviour.
104 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
105 mime_regex ACL conditions.
107 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
108 to DSN fail messages (bounces): remote IP, remote greeting, remote response
109 to HELO, local diagnostic string.
111 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
112 log line to debug. Even when configured with a tls authenticator many
113 client connections are expected to not authenticate in this way, so
114 an authenticate fail is not an error.
116 HS/02 Add the Exim version string to the process info. This way exiwhat
117 gives some more detail about the running daemon.
119 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
120 matter for fast-change records such as DNSBLs.
122 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
123 retry record, even if constant. There may be multiple transports with
124 different interface settings and the retry behaviour needs to be kept
127 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
129 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
131 JH/10 Bug 840: fix log_defer_output option of pipe transport
133 JH/11 Bug 830: use same host for all RCPTS of a message, even under
134 hosts_randomize. This matters a lot when combined with mua_wrapper.
136 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
137 ${quote_pgsql:<string>} operator.
139 JH/13 Bug 1708: avoid misaligned access in cached lookup.
141 JH/14 Change header file name for freeradius-client. Relevant if compiling
142 with Radius support; from the Gentoo tree and checked under Fedora.
144 JH/15 Bug 1712: Introduce $prdr_requested flag variable
146 JH/16 Bug 1714: Permit an empty string as expansion result for transport
147 option transport_filter, meaning no filtering.
149 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
151 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
152 defaults to "*" (all hosts). The variable is now available when not built
153 with TLS, default unset, mainly to enable keeping the testuite sane.
154 If a server certificate is not supplied (via tls_certificate) an error is
155 logged, and clients will find TLS connections fail on startup. Presumably
156 they will retry in-clear.
157 Packagers of Exim are strongly encouraged to create a server certificate
158 at installation time.
160 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
161 with the $config_file variable.
163 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
164 in transport context, after the attempt, and per-recipient. The latter type
165 is per host attempted. The event data is the error message, and the errno
166 information encodes the lookup type (A vs. MX) used for the (first) host,
167 and the trailing two digits of the smtp 4xx reponse.
169 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
170 to write to mainlog (or rejectlog, paniclog) in the window between file
171 creation and permissions/ownership being changed. Particularly affects
172 installations where exicyclog is run as root, rather than exim user;
173 result is that the running daemon panics and dies.
175 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
177 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
178 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
179 "pri" and "weight". Note that the previous implicit priority given by the
180 list order is no longer honoured.
182 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
185 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
186 by defining SUPPORT_SOCKS.
188 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
189 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
190 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
191 variables are renamed to proxy_{local,external}_{address,port}.
193 JH/27 Move Internationalisation support from Experimental to mainline, enabled
194 for a build by defining SUPPORT_I18N
196 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
197 of the query string, and make ${quote_redis:} do that quoting.
199 JH/29 Move Events support from Experimental to mainline, enabled by default
200 and removable for a build by defining DISABLE_EVENT.
202 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
204 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
205 cached by the daemon.
207 JH/32 Move Redis support from Experimental to mainline, enabled for a build
208 by defining LOOKUP_REDIS. The libhiredis library is required.
210 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
211 keys are given for lookup.
213 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
214 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
215 only supported when built with TLS support. The PolarSSL SHA routines
216 are still used when the TLS library is too old for convenient support.
218 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
219 openssl_options), for security. OpenSSL forces this from version 1.1.0
220 server-side so match that on older versions.
222 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
223 allocation for $value could be released as the expansion processing
224 concluded, but leaving the global pointer active for it.
226 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
227 and to use the domains and local_parts ACL conditions.
229 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
230 incorrectly not doubled on cutthrough transmission, hence seen as a
231 body-termination at the receiving system - resulting in truncated mails.
232 Commonly the sender saw a TCP-level error, and retransmitted the message
233 via the normal store-and-forward channel. This could result in duplicates
234 received - but deduplicating mailstores were liable to retain only the
235 initial truncated version.
237 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
239 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
241 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
242 we're in there, support oversigning also; bug 1309.
244 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
246 HS/04 Add support for keep_environment and add_environment options.
248 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
249 either intentional arithmetic overflow during PRNG, or testing config-
252 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
253 delivery resulted in actual delivery. Cancel cutthrough before DATA
256 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
257 rejects a recipient: pass the reject to the originator.
259 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
260 Many were false-positives and ignorable, but it's worth fixing the
263 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
264 for the new environment-manipulation done at startup. Move the routines
265 from being local to tls.c to being global via the os.c file.
267 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
268 an extract embedded as result-arg for a map, the first arg for extract
269 is unavailable so we cannot tell if this is a numbered or keyed
270 extraction. Accept either.
276 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
279 JH/02 The smtp transport option "multi_domain" is now expanded.
281 JH/03 The smtp transport now requests PRDR by default, if the server offers
284 JH/04 Certificate name checking on server certificates, when exim is a client,
285 is now done by default. The transport option tls_verify_cert_hostnames
286 can be used to disable this per-host. The build option
287 EXPERIMENTAL_CERTNAMES is withdrawn.
289 JH/05 The value of the tls_verify_certificates smtp transport and main options
290 default to the word "system" to access the system default CA bundle.
291 For GnuTLS, only version 3.0.20 or later.
293 JH/06 Verification of the server certificate for a TLS connection is now tried
294 (but not required) by default. The verification status is now logged by
295 default, for both outbound TLS and client-certificate supplying inbound
298 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
301 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
302 Status Notification (bounce) messages are now MIME format per RFC 3464.
303 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
304 under the control of the dsn_advertise_hosts option, and routers may
305 have a dsn_lasthop option.
307 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
308 default, modifiable by a malware= option. The list separator for
309 the options can now be changed in the usual way. Bug 68.
311 JH/10 The smtp_receive_timeout main option is now expanded before use.
313 JH/11 The incoming_interface log option now also enables logging of the
314 local interface on delivery outgoing connections.
316 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
317 if the interface and destination host and port all match.
319 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
322 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
323 Patch from Andrew Lewis.
325 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
326 now supports optional time-restrictions, weighting, and priority
327 modifiers per server. Patch originally by <rommer@active.by>.
329 JH/16 The spamd_address main option now supports a mixed list of local
330 and remote servers. Remote servers can be IPv6 addresses, and
331 specify a port-range.
333 JH/17 Bug 68: The spamd_address main option now supports an optional
334 timeout value per server.
336 JH/18 Bug 1581: Router and transport options headers_add/remove can
337 now have the list separator specified.
339 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
342 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
345 JH/21 Support for the A6 type of dns record is withdrawn.
347 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
348 rather than the verbs used.
350 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
351 from 255 to 1024 chars.
353 JH/24 Verification callouts now attempt to use TLS by default.
355 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
356 are generic router options now. The defaults didn't change.
358 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
359 Original patch from Alexander Shikoff, worked over by JH.
361 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
362 files if ZCAT_COMMAND is not executable.
364 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
366 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
368 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
369 Normally benign, it bites when the pair was led to by a CNAME;
370 modern usage is to not canoicalize the domain to a CNAME target
371 (and we were inconsistent anyway for A-only vs AAAA+A).
373 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
375 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
376 when evaluating $sender_host_dnssec.
378 JH/31 Check the HELO verification lookup for DNSSEC, adding new
379 $sender_helo_dnssec variable.
381 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
383 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
385 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
387 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
388 documented as working, but never had. Support all but $spam_report.
390 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
391 added for tls authenticator.
393 HS/03 Add perl_taintmode main config option
398 TL/01 When running the test suite, the README says that variables such as
399 no_msglog_check are global and can be placed anywhere in a specific
400 test's script, however it was observed that placement needed to be near
401 the beginning for it to behave that way. Changed the runtest perl
402 script to read through the entire script once to detect and set these
403 variables, reset to the beginning of the script, and then run through
404 the script parsing/test process like normal.
406 TL/02 The BSD's have an arc4random API. One of the functions to induce
407 adding randomness was arc4random_stir(), but it has been removed in
408 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
409 function when detected.
411 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
412 cause callback expansion.
414 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
415 syntax errors in an expansion can be treated as a string instead of
416 logging or causing an error, due to the internal use of bool_lax
417 instead of bool when processing it.
419 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
420 server certificates when making smtp deliveries.
422 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
424 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
426 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
428 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
429 Merged patch from Sebastian Wiedenroth.
431 JH/05 Fix results-pipe from transport process. Several recipients, combined
432 with certificate use, exposed issues where response data items split
433 over buffer boundaries were not parsed properly. This eventually
434 resulted in duplicates being sent. This issue only became common enough
435 to notice due to the introduction of conection certificate information,
436 the item size being so much larger. Found and fixed by Wolfgang Breyha.
438 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
439 size buffer was used, resulting in syntax errors when an expansion
442 JH/07 Add support for directories of certificates when compiled with a GnuTLS
443 version 3.3.6 or later.
445 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
446 is EXPERIMENTAL_EVENT, the main-configuration and transport options
447 both become "event_action", the variables become $event_name, $event_data
448 and $event_defer_errno. There is a new variable $verify_mode, usable in
449 routers, transports and related events. The tls:cert event is now also
450 raised for inbound connections, if the main configuration event_action
453 TL/06 In test suite, disable OCSP for old versions of openssl which contained
454 early OCSP support, but no stapling (appears to be less than 1.0.0).
456 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
457 server certificate names available under the smtp transport option
458 "tls_verify_cert_hostname" now do not permit multi-component wildcard
461 JH/10 Time-related extraction expansions from certificates now use the main
462 option "timezone" setting for output formatting, and are consistent
463 between OpenSSL and GnuTLS compilations. Bug 1541.
465 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
466 encoded parameter in the incoming message. Bug 1558.
468 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
469 include certificate info, eximon was claiming there were spoolfile
472 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
474 JH/14 Log delivery-related information more consistently, using the sequence
475 "H=<name> [<ip>]" wherever possible.
477 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
478 are problematic for Debian distribution, omit them from the release
481 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
483 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
486 JH/17 Fix a null-indirection in certextract expansions when a nondefault
487 output list separator was used.
492 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
493 checkers that were complaining about end of non-void function with no
496 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
497 This was a regression intruduced in 4.83 by another bugfix.
499 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
501 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
502 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
508 TF/01 Correctly close the server side of TLS when forking for delivery.
510 When a message was received over SMTP with TLS, Exim failed to clear up
511 the incoming connection properly after forking off the child process to
512 deliver the message. In some situations the subsequent outgoing
513 delivery connection happened to have the same fd number as the incoming
514 connection previously had. Exim would try to use TLS and fail, logging
515 a "Bad file descriptor" error.
517 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
518 utilities have not been installed.
520 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
521 temporary space as the ACL may create new global variables.
523 TL/01 LDAP support uses per connection or global context settings, depending
524 upon the detected version of the libraries at build time.
526 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
527 to extract and use the src ip:port in logging and expansions as if it
528 were a direct connection from the outside internet. PPv2 support was
529 updated based on HAProxy spec change in May 2014.
531 JH/02 Add ${listextract {number}{list}{success}{fail}}.
533 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
534 Properly escape header and check for NULL return.
536 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
539 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
541 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
542 characters in header names, implemented as a verify condition.
543 Contributed by Michael Fischer v. Mollard.
545 TL/05 Rename SPF condition results err_perm and err_temp to standardized
546 results permerror and temperror. Previous values are deprecated but
547 still accepted. In a future release, err_perm and err_temp will be
548 completely removed, which will be a backward incompatibility if the
549 ACL tests for either of these two old results. Patch contributed by
550 user bes-internal on the mailing list.
552 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
554 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
555 selectors, in both main and reject logs.
557 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
560 JH/07 Add malware type "sock" for talking to simple daemon.
562 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
564 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
565 routers/transports under cutthrough routing.
567 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
568 numbers. Touch up "bool" conditional to keep the same definition.
570 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
572 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
574 JH/12 Expand items in router/transport headers_add or headers_remove lists
575 individually rather than the list as a whole. Bug 1452.
577 Required for reasonable handling of multiple headers_ options when
578 they may be empty; requires that headers_remove items with embedded
579 colons must have them doubled (or the list-separator changed).
581 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
582 view the policy declared in the DMARC record. Currently, $dmarc_status
583 is a combined value of both the record presence and the result of the
586 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
588 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
589 dnslookup router and the smtp transport (applying to the forward
592 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
593 of ldap servers used for a specific lookup. Patch provided by Heiko
596 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
597 New variable $lookup_dnssec_authenticated for observability.
599 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
600 Patch submitted by Lars Timman.
602 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
604 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
605 Requires trusted mode and valid format message id, aborts otherwise.
606 Patch contributed by Heiko Schlichting.
608 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
609 certextract with support for various fields. Bug 1358.
611 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
612 is requested by default, modifiable by smtp transport option
615 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
616 operate on certificate variables to give certificate fingerprints
617 Also new ${sha256:cert_variable}.
619 JH/23 The PRDR feature is moved from being Experimental into the mainline.
621 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
622 Christian Aistleitner.
624 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
626 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
627 file. Patch from Wolfgang Breyha.
629 JH/25 Expand the coverage of the delivery $host and $host_address to
630 client authenticators run in verify callout. Bug 1476.
632 JH/26 Port service names are now accepted for tls_on_connect_ports, to
633 align with daemon_smtp_ports. Bug 72.
635 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
636 support and error reporting did not work properly.
638 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
639 and is readable. Patch from Andrew Colin Kissa.
641 TL/14 Enhance documentation of ${run expansion and how it parses the
642 commandline after expansion, particularly in the case when an
643 unquoted variable expansion results in an empty value.
645 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
647 PP/02 Fix internal collision of T_APL on systems which support RFC3123
648 by renaming away from it. Addresses GH issue 15, reported by
651 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
653 TL/15 SECURITY: prevent double expansion in math comparison functions
654 (can expand unsanitized data). Not remotely exploitable.
661 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
663 PP/02 Make -n do something, by making it not do something.
664 When combined with -bP, the name of an option is not output.
666 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
669 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
670 $sender_host_name and config options to manage this, and basic check
673 PP/05 DSCP support for outbound connections and control modifier for inbound.
675 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
676 (Only plugin which currently uses this is kerberos4, which nobody should
677 be using, but we should make it available and other future plugins might
678 conceivably use it, even though it would break NAT; stuff *should* be
679 using channel bindings instead).
681 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
682 name; added for Sendmail compatibility; requires admin caller.
683 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
684 just ignore it); requires trusted caller.
685 Also parse but ignore: -Ac -Am -X<logfile>
688 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
690 TL/02 Add +smtp_confirmation as a default logging option.
692 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
693 Patch by Magnus Holmgren from 2007-02-20.
695 TL/04 Bugzilla 1281 - Spec typo.
696 Bugzilla 1283 - Spec typo.
697 Bugzilla 1290 - Spec grammar fixes.
699 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
701 TL/06 Add Experimental DMARC support using libopendmarc libraries.
703 TL/07 Fix an out of order global option causing a segfault. Reported to dev
704 mailing list by by Dmitry Isaikin.
706 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
708 JH/02 Support "G" suffix to numbers in ${if comparisons.
710 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
712 NM/01 Bugzilla 1197 - Spec typo
713 Bugzilla 1196 - Spec examples corrections
715 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
717 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
718 gnutls_enable_pkcs11, but renamed to more accurately indicate its
721 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
722 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
724 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
725 "acl {{name}{arg}...}", and optional args on acl condition
728 JH/05 Permit multiple router/transport headers_add/remove lines.
730 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
732 JH/07 Avoid using a waiting database for a single-message-only transport.
733 Performance patch from Paul Fisher. Bugzilla 1262.
735 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
738 JH/09 Add $headers_added variable, with content from use of ACL modifier
739 add_header (but not yet added to the message). Bugzilla 199.
741 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
742 Pulled from Bugzilla 817 by Wolfgang Breyha.
744 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
746 (nb: this is the same fix as in Exim 4.80.1)
748 JH/11 Add A= logging on delivery lines, and a client_set_id option on
751 JH/12 Add optional authenticated_sender logging to A= and a log_selector
754 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
756 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
757 advertise SMTP AUTH mechanism to us, instead of a generic
758 protocol violation error. Also, make Exim more robust to bad
759 data from the Dovecot auth socket.
761 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
763 When a queue runner is handling a message, Exim first routes the
764 recipient addresses, during which it prunes them based on the retry
765 hints database. After that it attempts to deliver the message to
766 any remaining recipients. It then updates the hints database using
769 So if a recipient address works intermittently, it can get repeatedly
770 deferred at routing time. The retry hints record remains fresh so the
771 address never reaches the final cutoff time.
773 This is a fairly common occurrence when a user is bumping up against
774 their storage quota. Exim had some logic in its local delivery code
775 to deal with this. However it did not apply to per-recipient defers
776 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
778 This change adds a proper retry rule check during routing so that the
779 final cutoff time is checked against the message's age. We only do
780 this check if there is an address retry record and there is not a
781 domain retry record; this implies that previous attempts to handle
782 the address had the retry_use_local_parts option turned on. We use
783 this as an approximation for the destination being like a local
784 delivery, as in LMTP.
786 I suspect this new check makes the old local delivery cutoff check
787 redundant, but I have not verified this so I left the code in place.
789 TF/02 Correct gecos expansion when From: is a prefix of the username.
791 Test 0254 submits a message to Exim with the header
795 When I ran the test suite under the user fanf2, Exim expanded
796 the header to contain my full name, whereas it should have added
797 a Resent-Sender: header. It erroneously treats any prefix of the
798 username as equal to the username.
800 This change corrects that bug.
802 GF/01 DCC debug and logging tidyup
803 Error conditions log to paniclog rather than rejectlog.
804 Debug lines prefixed by "DCC: " to remove any ambiguity.
806 TF/03 Avoid unnecessary rebuilds of lookup-related code.
808 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
809 Bug spotted by Jeremy Harris; was flawed since initial commit.
810 Would have resulted in OCSP responses post-SNI triggering an Exim
811 NULL dereference and crash.
813 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
815 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
816 Bug detection, analysis and fix by Samuel Thibault.
817 Bugzilla 1331, Debian bug #698092.
819 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
821 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
822 Server implementation by Todd Lyons, client by JH.
823 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
824 config variable "prdr_enable" controls whether the server
825 advertises the facility. If the client requests PRDR a new
826 acl_data_smtp_prdr ACL is called once for each recipient, after
827 the body content is received and before the acl_smtp_data ACL.
828 The client is controlled by bolth of: a hosts_try_prdr option
829 on the smtp transport, and the server advertisement.
830 Default client logging of deliveries and rejections involving
831 PRDR are flagged with the string "PRDR".
833 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
834 fclose(). Diagnosis by Todd Lyons.
836 PP/17 Update configure.default to handle IPv6 localhost better.
837 Patch by Alain Williams (plus minor tweaks).
840 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
841 This is now consistent with GnuTLS, and is now documented: the
842 previous undocumented portable approach to treating the option as
843 unset was to force an expansion failure. That still works, and
844 an empty string is now equivalent.
846 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
847 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
848 not performing validation itself.
850 PP/20 Added force_command boolean option to pipe transport.
851 Patch from Nick Koston, of cPanel Inc.
853 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
856 TF/04 Added udpsend ACL modifer and hexquote expansion operator
858 PP/21 Fix eximon continuous updating with timestamped log-files.
859 Broken in a format-string cleanup in 4.80, missed when I repaired the
860 other false fix of the same issue.
861 Report and fix from Heiko Schlichting.
864 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
865 Report from Prashanth Katuri.
867 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
868 It's SecureTransport, so affects any MacOS clients which use the
869 system-integrated TLS libraries, including email clients.
871 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
872 using a MIME ACL for non-SMTP local injection.
873 Report and assistance in diagnosis by Warren Baker.
875 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
877 JH/16 Fix comparisons for 64b. Bugzilla 1385.
879 TL/09 Add expansion variable $authenticated_fail_id to keep track of
880 last id that failed so it may be referenced in subsequent ACL's.
882 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
885 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
886 ldap library initialization, allowing self-signed CA's to be
887 used. Also properly sets require_cert option later in code by
888 using NULL (global ldap config) instead of ldap handle (per
889 session). Bug diagnosis and testing by alxgomz.
891 TL/12 Enhanced documentation in the ratelimit.pl script provided in
892 the src/util/ subdirectory.
894 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
895 renamed to Transport Post Delivery Action by Jeremy Harris, as
898 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
899 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
900 redis_servers = needs to be configured which will be used by the redis
901 lookup. Patch from Warren Baker, of The Packet Hub.
903 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
905 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
906 hostname or reverse DNS when processing a host list. Used suggestions
907 from multiple comments on this bug.
909 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
911 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
912 Missed a few lines, added it to make the runtest require no keyboard
915 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
916 contains upper case chars. Make router use caseful_local_part.
918 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
919 support when GnuTLS has been built with p11-kit.
925 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
927 This, or similar/improved, will also be change PP/11 of 4.82.
933 PP/01 Handle short writes when writing local log-files.
934 In practice, only affects FreeBSD (8 onwards).
935 Bugzilla 1053, with thanks to Dmitry Isaikin.
937 NM/01 Bugzilla 949 - Documentation tweak
939 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
942 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
944 PP/02 Implemented gsasl authenticator.
946 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
948 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
949 `pkg-config foo` for cflags/libs.
951 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
952 with rest of GSASL and with heimdal_gssapi.
954 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
955 `pkg-config foo` for cflags/libs for the TLS implementation.
957 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
958 properties get this fed in as external SSF. A number of robustness
959 and debugging improvements to the cyrus_sasl authenticator.
961 PP/08 cyrus_sasl server now expands the server_realm option.
963 PP/09 Bugzilla 1214 - Log authentication information in reject log.
964 Patch by Jeremy Harris.
966 PP/10 Added dbmjz lookup type.
968 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
970 PP/12 MAIL args handles TAB as well as SP, for better interop with
971 non-compliant senders.
972 Analysis and variant patch by Todd Lyons.
974 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
975 Bug report from Lars Müller <lars@samba.org> (via SUSE),
976 Patch from Dirk Mueller <dmueller@suse.com>
978 PP/13 tls_peerdn now print-escaped for spool files.
979 Observed some $tls_peerdn in wild which contained \n, which resulted
980 in spool file corruption.
982 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
983 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
984 or write after TLS renegotiation, which otherwise led to messages
987 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
988 as a tracking header (ie: a signed header comes before the signature).
989 Patch from Wolfgang Breyha.
991 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
992 comma-sep list; embedded commas doubled.
994 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
996 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
998 Report and patch from Dmitry Banschikov.
1000 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1001 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1002 failures. We appear to now support TLS1.1+ with Exim.
1004 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1005 lets Exim select keys and certificates based upon TLS SNI from client.
1006 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1007 before an outbound SMTP session. New log_selector, +tls_sni.
1009 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1010 NULL dereference. Report and patch from Alun Jones.
1012 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1013 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1014 Not seeing resolver debug output on NetBSD, but suspect this is a
1015 resolver implementation change.
1017 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1018 Left warnings. Added "eximon gdb" invocation mode.
1020 PP/21 Defaulting "accept_8bitmime" to true, not false.
1022 PP/22 Added -bw for inetd wait mode support.
1024 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1025 locate the relevant includes and libraries. Made this the default.
1027 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1028 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1030 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1031 This may cause build issues on older platforms.
1033 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1034 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1035 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1036 Added SNI support via GnuTLS too.
1037 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1039 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1041 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1042 Applied second patch from Janne, implementing suggestion to default
1043 multiple-strings-in-record handling to match SPF spec.
1045 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1047 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1048 read-only, out of scope).
1049 Patch from Wolfgang Breyha, report from Stuart Northfield.
1051 PP/29 Fix three issues highlighted by clang analyser static analysis.
1052 Only crash-plausible issue would require the Cambridge-specific
1053 iplookup router and a misconfiguration.
1054 Report from Marcin Mirosław.
1056 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1058 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1059 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1060 As part of this, removing so much warning spew let me fix some minor
1061 real issues in debug logging.
1063 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1064 assignment on my part. Fixed.
1066 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1067 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1068 Janne Snabb (who went above and beyond: thank you).
1070 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1071 string otherwise requires a connection and a bunch more work and it's
1072 relatively easy to get wrong. Should also expose TLS library linkage
1075 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1076 64-bit ${eval} (JH/03).
1078 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1079 GNU libc to support some of the 64-bit stuff, should not lead to
1080 conflicts. Defined before os.h is pulled in, so if a given platform
1081 needs to override this, it can.
1083 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1084 protection layer was required, which is not implemented.
1085 Bugzilla 1254, patch from Wolfgang Breyha.
1087 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1088 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1089 tls_dhparam take prime identifiers. Also unbreak combination of
1090 OpenSSL+DH_params+TLSSNI.
1092 PP/39 Disable SSLv2 by default in OpenSSL support.
1098 PP/01 Solaris build fix for Oracle's LDAP libraries.
1099 Bugzilla 1109, patch from Stephen Usher.
1101 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1103 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1106 TF/02 Fix a couple more cases where we did not log the error message
1107 when unlink() failed. See also change 4.74-TF/03.
1109 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1110 lock up or crash if it happened to be inside a call to libc when it
1111 got a SIGUSR1 from exiwhat.
1113 The SIGUSR1 handler appends the current process status to the process
1114 log which is later printed by exiwhat. It used to use the general
1115 purpose logging code to do this, but several functions it calls are
1116 not safe for signals.
1118 The new output code in the SIGUSR1 handler is specific to the process
1119 log, and simple enough that it's easy to inspect for signal safety.
1120 Removing some special cases also simplifies the general logging code.
1121 Removing the spurious timestamps from the process log simplifies
1124 TF/04 Improved ratelimit ACL condition.
1126 The /noupdate option has been deprecated in favour of /readonly which
1127 has clearer semantics. The /leaky, /strict, and /readonly update modes
1128 are mutually exclusive. The update mode is no longer included in the
1129 database key; it just determines when the database is updated. (This
1130 means that when you upgrde Exim will forget old rate measurements.)
1132 Exim now checks that the per_* options are used with an update mode that
1133 makes sense for the current ACL. For example, when Exim is processing a
1134 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1135 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1136 must specify per_mail/readonly. If you omit the update mode it defaults to
1137 /leaky where that makes sense (as before) or /readonly where required.
1139 The /noupdate option is now undocumented but still supported for
1140 backwards compatibility. It is equivalent to /readonly except that in
1141 ACLs where /readonly is required you may specify /leaky/noupdate or
1142 /strict/noupdate which are treated the same as /readonly.
1144 A useful new feature is the /count= option. This is a generalization
1145 of the per_byte option, so that you can measure the throughput of other
1146 aggregate values. For example, the per_byte option is now equivalent
1147 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1149 The per_rcpt option has been generalized using the /count= mechanism
1150 (though it's more complicated than the per_byte equivalence). When it is
1151 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1152 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1153 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1154 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1155 Note that using per_rcpt with a non-readonly update mode in more than
1156 one ACL will cause the recipients to be double-counted. (The per_mail
1157 and per_byte options don't have this problem.)
1159 The handling of very low rates has changed slightly. If the computed rate
1160 is less than the event's count (usually one) then this event is the first
1161 after a long gap. In this case the rate is set to the same as this event's
1162 count, so that the first message of a spam run is counted properly.
1164 The major new feature is a mechanism for counting the rate of unique
1165 events. The new per_addr option counts the number of different
1166 recipients that someone has sent messages to in the last time period. It
1167 behaves like per_rcpt if all the recipient addresses are different, but
1168 duplicate recipient addresses do not increase the measured rate. Like
1169 the /count= option this is a general mechanism, so the per_addr option
1170 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1171 example, measure the rate that a client uses different sender addresses
1172 with the options per_mail/unique=$sender_address. There are further
1173 details in the main documentation.
1175 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1177 TF/06 Removed a few PCRE remnants.
1179 TF/07 Automatically extract Exim's version number from tags in the git
1180 repository when doing development or release builds.
1182 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1183 Bugzilla 879. Patch from Paul Fisher.
1185 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1186 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1189 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1191 PP/05 Variable $av_failed, true if the AV scanner deferred.
1192 Bugzilla 1078. Patch from John Horne.
1194 PP/06 Stop make process more reliably on build failure.
1195 Bugzilla 1087. Patch from Heiko Schlittermann.
1197 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1198 Bugzilla 1089. Patch from Heiko Schlittermann.
1200 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1201 Bugzilla 1131. Patch from Holger Weiß.
1203 PP/09 Handle IPv6 addresses with SPF.
1204 Bugzilla 860. Patch from Wolfgang Breyha.
1206 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1208 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1211 PP/11 match_* no longer expand right-hand-side by default.
1212 New compile-time build option, EXPAND_LISTMATCH_RHS.
1213 New expansion conditions, "inlist", "inlisti".
1215 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1217 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1219 PP/14 fix log_write() format string regression from TF/03.
1220 Bugzilla 1152. Patch from Dmitry Isaikin.
1226 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1228 PP/02 Harmonised TLS library version reporting; only show if debugging.
1229 Layout now matches that introduced for other libraries in 4.74 PP/03.
1231 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1233 PP/04 New "dns_use_edns0" global option.
1235 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1238 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1239 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1241 TK/01 Updated PolarSSL code to 0.14.2.
1242 Bugzilla 1097. Patch from Andreas Metzler.
1244 PP/07 Catch divide-by-zero in ${eval:...}.
1245 Fixes bugzilla 1102.
1247 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1250 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1251 format-string attack -- SECURITY: remote arbitrary code execution.
1253 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1254 time unintentionally subject to list matching rules, letting the header
1255 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1256 arbitrary string expansion). This allowed for information disclosure.
1258 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1259 INT_MIN/-1 -- value coerced to INT_MAX.
1265 NM/01 Workround for PCRE version dependency in version reporting
1268 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1269 This fixes portability to compilers other than gcc, notably
1270 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1272 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1273 makefiles for portability to HP-UX and POSIX correctness.
1275 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1276 Also via indented variable definition in the Makefile.
1277 (Debugging by Oliver Heesakkers).
1279 PP/02 Restore caching of spamd results with expanded spamd_address.
1280 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1282 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1283 Improves build reliability. Fix from: Frank Elsner
1285 NM/02 Fix wide character breakage in the rfc2047 coding
1286 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1288 NM/03 Allow underscore in dnslist lookups
1289 Fixes bug 1026. Patch from Graeme Fowler
1291 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1292 Code patches from Adam Ciarcinski of NetBSD.
1294 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1297 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1298 is logged, to avoid truncation. Patch from John Horne.
1300 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1301 Patch from Jakob Hirsch.
1303 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1304 SQL string expansion failure details.
1305 Patch from Andrey Oktyabrski.
1307 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1308 Patch from Simon Arlott.
1310 PP/09 New lookups functionality failed to compile on old gcc which rejects
1311 extern declarations in function scope.
1312 Patch from Oliver Fleischmann
1314 PP/10 Use sig_atomic_t for flags set from signal handlers.
1315 Check getgroups() return and improve debugging.
1316 Fixed developed for diagnosis in bug 927 (which turned out to be
1319 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1320 Patch from Mark Zealey.
1322 PP/12 Bugzilla 1056: Improved spamd server selection.
1323 Patch from Mark Zealey.
1325 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1326 Based on patch from Heiko Schlittermann.
1328 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1329 Patch from Uwe Doering, sign-off by Michael Haardt.
1331 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1332 variable declaration deep within a block. Bug and patch from
1335 PP/15 lookups-Makefile IRIX compatibilty coercion.
1337 PP/16 Make DISABLE_DKIM build knob functional.
1339 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1340 Patch by Simon Arlott
1342 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1343 variable argument macros. Our copy now differs from upstream.
1349 TF/01 Failure to get a lock on a hints database can have serious
1350 consequences so log it to the panic log.
1352 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1353 controlled using the smtp_confirmation log selector.
1355 TF/03 Include the error message when we fail to unlink a spool file.
1357 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1358 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1359 for maintaining out-of-tree patches for some time.
1361 PP/01 Bugzilla 139: Documentation and portability issues.
1362 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1363 Handle per-OS dynamic-module compilation flags.
1365 PP/02 Let /dev/null have normal permissions.
1366 The 4.73 fixes were a little too stringent and complained about the
1367 permissions on /dev/null. Exempt it from some checks.
1368 Reported by Andreas M. Kirchwitz.
1370 PP/03 Report version information for many libraries, including
1371 Exim version information for dynamically loaded libraries. Created
1372 version.h, now support a version extension string for distributors
1373 who patch heavily. Dynamic module ABI change.
1375 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1376 privilege escalation vulnerability whereby the Exim run-time user
1377 can cause root to append content of the attacker's choosing to
1380 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1383 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1384 If dropping privileges for untrusted macros, we disabled normal logging
1385 on the basis that it would fail; for the Exim run-time user, this is not
1386 the case, and it resulted in successful deliveries going unlogged.
1387 Fixed. Reported by Andreas Metzler.
1393 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1394 only prepend for the Resent-* case. Fixes regression introduced in
1395 Exim 4.70 by NM/22 for Bugzilla 607.
1397 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1398 increasing numbers of administrators be bitten by this.
1400 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1402 PP/03 Bugzilla 994: added openssl_options main configuration option.
1404 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1406 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1408 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1410 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1411 without a peer certificate, leading to a segfault because of an
1412 assumption that peers always have certificates. Be a little more
1413 paranoid. Problem reported by Martin Tscholak.
1415 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1416 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1417 NB: ClamAV planning to remove STREAM in "middle of 2010".
1418 CL also introduces -bmalware, various -d+acl logging additions and
1419 more caution in buffer sizes.
1421 PP/09 Implemented reverse_ip expansion operator.
1423 PP/10 Bugzilla 937: provide a "debug" ACL control.
1425 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1427 PP/12 Bugzilla 973: Implement --version.
1429 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1431 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1433 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1435 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1436 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1437 ignore trailing whitespace.
1439 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1441 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1444 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1445 Notification from Dr Andrew Aitchison.
1447 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1448 ExtendedDetectionInfo response format.
1449 Notification from John Horne.
1451 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1454 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1455 XSL and documented dependency on system catalogs, with examples of how
1458 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1461 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1462 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1463 configuration file which is writeable by the Exim user or group.
1465 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1466 of configuration files to cover files specified with the -C option if
1467 they are going to be used with root privileges, not just the default
1470 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1471 option (effectively making it always true).
1473 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1474 files to be used while preserving root privileges.
1476 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1477 that rogue child processes cannot use them.
1479 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1480 run-time user, instead of root.
1482 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1483 Exim run-time user without dropping privileges.
1485 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1486 result string, instead of calling string_vformat() twice with the same
1489 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1490 for other users. Others should always drop root privileges if they use
1491 -C on the command line, even for a whitelisted configure file.
1493 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1495 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1501 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1502 $data_path, and $header_path variables; fixed documentation bugs and
1505 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1506 exipick to access non-standard spools, including the "frozen" queue
1509 NM/01 Bugzilla 965: Support mysql stored procedures.
1510 Patch from Alain Williams
1512 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1514 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1515 Patch from Andreas Metzler
1517 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1518 Patch from Kirill Miazine
1520 NM/05 Bugzilla 671: Added umask to procmail example.
1522 JJ/03 installed exipick 20100323.0, fixing doc bug
1524 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1525 directory. Notification and patch from Dan Rosenberg.
1527 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1529 TK/02 Improve log output when DKIM signing operation fails.
1531 MH/01 Treat the transport option dkim_domain as a colon separated
1532 list, not as a single string, and sign the message with each element,
1533 omitting multiple occurences of the same signer.
1535 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1536 Bugzilla 985, 986. Patch by Simon Arlott
1538 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1539 Patch by Simon Arlott
1541 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1542 MBX locking. Notification from Dan Rosenberg.
1548 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1550 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1552 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1554 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1556 NM/04 Bugzilla 915: Flag broken perl installation during build.
1562 TK/01 Added patch by Johannes Berg that expands the main option
1563 "spamd_address" if it starts with a dollar sign.
1565 TK/02 Write list of recipients to X-Envelope-Sender header when building
1566 the mbox-format spool file for content scanning (suggested by Jakob
1569 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1570 (http://www.dcc-servers.net/) support via dccifd. Activated by
1571 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1573 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1574 by Mark Daniel Reidel <mr@df.eu>.
1576 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1577 When building exim an external PCRE library is now needed -
1578 PCRE is a system library on the majority of modern systems.
1579 See entry on PCRE_LIBS in EDITME file.
1581 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1582 conversation. Added nologin parameter to request.
1583 Patch contributed by Kirill Miazine.
1585 TF/01 Do not log submission mode rewrites if they do not change the address.
1587 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1589 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1590 log files in place. Contributed by Roberto Lima.
1592 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1594 TF/03 Bugzilla 615: When checking the local_parts router precondition
1595 after a local_part_suffix or local_part_prefix option, Exim now
1596 does not use the address's named list lookup cache, since this
1597 contains cached lookups for the whole local part.
1599 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1600 Robert Millan. Documentation is in experimental-spec.txt.
1602 TF/04 Bugzilla 668: Fix parallel build (make -j).
1604 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1606 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1607 Patch provided by Jan Srzednicki.
1609 TF/05 Leading white space used to be stripped from $spam_report which
1610 wrecked the formatting. Now it is preserved.
1612 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1613 that they are available at delivery time.
1615 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1617 TF/08 TLS error reporting now respects the incoming_interface and
1618 incoming_port log selectors.
1620 TF/09 Produce a more useful error message if an SMTP transport's hosts
1621 setting expands to an empty string.
1623 NM/06 Bugzilla 744: EXPN did not work under TLS.
1624 Patch provided by Phil Pennock.
1626 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1627 Patch provided by Richard Godbee.
1629 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1630 acl_smtp_notquit, added index entry.
1632 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1633 Patch provided by Eugene Bujak.
1635 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1636 accept(). Patch provided by Maxim Dounin.
1638 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1639 Patch provided by Phil Pennock.
1641 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1643 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1644 Patch provided by Brad "anomie" Jorsch.
1646 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1647 Patch provided by Dean Brooks.
1649 TK/05 Add native DKIM support (does not depend on external libraries).
1651 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1652 Patch provided by Graeme Fowler.
1654 NM/16 Bugzilla 851: Documentation example syntax fix.
1656 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1658 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1661 NM/19 Bugzilla 745: TLS version reporting.
1662 Patch provided by Phil Pennock.
1664 NM/20 Bugzilla 167: bool: condition support.
1665 Patch provided by Phil Pennock.
1667 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1668 clients. Patch provided by Phil Pennock.
1670 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1671 Patch provided by Brad "anomie" Jorsch.
1673 NM/23 Bugzilla 687: Fix misparses in eximstats.
1674 Patch provided by Heiko Schlittermann.
1676 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1677 Patch provided by Heiko Schlittermann.
1679 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1680 plus update to original patch.
1682 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1684 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1685 Patch provided by David Brownlee.
1687 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1689 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1691 NM/30 Bugzilla 888: TLS documentation bugfixes.
1693 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1695 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1696 Unlike the original bugzilla I have changed all shell scripts in src tree.
1698 NM/33 Bugzilla 898: Transport filter timeout fix.
1699 Patch by Todd Rinaldo.
1701 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1702 Patch by Serge Demonchaux.
1704 NM/35 Bugzilla 39: Base64 decode bug fixes.
1705 Patch by Jakob Hirsch.
1707 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1709 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1711 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1713 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1719 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1720 ALT-N's libdkim that I have put here:
1721 http://duncanthrax.net/exim-experimental/
1723 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1724 were called 'true' and it seems that C99 defines that as a reserved
1725 keyword to be used with 'bool' variable types. That means you could
1726 not include C99-style headers which use bools without triggering
1727 build errors in sieve.c.
1729 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1730 as mailq or other aliases. Changed the --help handling significantly
1731 to do whats expected. exim_usage() emits usage/help information.
1733 SC/01 Added the -bylocaldomain option to eximstats.
1735 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1737 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1739 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1745 PH/01 Another patch from the Sieve maintainer.
1747 PH/02 When an IPv6 address is converted to a string for single-key lookup
1748 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1749 dots are used instead of colons so that keys in lsearch files need not
1750 contain colons. This was done some time before quoting was made available
1751 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1752 (notated using the quote facility) so as to distinguish them from IPv4
1753 keys. This meant that lookups for IP addresses in host lists did not work
1754 for iplsearch lookups.
1756 This has been fixed by arranging for IPv6 addresses to be expressed with
1757 colons if the lookup type is iplsearch. This is not incompatible, because
1758 previously such lookups could never work.
1760 The situation is now rather anomolous, since one *can* have colons in
1761 ordinary lsearch keys. However, making the change in all cases is
1762 incompatible and would probably break a number of configurations.
1764 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1767 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1768 conversion specification without a maximum field width, thereby enabling
1769 a rogue spamd server to cause a buffer overflow. While nobody in their
1770 right mind would setup Exim to query an untrusted spamd server, an
1771 attacker that gains access to a server running spamd could potentially
1772 exploit this vulnerability to run arbitrary code as the Exim user.
1774 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1775 $primary_hostname instead of what libspf2 thinks the hosts name is.
1777 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1778 a directory entry by the name of the lookup key. Previously, if a
1779 symlink pointed to a non-existing file or a file in a directory that
1780 Exim lacked permissions to read, a lookup for a key matching that
1781 symlink would fail. Now it is enough that a matching directory entry
1782 exists, symlink or not. (Bugzilla 503.)
1784 PH/03 The body_linecount and body_zerocount variables are now exported in the
1787 PH/04 Added the $dnslist_matched variable.
1789 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1790 This means they are set thereafter only if the connection becomes
1793 PH/06 Added the client_condition to authenticators so that some can be skipped
1794 by clients under certain conditions.
1796 PH/07 The error message for a badly-placed control=no_multiline_responses left
1797 "_responses" off the end of the name.
1799 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1801 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1802 (without spaces) instead of just copying the configuration text.
1804 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1806 PH/11 Added $max_received_linelength.
1808 PH/12 Added +ignore_defer and +include_defer to host lists.
1810 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1811 way in which PCRE > 7.0 is built.
1813 PH/14 Implemented queue_only_load_latch.
1815 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1816 MAIL command. The effect was to mangle the value on 64-bit systems.
1818 PH/16 Another patch from the Sieve maintainer.
1820 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1822 PH/18 If a system quota error occurred while trying to create the file for
1823 a maildir delivery, the message "Mailbox is full" was not appended to the
1824 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1825 only to a quota excession during the actual writing of the file.
1827 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1828 characters?) which causes problems in log lines. The DN values are now
1829 passed through string_printing() before being added to log lines.
1831 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1832 and InterBase are left for another time.)
1834 PH/21 Added message_body_newlines option.
1836 PH/22 Guard against possible overflow in moan_check_errorcopy().
1838 PH/23 POSIX allows open() to be a macro; guard against that.
1840 PH/24 If the recipient of an error message contained an @ in the local part
1841 (suitably quoted, of course), incorrect values were put in $domain and
1842 $local_part during the evaluation of errors_copy.
1848 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1849 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1852 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1853 issue a MAIL command.
1855 PH/02 In an ACL statement such as
1857 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1859 if a client was not listed at all, or was listed with a value other than
1860 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1861 the condition was not true (as it should be), so access was not denied.
1862 The bug was that the ! inversion was incorrectly passed on to the second
1863 item. This has been fixed.
1865 PH/03 Added additional dnslists conditions == and =& which are different from
1866 = and & when the dns lookup returns more than one IP address.
1868 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1869 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1871 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1872 FSYNC, which compiles an option called disable_fsync that allows for
1873 bypassing fsync(). The documentation is heavily laced with warnings.
1875 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1877 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1878 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1879 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1880 including adding "make clean"; (3) Added -fPIC when compiling the test
1881 dynamically loaded module, to get rid of a warning.
1883 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1884 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1885 = 0s. The bug is otherwise harmless.
1887 PH/07 There was a bug in the dovecot authenticator such that the value of
1888 $auth1 could be overwritten, and so not correctly preserved, after a
1889 successful authentication. This usually meant that the value preserved by
1890 the server_setid option was incorrect.
1892 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1894 PH/09 Installed PCRE release 7.0.
1896 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1897 run for batched SMTP input. It is now run at the start of every message
1898 in the batch. While fixing this I discovered that the process information
1899 (output by running exiwhat) was not always getting set for -bs and -bS
1900 input. This is fixed, and it now also says "batched" for BSMTP.
1902 PH/11 Added control=no_pipelining.
1904 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1905 patch, slightly modified), and move the expansion of helo_data till after
1906 the connection is made in the smtp transport (so it can use these
1909 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1911 PH/14 Added log_selector = +pid.
1913 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1915 PH/16 Add ${if forany and ${if forall.
1917 PH/17 Added dsn_from option to vary the From: line in DSNs.
1919 PH/18 Flush SMTP output before performing a callout, unless control =
1920 no_callout_flush is set.
1922 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1923 was true (the default) a successful delivery failed to delete the retry
1924 item, thus causing premature timeout of the address. The bug is now
1927 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1929 PH/21 Long custom messages for fakedefer and fakereject are now split up
1930 into multiline reponses in the same way that messages for "deny" and
1931 other ACL rejections are.
1933 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1934 with slight modification.
1936 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1937 draft, changing the syntax and factoring some duplicate code".
1939 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1940 for deliveries of the second and subsequent messages over the same SMTP
1943 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1944 ${reduce, with only minor "tidies".
1946 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1948 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1949 expansion side effects.
1951 PH/27 When a message times out after an over-quota error from an Exim-imposed
1952 quota, the bounce message says "mailbox is full". This message was not
1953 being given when it was a system quota that was exceeded. It now should
1956 MH/03 Made $recipients available in local_scan(). local_scan() already has
1957 better access to the recipient list through recipients_list[], but
1958 $recipients can be useful in postmaster-provided expansion strings.
1960 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1961 in the case of a MAIL command with additional options following the
1962 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1963 were accidentally chopped off.
1965 PH/29 SMTP synchronization checks are implemented when a command is read -
1966 there is a check that no more input is waiting when there shouldn't be
1967 any. However, for some commands, a delay in an ACL can mean that it is
1968 some time before the response is written. In this time, more input might
1969 arrive, invalidly. So now there are extra checks after an ACL has run for
1970 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1971 pipelining has not been advertised.
1973 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1975 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1976 returned in the SMTP response when smtp_return_error_details was false.
1977 This has been fixed.
1979 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1980 instead of the C I/O that was originally supplied, because problems were
1981 reported on Solaris.
1983 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1984 Exim which did not show up earlier: it was assuming that a call to
1985 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1986 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1987 was a macro that became an assignment, so it seemed to work. This has
1988 changed to a proper function call with a void return, hence the compile
1989 error. Exim's code has been fixed.
1991 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1994 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1996 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1998 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1999 because it turns out to be too big or there is a timeout) there is no
2000 "Completed" line in the log. When some messages of this type were
2001 selected by exigrep, they were listed as "not completed". Others were
2002 picked up by some special patterns. I have improved the selection
2003 criteria to be more general.
2005 PH/38 The host_find_failed option in the manualroute router can now be set
2006 to "ignore", to completely ignore a host whose IP address cannot be
2007 found. If all hosts are ignored, the behaviour is controlled by the new
2008 host_all_ignored option.
2010 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2011 homing or because of multiple MX records with /mx) generated more than
2012 one IP address, and the following item turned out to be the local host,
2013 all the secondary addresses of the first item were incorrectly removed
2014 from the list, along with the local host and any following hosts (which
2015 is what is supposed to happen).
2017 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2018 whoever called Exim into the -H file. In the case of the daemon it was
2019 behaving confusingly. When first started, it used values for whoever
2020 started the daemon, but after a SIGHUP it used the Exim user (because it
2021 calls itself on a restart). I have changed the code so that it now always
2024 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2025 message are rejected with the same error (e.g. no authentication or bad
2026 sender address), and a DATA command is nevertheless sent (as can happen
2027 with PIPELINING or a stupid MUA), the error message that was given to the
2028 RCPT commands is included in the rejection of the DATA command. This is
2029 intended to be helpful for MUAs that show only the final error to their
2032 PH/42 Another patch from the Sieve maintainer.
2034 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2035 Eximstats - Fixed some broken HTML links and added missing column headers
2037 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2038 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2040 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2042 PH/43 Yet another patch from the Sieve maintainer.
2044 PH/44 I found a way to check for a TCP/IP connection going away before sending
2045 the response to the final '.' that terminates a message, but only in the
2046 case where the client has not sent further data following the '.'
2047 (unfortunately, this is allowed). However, in many cases there won't be
2048 any further data because there won't be any more messages to send. A call
2049 to select() can be used: if it shows that the input is "ready", there is
2050 either input waiting, or the socket has been closed. An attempt to read
2051 the next input character can distinguish the two cases. Previously, Exim
2052 would have sent an OK response which the client would never have see.
2053 This could lead to message repetition. This fix should cure that, at
2054 least in a lot of common cases.
2056 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2057 advertised in response to EHLO.
2063 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2064 fixed by 4.65/MH/01 (is this a record?) are fixed:
2066 (i) An empty string was always treated as zero by the numeric comparison
2067 operators. This behaviour has been restored.
2069 (ii) It is documented that the numeric comparison operators always treat
2070 their arguments as decimal numbers. This was broken in that numbers
2071 starting with 0 were being interpreted as octal.
2073 While fixing these problems I realized that there was another issue that
2074 hadn't been noticed. Values of message_size_limit (both the global option
2075 and the transport option) were treated as octal if they started with 0.
2076 The documentation was vague. These values are now always treated as
2077 decimal, and I will make that clear in the documentation.
2083 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2084 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2087 MH/01 Don't check that the operands of numeric comparison operators are
2088 integers when their expansion is in "skipping" mode (fixes bug
2089 introduced by 4.64-PH/07).
2091 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2092 child addresses, Exim now panics and dies. Previously, because the count
2093 is held in a short int, deliveries were likely to be lost. As such a
2094 large number of recipients for a single message is ridiculous
2095 (performance will be very, very poor), I have chosen to impose a limit
2096 rather than extend the field.
2102 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2103 leftover -K file (the existence of which was triggered by #402).
2104 While we were at it, introduced process PID as part of the -K
2105 filename. This should rule out race conditions when creating
2108 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2109 processing considerably. Previous code took too long for large mails,
2110 triggering a timeout which in turn triggers #401.
2112 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2113 in the DK code in transports.c. sendfile() is not really portable,
2114 hence the _LINUX specificness.
2116 TF/01 In the add_headers option to the mail command in an Exim filter,
2117 there was a bug that Exim would claim a syntax error in any
2118 header after the first one which had an odd number of characters
2121 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2122 callout verification, Exim cached a "reject" for the entire domain. This
2123 is correct for most verifications, but it is not correct for a recipient
2124 verification with use_sender or use_postmaster set, because in that case
2125 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2126 case of MAIL FROM:<> rejection from other early rejections (e.g.
2127 rejection of HELO). When verifying a recipient using a non-null MAIL
2128 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2129 Whatever the result of the callout, the value of the domain cache is
2130 left unchanged (for any other kind of callout, getting as far as trying
2131 RCPT means that the domain itself is ok).
2133 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2136 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2137 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2138 session, the daemon ignores ECONNECT errors and logs others; it now
2139 ignores EPIPE as well.
2141 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2142 (quoted-printable decoding).
2144 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2145 later the small subsequent patch to fix an introduced bug.
2147 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2149 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2151 PH/08 An error is now given if message_size_limit is specified negative.
2153 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2154 to be given (somewhat) arbitrary names.
2156 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2159 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2160 miscellaneous code fixes
2162 PH/10 Added the log_reject_target ACL modifier to specify where to log
2165 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2166 hostname. This is wrong, because it relates to the incoming message (and
2167 probably the interface on which it is arriving) and not to the outgoing
2168 callout (which could be using a different interface). This has been
2169 changed to use the value of the helo_data option from the smtp transport
2170 instead - this is what is used when a message is actually being sent. If
2171 there is no remote transport (possible with a router that sets up host
2172 addresses), $smtp_active_hostname is used.
2174 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2175 tweaks were necessary in order to get it to work (see also 21 below):
2176 (a) The code assumed that strncpy() returns a negative number on buffer
2177 overflow, which isn't the case. Replaced with Exim's string_format()
2179 (b) There were several signed/unsigned issues. I just did the minimum
2180 hacking in of casts. There is scope for a larger refactoring.
2181 (c) The code used strcasecmp() which is not a standard C function.
2182 Replaced with Exim's strcmpic() function.
2183 (d) The code set only $1; it now sets $auth1 as well.
2184 (e) A simple test gave the error "authentication client didn't specify
2185 service in request". It would seem that Dovecot has changed its
2186 interface. Fortunately there's a specification; I followed it and
2187 changed what the client sends and it appears to be working now.
2189 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2192 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2193 address A is aliased to B and C, where B exists and C does not. Without
2194 -v the output is "A verified" because verification stops after a
2195 successful redirection if more than one address is generated. However,
2196 with -v the child addresses are also verified. Exim was outputting "A
2197 failed to verify" and then showing the successful verification for C,
2198 with its parentage. It now outputs "B failed to verify", showing B's
2199 parentage before showing the successful verification of C.
2201 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2202 look up a TXT record in a specific list after matching in a combined
2205 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2206 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2207 they consult the DNS. I had assumed they would set it the way they
2208 wanted; and indeed my experiments on Linux seem to show that in some
2209 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2210 To be on the safe side, however, I have now made the interface to
2211 host_find_byname() similar to host_find_bydns(), with an argument
2212 containing the DNS resolver options. The host_find_byname() function now
2213 sets these options at its start, just as host_find_bydns() does. The smtp
2214 transport options dns_qualify_single and dns_search_parents are passed to
2215 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2216 of host_find_byname() use the default settings of RES_DEFNAMES
2217 (qualify_single) but not RES_DNSRCH (search_parents).
2219 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2220 spool_read_header() do less string testing, by means of a preliminary
2221 switch on the second character of optional "-foo" lines. (This is
2222 overdue, caused by the large number of possibilities that now exist.
2223 Originally there were few.) While I was there, I also converted the
2224 str(n)cmp tests so they don't re-test the leading "-" and the first
2225 character, in the hope this might squeeze out yet more improvement.
2227 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2228 flag allowing group syntax was set by the header_syntax check but not
2229 turned off, possible causing trouble later; (2) The flag was not being
2230 set at all for the header_verify test, causing "group"-style headers to
2231 be rejected. I have now set it in this case, and also caused header_
2232 verify to ignore an empty address taken from a group. While doing this, I
2233 came across some other cases where the code for allowing group syntax
2234 while scanning a header line wasn't quite right (mostly, not resetting
2235 the flag correctly in the right place). These bugs could have caused
2236 trouble for malformed header lines. I hope it is now all correct.
2238 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2239 with the "reply" argument non-NULL. The code, however (which originally
2240 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2241 but it didn't always do it. This confused somebody who was copying the
2242 code for some other use. I have removed all the tests.
2244 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2245 feature that was used to support insecure browsers during the U.S. crypto
2246 embargo. It requires special client support, and Exim is probably the
2247 only MTA that supported it -- and would never use it because real RSA is
2248 always available. This code has been removed, because it had the bad
2249 effect of slowing Exim down by computing (never used) parameters for the
2250 RSA_EXPORT functionality.
2252 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2253 authenticator to fail if there's a tab character in the incoming data
2254 (there should never be unless someone is messing about, as it's supposed
2255 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2256 if the connection is using TLS or if the remote IP is the same as the
2257 local IP, and the "valid-client-cert option" if a client certificate has
2260 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2261 authenticators. This can be used for authorization after authentication
2262 succeeds. (In the case of plaintext, it servers for both authentication
2265 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2266 if any retry times were supplied.
2268 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2269 connection, where there is no client IP address to check. In this
2270 situation, the verify now always succeeds.
2272 PH/25 Applied John Jetmore's -Mset patch.
2274 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2276 PH/27 In a string expansion for a processed (not raw) header when multiple
2277 headers of the same name were present, leading whitespace was being
2278 removed from all of them, but trailing whitespace was being removed only
2279 from the last one. Now trailing whitespace is removed from each header
2280 before concatenation. Completely empty headers in a concatenation (as
2281 before) are ignored.
2283 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2284 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2286 PH/29 [Removed. This was a change that I later backed out, and forgot to
2287 correct the ChangeLog entry (that I had efficiently created) before
2288 committing the later change.]
2290 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2291 address errors (4xx response to RCPT) over the same connection as other
2292 messages routed to the same hosts. Such deliveries are always "forced",
2293 so retry times are not inspected. This resulted in far too many retries
2294 for the affected addresses. The effect occurred only when there were more
2295 hosts than the hosts_max_try setting in the smtp transport when it had
2296 the 4xx errors. Those hosts that it had tried were not added to the list
2297 of hosts for which the message was waiting, so if all were tried, there
2298 was no problem. Two fixes have been applied:
2300 (i) If there are any address or message errors in an SMTP delivery, none
2301 of the hosts (tried or untried) are now added to the list of hosts
2302 for which the message is waiting, so the message should not be a
2303 candidate for sending over the same connection that was used for a
2304 successful delivery of some other message. This seems entirely
2305 reasonable: after all the message is NOT "waiting for some host".
2306 This is so "obvious" that I'm not sure why it wasn't done
2307 previously. Hope I haven't missed anything, but it can't do any
2308 harm, as the worst effect is to miss an optimization.
2310 (ii) If, despite (i), such a delivery is accidentally attempted, the
2311 routing retry time is respected, so at least it doesn't keep
2312 hammering the server.
2314 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2315 in ${readsocket because some servers need this prod.
2317 PH/32 Added some extra debug output when updating a wait-xxx database.
2319 PH/33 The hint "could be header name not terminated by colon", which has been
2320 given for certain expansion errors for a long time, was not being given
2321 for the ${if def:h_colon_omitted{... case.
2323 PH/34 The spec says: "With one important exception, whenever a domain list is
2324 being scanned, $domain contains the subject domain." There was at least
2325 one case where this was not true.
2327 PH/35 The error "getsockname() failed: connection reset by peer" was being
2328 written to the panic log as well as the main log, but it isn't really
2329 panic-worthy as it just means the connection died rather early on. I have
2330 removed the panic log writing for the ECONNRESET error when getsockname()
2333 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2334 runs only) independently of the message's sender address. This meant
2335 that, if the 4xx error was in fact related to the sender, a different
2336 message to the same recipient with a different sender could confuse
2337 things. In particualar, this can happen when sending to a greylisting
2338 server, but other circumstances could also provoke similar problems.
2339 I have changed the default so that the retry time for these errors is now
2340 based a combination of the sender and recipient addresses. This change
2341 can be overridden by setting address_retry_include_sender=false in the
2344 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2345 remote server are returned as part of bounce messages. This was not
2346 happening for LMTP over a pipe (the lmtp transport), but now it is the
2347 same for both kinds of LMTP.
2349 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2350 in header lines that were in fact CNAMEs. This is no longer the case.
2352 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2353 and queue runs started by the daemon processed all messages. This has
2354 been fixed so that -R and -S can now usefully be given with -q<time>.
2356 PH/40 Import PCRE release 6.7 (fixes some bugs).
2358 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2360 PH/42 Give an error if -q is specified more than once.
2362 PH/43 Renamed the variables $interface_address and $interface_port as
2363 $received_ip_address and $received_port, to make it clear that these
2364 values apply to message reception, and not to the outgoing interface when
2365 a message is delivered. (The old names remain recognized, of course.)
2367 PH/44 There was no timeout on the connect() call when using a Unix domain
2368 socket in the ${readsocket expansion. There now is.
2370 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2371 be meaningful with "accept".
2373 SC/01 Eximstats V1.43
2374 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2376 SC/02 Eximstats V1.44
2377 Use a glob alias rather than an array ref in the generated
2378 parser. This improves both readability and performance.
2380 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2381 Collect SpamAssassin and rejection statistics.
2382 Don't display local sender or destination tables unless
2383 there is data to show.
2384 Added average volumes into the top table text output.
2386 SC/04 Eximstats V1.46
2387 Collect data on the number of addresses (recipients)
2388 as well as the number of messages.
2390 SC/05 Eximstats V1.47
2391 Added 'Message too big' to the list of mail rejection
2392 reasons (thanks to Marco Gaiarin).
2394 SC/06 Eximstats V1.48
2395 Mainlog lines which have GMT offsets and are too short to
2396 have a flag are now skipped.
2398 SC/07 Eximstats V1.49 (Alain Williams)
2399 Added the -emptyok flag.
2401 SC/08 Eximstats V1.50
2402 Fixes for obtaining the IP address from reject messages.
2404 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2405 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2406 whitesspace changes from 4.64-PH/27
2408 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2411 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2412 are found, allow negative numbers in numeric criteria)
2414 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2416 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2417 to match changes made in 4.64-PH/43
2419 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2421 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2422 "domain" is actually an IP address, and force a failure. This locks out
2423 those revolvers/nameservers that support "A-for-A" lookups, in
2424 contravention of the specifications.
2426 PH/48 When a host name was looked up from an IP address, and the subsequent
2427 forward lookup of the name timed out, the host name was left in
2428 $sender_host_name, contrary to the specification.
2430 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2431 restricted to single-key lookups, Exim was not diagnosing an error if
2432 * or *@ was used with a query-style lookup.
2434 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2436 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2437 long ago, but noone interested enough thought of it. Let's just say that
2438 the "1.1" means that there are some new functions that weren't there at
2439 some point in the past.
2441 PH/51 Error processing for expansion failure of helo_data from an smtp
2442 transport during callout processing was broken.
2444 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2445 tested/used via the -bh/-bhc/-bs options.
2447 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2448 bug, fixed in subsequent PCRE releases).
2450 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2451 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2453 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2459 SC/01 Use a glob alias rather than an array ref in eximstats generated
2460 parser. This improves both readability and performance.
2462 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2463 Don't display local sender or destination tables in eximstats unless
2464 there is data to show.
2465 Added average volumes into the eximstats top table text output.
2467 SC/03 Collect data on the number of addresses (recipients) as well
2468 as the number of messages in eximstats.
2470 TF/01 Correct an error in the documentation for the redirect router. Exim
2471 does (usually) call initgroups() when daemonizing.
2473 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2474 with consistent privilege compared to when running as a daemon.
2476 TF/03 Note in the spec that $authenticated_id is not set for local
2477 submissions from trusted users.
2479 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2480 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2482 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2483 by adding some example configuration directives to the default
2484 configuration file. A little bit of work is required to uncomment the
2485 directives and define how usernames and passwords are checked, but
2486 there is now a framework to start from.
2488 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2489 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2490 without this. I don't know how relevant this is to other LDAP libraries.
2492 PH/02 Add the verb name to the "unknown ACL verb" error.
2494 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2496 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2498 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2499 directory not expanded when it should be if an expanded home directory
2500 was set for the address (which is overridden by the transport).
2502 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2505 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2506 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2507 because it is too late at that time, and has no effect.
2509 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2510 security issue with \' (bugzilla #107). I could not use the
2511 PQescapeStringConn() function, because it needs a PGconn value as one of
2514 PH/08 When testing addresses using -bt, indicate those final addresses that
2515 are duplicates that would not cause an additional delivery. At least one
2516 person was confused, thinking that -bt output corresponded to deliveries.
2517 (Suppressing duplicates isn't a good idea as you lose the information
2518 about possibly different redirections that led to the duplicates.)
2520 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2521 systems where poll() doesn't work, in particular OS X.
2523 PH/10 Added more information to debugging output for retry time not reached.
2525 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2526 operations in malware.c.
2528 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2531 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2532 syslog_duplication set false, log lines that would normally be written
2533 both the the main log and to the reject log were not written to syslog at
2536 PH/14 In the default configuration, change the use of "message" in ACL warn
2537 statements to "add_header".
2539 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2540 not followed by a command (e.g. "seen endif").
2542 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2543 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2546 PH/17 Added extra conditions to the default value of delay_warning_condition
2550 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2551 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2552 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2555 The Auto-Submitted: and various List- headers are standardised, whereas I
2556 don't think Precedence: ever was.
2558 PH/18 Refactored debugging code in route_finduser() to show more information,
2559 in particular, the error code if getpwnam() issues one.
2561 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2562 This is apparently needed in addition to the PH/07 change above to avoid
2563 any possible encoding problems.
2565 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2566 but not after initializing Perl.
2568 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2569 output them only if debugging. By default they are written stderr,
2570 apparently, which is not desirable.
2572 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2575 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2578 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2580 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2581 authenticated or an ident call has been made. Suppress the default
2582 values for $authenticated_id and $authenticated_sender (but permit -oMai
2583 and -oMas) when testing with -bh.
2585 PH/24 Re-jigged the order of the tests in the default configuration so that the
2586 tests for valid domains and recipients precede the DNS black list and CSA
2587 tests, on the grounds that those ones are more expensive.
2589 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2590 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2591 This bug exists in every version of Exim that I still have, right back to
2594 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2595 However, an attempt to turn on case-sensitivity in a regex key by
2596 including (?-i) didn't work because the subject string was already
2597 lowercased, and the effects were non-intuitive. It turns out that a
2598 one-line patch can be used to allow (?-i) to work as expected.
2604 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2605 other effects) broke the use of negated acl sub-conditions.
2607 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2610 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2611 "Deny" causes Exim to reject the incoming connection with a 554 error.
2612 Unfortunately, if there is a major crisis, such as a disk failure,
2613 tcp-wrappers gives "deny", whereas what one would like would be some
2614 kind of temporary error. A kludge has been added to help with this.
2615 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2616 554 error is used if errno is still zero or contains ENOENT (which occurs
2617 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2620 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2622 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2623 errors. Otherwise a message that provokes a temporary error (when other
2624 messages do not) can cause a whole host to time out.
2626 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2627 addresses were routed directly to files or pipes from a redirect router.
2628 File deliveries just didn't batch; pipe deliveries might have suffered
2631 PH/06 A failure to get a lock for a hints database would erroneously always say
2632 "Failed to get write lock", even when it was really a read lock.
2634 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2635 of 0600. This has been changed to use the value of the lockfile_mode
2636 option (which defaults to 0600).
2638 PH/08 Applied small patch from the Sieve maintainer.
2640 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2641 folder from quota calculations, a direct delivery into this folder messed
2642 up the contents of the maildirsize file. This was because the regex was
2643 used only to exclude .Trash (or whatever) when the size of the mailbox
2644 was calculated. There was no check that a delivery was happening into an
2645 excluded directory. This bug has been fixed by ignoring all quota
2646 processing for deliveries into excluded directories.
2648 PH/10 Added the maildirfolder_create_regex option to appendfile.
2654 PH/01 The code for finding all the local interface addresses on a FreeBSD
2655 system running IPv6 was broken. This may well have applied to all BSD
2656 systems, as well as to others that have similar system calls. The broken
2657 code found IPv4 interfaces correctly, but gave incorrect values for the
2658 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2659 that it would not match correctly against @[] and not recognize the IPv6
2662 PH/02 The ipliteral router was not recognizing addresses of the form user@
2663 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2665 PH/03 Added disable_ipv6.
2667 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2668 decoded form, because it is most often used to construct To: headers
2669 lines in autoreplies, and the decoded form may well be syntactically
2670 invalid. However, $reply_address has leading white space removed, and all
2671 newlines turned into spaces so that the autoreply transport does not
2674 PH/05 If group was specified without a user on a router, and no group or user
2675 was specified on a transport, the group from the router was ignored.
2677 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2678 for visible compile-time settings that can be used to change these
2679 numbers, for those that want even more. Backwards compatibility with old
2680 spool files has been maintained. However, going back to a previous Exim
2681 release will lost any variables that are in spool files.
2683 PH/07 Two small changes when running in the test harness: increase delay when
2684 passing a TCP/IP connection to a new process, in case the original
2685 process has to generate a bounce, and remove special handling of
2686 127.0.0.2 (sic), which is no longer necessary.
2688 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2689 be the same on different OS.
2691 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2694 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2695 whether --show-vars was specified or not
2697 JJ/02 exipick: Added support for new ACL variable spool format introduced
2700 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2701 syntactically invalid From: or Reply-to: line, and a filter used this to
2702 generate an autoreply, and therefore failed to obtain an address for the
2703 autoreply, Exim could try to deliver to a non-existent relative file
2704 name, causing unrelated and misleading errors. What now happens is that
2705 it logs this as a hard delivery error, but does not attempt to create a
2708 PH/11 The exinext utility has a -C option for testing purposes, but although
2709 the given file was scanned by exinext itself; it wasn't being passed on
2710 when Exim was called.
2712 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2713 an end-of-file indication when reading a command response.
2715 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2716 compiled. In many other places in Exim, IPv6 addresses are always
2717 recognized, so I have changed this. It also means that IPv4 domain
2718 literals of the form [IPV4:n.n.n.n] are now always recognized.
2720 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2721 used if the router is not running as root, for example, when verifying at
2722 ACL time, or when using -bh. The debugging output from this situation was
2723 non-existent - all you got was a failure to exec. I have made two
2726 (a) Failures to set uid/gid, the current directory, or a process leader
2727 in a subprocess such as that created by queryprogram now generate
2728 suitable debugging ouput when -d is set.
2730 (b) The queryprogram router detects when it is not running as root,
2731 outputs suitable debugging information if -d is set, and then runs
2732 the subprocess without attempting to change uid/gid.
2734 PH/15 Minor change to Makefile for building test_host (undocumented testing
2737 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2738 additional section of a DNS packet that returns MX or SRV records.
2739 Instead, it always explicitly searches for A/AAAA records. This avoids
2740 major problems that occur when a DNS server includes only records of one
2741 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2742 fixed another bug: if SRV records were looked up and the corresponding
2743 address records were *not* found in the additional section, the port
2744 values from the SRV records were lost.
2746 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2747 using the correct key (the original address) when searching the retry
2748 rules in order to find which one to use for generating the retry hint.
2750 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2751 adding the default one. Similarly, if it contains a Reply-To: header, the
2752 errors_reply_to option, if set, is not used.
2754 PH/19 When calculating a retry time, Exim used to measure the "time since
2755 failure" by looking at the "first failed" field in the retry record. Now
2756 it does not use this if it is later than than the arrival time of the
2757 message. Instead it uses the arrival time. This makes for better
2758 behaviour in cases where some deliveries succeed, thus re-setting the
2759 "first failed" field. An example is a quota failure for a huge message
2760 when small messages continue to be delivered. Without this change, the
2761 "time since failure" will always be short, possible causing more frequent
2762 delivery attempts for the huge message than are intended.
2763 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2765 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2766 $1, $2, $3) because the numerical variables can be reset during some
2767 expansion items (e.g. "match"), thereby losing the authentication data.
2769 PH/21 Make -bV show the size of off_t variables so that the test suite can
2770 decide whether to run tests for quotas > 2G.
2772 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2773 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2774 filecount value is greater than 2G or if a quota value is greater than 2G
2775 on a system where the size of off_t is not greater than 4, a panic error
2778 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2779 never match. The debug and -bh output now contains an explicit error
2780 message indicating a malformed IPv4 address or mask.
2782 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2783 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2784 PH/23 above applies.
2786 PH/25 Do not write to syslog when running in the test harness. The only
2787 occasion when this arises is a failure to open the main or panic logs
2788 (for which there is an explicit test).
2790 PH/26 Added the /no_tell option to "control=freeze".
2792 PH/27 If a host name lookup failed very early in a connection, for example, if
2793 the IP address matched host_lookup and the reverse lookup yielded a name
2794 that did not have a forward lookup, an error message of the form "no IP
2795 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2796 could be logged. Now it outputs the IP address instead of "NULL".
2798 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2799 allows the sender and the authenticated sender to be set when
2800 submitting a message from within Exim. Since child_open_exim() is
2801 documented for local_scan(), the new function should be too.
2803 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2804 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2805 results in an empty string is now treated as unset.
2807 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2809 PH/31 Added sender_verify_fail logging option.
2811 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2812 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2813 tidied the source and removed it altogether.
2815 PH/33 When a queue run was abandoned because the load average was too high, a
2816 log line was always written; now it is written only if the queue_run log
2817 selector is set. In addition, the log line for abandonment now contains
2818 information about the queue run such as the pid. This is always present
2819 in "start" and "stop" lines but was omitted from the "abandon" line.
2821 PH/34 Omit spaces between a header name and the colon in the error message that
2822 is given when verify = headers_syntax fails (if there are lots of them,
2823 the message gets confusing).
2825 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2826 names, as there are now some PTR records that contain slashes. This check
2827 is only to protect against broken name servers that fall over on strange
2828 characters, so the fact that it applies to all lookups doesn't matter.
2830 PH/36 Now that the new test suite is complete, we can remove some of the
2831 special code in Exim that was needed for the old test suite. For example,
2832 sorting DNS records because real resolvers return them in an arbitrary
2833 order. The new test suite's fake resolver always returns records in the
2836 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2837 bounces) except when queue_only is set, to avoid logging races between
2838 the different processes.
2840 PH/38 Panic-die if .include specifies a non-absolute path.
2842 PH/39 A tweak to the "H" retry rule from its user.
2844 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2845 a label. They prevented compilation on older perls.
2847 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2848 a warning to be raised on newish perls.
2850 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2851 on queue. Changes to match documented behaviour of showing count of
2852 messages matching specified criteria.
2854 PH/40 Changed the default ident timeout from 30s to 5s.
2856 PH/41 Added support for the use of login_cap features, on those BSD systems
2857 that have them, for controlling the resources used by pipe deliveries.
2859 PH/42 The content-scanning code uses fopen() to create files in which to put
2860 message data. Previously it was not paying any attention to the mode of
2861 the files. Exim runs with umask(0) because the rest of the code creates
2862 files with open(), and sets the required mode explicitly. Thus, these
2863 files were ending up world-writeable. This was not a big issue, because,
2864 being within the spool directory, they were not world-accessible. I have
2865 created a function called modefopen, which takes an additional mode
2866 argument. It sets umask(777), creates the file, chmods it to the required
2867 mode, then resets the umask. All the relevant calls to fopen() in the
2868 content scanning code have been changed to use this function.
2870 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2871 to 24 hours. This avoids potential overflow problems when processing G
2872 and H retry rules. I suspect nobody ever tinkers with this value.
2874 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2876 PH/45 When the plaintext authenticator is running as a client, the server's
2877 challenges are checked to ensure they are valid base64 strings. By
2878 default, the authentication attempt is cancelled if an invalid string is
2879 received. Setting client_ignore_invalid_base64 true ignores these errors.
2880 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2881 they are received. Thus, the responses can be made to depend on the
2882 challenges. If an invalid string is ignored, an empty string is placed in
2885 PH/46 Messages that are created by the autoreply transport now contains a
2886 References: header, in accordance with RFCs 2822 and 3834.
2888 PH/47 Added authenticated_sender_force to the smtp transport.
2890 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2892 PH/49 Installed latest patch from the Sieve maintainer.
2894 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2895 was also set, the appendfile transport was unnecessarily scanning a
2896 directory of message files (e.g. for maildir delivery) to find the count
2897 of files (along with the size), even though it did not need this
2898 information. It now does the scan only if it needs to find either the
2899 size of the count of files.
2901 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2903 PH/52 Two bugs concerned with error handling when the smtp transport is
2906 (i) Exim was not creating retry information for temporary errors given
2907 for individual recipients after the DATA command when the smtp transport
2908 was used in LMTP mode. This meant that they could be retried too
2909 frequently, and not timed out correctly.
2911 (ii) Exim was setting the flag that allows error details to be returned
2912 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2913 recipients that were returned after the DATA command.
2915 PH/53 This is related to PH/52, but is more general: for any failing address,
2916 when detailed error information was permitted to be returned to the
2917 sender, but the error was temporary, then after the final timeout, only
2918 "retry timeout exceeded" was returned. Now it returns the full error as
2919 well as "retry timeout exceeded".
2921 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2922 do this, and (what is worse) MTAs that accept it.
2924 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2925 will now be deprecated.
2927 PH/56 New os.c-cygwin from the Cygwin maintainer.
2929 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2930 formats (previously only available in exim formats via -bpr, -bpru,
2931 and -bpra. Now also available in native and exiqgrep formats)
2933 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2934 with very large, slow to parse queues
2936 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2938 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2940 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2941 responses to authentication challenges, though it was showing the
2942 challenges; (ii) I've removed the CR characters from the debug output for
2945 PH/58 Allow for the insertion of a newline as well as a space when a string
2946 is turned into more than one encoded-word during RFC 2047 encoding. The
2947 Sieve code now uses this.
2949 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2950 data_4xx, lost_connection, tls_required.
2952 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2953 message was being sent as an SMTP response.
2955 PH/61 Add -l and -k options to exicyclog.
2957 PH/62 When verifying, if an address was redirected to one new address, so that
2958 verification continued, and the new address failed or deferred after
2959 having set something in $address_data, the value of $address_data was not
2960 passed back to the ACL. This was different to the case when no
2961 redirection occurred. The value is now passed back in both cases.
2963 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2964 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2965 use login_cap.h, so on its own it isn't the distinguishing feature. The
2966 new name refers directly to the setclassresources() function.
2968 PH/65 Added configuration files for NetBSD3.
2970 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2972 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2973 is preferred over IPv4.
2975 PH/68 The bounce_return_message and bounce_return_body options were not being
2976 honoured for bounces generated during the reception of non-SMTP messages.
2977 In particular, this applied to messages rejected by the ACL. This bug has
2978 been fixed. However, if bounce_return_message is true and bounce_return_
2979 body is false, the headers that are returned for a non-SMTP message
2980 include only those that have been read before the error was detected.
2981 (In the case of an ACL rejection, they have all been read.)
2983 PH/69 The HTML version of the specification is now built in a directory called
2984 spec_html instead of spec.html, because the latter looks like a path with
2985 a MIME-type, and this confuses some software.
2987 PH/70 Catch two compiler warnings in sieve.c.
2989 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2990 function verify_get_ident() calls ip_connect() to connect a socket, but
2991 if the "connect()" function timed out, ip_connect() used to close the
2992 socket. However, verify_get_ident() also closes the socket later, and in
2993 between Exim writes to the log, which may get opened at this point. When
2994 the socket was closed in ip_connect(), the log could get the same file
2995 descriptor number as the socket. This naturally causes chaos. The fix is
2996 not to close the socket in ip_connect(); the socket should be closed by
2997 the function that creates it. There was only one place in the code where
2998 this was missing, in the iplookup router, which I don't think anybody now
2999 uses, but I've fixed it anyway.
3001 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3002 well as to direct DNS lookups. Otherwise the handling of names in host
3003 lists is inconsistent and therefore confusing.
3009 PH/01 Two changes to the default runtime configuration:
3011 (1) Move the checks for relay_from_hosts and authenticated clients from
3012 after to before the (commented out) DNS black list checks.
3014 (2) Add control=submission to the relay_from_hosts and authenticated
3015 clients checks, on the grounds that messages accepted by these
3016 statements are most likely to be submissions.
3018 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3020 (1) Generate an error if the third argument for the ${prvs expansion is
3023 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3026 (3) Reset the variables that are obtained from the first argument of
3027 ${prvscheck and used in the second argument before leaving the code,
3028 because their memory is reclaimed, so using them afterwards may do
3031 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3032 one (it's much easier than Tom thought :-).
3034 (5) Because of (4), we can now allow for the use of $prvscheck_result
3035 inside the third argument.
3037 PH/03 For some reason, the default setting of PATH when running a command from
3038 a pipe transport was just "/usr/bin". I have changed it to
3041 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3042 anything to be listed in the output from -bV.
3044 PH/05 When a filter generated an autoreply, the entire To: header line was
3045 quoted in the delivery log line, like this:
3047 => >A.N.Other <ano@some.domain> <original@ddress> ...
3049 This has been changed so that it extracts the operative address. There
3050 may be more than one such address. If so, they are comma-separated, like
3053 => >ano@some.domain,ona@other.domain <original@ddress> ...
3055 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3056 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3057 not being looked up in the rDNS to get a host name, Exim was showing the
3058 IP address twice in Received: lines, even though the IP addresses were
3059 identical. For example:
3061 Received: from [1.2.3.4] (helo=[1.2.3.4])
3063 However, if the real host name was known, it was omitting the HELO data
3064 if it matched the actual IP address. This has been tidied up so that it
3065 doesn't show the same IP address twice.
3067 PH/07 When both +timestamp and +memory debugging was on, the value given by
3068 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3069 called by the debug printing, thereby overwriting the timestamp buffer.
3070 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3072 PH/08 When the original message was included in an autoreply transport, it
3073 always said "this is a copy of the message, including all the headers",
3074 even if body_only or headers_only was set. It now gives an appropriate
3077 PH/09 Applied a patch from the Sieve maintainer which:
3079 o fixes some comments
3080 o adds the (disabled) notify extension core
3081 o adds some debug output for the result of if/elsif tests
3082 o points to the current vacation draft in the documentation
3083 and documents the missing references header update
3087 o fixes a bug in processing the envelope test (when testing
3088 multiple envelope elements, the last element determinted the
3091 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3092 Electronic Mail") by including:
3094 Auto-submitted: auto-generated
3096 in the messages that it generates (bounce messages and others, such as
3097 warnings). In the case of bounce messages for non-SMTP mesages, there was
3098 also a typo: it was using "Auto_submitted" (underscore instead of
3099 hyphen). Since every message generated by Exim is necessarily in response
3100 to another message, thes have all been changed to:
3102 Auto-Submitted: auto-replied
3104 in accordance with these statements in the RFC:
3106 The auto-replied keyword:
3108 - SHOULD be used on messages sent in direct response to another
3109 message by an automatic process,
3111 - MUST NOT be used on manually-generated messages,
3113 - MAY be used on Delivery Status Notifications (DSNs) and Message
3114 Disposition Notifications (MDNs),
3116 - MUST NOT be used on messages generated by automatic or periodic
3117 processes, except for messages which are automatic responses to
3120 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3121 to the default Received: header definition.
3123 PH/12 Added log selector acl_warn_skipped (default on).
3125 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3126 variables because (a) they are in the wrong storage pool and (b) even if
3127 they were copied, it wouldn't work properly because of the caching.
3129 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3130 checking when decoding. Apparently there are clients that generate
3131 overlong encoded strings. Why am I not surprised?
3133 PH/15 If the first argument of "${if match_address" was not empty, but did not
3134 contain an "@" character, Exim crashed. Now it writes a panic log message
3135 and treats the condition as false.
3137 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3139 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3140 extension "envelope-auth". The code is finished and in agreement with
3141 other implementations, but there is no documentation so far and in fact,
3142 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3143 not changing the active code.
3145 Print executed "if" and "elsif" statements when debugging is used. This
3146 helps a great deal to understand what a filter does.
3148 Document more things not specified clearly in RFC3028. I had all this
3149 sorted out, when out of a sudden new issues came to my mind. Oops."
3151 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3154 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3155 canonical form (as documented). However, after a host name lookup from
3156 the IP address, check_host() was doing a simple string comparison with
3157 addresses acquired from the DNS when checking that the found name did
3158 have the original IP as one of its addresses. Since any found IPv6
3159 addresses are likely to be in abbreviated form, the comparison could
3160 fail. Luckily, there already exists a function for doing the comparison
3161 by converting both addresses to binary, so now that is used instead of
3162 the text comparison.
3164 PH/20 There was another similar case to PH/19, when a complete host name was
3165 given in a host list; looking up its IP address could give an abbreviated
3166 form, whereas the current host's name might or might not be abbreviated.
3167 The same fix has been applied.
3173 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3174 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3177 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3178 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3180 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3182 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3183 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3184 Also, for "Auto-Submitted", treat anything other than "no" as
3185 non-personal, in accordance with RFC 3834. (Previously it treated
3186 anything starting "auto-" as non-personal.)
3188 TF/01 The control=submission/name=... option had a problem with syntax
3189 errors if the name included a slash character. The /name= option
3190 now slurps the rest of the string, so it can include any characters
3191 but it must come last in the list of options (after /sender_retain
3194 PH/05 Some modifications to the interface to the fake nameserver for the new
3202 TK/01 Added the "success_on_redirect" address verification option. See
3203 NewStuff for rationale and an example.
3205 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3207 PH/02 Patch to exigrep to allow it to work on syslog lines.
3209 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3210 fread() to skip over the body file's header line, because in Cygwin the
3211 header line is locked and is inaccessible.
3213 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3214 co-exist for some time) to make it clear that it is the Exim ID that is
3215 referenced, not the Message-ID: header line.
3217 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3218 string_format() function, because snprintf() does not exist on all
3221 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3222 ${stat: expansion item.
3224 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3225 protocol synchronization error", to keep the pedants happy.
3227 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3228 well as for IRIX systems, when gcc is being used. See the host.c source
3231 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3233 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3236 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3237 transport and to the smtp transport in LMTP mode.
3239 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3241 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3242 run a filter in a subprocess. This could lead to confusion in subsequent
3243 lookups in the parent process. There should also be a search_tidyup() at
3244 the end of the subprocess.
3246 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3247 only if the host matched helo_try_verify_hosts, which caused the
3248 verification to occur when the EHLO/HELO command was issued. The ACL just
3249 tested the remembered result. Now, if a previous verification attempt has
3250 not happened, "verify = helo" does it there and then.
3252 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3254 TK/03 Fix log output including CR from clamd.
3256 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3257 exist provoked a memory error which could cause a segfault.
3259 PH/15 Installed PCRE 6.2
3261 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3263 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3264 of the problem. Specifically, suggested +O2 rather than +O1 for the
3267 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3269 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3270 record, the port was not correctly logged when the outgoing_port log
3271 selector was set (it logged the transort's default port).
3273 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3274 fallback_hosts, and "hosts" in the smtp transport.
3276 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3277 host errors such as "Connection refused".
3279 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3280 authentication with radiusclient 0.4.9:
3282 - Error returned from rc_read_config was caught wrongly
3283 - Username/password not passed on to radius server due to wrong length.
3285 The presumption is that some radiusclient API changes for 4.51/PH/17
3286 were not taken care of correctly. The code is still untested by me (my
3287 Linux distribution still has 0.3.2 of radiusclient), but it was
3288 contributed by a Radius user.
3290 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3291 expanding the "port" option of the smtp transport.
3293 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3294 while reading a MIME header. Thanks to Tom Hughes for a patch.
3296 PH/24 Include config.h inside local_scan.h so that configuration settings are
3299 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3300 that in an ACL for RCPT (for example), you can examine exactly what was
3303 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3304 commands, but it was not correctly comparing the address with the actual
3305 client host address. Thus, it would show the EHLO address in Received:
3306 header lines when this was not necessary.
3308 PH/27 Added the % operator to ${eval:}.
3310 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3311 it should be ignoring failures (because with -C, for example, it has lost
3312 privilege). It wasn't ignoring creation failures other than "already
3315 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3318 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3319 because an input line was too long, either on its own, or by virtue of
3320 too many continuations, the temporary file was not being removed, and the
3321 return code was incorrect.
3323 PH/31 Missing "BOOL" in function definition in filtertest.c.
3325 PH/32 Applied Sieve patches from the maintainer.
3327 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3329 PH/33 Added "verify = not_blind".
3331 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3332 Local/Makefile (with some defaults set). These are used in built scripts
3333 such as exicyclog, but they have never been used in the exim_install
3334 script (though there are many overriding facilities there). I have
3335 arranged that the exim_install script now takes note of these two
3338 PH/35 Installed configuration files for Dragonfly.
3340 PH/36 When a locally submitted message by a trusted user did not contain a
3341 From: header, and the sender address was obtained from -f or from an SMTP
3342 MAIL command, and the trusted user did not use -F to supply a sender
3343 name, $originator_name was incorrectly used when constructing a From:
3344 header. Furthermore, $originator_name was used for submission mode
3345 messages from external hosts without From: headers in a similar way,
3346 which is clearly wrong.
3348 PH/37 Added control=suppress_local_fixups.
3350 PH/38 When log_selector = +received_sender was set, and the addition of the
3351 sender made the log line's construction buffer exactly full, or one byte
3352 less than full, an overflow happened when the terminating "\n" was
3355 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3356 when the result of a list match is failure because a DNS lookup failed.
3358 PH/40 RM_COMMAND is now used in the building process.
3360 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3361 the "build-* directories that it finds.
3363 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3364 address in a domain literal was a prefix of an interface address.
3366 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3367 when verifying a sender address, unless rewrite_headers is false.
3369 PH/44 Wrote a long comment about why errors_to addresses are verified as
3370 recipients, not senders.
3372 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3373 the ratelimit ACL was added.
3375 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3377 PH/46 Added extra information about PostgreSQL errors to the error string.
3379 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3380 avoiding the need to install special zones in a real server. This is
3381 backwards compatible; if it can't find the fake resolver, it drops back.
3382 Thus, both old and new test suites can be run.
3384 TF/02 Added util/ratelimit.pl
3386 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3387 clock is set back in time.
3389 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3390 Candler <B.Candler@pobox.com>.
3392 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3393 OK for addresses that are the result of redirections.
3395 PH/48 A number of further additions for the benefit of the new test suite,
3396 including a fake gethostbyname() that interfaces to the fake DNS resolver
3399 TF/06 The fix for widen_domains has also been applied to qualify_single and
3400 search_parents which are the other dnslookup options that can cause
3403 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3406 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3408 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3409 service" if it encountered an MX record with an empty target hostname.
3410 The message is now "an MX or SRV record indicated no SMTP service".
3412 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3413 if the DNS of the sending site is misconfigured. This is quite a
3414 common situation. This change restores the behaviour of treating a
3415 helo verification defer as a failure.
3417 PH/51 If self=fail was set on a router, the bounce message did not include the
3418 actual error message.
3424 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3426 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3427 command itself ended in error, the underlying message about the transport
3428 filter timeout was being overwritten with the pipe command error. Now the
3429 underlying error message should be appended to the second error message.
3431 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3433 PH/02 Reduce the amount of output that "make" produces by default. Full output
3434 can still be requested.
3436 PH/03 The warning log line about a condition test deferring for a "warn" verb
3437 was being output only once per connection, rather than after each
3438 occurrence (because it was using the same function as for successful
3439 "warn" verbs). This seems wrong, so I have changed it.
3441 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3442 it should not have, which might have caused a crash in the right
3443 circumstances, but probably never did.
3445 PH/04 Installed a modified version of Tony Finch's patch to make submission
3446 mode fix the return path as well as the Sender: header line, and to
3447 add a /name= option so that you can make the user's friendly name appear
3450 TF/03 Added the control = fakedefer ACL modifier.
3452 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3453 Mark Lowes for thorough testing.
3455 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3457 TK/03 Merged latest SRS patch from Miles Wilton.
3459 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3460 with the definition in sysexits.h (which is #included earlier).
3461 Fortunately, Exim does not actually use EX_OK. The code used to try to
3462 preserve the sysexits.h value, by assumimg that macro definitions were
3463 scanned for macro replacements. I have been disabused of this notion,
3464 so now the code just undefines EX_OK before #including unistd.h.
3466 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3467 in the smtp transport. When a block could not be written in a single
3468 write() function, the timeout was being re-applied to each part-write.
3469 This seems wrong - if the receiver was accepting one byte at a time it
3470 would take for ever. The timeout is now adjusted when this happens. It
3471 doesn't have to be particularly precise.
3473 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3474 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3476 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3477 without a domain if the check to <postmaster@domain> fails.
3479 SC/01 Eximstats: added -xls and the ability to specify output files
3480 (patch written by Frank Heydlauf).
3482 SC/02 Eximstats: use FileHandles for outputing results.
3484 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3486 SC/04 Eximstats: fixed display of large numbers with -nvr option
3488 SC/05 Eximstats: fixed merging of reports with empty tables.
3490 SC/06 Eximstats: added the -include_original_destination flag
3492 SC/07 Eximstats: removed tabs and trailing whitespace.
3494 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3496 TK/06 MBOX spool code: Add real "From " MBOX separator line
3497 so the .eml file is really in mbox format (even though
3498 most programs do not really care). Patch from Alex Miller.
3500 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3501 The latter is generated from $received_to and is only set if the
3502 message has one envelope recipient. SA can use these headers,
3503 obviously out-of-the-box. Patch from Alex Miller.
3505 PH/08 The ${def test on a variable was returning false if the variable's
3506 value was "0", contrary to what the specification has always said!
3507 The result should be true unless the variable is empty.
3509 PH/09 The syntax error of a character other than { following "${if
3510 def:variable_name" (after optional whitespace) was not being diagnosed.
3511 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3512 accidental colon was present, for example, could give incorrect results.
3514 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3515 result is used (not including appendfile, where other changes are about
3518 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3519 This involved changing a lot of size variables from int to off_t. It
3520 should work with maildirs and everything.
3522 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3523 spamd dying while we are connected to it.
3525 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3528 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3529 function for BDB 4.3.
3531 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3533 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3534 the sources for PCRE have been re-arranged and more files are now
3537 PH/15 The code I had for printing potentially long long variables in PH/11
3538 above was not the best (it lost precision). The length of off_t variables
3539 is now inspected at build time, and an appropriate printing format (%ld
3540 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3541 to be "long long int" or "long int". This is needed for the internal
3542 formatting function string_vformat().
3544 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3545 the configuration file to be ":syslog", then the script "guesses" where
3546 the logs files are, rather than using the compiled in default. In our
3547 case the guess is not the same as the compiled default, so the script
3548 suddenly stopped working when I started to use syslog. The patch checks
3549 to see if log_file_path is "". If so, it attempts to read it from exim
3550 with no configuration file to get the compiled in version, before it
3551 falls back to the previous guessing code."
3553 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3554 implementing BATV in an Exim configuration. See NewStuff for the gory
3557 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3558 Makefile that are specific to HP-UX.
3560 PH/18 If the "use_postmaster" option was set for a recipient callout together
3561 with the "random" option, the postmaster address was used as the MAIL
3562 FROM address for the random test, but not for the subsequent recipient
3563 test. It is now used for both.
3565 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3566 patch removes a few documentation additions to RFC 3028, because the
3567 latest draft now contains them. It adds the new en;ascii-case comparator
3568 and a new error check for 8bit text in MIME parts. Comparator and
3569 require names are now matched exactly. I enabled the subaddress
3570 extension, but it is not well tested yet (read: it works for me)."
3572 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3573 rework some of the code of TK/09 above to avoid the hardwired use of
3574 "%lld" and "long long". Replaced the call to snprintf() with a call to
3577 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3578 records point to non-existent hosts", "retry timeout exceeded", and
3579 "retry time not reached for any host after a long failure period".
3581 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3582 experimental DomainKeys support:
3584 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3585 (2) On an error such as an illegally used "control", the wrong name for
3586 the control was given.
3588 These problems did NOT occur unless DomainKeys support was compiled.
3590 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3592 PH/24 Added ${if match_ip condition.
3594 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3595 databases so that it will be absolutely obvious if a crash occurs in the
3596 DB library. This is a regular occurrence (often caused by mis-matched
3599 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3600 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3601 picked up on a user's system that detects such things. There doesn't seem
3602 to be a gcc warning option for this - only an attribute that has to be
3603 put on the function's prototype. It seems that in Fedora Core 4 they have
3604 set this on a number of new functions. No doubt there will be more in due
3607 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3608 specify a transport. However, if an address that was verified by such a
3609 router was the subject of a callout, Exim crashed because it tried to
3610 read the rcpt_include_affixes from the non-existent transport. Now it
3611 just assumes that the setting of that option is false. This bug was
3612 introduced by 4.51/PH/31.
3614 PH/28 Changed -d+all to exclude +memory, because that information is very
3615 rarely of interest, but it makes the output a lot bigger. People tend to
3616 do -d+all out of habit.
3618 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3619 code in os-type was giving problems when libc.so lives in lib64, like on
3622 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3623 aren't the modern standard, and it seems that some systems' include files
3624 don't always have them. Exim was already checking for some of the newer
3625 ones like T_AAAA, and defining it itself. I've added checks for all the
3626 record types that Exim uses.
3628 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3629 not automatically generating a new one, as it is supposed to. This
3630 prevented TLS from working. If the file did exist, but contained invalid
3631 data, a new version was generated, as expected. It was only the case of a
3632 non-existent file that was broken.
3634 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3635 with a change in libdomainkeys > 0.64.
3637 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3638 from DNS. If the selector record carries the flag, it now has
3639 precedence over the domain-wide flag.
3641 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3643 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3644 the use of a port name that isn't defined in /etc/services) occurred, the
3645 message was deferred as in a normal delivery, and thus remained on the
3646 spool, instead of being failed because of the mua_wrapper setting. This
3647 is now fixed, and I tidied up some of the mua_wrapper messages at the
3650 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3651 the messages in a hash of arrays rather than using individual hashes.
3652 This is a bit cleaner and results in dramatic memory savings, albeit
3653 at a slight CPU cost.
3655 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3656 as requested by Marc Sherman.
3658 SC/10 Eximstats: added histograms for user specified patterns as requested
3661 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3663 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3664 fopen() in the content-scanning modules that did not already have it.
3670 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3671 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3673 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3675 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3677 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3678 to test Sieve filters that use "vacation".
3680 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3681 that changes the way the GnuTLS parameters are stored in the cache file.
3682 The new format can be generated externally. For backward compatibility,
3683 if the data in the cache doesn't make sense, Exim assumes it has read an
3684 old-format file, and it generates new data and writes a new file. This
3685 means that you can't go back to an older release without removing the
3688 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3689 work if there are any delivery delays because "one_time" forces the
3690 parent to be marked "delivered", so its unseen clone is never tried
3691 again. For this reason, Exim now forbids the simultaneous setting of
3694 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3695 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3696 entry if you want to know the details. The fix, however, neglected to
3697 consider the case when local delivery batching is involved. The test for
3698 "previously delivered" was not happening when checking to see if an
3699 address could be batched with a previous (undelivered) one; under
3700 certain circumstances this could lead to multiple deliveries to the same
3703 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3704 in its include files, and this causes problems building Exim.
3706 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3707 header_syntax) but Exim was just ignoring anything given after a slash.
3708 In particular, this caused confusion with an attempt to use "verify =
3709 reverse_host_lookup/defer_ok". An error is now given when options are
3710 supplied for verify items that do not have them. (Maybe reverse_host_
3711 lookup should have a defer_ok option, but that's a different point.)
3713 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3714 defined by RFC 821) to 2048, because there were problems with some AUTH
3715 commands, and RFC 1869 says the size should be increased for extended
3716 SMTP commands that take arguments.
3718 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3721 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3722 "unknown" error; now it says that the functionality isn't in the binary.
3724 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3725 an address' error message when a string expansion fails (syntax or
3726 whatever). Otherwise the password may appear in the log. Following change
3727 PH/42 below, there is no longer a chance of it appearing in a bounce
3730 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3732 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3733 of its addresses was ever tried. (Bugzilla bug #2.)
3735 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3736 the result incorrectly in the debug output. (It correctly added a newline
3737 to what was transported.)
3739 TF/01 Added $received_time.
3741 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3742 commented out examples of how to interface to a virus scanner and to
3743 SpamAssassin. Also added commented examples of av_scanner and
3744 spamd_address settings.
3746 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3747 and controls are allowed in which ACLs. There were a couple of minor
3748 errors. Some of the entries in the conditions table (which is a table of
3749 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3750 negation of where the condition IS allowed.
3752 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3754 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3755 header file does not have a version number, so I've had to invent a new
3756 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3757 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3758 radiusclient), but it was contributed by a Radius user.
3760 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3761 files or directories, for OpenSSL.
3763 PH/19 When an Exim process that is running as root has to create an Exim log
3764 file, it does so in a subprocess that runs as exim:exim so as to get the
3765 ownership right at creation (otherwise, other Exim processes might see
3766 the file with the wrong ownership). There was no test for failure of this
3767 fork() call, which would lead to the process getting stuck as it waited
3768 for a non-existent subprocess. Forks do occasionally fail when resources
3769 run out. I reviewed all the other calls to fork(); they all seem to check
3772 PH/20 When checking for unexpected SMTP input at connect time (before writing
3773 the banner), Exim was not dealing correctly with a non-positive return
3774 from the read() function. If the client had disconnected by this time,
3775 the result was a log entry for a synchronization error with an empty
3776 string after "input=" when read() returned zero. If read() returned -1
3777 (an event I could not check), uninitialized data bytes were printed.
3778 There were reports of junk text (parts of files, etc) appearing after
3781 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3783 PH/22 Added support for macro redefinition, and (re)definition in between
3784 driver and ACL definitions.
3786 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3787 forgetting to use the resulting value; it was using the unexpanded value.
3789 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3790 hadn't been configured. The fix is from Juergen Kreileder, who
3791 understands it better than I do:
3793 "Here's what I see happening with three configured cyrus_sasl
3794 authenticators configured (plain, login, cram-md5):
3796 On startup auth_cyrus_sasl_init() gets called for each of these.
3797 This means three calls to sasl_listmech() without a specified mech_list.
3798 => SASL tests which mechs of all available mechs actually work
3799 => three warnings about OTP not working
3800 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3802 With the patch, sasl_listmech() also gets called three times. But now
3803 SASL's mech_list option is set to the server_mech specified in the the
3804 authenticator. Or in other words, the answer from sasl_listmech()
3805 gets limited to just the mech you're testing for (which is different
3807 => the return list contains just 'plain' or 'login', 'cram-md5' or
3808 nothing depending on the value of ob->server_mech.
3810 I've just tested the patch: Authentication still works fine,
3811 unavailable mechs specified in the exim configuration are still
3812 caught, and the auth.log warnings about OTP are gone."
3814 PH/25 When debugging is enabled, the contents of the command line are added
3815 to the debugging output, even when log_selector=+arguments is not
3818 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3819 answer is "GNU", and only if the return is "GNU/something" is the answer
3822 PH/27 $acl_verify_message is now set immediately after the failure of a
3823 verification in an ACL, and so is available in subsequent modifiers. In
3824 particular, the message can be preserved by coding like this:
3826 warn !verify = sender
3827 set acl_m0 = $acl_verify_message
3829 Previously, $acl_verify_message was set only while expanding "message"
3830 and "log_message" when a very denied access.
3832 PH/28 Modified OS/os.c-Linux with
3834 -#ifndef OS_LOAD_AVERAGE
3835 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3837 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3838 nomenclature these days.)
3840 PH/29 Installed patch from the Sieve maintainer that adds the options
3841 sieve_useraddress and sieve_subaddress to the redirect router.
3843 PH/30 In these circumstances:
3844 . Two addresses routed to the same list of hosts;
3845 . First host does not offer TLS;
3846 . First host accepts first address;
3847 . First host gives temporary error to second address;
3848 . Second host offers TLS and a TLS session is established;
3849 . Second host accepts second address.
3850 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3851 and peerdn, if requested) that were in fact used only for the second
3854 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3855 attention to any local part prefix or suffix that was matched by the
3856 router that accepted the address. It now behaves in the same way as it
3857 does for delivery: the affixes are removed from the local part unless
3858 rcpt_include_affixes is set on the transport.
3860 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3861 timeout during the DATA phase of an incoming message.
3863 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3864 applied a patch sanctioned by the Sieve maintainer.
3866 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3867 the uid or gid is negative. A case of a negative gid caused this to be
3868 noticed. The fix allows for either to be negative.
3870 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3871 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3874 PH/36 The change PH/12 above was broken. Fixed it.
3876 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3877 the grounds that routing the same address twice would always produce the
3878 same answer. This might have been true once, but it is certainly no
3879 longer true now. Routing a child address may depend on the previous
3880 routing that produced that child. Some complicated redirection strategies
3881 went wrong when messages had multiple recipients, and made Exim's
3882 behaviour dependent on the order in which the addresses were given.
3884 I have moved the duplicate checking until after the routing is complete.
3885 Exim scans the addresses that are assigned to local and remote
3886 transports, and removes any duplicates. This means that more work will be
3887 done, as duplicates will always all be routed, but duplicates are
3888 presumably rare, so I don't expect this is of any significance.
3890 For deliveries to pipes, files, and autoreplies, the duplicate checking
3891 still happens during the routing process, since they are not going to be
3894 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3895 It corrects a timeout issue with spamd. This is Ian's comment: "The
3896 background is that sometimes spamd either never reads data from a
3897 connection it has accepted, or it never writes response data. The exiscan
3898 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3899 blindly assumes that writes won't block so it may never time out."
3901 PH/39 Allow G after quota size as well as K and M.
3903 PH/40 The value set for $authenticated_id in an authenticator may not contain
3904 binary zeroes or newlines because the value is written to log lines and
3905 to spool files. There was no check on this. Now the value is run through
3906 the string_printing() function so that such characters are converted to
3907 printable escape sequences.
3909 PH/41 $message_linecount is a new variable that contains the total number of
3910 lines in the message. Compare $body_linecount, which is the count for the
3913 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3914 bounce and delay warning messages, except in certain special cases, which
3917 (a) An SMTP error message from a remote host;
3918 (b) A message specified in a :fail: redirection;
3919 (c) A message specified in a "fail" command in a system filter;
3920 (d) A message specified in a FAIL return from the queryprogram router;
3921 (e) A message specified by the cannot_route_message router option.
3923 In these cases only, Exim does include the error details in bounce and
3924 warning messages. There are also a few cases where bland messages such
3925 as "unrouteable address" or "local delivery error" are given.
3927 PH/43 $value is now also set for the "else" part of a ${run expansion.
3929 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3930 being worked on, but at least Exim now implements the latest version to
3933 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3934 process to complete was treated as a delivery failure, a timeout while
3935 writing the message to the pipe was logged, but erroneously treated as a
3936 successful delivery. Such timeouts include transport filter timeouts. For
3937 consistency with the overall process timeout, these timeouts are now
3938 treated as errors, giving rise to delivery failures by default. However,
3939 there is now a new Boolean option for the pipe transport called
3940 timeout_defer, which, if set TRUE, converts the failures into defers for
3941 both kinds of timeout. A transport filter timeout is now identified in
3944 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3945 systems where "make" and "gmake" are different, calling "gmake" at top
3946 level broke things. I've arranged for the value of $(MAKE) to be passed
3947 from the Makefile to this script so that it can call the same version of
3951 A note about Exim versions 4.44 and 4.50
3952 ----------------------------------------
3954 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3955 changes of various kinds. As a consequence, a big documentation update was
3956 needed. This delayed the release for rather longer than seemed good, especially
3957 in the light of a couple of (minor) security issues. Therefore, the changes
3958 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3959 release. So 4.44 and 4.50 are in effect two different branches that both start
3962 I have left the 4.50 change log unchanged; it contains all the changes since
3963 4.43. The change log for 4.44 is below; many of its items are identical to
3964 those for 4.50. This seems to be the most sensible way to preserve the
3965 historical information.
3971 1. Minor wording change to the doc/README.SIEVE file.
3973 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3974 computation of the current number of files was incorrect.
3976 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3977 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3980 4. Give more explanation in the error message when the command for a transport
3981 filter fails to execute.
3983 5. There are several places where Exim runs a non-Exim command in a
3984 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3985 was being done only for the command run by the queryprogram router. It is
3986 now done for all such subprocesses. The other cases are: ${run, transport
3987 filters, and the commands run by the lmtp and pipe transports.
3989 6. Added CONFIGURE_GROUP build-time option.
3991 7. Some older OS have a limit of 256 on the maximum number of file
3992 descriptors. Exim was using setrlimit() to set 1000 as a large value
3993 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3994 systems. I've change it so that if it can't get 1000, it tries for 256.
3996 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3997 was an oversight, and furthermore, ever since the addition of extra
3998 controls (e.g. 4.43/32), the checks on when to allow different forms of
3999 "control" were broken. There should now be diagnostics for all cases when a
4000 control that does not make sense is encountered.
4002 9. Added the /retain_sender option to "control=submission".
4004 10. $recipients is now available in the predata ACL (oversight).
4006 11. Tidy the search cache before the fork to do a delivery from a message
4007 received from the command line. Otherwise the child will trigger a lookup
4008 failure and thereby defer the delivery if it tries to use (for example) a
4009 cached ldap connection that the parent has called unbind on.
4011 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4012 of $address_data from the recipient verification was clobbered by the
4013 sender verification.
4015 13. The value of address_data from a sender verification is now available in
4016 $sender_address_data in subsequent conditions in the ACL statement.
4018 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4020 15. Added a new option "connect=<time>" to callout options, to set a different
4023 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4024 was its contents. (It was OK if the option was not defined at all.)
4026 17. A "Completed" log line is now written for messages that are removed from
4027 the spool by the -Mrm option.
4029 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4030 information about exactly what failed.
4032 19. Added -dd to debug only the daemon process.
4034 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4035 handles timeouts, both on the server side and network timeouts. Renamed the
4036 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4038 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4039 It is now set to "smtps".
4041 22. $host_address is now set to the target address during the checking of
4042 ignore_target_hosts.
4044 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4045 being passed; this would have caused $sender_host_name to have been used if
4046 matching the list had actually called for a host name (not very likely,
4047 since this list is usually IP addresses). A host name is now passed as
4050 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4051 code that specifies a non-restarting handler (typically sigaction() in
4052 modern systems) in an attempt to fix a rare and obscure crash bug.
4054 25. Narrowed the window for a race in the daemon that could cause it to ignore
4055 SIGCHLD signals. This is not a major problem, because they are used only to
4056 wake it up if nothing else does.
4058 26. A malformed maildirsize file could cause Exim to calculate negative values
4059 for the mailbox size or file count. Odd effects could occur as a result.
4060 The maildirsize information is now recalculated if the size or filecount
4063 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4064 support for a long time. Removed HAVE_SYS_VFS_H.
4066 28. Installed the latest version of exipick from John Jetmore.
4068 29. In an address list, if the pattern was not a regular expression, an empty
4069 subject address (from a bounce message) matched only if the pattern was an
4070 empty string. Non-empty patterns were not even tested. This was the wrong
4071 because it is perfectly reasonable to use an empty address as part of a
4072 database query. An empty address is now tested by patterns that are
4073 lookups. However, all the other forms of pattern expect the subject to
4074 contain a local part and a domain, and therefore, for them, an empty
4075 address still always fails if the pattern is not itself empty.
4077 30. Exim went into a mad DNS loop when attempting to do a callout where the
4078 host was specified on an smtp transport, and looking it up yielded more
4079 than one IP address.
4081 31. Re-factored the code for checking spool and log partition space into a
4082 function that finds that data and another that does the check. The former
4083 is then used to implement four new variables: $spool_space, $log_space,
4084 $spool_inodes, and $log_inodes.
4086 32. The RFC2047 encoding function was originally intended for short strings
4087 such as real names; it was not keeping to the 75-character limit for
4088 encoded words that the RFC imposes. It now respects the limit, and
4089 generates multiple encoded words if necessary. To be on the safe side, I
4090 have increased the buffer size for the ${rfc2047: expansion operator from
4093 33. It is now permitted to omit both strings after an "if" condition; if the
4094 condition is true, the result is "true". As before, when the second string
4095 is omitted, a false condition yields an empty string. This makes it less
4096 cumbersome to write custom ACL and router conditions.
4098 34. Failure to deliver a bounce message always caused it to be frozen, even if
4099 there was an errors_to setting on the router. The errors_to setting is now
4102 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4103 canonical form (fully expanded) before being placed in
4104 $sender_host_address.
4106 36. The table in the code that translates DNS record types into text (T_A to
4107 "A" for instance) was missing entries for NS and CNAME. It is just possible
4108 that this could have caused confusion if both these types were looked up
4109 for the same domain, because the text type is used as part of Exim's
4110 per-process caching. But the chance of anyone hitting this buglet seems
4113 37. The dnsdb lookup has been extended in a number of ways.
4115 (1) There is a new type, "zns", which walks up the domain tree until it
4116 finds some nameserver records. It should be used with care.
4118 (2) There is a new type, "mxh", which is like "mx" except that it returns
4119 just the host names, not the priorities.
4121 (3) It is now possible to give a list of domains (or IP addresses) to be
4122 looked up. The behaviour when one of the lookups defers can be
4123 controlled by a keyword.
4125 (4) It is now possible to specify the separator character for use when
4126 multiple records are returned.
4128 38. The dnslists ACL condition has been extended: it is now possible to supply
4129 a list of IP addresses and/or domains to be looked up in a particular DNS
4132 39. Added log_selector=+queue_time_overall.
4134 40. When running the queue in the test harness, wait just a tad after forking a
4135 delivery process, to get repeatability of debugging output.
4137 41. Include certificate and key file names in error message when GnuTLS fails
4138 to set them up, because the GnuTLS error message doesn't include the name
4139 of the failing file when there is a problem reading it.
4141 42. Allow both -bf and -bF in the same test run.
4143 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4145 44. The "Exiscan patch" is now merged into the mainline Exim source.
4147 45. Sometimes the final signoff response after QUIT could fail to get
4148 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4149 < 0 before doing a fflush(). This bug looks as though it goes back to the
4150 introduction of TLS in release 3.20, but "sometimes" must have been rare
4151 because the tests only now provoked it.
4153 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4154 (this can affect the format of dates).
4156 47. exim_tidydb, when checking for the continued existence of a message for
4157 which it has found a message-specific retry record, was not finding
4158 messages that were in split spool directories. Consequently, it was
4159 deleting retry records that should have stayed in existence.
4161 48. Steve fixed some bugs in eximstats.
4163 49. The SPA authentication driver was not abandoning authentication and moving
4164 on to the next authenticator when an expansion was forced to fail,
4165 contradicting the general specification for all authenticators. Instead it
4166 was generating a temporary error. It now behaves as specified.
4168 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4169 (the order specifies the preference for clients). The order is now AES256,
4170 AES128, 3DES, ARCFOUR128.
4172 51. Small patch to Sieve code - explicitly set From: when generating an
4175 52. Exim crashed if a remote delivery caused a very long error message to be
4176 recorded - for instance if somebody sent an entire SpamAssassin report back
4177 as a large number of 550 error lines. This bug was coincidentally fixed by
4178 increasing the size of one of Exim's internal buffers (big_buffer) that
4179 happened as part of the Exiscan merge. However, to be on the safe side, I
4180 have made the code more robust (and fixed the comments that describe what
4183 53. Now that there can be additional text after "Completed" in log lines (if
4184 the queue_time_overall log selector is set), a one-byte patch to exigrep
4185 was needed to allow it to recognize "Completed" as not the last thing in
4188 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4189 patch that reportedly fixes this has been added. I am not expert enough to
4190 create a test for it. This is what the patch creator wrote:
4192 "I found a little strange behaviour of ldap code when working with
4193 Windows 2003 AD Domain, where users was placed in more than one
4194 Organization Units. When I tried to give exim partial DN, the exit code
4195 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4196 But simultaneously result of request was absolutely normal ldap result,
4197 so I produce this patch..."
4199 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4200 REFERENCE, so I have modified the code to exclude the patch when that macro
4203 55. Some experimental protocols are using DNS PTR records for new purposes. The
4204 keys for these records are domain names, not reversed IP addresses. The
4205 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4206 leaves it alone. Component reversal etc. now happens only for IP addresses.
4209 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4211 57. Double the size of the debug message buffer (to 2048) so that more of very
4212 long debug lines gets shown.
4214 58. The exicyclog utility now does better if the number of log files to keep
4215 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4217 59. Two changes related to the smtp_active_hostname option:
4219 (1) $smtp_active_hostname is now available as a variable.
4220 (2) The default for smtp_banner uses $smtp_active_hostname instead
4221 of $primary_hostname.
4223 60. The host_aton() function is supposed to be passed a string that is known
4224 to be a valid IP address. However, in the case of IPv6 addresses, it was
4225 not checking this. This is a hostage to fortune. Exim now panics and dies
4226 if the condition is not met. A case was found where this could be provoked
4227 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4228 components; fortuitously, this particular loophole had already been fixed
4229 by change 4.50/55 above.
4231 If there are any other similar loopholes, the new check in host_aton()
4232 itself should stop them being exploited. The report I received stated that
4233 data on the command line could provoke the exploit when Exim was running as
4234 exim, but did not say which command line option was involved. All I could
4235 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4236 running as the user.
4239 61. There was a buffer overflow vulnerability in the SPA authentication code
4240 (which came originally from the Samba project). I have added a test to the
4241 spa_base64_to_bits() function which I hope fixes it.
4244 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4245 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4247 63. The daemon start-up calls getloadavg() while still root for those OS that
4248 need the first call to be done as root, but it missed one case: when
4249 deliver_queue_load_max is set with deliver_drop_privilege. This is
4250 necessary for the benefit of the queue runner, because there is no re-exec
4251 when deliver_drop_privilege is set.
4253 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4254 This has been fixed.
4256 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4257 was in use, was not putting the data itself into the right store pool;
4258 consequently, it could be overwritten for a subsequent message in the same
4259 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4262 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4264 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4265 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4266 the function were treating the return as a boolean value, which happened to
4267 work because 0=false and not-0=true, but is not correct code.
4269 68. The host_aton() function was not handling scoped IPv6 addresses (those
4270 with, for example, "%eth0" on the end) correctly.
4272 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4273 negated items (that is, ~something) in unsigned ints. Some compilers
4274 apparently mutter when there is no cast.
4276 70. If an address verification called from an ACL failed, and did not produce a
4277 user-specific message (i.e. there was only a "system" message), nothing was
4278 put in $acl_verify_message. In this situation, it now puts the system
4281 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4282 session; change 4.31/43 added the unwanted input to the log line - except
4283 that it did not do this in the start of session case. It now does.
4285 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4286 This is wrong and can cause the other end to generate a synchronization
4287 error if it is another Exim or anything else that does the synchronization
4288 check. A QUIT command is no longer sent after a timeout.
4290 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4291 during host lookups.
4293 74. The defer_ok option of callout verification was not working if it was used
4294 when verifying addresses in header lines, that is, for this case:
4296 verify = header_sender/callout=defer_ok
4298 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4299 those file descriptors could be used for SMTP connections. If anything
4300 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4301 could be sent to the SMTP client, causing chaos. The daemon now opens
4302 stdin, stdout, and stderr to /dev/null when it puts itself into the
4305 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4306 log by default. The user can override this with suitable Perl magic.
4308 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4309 the log message when discard triggers, was not working for the DATA ACL or
4310 for the non-SMTP ACL.
4312 78. Error message wording change in sieve.c.
4314 79. If smtp_accept_max_per_host was set, the number of connections could be
4315 restricted to fewer than expected, because the daemon was trying to set up
4316 a new connection before checking whether the processes handling previous
4317 connections had finished. The check for completed processes is now done
4318 earlier. On busy systems, this bug wouldn't be noticed because something
4319 else would have woken the daemon, and it would have reaped the completed
4322 80. If a message was submitted locally by a user whose login name contained one
4323 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4324 It caused a spool format error. I have fixed the spool reading code. A
4325 related problem was that the "from" clause in the Received: line became
4326 illegal because of the space(s). It is now covered by ${quote_local_part.
4328 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4331 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4333 83. Patch from Sieve maintainer to fix unterminated string problem in
4334 "vacation" handling.
4336 84. Some minor changes to the Linux configuration files to help with other
4337 OS variants using glibc.
4339 85. One more patch for Sieve to update vacation handling to latest spec.
4342 ----------------------------------------------------
4343 See the note above about the 4.44 and 4.50 releases.
4344 ----------------------------------------------------
4350 1. Change 4.43/35 introduced a bug that caused file counts to be
4351 incorrectly computed when quota_filecount was set in an appendfile
4354 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4355 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4358 3. Give more explanation in the error message when the command for a transport
4359 filter fails to execute.
4361 4. There are several places where Exim runs a non-Exim command in a
4362 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4363 was being done only for the command run by the queryprogram router. It is
4364 now done for all such subprocesses. The other cases are: ${run, transport
4365 filters, and the commands run by the lmtp and pipe transports.
4367 5. Some older OS have a limit of 256 on the maximum number of file
4368 descriptors. Exim was using setrlimit() to set 1000 as a large value
4369 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4370 systems. I've change it so that if it can't get 1000, it tries for 256.
4372 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4373 was an oversight, and furthermore, ever since the addition of extra
4374 controls (e.g. 4.43/32), the checks on when to allow different forms of
4375 "control" were broken. There should now be diagnostics for all cases when a
4376 control that does not make sense is encountered.
4378 7. $recipients is now available in the predata ACL (oversight).
4380 8. Tidy the search cache before the fork to do a delivery from a message
4381 received from the command line. Otherwise the child will trigger a lookup
4382 failure and thereby defer the delivery if it tries to use (for example) a
4383 cached ldap connection that the parent has called unbind on.
4385 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4386 of $address_data from the recipient verification was clobbered by the
4387 sender verification.
4389 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4390 was its contents. (It was OK if the option was not defined at all.)
4392 11. A "Completed" log line is now written for messages that are removed from
4393 the spool by the -Mrm option.
4395 12. $host_address is now set to the target address during the checking of
4396 ignore_target_hosts.
4398 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4399 being passed; this would have caused $sender_host_name to have been used if
4400 matching the list had actually called for a host name (not very likely,
4401 since this list is usually IP addresses). A host name is now passed as
4404 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4405 code that specifies a non-restarting handler (typically sigaction() in
4406 modern systems) in an attempt to fix a rare and obscure crash bug.
4408 15. Narrowed the window for a race in the daemon that could cause it to ignore
4409 SIGCHLD signals. This is not a major problem, because they are used only to
4410 wake it up if nothing else does.
4412 16. A malformed maildirsize file could cause Exim to calculate negative values
4413 for the mailbox size or file count. Odd effects could occur as a result.
4414 The maildirsize information is now recalculated if the size or filecount
4417 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4418 support for a long time. Removed HAVE_SYS_VFS_H.
4420 18. Updated exipick to current release from John Jetmore.
4422 19. Allow an empty sender to be matched against a lookup in an address list.
4423 Previously the only cases considered were a regular expression, or an
4426 20. Exim went into a mad DNS lookup loop when doing a callout where the
4427 host was specified on the transport, if the DNS lookup yielded more than
4430 21. The RFC2047 encoding function was originally intended for short strings
4431 such as real names; it was not keeping to the 75-character limit for
4432 encoded words that the RFC imposes. It now respects the limit, and
4433 generates multiple encoded words if necessary. To be on the safe side, I
4434 have increased the buffer size for the ${rfc2047: expansion operator from
4437 22. Failure to deliver a bounce message always caused it to be frozen, even if
4438 there was an errors_to setting on the router. The errors_to setting is now
4441 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4442 canonical form (fully expanded) before being placed in
4443 $sender_host_address.
4445 24. Updated eximstats to version 1.33
4447 25. Include certificate and key file names in error message when GnuTLS fails
4448 to set them up, because the GnuTLS error message doesn't include the name
4449 of the failing file when there is a problem reading it.
4451 26. Expand error message when OpenSSL has problems setting up cert/key files.
4454 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4455 (this can affect the format of dates).
4457 28. exim_tidydb, when checking for the continued existence of a message for
4458 which it has found a message-specific retry record, was not finding
4459 messages that were in split spool directories. Consequently, it was
4460 deleting retry records that should have stayed in existence.
4462 29. eximstats updated to version 1.35
4463 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4464 1.35 - bugfix such that pie charts by volume are generated correctly
4466 30. The SPA authentication driver was not abandoning authentication and moving
4467 on to the next authenticator when an expansion was forced to fail,
4468 contradicting the general specification for all authenticators. Instead it
4469 was generating a temporary error. It now behaves as specified.
4471 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4472 (the order specifies the preference for clients). The order is now AES256,
4473 AES128, 3DES, ARCFOUR128.
4475 31. Small patch to Sieve code - explicitly set From: when generating an
4478 32. Exim crashed if a remote delivery caused a very long error message to be
4479 recorded - for instance if somebody sent an entire SpamAssassin report back
4480 as a large number of 550 error lines. This bug was coincidentally fixed by
4481 increasing the size of one of Exim's internal buffers (big_buffer) that
4482 happened as part of the Exiscan merge. However, to be on the safe side, I
4483 have made the code more robust (and fixed the comments that describe what
4486 33. Some experimental protocols are using DNS PTR records for new purposes. The
4487 keys for these records are domain names, not reversed IP addresses. The
4488 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4489 leaves it alone. Component reversal etc. now happens only for IP addresses.
4492 34. The host_aton() function is supposed to be passed a string that is known
4493 to be a valid IP address. However, in the case of IPv6 addresses, it was
4494 not checking this. This is a hostage to fortune. Exim now panics and dies
4495 if the condition is not met. A case was found where this could be provoked
4496 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4497 components; fortuitously, this particular loophole had already been fixed
4498 by change 4.50/55 or 4.44/33 above.
4500 If there are any other similar loopholes, the new check in host_aton()
4501 itself should stop them being exploited. The report I received stated that
4502 data on the command line could provoke the exploit when Exim was running as
4503 exim, but did not say which command line option was involved. All I could
4504 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4505 running as the user.
4508 35. There was a buffer overflow vulnerability in the SPA authentication code
4509 (which came originally from the Samba project). I have added a test to the
4510 spa_base64_to_bits() function which I hope fixes it.
4513 36. The daemon start-up calls getloadavg() while still root for those OS that
4514 need the first call to be done as root, but it missed one case: when
4515 deliver_queue_load_max is set with deliver_drop_privilege. This is
4516 necessary for the benefit of the queue runner, because there is no re-exec
4517 when deliver_drop_privilege is set.
4519 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4520 was in use, was not putting the data itself into the right store pool;
4521 consequently, it could be overwritten for a subsequent message in the same
4522 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4525 38. Sometimes the final signoff response after QUIT could fail to get
4526 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4527 < 0 before doing a fflush(). This bug looks as though it goes back to the
4528 introduction of TLS in release 3.20, but "sometimes" must have been rare
4529 because the tests only now provoked it.
4535 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4536 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4537 Now it returns an integer. A number of calls were still expecting a T/F
4538 return. Fortuitously, in all cases, the tests worked in OK situations,
4539 which is the norm. However, things would have gone wrong on any write
4540 failures on the smtp file descriptor. This function is used when sending
4541 messages over SMTP and also when doing verify callouts.
4543 2. When Exim is called to do synchronous delivery of a locally submitted
4544 message (the -odf or -odi options), it no longer closes stderr before doing
4547 3. Implemented the mua_wrapper option.
4549 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4551 5. Implemented the functions header_remove(), header_testname(),
4552 header_add_at_position(), and receive_remove_recipient(), and exported them
4555 6. If an ACL "warn" statement specified the addition of headers, Exim already
4556 inserted X-ACL-Warn: at the start if there was no header name. However, it
4557 was not making this test for the second and subsequent header lines if
4558 there were newlines in the string. This meant that an invalid header could
4559 be inserted if Exim was badly configured.
4561 7. Allow an ACL "warn" statement to add header lines at the start or after all
4562 the Received: headers, as well as at the end.
4564 8. Added the rcpt_4xx retry error code.
4566 9. Added postmaster_mailfrom=xxx to callout verification option.
4568 10. Added mailfrom=xxxx to the callout verification option, for verify=
4571 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4572 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4574 12. Inserted some casts to stop certain compilers warning when using pointer
4575 differences as field lengths or precisions in printf-type calls (mostly
4576 affecting debugging statements).
4578 13. Added optional readline() support for -be (dynamically loaded).
4580 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4581 same clock tick as a message's arrival, so that its received time was the
4582 same as the "first fail" time on the retry record, and that message
4583 remained on the queue past the ultimate address timeout, every queue runner
4584 would try a delivery (because it was past the ultimate address timeout) but
4585 after another failure, the ultimate address timeout, which should have then
4586 bounced the address, did not kick in. This was a "< instead of <=" error;
4587 in most cases the first failure would have been in the next clock tick
4588 after the received time, and all would be well.
4590 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4591 being recognized when the domain list was tested by the match_domain
4592 condition in an expansion string.
4594 16. Added the ${str2b64: operator.
4596 17. Exim was always calling setrlimit() to set a large limit for the number of
4597 processes, without checking whether the existing limit was already
4598 adequate. (It did check for the limit on file descriptors.) Furthermore,
4599 errors from getrlimit() and setrlimit() were being ignored. Now they are
4600 logged to the main and panic logs, but Exim does carry on, to try to do its
4601 job under whatever limits there are.
4603 18. Imported PCRE 5.0.
4605 19. Trivial typo in log message " temporarily refused connection" (the leading
4608 20. If the log selector return_path_on_delivery was set and an address was
4609 redirected to /dev/null, the delivery process crashed because it assumed
4610 that a return path would always be set for a "successful" delivery. In this
4611 case, the whole delivery is bypassed as an optimization, and therefore no
4614 21. Internal re-arrangement: the function for sending a challenge and reading
4615 a response while authentication was assuming a zero-terminated challenge
4616 string. It's now changed to take a pointer and a length, to allow for
4617 binary data in such strings.
4619 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4621 23. Exim was not respecting finduser_retries when seeking the login of the
4622 uid under which it was called; it was always trying 10 times. (The default
4623 setting of finduser_retries is zero.) Also, it was sleeping after the final
4624 failure, which is pointless.
4626 24. Implemented tls_on_connect_ports.
4628 25. Implemented acl_smtp_predata.
4630 26. If the domain in control=submission is set empty, Exim assumes that the
4631 authenticated id is a complete email address when it generates From: or
4632 Sender: header lines.
4634 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4635 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4636 chown and chgrp in /bin and hostname in /usr/bin.
4638 28. Exim was keeping the "process log" file open after each use, just as it
4639 does for the main log. This opens the possibility of it remaining open for
4640 long periods when the USR1 signal hits a daemon. Occasional processlog
4641 errors were reported, that could have been caused by this. Anyway, it seems
4642 much more sensible not to leave this file open at all, so that is what now
4645 29. The long-running daemon process does not normally write to the log once it
4646 has entered its main loop, and it closes the log before doing so. This is
4647 so that log files can straightforwardly be renamed and moved. However,
4648 there are a couple of unusual error situations where the daemon does write
4649 log entries, and I had neglected to close the log afterwards.
4651 30. The text of an SMTP error response that was received during a remote
4652 delivery was being truncated at 512 bytes. This is too short for some of
4653 the long messages that one sometimes sees. I've increased the limit to
4656 31. It is now possible to make retry rules that apply only when a message has a
4657 specific sender, in particular, an empty sender.
4659 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4660 it possible to be selective about when SMTP synchronization is enforced.
4662 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4664 32. Implemented hosts_connection_nolog.
4666 33. Added an ACL for QUIT.
4668 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4671 35. Added mailbox_size and mailbox_filecount to appendfile.
4673 36. Added control = no_multiline_responses to ACLs.
4675 37. There was a bug in the logic of the code that waits for the clock to tick
4676 in the case where the clock went backwards by a substantial amount such
4677 that the microsecond fraction of "now" was more than the microsecond
4678 fraction of "then" (but the whole seconds number was less).
4680 38. Added support for the libradius Radius client library this is found on
4681 FreeBSD (previously only the radiusclient library was supported).
4687 1. When certain lookups returned multiple values in the form name=value, the
4688 quoting of the values was not always being done properly. Specifically:
4689 (a) If the value started with a double quote, but contained no whitespace,
4691 (b) If the value contained whitespace other than a space character (i.e.
4692 tabs or newlines or carriage returns) it was not quoted.
4693 This fix has been applied to the mysql and pgsql lookups by writing a
4694 separate quoting function and calling it from the lookup code. The fix
4695 should probably also be applied to nisplus, ibase and oracle lookups, but
4696 since I cannot test any of those, I have not disturbed their existing code.
4698 2. A hit in the callout cache for a specific address caused a log line with no
4699 reason for rejecting RCPT. Now it says "Previous (cached) callout
4700 verification failure".
4702 3. There was an off-by-one bug in the queryprogram router. An over-long
4703 return line was truncated at 256 instead of 255 characters, thereby
4704 overflowing its buffer with the terminating zero. As well as fixing this, I
4705 have increased the buffer size to 1024 (and made a note to document this).
4707 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4708 when Exim is waiting for an SMTP response from a remote server, Exim
4709 restarts its select() call on the socket, thereby resetting its timeout.
4710 This is not a problem when such interrupts are rare. Somebody set up a cron
4711 job to run exiwhat every 2 minutes, which is less than the normal select()
4712 timeout (5 or 10 minutes). This meant that the select() timeout never
4713 kicked in because it was always reset. I have fixed this by comparing the
4714 time when an interrupt arrives with the time at the start of the first call
4715 to select(). If more time than the timeout has elapsed, the interrupt is
4716 treated as a timeout.
4718 5. Some internal re-factoring in preparation for the addition of Sieve
4719 extensions (by MH). In particular, the "personal" test is moved to a
4720 separate function, and given an option for scanning Cc: and Bcc: (which is
4721 not set for Exim filters).
4723 6. When Exim created an email address using the login of the caller as the
4724 local part (e.g. when creating a From: or Sender: header line), it was not
4725 quoting the local part when it contained special characters such as @.
4727 7. Installed new OpenBSD configuration files.
4729 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4730 try to make them clearer.
4732 9. Callout options, other than the timeout value, were being ignored when
4733 verifying sender addresses in header lines. For example, when using
4735 verify = header_sender/callout=no_cache
4737 the cache was (incorrectly) being used.
4739 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4740 only the Cygwin environment.
4742 11. When return_path_on_delivery was set as a log selector, if different remote
4743 addresses in the same message used different return paths and parallel
4744 remote delivery occurred, the wrong values would sometimes be logged.
4745 (Whenever a remote delivery process finished, the return path value from
4746 the most recently started remote delivery process was logged.)
4748 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4749 lines when AUTH and/or TLS are in use. This is the "received protocol"
4750 field. Exim used to use "asmtp" for authenticated SMTP, without any
4751 indication (in the protocol name) for TLS use. Now it follows the RFC and
4752 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4753 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4754 names appear in log lines as well as in Received: header lines.
4756 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4757 extensions, and comparison tests, and to fix some bugs.
4759 14. Changes to the "personal" filter test:
4761 (1) The test was buggy in that it was just doing the equivalent of
4762 "contains" tests on header lines. For example, if a user's address was
4763 anne@some.where, the "personal" test would incorrectly be true for
4765 To: susanne@some.where
4767 This test is now done by extracting each address from the header in turn,
4768 and checking the entire address. Other tests that are part of "personal"
4769 are now done using regular expressions (for example, to check local parts
4770 of addresses in From: header lines).
4772 (2) The list of non-personal local parts in From: addresses has been
4773 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4774 taken from the Sieve specification recommendations.
4776 (3) If the message contains any header line starting with "List-" it is
4777 treated as non-personal.
4779 (4) The test for "circular" in the Subject: header line has been removed
4780 because it now seems ill-conceived.
4782 15. Minor typos in src/EDITME comments corrected.
4784 16. Installed latest exipick from John Jetmore.
4786 17. If headers_add on a router specified a text string that was too long for
4787 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4788 of string_sprintf() is now avoided.
4790 18. $message_body_size was not set (it was always zero) when running the DATA
4791 ACL and the local_scan() function.
4793 19. For the "mail" command in an Exim filter, no default was being set for
4794 the once_repeat time, causing a random time value to be used if "once" was
4795 specified. (If the value happened to be <= 0, no repeat happened.) The
4796 default is now 0s, meaning "never repeat". The "vacation" command was OK
4797 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4798 (I found it when inspecting the code).
4800 20. There is now an overall timeout for performing a callout verification. It
4801 defaults to 4 times the callout timeout, which applies to individual SMTP
4802 commands during the callout. The overall timeout applies when there is more
4803 than one host that can be tried. The timeout is checked before trying the
4804 next host. This prevents very long delays if there are a large number of
4805 hosts and all are timing out (e.g. when the network connections are timing
4806 out). The value of the overall timeout can be changed by specifying an
4807 additional sub-option for "callout", called "maxwait". For example:
4809 verify = sender/callout=5s,maxwait=20s
4811 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4812 to the end before writing, but this should make it even safer).
4814 22. Exim was forgetting that it had advertised PIPELINING for the second and
4815 subsequent messages on an SMTP connection. It was also not resetting its
4816 memory on STARTTLS and an internal HELO.
4818 23. When Exim logs an SMTP synchronization error within a session, it now
4819 records whether PIPELINING has been advertised or not.
4821 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4822 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4823 rather than long int.
4825 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4827 26. Added the never_mail option to autoreply.
4833 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4834 crash if the getsockname() call failed; this can happen if a connection is
4835 closed very soon after it is established. The problem was simply in the
4836 order in which certain operations were done, causing Exim to try to write
4837 to the SMTP stream before it had set up the file descriptor. The bug has
4838 been fixed by making things happen in the correct order.
4844 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4845 before the connection was closed, thus losing the rejection response.
4847 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4848 some early Solaris releases, but causes trouble in current releases where
4849 socklen_t is defined.
4851 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4854 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4857 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4858 that releasing the top of it at the end releases what was used for sub-
4859 expansions (unless the block got too big). However, discard this block if
4860 the first thing is a variable or header, so that we can use its block when
4861 it is dynamic (useful for very large $message_headers, for example).
4863 6. Lookups now cache *every* query, not just the most recent. A new, separate
4864 store pool is used for this. It can be recovered when all lookup caches are
4865 flushed. Lookups now release memory at the end of their result strings.
4866 This has involved some general refactoring of the lookup sources.
4868 7. Some code has been added to the store_xxx() functions to reduce the amount
4869 of flapping under certain conditions.
4871 8. log_incoming_interface used to affect only the <= reception log lines. Now
4872 it causes the local interface and port to be added to several more SMTP log
4873 lines, for example "SMTP connection from", and rejection lines.
4875 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4877 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4879 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4880 could be overwritten at the end of the current message (or the start of a
4881 new message if it was set in a HELO ACL). The value is now preserved for
4882 the duration of the SMTP connection.
4884 12. If a transport had a headers_rewrite setting, and a matching header line
4885 contained an unqualified address, that address was qualified, even if it
4886 did not match any rewriting rules. The underlying bug was that the values
4887 of the flags that permit the existence of unqualified sender and recipient
4888 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4889 non-local messages, and by -bnq for local messages) were not being
4890 preserved with the message after it was received.
4892 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4893 "next input=" as part of the text comprising the host identity instead of
4894 the correct text. The code was using the same buffer for two different
4895 strings. However, depending on which order the printing function evaluated
4896 its arguments, the bug did not always show up. Under Linux, for example, my
4897 test suite worked just fine.
4899 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4900 doesn't work with some older versions of Perl. It has been changed to "my",
4901 which in any case is probably the better facility to use.
4903 15. A really picky compiler found some instances of statements for creating
4904 error messages that either had too many or two few arguments for the format
4907 16. The size of the buffer for calls to the DNS resolver has been increased
4908 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4909 for addresses that have a lot of PTR records. This alleviates a problem; it
4910 does not fully solve it.
4912 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4913 buffer now truncates the list and logs the incident, which is the same
4914 action as happens when Exim is looking up a host name and its aliases.
4915 Previously in this situation something unpredictable would happen;
4916 sometimes it was "internal error: store_reset failed".
4918 18. If a server dropped the connection unexpectedly when an Exim client was
4919 using GnuTLS and trying to read a response, the client delivery process
4920 crashed while trying to generate an error log message.
4922 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4923 string, for example:
4925 warn message = H1: something\nH2: something
4927 the text was added as a single header line from Exim's point of view
4928 though it ended up OK in the delivered message. However, searching for the
4929 second and subsequent header lines using $h_h2: did not work. This has been
4930 fixed. Similarly, if a system filter added multiple headers in this way,
4931 the routers could not see them.
4933 20. Expanded the error message when iplsearch is called with an invalid key to
4934 suggest using net-iplsearch in a host list.
4936 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4937 ACLs are no longer actually imposed (and a message to that effect is
4940 22. If a "gecos" field in a passwd entry contained escaped characters, in
4941 particular, if it contained a \" sequence, Exim got it wrong when building
4942 a From: or a Sender: header from that name. A second bug also caused
4943 incorrect handling when an unquoted " was present following a character
4944 that needed quoting.
4946 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4947 was not being matched caselessly.
4949 24. Arranged for all hyphens in the exim.8 source to be escaped with
4952 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4953 properly. Recipient callout cache records were still being keyed to include
4954 the sender, even when use_sender was set false. This led to far more
4955 callouts that were necessary. The sender is no longer included in the key
4956 when use_sender is false.
4958 26. Added "control = submission" modifier to ACLs.
4960 27. Added the ${base62d: operator to decode base 62 numbers.
4962 28. dnsdb lookups can now access SRV records.
4964 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4965 the configuration file.
4967 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4968 (-v) mode. This makes the output for a verbose queue run more intelligible.
4970 31. Added a use_postmaster feature to recipient callouts.
4972 32. Added the $body_zerocount variable, containing the number of binary zero
4973 bytes in the message body.
4975 33. The time of last modification of the "new" subdirectory is now used as the
4976 "mailbox time last read" when there is a quota error for a maildir
4979 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4981 35. Added +ignore_unknown as a special item in host lists.
4983 36. Code for decoding IPv6 addresses in host lists is now included, even if
4984 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4985 address was recognized as an IP address, but was then not correctly decoded
4986 into binary, causing unexpected and incorrect effects when compared with
4993 1. Very minor rewording of debugging text in manualroute to say "list of
4994 hosts" instead of "hostlist".
4996 2. If verify=header_syntax was set, and a header line with an unqualified
4997 address (no domain) and a large number of spaces between the end of the
4998 name and the colon was received, the reception process suffered a buffer
4999 overflow, and (when I tested it) crashed. This was caused by some obsolete
5000 code that should have been removed. The fix is to remove it!
5002 3. When running in the test harness, delay a bit after writing a bounce
5003 message to get a bit more predictability in the log output.
5005 4. Added a call to search_tidyup() just before forking a reception process. In
5006 theory, someone could use a lookup in the expansion of smtp_accept_max_
5007 per_host which, without the tidyup, could leave open a database connection.
5009 5. Added the variables $recipient_data and $sender_data which get set from a
5010 lookup success in an ACL "recipients" or "senders" condition, or a router
5011 "senders" option, similar to $domain_data and $local_part_data.
5013 6. Moved the writing of debug_print from before to after the "senders" test
5016 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5017 problems for message scanning, either using a data ACL, or using
5018 local_scan() because the Received: header was not generated till after they
5019 were called (in order to set the time as the time of reception completion).
5020 I have revised the way this works. The header is now generated after the
5021 body is received, but before the ACL or local_scan() are called. After they
5022 are run, the timestamp in the header is updated.
5028 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5029 before starting a queue runner without re-exec. This happened only when
5030 deliver_drop_privilege was set or when the Exim user was set to root. The
5031 effect of the bug was that timeouts during subsequent deliveries caused
5032 crashes instead of being properly handled. The handler is now left at its
5033 default (and expected) setting.
5035 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5036 message, again when deliver_drop_privilege is set or Exim is run as root.
5037 The bug described in (1) was not present in this case, but the tidying up
5038 of the other signals was missing. I have made the two cases consistent.
5040 3. The ignore_target_hosts setting on a manualroute router was being ignored
5041 for hosts that were looked up using the /MX notation.
5043 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5046 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5047 operated on the sender address. After changing the $sender_address to <>
5048 for the sender address verify, Exim was re-instated it as the original
5049 (before rewriting) address, but remembering that it had rewritten it, so it
5050 wasn't rewriting it again. This bug also had the effect of breaking the
5051 sender address verification caching when the sender address was rewritten.
5053 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5054 This has been changed so that if the ip literal address matches
5055 ignore_target_hosts, the router declines.
5057 7. Added expansion conditions match_domain, match_address, and match_local_
5058 part (NOT match_host).
5060 8. The placeholder for the Received: header didn't have a length field set.
5062 9. Added code to Exim itself and to exim_lock to test for a specific race
5063 condition that could lead to file corruption when using MBX delivery. The
5064 issue is with the lockfile that is created in /tmp. If this file is removed
5065 after a process has opened it but before that process has acquired a lock,
5066 there is the potential for a second process to recreate the file and also
5067 acquire a lock. This could lead to two Exim processes writing to the file
5068 at the same time. The added code performs the same test as UW imapd; it
5069 checks after acquiring the lock that its file descriptor still refers to
5070 the same named file.
5072 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5073 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5076 11. Added the smtp_active_hostname option. If used, this will typically be made
5077 to depend on the incoming interface address. Because $interface_address is
5078 not set up until the daemon has forked a reception process, error responses
5079 that can happen earlier (such as "too many connections") no longer contain
5082 12. If an expansion in a condition on a "warn" statement fails because a lookup
5083 defers, the "warn" statement is abandoned, and the next ACL statement is
5084 processed. Previously this caused the whole ACL to be aborted.
5086 13. Added the iplsearch lookup type.
5088 14. Added ident_timeout as a log selector.
5090 15. Added tls_certificate_verified as a log selector.
5092 16. Added a global option tls_require_ciphers (compare the smtp transport
5093 option of the same name). This controls incoming TLS connections.
5095 17. I finally figured out how to make tls_require_ciphers do a similar thing
5096 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5097 before starting the TLS session.
5099 18. Tabs are now shown as \t in -bP output.
5101 19. If the log selector return_path_on_delivery was set, Exim crashed when
5102 bouncing a message because it had too many Received: header lines.
5104 20. If two routers both had headers_remove settings, and the first one included
5105 a superfluous trailing colon, the final name in the first list and the
5106 first name in the second list were incorrectly joined into one item (with a
5107 colon in the middle).
5113 1. Added -C and -D options to the exinext utility, mainly to make it easier
5114 to include in the automated testing, but these could be helpful when
5115 multiple configurations are in use.
5117 2. The exinext utility was not formatting the output nicely when there was
5118 an alternate port involved in the retry record key, nor when there was a
5119 message id as well (for retries that were specific to a specific message
5120 and a specific host). It was also confused by IPv6 addresses, because of
5121 the additional colons they contain. I have fixed the IPv4 problem, and
5122 patched it up to do a reasonable job for IPv6.
5124 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5125 delivery, the log line now contains "pipelined" if PIPELINING was used.
5127 4. An SMTP transport process used to panic and die if the bind() call to set
5128 an explicit outgoing interface failed. This has been changed; it is now
5129 treated in the same way as a connect() failure.
5131 5. A reference to $sender_host_name in the part of a conditional expansion
5132 that was being skipped was still causing a DNS lookup. This no longer
5135 6. The def: expansion condition was not recognizing references to header lines
5136 that used bh_ and bheader_.
5138 7. Added the _cache feature to named lists.
5140 8. The code for checking quota_filecount in the appendfile transport was
5141 allowing one more file than it should have been.
5143 9. For compatibility with Sendmail, the command line option
5151 and sets the incoming protocol and host name (for trusted callers). The
5152 host name and its colon can be omitted when only the protocol is to be set.
5153 Note the Exim already has two private options, -pd and -ps, that refer to
5154 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5155 "s", but I don't think that's a major issue.
5157 10. A number of refactoring changes to the code, none of which should affect
5160 (a) The number of logging options was getting close to filling up the
5161 32-bit word that was used as a bit map. I have split them into two classes:
5162 those that are passed in the argument to log_write(), and those that are
5163 only ever tested independently outside of that function. These are now in
5164 separate 32-bit words, so there is plenty of room for expansion again.
5165 There is no change in the user interface or the logging behaviour.
5167 (b) When building, for example, log lines, the code previously used a
5168 macro that called string_cat() twice, in order to add two strings. This is
5169 not really sufficiently general. Furthermore, there was one instance where
5170 it was actually wrong because one of the argument was used twice, and in
5171 one call a function was used. (As it happened, calling the function twice
5172 did not affect the overall behaviour.) The macro has been replaced by a
5173 function that can join an arbitrary number of extra strings onto a growing
5176 (c) The code for expansion conditions now uses a table and a binary chop
5177 instead of a serial search (which was left over from when there were very
5178 few conditions). Also, it now recognizes conditions like "pam" even when
5179 the relevant support is not compiled in: a suitably worded error message is
5180 given if an attempt is made to use such a condition.
5182 11. Added ${time_interval:xxxxx}.
5184 12. A bug was causing one of the ddress fields not to be passed back correctly
5185 from remote delivery subprocesses. The field in question was not being
5186 subsequently used, so this caused to problems in practice.
5188 13. Added new log selectors queue_time and deliver_time.
5190 14. Might have fixed a bug in maildirsizefile handling that threw up
5191 "unexpected character" debug warnings, and recalculated the data
5192 unnecessarily. In any case, I expanded the warning message to give more
5195 15. Added the message "Restricted characters in address" to the statements in
5196 the default ACL that block characters like @ and % in local parts.
5198 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5199 Three changes have been made:
5201 (a) There was a serious bug; a negative response to MAIL caused the whole
5202 recipient domain to be cached as invalid, thereby blocking all messages
5203 to all local parts at the same domain, from all senders. This bug has
5204 been fixed. The domain is no longer cached after a negative response to
5205 MAIL if the sender used is not empty.
5207 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5210 (c) A new callout option, "use_sender" has been added for people who want
5211 the modified behaviour.
5217 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5220 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5221 indeed breaks things for older releases.
5223 3. Added additional logging to the case where there is a problem reading data
5224 from a filter that is running in a subprocess using a pipe, in order to
5225 try to track down a specific problem.
5227 4. Testing facility fudge: when running in the test harness and attempting
5228 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5229 getting "No route to host". Convert this to a timeout.
5231 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5234 6. Some OS don't have socklen_t but use size_t instead. This affects the
5235 fifth argument of getsockopt() amongst other things. This is now
5236 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5237 can be set for individual OS. I have set it for SunOS5, OSF1, and
5238 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5239 some earlier ones do not.
5241 7. Change 4.30/15 was not doing the test caselessly.
5243 8. The standard form for an IPv6 address literal was being rejected by address
5244 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5245 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5246 this, as well as the form without the "IPv6" on the front (but only when
5247 address literals are enabled, of course).
5249 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5251 10. Exim crashed if a message with an empty sender address specified by -f
5252 encountered a router with an errors_to setting. This could be provoked only
5253 by a command such as
5257 where an empty string was supplied; "<>" did not hit this bug.
5259 11. Installed PCRE release 4.5.
5261 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5262 remained set. It is now erased.
5264 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5265 times from message ids (which are base 36 rather than the normal 62).
5267 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5268 were being counted as actual protocol errors, and logged if the log
5269 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5270 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5271 and DATA following a set of rejected RCPTs do not count as protocol errors.
5272 In other words, Exim assumes they were pipelined, though this may not
5273 actually be the case. Of course, in all cases the client gets an
5274 appropriate error code.
5276 15. If a lookup fails in an ACL condition, a message about the failure may
5277 be available; it is used if testing the ACL cannot continue, because most
5278 such messages specify what the cause of the deferral is. However, some
5279 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5280 that caused an old message to be retained and used if a later statement
5281 caused a defer, replacing the real cause of the deferral.
5283 16. If an IP address had so many PTR records that the DNS lookup buffer
5284 was not large enough to hold them, Exim could crash while trying to process
5285 the truncated data. It now detects and logs this case.
5287 17. Further to 4.21/58, another change has been made: if (and only if) the
5288 first line of a message (the first header line) ends with CRLF, a bare LF
5289 in a subsequent header line has a space inserted after it, so as not to
5290 terminate the header.
5292 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5293 unnecessarily, used atoi() instead of strtol(), and didn't check the
5294 termination when getting file sizes from file names by regex.
5296 19. Completely re-implemented the support for maildirsize files, in the light
5297 of a number of problems with the previous contributed implementation
5298 (4.30/29). In particular:
5300 . If the quota is zero, the maildirsize file is maintained, but no quota is
5303 . If the maildir directory does not exist, it is created before any attempt
5304 to write a maildirsize file.
5306 . The quota value in the file is just a cache; if the quota is changed in
5307 the transport, the new value overrides.
5309 . A regular expression is available for excluding directories from the
5312 20. The autoreply transport checks the characters in options that define the
5313 message's headers; it allows continued headers, but it was checking with
5314 isspace() after an embedded newline instead of explicitly looking for a
5317 21. If all the "regular" hosts to which an address was routed had passed their
5318 expiry times, and had not reached their retry times, the address was
5319 bounced, even if fallback hosts were defined. Now Exim should go on to try
5322 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5323 equivalent code in the SMTP transport. Some hosts send humungous responses
5324 to HELO/EHLO, more than 1024 it seems.
5326 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5327 gives compiler warnings in some environments. I've now done it "properly",
5330 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5331 (because of problems with the built-in one) was declared to return uschar *
5332 instead of char *, causing compiler failure.
5334 25. Fixed a file descriptor leak when processing alias/forward files.
5336 26. Fixed a minor format string issue in dbfn.c.
5338 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5340 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5341 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5344 29. When a sender address is verified, it is cached, to save repeating the test
5345 when there is more than one recipient in a message. However, when the
5346 verification involves a callout, it is possible for different callout
5347 options to be set for different recipients. It is too complicated to keep
5348 track of this in the cache, so now Exim always runs a verification when a
5349 callout is required, relying on the callout cache for the optimization.
5350 The overhead is duplication of the address routing, but this should not be
5353 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5354 to be verified with callout=postmaster, and the main callout worked but the
5355 postmaster check failed, the verification correctly failed. However, if a
5356 subsequent RCPT command asked for sender verification *without* the
5357 postmaster check, incorrect caching caused this verification also to fail,
5360 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5361 it was not caching the DNS options (qualify_single, search_parents) that
5362 were used when the lookup failed. A subsequent lookup with different
5363 options therefore always gave the same answer, though there were cases
5364 where it should not have. (Example: a "domains = !$mx_any" option on a
5365 dnslookup router: the "domains" option is always processed without any
5366 widening, but the router might have qualify_single set.) Now Exim uses the
5367 cached value only when the same options are set.
5369 32. Added John Jetmore's "exipick" utility to the distribution.
5371 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5372 than a timeout (e.g. a certificate is required, and is not provided), an
5373 Exim server now closes the connection immediately. Previously it waited for
5374 the client to close - but if the client is SSL, it seems that they each
5375 wait for each other, leading to a delay before one of them times out.
5377 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5378 maintained 0.8.x compatibility because I don't think many are using it, and
5379 it is clearly obsolete.
5381 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5384 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5385 client certificate was expired. A simple patch fixes this, though I don't
5386 understand the full logic of why the verify callback is called multiple
5389 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5390 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5391 which causes problems with some clients (such as the Certicom SSL Plus
5392 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5393 disables the coutermeasure allowing Eudora to connect."
5395 38. Exim was not checking that a write() to a log file succeeded. This could
5396 lead to Bad Things if a log got too big, in particular if it hit a file
5397 size limit. Exim now panics and dies if it cannot write to a log file, just
5398 as it does if it cannot open a log file.
5400 39. Modified OS/Makefile-Linux so that it now contains
5402 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5404 The two -D definitions ensure that Exim is compiled with large file
5405 support, which makes it possible to handle log files that are bigger than
5408 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5409 instance) a domain was checked against a named list that involved a lookup,
5410 causing $domain_data to be set, then another domain was checked against the
5411 same list, then the first domain was re-checked, the value of $domain_data
5412 after the final check could be wrong. In particular, if the second check
5413 failed, it could be set empty. This bug probably also applied to
5416 41. The strip_trailing_dot option was not being applied to the address given
5417 with the -f command-line option.
5419 42. The code for reading a message's header from the spool was incrementing
5420 $received_count, but never initializing it. This meant that the value was
5421 incorrect (doubled) while delivering a message in the same process in which
5422 it was received. In the most common configuration of Exim, this never
5423 happens - a fresh exec is done - but it can happen when
5424 deliver_drop_privilege is set.
5426 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5427 it now includes up to 150 characters of the unexpected data in the log
5430 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5431 and building data strings. The size of both of these buffers was 10 000
5432 bytes - far larger than anybody would *ever* want, thought I. Needless to
5433 say, somebody hit the limit. I have increased the maximum line length to
5434 20 000 and the maximum data length of concatenated lines to 100 000. I have
5435 also fixed two bugs, because there was no checking on these buffers. Tsk,
5436 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5437 buffer is too small.
5439 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5440 lsearch lookups. Now it does.
5442 46. When parsing a route_list item in a manualroute router, a fixed-length
5443 buffer was used for the list of hosts. I made this 1024 bytes long,
5444 thinking that nobody would ever have a list of hosts that long. Wrong.
5445 Somebody had a whole pile of complicated expansion conditions, and the
5446 string was silently truncated, leading to an expansion error. It turns out
5447 that it is easier to change to an unlimited length (owing to other changes
5448 that have happened since this code was originally written) than to build
5449 structure for giving a limitation error. The length of the item that
5450 expands into the list of hosts is now unlimited.
5452 47. The lsearch lookup could not handle data where the length of text line was
5453 more than 4095 characters. Such lines were truncated, leading to shortened
5454 data being returned. It should now handle lines of any length.
5456 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5457 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5460 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5461 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5462 no longer happens in comment lines. A list of replacements is now placed
5463 at the head of all of the source files, except those whose only change is
5464 to replace PERL_COMMAND in the very first #! line.
5466 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5467 messages on the queue, with a bottom-up merge sort, using code contributed
5468 by Michael Haardt. This should make operations like -bp somewhat faster on
5469 large queues. It won't affect queue runners, except when queue_run_in_order
5472 51. Installed eximstats 1.31 in the distribution.
5474 52. Added support for SRV lookups to the dnslookup router.
5476 53. If an ACL referred to $message_body or $message_body_end, the value was not
5477 reset for any messages that followed in the same SMTP session.
5479 54. The store-handling optimization for building very long strings was not
5480 differentiating between the different store pools. I don't think this
5481 actually made any difference in practice, but I've tidied it.
5483 55. While running the routers to verify a sender address, $sender_address
5484 was still set to the sender address. This is wrong, because when routing to
5485 send a bounce to the sender, it would be empty. Therefore, I have changed
5486 it so that, while verifying a sender address, $sender_address is set to <>.
5487 (There is no change to what happens when verifying a recipient address.)
5489 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5490 target A or AAAA records (if not already returned) without resetting the
5491 qualify_single or search_parents options of the DNS resolver. These are
5492 inappropriate in this case because the targets of MX and SRV records must
5493 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5494 target that, when qualified, matched something in the local domain. These
5495 two options are now turned off when doing these lookups.
5497 57. It seems that at least some releases of Reiserfs (which does not have the
5498 concept of a fixed number of inodes) returns zero and not -1 for the
5499 number of available inodes. This interacted badly with check_spool_inodes,
5500 which assumed that -1 was the "no such thing" setting. What I have done is
5501 to check that the total number of inodes is greater than zero before doing
5502 the test of how many are available.
5504 58. When a "warn" ACL statement has a log_message modifier, the message is
5505 remembered, and not repeated. This is to avoid a lot of repetition when a
5506 message has many recipients that cause the same warning to be written.
5507 Howewer, Exim was preserving the list of already written lines for an
5508 entire SMTP session, which doesn't seem right. The memory is now reset if a
5509 new message is started.
5511 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5512 debugging output unless log_rewrite was also set.
5514 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5515 of (int)(handle) when we know that handle contains (void *)(-1).
5517 61. The Exim daemon panic-logs an error return when it closes the incoming
5518 connection. However "connection reset by peer" seems to be common, and
5519 isn't really an error worthy of noting specially, so that particular error
5522 62. When Exim is trying to find all the local interfaces, it used to panic and
5523 die if the ioctl to get the interface flags failed. However, it seems that
5524 on at least one OS (Solaris 9) it is possible to have an interface that is
5525 included in the list of interfaces, but for which you get a failure error
5526 for this call. This happens when the interface is not "plumbed" into a
5527 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5528 failure of the "get flags" call assumes that the interface is down.
5530 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5531 makes life easier for people who are doing arithmetic on fields extracted
5532 from dates, where you often get leading zeros that should not be
5533 interpreted as octal.
5535 64. Added qualify_domain to the redirect router, to override the global
5538 65. If a pathologically long header line contained very many addresses (the
5539 report of this problem mentioned 10 000) and each of them was rewritten,
5540 Exim could use up a very large amount of memory. (It kept on making new
5541 copies of the header line as it rewrote, and never released the old ones.)
5542 At the expense of a bit more processing, the header rewriting function has
5543 been changed so that it no longer eats memory in this way.
5545 66. The generation of the Received: header has been moved from the time that a
5546 message starts to be received, to the time that it finishes. The timestamp
5547 in the Received: header should now be very close to that of the <= log
5548 line. There are two side-effects of this change:
5550 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5551 logged header lines no longer include the local Received: line, because
5552 it has not yet been created. The same applies to a copy of the message
5553 that is returned to a non-SMTP sender when a message is rejected.
5555 (b) When a filter file is tested using -bf, no additional Received: header
5556 is added to the test message. After some thought, I decided that this
5559 This change does not affect the value of $received_for. It is still set
5560 after address rewriting, but before local_scan() is called.
5562 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5564 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5565 gave an unhelpful panic error message, and a defer error. I have managed to
5566 change this behaviour so that it now rejects any supplied certificate,
5567 which seems right, as the list of acceptable certificates is empty.
5569 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5570 gave an unhelpful defer error. I have not managed to make this reject any
5571 supplied certificates, but the error message it gives is "no certificate
5572 supplied", which is not helpful.
5574 70. exigrep's output now also includes lines that are not associated with any
5575 message, but which match the given pattern. Implemented by a patch from
5576 Martin Sluka, which also tidied up the Perl a bit.
5578 71. Recipient callout verification, like sender verification, was using <> in
5579 the MAIL FROM command. This isn't really the right thing, since the actual
5580 sender may affect whether the remote host accepts the recipient or not. I
5581 have changed it to use the actual sender in the callout; this means that
5582 the cache record is now keyed on a recipient/sender pair, not just the
5583 recipient address. There doesn't seem to be a real danger of callout loops,
5584 since a callout by the remote host to check the sender would use <>.
5585 [SEE ABOVE: changed after hitting problems.]
5587 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5588 temporary errors. However, in the case of such a code being given after
5589 the end of a data transmission (i.e. after ".") Exim was failing to write
5590 a retry record for the message. (Yes, there was some broken host that was
5591 actually sending 8xx at this point.)
5593 73. An unknown lookup type in a host list could cause Exim to panic-die when
5594 the list was checked. (An example that provoked this was putting <; in the
5595 middle of a list instead of at the start.) If this happened during a DATA
5596 ACL check, a -D file could be left lying around. This kind of configuration
5597 error no longer causes Exim to die; instead it causes a defer errror. The
5598 incident is still logged to the main and panic logs.
5600 74. Buglet left over from Exim 3 conversion. The message "too many messages
5601 in one connection" was written to the rejectlog but not the mainlog, except
5602 when address rewriting (yes!) was being logged.
5604 75. Added write_rejectlog option.
5606 76. When a system filter was run not as root (that is, when system_filter_user
5607 was set), the values of the $n variables were not being returned to the
5608 main process; thus, they were not subsequently available in the $sn
5611 77. Added +return_path_on_delivery log selector.
5613 78. A connection timeout was being treated differently from recipients deferred
5614 when testing hosts_max_try with a message that was older than the host's
5615 retry timeout. (The host should not be counted, thus allowing all hosts to
5616 be tried at least once before bouncing.) This may have been the cause of an
5617 occasionally reported bug whereby a message would remain on the queue
5618 longer than the retry timeout, but would be bounced if a delivery was
5619 forced. I say "may" because I never totally pinned down the problem;
5620 setting up timeout/retry tests is difficult. See also the next item.
5622 79. The ultimate address timeout was not being applied to errors that involved
5623 a combination of host plus message (for example, a timeout on a MAIL
5624 command). When an address resolved to a number of possible hosts, and they
5625 were not all tried for each delivery (e.g. because of hosts_max_try), a
5626 message could remain on the queue longer than the retry timeout.
5628 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5631 81. Fixed an obscure SMTP outgoing bug which required at least the following
5632 conditions: (a) there was another message waiting for the same server;
5633 (b) the server returned 5xx to all RCPT commands in the first message so
5634 that the message was not completed; (c) the server dropped the connection
5635 or gave a negative response to the RSET that Exim sends to abort the
5636 transaction. The observed case was a dropped connection after DATA that had
5637 been sent in pipelining mode. That is, the server had advertised PIPELINING
5638 but was not implementing it correctly. The effect of the bug was incorrect
5639 behaviour, such as trying another host, and this could lead to a crash.
5645 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5646 and daemon.c were passed as pointers to ints; they should have been
5647 pointers to socklen_t variables (which are typically unsigned ints).
5649 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5652 3. Fixed a really odd bug that affected only the testing scheme; patching a
5653 certain fixed string in the binary changed the value of another string that
5654 happened to be identical to the end of the original first string.
5656 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5657 name", it returns that address as the IP address. On some operating
5658 systems (e.g. Solaris), it also passes back the IP address string as the
5659 "host name". However, on others (e.g. Linux), it passes back an empty
5660 string. Exim wasn't checking for this, and was changing the host name to an
5661 empty string, assuming it had been canonicized.
5663 5. Although rare, it is permitted to have more than one PTR record for a given
5664 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5665 all the names associated with an address, because they do in Solaris.
5666 However, it seems that they do not in Linux for data that comes from the
5667 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5668 I found this out when I moved to a new Linux workstation and tried to run
5669 the Exim test suite.
5671 To get round this problem I have changed the code so that it now does its
5672 own call to the DNS to look up PTR records when searching for a host name.
5673 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5674 addresses that are only in /etc/hosts are still found.
5676 This behaviour is, however, controlled by an option called host_lookup_
5677 order, which defaults to "bydns:byaddr". If people want to use the other
5678 order, or indeed, just use one or the other means of lookup, they can
5679 specify it in this variable.
5681 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5682 some operating systems, this comes back from gethostbyaddr() as an empty
5683 string, and this is what Exim used to test for. However, it seems that in
5684 other systems, "." is yielded. Exim now tests for this case too.
5686 7. The values of check_spool_space and check_log_space are now held internally
5687 as a number of kilobytes instead of an absolute number of bytes. If a
5688 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5689 kilobyte. This means that much larger values can be stored.
5691 8. Exim monitor: an attempt to get the action menu when not actually pointing
5692 at a message produces an empty menu entitled "No message selected". This
5693 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5694 no entries in it ("Shell widget menu has zero width and/or height"). So I
5695 have added a single, blank menu entry in this case.
5697 9. Added ${quote_local_part.
5699 10. MIME decoding is now applied to the contents of Subject: header lines when
5702 11. Now that a reference to $sender_host_address automatically causes a reverse
5703 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5704 host lookup before query-style lookups in lists that might use this
5705 variable. This has therefore been abolished, and the "net-" prefix is no
5706 longer necessary for query-style lookups.
5708 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5709 have been a typo for LFLAGS, so it has been changed.
5711 13. The install script calls Exim with "-C /dev/null" in order to find the
5712 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5713 to be output. Howeve, since Exim outputs its version number before the
5714 error, it didn't break the script. It just looked ugly. I fixed this by
5715 always allowing "-C /dev/null" if the caller is root.
5717 14. Ignore overlarge ACL variable number when reading spool file - insurance
5718 against a later release with more variables having written the file.
5720 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5721 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5722 this, as well as the form without the "IPv6" on the front.
5724 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5725 OS/Makefile-Darwin file.
5727 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5728 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5730 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5731 variable. However, some OS use ints for inodes. I've added cast to long int
5732 to get rid of the compiler warning.
5734 19. I had forgotten to lock out "/../" in configuration file names when
5735 ALT_CONFIG_PREFIX was set.
5737 20. Routers used for verification do not need to specify transports. However,
5738 if such a router generated a host list, and callout was configured, Exim
5739 crashed, because it could not find a port number from the (non-existent)
5740 transport. It now assumes port 25 in this circumstance.
5742 21. Added the -t option to exigrep.
5744 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5745 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5746 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5747 them is set, LOOKUP_LSEARCH is forced.
5749 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5751 24. Added sender and host information to the "rejected by local_scan()" log
5752 line; previously there was no indication of these.
5754 25. Added .include_if_exists.
5756 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5757 Linux. It turns out that not all file systems support this. Apparently some
5758 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5759 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5760 not supported on the file descriptor, is now ignored when Exim is trying to
5761 sync a directory. This applies only to Linux.
5763 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5765 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5766 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5769 29. Added support for maildirsize files from supplied patch (modified a bit).
5771 30. The use of :fail: followed by an empty string could lead Exim to respond to
5772 sender verification failures with (e.g.):
5774 550 Verification failed for <xxx>
5775 550 Sender verify failed
5777 where the first response line was missing the '-' that indicates it is not
5778 the final line of the response.
5780 31. The loop for finding the name of the user that called Exim had a hardwired
5781 limit of 10; it now uses the value of finduser_retries, which is used for
5782 all other user lookups.
5784 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5787 33. Exim was neglecting to zero errno before one call of strtol() when
5788 expanding a string and expecting an integer value. On some systems this
5789 resulted in spurious "integer overflow" errors. Also, it was casting the
5790 result into an int without checking.
5792 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5793 did not work. The code looks as if it has *never* worked, though it appears
5794 to have been documented since at least releast 1.62. I have made it work.
5796 35. The "timeout_DNS" error in retry rules, also documented since at least
5797 1.62, also never worked. As it isn't clear exactly what this means, and
5798 clearly it isn't a major issue, I have abolished the feature by treating it
5799 as "timeout", and writing a warning to the main and panic logs.
5801 36. The display of retry rules for -brt wasn't always showing the error code
5804 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5805 timeout_connect_A, timeout_connect_MX.
5807 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5808 to the empty sender.
5810 39. The daemon was not analysing the content of -oX till after it had closed
5811 stderr and disconnected from the controlling terminal. This meant that any
5812 syntax errors were only noted on the panic log, and the return code from
5813 the command was 0. By re-arranging the code a little, I've made the
5814 decoding happen first, so such errors now appear on stderr, and the return
5815 code is 1. However, the actual setting up of the sockets still happens in
5816 the disconnected process, so errors there are still only recorded on the
5819 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5820 connections (as happens on some IP stacks) was logged at start up time as
5821 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5822 it from "IPv6 and IPv4", which means that two separate sockets are being
5825 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5826 says whether AF_INET or AF_INET6 was passed as an argument.
5828 42. Exiwhat output was messed up when time zones were included in log
5831 43. Exiwhat now gives more information about the daemon's listening ports,
5832 and whether -tls-on-connect was used.
5834 44. The "port" option of the smtp transport is now expanded.
5836 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5837 silently ignored. Now an error message is written to the main and panic
5840 46. There's a new ACL modifier called "logwrite" which writes to a log file
5841 as soon as it is encountered.
5843 47. Added $local_user_uid and $local_user_gid at routing time.
5845 48. Exim crashed when trying to verify a sender address that was being
5848 49. Exim was recognizing only a space character after ".include". It now also
5849 recognizes a tab character.
5851 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5852 extracting the relevant information from the specification. The man page no
5853 longer contains scrambled data for the -d option, and I've added a section
5854 at the front about calling Exim under different names.
5856 51. Added "extra_headers" argument to the "mail" command in filter files.
5858 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5861 53. Installed eximstats 1.29.
5863 54. Added transport_filter_timeout as a generic transport option.
5865 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5866 Cc: header lines. This was required by RFC 822, but it not required by RFC
5869 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5870 incoming messages that did not have them. Now it does so only if the
5871 message originates locally, that is, if there is no associated remote host
5872 address. When Resent- header lines are present, this applies to the Resent-
5873 lines rather than the non-Resent- lines.
5875 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5876 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5878 58. Messages for configuration errors now include the name of the main
5879 configuration file - useful now that there may be more than one file in a
5880 list (.included file names were always shown).
5882 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5883 for those rare installations that do not start the daemon as root or run it
5884 setuid root. I've cut out the call to initgroups() if the daemon is not
5887 60. The Exim user and group can now be bound into the binary as text strings
5888 that are looked up at the start of Exim's processing.
5890 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5892 62. Added $mailstore_basename variable.
5894 63. Installed patch to sieve.c from Michael Haardt.
5896 64. When Exim failed to open the panic log after failing to open the main log,
5897 the original message it was trying to log was written to stderr and debug
5898 output, but if they were not available (the usual case in production), it
5899 was lost. Now it is written to syslog before the two lines that record the
5900 failures to open the logs.
5902 65. Users' Exim filters run in subprocesses under the user's uid. It is
5903 possible for a "deliver" command or an alias in a "personal" command to
5904 provoke an address rewrite. If logging of address rewriting is configured,
5905 this fails because the process is not running as root or exim. There may be
5906 a better way of dealing with this, but for the moment (because 4.30 needs
5907 to be released), I have disabled address rewrite logging when running a
5908 filter in a non-root, non-exim process.
5914 1. The buildconfig auxiliary program wasn't quoting the value set for
5915 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5916 not defined. This bug was masked in 4.22 by the effect that was fixed in
5919 2. Some messages that were rejected after a message id was allocated were
5920 shown as "incomplete" by exigrep. It no longer does this for messages that
5921 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5923 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5924 have allow_domain_literals set, the ID did not get logged in the <= line.
5925 Domain literals are now always recognized in Message-ID: header lines.
5927 4. The first argument for a ${extract expansion item is the key name or field
5928 number. Leading and trailing spaces in this item were not being ignored,
5929 causing some misleading effects.
5931 5. When deliver_drop_privilege was set, single queue runner processes started
5932 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5933 same command in the process it spins off) were not dropping privilege.
5935 6. When the daemon running as "exim" started a queue runner, it always
5936 re-executed Exim in the spun-off process. This is a waste of effort when
5937 deliver_drop_privilege is set. The new process now just calls the
5938 queue-runner function directly.
5944 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5947 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5948 ignored. Though the use of -oP was forcing the writing of a pid file, it
5949 was always written to the default place.
5951 3. If the message "no IP address found for host xxxx" is generated during
5952 incoming verification, it is now followed by identification of the incoming
5953 connection (so you can more easily find what provoked it).
5955 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5957 5. Added some features to "harden" Exim a bit more against certain attacks:
5959 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5960 be put in Local/Makefile. This is like the never_users runtime option,
5961 but it cannot be overridden. The default setting is "root".
5963 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5964 prefix string with which any file named in a -C command line option
5967 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5968 is retained for -C and -D only if the caller of Exim is root. Without
5969 it, the exim user may also use -C and -D and retain privilege.
5971 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5972 command line option is disabled.
5974 6. Macro names set by the -D option must start with an upper case letter, just
5975 like macro names defined in the configuration file.
5977 7. Added "dereference=" facility to LDAP.
5979 8. Two instances of the typo "uknown" in the source files are fixed.
5981 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5982 the Configure-Makefile script screwed up while processing it.
5984 10. Incorporated PCRE 4.4.
5986 11. The SMTP synchronization check was not operating right at the start of an
5987 SMTP session. For example, it could not catch a HELO sent before the client
5988 waited for the greeting. There is now a check for outstanding input at the
5989 point when the greeting is written. Because of the duplex, asynchronous
5990 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5991 way, but not yet received, when the check is performed.
5993 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5994 on TCP/IP sockets, because this apparently causes some broken clients to
5997 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5998 unchanged) from the Cygwin maintainer.
6000 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6001 was supported instead of testing for mbx. Effectively a typo.
6003 15. The spa authenticator server code was not checking that the input it
6004 received was valid base64.
6006 16. The debug output line for the "set" modifier in ACLs was not showing the
6007 name of the variable that was being set.
6009 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6011 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6012 lookup on its own. Something else was needed to trigger the lookup. For
6013 example, a match in host_lookup or the need for a host name in a host list.
6014 Now, if $sender_host_name is referenced and the host name has not yet been
6015 looked up, a lookup is performed. If the lookup fails, the variable remains
6016 empty, and $host_lookup_failed is set to "1".
6018 19. Added "eqi" as a case-independent comparison operator.
6020 20. The saslauthd authentication condition could segfault if neither service
6021 nor realm was specified.
6023 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6024 error message that was logged was misleading, and incoming SMTP
6025 connections were dropped. The message is now more accurate, and temporary
6026 errors are given to SMTP connections.
6028 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6029 (except RSET) with a 421 error, until QUIT is received. However, it was
6030 failing to send a response to QUIT.
6032 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6033 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6034 ACL is now run after the helo_try_verify_hosts code.
6036 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6037 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6038 case-independent, so other case variants are also recognized). Apparently
6039 some systems use these upper case variants.
6041 25. If more than two messages were waiting for the same host, and a transport
6042 filter was specified for the transport, Exim sent two messages over the
6043 same TCP/IP connection, and then failed with "socket operation on non-
6044 socket" when it tried to send the third.
6046 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6048 27. The extern definition of crypt16() in expand.c was not being excluded when
6049 the OS had its own crypt16() function.
6051 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6052 as a preferred synonym for return_size_limit, both as an option and as an
6055 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6057 30. Changed the default configuration ACL to relax the local part checking rule
6058 for addresses that are not in any local domains. For these addresses,
6059 slashes and pipe symbols are allowed within local parts, but the sequence
6060 /../ is explicitly forbidden.
6062 31. SPA server authentication was not clearing the challenge buffer before
6065 32. log_message in a "warn" ACL statement was writing to the reject log as
6066 well as to the main log, which contradicts the documentation and doesn't
6067 seem right (because no rejection is happening). So I have stopped it.
6069 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6070 However, I am unable to do any testing of this.
6072 34. Fixed an infelicity in the appendfile transport. When checking directories
6073 for a mailbox, to see if any needed to be created, it was accidentally
6074 using path names with one or more superfluous leading slashes; tracing
6075 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6077 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6078 contents are added to the log line that is written for every discarded
6079 recipient. (Previously a log_message setting was ignored.)
6081 36. The ${quote: operator now quotes the string if it is empty.
6083 37. The install script runs exim in order to find its version number. If for
6084 some reason other than non-existence or emptiness, which it checks, it
6085 could not run './exim', it was installing it with an empty version number,
6086 i.e. as "exim-". This error state is now caught, and the installation is
6089 38. An argument was missing from the function that creates an error message
6090 when Exim fails to connect to the socket for saslauthd authentication.
6091 This could cause Exim to crash, or give a corrupted message.
6093 39. Added isip, isip4, and isip6 to ${if conditions.
6095 40. The ACL variables $acl_xx are now saved with the message, and can be
6096 accessed later in routers, transports, and filters.
6098 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6099 strings in the file are not string-expanded.
6101 42. If a MAIL command specified a SIZE value that was too large to fit into an
6102 int variable, the check against message_size_limit failed. Such values are
6103 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6104 day this will have to be increased, but I don't think I want to be around
6105 when emails are that large.
6112 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6113 iconv() is not standard in FreeBSD.
6115 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6116 IPv6 enabled. The observed symptom was a segmentation fault on return from
6117 the function os_common_find_running_interfaces() in src/os.c.
6119 3. In the check_special_case() function in daemon.c I had used "errno" as an
6120 argument name, which causes warnings on some systems. This was basically a
6121 typo, since it was named "eno" in the comments!
6123 4. The code that waits for the clock to tick (at a resolution of some fraction
6124 of a second) so as to ensure message-id uniqueness was always waiting for
6125 at least one whole tick, when it could have waited for less. [This is
6126 almost certainly not relevant at current processor speeds, where it is
6127 unlikely to ever wait at all. But we try to future-proof.]
6129 5. The function that sleeps for a time interval that includes fractions of a
6130 second contained a race. It did not block SIGALRM between setting the
6131 timer, and suspending (a couple of lines later). If the interval was short
6132 and the sigsuspend() was delayed until after it had expired, the suspension
6133 never ended. On busy systems this could lead to processes getting stuck for
6136 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6137 process, before it forks any delivery processes. The open lookup caching
6138 mechanism meant that the open file or database connection was passed into
6139 the delivery process. The problem was that delivery processes always tidy
6140 up cached lookup data. This could cause a problem for the next delivery
6141 process started by the queue runner, because the external queue runner
6142 process does not know about the closure. So the next delivery process
6143 still has data in the lookup cache. In the case of a file lookup, there was
6144 no problem because closing a file descriptor in a subprocess doesn't affect
6145 the parent. However, if the lookup was caching a connection to a database,
6146 the connection was closed, and the second delivery process was likely to
6147 see errors such as "PGSQL: query failed: server closed the connection
6148 unexpectedly". The problem has been fixed by closing all cached lookups
6149 in a queue runner before running a delivery process.
6151 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6152 seem to have the "const" qualifier which it has on other OS. I've
6155 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6156 *of the same type* (client or server) with the same public name that an
6157 error should be diagnosed.
6159 9. When Exim looked up a host name for an IP address, but failed to find the
6160 original IP address when looking up the host name (a safety check), it
6161 output the message "<ip address> does not match any IP for NULL", which was
6162 confusing, to say the least. The bug was that the host name should have
6163 appeared instead of "NULL".
6165 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6166 user that is built into the binary, and the -C or -D options is used, root
6167 privilege is dropped before the configuration file is read. In addition,
6168 logging is switched to stderr instead of the normal log files. If the
6169 configuration then re-defines the Exim user, the unprivileged environment
6170 is probably not what is expected, so Exim logs a panic warning message (but
6173 However, if deliver_drop_privilege is set, the unprivileged state may well
6174 be exactly what is intended, so the warning has been cut out in that case,
6175 and Exim is allowed to try to write to its normal log files.
6181 1. smtp_return_error_details was not giving details for temporary sender
6182 or receiver verification errors.
6184 2. Diagnose a configuration error if two authenticators have the same public
6187 3. Exim used not to create the message log file for a message until the first
6188 delivery attempt. This could be confusing when incoming messages were held
6189 for policy or load reasons. The message log file is now created at the time
6190 the message is received, and an initial "Received" line is written to it.
6192 4. The automatically generated man page for command line options had a minor
6193 bug that caused no ill effects; however, a more serious problem was that
6194 the procedure for building the man page automatically didn't always
6195 operate. Consequently, release 4.20 contains an out-of-date version. This
6196 shouldn't happen again.
6198 5. When building Exim with embedded Perl support, the script that builds the
6199 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6200 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6202 6. The freeze_tell option was not being used for messages that were frozen on
6203 arrival, either by an ACL or by local_scan().
6205 7. Added the smtp_incomplete_transaction log selector.
6207 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6208 was accepting AUTH without a new EHLO.
6210 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6211 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6214 10. Logging of TCP/IP connections (when configured) now happens in the main
6215 daemon process instead of the child process, so that the TCP/IP connection
6216 count is more accurate (but it can never be perfect).
6218 11. The use of "drop" in a nested ACL was not being handled correctly in the
6219 outer ACL. Now, if condition failure induced by the nested "drop" causes
6220 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6221 or "require"), the connection is dropped.
6223 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6224 that yield "discard" can now be used with an "accept" or a "discard" verb,
6225 but an error is generated for any others (because I can't see a useful way
6226 to define what should happen).
6228 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6229 are now processed in the same way as lines in the Exim configuration file.
6230 In particular, continuation lines are supported.
6232 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6234 15. Added -ti meaning -t -i.
6236 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6237 domains, and warn by logging if others are found.
6239 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6240 returned from GIFCONF when Exim is trying to find the list of interfaces on
6241 a host. The code in os.c has been modified to copy each ifreq to an aligned
6242 structure in all cases.
6244 Also, in some cases, the returned ifreq's were being copied to a 'struct
6245 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6246 means the last couple of bytes of an IPv6 address could be chopped if the
6247 ifreq contained only a normal sockaddr (14 bytes storage).
6249 18. Named domain lists were not supported in the hosts_treat_as_local option.
6250 An entry such as +xxxx was not recognized, and was treated as a literal
6253 19. Ensure that header lines added by a DATA ACL are included in the reject log
6254 if the ACL subsequently rejects the message.
6256 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6257 MD5 (which is deprecated).
6259 21. When testing a filter file using -bf, Exim was writing a message when it
6260 took the sender from a "From " line in the message, but it was not doing so
6261 when it took $return_path from a Return-Path: header line. It now does.
6263 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6264 with a valid header line field name (a series of printing characters
6265 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6267 23. Changed "disc" in the source to "disk" to conform to the documentation and
6268 the book and for uniformity.
6270 24. Ignore Sendmail's -Ooption=value command line item.
6272 25. When execve() failed while trying to run a command in a pipe transport,
6273 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6274 could be confused with a return value of 69 from the command itself. This
6275 has been changed to 127, the value the shell returns if it is asked to run
6276 a non-existent command. The wording for the related log line suggests a
6277 non-existent command as the problem.
6279 26. If received_header_text expands to an empty string, do not add a Received:
6280 header line to the message. (Well, it adds a token one on the spool, but
6281 marks it "old" so that it doesn't get used or transmitted.)
6283 27. Installed eximstats 1.28 (addition of -nt option).
6285 28. There was no check for failure on the call to getsockname() in the daemon
6286 code. This can fail if there is a shortage of resources on the system, with
6287 ENOMEM, for example. A temporary error is now given on failure.
6289 29. Contrary to the C standard, it seems that in some environments, the
6290 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6291 program. Exim now does this explicitly; it affects the formatting of
6292 timestamps using strftime().
6294 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6295 complaints. I've now initialized all the variables, to avoid this.
6297 32. Header lines added by a system filter were not being "seen" during
6298 transport-time rewrites.
6300 33. The info_callback() function passed to OpenSSL is set up with type void
6301 (*)(SSL *, int, int), as described somewhere. However, when calling the
6302 function (actually a macro) that sets it up, the type void(*)() is
6303 expected. I've put in a cast to prevent warnings from picky compilers.
6305 34. If a DNS black list lookup found a CNAME record, but there were no A
6306 records associated with the domain it pointed at, Exim crashed.
6308 35. If a DNS black list lookup returned more than one A record, Exim ignored
6309 all but the first. It now scans all returned addresses if a particular IP
6310 value is being sought. In this situation, the contents of the
6311 $dnslist_value variable are a list of all the addresses, separated by a
6314 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6315 to accept a host name and all its aliases if the forward lookup for any of
6316 them yielded the IP address of the incoming connection. Now it accepts only
6317 those names whose forward lookup yields the correct IP address. Any other
6318 names are discarded. This closes a loophole whereby a rogue DNS
6319 administrator could create reverse DNS records to break through a
6320 wildcarded host restriction in an ACL.
6322 37. If a user filter or a system filter that ran in a subprocess used any of
6323 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6324 the wrong values were passed to the pipe command ($thisaddress had the
6325 value of $0, $0 had the value of $1, etc). This bug was introduced by
6326 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6328 38. Improved the line breaking for long SMTP error messages from ACLs.
6329 Previously, if there was no break point between 40 and 75 characters, Exim
6330 left the rest of the message alone. Two changes have been made: (a) I've
6331 reduced the minimum length to 35 characters; (b) if it can't find a break
6332 point between 35 and 75 characters, it looks ahead and uses the first one
6333 that it finds. This may give the occasional overlong line, but at least the
6334 remaining text gets split now.
6336 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6337 file descriptors might be low, and that setting 1000 would always raise it.
6338 It turns out that in some environments, the limit is already over 1000 and
6339 that lowering it causes trouble. So now Exim takes care not to decrease it.
6341 40. When delivering a message, the value of $return_path is set to $sender_
6342 address at the start of routing (routers may change the value). By an
6343 oversight, this default was not being set up when an address was tested by
6344 -bt or -bv, which affected the outcome if any router or filter referred to
6347 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6348 log only when "message" or "log_message" are set. However, if one of the
6349 conditions was an address verification, or a call to a nested ACL, the
6350 messages generated by the underlying test were being passed through. This
6351 no longer happens. The underlying message is available in $acl_verify_
6352 message for both "message" and "log_message" expansions, so it can be
6353 passed through if needed.
6355 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6356 new expansion $bh_ to give the encoded byte string without charset
6357 translation. Translation happens only if iconv() is available; HAVE_ICONV
6358 indicates this at build time. HEADERS_CHARSET gives the charset to
6359 translate to; headers_charset can change it in the configuration, and
6360 "headers charset" can change it in an individual filter file.
6362 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6363 that creates RFC 2047 encoded "words" labels them as that charset instead
6364 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6365 expansion operator; (ii) when Exim creates a From: line for a local
6366 message; (iii) when a header line is rewritten to include a "phrase" part.
6368 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6369 buggy, causing it to skip the first lines of messages whose message ID
6370 ended in 'D'. This would not have bitten before Exim release 4.14, because
6371 message IDs were unlikely to end in 'D' before then. The effect was to have
6372 incorrect size information for certain domains.
6374 45. #include "config.h" was missing at the start of the crypt16.c module. This
6375 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6378 46. If there was a timeout during a "random" callout check, Exim treated it as
6379 a failure of the random address, and carried on sending RSET and the real
6380 address. If the delay was just some slowness somewhere, the response to the
6381 original RCPT would be taken as a response to RSET and so on, causing
6382 mayhem of various kinds.
6384 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6385 when I implemented it. It didn't allow for the fact that some option values
6386 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6387 the right test for positive values.
6389 48. Domain names in DNS records are case-independent. Exim always looks them up
6390 in lower case. Some resolvers return domain names in exactly the case they
6391 appear in the zone file, that is, they may contain uppercase letters. Not
6392 all resolvers do this - some return always lower case. Exim was treating a
6393 change of case by a resolver as a change of domain, similar to a widening
6394 of a domain abbreviation. This triggered its re-routing code and so it was
6395 trying to route what was effectively the same domain again. This normally
6396 caused routing to fail (because the router wouldn't handle the domain
6397 twice). Now Exim checks for this case specially, and just changes the
6398 casing of the domain that it ultimately uses when it transmits the message
6401 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6404 50. If a filter generated a file delivery with a non-absolute name (possible if
6405 no home directory exists for the router), the forbid_file option was not
6408 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6409 the existing equality matching.
6411 52. Exim was using ints instead of ino_t variables in some places where it was
6412 dealing with inode numbers.
6414 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6415 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6416 called TMPDIR, and if it finds it is different, it changes its value.
6418 54. The smtp_printf() function is now made available to local_scan() so
6419 additional output lines can be written before returning. There is also an
6420 smtp_fflush() function to enable the detection of a dropped connection.
6421 The variables smtp_input and smtp_batched_input are exported to
6424 55. Changed the default runtime configuration: the message "Unknown user"
6425 has been removed from the ACL, and instead placed on the localuser router,
6426 using the cannot_route_message feature. This means that any verification
6427 failures that generate their own messages won't get overridden. Similarly,
6428 the "Unrouteable address" message that was in the ACL for unverifiable
6429 relay addresses has also been removed.
6431 56. Added hosts_avoid_esmtp to the smtp transport.
6433 57. The exicyclog script was not checking for the esoteric option
6434 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6435 will work only if exicyclog is run under the appropriate euid.
6437 58. Following a discussion on the list, the rules by which Exim recognises line
6438 endings on incoming messages have been changed. The -dropcr and drop_cr
6439 options are now no-ops, retained only for backwards compatibility. The
6440 following line terminators are recognized: LF CRLF CR. However, special
6441 processing applies to CR:
6443 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6444 nor a local message in the state where . is a terminator.
6446 (ii) If a bare CR is encountered in a header line, an extra space is added
6447 after the line terminator so as not to end the header. The reasoning
6448 behind this is that bare CRs in header lines are most likely either
6449 to be mistakes, or people trying to play silly games.
6451 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6452 was being incorrectly given as 18 bytes larger than it should have been.
6453 This is a VOB (very old bug).
6455 60. This may never have affected anything current, but just in case it has:
6456 When the local host is found other than at the start of a list of hosts,
6457 the local host, those with the same MX, and any that follow, are discarded.
6458 When the list in question was part of a longer list of hosts, the following
6459 hosts (not currently being processed) were also being discarded. This no
6460 longer happens. I'm not sure if this situation could ever has previously
6463 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6466 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6467 header. This is something that is recommended in a new Internet Draft, and
6468 is something that is documented as being done by Sendmail. There are two
6469 possible values. For messages generated by the autoreply transport, Exim
6472 Auto-Submitted: auto-replied
6474 whereas for all other generated messages (e.g. bounces) it adds
6476 Auto-Submitted: auto-generated
6478 63. The "personal" condition in filters now includes a test for the
6479 Auto-Submitted: header. If it contains the string "auto-" the message it
6480 not considered personal.
6482 64. Added rcpt_include_affixes as a generic transport option.
6484 65. Added queue_only_override (default true).
6486 66. Added the syslog_duplication option.
6488 67. If what should have been the first header line of a message consisted of
6489 a space followed by a colon, Exim was mis-interpreting it as a header line.
6490 It isn't of course - it is syntactically invalid and should therefore be
6491 treated as the start of the message body. The misbehaviour could have
6492 caused a number of strange effects, including loss of data in subsequent
6493 header lines, and spool format errors.
6495 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6496 client host had authenticated. This control can now be exercised by an ACL
6497 for more flexibility.
6499 69. By default, callouts do not happen when testing with -bh. There is now a
6500 variant, -bhc, which does actually run the callout code, including
6501 consulting and updating the callout cache.
6503 70. Added support for saslauthd authentication, courtesy of Alexander
6506 71. If statvfs() failed on the spool or log directories while checking their
6507 size for availability, Exim confusingly gave the error "space shortage".
6508 Furthermore, in debugging mode it crashed with a floating point exception.
6509 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6510 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6511 serious problem, Exim now writes to the main and panic logs when this
6512 happens, with details of the failure. It then refuses to accept the
6513 incoming message, giving the message "spool directory problem" or "log
6514 directory problem" with a 421 code for SMTP messages.
6516 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6517 0, 1, and 2 exist, because some OS complain for execs without them (see
6518 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6519 descriptors. However, the code omitted to check that the open succeeded,
6520 causing mysterious errors if for some reason the permissions on /dev/null
6521 got screwed. Now Exim writes a message to the main and panic logs, and
6522 bombs out if it can't open /dev/null.
6524 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6525 interact so that it is all more flexible. It is supposed to remain
6526 backwards compatible. Also added extra_local_interfaces.
6528 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6529 to bomb out with an assertion failure - to the client this appears as a
6530 connection drop. This problem occurs in the part of the code that was taken
6531 from the Samba project. Fortunately, the assertion is in a very simple
6532 function, so I have fixed this by reproducing the function inline in the
6533 one place where it is called, and arranging for authentication to fail
6534 instead of killing the process with assert().
6536 75. The SPA client code was not working when the server requested OEM rather
6537 than Unicode encoding.
6539 76. Added code to make require_files with a specific uid setting more usable in
6540 the case where statting the file as root fails - usually a non-root-mounted
6541 NFS file system. When this happens and the failure is EACCES, Exim now
6542 forks a subprocess and does the per-uid checking as the relevant uid.
6544 77. Added process_log_path.
6546 78. If log_file_path was not explicitly set, a setting of check_log_space or
6547 check_log_inodes was ignored.
6549 79. If a space check for the spool or log partitions fails, the incident is now
6550 logged. Of course, in the latter case the data may get lost...
6552 80. Added the %p formatting code to string_format() so that it can be used to
6553 print addresses in debug_print(). Adjusted all the address printing in the
6554 debugging in store.c to use %p rather than %d.
6556 81. There was a concern that a line of code in smtp_in.c could overflow a
6557 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6558 initially expressed, the concern was not well-founded, because trailing
6559 spaces are removed early. However, if the trailing spaces were followed by
6560 a NULL, they did not get removed, so the overflow was possible. Two fixes
6563 (a) I re-wrote the offending code in a cleaner fashion.
6564 (b) If an incoming SMTP command contains a NULL character, it is rejected
6567 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6568 runs initgroups(), so that if the Exim user is in any additional groups,
6569 they will be used during message reception.
6575 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob