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.
74 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
76 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
77 after the data-go-ahead and data-ack. Patch from Jason Betts.
79 JH/21 Bug 1846: Send DMARC forensic reports for reject and quaratine results,
80 even for a "none" policy. Patch from Tony Meyer.
82 JH/22 Fix continued use of a connection for further deliveries. If a port was
83 specified by a router, it must also match for the delivery to be
86 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
87 When one of the recipients of a message was unsuitable for the connection
88 (has no matching addresses), we lost track of needing to mark it
89 deferred. As a result mail would be lost.
91 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
93 JH/25 Decoding ACL controls is now done using a binary search; the sourcecode
94 takes up less space and should be simpler to maintain.
99 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
100 and 3.4.4 - once the server is enabled to respond to an OCSP request
101 it does even when not requested, resulting in a stapling non-aware
102 client dropping the TLS connection.
104 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
105 support variable-length bit vectors. No functional change.
107 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
108 The I= interface field on outgoing lines is now after the H= remote
109 host field, same as incoming lines. There is a separate
110 outgoing_interface log selector which allows you to disable the
113 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
114 If not running log_selector +smtp_connection the mainlog would be held
115 open indefinitely after a "too many connections" event, including to a
116 deleted file after a log rotate. Leave the per net connection logging
117 leaving it open for efficiency as that will be quickly detected by the
118 check on the next write.
120 HS/01 Bug 1671: Fix post transport crash.
121 Processing the wait-<transport> messages could crash the delivery
122 process if the message IDs didn't exist for some reason. When
123 using 'split_spool_directory=yes' the construction of the spool
124 file name failed already, exposing the same netto behaviour.
126 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
127 mime_regex ACL conditions.
129 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
130 to DSN fail messages (bounces): remote IP, remote greeting, remote response
131 to HELO, local diagnostic string.
133 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
134 log line to debug. Even when configured with a tls authenticator many
135 client connections are expected to not authenticate in this way, so
136 an authenticate fail is not an error.
138 HS/02 Add the Exim version string to the process info. This way exiwhat
139 gives some more detail about the running daemon.
141 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
142 matter for fast-change records such as DNSBLs.
144 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
145 retry record, even if constant. There may be multiple transports with
146 different interface settings and the retry behaviour needs to be kept
149 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
151 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
153 JH/10 Bug 840: fix log_defer_output option of pipe transport
155 JH/11 Bug 830: use same host for all RCPTS of a message, even under
156 hosts_randomize. This matters a lot when combined with mua_wrapper.
158 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
159 ${quote_pgsql:<string>} operator.
161 JH/13 Bug 1708: avoid misaligned access in cached lookup.
163 JH/14 Change header file name for freeradius-client. Relevant if compiling
164 with Radius support; from the Gentoo tree and checked under Fedora.
166 JH/15 Bug 1712: Introduce $prdr_requested flag variable
168 JH/16 Bug 1714: Permit an empty string as expansion result for transport
169 option transport_filter, meaning no filtering.
171 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
173 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
174 defaults to "*" (all hosts). The variable is now available when not built
175 with TLS, default unset, mainly to enable keeping the testuite sane.
176 If a server certificate is not supplied (via tls_certificate) an error is
177 logged, and clients will find TLS connections fail on startup. Presumably
178 they will retry in-clear.
179 Packagers of Exim are strongly encouraged to create a server certificate
180 at installation time.
182 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
183 with the $config_file variable.
185 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
186 in transport context, after the attempt, and per-recipient. The latter type
187 is per host attempted. The event data is the error message, and the errno
188 information encodes the lookup type (A vs. MX) used for the (first) host,
189 and the trailing two digits of the smtp 4xx reponse.
191 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
192 to write to mainlog (or rejectlog, paniclog) in the window between file
193 creation and permissions/ownership being changed. Particularly affects
194 installations where exicyclog is run as root, rather than exim user;
195 result is that the running daemon panics and dies.
197 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
199 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
200 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
201 "pri" and "weight". Note that the previous implicit priority given by the
202 list order is no longer honoured.
204 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
207 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
208 by defining SUPPORT_SOCKS.
210 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
211 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
212 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
213 variables are renamed to proxy_{local,external}_{address,port}.
215 JH/27 Move Internationalisation support from Experimental to mainline, enabled
216 for a build by defining SUPPORT_I18N
218 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
219 of the query string, and make ${quote_redis:} do that quoting.
221 JH/29 Move Events support from Experimental to mainline, enabled by default
222 and removable for a build by defining DISABLE_EVENT.
224 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
226 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
227 cached by the daemon.
229 JH/32 Move Redis support from Experimental to mainline, enabled for a build
230 by defining LOOKUP_REDIS. The libhiredis library is required.
232 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
233 keys are given for lookup.
235 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
236 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
237 only supported when built with TLS support. The PolarSSL SHA routines
238 are still used when the TLS library is too old for convenient support.
240 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
241 openssl_options), for security. OpenSSL forces this from version 1.1.0
242 server-side so match that on older versions.
244 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
245 allocation for $value could be released as the expansion processing
246 concluded, but leaving the global pointer active for it.
248 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
249 and to use the domains and local_parts ACL conditions.
251 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
252 incorrectly not doubled on cutthrough transmission, hence seen as a
253 body-termination at the receiving system - resulting in truncated mails.
254 Commonly the sender saw a TCP-level error, and retransmitted the message
255 via the normal store-and-forward channel. This could result in duplicates
256 received - but deduplicating mailstores were liable to retain only the
257 initial truncated version.
259 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
261 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
263 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
264 we're in there, support oversigning also; bug 1309.
266 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
268 HS/04 Add support for keep_environment and add_environment options.
270 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
271 either intentional arithmetic overflow during PRNG, or testing config-
274 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
275 delivery resulted in actual delivery. Cancel cutthrough before DATA
278 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
279 rejects a recipient: pass the reject to the originator.
281 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
282 Many were false-positives and ignorable, but it's worth fixing the
285 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
286 for the new environment-manipulation done at startup. Move the routines
287 from being local to tls.c to being global via the os.c file.
289 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
290 an extract embedded as result-arg for a map, the first arg for extract
291 is unavailable so we cannot tell if this is a numbered or keyed
292 extraction. Accept either.
298 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
301 JH/02 The smtp transport option "multi_domain" is now expanded.
303 JH/03 The smtp transport now requests PRDR by default, if the server offers
306 JH/04 Certificate name checking on server certificates, when exim is a client,
307 is now done by default. The transport option tls_verify_cert_hostnames
308 can be used to disable this per-host. The build option
309 EXPERIMENTAL_CERTNAMES is withdrawn.
311 JH/05 The value of the tls_verify_certificates smtp transport and main options
312 default to the word "system" to access the system default CA bundle.
313 For GnuTLS, only version 3.0.20 or later.
315 JH/06 Verification of the server certificate for a TLS connection is now tried
316 (but not required) by default. The verification status is now logged by
317 default, for both outbound TLS and client-certificate supplying inbound
320 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
323 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
324 Status Notification (bounce) messages are now MIME format per RFC 3464.
325 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
326 under the control of the dsn_advertise_hosts option, and routers may
327 have a dsn_lasthop option.
329 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
330 default, modifiable by a malware= option. The list separator for
331 the options can now be changed in the usual way. Bug 68.
333 JH/10 The smtp_receive_timeout main option is now expanded before use.
335 JH/11 The incoming_interface log option now also enables logging of the
336 local interface on delivery outgoing connections.
338 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
339 if the interface and destination host and port all match.
341 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
344 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
345 Patch from Andrew Lewis.
347 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
348 now supports optional time-restrictions, weighting, and priority
349 modifiers per server. Patch originally by <rommer@active.by>.
351 JH/16 The spamd_address main option now supports a mixed list of local
352 and remote servers. Remote servers can be IPv6 addresses, and
353 specify a port-range.
355 JH/17 Bug 68: The spamd_address main option now supports an optional
356 timeout value per server.
358 JH/18 Bug 1581: Router and transport options headers_add/remove can
359 now have the list separator specified.
361 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
364 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
367 JH/21 Support for the A6 type of dns record is withdrawn.
369 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
370 rather than the verbs used.
372 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
373 from 255 to 1024 chars.
375 JH/24 Verification callouts now attempt to use TLS by default.
377 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
378 are generic router options now. The defaults didn't change.
380 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
381 Original patch from Alexander Shikoff, worked over by JH.
383 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
384 files if ZCAT_COMMAND is not executable.
386 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
388 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
390 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
391 Normally benign, it bites when the pair was led to by a CNAME;
392 modern usage is to not canoicalize the domain to a CNAME target
393 (and we were inconsistent anyway for A-only vs AAAA+A).
395 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
397 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
398 when evaluating $sender_host_dnssec.
400 JH/31 Check the HELO verification lookup for DNSSEC, adding new
401 $sender_helo_dnssec variable.
403 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
405 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
407 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
409 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
410 documented as working, but never had. Support all but $spam_report.
412 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
413 added for tls authenticator.
415 HS/03 Add perl_taintmode main config option
420 TL/01 When running the test suite, the README says that variables such as
421 no_msglog_check are global and can be placed anywhere in a specific
422 test's script, however it was observed that placement needed to be near
423 the beginning for it to behave that way. Changed the runtest perl
424 script to read through the entire script once to detect and set these
425 variables, reset to the beginning of the script, and then run through
426 the script parsing/test process like normal.
428 TL/02 The BSD's have an arc4random API. One of the functions to induce
429 adding randomness was arc4random_stir(), but it has been removed in
430 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
431 function when detected.
433 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
434 cause callback expansion.
436 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
437 syntax errors in an expansion can be treated as a string instead of
438 logging or causing an error, due to the internal use of bool_lax
439 instead of bool when processing it.
441 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
442 server certificates when making smtp deliveries.
444 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
446 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
448 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
450 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
451 Merged patch from Sebastian Wiedenroth.
453 JH/05 Fix results-pipe from transport process. Several recipients, combined
454 with certificate use, exposed issues where response data items split
455 over buffer boundaries were not parsed properly. This eventually
456 resulted in duplicates being sent. This issue only became common enough
457 to notice due to the introduction of conection certificate information,
458 the item size being so much larger. Found and fixed by Wolfgang Breyha.
460 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
461 size buffer was used, resulting in syntax errors when an expansion
464 JH/07 Add support for directories of certificates when compiled with a GnuTLS
465 version 3.3.6 or later.
467 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
468 is EXPERIMENTAL_EVENT, the main-configuration and transport options
469 both become "event_action", the variables become $event_name, $event_data
470 and $event_defer_errno. There is a new variable $verify_mode, usable in
471 routers, transports and related events. The tls:cert event is now also
472 raised for inbound connections, if the main configuration event_action
475 TL/06 In test suite, disable OCSP for old versions of openssl which contained
476 early OCSP support, but no stapling (appears to be less than 1.0.0).
478 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
479 server certificate names available under the smtp transport option
480 "tls_verify_cert_hostname" now do not permit multi-component wildcard
483 JH/10 Time-related extraction expansions from certificates now use the main
484 option "timezone" setting for output formatting, and are consistent
485 between OpenSSL and GnuTLS compilations. Bug 1541.
487 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
488 encoded parameter in the incoming message. Bug 1558.
490 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
491 include certificate info, eximon was claiming there were spoolfile
494 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
496 JH/14 Log delivery-related information more consistently, using the sequence
497 "H=<name> [<ip>]" wherever possible.
499 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
500 are problematic for Debian distribution, omit them from the release
503 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
505 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
508 JH/17 Fix a null-indirection in certextract expansions when a nondefault
509 output list separator was used.
514 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
515 checkers that were complaining about end of non-void function with no
518 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
519 This was a regression intruduced in 4.83 by another bugfix.
521 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
523 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
524 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
530 TF/01 Correctly close the server side of TLS when forking for delivery.
532 When a message was received over SMTP with TLS, Exim failed to clear up
533 the incoming connection properly after forking off the child process to
534 deliver the message. In some situations the subsequent outgoing
535 delivery connection happened to have the same fd number as the incoming
536 connection previously had. Exim would try to use TLS and fail, logging
537 a "Bad file descriptor" error.
539 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
540 utilities have not been installed.
542 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
543 temporary space as the ACL may create new global variables.
545 TL/01 LDAP support uses per connection or global context settings, depending
546 upon the detected version of the libraries at build time.
548 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
549 to extract and use the src ip:port in logging and expansions as if it
550 were a direct connection from the outside internet. PPv2 support was
551 updated based on HAProxy spec change in May 2014.
553 JH/02 Add ${listextract {number}{list}{success}{fail}}.
555 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
556 Properly escape header and check for NULL return.
558 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
561 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
563 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
564 characters in header names, implemented as a verify condition.
565 Contributed by Michael Fischer v. Mollard.
567 TL/05 Rename SPF condition results err_perm and err_temp to standardized
568 results permerror and temperror. Previous values are deprecated but
569 still accepted. In a future release, err_perm and err_temp will be
570 completely removed, which will be a backward incompatibility if the
571 ACL tests for either of these two old results. Patch contributed by
572 user bes-internal on the mailing list.
574 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
576 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
577 selectors, in both main and reject logs.
579 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
582 JH/07 Add malware type "sock" for talking to simple daemon.
584 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
586 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
587 routers/transports under cutthrough routing.
589 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
590 numbers. Touch up "bool" conditional to keep the same definition.
592 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
594 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
596 JH/12 Expand items in router/transport headers_add or headers_remove lists
597 individually rather than the list as a whole. Bug 1452.
599 Required for reasonable handling of multiple headers_ options when
600 they may be empty; requires that headers_remove items with embedded
601 colons must have them doubled (or the list-separator changed).
603 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
604 view the policy declared in the DMARC record. Currently, $dmarc_status
605 is a combined value of both the record presence and the result of the
608 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
610 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
611 dnslookup router and the smtp transport (applying to the forward
614 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
615 of ldap servers used for a specific lookup. Patch provided by Heiko
618 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
619 New variable $lookup_dnssec_authenticated for observability.
621 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
622 Patch submitted by Lars Timman.
624 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
626 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
627 Requires trusted mode and valid format message id, aborts otherwise.
628 Patch contributed by Heiko Schlichting.
630 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
631 certextract with support for various fields. Bug 1358.
633 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
634 is requested by default, modifiable by smtp transport option
637 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
638 operate on certificate variables to give certificate fingerprints
639 Also new ${sha256:cert_variable}.
641 JH/23 The PRDR feature is moved from being Experimental into the mainline.
643 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
644 Christian Aistleitner.
646 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
648 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
649 file. Patch from Wolfgang Breyha.
651 JH/25 Expand the coverage of the delivery $host and $host_address to
652 client authenticators run in verify callout. Bug 1476.
654 JH/26 Port service names are now accepted for tls_on_connect_ports, to
655 align with daemon_smtp_ports. Bug 72.
657 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
658 support and error reporting did not work properly.
660 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
661 and is readable. Patch from Andrew Colin Kissa.
663 TL/14 Enhance documentation of ${run expansion and how it parses the
664 commandline after expansion, particularly in the case when an
665 unquoted variable expansion results in an empty value.
667 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
669 PP/02 Fix internal collision of T_APL on systems which support RFC3123
670 by renaming away from it. Addresses GH issue 15, reported by
673 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
675 TL/15 SECURITY: prevent double expansion in math comparison functions
676 (can expand unsanitized data). Not remotely exploitable.
683 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
685 PP/02 Make -n do something, by making it not do something.
686 When combined with -bP, the name of an option is not output.
688 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
691 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
692 $sender_host_name and config options to manage this, and basic check
695 PP/05 DSCP support for outbound connections and control modifier for inbound.
697 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
698 (Only plugin which currently uses this is kerberos4, which nobody should
699 be using, but we should make it available and other future plugins might
700 conceivably use it, even though it would break NAT; stuff *should* be
701 using channel bindings instead).
703 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
704 name; added for Sendmail compatibility; requires admin caller.
705 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
706 just ignore it); requires trusted caller.
707 Also parse but ignore: -Ac -Am -X<logfile>
710 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
712 TL/02 Add +smtp_confirmation as a default logging option.
714 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
715 Patch by Magnus Holmgren from 2007-02-20.
717 TL/04 Bugzilla 1281 - Spec typo.
718 Bugzilla 1283 - Spec typo.
719 Bugzilla 1290 - Spec grammar fixes.
721 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
723 TL/06 Add Experimental DMARC support using libopendmarc libraries.
725 TL/07 Fix an out of order global option causing a segfault. Reported to dev
726 mailing list by by Dmitry Isaikin.
728 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
730 JH/02 Support "G" suffix to numbers in ${if comparisons.
732 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
734 NM/01 Bugzilla 1197 - Spec typo
735 Bugzilla 1196 - Spec examples corrections
737 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
739 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
740 gnutls_enable_pkcs11, but renamed to more accurately indicate its
743 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
744 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
746 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
747 "acl {{name}{arg}...}", and optional args on acl condition
750 JH/05 Permit multiple router/transport headers_add/remove lines.
752 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
754 JH/07 Avoid using a waiting database for a single-message-only transport.
755 Performance patch from Paul Fisher. Bugzilla 1262.
757 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
760 JH/09 Add $headers_added variable, with content from use of ACL modifier
761 add_header (but not yet added to the message). Bugzilla 199.
763 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
764 Pulled from Bugzilla 817 by Wolfgang Breyha.
766 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
768 (nb: this is the same fix as in Exim 4.80.1)
770 JH/11 Add A= logging on delivery lines, and a client_set_id option on
773 JH/12 Add optional authenticated_sender logging to A= and a log_selector
776 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
778 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
779 advertise SMTP AUTH mechanism to us, instead of a generic
780 protocol violation error. Also, make Exim more robust to bad
781 data from the Dovecot auth socket.
783 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
785 When a queue runner is handling a message, Exim first routes the
786 recipient addresses, during which it prunes them based on the retry
787 hints database. After that it attempts to deliver the message to
788 any remaining recipients. It then updates the hints database using
791 So if a recipient address works intermittently, it can get repeatedly
792 deferred at routing time. The retry hints record remains fresh so the
793 address never reaches the final cutoff time.
795 This is a fairly common occurrence when a user is bumping up against
796 their storage quota. Exim had some logic in its local delivery code
797 to deal with this. However it did not apply to per-recipient defers
798 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
800 This change adds a proper retry rule check during routing so that the
801 final cutoff time is checked against the message's age. We only do
802 this check if there is an address retry record and there is not a
803 domain retry record; this implies that previous attempts to handle
804 the address had the retry_use_local_parts option turned on. We use
805 this as an approximation for the destination being like a local
806 delivery, as in LMTP.
808 I suspect this new check makes the old local delivery cutoff check
809 redundant, but I have not verified this so I left the code in place.
811 TF/02 Correct gecos expansion when From: is a prefix of the username.
813 Test 0254 submits a message to Exim with the header
817 When I ran the test suite under the user fanf2, Exim expanded
818 the header to contain my full name, whereas it should have added
819 a Resent-Sender: header. It erroneously treats any prefix of the
820 username as equal to the username.
822 This change corrects that bug.
824 GF/01 DCC debug and logging tidyup
825 Error conditions log to paniclog rather than rejectlog.
826 Debug lines prefixed by "DCC: " to remove any ambiguity.
828 TF/03 Avoid unnecessary rebuilds of lookup-related code.
830 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
831 Bug spotted by Jeremy Harris; was flawed since initial commit.
832 Would have resulted in OCSP responses post-SNI triggering an Exim
833 NULL dereference and crash.
835 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
837 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
838 Bug detection, analysis and fix by Samuel Thibault.
839 Bugzilla 1331, Debian bug #698092.
841 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
843 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
844 Server implementation by Todd Lyons, client by JH.
845 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
846 config variable "prdr_enable" controls whether the server
847 advertises the facility. If the client requests PRDR a new
848 acl_data_smtp_prdr ACL is called once for each recipient, after
849 the body content is received and before the acl_smtp_data ACL.
850 The client is controlled by bolth of: a hosts_try_prdr option
851 on the smtp transport, and the server advertisement.
852 Default client logging of deliveries and rejections involving
853 PRDR are flagged with the string "PRDR".
855 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
856 fclose(). Diagnosis by Todd Lyons.
858 PP/17 Update configure.default to handle IPv6 localhost better.
859 Patch by Alain Williams (plus minor tweaks).
862 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
863 This is now consistent with GnuTLS, and is now documented: the
864 previous undocumented portable approach to treating the option as
865 unset was to force an expansion failure. That still works, and
866 an empty string is now equivalent.
868 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
869 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
870 not performing validation itself.
872 PP/20 Added force_command boolean option to pipe transport.
873 Patch from Nick Koston, of cPanel Inc.
875 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
878 TF/04 Added udpsend ACL modifer and hexquote expansion operator
880 PP/21 Fix eximon continuous updating with timestamped log-files.
881 Broken in a format-string cleanup in 4.80, missed when I repaired the
882 other false fix of the same issue.
883 Report and fix from Heiko Schlichting.
886 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
887 Report from Prashanth Katuri.
889 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
890 It's SecureTransport, so affects any MacOS clients which use the
891 system-integrated TLS libraries, including email clients.
893 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
894 using a MIME ACL for non-SMTP local injection.
895 Report and assistance in diagnosis by Warren Baker.
897 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
899 JH/16 Fix comparisons for 64b. Bugzilla 1385.
901 TL/09 Add expansion variable $authenticated_fail_id to keep track of
902 last id that failed so it may be referenced in subsequent ACL's.
904 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
907 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
908 ldap library initialization, allowing self-signed CA's to be
909 used. Also properly sets require_cert option later in code by
910 using NULL (global ldap config) instead of ldap handle (per
911 session). Bug diagnosis and testing by alxgomz.
913 TL/12 Enhanced documentation in the ratelimit.pl script provided in
914 the src/util/ subdirectory.
916 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
917 renamed to Transport Post Delivery Action by Jeremy Harris, as
920 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
921 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
922 redis_servers = needs to be configured which will be used by the redis
923 lookup. Patch from Warren Baker, of The Packet Hub.
925 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
927 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
928 hostname or reverse DNS when processing a host list. Used suggestions
929 from multiple comments on this bug.
931 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
933 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
934 Missed a few lines, added it to make the runtest require no keyboard
937 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
938 contains upper case chars. Make router use caseful_local_part.
940 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
941 support when GnuTLS has been built with p11-kit.
947 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
949 This, or similar/improved, will also be change PP/11 of 4.82.
955 PP/01 Handle short writes when writing local log-files.
956 In practice, only affects FreeBSD (8 onwards).
957 Bugzilla 1053, with thanks to Dmitry Isaikin.
959 NM/01 Bugzilla 949 - Documentation tweak
961 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
964 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
966 PP/02 Implemented gsasl authenticator.
968 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
970 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
971 `pkg-config foo` for cflags/libs.
973 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
974 with rest of GSASL and with heimdal_gssapi.
976 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
977 `pkg-config foo` for cflags/libs for the TLS implementation.
979 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
980 properties get this fed in as external SSF. A number of robustness
981 and debugging improvements to the cyrus_sasl authenticator.
983 PP/08 cyrus_sasl server now expands the server_realm option.
985 PP/09 Bugzilla 1214 - Log authentication information in reject log.
986 Patch by Jeremy Harris.
988 PP/10 Added dbmjz lookup type.
990 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
992 PP/12 MAIL args handles TAB as well as SP, for better interop with
993 non-compliant senders.
994 Analysis and variant patch by Todd Lyons.
996 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
997 Bug report from Lars Müller <lars@samba.org> (via SUSE),
998 Patch from Dirk Mueller <dmueller@suse.com>
1000 PP/13 tls_peerdn now print-escaped for spool files.
1001 Observed some $tls_peerdn in wild which contained \n, which resulted
1002 in spool file corruption.
1004 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1005 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1006 or write after TLS renegotiation, which otherwise led to messages
1009 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1010 as a tracking header (ie: a signed header comes before the signature).
1011 Patch from Wolfgang Breyha.
1013 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1014 comma-sep list; embedded commas doubled.
1016 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1018 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1020 Report and patch from Dmitry Banschikov.
1022 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1023 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1024 failures. We appear to now support TLS1.1+ with Exim.
1026 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1027 lets Exim select keys and certificates based upon TLS SNI from client.
1028 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1029 before an outbound SMTP session. New log_selector, +tls_sni.
1031 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1032 NULL dereference. Report and patch from Alun Jones.
1034 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1035 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1036 Not seeing resolver debug output on NetBSD, but suspect this is a
1037 resolver implementation change.
1039 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1040 Left warnings. Added "eximon gdb" invocation mode.
1042 PP/21 Defaulting "accept_8bitmime" to true, not false.
1044 PP/22 Added -bw for inetd wait mode support.
1046 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1047 locate the relevant includes and libraries. Made this the default.
1049 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1050 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1052 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1053 This may cause build issues on older platforms.
1055 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1056 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1057 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1058 Added SNI support via GnuTLS too.
1059 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1061 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1063 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1064 Applied second patch from Janne, implementing suggestion to default
1065 multiple-strings-in-record handling to match SPF spec.
1067 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1069 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1070 read-only, out of scope).
1071 Patch from Wolfgang Breyha, report from Stuart Northfield.
1073 PP/29 Fix three issues highlighted by clang analyser static analysis.
1074 Only crash-plausible issue would require the Cambridge-specific
1075 iplookup router and a misconfiguration.
1076 Report from Marcin Mirosław.
1078 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1080 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1081 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1082 As part of this, removing so much warning spew let me fix some minor
1083 real issues in debug logging.
1085 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1086 assignment on my part. Fixed.
1088 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1089 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1090 Janne Snabb (who went above and beyond: thank you).
1092 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1093 string otherwise requires a connection and a bunch more work and it's
1094 relatively easy to get wrong. Should also expose TLS library linkage
1097 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1098 64-bit ${eval} (JH/03).
1100 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1101 GNU libc to support some of the 64-bit stuff, should not lead to
1102 conflicts. Defined before os.h is pulled in, so if a given platform
1103 needs to override this, it can.
1105 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1106 protection layer was required, which is not implemented.
1107 Bugzilla 1254, patch from Wolfgang Breyha.
1109 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1110 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1111 tls_dhparam take prime identifiers. Also unbreak combination of
1112 OpenSSL+DH_params+TLSSNI.
1114 PP/39 Disable SSLv2 by default in OpenSSL support.
1120 PP/01 Solaris build fix for Oracle's LDAP libraries.
1121 Bugzilla 1109, patch from Stephen Usher.
1123 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1125 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1128 TF/02 Fix a couple more cases where we did not log the error message
1129 when unlink() failed. See also change 4.74-TF/03.
1131 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1132 lock up or crash if it happened to be inside a call to libc when it
1133 got a SIGUSR1 from exiwhat.
1135 The SIGUSR1 handler appends the current process status to the process
1136 log which is later printed by exiwhat. It used to use the general
1137 purpose logging code to do this, but several functions it calls are
1138 not safe for signals.
1140 The new output code in the SIGUSR1 handler is specific to the process
1141 log, and simple enough that it's easy to inspect for signal safety.
1142 Removing some special cases also simplifies the general logging code.
1143 Removing the spurious timestamps from the process log simplifies
1146 TF/04 Improved ratelimit ACL condition.
1148 The /noupdate option has been deprecated in favour of /readonly which
1149 has clearer semantics. The /leaky, /strict, and /readonly update modes
1150 are mutually exclusive. The update mode is no longer included in the
1151 database key; it just determines when the database is updated. (This
1152 means that when you upgrde Exim will forget old rate measurements.)
1154 Exim now checks that the per_* options are used with an update mode that
1155 makes sense for the current ACL. For example, when Exim is processing a
1156 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1157 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1158 must specify per_mail/readonly. If you omit the update mode it defaults to
1159 /leaky where that makes sense (as before) or /readonly where required.
1161 The /noupdate option is now undocumented but still supported for
1162 backwards compatibility. It is equivalent to /readonly except that in
1163 ACLs where /readonly is required you may specify /leaky/noupdate or
1164 /strict/noupdate which are treated the same as /readonly.
1166 A useful new feature is the /count= option. This is a generalization
1167 of the per_byte option, so that you can measure the throughput of other
1168 aggregate values. For example, the per_byte option is now equivalent
1169 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1171 The per_rcpt option has been generalized using the /count= mechanism
1172 (though it's more complicated than the per_byte equivalence). When it is
1173 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1174 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1175 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1176 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1177 Note that using per_rcpt with a non-readonly update mode in more than
1178 one ACL will cause the recipients to be double-counted. (The per_mail
1179 and per_byte options don't have this problem.)
1181 The handling of very low rates has changed slightly. If the computed rate
1182 is less than the event's count (usually one) then this event is the first
1183 after a long gap. In this case the rate is set to the same as this event's
1184 count, so that the first message of a spam run is counted properly.
1186 The major new feature is a mechanism for counting the rate of unique
1187 events. The new per_addr option counts the number of different
1188 recipients that someone has sent messages to in the last time period. It
1189 behaves like per_rcpt if all the recipient addresses are different, but
1190 duplicate recipient addresses do not increase the measured rate. Like
1191 the /count= option this is a general mechanism, so the per_addr option
1192 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1193 example, measure the rate that a client uses different sender addresses
1194 with the options per_mail/unique=$sender_address. There are further
1195 details in the main documentation.
1197 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1199 TF/06 Removed a few PCRE remnants.
1201 TF/07 Automatically extract Exim's version number from tags in the git
1202 repository when doing development or release builds.
1204 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1205 Bugzilla 879. Patch from Paul Fisher.
1207 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1208 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1211 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1213 PP/05 Variable $av_failed, true if the AV scanner deferred.
1214 Bugzilla 1078. Patch from John Horne.
1216 PP/06 Stop make process more reliably on build failure.
1217 Bugzilla 1087. Patch from Heiko Schlittermann.
1219 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1220 Bugzilla 1089. Patch from Heiko Schlittermann.
1222 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1223 Bugzilla 1131. Patch from Holger Weiß.
1225 PP/09 Handle IPv6 addresses with SPF.
1226 Bugzilla 860. Patch from Wolfgang Breyha.
1228 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1230 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1233 PP/11 match_* no longer expand right-hand-side by default.
1234 New compile-time build option, EXPAND_LISTMATCH_RHS.
1235 New expansion conditions, "inlist", "inlisti".
1237 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1239 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1241 PP/14 fix log_write() format string regression from TF/03.
1242 Bugzilla 1152. Patch from Dmitry Isaikin.
1248 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1250 PP/02 Harmonised TLS library version reporting; only show if debugging.
1251 Layout now matches that introduced for other libraries in 4.74 PP/03.
1253 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1255 PP/04 New "dns_use_edns0" global option.
1257 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1260 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1261 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1263 TK/01 Updated PolarSSL code to 0.14.2.
1264 Bugzilla 1097. Patch from Andreas Metzler.
1266 PP/07 Catch divide-by-zero in ${eval:...}.
1267 Fixes bugzilla 1102.
1269 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1272 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1273 format-string attack -- SECURITY: remote arbitrary code execution.
1275 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1276 time unintentionally subject to list matching rules, letting the header
1277 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1278 arbitrary string expansion). This allowed for information disclosure.
1280 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1281 INT_MIN/-1 -- value coerced to INT_MAX.
1287 NM/01 Workround for PCRE version dependency in version reporting
1290 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1291 This fixes portability to compilers other than gcc, notably
1292 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1294 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1295 makefiles for portability to HP-UX and POSIX correctness.
1297 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1298 Also via indented variable definition in the Makefile.
1299 (Debugging by Oliver Heesakkers).
1301 PP/02 Restore caching of spamd results with expanded spamd_address.
1302 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1304 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1305 Improves build reliability. Fix from: Frank Elsner
1307 NM/02 Fix wide character breakage in the rfc2047 coding
1308 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1310 NM/03 Allow underscore in dnslist lookups
1311 Fixes bug 1026. Patch from Graeme Fowler
1313 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1314 Code patches from Adam Ciarcinski of NetBSD.
1316 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1319 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1320 is logged, to avoid truncation. Patch from John Horne.
1322 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1323 Patch from Jakob Hirsch.
1325 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1326 SQL string expansion failure details.
1327 Patch from Andrey Oktyabrski.
1329 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1330 Patch from Simon Arlott.
1332 PP/09 New lookups functionality failed to compile on old gcc which rejects
1333 extern declarations in function scope.
1334 Patch from Oliver Fleischmann
1336 PP/10 Use sig_atomic_t for flags set from signal handlers.
1337 Check getgroups() return and improve debugging.
1338 Fixed developed for diagnosis in bug 927 (which turned out to be
1341 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1342 Patch from Mark Zealey.
1344 PP/12 Bugzilla 1056: Improved spamd server selection.
1345 Patch from Mark Zealey.
1347 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1348 Based on patch from Heiko Schlittermann.
1350 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1351 Patch from Uwe Doering, sign-off by Michael Haardt.
1353 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1354 variable declaration deep within a block. Bug and patch from
1357 PP/15 lookups-Makefile IRIX compatibilty coercion.
1359 PP/16 Make DISABLE_DKIM build knob functional.
1361 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1362 Patch by Simon Arlott
1364 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1365 variable argument macros. Our copy now differs from upstream.
1371 TF/01 Failure to get a lock on a hints database can have serious
1372 consequences so log it to the panic log.
1374 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1375 controlled using the smtp_confirmation log selector.
1377 TF/03 Include the error message when we fail to unlink a spool file.
1379 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1380 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1381 for maintaining out-of-tree patches for some time.
1383 PP/01 Bugzilla 139: Documentation and portability issues.
1384 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1385 Handle per-OS dynamic-module compilation flags.
1387 PP/02 Let /dev/null have normal permissions.
1388 The 4.73 fixes were a little too stringent and complained about the
1389 permissions on /dev/null. Exempt it from some checks.
1390 Reported by Andreas M. Kirchwitz.
1392 PP/03 Report version information for many libraries, including
1393 Exim version information for dynamically loaded libraries. Created
1394 version.h, now support a version extension string for distributors
1395 who patch heavily. Dynamic module ABI change.
1397 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1398 privilege escalation vulnerability whereby the Exim run-time user
1399 can cause root to append content of the attacker's choosing to
1402 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1405 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1406 If dropping privileges for untrusted macros, we disabled normal logging
1407 on the basis that it would fail; for the Exim run-time user, this is not
1408 the case, and it resulted in successful deliveries going unlogged.
1409 Fixed. Reported by Andreas Metzler.
1415 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1416 only prepend for the Resent-* case. Fixes regression introduced in
1417 Exim 4.70 by NM/22 for Bugzilla 607.
1419 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1420 increasing numbers of administrators be bitten by this.
1422 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1424 PP/03 Bugzilla 994: added openssl_options main configuration option.
1426 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1428 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1430 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1432 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1433 without a peer certificate, leading to a segfault because of an
1434 assumption that peers always have certificates. Be a little more
1435 paranoid. Problem reported by Martin Tscholak.
1437 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1438 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1439 NB: ClamAV planning to remove STREAM in "middle of 2010".
1440 CL also introduces -bmalware, various -d+acl logging additions and
1441 more caution in buffer sizes.
1443 PP/09 Implemented reverse_ip expansion operator.
1445 PP/10 Bugzilla 937: provide a "debug" ACL control.
1447 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1449 PP/12 Bugzilla 973: Implement --version.
1451 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1453 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1455 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1457 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1458 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1459 ignore trailing whitespace.
1461 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1463 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1466 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1467 Notification from Dr Andrew Aitchison.
1469 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1470 ExtendedDetectionInfo response format.
1471 Notification from John Horne.
1473 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1476 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1477 XSL and documented dependency on system catalogs, with examples of how
1480 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1483 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1484 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1485 configuration file which is writeable by the Exim user or group.
1487 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1488 of configuration files to cover files specified with the -C option if
1489 they are going to be used with root privileges, not just the default
1492 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1493 option (effectively making it always true).
1495 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1496 files to be used while preserving root privileges.
1498 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1499 that rogue child processes cannot use them.
1501 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1502 run-time user, instead of root.
1504 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1505 Exim run-time user without dropping privileges.
1507 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1508 result string, instead of calling string_vformat() twice with the same
1511 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1512 for other users. Others should always drop root privileges if they use
1513 -C on the command line, even for a whitelisted configure file.
1515 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1517 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1523 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1524 $data_path, and $header_path variables; fixed documentation bugs and
1527 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1528 exipick to access non-standard spools, including the "frozen" queue
1531 NM/01 Bugzilla 965: Support mysql stored procedures.
1532 Patch from Alain Williams
1534 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1536 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1537 Patch from Andreas Metzler
1539 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1540 Patch from Kirill Miazine
1542 NM/05 Bugzilla 671: Added umask to procmail example.
1544 JJ/03 installed exipick 20100323.0, fixing doc bug
1546 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1547 directory. Notification and patch from Dan Rosenberg.
1549 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1551 TK/02 Improve log output when DKIM signing operation fails.
1553 MH/01 Treat the transport option dkim_domain as a colon separated
1554 list, not as a single string, and sign the message with each element,
1555 omitting multiple occurences of the same signer.
1557 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1558 Bugzilla 985, 986. Patch by Simon Arlott
1560 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1561 Patch by Simon Arlott
1563 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1564 MBX locking. Notification from Dan Rosenberg.
1570 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1572 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1574 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1576 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1578 NM/04 Bugzilla 915: Flag broken perl installation during build.
1584 TK/01 Added patch by Johannes Berg that expands the main option
1585 "spamd_address" if it starts with a dollar sign.
1587 TK/02 Write list of recipients to X-Envelope-Sender header when building
1588 the mbox-format spool file for content scanning (suggested by Jakob
1591 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1592 (http://www.dcc-servers.net/) support via dccifd. Activated by
1593 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1595 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1596 by Mark Daniel Reidel <mr@df.eu>.
1598 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1599 When building exim an external PCRE library is now needed -
1600 PCRE is a system library on the majority of modern systems.
1601 See entry on PCRE_LIBS in EDITME file.
1603 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1604 conversation. Added nologin parameter to request.
1605 Patch contributed by Kirill Miazine.
1607 TF/01 Do not log submission mode rewrites if they do not change the address.
1609 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1611 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1612 log files in place. Contributed by Roberto Lima.
1614 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1616 TF/03 Bugzilla 615: When checking the local_parts router precondition
1617 after a local_part_suffix or local_part_prefix option, Exim now
1618 does not use the address's named list lookup cache, since this
1619 contains cached lookups for the whole local part.
1621 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1622 Robert Millan. Documentation is in experimental-spec.txt.
1624 TF/04 Bugzilla 668: Fix parallel build (make -j).
1626 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1628 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1629 Patch provided by Jan Srzednicki.
1631 TF/05 Leading white space used to be stripped from $spam_report which
1632 wrecked the formatting. Now it is preserved.
1634 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1635 that they are available at delivery time.
1637 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1639 TF/08 TLS error reporting now respects the incoming_interface and
1640 incoming_port log selectors.
1642 TF/09 Produce a more useful error message if an SMTP transport's hosts
1643 setting expands to an empty string.
1645 NM/06 Bugzilla 744: EXPN did not work under TLS.
1646 Patch provided by Phil Pennock.
1648 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1649 Patch provided by Richard Godbee.
1651 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1652 acl_smtp_notquit, added index entry.
1654 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1655 Patch provided by Eugene Bujak.
1657 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1658 accept(). Patch provided by Maxim Dounin.
1660 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1661 Patch provided by Phil Pennock.
1663 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1665 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1666 Patch provided by Brad "anomie" Jorsch.
1668 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1669 Patch provided by Dean Brooks.
1671 TK/05 Add native DKIM support (does not depend on external libraries).
1673 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1674 Patch provided by Graeme Fowler.
1676 NM/16 Bugzilla 851: Documentation example syntax fix.
1678 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1680 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1683 NM/19 Bugzilla 745: TLS version reporting.
1684 Patch provided by Phil Pennock.
1686 NM/20 Bugzilla 167: bool: condition support.
1687 Patch provided by Phil Pennock.
1689 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1690 clients. Patch provided by Phil Pennock.
1692 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1693 Patch provided by Brad "anomie" Jorsch.
1695 NM/23 Bugzilla 687: Fix misparses in eximstats.
1696 Patch provided by Heiko Schlittermann.
1698 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1699 Patch provided by Heiko Schlittermann.
1701 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1702 plus update to original patch.
1704 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1706 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1707 Patch provided by David Brownlee.
1709 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1711 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1713 NM/30 Bugzilla 888: TLS documentation bugfixes.
1715 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1717 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1718 Unlike the original bugzilla I have changed all shell scripts in src tree.
1720 NM/33 Bugzilla 898: Transport filter timeout fix.
1721 Patch by Todd Rinaldo.
1723 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1724 Patch by Serge Demonchaux.
1726 NM/35 Bugzilla 39: Base64 decode bug fixes.
1727 Patch by Jakob Hirsch.
1729 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1731 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1733 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1735 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1741 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1742 ALT-N's libdkim that I have put here:
1743 http://duncanthrax.net/exim-experimental/
1745 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1746 were called 'true' and it seems that C99 defines that as a reserved
1747 keyword to be used with 'bool' variable types. That means you could
1748 not include C99-style headers which use bools without triggering
1749 build errors in sieve.c.
1751 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1752 as mailq or other aliases. Changed the --help handling significantly
1753 to do whats expected. exim_usage() emits usage/help information.
1755 SC/01 Added the -bylocaldomain option to eximstats.
1757 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1759 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1761 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1767 PH/01 Another patch from the Sieve maintainer.
1769 PH/02 When an IPv6 address is converted to a string for single-key lookup
1770 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1771 dots are used instead of colons so that keys in lsearch files need not
1772 contain colons. This was done some time before quoting was made available
1773 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1774 (notated using the quote facility) so as to distinguish them from IPv4
1775 keys. This meant that lookups for IP addresses in host lists did not work
1776 for iplsearch lookups.
1778 This has been fixed by arranging for IPv6 addresses to be expressed with
1779 colons if the lookup type is iplsearch. This is not incompatible, because
1780 previously such lookups could never work.
1782 The situation is now rather anomolous, since one *can* have colons in
1783 ordinary lsearch keys. However, making the change in all cases is
1784 incompatible and would probably break a number of configurations.
1786 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1789 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1790 conversion specification without a maximum field width, thereby enabling
1791 a rogue spamd server to cause a buffer overflow. While nobody in their
1792 right mind would setup Exim to query an untrusted spamd server, an
1793 attacker that gains access to a server running spamd could potentially
1794 exploit this vulnerability to run arbitrary code as the Exim user.
1796 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1797 $primary_hostname instead of what libspf2 thinks the hosts name is.
1799 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1800 a directory entry by the name of the lookup key. Previously, if a
1801 symlink pointed to a non-existing file or a file in a directory that
1802 Exim lacked permissions to read, a lookup for a key matching that
1803 symlink would fail. Now it is enough that a matching directory entry
1804 exists, symlink or not. (Bugzilla 503.)
1806 PH/03 The body_linecount and body_zerocount variables are now exported in the
1809 PH/04 Added the $dnslist_matched variable.
1811 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1812 This means they are set thereafter only if the connection becomes
1815 PH/06 Added the client_condition to authenticators so that some can be skipped
1816 by clients under certain conditions.
1818 PH/07 The error message for a badly-placed control=no_multiline_responses left
1819 "_responses" off the end of the name.
1821 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1823 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1824 (without spaces) instead of just copying the configuration text.
1826 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1828 PH/11 Added $max_received_linelength.
1830 PH/12 Added +ignore_defer and +include_defer to host lists.
1832 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1833 way in which PCRE > 7.0 is built.
1835 PH/14 Implemented queue_only_load_latch.
1837 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1838 MAIL command. The effect was to mangle the value on 64-bit systems.
1840 PH/16 Another patch from the Sieve maintainer.
1842 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1844 PH/18 If a system quota error occurred while trying to create the file for
1845 a maildir delivery, the message "Mailbox is full" was not appended to the
1846 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1847 only to a quota excession during the actual writing of the file.
1849 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1850 characters?) which causes problems in log lines. The DN values are now
1851 passed through string_printing() before being added to log lines.
1853 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1854 and InterBase are left for another time.)
1856 PH/21 Added message_body_newlines option.
1858 PH/22 Guard against possible overflow in moan_check_errorcopy().
1860 PH/23 POSIX allows open() to be a macro; guard against that.
1862 PH/24 If the recipient of an error message contained an @ in the local part
1863 (suitably quoted, of course), incorrect values were put in $domain and
1864 $local_part during the evaluation of errors_copy.
1870 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1871 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1874 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1875 issue a MAIL command.
1877 PH/02 In an ACL statement such as
1879 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1881 if a client was not listed at all, or was listed with a value other than
1882 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1883 the condition was not true (as it should be), so access was not denied.
1884 The bug was that the ! inversion was incorrectly passed on to the second
1885 item. This has been fixed.
1887 PH/03 Added additional dnslists conditions == and =& which are different from
1888 = and & when the dns lookup returns more than one IP address.
1890 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1891 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1893 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1894 FSYNC, which compiles an option called disable_fsync that allows for
1895 bypassing fsync(). The documentation is heavily laced with warnings.
1897 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1899 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1900 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1901 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1902 including adding "make clean"; (3) Added -fPIC when compiling the test
1903 dynamically loaded module, to get rid of a warning.
1905 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1906 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1907 = 0s. The bug is otherwise harmless.
1909 PH/07 There was a bug in the dovecot authenticator such that the value of
1910 $auth1 could be overwritten, and so not correctly preserved, after a
1911 successful authentication. This usually meant that the value preserved by
1912 the server_setid option was incorrect.
1914 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1916 PH/09 Installed PCRE release 7.0.
1918 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1919 run for batched SMTP input. It is now run at the start of every message
1920 in the batch. While fixing this I discovered that the process information
1921 (output by running exiwhat) was not always getting set for -bs and -bS
1922 input. This is fixed, and it now also says "batched" for BSMTP.
1924 PH/11 Added control=no_pipelining.
1926 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1927 patch, slightly modified), and move the expansion of helo_data till after
1928 the connection is made in the smtp transport (so it can use these
1931 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1933 PH/14 Added log_selector = +pid.
1935 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1937 PH/16 Add ${if forany and ${if forall.
1939 PH/17 Added dsn_from option to vary the From: line in DSNs.
1941 PH/18 Flush SMTP output before performing a callout, unless control =
1942 no_callout_flush is set.
1944 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1945 was true (the default) a successful delivery failed to delete the retry
1946 item, thus causing premature timeout of the address. The bug is now
1949 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1951 PH/21 Long custom messages for fakedefer and fakereject are now split up
1952 into multiline reponses in the same way that messages for "deny" and
1953 other ACL rejections are.
1955 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1956 with slight modification.
1958 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1959 draft, changing the syntax and factoring some duplicate code".
1961 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1962 for deliveries of the second and subsequent messages over the same SMTP
1965 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1966 ${reduce, with only minor "tidies".
1968 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1970 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1971 expansion side effects.
1973 PH/27 When a message times out after an over-quota error from an Exim-imposed
1974 quota, the bounce message says "mailbox is full". This message was not
1975 being given when it was a system quota that was exceeded. It now should
1978 MH/03 Made $recipients available in local_scan(). local_scan() already has
1979 better access to the recipient list through recipients_list[], but
1980 $recipients can be useful in postmaster-provided expansion strings.
1982 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1983 in the case of a MAIL command with additional options following the
1984 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1985 were accidentally chopped off.
1987 PH/29 SMTP synchronization checks are implemented when a command is read -
1988 there is a check that no more input is waiting when there shouldn't be
1989 any. However, for some commands, a delay in an ACL can mean that it is
1990 some time before the response is written. In this time, more input might
1991 arrive, invalidly. So now there are extra checks after an ACL has run for
1992 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1993 pipelining has not been advertised.
1995 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1997 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1998 returned in the SMTP response when smtp_return_error_details was false.
1999 This has been fixed.
2001 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2002 instead of the C I/O that was originally supplied, because problems were
2003 reported on Solaris.
2005 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2006 Exim which did not show up earlier: it was assuming that a call to
2007 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2008 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2009 was a macro that became an assignment, so it seemed to work. This has
2010 changed to a proper function call with a void return, hence the compile
2011 error. Exim's code has been fixed.
2013 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2016 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2018 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2020 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2021 because it turns out to be too big or there is a timeout) there is no
2022 "Completed" line in the log. When some messages of this type were
2023 selected by exigrep, they were listed as "not completed". Others were
2024 picked up by some special patterns. I have improved the selection
2025 criteria to be more general.
2027 PH/38 The host_find_failed option in the manualroute router can now be set
2028 to "ignore", to completely ignore a host whose IP address cannot be
2029 found. If all hosts are ignored, the behaviour is controlled by the new
2030 host_all_ignored option.
2032 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2033 homing or because of multiple MX records with /mx) generated more than
2034 one IP address, and the following item turned out to be the local host,
2035 all the secondary addresses of the first item were incorrectly removed
2036 from the list, along with the local host and any following hosts (which
2037 is what is supposed to happen).
2039 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2040 whoever called Exim into the -H file. In the case of the daemon it was
2041 behaving confusingly. When first started, it used values for whoever
2042 started the daemon, but after a SIGHUP it used the Exim user (because it
2043 calls itself on a restart). I have changed the code so that it now always
2046 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2047 message are rejected with the same error (e.g. no authentication or bad
2048 sender address), and a DATA command is nevertheless sent (as can happen
2049 with PIPELINING or a stupid MUA), the error message that was given to the
2050 RCPT commands is included in the rejection of the DATA command. This is
2051 intended to be helpful for MUAs that show only the final error to their
2054 PH/42 Another patch from the Sieve maintainer.
2056 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2057 Eximstats - Fixed some broken HTML links and added missing column headers
2059 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2060 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2062 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2064 PH/43 Yet another patch from the Sieve maintainer.
2066 PH/44 I found a way to check for a TCP/IP connection going away before sending
2067 the response to the final '.' that terminates a message, but only in the
2068 case where the client has not sent further data following the '.'
2069 (unfortunately, this is allowed). However, in many cases there won't be
2070 any further data because there won't be any more messages to send. A call
2071 to select() can be used: if it shows that the input is "ready", there is
2072 either input waiting, or the socket has been closed. An attempt to read
2073 the next input character can distinguish the two cases. Previously, Exim
2074 would have sent an OK response which the client would never have see.
2075 This could lead to message repetition. This fix should cure that, at
2076 least in a lot of common cases.
2078 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2079 advertised in response to EHLO.
2085 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2086 fixed by 4.65/MH/01 (is this a record?) are fixed:
2088 (i) An empty string was always treated as zero by the numeric comparison
2089 operators. This behaviour has been restored.
2091 (ii) It is documented that the numeric comparison operators always treat
2092 their arguments as decimal numbers. This was broken in that numbers
2093 starting with 0 were being interpreted as octal.
2095 While fixing these problems I realized that there was another issue that
2096 hadn't been noticed. Values of message_size_limit (both the global option
2097 and the transport option) were treated as octal if they started with 0.
2098 The documentation was vague. These values are now always treated as
2099 decimal, and I will make that clear in the documentation.
2105 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2106 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2109 MH/01 Don't check that the operands of numeric comparison operators are
2110 integers when their expansion is in "skipping" mode (fixes bug
2111 introduced by 4.64-PH/07).
2113 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2114 child addresses, Exim now panics and dies. Previously, because the count
2115 is held in a short int, deliveries were likely to be lost. As such a
2116 large number of recipients for a single message is ridiculous
2117 (performance will be very, very poor), I have chosen to impose a limit
2118 rather than extend the field.
2124 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2125 leftover -K file (the existence of which was triggered by #402).
2126 While we were at it, introduced process PID as part of the -K
2127 filename. This should rule out race conditions when creating
2130 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2131 processing considerably. Previous code took too long for large mails,
2132 triggering a timeout which in turn triggers #401.
2134 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2135 in the DK code in transports.c. sendfile() is not really portable,
2136 hence the _LINUX specificness.
2138 TF/01 In the add_headers option to the mail command in an Exim filter,
2139 there was a bug that Exim would claim a syntax error in any
2140 header after the first one which had an odd number of characters
2143 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2144 callout verification, Exim cached a "reject" for the entire domain. This
2145 is correct for most verifications, but it is not correct for a recipient
2146 verification with use_sender or use_postmaster set, because in that case
2147 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2148 case of MAIL FROM:<> rejection from other early rejections (e.g.
2149 rejection of HELO). When verifying a recipient using a non-null MAIL
2150 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2151 Whatever the result of the callout, the value of the domain cache is
2152 left unchanged (for any other kind of callout, getting as far as trying
2153 RCPT means that the domain itself is ok).
2155 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2158 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2159 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2160 session, the daemon ignores ECONNECT errors and logs others; it now
2161 ignores EPIPE as well.
2163 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2164 (quoted-printable decoding).
2166 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2167 later the small subsequent patch to fix an introduced bug.
2169 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2171 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2173 PH/08 An error is now given if message_size_limit is specified negative.
2175 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2176 to be given (somewhat) arbitrary names.
2178 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2181 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2182 miscellaneous code fixes
2184 PH/10 Added the log_reject_target ACL modifier to specify where to log
2187 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2188 hostname. This is wrong, because it relates to the incoming message (and
2189 probably the interface on which it is arriving) and not to the outgoing
2190 callout (which could be using a different interface). This has been
2191 changed to use the value of the helo_data option from the smtp transport
2192 instead - this is what is used when a message is actually being sent. If
2193 there is no remote transport (possible with a router that sets up host
2194 addresses), $smtp_active_hostname is used.
2196 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2197 tweaks were necessary in order to get it to work (see also 21 below):
2198 (a) The code assumed that strncpy() returns a negative number on buffer
2199 overflow, which isn't the case. Replaced with Exim's string_format()
2201 (b) There were several signed/unsigned issues. I just did the minimum
2202 hacking in of casts. There is scope for a larger refactoring.
2203 (c) The code used strcasecmp() which is not a standard C function.
2204 Replaced with Exim's strcmpic() function.
2205 (d) The code set only $1; it now sets $auth1 as well.
2206 (e) A simple test gave the error "authentication client didn't specify
2207 service in request". It would seem that Dovecot has changed its
2208 interface. Fortunately there's a specification; I followed it and
2209 changed what the client sends and it appears to be working now.
2211 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2214 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2215 address A is aliased to B and C, where B exists and C does not. Without
2216 -v the output is "A verified" because verification stops after a
2217 successful redirection if more than one address is generated. However,
2218 with -v the child addresses are also verified. Exim was outputting "A
2219 failed to verify" and then showing the successful verification for C,
2220 with its parentage. It now outputs "B failed to verify", showing B's
2221 parentage before showing the successful verification of C.
2223 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2224 look up a TXT record in a specific list after matching in a combined
2227 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2228 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2229 they consult the DNS. I had assumed they would set it the way they
2230 wanted; and indeed my experiments on Linux seem to show that in some
2231 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2232 To be on the safe side, however, I have now made the interface to
2233 host_find_byname() similar to host_find_bydns(), with an argument
2234 containing the DNS resolver options. The host_find_byname() function now
2235 sets these options at its start, just as host_find_bydns() does. The smtp
2236 transport options dns_qualify_single and dns_search_parents are passed to
2237 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2238 of host_find_byname() use the default settings of RES_DEFNAMES
2239 (qualify_single) but not RES_DNSRCH (search_parents).
2241 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2242 spool_read_header() do less string testing, by means of a preliminary
2243 switch on the second character of optional "-foo" lines. (This is
2244 overdue, caused by the large number of possibilities that now exist.
2245 Originally there were few.) While I was there, I also converted the
2246 str(n)cmp tests so they don't re-test the leading "-" and the first
2247 character, in the hope this might squeeze out yet more improvement.
2249 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2250 flag allowing group syntax was set by the header_syntax check but not
2251 turned off, possible causing trouble later; (2) The flag was not being
2252 set at all for the header_verify test, causing "group"-style headers to
2253 be rejected. I have now set it in this case, and also caused header_
2254 verify to ignore an empty address taken from a group. While doing this, I
2255 came across some other cases where the code for allowing group syntax
2256 while scanning a header line wasn't quite right (mostly, not resetting
2257 the flag correctly in the right place). These bugs could have caused
2258 trouble for malformed header lines. I hope it is now all correct.
2260 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2261 with the "reply" argument non-NULL. The code, however (which originally
2262 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2263 but it didn't always do it. This confused somebody who was copying the
2264 code for some other use. I have removed all the tests.
2266 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2267 feature that was used to support insecure browsers during the U.S. crypto
2268 embargo. It requires special client support, and Exim is probably the
2269 only MTA that supported it -- and would never use it because real RSA is
2270 always available. This code has been removed, because it had the bad
2271 effect of slowing Exim down by computing (never used) parameters for the
2272 RSA_EXPORT functionality.
2274 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2275 authenticator to fail if there's a tab character in the incoming data
2276 (there should never be unless someone is messing about, as it's supposed
2277 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2278 if the connection is using TLS or if the remote IP is the same as the
2279 local IP, and the "valid-client-cert option" if a client certificate has
2282 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2283 authenticators. This can be used for authorization after authentication
2284 succeeds. (In the case of plaintext, it servers for both authentication
2287 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2288 if any retry times were supplied.
2290 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2291 connection, where there is no client IP address to check. In this
2292 situation, the verify now always succeeds.
2294 PH/25 Applied John Jetmore's -Mset patch.
2296 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2298 PH/27 In a string expansion for a processed (not raw) header when multiple
2299 headers of the same name were present, leading whitespace was being
2300 removed from all of them, but trailing whitespace was being removed only
2301 from the last one. Now trailing whitespace is removed from each header
2302 before concatenation. Completely empty headers in a concatenation (as
2303 before) are ignored.
2305 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2306 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2308 PH/29 [Removed. This was a change that I later backed out, and forgot to
2309 correct the ChangeLog entry (that I had efficiently created) before
2310 committing the later change.]
2312 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2313 address errors (4xx response to RCPT) over the same connection as other
2314 messages routed to the same hosts. Such deliveries are always "forced",
2315 so retry times are not inspected. This resulted in far too many retries
2316 for the affected addresses. The effect occurred only when there were more
2317 hosts than the hosts_max_try setting in the smtp transport when it had
2318 the 4xx errors. Those hosts that it had tried were not added to the list
2319 of hosts for which the message was waiting, so if all were tried, there
2320 was no problem. Two fixes have been applied:
2322 (i) If there are any address or message errors in an SMTP delivery, none
2323 of the hosts (tried or untried) are now added to the list of hosts
2324 for which the message is waiting, so the message should not be a
2325 candidate for sending over the same connection that was used for a
2326 successful delivery of some other message. This seems entirely
2327 reasonable: after all the message is NOT "waiting for some host".
2328 This is so "obvious" that I'm not sure why it wasn't done
2329 previously. Hope I haven't missed anything, but it can't do any
2330 harm, as the worst effect is to miss an optimization.
2332 (ii) If, despite (i), such a delivery is accidentally attempted, the
2333 routing retry time is respected, so at least it doesn't keep
2334 hammering the server.
2336 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2337 in ${readsocket because some servers need this prod.
2339 PH/32 Added some extra debug output when updating a wait-xxx database.
2341 PH/33 The hint "could be header name not terminated by colon", which has been
2342 given for certain expansion errors for a long time, was not being given
2343 for the ${if def:h_colon_omitted{... case.
2345 PH/34 The spec says: "With one important exception, whenever a domain list is
2346 being scanned, $domain contains the subject domain." There was at least
2347 one case where this was not true.
2349 PH/35 The error "getsockname() failed: connection reset by peer" was being
2350 written to the panic log as well as the main log, but it isn't really
2351 panic-worthy as it just means the connection died rather early on. I have
2352 removed the panic log writing for the ECONNRESET error when getsockname()
2355 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2356 runs only) independently of the message's sender address. This meant
2357 that, if the 4xx error was in fact related to the sender, a different
2358 message to the same recipient with a different sender could confuse
2359 things. In particualar, this can happen when sending to a greylisting
2360 server, but other circumstances could also provoke similar problems.
2361 I have changed the default so that the retry time for these errors is now
2362 based a combination of the sender and recipient addresses. This change
2363 can be overridden by setting address_retry_include_sender=false in the
2366 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2367 remote server are returned as part of bounce messages. This was not
2368 happening for LMTP over a pipe (the lmtp transport), but now it is the
2369 same for both kinds of LMTP.
2371 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2372 in header lines that were in fact CNAMEs. This is no longer the case.
2374 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2375 and queue runs started by the daemon processed all messages. This has
2376 been fixed so that -R and -S can now usefully be given with -q<time>.
2378 PH/40 Import PCRE release 6.7 (fixes some bugs).
2380 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2382 PH/42 Give an error if -q is specified more than once.
2384 PH/43 Renamed the variables $interface_address and $interface_port as
2385 $received_ip_address and $received_port, to make it clear that these
2386 values apply to message reception, and not to the outgoing interface when
2387 a message is delivered. (The old names remain recognized, of course.)
2389 PH/44 There was no timeout on the connect() call when using a Unix domain
2390 socket in the ${readsocket expansion. There now is.
2392 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2393 be meaningful with "accept".
2395 SC/01 Eximstats V1.43
2396 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2398 SC/02 Eximstats V1.44
2399 Use a glob alias rather than an array ref in the generated
2400 parser. This improves both readability and performance.
2402 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2403 Collect SpamAssassin and rejection statistics.
2404 Don't display local sender or destination tables unless
2405 there is data to show.
2406 Added average volumes into the top table text output.
2408 SC/04 Eximstats V1.46
2409 Collect data on the number of addresses (recipients)
2410 as well as the number of messages.
2412 SC/05 Eximstats V1.47
2413 Added 'Message too big' to the list of mail rejection
2414 reasons (thanks to Marco Gaiarin).
2416 SC/06 Eximstats V1.48
2417 Mainlog lines which have GMT offsets and are too short to
2418 have a flag are now skipped.
2420 SC/07 Eximstats V1.49 (Alain Williams)
2421 Added the -emptyok flag.
2423 SC/08 Eximstats V1.50
2424 Fixes for obtaining the IP address from reject messages.
2426 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2427 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2428 whitesspace changes from 4.64-PH/27
2430 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2433 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2434 are found, allow negative numbers in numeric criteria)
2436 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2438 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2439 to match changes made in 4.64-PH/43
2441 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2443 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2444 "domain" is actually an IP address, and force a failure. This locks out
2445 those revolvers/nameservers that support "A-for-A" lookups, in
2446 contravention of the specifications.
2448 PH/48 When a host name was looked up from an IP address, and the subsequent
2449 forward lookup of the name timed out, the host name was left in
2450 $sender_host_name, contrary to the specification.
2452 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2453 restricted to single-key lookups, Exim was not diagnosing an error if
2454 * or *@ was used with a query-style lookup.
2456 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2458 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2459 long ago, but noone interested enough thought of it. Let's just say that
2460 the "1.1" means that there are some new functions that weren't there at
2461 some point in the past.
2463 PH/51 Error processing for expansion failure of helo_data from an smtp
2464 transport during callout processing was broken.
2466 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2467 tested/used via the -bh/-bhc/-bs options.
2469 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2470 bug, fixed in subsequent PCRE releases).
2472 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2473 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2475 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2481 SC/01 Use a glob alias rather than an array ref in eximstats generated
2482 parser. This improves both readability and performance.
2484 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2485 Don't display local sender or destination tables in eximstats unless
2486 there is data to show.
2487 Added average volumes into the eximstats top table text output.
2489 SC/03 Collect data on the number of addresses (recipients) as well
2490 as the number of messages in eximstats.
2492 TF/01 Correct an error in the documentation for the redirect router. Exim
2493 does (usually) call initgroups() when daemonizing.
2495 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2496 with consistent privilege compared to when running as a daemon.
2498 TF/03 Note in the spec that $authenticated_id is not set for local
2499 submissions from trusted users.
2501 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2502 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2504 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2505 by adding some example configuration directives to the default
2506 configuration file. A little bit of work is required to uncomment the
2507 directives and define how usernames and passwords are checked, but
2508 there is now a framework to start from.
2510 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2511 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2512 without this. I don't know how relevant this is to other LDAP libraries.
2514 PH/02 Add the verb name to the "unknown ACL verb" error.
2516 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2518 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2520 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2521 directory not expanded when it should be if an expanded home directory
2522 was set for the address (which is overridden by the transport).
2524 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2527 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2528 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2529 because it is too late at that time, and has no effect.
2531 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2532 security issue with \' (bugzilla #107). I could not use the
2533 PQescapeStringConn() function, because it needs a PGconn value as one of
2536 PH/08 When testing addresses using -bt, indicate those final addresses that
2537 are duplicates that would not cause an additional delivery. At least one
2538 person was confused, thinking that -bt output corresponded to deliveries.
2539 (Suppressing duplicates isn't a good idea as you lose the information
2540 about possibly different redirections that led to the duplicates.)
2542 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2543 systems where poll() doesn't work, in particular OS X.
2545 PH/10 Added more information to debugging output for retry time not reached.
2547 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2548 operations in malware.c.
2550 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2553 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2554 syslog_duplication set false, log lines that would normally be written
2555 both the the main log and to the reject log were not written to syslog at
2558 PH/14 In the default configuration, change the use of "message" in ACL warn
2559 statements to "add_header".
2561 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2562 not followed by a command (e.g. "seen endif").
2564 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2565 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2568 PH/17 Added extra conditions to the default value of delay_warning_condition
2572 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2573 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2574 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2577 The Auto-Submitted: and various List- headers are standardised, whereas I
2578 don't think Precedence: ever was.
2580 PH/18 Refactored debugging code in route_finduser() to show more information,
2581 in particular, the error code if getpwnam() issues one.
2583 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2584 This is apparently needed in addition to the PH/07 change above to avoid
2585 any possible encoding problems.
2587 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2588 but not after initializing Perl.
2590 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2591 output them only if debugging. By default they are written stderr,
2592 apparently, which is not desirable.
2594 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2597 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2600 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2602 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2603 authenticated or an ident call has been made. Suppress the default
2604 values for $authenticated_id and $authenticated_sender (but permit -oMai
2605 and -oMas) when testing with -bh.
2607 PH/24 Re-jigged the order of the tests in the default configuration so that the
2608 tests for valid domains and recipients precede the DNS black list and CSA
2609 tests, on the grounds that those ones are more expensive.
2611 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2612 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2613 This bug exists in every version of Exim that I still have, right back to
2616 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2617 However, an attempt to turn on case-sensitivity in a regex key by
2618 including (?-i) didn't work because the subject string was already
2619 lowercased, and the effects were non-intuitive. It turns out that a
2620 one-line patch can be used to allow (?-i) to work as expected.
2626 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2627 other effects) broke the use of negated acl sub-conditions.
2629 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2632 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2633 "Deny" causes Exim to reject the incoming connection with a 554 error.
2634 Unfortunately, if there is a major crisis, such as a disk failure,
2635 tcp-wrappers gives "deny", whereas what one would like would be some
2636 kind of temporary error. A kludge has been added to help with this.
2637 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2638 554 error is used if errno is still zero or contains ENOENT (which occurs
2639 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2642 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2644 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2645 errors. Otherwise a message that provokes a temporary error (when other
2646 messages do not) can cause a whole host to time out.
2648 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2649 addresses were routed directly to files or pipes from a redirect router.
2650 File deliveries just didn't batch; pipe deliveries might have suffered
2653 PH/06 A failure to get a lock for a hints database would erroneously always say
2654 "Failed to get write lock", even when it was really a read lock.
2656 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2657 of 0600. This has been changed to use the value of the lockfile_mode
2658 option (which defaults to 0600).
2660 PH/08 Applied small patch from the Sieve maintainer.
2662 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2663 folder from quota calculations, a direct delivery into this folder messed
2664 up the contents of the maildirsize file. This was because the regex was
2665 used only to exclude .Trash (or whatever) when the size of the mailbox
2666 was calculated. There was no check that a delivery was happening into an
2667 excluded directory. This bug has been fixed by ignoring all quota
2668 processing for deliveries into excluded directories.
2670 PH/10 Added the maildirfolder_create_regex option to appendfile.
2676 PH/01 The code for finding all the local interface addresses on a FreeBSD
2677 system running IPv6 was broken. This may well have applied to all BSD
2678 systems, as well as to others that have similar system calls. The broken
2679 code found IPv4 interfaces correctly, but gave incorrect values for the
2680 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2681 that it would not match correctly against @[] and not recognize the IPv6
2684 PH/02 The ipliteral router was not recognizing addresses of the form user@
2685 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2687 PH/03 Added disable_ipv6.
2689 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2690 decoded form, because it is most often used to construct To: headers
2691 lines in autoreplies, and the decoded form may well be syntactically
2692 invalid. However, $reply_address has leading white space removed, and all
2693 newlines turned into spaces so that the autoreply transport does not
2696 PH/05 If group was specified without a user on a router, and no group or user
2697 was specified on a transport, the group from the router was ignored.
2699 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2700 for visible compile-time settings that can be used to change these
2701 numbers, for those that want even more. Backwards compatibility with old
2702 spool files has been maintained. However, going back to a previous Exim
2703 release will lost any variables that are in spool files.
2705 PH/07 Two small changes when running in the test harness: increase delay when
2706 passing a TCP/IP connection to a new process, in case the original
2707 process has to generate a bounce, and remove special handling of
2708 127.0.0.2 (sic), which is no longer necessary.
2710 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2711 be the same on different OS.
2713 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2716 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2717 whether --show-vars was specified or not
2719 JJ/02 exipick: Added support for new ACL variable spool format introduced
2722 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2723 syntactically invalid From: or Reply-to: line, and a filter used this to
2724 generate an autoreply, and therefore failed to obtain an address for the
2725 autoreply, Exim could try to deliver to a non-existent relative file
2726 name, causing unrelated and misleading errors. What now happens is that
2727 it logs this as a hard delivery error, but does not attempt to create a
2730 PH/11 The exinext utility has a -C option for testing purposes, but although
2731 the given file was scanned by exinext itself; it wasn't being passed on
2732 when Exim was called.
2734 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2735 an end-of-file indication when reading a command response.
2737 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2738 compiled. In many other places in Exim, IPv6 addresses are always
2739 recognized, so I have changed this. It also means that IPv4 domain
2740 literals of the form [IPV4:n.n.n.n] are now always recognized.
2742 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2743 used if the router is not running as root, for example, when verifying at
2744 ACL time, or when using -bh. The debugging output from this situation was
2745 non-existent - all you got was a failure to exec. I have made two
2748 (a) Failures to set uid/gid, the current directory, or a process leader
2749 in a subprocess such as that created by queryprogram now generate
2750 suitable debugging ouput when -d is set.
2752 (b) The queryprogram router detects when it is not running as root,
2753 outputs suitable debugging information if -d is set, and then runs
2754 the subprocess without attempting to change uid/gid.
2756 PH/15 Minor change to Makefile for building test_host (undocumented testing
2759 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2760 additional section of a DNS packet that returns MX or SRV records.
2761 Instead, it always explicitly searches for A/AAAA records. This avoids
2762 major problems that occur when a DNS server includes only records of one
2763 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2764 fixed another bug: if SRV records were looked up and the corresponding
2765 address records were *not* found in the additional section, the port
2766 values from the SRV records were lost.
2768 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2769 using the correct key (the original address) when searching the retry
2770 rules in order to find which one to use for generating the retry hint.
2772 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2773 adding the default one. Similarly, if it contains a Reply-To: header, the
2774 errors_reply_to option, if set, is not used.
2776 PH/19 When calculating a retry time, Exim used to measure the "time since
2777 failure" by looking at the "first failed" field in the retry record. Now
2778 it does not use this if it is later than than the arrival time of the
2779 message. Instead it uses the arrival time. This makes for better
2780 behaviour in cases where some deliveries succeed, thus re-setting the
2781 "first failed" field. An example is a quota failure for a huge message
2782 when small messages continue to be delivered. Without this change, the
2783 "time since failure" will always be short, possible causing more frequent
2784 delivery attempts for the huge message than are intended.
2785 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2787 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2788 $1, $2, $3) because the numerical variables can be reset during some
2789 expansion items (e.g. "match"), thereby losing the authentication data.
2791 PH/21 Make -bV show the size of off_t variables so that the test suite can
2792 decide whether to run tests for quotas > 2G.
2794 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2795 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2796 filecount value is greater than 2G or if a quota value is greater than 2G
2797 on a system where the size of off_t is not greater than 4, a panic error
2800 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2801 never match. The debug and -bh output now contains an explicit error
2802 message indicating a malformed IPv4 address or mask.
2804 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2805 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2806 PH/23 above applies.
2808 PH/25 Do not write to syslog when running in the test harness. The only
2809 occasion when this arises is a failure to open the main or panic logs
2810 (for which there is an explicit test).
2812 PH/26 Added the /no_tell option to "control=freeze".
2814 PH/27 If a host name lookup failed very early in a connection, for example, if
2815 the IP address matched host_lookup and the reverse lookup yielded a name
2816 that did not have a forward lookup, an error message of the form "no IP
2817 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2818 could be logged. Now it outputs the IP address instead of "NULL".
2820 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2821 allows the sender and the authenticated sender to be set when
2822 submitting a message from within Exim. Since child_open_exim() is
2823 documented for local_scan(), the new function should be too.
2825 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2826 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2827 results in an empty string is now treated as unset.
2829 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2831 PH/31 Added sender_verify_fail logging option.
2833 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2834 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2835 tidied the source and removed it altogether.
2837 PH/33 When a queue run was abandoned because the load average was too high, a
2838 log line was always written; now it is written only if the queue_run log
2839 selector is set. In addition, the log line for abandonment now contains
2840 information about the queue run such as the pid. This is always present
2841 in "start" and "stop" lines but was omitted from the "abandon" line.
2843 PH/34 Omit spaces between a header name and the colon in the error message that
2844 is given when verify = headers_syntax fails (if there are lots of them,
2845 the message gets confusing).
2847 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2848 names, as there are now some PTR records that contain slashes. This check
2849 is only to protect against broken name servers that fall over on strange
2850 characters, so the fact that it applies to all lookups doesn't matter.
2852 PH/36 Now that the new test suite is complete, we can remove some of the
2853 special code in Exim that was needed for the old test suite. For example,
2854 sorting DNS records because real resolvers return them in an arbitrary
2855 order. The new test suite's fake resolver always returns records in the
2858 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2859 bounces) except when queue_only is set, to avoid logging races between
2860 the different processes.
2862 PH/38 Panic-die if .include specifies a non-absolute path.
2864 PH/39 A tweak to the "H" retry rule from its user.
2866 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2867 a label. They prevented compilation on older perls.
2869 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2870 a warning to be raised on newish perls.
2872 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2873 on queue. Changes to match documented behaviour of showing count of
2874 messages matching specified criteria.
2876 PH/40 Changed the default ident timeout from 30s to 5s.
2878 PH/41 Added support for the use of login_cap features, on those BSD systems
2879 that have them, for controlling the resources used by pipe deliveries.
2881 PH/42 The content-scanning code uses fopen() to create files in which to put
2882 message data. Previously it was not paying any attention to the mode of
2883 the files. Exim runs with umask(0) because the rest of the code creates
2884 files with open(), and sets the required mode explicitly. Thus, these
2885 files were ending up world-writeable. This was not a big issue, because,
2886 being within the spool directory, they were not world-accessible. I have
2887 created a function called modefopen, which takes an additional mode
2888 argument. It sets umask(777), creates the file, chmods it to the required
2889 mode, then resets the umask. All the relevant calls to fopen() in the
2890 content scanning code have been changed to use this function.
2892 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2893 to 24 hours. This avoids potential overflow problems when processing G
2894 and H retry rules. I suspect nobody ever tinkers with this value.
2896 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2898 PH/45 When the plaintext authenticator is running as a client, the server's
2899 challenges are checked to ensure they are valid base64 strings. By
2900 default, the authentication attempt is cancelled if an invalid string is
2901 received. Setting client_ignore_invalid_base64 true ignores these errors.
2902 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2903 they are received. Thus, the responses can be made to depend on the
2904 challenges. If an invalid string is ignored, an empty string is placed in
2907 PH/46 Messages that are created by the autoreply transport now contains a
2908 References: header, in accordance with RFCs 2822 and 3834.
2910 PH/47 Added authenticated_sender_force to the smtp transport.
2912 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2914 PH/49 Installed latest patch from the Sieve maintainer.
2916 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2917 was also set, the appendfile transport was unnecessarily scanning a
2918 directory of message files (e.g. for maildir delivery) to find the count
2919 of files (along with the size), even though it did not need this
2920 information. It now does the scan only if it needs to find either the
2921 size of the count of files.
2923 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2925 PH/52 Two bugs concerned with error handling when the smtp transport is
2928 (i) Exim was not creating retry information for temporary errors given
2929 for individual recipients after the DATA command when the smtp transport
2930 was used in LMTP mode. This meant that they could be retried too
2931 frequently, and not timed out correctly.
2933 (ii) Exim was setting the flag that allows error details to be returned
2934 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2935 recipients that were returned after the DATA command.
2937 PH/53 This is related to PH/52, but is more general: for any failing address,
2938 when detailed error information was permitted to be returned to the
2939 sender, but the error was temporary, then after the final timeout, only
2940 "retry timeout exceeded" was returned. Now it returns the full error as
2941 well as "retry timeout exceeded".
2943 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2944 do this, and (what is worse) MTAs that accept it.
2946 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2947 will now be deprecated.
2949 PH/56 New os.c-cygwin from the Cygwin maintainer.
2951 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2952 formats (previously only available in exim formats via -bpr, -bpru,
2953 and -bpra. Now also available in native and exiqgrep formats)
2955 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2956 with very large, slow to parse queues
2958 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2960 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2962 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2963 responses to authentication challenges, though it was showing the
2964 challenges; (ii) I've removed the CR characters from the debug output for
2967 PH/58 Allow for the insertion of a newline as well as a space when a string
2968 is turned into more than one encoded-word during RFC 2047 encoding. The
2969 Sieve code now uses this.
2971 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2972 data_4xx, lost_connection, tls_required.
2974 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2975 message was being sent as an SMTP response.
2977 PH/61 Add -l and -k options to exicyclog.
2979 PH/62 When verifying, if an address was redirected to one new address, so that
2980 verification continued, and the new address failed or deferred after
2981 having set something in $address_data, the value of $address_data was not
2982 passed back to the ACL. This was different to the case when no
2983 redirection occurred. The value is now passed back in both cases.
2985 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2986 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2987 use login_cap.h, so on its own it isn't the distinguishing feature. The
2988 new name refers directly to the setclassresources() function.
2990 PH/65 Added configuration files for NetBSD3.
2992 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2994 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2995 is preferred over IPv4.
2997 PH/68 The bounce_return_message and bounce_return_body options were not being
2998 honoured for bounces generated during the reception of non-SMTP messages.
2999 In particular, this applied to messages rejected by the ACL. This bug has
3000 been fixed. However, if bounce_return_message is true and bounce_return_
3001 body is false, the headers that are returned for a non-SMTP message
3002 include only those that have been read before the error was detected.
3003 (In the case of an ACL rejection, they have all been read.)
3005 PH/69 The HTML version of the specification is now built in a directory called
3006 spec_html instead of spec.html, because the latter looks like a path with
3007 a MIME-type, and this confuses some software.
3009 PH/70 Catch two compiler warnings in sieve.c.
3011 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3012 function verify_get_ident() calls ip_connect() to connect a socket, but
3013 if the "connect()" function timed out, ip_connect() used to close the
3014 socket. However, verify_get_ident() also closes the socket later, and in
3015 between Exim writes to the log, which may get opened at this point. When
3016 the socket was closed in ip_connect(), the log could get the same file
3017 descriptor number as the socket. This naturally causes chaos. The fix is
3018 not to close the socket in ip_connect(); the socket should be closed by
3019 the function that creates it. There was only one place in the code where
3020 this was missing, in the iplookup router, which I don't think anybody now
3021 uses, but I've fixed it anyway.
3023 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3024 well as to direct DNS lookups. Otherwise the handling of names in host
3025 lists is inconsistent and therefore confusing.
3031 PH/01 Two changes to the default runtime configuration:
3033 (1) Move the checks for relay_from_hosts and authenticated clients from
3034 after to before the (commented out) DNS black list checks.
3036 (2) Add control=submission to the relay_from_hosts and authenticated
3037 clients checks, on the grounds that messages accepted by these
3038 statements are most likely to be submissions.
3040 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3042 (1) Generate an error if the third argument for the ${prvs expansion is
3045 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3048 (3) Reset the variables that are obtained from the first argument of
3049 ${prvscheck and used in the second argument before leaving the code,
3050 because their memory is reclaimed, so using them afterwards may do
3053 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3054 one (it's much easier than Tom thought :-).
3056 (5) Because of (4), we can now allow for the use of $prvscheck_result
3057 inside the third argument.
3059 PH/03 For some reason, the default setting of PATH when running a command from
3060 a pipe transport was just "/usr/bin". I have changed it to
3063 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3064 anything to be listed in the output from -bV.
3066 PH/05 When a filter generated an autoreply, the entire To: header line was
3067 quoted in the delivery log line, like this:
3069 => >A.N.Other <ano@some.domain> <original@ddress> ...
3071 This has been changed so that it extracts the operative address. There
3072 may be more than one such address. If so, they are comma-separated, like
3075 => >ano@some.domain,ona@other.domain <original@ddress> ...
3077 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3078 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3079 not being looked up in the rDNS to get a host name, Exim was showing the
3080 IP address twice in Received: lines, even though the IP addresses were
3081 identical. For example:
3083 Received: from [1.2.3.4] (helo=[1.2.3.4])
3085 However, if the real host name was known, it was omitting the HELO data
3086 if it matched the actual IP address. This has been tidied up so that it
3087 doesn't show the same IP address twice.
3089 PH/07 When both +timestamp and +memory debugging was on, the value given by
3090 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3091 called by the debug printing, thereby overwriting the timestamp buffer.
3092 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3094 PH/08 When the original message was included in an autoreply transport, it
3095 always said "this is a copy of the message, including all the headers",
3096 even if body_only or headers_only was set. It now gives an appropriate
3099 PH/09 Applied a patch from the Sieve maintainer which:
3101 o fixes some comments
3102 o adds the (disabled) notify extension core
3103 o adds some debug output for the result of if/elsif tests
3104 o points to the current vacation draft in the documentation
3105 and documents the missing references header update
3109 o fixes a bug in processing the envelope test (when testing
3110 multiple envelope elements, the last element determinted the
3113 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3114 Electronic Mail") by including:
3116 Auto-submitted: auto-generated
3118 in the messages that it generates (bounce messages and others, such as
3119 warnings). In the case of bounce messages for non-SMTP mesages, there was
3120 also a typo: it was using "Auto_submitted" (underscore instead of
3121 hyphen). Since every message generated by Exim is necessarily in response
3122 to another message, thes have all been changed to:
3124 Auto-Submitted: auto-replied
3126 in accordance with these statements in the RFC:
3128 The auto-replied keyword:
3130 - SHOULD be used on messages sent in direct response to another
3131 message by an automatic process,
3133 - MUST NOT be used on manually-generated messages,
3135 - MAY be used on Delivery Status Notifications (DSNs) and Message
3136 Disposition Notifications (MDNs),
3138 - MUST NOT be used on messages generated by automatic or periodic
3139 processes, except for messages which are automatic responses to
3142 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3143 to the default Received: header definition.
3145 PH/12 Added log selector acl_warn_skipped (default on).
3147 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3148 variables because (a) they are in the wrong storage pool and (b) even if
3149 they were copied, it wouldn't work properly because of the caching.
3151 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3152 checking when decoding. Apparently there are clients that generate
3153 overlong encoded strings. Why am I not surprised?
3155 PH/15 If the first argument of "${if match_address" was not empty, but did not
3156 contain an "@" character, Exim crashed. Now it writes a panic log message
3157 and treats the condition as false.
3159 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3161 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3162 extension "envelope-auth". The code is finished and in agreement with
3163 other implementations, but there is no documentation so far and in fact,
3164 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3165 not changing the active code.
3167 Print executed "if" and "elsif" statements when debugging is used. This
3168 helps a great deal to understand what a filter does.
3170 Document more things not specified clearly in RFC3028. I had all this
3171 sorted out, when out of a sudden new issues came to my mind. Oops."
3173 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3176 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3177 canonical form (as documented). However, after a host name lookup from
3178 the IP address, check_host() was doing a simple string comparison with
3179 addresses acquired from the DNS when checking that the found name did
3180 have the original IP as one of its addresses. Since any found IPv6
3181 addresses are likely to be in abbreviated form, the comparison could
3182 fail. Luckily, there already exists a function for doing the comparison
3183 by converting both addresses to binary, so now that is used instead of
3184 the text comparison.
3186 PH/20 There was another similar case to PH/19, when a complete host name was
3187 given in a host list; looking up its IP address could give an abbreviated
3188 form, whereas the current host's name might or might not be abbreviated.
3189 The same fix has been applied.
3195 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3196 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3199 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3200 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3202 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3204 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3205 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3206 Also, for "Auto-Submitted", treat anything other than "no" as
3207 non-personal, in accordance with RFC 3834. (Previously it treated
3208 anything starting "auto-" as non-personal.)
3210 TF/01 The control=submission/name=... option had a problem with syntax
3211 errors if the name included a slash character. The /name= option
3212 now slurps the rest of the string, so it can include any characters
3213 but it must come last in the list of options (after /sender_retain
3216 PH/05 Some modifications to the interface to the fake nameserver for the new
3224 TK/01 Added the "success_on_redirect" address verification option. See
3225 NewStuff for rationale and an example.
3227 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3229 PH/02 Patch to exigrep to allow it to work on syslog lines.
3231 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3232 fread() to skip over the body file's header line, because in Cygwin the
3233 header line is locked and is inaccessible.
3235 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3236 co-exist for some time) to make it clear that it is the Exim ID that is
3237 referenced, not the Message-ID: header line.
3239 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3240 string_format() function, because snprintf() does not exist on all
3243 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3244 ${stat: expansion item.
3246 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3247 protocol synchronization error", to keep the pedants happy.
3249 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3250 well as for IRIX systems, when gcc is being used. See the host.c source
3253 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3255 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3258 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3259 transport and to the smtp transport in LMTP mode.
3261 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3263 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3264 run a filter in a subprocess. This could lead to confusion in subsequent
3265 lookups in the parent process. There should also be a search_tidyup() at
3266 the end of the subprocess.
3268 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3269 only if the host matched helo_try_verify_hosts, which caused the
3270 verification to occur when the EHLO/HELO command was issued. The ACL just
3271 tested the remembered result. Now, if a previous verification attempt has
3272 not happened, "verify = helo" does it there and then.
3274 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3276 TK/03 Fix log output including CR from clamd.
3278 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3279 exist provoked a memory error which could cause a segfault.
3281 PH/15 Installed PCRE 6.2
3283 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3285 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3286 of the problem. Specifically, suggested +O2 rather than +O1 for the
3289 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3291 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3292 record, the port was not correctly logged when the outgoing_port log
3293 selector was set (it logged the transort's default port).
3295 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3296 fallback_hosts, and "hosts" in the smtp transport.
3298 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3299 host errors such as "Connection refused".
3301 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3302 authentication with radiusclient 0.4.9:
3304 - Error returned from rc_read_config was caught wrongly
3305 - Username/password not passed on to radius server due to wrong length.
3307 The presumption is that some radiusclient API changes for 4.51/PH/17
3308 were not taken care of correctly. The code is still untested by me (my
3309 Linux distribution still has 0.3.2 of radiusclient), but it was
3310 contributed by a Radius user.
3312 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3313 expanding the "port" option of the smtp transport.
3315 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3316 while reading a MIME header. Thanks to Tom Hughes for a patch.
3318 PH/24 Include config.h inside local_scan.h so that configuration settings are
3321 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3322 that in an ACL for RCPT (for example), you can examine exactly what was
3325 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3326 commands, but it was not correctly comparing the address with the actual
3327 client host address. Thus, it would show the EHLO address in Received:
3328 header lines when this was not necessary.
3330 PH/27 Added the % operator to ${eval:}.
3332 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3333 it should be ignoring failures (because with -C, for example, it has lost
3334 privilege). It wasn't ignoring creation failures other than "already
3337 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3340 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3341 because an input line was too long, either on its own, or by virtue of
3342 too many continuations, the temporary file was not being removed, and the
3343 return code was incorrect.
3345 PH/31 Missing "BOOL" in function definition in filtertest.c.
3347 PH/32 Applied Sieve patches from the maintainer.
3349 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3351 PH/33 Added "verify = not_blind".
3353 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3354 Local/Makefile (with some defaults set). These are used in built scripts
3355 such as exicyclog, but they have never been used in the exim_install
3356 script (though there are many overriding facilities there). I have
3357 arranged that the exim_install script now takes note of these two
3360 PH/35 Installed configuration files for Dragonfly.
3362 PH/36 When a locally submitted message by a trusted user did not contain a
3363 From: header, and the sender address was obtained from -f or from an SMTP
3364 MAIL command, and the trusted user did not use -F to supply a sender
3365 name, $originator_name was incorrectly used when constructing a From:
3366 header. Furthermore, $originator_name was used for submission mode
3367 messages from external hosts without From: headers in a similar way,
3368 which is clearly wrong.
3370 PH/37 Added control=suppress_local_fixups.
3372 PH/38 When log_selector = +received_sender was set, and the addition of the
3373 sender made the log line's construction buffer exactly full, or one byte
3374 less than full, an overflow happened when the terminating "\n" was
3377 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3378 when the result of a list match is failure because a DNS lookup failed.
3380 PH/40 RM_COMMAND is now used in the building process.
3382 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3383 the "build-* directories that it finds.
3385 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3386 address in a domain literal was a prefix of an interface address.
3388 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3389 when verifying a sender address, unless rewrite_headers is false.
3391 PH/44 Wrote a long comment about why errors_to addresses are verified as
3392 recipients, not senders.
3394 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3395 the ratelimit ACL was added.
3397 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3399 PH/46 Added extra information about PostgreSQL errors to the error string.
3401 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3402 avoiding the need to install special zones in a real server. This is
3403 backwards compatible; if it can't find the fake resolver, it drops back.
3404 Thus, both old and new test suites can be run.
3406 TF/02 Added util/ratelimit.pl
3408 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3409 clock is set back in time.
3411 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3412 Candler <B.Candler@pobox.com>.
3414 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3415 OK for addresses that are the result of redirections.
3417 PH/48 A number of further additions for the benefit of the new test suite,
3418 including a fake gethostbyname() that interfaces to the fake DNS resolver
3421 TF/06 The fix for widen_domains has also been applied to qualify_single and
3422 search_parents which are the other dnslookup options that can cause
3425 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3428 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3430 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3431 service" if it encountered an MX record with an empty target hostname.
3432 The message is now "an MX or SRV record indicated no SMTP service".
3434 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3435 if the DNS of the sending site is misconfigured. This is quite a
3436 common situation. This change restores the behaviour of treating a
3437 helo verification defer as a failure.
3439 PH/51 If self=fail was set on a router, the bounce message did not include the
3440 actual error message.
3446 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3448 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3449 command itself ended in error, the underlying message about the transport
3450 filter timeout was being overwritten with the pipe command error. Now the
3451 underlying error message should be appended to the second error message.
3453 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3455 PH/02 Reduce the amount of output that "make" produces by default. Full output
3456 can still be requested.
3458 PH/03 The warning log line about a condition test deferring for a "warn" verb
3459 was being output only once per connection, rather than after each
3460 occurrence (because it was using the same function as for successful
3461 "warn" verbs). This seems wrong, so I have changed it.
3463 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3464 it should not have, which might have caused a crash in the right
3465 circumstances, but probably never did.
3467 PH/04 Installed a modified version of Tony Finch's patch to make submission
3468 mode fix the return path as well as the Sender: header line, and to
3469 add a /name= option so that you can make the user's friendly name appear
3472 TF/03 Added the control = fakedefer ACL modifier.
3474 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3475 Mark Lowes for thorough testing.
3477 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3479 TK/03 Merged latest SRS patch from Miles Wilton.
3481 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3482 with the definition in sysexits.h (which is #included earlier).
3483 Fortunately, Exim does not actually use EX_OK. The code used to try to
3484 preserve the sysexits.h value, by assumimg that macro definitions were
3485 scanned for macro replacements. I have been disabused of this notion,
3486 so now the code just undefines EX_OK before #including unistd.h.
3488 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3489 in the smtp transport. When a block could not be written in a single
3490 write() function, the timeout was being re-applied to each part-write.
3491 This seems wrong - if the receiver was accepting one byte at a time it
3492 would take for ever. The timeout is now adjusted when this happens. It
3493 doesn't have to be particularly precise.
3495 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3496 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3498 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3499 without a domain if the check to <postmaster@domain> fails.
3501 SC/01 Eximstats: added -xls and the ability to specify output files
3502 (patch written by Frank Heydlauf).
3504 SC/02 Eximstats: use FileHandles for outputing results.
3506 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3508 SC/04 Eximstats: fixed display of large numbers with -nvr option
3510 SC/05 Eximstats: fixed merging of reports with empty tables.
3512 SC/06 Eximstats: added the -include_original_destination flag
3514 SC/07 Eximstats: removed tabs and trailing whitespace.
3516 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3518 TK/06 MBOX spool code: Add real "From " MBOX separator line
3519 so the .eml file is really in mbox format (even though
3520 most programs do not really care). Patch from Alex Miller.
3522 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3523 The latter is generated from $received_to and is only set if the
3524 message has one envelope recipient. SA can use these headers,
3525 obviously out-of-the-box. Patch from Alex Miller.
3527 PH/08 The ${def test on a variable was returning false if the variable's
3528 value was "0", contrary to what the specification has always said!
3529 The result should be true unless the variable is empty.
3531 PH/09 The syntax error of a character other than { following "${if
3532 def:variable_name" (after optional whitespace) was not being diagnosed.
3533 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3534 accidental colon was present, for example, could give incorrect results.
3536 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3537 result is used (not including appendfile, where other changes are about
3540 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3541 This involved changing a lot of size variables from int to off_t. It
3542 should work with maildirs and everything.
3544 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3545 spamd dying while we are connected to it.
3547 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3550 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3551 function for BDB 4.3.
3553 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3555 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3556 the sources for PCRE have been re-arranged and more files are now
3559 PH/15 The code I had for printing potentially long long variables in PH/11
3560 above was not the best (it lost precision). The length of off_t variables
3561 is now inspected at build time, and an appropriate printing format (%ld
3562 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3563 to be "long long int" or "long int". This is needed for the internal
3564 formatting function string_vformat().
3566 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3567 the configuration file to be ":syslog", then the script "guesses" where
3568 the logs files are, rather than using the compiled in default. In our
3569 case the guess is not the same as the compiled default, so the script
3570 suddenly stopped working when I started to use syslog. The patch checks
3571 to see if log_file_path is "". If so, it attempts to read it from exim
3572 with no configuration file to get the compiled in version, before it
3573 falls back to the previous guessing code."
3575 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3576 implementing BATV in an Exim configuration. See NewStuff for the gory
3579 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3580 Makefile that are specific to HP-UX.
3582 PH/18 If the "use_postmaster" option was set for a recipient callout together
3583 with the "random" option, the postmaster address was used as the MAIL
3584 FROM address for the random test, but not for the subsequent recipient
3585 test. It is now used for both.
3587 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3588 patch removes a few documentation additions to RFC 3028, because the
3589 latest draft now contains them. It adds the new en;ascii-case comparator
3590 and a new error check for 8bit text in MIME parts. Comparator and
3591 require names are now matched exactly. I enabled the subaddress
3592 extension, but it is not well tested yet (read: it works for me)."
3594 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3595 rework some of the code of TK/09 above to avoid the hardwired use of
3596 "%lld" and "long long". Replaced the call to snprintf() with a call to
3599 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3600 records point to non-existent hosts", "retry timeout exceeded", and
3601 "retry time not reached for any host after a long failure period".
3603 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3604 experimental DomainKeys support:
3606 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3607 (2) On an error such as an illegally used "control", the wrong name for
3608 the control was given.
3610 These problems did NOT occur unless DomainKeys support was compiled.
3612 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3614 PH/24 Added ${if match_ip condition.
3616 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3617 databases so that it will be absolutely obvious if a crash occurs in the
3618 DB library. This is a regular occurrence (often caused by mis-matched
3621 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3622 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3623 picked up on a user's system that detects such things. There doesn't seem
3624 to be a gcc warning option for this - only an attribute that has to be
3625 put on the function's prototype. It seems that in Fedora Core 4 they have
3626 set this on a number of new functions. No doubt there will be more in due
3629 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3630 specify a transport. However, if an address that was verified by such a
3631 router was the subject of a callout, Exim crashed because it tried to
3632 read the rcpt_include_affixes from the non-existent transport. Now it
3633 just assumes that the setting of that option is false. This bug was
3634 introduced by 4.51/PH/31.
3636 PH/28 Changed -d+all to exclude +memory, because that information is very
3637 rarely of interest, but it makes the output a lot bigger. People tend to
3638 do -d+all out of habit.
3640 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3641 code in os-type was giving problems when libc.so lives in lib64, like on
3644 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3645 aren't the modern standard, and it seems that some systems' include files
3646 don't always have them. Exim was already checking for some of the newer
3647 ones like T_AAAA, and defining it itself. I've added checks for all the
3648 record types that Exim uses.
3650 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3651 not automatically generating a new one, as it is supposed to. This
3652 prevented TLS from working. If the file did exist, but contained invalid
3653 data, a new version was generated, as expected. It was only the case of a
3654 non-existent file that was broken.
3656 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3657 with a change in libdomainkeys > 0.64.
3659 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3660 from DNS. If the selector record carries the flag, it now has
3661 precedence over the domain-wide flag.
3663 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3665 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3666 the use of a port name that isn't defined in /etc/services) occurred, the
3667 message was deferred as in a normal delivery, and thus remained on the
3668 spool, instead of being failed because of the mua_wrapper setting. This
3669 is now fixed, and I tidied up some of the mua_wrapper messages at the
3672 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3673 the messages in a hash of arrays rather than using individual hashes.
3674 This is a bit cleaner and results in dramatic memory savings, albeit
3675 at a slight CPU cost.
3677 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3678 as requested by Marc Sherman.
3680 SC/10 Eximstats: added histograms for user specified patterns as requested
3683 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3685 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3686 fopen() in the content-scanning modules that did not already have it.
3692 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3693 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3695 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3697 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3699 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3700 to test Sieve filters that use "vacation".
3702 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3703 that changes the way the GnuTLS parameters are stored in the cache file.
3704 The new format can be generated externally. For backward compatibility,
3705 if the data in the cache doesn't make sense, Exim assumes it has read an
3706 old-format file, and it generates new data and writes a new file. This
3707 means that you can't go back to an older release without removing the
3710 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3711 work if there are any delivery delays because "one_time" forces the
3712 parent to be marked "delivered", so its unseen clone is never tried
3713 again. For this reason, Exim now forbids the simultaneous setting of
3716 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3717 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3718 entry if you want to know the details. The fix, however, neglected to
3719 consider the case when local delivery batching is involved. The test for
3720 "previously delivered" was not happening when checking to see if an
3721 address could be batched with a previous (undelivered) one; under
3722 certain circumstances this could lead to multiple deliveries to the same
3725 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3726 in its include files, and this causes problems building Exim.
3728 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3729 header_syntax) but Exim was just ignoring anything given after a slash.
3730 In particular, this caused confusion with an attempt to use "verify =
3731 reverse_host_lookup/defer_ok". An error is now given when options are
3732 supplied for verify items that do not have them. (Maybe reverse_host_
3733 lookup should have a defer_ok option, but that's a different point.)
3735 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3736 defined by RFC 821) to 2048, because there were problems with some AUTH
3737 commands, and RFC 1869 says the size should be increased for extended
3738 SMTP commands that take arguments.
3740 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3743 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3744 "unknown" error; now it says that the functionality isn't in the binary.
3746 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3747 an address' error message when a string expansion fails (syntax or
3748 whatever). Otherwise the password may appear in the log. Following change
3749 PH/42 below, there is no longer a chance of it appearing in a bounce
3752 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3754 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3755 of its addresses was ever tried. (Bugzilla bug #2.)
3757 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3758 the result incorrectly in the debug output. (It correctly added a newline
3759 to what was transported.)
3761 TF/01 Added $received_time.
3763 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3764 commented out examples of how to interface to a virus scanner and to
3765 SpamAssassin. Also added commented examples of av_scanner and
3766 spamd_address settings.
3768 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3769 and controls are allowed in which ACLs. There were a couple of minor
3770 errors. Some of the entries in the conditions table (which is a table of
3771 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3772 negation of where the condition IS allowed.
3774 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3776 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3777 header file does not have a version number, so I've had to invent a new
3778 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3779 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3780 radiusclient), but it was contributed by a Radius user.
3782 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3783 files or directories, for OpenSSL.
3785 PH/19 When an Exim process that is running as root has to create an Exim log
3786 file, it does so in a subprocess that runs as exim:exim so as to get the
3787 ownership right at creation (otherwise, other Exim processes might see
3788 the file with the wrong ownership). There was no test for failure of this
3789 fork() call, which would lead to the process getting stuck as it waited
3790 for a non-existent subprocess. Forks do occasionally fail when resources
3791 run out. I reviewed all the other calls to fork(); they all seem to check
3794 PH/20 When checking for unexpected SMTP input at connect time (before writing
3795 the banner), Exim was not dealing correctly with a non-positive return
3796 from the read() function. If the client had disconnected by this time,
3797 the result was a log entry for a synchronization error with an empty
3798 string after "input=" when read() returned zero. If read() returned -1
3799 (an event I could not check), uninitialized data bytes were printed.
3800 There were reports of junk text (parts of files, etc) appearing after
3803 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3805 PH/22 Added support for macro redefinition, and (re)definition in between
3806 driver and ACL definitions.
3808 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3809 forgetting to use the resulting value; it was using the unexpanded value.
3811 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3812 hadn't been configured. The fix is from Juergen Kreileder, who
3813 understands it better than I do:
3815 "Here's what I see happening with three configured cyrus_sasl
3816 authenticators configured (plain, login, cram-md5):
3818 On startup auth_cyrus_sasl_init() gets called for each of these.
3819 This means three calls to sasl_listmech() without a specified mech_list.
3820 => SASL tests which mechs of all available mechs actually work
3821 => three warnings about OTP not working
3822 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3824 With the patch, sasl_listmech() also gets called three times. But now
3825 SASL's mech_list option is set to the server_mech specified in the the
3826 authenticator. Or in other words, the answer from sasl_listmech()
3827 gets limited to just the mech you're testing for (which is different
3829 => the return list contains just 'plain' or 'login', 'cram-md5' or
3830 nothing depending on the value of ob->server_mech.
3832 I've just tested the patch: Authentication still works fine,
3833 unavailable mechs specified in the exim configuration are still
3834 caught, and the auth.log warnings about OTP are gone."
3836 PH/25 When debugging is enabled, the contents of the command line are added
3837 to the debugging output, even when log_selector=+arguments is not
3840 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3841 answer is "GNU", and only if the return is "GNU/something" is the answer
3844 PH/27 $acl_verify_message is now set immediately after the failure of a
3845 verification in an ACL, and so is available in subsequent modifiers. In
3846 particular, the message can be preserved by coding like this:
3848 warn !verify = sender
3849 set acl_m0 = $acl_verify_message
3851 Previously, $acl_verify_message was set only while expanding "message"
3852 and "log_message" when a very denied access.
3854 PH/28 Modified OS/os.c-Linux with
3856 -#ifndef OS_LOAD_AVERAGE
3857 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3859 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3860 nomenclature these days.)
3862 PH/29 Installed patch from the Sieve maintainer that adds the options
3863 sieve_useraddress and sieve_subaddress to the redirect router.
3865 PH/30 In these circumstances:
3866 . Two addresses routed to the same list of hosts;
3867 . First host does not offer TLS;
3868 . First host accepts first address;
3869 . First host gives temporary error to second address;
3870 . Second host offers TLS and a TLS session is established;
3871 . Second host accepts second address.
3872 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3873 and peerdn, if requested) that were in fact used only for the second
3876 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3877 attention to any local part prefix or suffix that was matched by the
3878 router that accepted the address. It now behaves in the same way as it
3879 does for delivery: the affixes are removed from the local part unless
3880 rcpt_include_affixes is set on the transport.
3882 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3883 timeout during the DATA phase of an incoming message.
3885 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3886 applied a patch sanctioned by the Sieve maintainer.
3888 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3889 the uid or gid is negative. A case of a negative gid caused this to be
3890 noticed. The fix allows for either to be negative.
3892 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3893 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3896 PH/36 The change PH/12 above was broken. Fixed it.
3898 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3899 the grounds that routing the same address twice would always produce the
3900 same answer. This might have been true once, but it is certainly no
3901 longer true now. Routing a child address may depend on the previous
3902 routing that produced that child. Some complicated redirection strategies
3903 went wrong when messages had multiple recipients, and made Exim's
3904 behaviour dependent on the order in which the addresses were given.
3906 I have moved the duplicate checking until after the routing is complete.
3907 Exim scans the addresses that are assigned to local and remote
3908 transports, and removes any duplicates. This means that more work will be
3909 done, as duplicates will always all be routed, but duplicates are
3910 presumably rare, so I don't expect this is of any significance.
3912 For deliveries to pipes, files, and autoreplies, the duplicate checking
3913 still happens during the routing process, since they are not going to be
3916 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3917 It corrects a timeout issue with spamd. This is Ian's comment: "The
3918 background is that sometimes spamd either never reads data from a
3919 connection it has accepted, or it never writes response data. The exiscan
3920 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3921 blindly assumes that writes won't block so it may never time out."
3923 PH/39 Allow G after quota size as well as K and M.
3925 PH/40 The value set for $authenticated_id in an authenticator may not contain
3926 binary zeroes or newlines because the value is written to log lines and
3927 to spool files. There was no check on this. Now the value is run through
3928 the string_printing() function so that such characters are converted to
3929 printable escape sequences.
3931 PH/41 $message_linecount is a new variable that contains the total number of
3932 lines in the message. Compare $body_linecount, which is the count for the
3935 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3936 bounce and delay warning messages, except in certain special cases, which
3939 (a) An SMTP error message from a remote host;
3940 (b) A message specified in a :fail: redirection;
3941 (c) A message specified in a "fail" command in a system filter;
3942 (d) A message specified in a FAIL return from the queryprogram router;
3943 (e) A message specified by the cannot_route_message router option.
3945 In these cases only, Exim does include the error details in bounce and
3946 warning messages. There are also a few cases where bland messages such
3947 as "unrouteable address" or "local delivery error" are given.
3949 PH/43 $value is now also set for the "else" part of a ${run expansion.
3951 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3952 being worked on, but at least Exim now implements the latest version to
3955 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3956 process to complete was treated as a delivery failure, a timeout while
3957 writing the message to the pipe was logged, but erroneously treated as a
3958 successful delivery. Such timeouts include transport filter timeouts. For
3959 consistency with the overall process timeout, these timeouts are now
3960 treated as errors, giving rise to delivery failures by default. However,
3961 there is now a new Boolean option for the pipe transport called
3962 timeout_defer, which, if set TRUE, converts the failures into defers for
3963 both kinds of timeout. A transport filter timeout is now identified in
3966 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3967 systems where "make" and "gmake" are different, calling "gmake" at top
3968 level broke things. I've arranged for the value of $(MAKE) to be passed
3969 from the Makefile to this script so that it can call the same version of
3973 A note about Exim versions 4.44 and 4.50
3974 ----------------------------------------
3976 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3977 changes of various kinds. As a consequence, a big documentation update was
3978 needed. This delayed the release for rather longer than seemed good, especially
3979 in the light of a couple of (minor) security issues. Therefore, the changes
3980 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3981 release. So 4.44 and 4.50 are in effect two different branches that both start
3984 I have left the 4.50 change log unchanged; it contains all the changes since
3985 4.43. The change log for 4.44 is below; many of its items are identical to
3986 those for 4.50. This seems to be the most sensible way to preserve the
3987 historical information.
3993 1. Minor wording change to the doc/README.SIEVE file.
3995 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3996 computation of the current number of files was incorrect.
3998 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3999 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4002 4. Give more explanation in the error message when the command for a transport
4003 filter fails to execute.
4005 5. There are several places where Exim runs a non-Exim command in a
4006 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4007 was being done only for the command run by the queryprogram router. It is
4008 now done for all such subprocesses. The other cases are: ${run, transport
4009 filters, and the commands run by the lmtp and pipe transports.
4011 6. Added CONFIGURE_GROUP build-time option.
4013 7. Some older OS have a limit of 256 on the maximum number of file
4014 descriptors. Exim was using setrlimit() to set 1000 as a large value
4015 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4016 systems. I've change it so that if it can't get 1000, it tries for 256.
4018 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4019 was an oversight, and furthermore, ever since the addition of extra
4020 controls (e.g. 4.43/32), the checks on when to allow different forms of
4021 "control" were broken. There should now be diagnostics for all cases when a
4022 control that does not make sense is encountered.
4024 9. Added the /retain_sender option to "control=submission".
4026 10. $recipients is now available in the predata ACL (oversight).
4028 11. Tidy the search cache before the fork to do a delivery from a message
4029 received from the command line. Otherwise the child will trigger a lookup
4030 failure and thereby defer the delivery if it tries to use (for example) a
4031 cached ldap connection that the parent has called unbind on.
4033 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4034 of $address_data from the recipient verification was clobbered by the
4035 sender verification.
4037 13. The value of address_data from a sender verification is now available in
4038 $sender_address_data in subsequent conditions in the ACL statement.
4040 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4042 15. Added a new option "connect=<time>" to callout options, to set a different
4045 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4046 was its contents. (It was OK if the option was not defined at all.)
4048 17. A "Completed" log line is now written for messages that are removed from
4049 the spool by the -Mrm option.
4051 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4052 information about exactly what failed.
4054 19. Added -dd to debug only the daemon process.
4056 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4057 handles timeouts, both on the server side and network timeouts. Renamed the
4058 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4060 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4061 It is now set to "smtps".
4063 22. $host_address is now set to the target address during the checking of
4064 ignore_target_hosts.
4066 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4067 being passed; this would have caused $sender_host_name to have been used if
4068 matching the list had actually called for a host name (not very likely,
4069 since this list is usually IP addresses). A host name is now passed as
4072 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4073 code that specifies a non-restarting handler (typically sigaction() in
4074 modern systems) in an attempt to fix a rare and obscure crash bug.
4076 25. Narrowed the window for a race in the daemon that could cause it to ignore
4077 SIGCHLD signals. This is not a major problem, because they are used only to
4078 wake it up if nothing else does.
4080 26. A malformed maildirsize file could cause Exim to calculate negative values
4081 for the mailbox size or file count. Odd effects could occur as a result.
4082 The maildirsize information is now recalculated if the size or filecount
4085 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4086 support for a long time. Removed HAVE_SYS_VFS_H.
4088 28. Installed the latest version of exipick from John Jetmore.
4090 29. In an address list, if the pattern was not a regular expression, an empty
4091 subject address (from a bounce message) matched only if the pattern was an
4092 empty string. Non-empty patterns were not even tested. This was the wrong
4093 because it is perfectly reasonable to use an empty address as part of a
4094 database query. An empty address is now tested by patterns that are
4095 lookups. However, all the other forms of pattern expect the subject to
4096 contain a local part and a domain, and therefore, for them, an empty
4097 address still always fails if the pattern is not itself empty.
4099 30. Exim went into a mad DNS loop when attempting to do a callout where the
4100 host was specified on an smtp transport, and looking it up yielded more
4101 than one IP address.
4103 31. Re-factored the code for checking spool and log partition space into a
4104 function that finds that data and another that does the check. The former
4105 is then used to implement four new variables: $spool_space, $log_space,
4106 $spool_inodes, and $log_inodes.
4108 32. The RFC2047 encoding function was originally intended for short strings
4109 such as real names; it was not keeping to the 75-character limit for
4110 encoded words that the RFC imposes. It now respects the limit, and
4111 generates multiple encoded words if necessary. To be on the safe side, I
4112 have increased the buffer size for the ${rfc2047: expansion operator from
4115 33. It is now permitted to omit both strings after an "if" condition; if the
4116 condition is true, the result is "true". As before, when the second string
4117 is omitted, a false condition yields an empty string. This makes it less
4118 cumbersome to write custom ACL and router conditions.
4120 34. Failure to deliver a bounce message always caused it to be frozen, even if
4121 there was an errors_to setting on the router. The errors_to setting is now
4124 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4125 canonical form (fully expanded) before being placed in
4126 $sender_host_address.
4128 36. The table in the code that translates DNS record types into text (T_A to
4129 "A" for instance) was missing entries for NS and CNAME. It is just possible
4130 that this could have caused confusion if both these types were looked up
4131 for the same domain, because the text type is used as part of Exim's
4132 per-process caching. But the chance of anyone hitting this buglet seems
4135 37. The dnsdb lookup has been extended in a number of ways.
4137 (1) There is a new type, "zns", which walks up the domain tree until it
4138 finds some nameserver records. It should be used with care.
4140 (2) There is a new type, "mxh", which is like "mx" except that it returns
4141 just the host names, not the priorities.
4143 (3) It is now possible to give a list of domains (or IP addresses) to be
4144 looked up. The behaviour when one of the lookups defers can be
4145 controlled by a keyword.
4147 (4) It is now possible to specify the separator character for use when
4148 multiple records are returned.
4150 38. The dnslists ACL condition has been extended: it is now possible to supply
4151 a list of IP addresses and/or domains to be looked up in a particular DNS
4154 39. Added log_selector=+queue_time_overall.
4156 40. When running the queue in the test harness, wait just a tad after forking a
4157 delivery process, to get repeatability of debugging output.
4159 41. Include certificate and key file names in error message when GnuTLS fails
4160 to set them up, because the GnuTLS error message doesn't include the name
4161 of the failing file when there is a problem reading it.
4163 42. Allow both -bf and -bF in the same test run.
4165 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4167 44. The "Exiscan patch" is now merged into the mainline Exim source.
4169 45. Sometimes the final signoff response after QUIT could fail to get
4170 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4171 < 0 before doing a fflush(). This bug looks as though it goes back to the
4172 introduction of TLS in release 3.20, but "sometimes" must have been rare
4173 because the tests only now provoked it.
4175 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4176 (this can affect the format of dates).
4178 47. exim_tidydb, when checking for the continued existence of a message for
4179 which it has found a message-specific retry record, was not finding
4180 messages that were in split spool directories. Consequently, it was
4181 deleting retry records that should have stayed in existence.
4183 48. Steve fixed some bugs in eximstats.
4185 49. The SPA authentication driver was not abandoning authentication and moving
4186 on to the next authenticator when an expansion was forced to fail,
4187 contradicting the general specification for all authenticators. Instead it
4188 was generating a temporary error. It now behaves as specified.
4190 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4191 (the order specifies the preference for clients). The order is now AES256,
4192 AES128, 3DES, ARCFOUR128.
4194 51. Small patch to Sieve code - explicitly set From: when generating an
4197 52. Exim crashed if a remote delivery caused a very long error message to be
4198 recorded - for instance if somebody sent an entire SpamAssassin report back
4199 as a large number of 550 error lines. This bug was coincidentally fixed by
4200 increasing the size of one of Exim's internal buffers (big_buffer) that
4201 happened as part of the Exiscan merge. However, to be on the safe side, I
4202 have made the code more robust (and fixed the comments that describe what
4205 53. Now that there can be additional text after "Completed" in log lines (if
4206 the queue_time_overall log selector is set), a one-byte patch to exigrep
4207 was needed to allow it to recognize "Completed" as not the last thing in
4210 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4211 patch that reportedly fixes this has been added. I am not expert enough to
4212 create a test for it. This is what the patch creator wrote:
4214 "I found a little strange behaviour of ldap code when working with
4215 Windows 2003 AD Domain, where users was placed in more than one
4216 Organization Units. When I tried to give exim partial DN, the exit code
4217 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4218 But simultaneously result of request was absolutely normal ldap result,
4219 so I produce this patch..."
4221 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4222 REFERENCE, so I have modified the code to exclude the patch when that macro
4225 55. Some experimental protocols are using DNS PTR records for new purposes. The
4226 keys for these records are domain names, not reversed IP addresses. The
4227 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4228 leaves it alone. Component reversal etc. now happens only for IP addresses.
4231 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4233 57. Double the size of the debug message buffer (to 2048) so that more of very
4234 long debug lines gets shown.
4236 58. The exicyclog utility now does better if the number of log files to keep
4237 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4239 59. Two changes related to the smtp_active_hostname option:
4241 (1) $smtp_active_hostname is now available as a variable.
4242 (2) The default for smtp_banner uses $smtp_active_hostname instead
4243 of $primary_hostname.
4245 60. The host_aton() function is supposed to be passed a string that is known
4246 to be a valid IP address. However, in the case of IPv6 addresses, it was
4247 not checking this. This is a hostage to fortune. Exim now panics and dies
4248 if the condition is not met. A case was found where this could be provoked
4249 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4250 components; fortuitously, this particular loophole had already been fixed
4251 by change 4.50/55 above.
4253 If there are any other similar loopholes, the new check in host_aton()
4254 itself should stop them being exploited. The report I received stated that
4255 data on the command line could provoke the exploit when Exim was running as
4256 exim, but did not say which command line option was involved. All I could
4257 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4258 running as the user.
4261 61. There was a buffer overflow vulnerability in the SPA authentication code
4262 (which came originally from the Samba project). I have added a test to the
4263 spa_base64_to_bits() function which I hope fixes it.
4266 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4267 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4269 63. The daemon start-up calls getloadavg() while still root for those OS that
4270 need the first call to be done as root, but it missed one case: when
4271 deliver_queue_load_max is set with deliver_drop_privilege. This is
4272 necessary for the benefit of the queue runner, because there is no re-exec
4273 when deliver_drop_privilege is set.
4275 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4276 This has been fixed.
4278 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4279 was in use, was not putting the data itself into the right store pool;
4280 consequently, it could be overwritten for a subsequent message in the same
4281 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4284 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4286 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4287 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4288 the function were treating the return as a boolean value, which happened to
4289 work because 0=false and not-0=true, but is not correct code.
4291 68. The host_aton() function was not handling scoped IPv6 addresses (those
4292 with, for example, "%eth0" on the end) correctly.
4294 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4295 negated items (that is, ~something) in unsigned ints. Some compilers
4296 apparently mutter when there is no cast.
4298 70. If an address verification called from an ACL failed, and did not produce a
4299 user-specific message (i.e. there was only a "system" message), nothing was
4300 put in $acl_verify_message. In this situation, it now puts the system
4303 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4304 session; change 4.31/43 added the unwanted input to the log line - except
4305 that it did not do this in the start of session case. It now does.
4307 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4308 This is wrong and can cause the other end to generate a synchronization
4309 error if it is another Exim or anything else that does the synchronization
4310 check. A QUIT command is no longer sent after a timeout.
4312 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4313 during host lookups.
4315 74. The defer_ok option of callout verification was not working if it was used
4316 when verifying addresses in header lines, that is, for this case:
4318 verify = header_sender/callout=defer_ok
4320 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4321 those file descriptors could be used for SMTP connections. If anything
4322 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4323 could be sent to the SMTP client, causing chaos. The daemon now opens
4324 stdin, stdout, and stderr to /dev/null when it puts itself into the
4327 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4328 log by default. The user can override this with suitable Perl magic.
4330 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4331 the log message when discard triggers, was not working for the DATA ACL or
4332 for the non-SMTP ACL.
4334 78. Error message wording change in sieve.c.
4336 79. If smtp_accept_max_per_host was set, the number of connections could be
4337 restricted to fewer than expected, because the daemon was trying to set up
4338 a new connection before checking whether the processes handling previous
4339 connections had finished. The check for completed processes is now done
4340 earlier. On busy systems, this bug wouldn't be noticed because something
4341 else would have woken the daemon, and it would have reaped the completed
4344 80. If a message was submitted locally by a user whose login name contained one
4345 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4346 It caused a spool format error. I have fixed the spool reading code. A
4347 related problem was that the "from" clause in the Received: line became
4348 illegal because of the space(s). It is now covered by ${quote_local_part.
4350 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4353 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4355 83. Patch from Sieve maintainer to fix unterminated string problem in
4356 "vacation" handling.
4358 84. Some minor changes to the Linux configuration files to help with other
4359 OS variants using glibc.
4361 85. One more patch for Sieve to update vacation handling to latest spec.
4364 ----------------------------------------------------
4365 See the note above about the 4.44 and 4.50 releases.
4366 ----------------------------------------------------
4372 1. Change 4.43/35 introduced a bug that caused file counts to be
4373 incorrectly computed when quota_filecount was set in an appendfile
4376 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4377 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4380 3. Give more explanation in the error message when the command for a transport
4381 filter fails to execute.
4383 4. There are several places where Exim runs a non-Exim command in a
4384 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4385 was being done only for the command run by the queryprogram router. It is
4386 now done for all such subprocesses. The other cases are: ${run, transport
4387 filters, and the commands run by the lmtp and pipe transports.
4389 5. Some older OS have a limit of 256 on the maximum number of file
4390 descriptors. Exim was using setrlimit() to set 1000 as a large value
4391 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4392 systems. I've change it so that if it can't get 1000, it tries for 256.
4394 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4395 was an oversight, and furthermore, ever since the addition of extra
4396 controls (e.g. 4.43/32), the checks on when to allow different forms of
4397 "control" were broken. There should now be diagnostics for all cases when a
4398 control that does not make sense is encountered.
4400 7. $recipients is now available in the predata ACL (oversight).
4402 8. Tidy the search cache before the fork to do a delivery from a message
4403 received from the command line. Otherwise the child will trigger a lookup
4404 failure and thereby defer the delivery if it tries to use (for example) a
4405 cached ldap connection that the parent has called unbind on.
4407 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4408 of $address_data from the recipient verification was clobbered by the
4409 sender verification.
4411 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4412 was its contents. (It was OK if the option was not defined at all.)
4414 11. A "Completed" log line is now written for messages that are removed from
4415 the spool by the -Mrm option.
4417 12. $host_address is now set to the target address during the checking of
4418 ignore_target_hosts.
4420 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4421 being passed; this would have caused $sender_host_name to have been used if
4422 matching the list had actually called for a host name (not very likely,
4423 since this list is usually IP addresses). A host name is now passed as
4426 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4427 code that specifies a non-restarting handler (typically sigaction() in
4428 modern systems) in an attempt to fix a rare and obscure crash bug.
4430 15. Narrowed the window for a race in the daemon that could cause it to ignore
4431 SIGCHLD signals. This is not a major problem, because they are used only to
4432 wake it up if nothing else does.
4434 16. A malformed maildirsize file could cause Exim to calculate negative values
4435 for the mailbox size or file count. Odd effects could occur as a result.
4436 The maildirsize information is now recalculated if the size or filecount
4439 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4440 support for a long time. Removed HAVE_SYS_VFS_H.
4442 18. Updated exipick to current release from John Jetmore.
4444 19. Allow an empty sender to be matched against a lookup in an address list.
4445 Previously the only cases considered were a regular expression, or an
4448 20. Exim went into a mad DNS lookup loop when doing a callout where the
4449 host was specified on the transport, if the DNS lookup yielded more than
4452 21. The RFC2047 encoding function was originally intended for short strings
4453 such as real names; it was not keeping to the 75-character limit for
4454 encoded words that the RFC imposes. It now respects the limit, and
4455 generates multiple encoded words if necessary. To be on the safe side, I
4456 have increased the buffer size for the ${rfc2047: expansion operator from
4459 22. Failure to deliver a bounce message always caused it to be frozen, even if
4460 there was an errors_to setting on the router. The errors_to setting is now
4463 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4464 canonical form (fully expanded) before being placed in
4465 $sender_host_address.
4467 24. Updated eximstats to version 1.33
4469 25. Include certificate and key file names in error message when GnuTLS fails
4470 to set them up, because the GnuTLS error message doesn't include the name
4471 of the failing file when there is a problem reading it.
4473 26. Expand error message when OpenSSL has problems setting up cert/key files.
4476 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4477 (this can affect the format of dates).
4479 28. exim_tidydb, when checking for the continued existence of a message for
4480 which it has found a message-specific retry record, was not finding
4481 messages that were in split spool directories. Consequently, it was
4482 deleting retry records that should have stayed in existence.
4484 29. eximstats updated to version 1.35
4485 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4486 1.35 - bugfix such that pie charts by volume are generated correctly
4488 30. The SPA authentication driver was not abandoning authentication and moving
4489 on to the next authenticator when an expansion was forced to fail,
4490 contradicting the general specification for all authenticators. Instead it
4491 was generating a temporary error. It now behaves as specified.
4493 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4494 (the order specifies the preference for clients). The order is now AES256,
4495 AES128, 3DES, ARCFOUR128.
4497 31. Small patch to Sieve code - explicitly set From: when generating an
4500 32. Exim crashed if a remote delivery caused a very long error message to be
4501 recorded - for instance if somebody sent an entire SpamAssassin report back
4502 as a large number of 550 error lines. This bug was coincidentally fixed by
4503 increasing the size of one of Exim's internal buffers (big_buffer) that
4504 happened as part of the Exiscan merge. However, to be on the safe side, I
4505 have made the code more robust (and fixed the comments that describe what
4508 33. Some experimental protocols are using DNS PTR records for new purposes. The
4509 keys for these records are domain names, not reversed IP addresses. The
4510 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4511 leaves it alone. Component reversal etc. now happens only for IP addresses.
4514 34. The host_aton() function is supposed to be passed a string that is known
4515 to be a valid IP address. However, in the case of IPv6 addresses, it was
4516 not checking this. This is a hostage to fortune. Exim now panics and dies
4517 if the condition is not met. A case was found where this could be provoked
4518 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4519 components; fortuitously, this particular loophole had already been fixed
4520 by change 4.50/55 or 4.44/33 above.
4522 If there are any other similar loopholes, the new check in host_aton()
4523 itself should stop them being exploited. The report I received stated that
4524 data on the command line could provoke the exploit when Exim was running as
4525 exim, but did not say which command line option was involved. All I could
4526 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4527 running as the user.
4530 35. There was a buffer overflow vulnerability in the SPA authentication code
4531 (which came originally from the Samba project). I have added a test to the
4532 spa_base64_to_bits() function which I hope fixes it.
4535 36. The daemon start-up calls getloadavg() while still root for those OS that
4536 need the first call to be done as root, but it missed one case: when
4537 deliver_queue_load_max is set with deliver_drop_privilege. This is
4538 necessary for the benefit of the queue runner, because there is no re-exec
4539 when deliver_drop_privilege is set.
4541 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4542 was in use, was not putting the data itself into the right store pool;
4543 consequently, it could be overwritten for a subsequent message in the same
4544 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4547 38. Sometimes the final signoff response after QUIT could fail to get
4548 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4549 < 0 before doing a fflush(). This bug looks as though it goes back to the
4550 introduction of TLS in release 3.20, but "sometimes" must have been rare
4551 because the tests only now provoked it.
4557 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4558 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4559 Now it returns an integer. A number of calls were still expecting a T/F
4560 return. Fortuitously, in all cases, the tests worked in OK situations,
4561 which is the norm. However, things would have gone wrong on any write
4562 failures on the smtp file descriptor. This function is used when sending
4563 messages over SMTP and also when doing verify callouts.
4565 2. When Exim is called to do synchronous delivery of a locally submitted
4566 message (the -odf or -odi options), it no longer closes stderr before doing
4569 3. Implemented the mua_wrapper option.
4571 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4573 5. Implemented the functions header_remove(), header_testname(),
4574 header_add_at_position(), and receive_remove_recipient(), and exported them
4577 6. If an ACL "warn" statement specified the addition of headers, Exim already
4578 inserted X-ACL-Warn: at the start if there was no header name. However, it
4579 was not making this test for the second and subsequent header lines if
4580 there were newlines in the string. This meant that an invalid header could
4581 be inserted if Exim was badly configured.
4583 7. Allow an ACL "warn" statement to add header lines at the start or after all
4584 the Received: headers, as well as at the end.
4586 8. Added the rcpt_4xx retry error code.
4588 9. Added postmaster_mailfrom=xxx to callout verification option.
4590 10. Added mailfrom=xxxx to the callout verification option, for verify=
4593 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4594 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4596 12. Inserted some casts to stop certain compilers warning when using pointer
4597 differences as field lengths or precisions in printf-type calls (mostly
4598 affecting debugging statements).
4600 13. Added optional readline() support for -be (dynamically loaded).
4602 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4603 same clock tick as a message's arrival, so that its received time was the
4604 same as the "first fail" time on the retry record, and that message
4605 remained on the queue past the ultimate address timeout, every queue runner
4606 would try a delivery (because it was past the ultimate address timeout) but
4607 after another failure, the ultimate address timeout, which should have then
4608 bounced the address, did not kick in. This was a "< instead of <=" error;
4609 in most cases the first failure would have been in the next clock tick
4610 after the received time, and all would be well.
4612 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4613 being recognized when the domain list was tested by the match_domain
4614 condition in an expansion string.
4616 16. Added the ${str2b64: operator.
4618 17. Exim was always calling setrlimit() to set a large limit for the number of
4619 processes, without checking whether the existing limit was already
4620 adequate. (It did check for the limit on file descriptors.) Furthermore,
4621 errors from getrlimit() and setrlimit() were being ignored. Now they are
4622 logged to the main and panic logs, but Exim does carry on, to try to do its
4623 job under whatever limits there are.
4625 18. Imported PCRE 5.0.
4627 19. Trivial typo in log message " temporarily refused connection" (the leading
4630 20. If the log selector return_path_on_delivery was set and an address was
4631 redirected to /dev/null, the delivery process crashed because it assumed
4632 that a return path would always be set for a "successful" delivery. In this
4633 case, the whole delivery is bypassed as an optimization, and therefore no
4636 21. Internal re-arrangement: the function for sending a challenge and reading
4637 a response while authentication was assuming a zero-terminated challenge
4638 string. It's now changed to take a pointer and a length, to allow for
4639 binary data in such strings.
4641 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4643 23. Exim was not respecting finduser_retries when seeking the login of the
4644 uid under which it was called; it was always trying 10 times. (The default
4645 setting of finduser_retries is zero.) Also, it was sleeping after the final
4646 failure, which is pointless.
4648 24. Implemented tls_on_connect_ports.
4650 25. Implemented acl_smtp_predata.
4652 26. If the domain in control=submission is set empty, Exim assumes that the
4653 authenticated id is a complete email address when it generates From: or
4654 Sender: header lines.
4656 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4657 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4658 chown and chgrp in /bin and hostname in /usr/bin.
4660 28. Exim was keeping the "process log" file open after each use, just as it
4661 does for the main log. This opens the possibility of it remaining open for
4662 long periods when the USR1 signal hits a daemon. Occasional processlog
4663 errors were reported, that could have been caused by this. Anyway, it seems
4664 much more sensible not to leave this file open at all, so that is what now
4667 29. The long-running daemon process does not normally write to the log once it
4668 has entered its main loop, and it closes the log before doing so. This is
4669 so that log files can straightforwardly be renamed and moved. However,
4670 there are a couple of unusual error situations where the daemon does write
4671 log entries, and I had neglected to close the log afterwards.
4673 30. The text of an SMTP error response that was received during a remote
4674 delivery was being truncated at 512 bytes. This is too short for some of
4675 the long messages that one sometimes sees. I've increased the limit to
4678 31. It is now possible to make retry rules that apply only when a message has a
4679 specific sender, in particular, an empty sender.
4681 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4682 it possible to be selective about when SMTP synchronization is enforced.
4684 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4686 32. Implemented hosts_connection_nolog.
4688 33. Added an ACL for QUIT.
4690 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4693 35. Added mailbox_size and mailbox_filecount to appendfile.
4695 36. Added control = no_multiline_responses to ACLs.
4697 37. There was a bug in the logic of the code that waits for the clock to tick
4698 in the case where the clock went backwards by a substantial amount such
4699 that the microsecond fraction of "now" was more than the microsecond
4700 fraction of "then" (but the whole seconds number was less).
4702 38. Added support for the libradius Radius client library this is found on
4703 FreeBSD (previously only the radiusclient library was supported).
4709 1. When certain lookups returned multiple values in the form name=value, the
4710 quoting of the values was not always being done properly. Specifically:
4711 (a) If the value started with a double quote, but contained no whitespace,
4713 (b) If the value contained whitespace other than a space character (i.e.
4714 tabs or newlines or carriage returns) it was not quoted.
4715 This fix has been applied to the mysql and pgsql lookups by writing a
4716 separate quoting function and calling it from the lookup code. The fix
4717 should probably also be applied to nisplus, ibase and oracle lookups, but
4718 since I cannot test any of those, I have not disturbed their existing code.
4720 2. A hit in the callout cache for a specific address caused a log line with no
4721 reason for rejecting RCPT. Now it says "Previous (cached) callout
4722 verification failure".
4724 3. There was an off-by-one bug in the queryprogram router. An over-long
4725 return line was truncated at 256 instead of 255 characters, thereby
4726 overflowing its buffer with the terminating zero. As well as fixing this, I
4727 have increased the buffer size to 1024 (and made a note to document this).
4729 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4730 when Exim is waiting for an SMTP response from a remote server, Exim
4731 restarts its select() call on the socket, thereby resetting its timeout.
4732 This is not a problem when such interrupts are rare. Somebody set up a cron
4733 job to run exiwhat every 2 minutes, which is less than the normal select()
4734 timeout (5 or 10 minutes). This meant that the select() timeout never
4735 kicked in because it was always reset. I have fixed this by comparing the
4736 time when an interrupt arrives with the time at the start of the first call
4737 to select(). If more time than the timeout has elapsed, the interrupt is
4738 treated as a timeout.
4740 5. Some internal re-factoring in preparation for the addition of Sieve
4741 extensions (by MH). In particular, the "personal" test is moved to a
4742 separate function, and given an option for scanning Cc: and Bcc: (which is
4743 not set for Exim filters).
4745 6. When Exim created an email address using the login of the caller as the
4746 local part (e.g. when creating a From: or Sender: header line), it was not
4747 quoting the local part when it contained special characters such as @.
4749 7. Installed new OpenBSD configuration files.
4751 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4752 try to make them clearer.
4754 9. Callout options, other than the timeout value, were being ignored when
4755 verifying sender addresses in header lines. For example, when using
4757 verify = header_sender/callout=no_cache
4759 the cache was (incorrectly) being used.
4761 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4762 only the Cygwin environment.
4764 11. When return_path_on_delivery was set as a log selector, if different remote
4765 addresses in the same message used different return paths and parallel
4766 remote delivery occurred, the wrong values would sometimes be logged.
4767 (Whenever a remote delivery process finished, the return path value from
4768 the most recently started remote delivery process was logged.)
4770 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4771 lines when AUTH and/or TLS are in use. This is the "received protocol"
4772 field. Exim used to use "asmtp" for authenticated SMTP, without any
4773 indication (in the protocol name) for TLS use. Now it follows the RFC and
4774 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4775 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4776 names appear in log lines as well as in Received: header lines.
4778 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4779 extensions, and comparison tests, and to fix some bugs.
4781 14. Changes to the "personal" filter test:
4783 (1) The test was buggy in that it was just doing the equivalent of
4784 "contains" tests on header lines. For example, if a user's address was
4785 anne@some.where, the "personal" test would incorrectly be true for
4787 To: susanne@some.where
4789 This test is now done by extracting each address from the header in turn,
4790 and checking the entire address. Other tests that are part of "personal"
4791 are now done using regular expressions (for example, to check local parts
4792 of addresses in From: header lines).
4794 (2) The list of non-personal local parts in From: addresses has been
4795 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4796 taken from the Sieve specification recommendations.
4798 (3) If the message contains any header line starting with "List-" it is
4799 treated as non-personal.
4801 (4) The test for "circular" in the Subject: header line has been removed
4802 because it now seems ill-conceived.
4804 15. Minor typos in src/EDITME comments corrected.
4806 16. Installed latest exipick from John Jetmore.
4808 17. If headers_add on a router specified a text string that was too long for
4809 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4810 of string_sprintf() is now avoided.
4812 18. $message_body_size was not set (it was always zero) when running the DATA
4813 ACL and the local_scan() function.
4815 19. For the "mail" command in an Exim filter, no default was being set for
4816 the once_repeat time, causing a random time value to be used if "once" was
4817 specified. (If the value happened to be <= 0, no repeat happened.) The
4818 default is now 0s, meaning "never repeat". The "vacation" command was OK
4819 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4820 (I found it when inspecting the code).
4822 20. There is now an overall timeout for performing a callout verification. It
4823 defaults to 4 times the callout timeout, which applies to individual SMTP
4824 commands during the callout. The overall timeout applies when there is more
4825 than one host that can be tried. The timeout is checked before trying the
4826 next host. This prevents very long delays if there are a large number of
4827 hosts and all are timing out (e.g. when the network connections are timing
4828 out). The value of the overall timeout can be changed by specifying an
4829 additional sub-option for "callout", called "maxwait". For example:
4831 verify = sender/callout=5s,maxwait=20s
4833 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4834 to the end before writing, but this should make it even safer).
4836 22. Exim was forgetting that it had advertised PIPELINING for the second and
4837 subsequent messages on an SMTP connection. It was also not resetting its
4838 memory on STARTTLS and an internal HELO.
4840 23. When Exim logs an SMTP synchronization error within a session, it now
4841 records whether PIPELINING has been advertised or not.
4843 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4844 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4845 rather than long int.
4847 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4849 26. Added the never_mail option to autoreply.
4855 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4856 crash if the getsockname() call failed; this can happen if a connection is
4857 closed very soon after it is established. The problem was simply in the
4858 order in which certain operations were done, causing Exim to try to write
4859 to the SMTP stream before it had set up the file descriptor. The bug has
4860 been fixed by making things happen in the correct order.
4866 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4867 before the connection was closed, thus losing the rejection response.
4869 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4870 some early Solaris releases, but causes trouble in current releases where
4871 socklen_t is defined.
4873 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4876 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4879 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4880 that releasing the top of it at the end releases what was used for sub-
4881 expansions (unless the block got too big). However, discard this block if
4882 the first thing is a variable or header, so that we can use its block when
4883 it is dynamic (useful for very large $message_headers, for example).
4885 6. Lookups now cache *every* query, not just the most recent. A new, separate
4886 store pool is used for this. It can be recovered when all lookup caches are
4887 flushed. Lookups now release memory at the end of their result strings.
4888 This has involved some general refactoring of the lookup sources.
4890 7. Some code has been added to the store_xxx() functions to reduce the amount
4891 of flapping under certain conditions.
4893 8. log_incoming_interface used to affect only the <= reception log lines. Now
4894 it causes the local interface and port to be added to several more SMTP log
4895 lines, for example "SMTP connection from", and rejection lines.
4897 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4899 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4901 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4902 could be overwritten at the end of the current message (or the start of a
4903 new message if it was set in a HELO ACL). The value is now preserved for
4904 the duration of the SMTP connection.
4906 12. If a transport had a headers_rewrite setting, and a matching header line
4907 contained an unqualified address, that address was qualified, even if it
4908 did not match any rewriting rules. The underlying bug was that the values
4909 of the flags that permit the existence of unqualified sender and recipient
4910 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4911 non-local messages, and by -bnq for local messages) were not being
4912 preserved with the message after it was received.
4914 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4915 "next input=" as part of the text comprising the host identity instead of
4916 the correct text. The code was using the same buffer for two different
4917 strings. However, depending on which order the printing function evaluated
4918 its arguments, the bug did not always show up. Under Linux, for example, my
4919 test suite worked just fine.
4921 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4922 doesn't work with some older versions of Perl. It has been changed to "my",
4923 which in any case is probably the better facility to use.
4925 15. A really picky compiler found some instances of statements for creating
4926 error messages that either had too many or two few arguments for the format
4929 16. The size of the buffer for calls to the DNS resolver has been increased
4930 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4931 for addresses that have a lot of PTR records. This alleviates a problem; it
4932 does not fully solve it.
4934 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4935 buffer now truncates the list and logs the incident, which is the same
4936 action as happens when Exim is looking up a host name and its aliases.
4937 Previously in this situation something unpredictable would happen;
4938 sometimes it was "internal error: store_reset failed".
4940 18. If a server dropped the connection unexpectedly when an Exim client was
4941 using GnuTLS and trying to read a response, the client delivery process
4942 crashed while trying to generate an error log message.
4944 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4945 string, for example:
4947 warn message = H1: something\nH2: something
4949 the text was added as a single header line from Exim's point of view
4950 though it ended up OK in the delivered message. However, searching for the
4951 second and subsequent header lines using $h_h2: did not work. This has been
4952 fixed. Similarly, if a system filter added multiple headers in this way,
4953 the routers could not see them.
4955 20. Expanded the error message when iplsearch is called with an invalid key to
4956 suggest using net-iplsearch in a host list.
4958 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4959 ACLs are no longer actually imposed (and a message to that effect is
4962 22. If a "gecos" field in a passwd entry contained escaped characters, in
4963 particular, if it contained a \" sequence, Exim got it wrong when building
4964 a From: or a Sender: header from that name. A second bug also caused
4965 incorrect handling when an unquoted " was present following a character
4966 that needed quoting.
4968 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4969 was not being matched caselessly.
4971 24. Arranged for all hyphens in the exim.8 source to be escaped with
4974 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4975 properly. Recipient callout cache records were still being keyed to include
4976 the sender, even when use_sender was set false. This led to far more
4977 callouts that were necessary. The sender is no longer included in the key
4978 when use_sender is false.
4980 26. Added "control = submission" modifier to ACLs.
4982 27. Added the ${base62d: operator to decode base 62 numbers.
4984 28. dnsdb lookups can now access SRV records.
4986 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4987 the configuration file.
4989 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4990 (-v) mode. This makes the output for a verbose queue run more intelligible.
4992 31. Added a use_postmaster feature to recipient callouts.
4994 32. Added the $body_zerocount variable, containing the number of binary zero
4995 bytes in the message body.
4997 33. The time of last modification of the "new" subdirectory is now used as the
4998 "mailbox time last read" when there is a quota error for a maildir
5001 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5003 35. Added +ignore_unknown as a special item in host lists.
5005 36. Code for decoding IPv6 addresses in host lists is now included, even if
5006 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5007 address was recognized as an IP address, but was then not correctly decoded
5008 into binary, causing unexpected and incorrect effects when compared with
5015 1. Very minor rewording of debugging text in manualroute to say "list of
5016 hosts" instead of "hostlist".
5018 2. If verify=header_syntax was set, and a header line with an unqualified
5019 address (no domain) and a large number of spaces between the end of the
5020 name and the colon was received, the reception process suffered a buffer
5021 overflow, and (when I tested it) crashed. This was caused by some obsolete
5022 code that should have been removed. The fix is to remove it!
5024 3. When running in the test harness, delay a bit after writing a bounce
5025 message to get a bit more predictability in the log output.
5027 4. Added a call to search_tidyup() just before forking a reception process. In
5028 theory, someone could use a lookup in the expansion of smtp_accept_max_
5029 per_host which, without the tidyup, could leave open a database connection.
5031 5. Added the variables $recipient_data and $sender_data which get set from a
5032 lookup success in an ACL "recipients" or "senders" condition, or a router
5033 "senders" option, similar to $domain_data and $local_part_data.
5035 6. Moved the writing of debug_print from before to after the "senders" test
5038 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5039 problems for message scanning, either using a data ACL, or using
5040 local_scan() because the Received: header was not generated till after they
5041 were called (in order to set the time as the time of reception completion).
5042 I have revised the way this works. The header is now generated after the
5043 body is received, but before the ACL or local_scan() are called. After they
5044 are run, the timestamp in the header is updated.
5050 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5051 before starting a queue runner without re-exec. This happened only when
5052 deliver_drop_privilege was set or when the Exim user was set to root. The
5053 effect of the bug was that timeouts during subsequent deliveries caused
5054 crashes instead of being properly handled. The handler is now left at its
5055 default (and expected) setting.
5057 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5058 message, again when deliver_drop_privilege is set or Exim is run as root.
5059 The bug described in (1) was not present in this case, but the tidying up
5060 of the other signals was missing. I have made the two cases consistent.
5062 3. The ignore_target_hosts setting on a manualroute router was being ignored
5063 for hosts that were looked up using the /MX notation.
5065 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5068 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5069 operated on the sender address. After changing the $sender_address to <>
5070 for the sender address verify, Exim was re-instated it as the original
5071 (before rewriting) address, but remembering that it had rewritten it, so it
5072 wasn't rewriting it again. This bug also had the effect of breaking the
5073 sender address verification caching when the sender address was rewritten.
5075 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5076 This has been changed so that if the ip literal address matches
5077 ignore_target_hosts, the router declines.
5079 7. Added expansion conditions match_domain, match_address, and match_local_
5080 part (NOT match_host).
5082 8. The placeholder for the Received: header didn't have a length field set.
5084 9. Added code to Exim itself and to exim_lock to test for a specific race
5085 condition that could lead to file corruption when using MBX delivery. The
5086 issue is with the lockfile that is created in /tmp. If this file is removed
5087 after a process has opened it but before that process has acquired a lock,
5088 there is the potential for a second process to recreate the file and also
5089 acquire a lock. This could lead to two Exim processes writing to the file
5090 at the same time. The added code performs the same test as UW imapd; it
5091 checks after acquiring the lock that its file descriptor still refers to
5092 the same named file.
5094 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5095 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5098 11. Added the smtp_active_hostname option. If used, this will typically be made
5099 to depend on the incoming interface address. Because $interface_address is
5100 not set up until the daemon has forked a reception process, error responses
5101 that can happen earlier (such as "too many connections") no longer contain
5104 12. If an expansion in a condition on a "warn" statement fails because a lookup
5105 defers, the "warn" statement is abandoned, and the next ACL statement is
5106 processed. Previously this caused the whole ACL to be aborted.
5108 13. Added the iplsearch lookup type.
5110 14. Added ident_timeout as a log selector.
5112 15. Added tls_certificate_verified as a log selector.
5114 16. Added a global option tls_require_ciphers (compare the smtp transport
5115 option of the same name). This controls incoming TLS connections.
5117 17. I finally figured out how to make tls_require_ciphers do a similar thing
5118 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5119 before starting the TLS session.
5121 18. Tabs are now shown as \t in -bP output.
5123 19. If the log selector return_path_on_delivery was set, Exim crashed when
5124 bouncing a message because it had too many Received: header lines.
5126 20. If two routers both had headers_remove settings, and the first one included
5127 a superfluous trailing colon, the final name in the first list and the
5128 first name in the second list were incorrectly joined into one item (with a
5129 colon in the middle).
5135 1. Added -C and -D options to the exinext utility, mainly to make it easier
5136 to include in the automated testing, but these could be helpful when
5137 multiple configurations are in use.
5139 2. The exinext utility was not formatting the output nicely when there was
5140 an alternate port involved in the retry record key, nor when there was a
5141 message id as well (for retries that were specific to a specific message
5142 and a specific host). It was also confused by IPv6 addresses, because of
5143 the additional colons they contain. I have fixed the IPv4 problem, and
5144 patched it up to do a reasonable job for IPv6.
5146 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5147 delivery, the log line now contains "pipelined" if PIPELINING was used.
5149 4. An SMTP transport process used to panic and die if the bind() call to set
5150 an explicit outgoing interface failed. This has been changed; it is now
5151 treated in the same way as a connect() failure.
5153 5. A reference to $sender_host_name in the part of a conditional expansion
5154 that was being skipped was still causing a DNS lookup. This no longer
5157 6. The def: expansion condition was not recognizing references to header lines
5158 that used bh_ and bheader_.
5160 7. Added the _cache feature to named lists.
5162 8. The code for checking quota_filecount in the appendfile transport was
5163 allowing one more file than it should have been.
5165 9. For compatibility with Sendmail, the command line option
5173 and sets the incoming protocol and host name (for trusted callers). The
5174 host name and its colon can be omitted when only the protocol is to be set.
5175 Note the Exim already has two private options, -pd and -ps, that refer to
5176 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5177 "s", but I don't think that's a major issue.
5179 10. A number of refactoring changes to the code, none of which should affect
5182 (a) The number of logging options was getting close to filling up the
5183 32-bit word that was used as a bit map. I have split them into two classes:
5184 those that are passed in the argument to log_write(), and those that are
5185 only ever tested independently outside of that function. These are now in
5186 separate 32-bit words, so there is plenty of room for expansion again.
5187 There is no change in the user interface or the logging behaviour.
5189 (b) When building, for example, log lines, the code previously used a
5190 macro that called string_cat() twice, in order to add two strings. This is
5191 not really sufficiently general. Furthermore, there was one instance where
5192 it was actually wrong because one of the argument was used twice, and in
5193 one call a function was used. (As it happened, calling the function twice
5194 did not affect the overall behaviour.) The macro has been replaced by a
5195 function that can join an arbitrary number of extra strings onto a growing
5198 (c) The code for expansion conditions now uses a table and a binary chop
5199 instead of a serial search (which was left over from when there were very
5200 few conditions). Also, it now recognizes conditions like "pam" even when
5201 the relevant support is not compiled in: a suitably worded error message is
5202 given if an attempt is made to use such a condition.
5204 11. Added ${time_interval:xxxxx}.
5206 12. A bug was causing one of the ddress fields not to be passed back correctly
5207 from remote delivery subprocesses. The field in question was not being
5208 subsequently used, so this caused to problems in practice.
5210 13. Added new log selectors queue_time and deliver_time.
5212 14. Might have fixed a bug in maildirsizefile handling that threw up
5213 "unexpected character" debug warnings, and recalculated the data
5214 unnecessarily. In any case, I expanded the warning message to give more
5217 15. Added the message "Restricted characters in address" to the statements in
5218 the default ACL that block characters like @ and % in local parts.
5220 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5221 Three changes have been made:
5223 (a) There was a serious bug; a negative response to MAIL caused the whole
5224 recipient domain to be cached as invalid, thereby blocking all messages
5225 to all local parts at the same domain, from all senders. This bug has
5226 been fixed. The domain is no longer cached after a negative response to
5227 MAIL if the sender used is not empty.
5229 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5232 (c) A new callout option, "use_sender" has been added for people who want
5233 the modified behaviour.
5239 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5242 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5243 indeed breaks things for older releases.
5245 3. Added additional logging to the case where there is a problem reading data
5246 from a filter that is running in a subprocess using a pipe, in order to
5247 try to track down a specific problem.
5249 4. Testing facility fudge: when running in the test harness and attempting
5250 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5251 getting "No route to host". Convert this to a timeout.
5253 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5256 6. Some OS don't have socklen_t but use size_t instead. This affects the
5257 fifth argument of getsockopt() amongst other things. This is now
5258 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5259 can be set for individual OS. I have set it for SunOS5, OSF1, and
5260 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5261 some earlier ones do not.
5263 7. Change 4.30/15 was not doing the test caselessly.
5265 8. The standard form for an IPv6 address literal was being rejected by address
5266 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5267 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5268 this, as well as the form without the "IPv6" on the front (but only when
5269 address literals are enabled, of course).
5271 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5273 10. Exim crashed if a message with an empty sender address specified by -f
5274 encountered a router with an errors_to setting. This could be provoked only
5275 by a command such as
5279 where an empty string was supplied; "<>" did not hit this bug.
5281 11. Installed PCRE release 4.5.
5283 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5284 remained set. It is now erased.
5286 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5287 times from message ids (which are base 36 rather than the normal 62).
5289 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5290 were being counted as actual protocol errors, and logged if the log
5291 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5292 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5293 and DATA following a set of rejected RCPTs do not count as protocol errors.
5294 In other words, Exim assumes they were pipelined, though this may not
5295 actually be the case. Of course, in all cases the client gets an
5296 appropriate error code.
5298 15. If a lookup fails in an ACL condition, a message about the failure may
5299 be available; it is used if testing the ACL cannot continue, because most
5300 such messages specify what the cause of the deferral is. However, some
5301 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5302 that caused an old message to be retained and used if a later statement
5303 caused a defer, replacing the real cause of the deferral.
5305 16. If an IP address had so many PTR records that the DNS lookup buffer
5306 was not large enough to hold them, Exim could crash while trying to process
5307 the truncated data. It now detects and logs this case.
5309 17. Further to 4.21/58, another change has been made: if (and only if) the
5310 first line of a message (the first header line) ends with CRLF, a bare LF
5311 in a subsequent header line has a space inserted after it, so as not to
5312 terminate the header.
5314 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5315 unnecessarily, used atoi() instead of strtol(), and didn't check the
5316 termination when getting file sizes from file names by regex.
5318 19. Completely re-implemented the support for maildirsize files, in the light
5319 of a number of problems with the previous contributed implementation
5320 (4.30/29). In particular:
5322 . If the quota is zero, the maildirsize file is maintained, but no quota is
5325 . If the maildir directory does not exist, it is created before any attempt
5326 to write a maildirsize file.
5328 . The quota value in the file is just a cache; if the quota is changed in
5329 the transport, the new value overrides.
5331 . A regular expression is available for excluding directories from the
5334 20. The autoreply transport checks the characters in options that define the
5335 message's headers; it allows continued headers, but it was checking with
5336 isspace() after an embedded newline instead of explicitly looking for a
5339 21. If all the "regular" hosts to which an address was routed had passed their
5340 expiry times, and had not reached their retry times, the address was
5341 bounced, even if fallback hosts were defined. Now Exim should go on to try
5344 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5345 equivalent code in the SMTP transport. Some hosts send humungous responses
5346 to HELO/EHLO, more than 1024 it seems.
5348 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5349 gives compiler warnings in some environments. I've now done it "properly",
5352 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5353 (because of problems with the built-in one) was declared to return uschar *
5354 instead of char *, causing compiler failure.
5356 25. Fixed a file descriptor leak when processing alias/forward files.
5358 26. Fixed a minor format string issue in dbfn.c.
5360 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5362 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5363 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5366 29. When a sender address is verified, it is cached, to save repeating the test
5367 when there is more than one recipient in a message. However, when the
5368 verification involves a callout, it is possible for different callout
5369 options to be set for different recipients. It is too complicated to keep
5370 track of this in the cache, so now Exim always runs a verification when a
5371 callout is required, relying on the callout cache for the optimization.
5372 The overhead is duplication of the address routing, but this should not be
5375 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5376 to be verified with callout=postmaster, and the main callout worked but the
5377 postmaster check failed, the verification correctly failed. However, if a
5378 subsequent RCPT command asked for sender verification *without* the
5379 postmaster check, incorrect caching caused this verification also to fail,
5382 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5383 it was not caching the DNS options (qualify_single, search_parents) that
5384 were used when the lookup failed. A subsequent lookup with different
5385 options therefore always gave the same answer, though there were cases
5386 where it should not have. (Example: a "domains = !$mx_any" option on a
5387 dnslookup router: the "domains" option is always processed without any
5388 widening, but the router might have qualify_single set.) Now Exim uses the
5389 cached value only when the same options are set.
5391 32. Added John Jetmore's "exipick" utility to the distribution.
5393 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5394 than a timeout (e.g. a certificate is required, and is not provided), an
5395 Exim server now closes the connection immediately. Previously it waited for
5396 the client to close - but if the client is SSL, it seems that they each
5397 wait for each other, leading to a delay before one of them times out.
5399 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5400 maintained 0.8.x compatibility because I don't think many are using it, and
5401 it is clearly obsolete.
5403 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5406 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5407 client certificate was expired. A simple patch fixes this, though I don't
5408 understand the full logic of why the verify callback is called multiple
5411 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5412 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5413 which causes problems with some clients (such as the Certicom SSL Plus
5414 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5415 disables the coutermeasure allowing Eudora to connect."
5417 38. Exim was not checking that a write() to a log file succeeded. This could
5418 lead to Bad Things if a log got too big, in particular if it hit a file
5419 size limit. Exim now panics and dies if it cannot write to a log file, just
5420 as it does if it cannot open a log file.
5422 39. Modified OS/Makefile-Linux so that it now contains
5424 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5426 The two -D definitions ensure that Exim is compiled with large file
5427 support, which makes it possible to handle log files that are bigger than
5430 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5431 instance) a domain was checked against a named list that involved a lookup,
5432 causing $domain_data to be set, then another domain was checked against the
5433 same list, then the first domain was re-checked, the value of $domain_data
5434 after the final check could be wrong. In particular, if the second check
5435 failed, it could be set empty. This bug probably also applied to
5438 41. The strip_trailing_dot option was not being applied to the address given
5439 with the -f command-line option.
5441 42. The code for reading a message's header from the spool was incrementing
5442 $received_count, but never initializing it. This meant that the value was
5443 incorrect (doubled) while delivering a message in the same process in which
5444 it was received. In the most common configuration of Exim, this never
5445 happens - a fresh exec is done - but it can happen when
5446 deliver_drop_privilege is set.
5448 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5449 it now includes up to 150 characters of the unexpected data in the log
5452 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5453 and building data strings. The size of both of these buffers was 10 000
5454 bytes - far larger than anybody would *ever* want, thought I. Needless to
5455 say, somebody hit the limit. I have increased the maximum line length to
5456 20 000 and the maximum data length of concatenated lines to 100 000. I have
5457 also fixed two bugs, because there was no checking on these buffers. Tsk,
5458 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5459 buffer is too small.
5461 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5462 lsearch lookups. Now it does.
5464 46. When parsing a route_list item in a manualroute router, a fixed-length
5465 buffer was used for the list of hosts. I made this 1024 bytes long,
5466 thinking that nobody would ever have a list of hosts that long. Wrong.
5467 Somebody had a whole pile of complicated expansion conditions, and the
5468 string was silently truncated, leading to an expansion error. It turns out
5469 that it is easier to change to an unlimited length (owing to other changes
5470 that have happened since this code was originally written) than to build
5471 structure for giving a limitation error. The length of the item that
5472 expands into the list of hosts is now unlimited.
5474 47. The lsearch lookup could not handle data where the length of text line was
5475 more than 4095 characters. Such lines were truncated, leading to shortened
5476 data being returned. It should now handle lines of any length.
5478 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5479 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5482 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5483 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5484 no longer happens in comment lines. A list of replacements is now placed
5485 at the head of all of the source files, except those whose only change is
5486 to replace PERL_COMMAND in the very first #! line.
5488 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5489 messages on the queue, with a bottom-up merge sort, using code contributed
5490 by Michael Haardt. This should make operations like -bp somewhat faster on
5491 large queues. It won't affect queue runners, except when queue_run_in_order
5494 51. Installed eximstats 1.31 in the distribution.
5496 52. Added support for SRV lookups to the dnslookup router.
5498 53. If an ACL referred to $message_body or $message_body_end, the value was not
5499 reset for any messages that followed in the same SMTP session.
5501 54. The store-handling optimization for building very long strings was not
5502 differentiating between the different store pools. I don't think this
5503 actually made any difference in practice, but I've tidied it.
5505 55. While running the routers to verify a sender address, $sender_address
5506 was still set to the sender address. This is wrong, because when routing to
5507 send a bounce to the sender, it would be empty. Therefore, I have changed
5508 it so that, while verifying a sender address, $sender_address is set to <>.
5509 (There is no change to what happens when verifying a recipient address.)
5511 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5512 target A or AAAA records (if not already returned) without resetting the
5513 qualify_single or search_parents options of the DNS resolver. These are
5514 inappropriate in this case because the targets of MX and SRV records must
5515 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5516 target that, when qualified, matched something in the local domain. These
5517 two options are now turned off when doing these lookups.
5519 57. It seems that at least some releases of Reiserfs (which does not have the
5520 concept of a fixed number of inodes) returns zero and not -1 for the
5521 number of available inodes. This interacted badly with check_spool_inodes,
5522 which assumed that -1 was the "no such thing" setting. What I have done is
5523 to check that the total number of inodes is greater than zero before doing
5524 the test of how many are available.
5526 58. When a "warn" ACL statement has a log_message modifier, the message is
5527 remembered, and not repeated. This is to avoid a lot of repetition when a
5528 message has many recipients that cause the same warning to be written.
5529 Howewer, Exim was preserving the list of already written lines for an
5530 entire SMTP session, which doesn't seem right. The memory is now reset if a
5531 new message is started.
5533 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5534 debugging output unless log_rewrite was also set.
5536 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5537 of (int)(handle) when we know that handle contains (void *)(-1).
5539 61. The Exim daemon panic-logs an error return when it closes the incoming
5540 connection. However "connection reset by peer" seems to be common, and
5541 isn't really an error worthy of noting specially, so that particular error
5544 62. When Exim is trying to find all the local interfaces, it used to panic and
5545 die if the ioctl to get the interface flags failed. However, it seems that
5546 on at least one OS (Solaris 9) it is possible to have an interface that is
5547 included in the list of interfaces, but for which you get a failure error
5548 for this call. This happens when the interface is not "plumbed" into a
5549 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5550 failure of the "get flags" call assumes that the interface is down.
5552 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5553 makes life easier for people who are doing arithmetic on fields extracted
5554 from dates, where you often get leading zeros that should not be
5555 interpreted as octal.
5557 64. Added qualify_domain to the redirect router, to override the global
5560 65. If a pathologically long header line contained very many addresses (the
5561 report of this problem mentioned 10 000) and each of them was rewritten,
5562 Exim could use up a very large amount of memory. (It kept on making new
5563 copies of the header line as it rewrote, and never released the old ones.)
5564 At the expense of a bit more processing, the header rewriting function has
5565 been changed so that it no longer eats memory in this way.
5567 66. The generation of the Received: header has been moved from the time that a
5568 message starts to be received, to the time that it finishes. The timestamp
5569 in the Received: header should now be very close to that of the <= log
5570 line. There are two side-effects of this change:
5572 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5573 logged header lines no longer include the local Received: line, because
5574 it has not yet been created. The same applies to a copy of the message
5575 that is returned to a non-SMTP sender when a message is rejected.
5577 (b) When a filter file is tested using -bf, no additional Received: header
5578 is added to the test message. After some thought, I decided that this
5581 This change does not affect the value of $received_for. It is still set
5582 after address rewriting, but before local_scan() is called.
5584 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5586 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5587 gave an unhelpful panic error message, and a defer error. I have managed to
5588 change this behaviour so that it now rejects any supplied certificate,
5589 which seems right, as the list of acceptable certificates is empty.
5591 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5592 gave an unhelpful defer error. I have not managed to make this reject any
5593 supplied certificates, but the error message it gives is "no certificate
5594 supplied", which is not helpful.
5596 70. exigrep's output now also includes lines that are not associated with any
5597 message, but which match the given pattern. Implemented by a patch from
5598 Martin Sluka, which also tidied up the Perl a bit.
5600 71. Recipient callout verification, like sender verification, was using <> in
5601 the MAIL FROM command. This isn't really the right thing, since the actual
5602 sender may affect whether the remote host accepts the recipient or not. I
5603 have changed it to use the actual sender in the callout; this means that
5604 the cache record is now keyed on a recipient/sender pair, not just the
5605 recipient address. There doesn't seem to be a real danger of callout loops,
5606 since a callout by the remote host to check the sender would use <>.
5607 [SEE ABOVE: changed after hitting problems.]
5609 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5610 temporary errors. However, in the case of such a code being given after
5611 the end of a data transmission (i.e. after ".") Exim was failing to write
5612 a retry record for the message. (Yes, there was some broken host that was
5613 actually sending 8xx at this point.)
5615 73. An unknown lookup type in a host list could cause Exim to panic-die when
5616 the list was checked. (An example that provoked this was putting <; in the
5617 middle of a list instead of at the start.) If this happened during a DATA
5618 ACL check, a -D file could be left lying around. This kind of configuration
5619 error no longer causes Exim to die; instead it causes a defer errror. The
5620 incident is still logged to the main and panic logs.
5622 74. Buglet left over from Exim 3 conversion. The message "too many messages
5623 in one connection" was written to the rejectlog but not the mainlog, except
5624 when address rewriting (yes!) was being logged.
5626 75. Added write_rejectlog option.
5628 76. When a system filter was run not as root (that is, when system_filter_user
5629 was set), the values of the $n variables were not being returned to the
5630 main process; thus, they were not subsequently available in the $sn
5633 77. Added +return_path_on_delivery log selector.
5635 78. A connection timeout was being treated differently from recipients deferred
5636 when testing hosts_max_try with a message that was older than the host's
5637 retry timeout. (The host should not be counted, thus allowing all hosts to
5638 be tried at least once before bouncing.) This may have been the cause of an
5639 occasionally reported bug whereby a message would remain on the queue
5640 longer than the retry timeout, but would be bounced if a delivery was
5641 forced. I say "may" because I never totally pinned down the problem;
5642 setting up timeout/retry tests is difficult. See also the next item.
5644 79. The ultimate address timeout was not being applied to errors that involved
5645 a combination of host plus message (for example, a timeout on a MAIL
5646 command). When an address resolved to a number of possible hosts, and they
5647 were not all tried for each delivery (e.g. because of hosts_max_try), a
5648 message could remain on the queue longer than the retry timeout.
5650 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5653 81. Fixed an obscure SMTP outgoing bug which required at least the following
5654 conditions: (a) there was another message waiting for the same server;
5655 (b) the server returned 5xx to all RCPT commands in the first message so
5656 that the message was not completed; (c) the server dropped the connection
5657 or gave a negative response to the RSET that Exim sends to abort the
5658 transaction. The observed case was a dropped connection after DATA that had
5659 been sent in pipelining mode. That is, the server had advertised PIPELINING
5660 but was not implementing it correctly. The effect of the bug was incorrect
5661 behaviour, such as trying another host, and this could lead to a crash.
5667 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5668 and daemon.c were passed as pointers to ints; they should have been
5669 pointers to socklen_t variables (which are typically unsigned ints).
5671 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5674 3. Fixed a really odd bug that affected only the testing scheme; patching a
5675 certain fixed string in the binary changed the value of another string that
5676 happened to be identical to the end of the original first string.
5678 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5679 name", it returns that address as the IP address. On some operating
5680 systems (e.g. Solaris), it also passes back the IP address string as the
5681 "host name". However, on others (e.g. Linux), it passes back an empty
5682 string. Exim wasn't checking for this, and was changing the host name to an
5683 empty string, assuming it had been canonicized.
5685 5. Although rare, it is permitted to have more than one PTR record for a given
5686 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5687 all the names associated with an address, because they do in Solaris.
5688 However, it seems that they do not in Linux for data that comes from the
5689 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5690 I found this out when I moved to a new Linux workstation and tried to run
5691 the Exim test suite.
5693 To get round this problem I have changed the code so that it now does its
5694 own call to the DNS to look up PTR records when searching for a host name.
5695 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5696 addresses that are only in /etc/hosts are still found.
5698 This behaviour is, however, controlled by an option called host_lookup_
5699 order, which defaults to "bydns:byaddr". If people want to use the other
5700 order, or indeed, just use one or the other means of lookup, they can
5701 specify it in this variable.
5703 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5704 some operating systems, this comes back from gethostbyaddr() as an empty
5705 string, and this is what Exim used to test for. However, it seems that in
5706 other systems, "." is yielded. Exim now tests for this case too.
5708 7. The values of check_spool_space and check_log_space are now held internally
5709 as a number of kilobytes instead of an absolute number of bytes. If a
5710 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5711 kilobyte. This means that much larger values can be stored.
5713 8. Exim monitor: an attempt to get the action menu when not actually pointing
5714 at a message produces an empty menu entitled "No message selected". This
5715 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5716 no entries in it ("Shell widget menu has zero width and/or height"). So I
5717 have added a single, blank menu entry in this case.
5719 9. Added ${quote_local_part.
5721 10. MIME decoding is now applied to the contents of Subject: header lines when
5724 11. Now that a reference to $sender_host_address automatically causes a reverse
5725 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5726 host lookup before query-style lookups in lists that might use this
5727 variable. This has therefore been abolished, and the "net-" prefix is no
5728 longer necessary for query-style lookups.
5730 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5731 have been a typo for LFLAGS, so it has been changed.
5733 13. The install script calls Exim with "-C /dev/null" in order to find the
5734 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5735 to be output. Howeve, since Exim outputs its version number before the
5736 error, it didn't break the script. It just looked ugly. I fixed this by
5737 always allowing "-C /dev/null" if the caller is root.
5739 14. Ignore overlarge ACL variable number when reading spool file - insurance
5740 against a later release with more variables having written the file.
5742 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5743 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5744 this, as well as the form without the "IPv6" on the front.
5746 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5747 OS/Makefile-Darwin file.
5749 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5750 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5752 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5753 variable. However, some OS use ints for inodes. I've added cast to long int
5754 to get rid of the compiler warning.
5756 19. I had forgotten to lock out "/../" in configuration file names when
5757 ALT_CONFIG_PREFIX was set.
5759 20. Routers used for verification do not need to specify transports. However,
5760 if such a router generated a host list, and callout was configured, Exim
5761 crashed, because it could not find a port number from the (non-existent)
5762 transport. It now assumes port 25 in this circumstance.
5764 21. Added the -t option to exigrep.
5766 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5767 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5768 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5769 them is set, LOOKUP_LSEARCH is forced.
5771 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5773 24. Added sender and host information to the "rejected by local_scan()" log
5774 line; previously there was no indication of these.
5776 25. Added .include_if_exists.
5778 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5779 Linux. It turns out that not all file systems support this. Apparently some
5780 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5781 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5782 not supported on the file descriptor, is now ignored when Exim is trying to
5783 sync a directory. This applies only to Linux.
5785 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5787 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5788 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5791 29. Added support for maildirsize files from supplied patch (modified a bit).
5793 30. The use of :fail: followed by an empty string could lead Exim to respond to
5794 sender verification failures with (e.g.):
5796 550 Verification failed for <xxx>
5797 550 Sender verify failed
5799 where the first response line was missing the '-' that indicates it is not
5800 the final line of the response.
5802 31. The loop for finding the name of the user that called Exim had a hardwired
5803 limit of 10; it now uses the value of finduser_retries, which is used for
5804 all other user lookups.
5806 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5809 33. Exim was neglecting to zero errno before one call of strtol() when
5810 expanding a string and expecting an integer value. On some systems this
5811 resulted in spurious "integer overflow" errors. Also, it was casting the
5812 result into an int without checking.
5814 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5815 did not work. The code looks as if it has *never* worked, though it appears
5816 to have been documented since at least releast 1.62. I have made it work.
5818 35. The "timeout_DNS" error in retry rules, also documented since at least
5819 1.62, also never worked. As it isn't clear exactly what this means, and
5820 clearly it isn't a major issue, I have abolished the feature by treating it
5821 as "timeout", and writing a warning to the main and panic logs.
5823 36. The display of retry rules for -brt wasn't always showing the error code
5826 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5827 timeout_connect_A, timeout_connect_MX.
5829 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5830 to the empty sender.
5832 39. The daemon was not analysing the content of -oX till after it had closed
5833 stderr and disconnected from the controlling terminal. This meant that any
5834 syntax errors were only noted on the panic log, and the return code from
5835 the command was 0. By re-arranging the code a little, I've made the
5836 decoding happen first, so such errors now appear on stderr, and the return
5837 code is 1. However, the actual setting up of the sockets still happens in
5838 the disconnected process, so errors there are still only recorded on the
5841 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5842 connections (as happens on some IP stacks) was logged at start up time as
5843 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5844 it from "IPv6 and IPv4", which means that two separate sockets are being
5847 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5848 says whether AF_INET or AF_INET6 was passed as an argument.
5850 42. Exiwhat output was messed up when time zones were included in log
5853 43. Exiwhat now gives more information about the daemon's listening ports,
5854 and whether -tls-on-connect was used.
5856 44. The "port" option of the smtp transport is now expanded.
5858 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5859 silently ignored. Now an error message is written to the main and panic
5862 46. There's a new ACL modifier called "logwrite" which writes to a log file
5863 as soon as it is encountered.
5865 47. Added $local_user_uid and $local_user_gid at routing time.
5867 48. Exim crashed when trying to verify a sender address that was being
5870 49. Exim was recognizing only a space character after ".include". It now also
5871 recognizes a tab character.
5873 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5874 extracting the relevant information from the specification. The man page no
5875 longer contains scrambled data for the -d option, and I've added a section
5876 at the front about calling Exim under different names.
5878 51. Added "extra_headers" argument to the "mail" command in filter files.
5880 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5883 53. Installed eximstats 1.29.
5885 54. Added transport_filter_timeout as a generic transport option.
5887 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5888 Cc: header lines. This was required by RFC 822, but it not required by RFC
5891 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5892 incoming messages that did not have them. Now it does so only if the
5893 message originates locally, that is, if there is no associated remote host
5894 address. When Resent- header lines are present, this applies to the Resent-
5895 lines rather than the non-Resent- lines.
5897 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5898 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5900 58. Messages for configuration errors now include the name of the main
5901 configuration file - useful now that there may be more than one file in a
5902 list (.included file names were always shown).
5904 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5905 for those rare installations that do not start the daemon as root or run it
5906 setuid root. I've cut out the call to initgroups() if the daemon is not
5909 60. The Exim user and group can now be bound into the binary as text strings
5910 that are looked up at the start of Exim's processing.
5912 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5914 62. Added $mailstore_basename variable.
5916 63. Installed patch to sieve.c from Michael Haardt.
5918 64. When Exim failed to open the panic log after failing to open the main log,
5919 the original message it was trying to log was written to stderr and debug
5920 output, but if they were not available (the usual case in production), it
5921 was lost. Now it is written to syslog before the two lines that record the
5922 failures to open the logs.
5924 65. Users' Exim filters run in subprocesses under the user's uid. It is
5925 possible for a "deliver" command or an alias in a "personal" command to
5926 provoke an address rewrite. If logging of address rewriting is configured,
5927 this fails because the process is not running as root or exim. There may be
5928 a better way of dealing with this, but for the moment (because 4.30 needs
5929 to be released), I have disabled address rewrite logging when running a
5930 filter in a non-root, non-exim process.
5936 1. The buildconfig auxiliary program wasn't quoting the value set for
5937 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5938 not defined. This bug was masked in 4.22 by the effect that was fixed in
5941 2. Some messages that were rejected after a message id was allocated were
5942 shown as "incomplete" by exigrep. It no longer does this for messages that
5943 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5945 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5946 have allow_domain_literals set, the ID did not get logged in the <= line.
5947 Domain literals are now always recognized in Message-ID: header lines.
5949 4. The first argument for a ${extract expansion item is the key name or field
5950 number. Leading and trailing spaces in this item were not being ignored,
5951 causing some misleading effects.
5953 5. When deliver_drop_privilege was set, single queue runner processes started
5954 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5955 same command in the process it spins off) were not dropping privilege.
5957 6. When the daemon running as "exim" started a queue runner, it always
5958 re-executed Exim in the spun-off process. This is a waste of effort when
5959 deliver_drop_privilege is set. The new process now just calls the
5960 queue-runner function directly.
5966 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5969 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5970 ignored. Though the use of -oP was forcing the writing of a pid file, it
5971 was always written to the default place.
5973 3. If the message "no IP address found for host xxxx" is generated during
5974 incoming verification, it is now followed by identification of the incoming
5975 connection (so you can more easily find what provoked it).
5977 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5979 5. Added some features to "harden" Exim a bit more against certain attacks:
5981 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5982 be put in Local/Makefile. This is like the never_users runtime option,
5983 but it cannot be overridden. The default setting is "root".
5985 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5986 prefix string with which any file named in a -C command line option
5989 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5990 is retained for -C and -D only if the caller of Exim is root. Without
5991 it, the exim user may also use -C and -D and retain privilege.
5993 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5994 command line option is disabled.
5996 6. Macro names set by the -D option must start with an upper case letter, just
5997 like macro names defined in the configuration file.
5999 7. Added "dereference=" facility to LDAP.
6001 8. Two instances of the typo "uknown" in the source files are fixed.
6003 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6004 the Configure-Makefile script screwed up while processing it.
6006 10. Incorporated PCRE 4.4.
6008 11. The SMTP synchronization check was not operating right at the start of an
6009 SMTP session. For example, it could not catch a HELO sent before the client
6010 waited for the greeting. There is now a check for outstanding input at the
6011 point when the greeting is written. Because of the duplex, asynchronous
6012 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6013 way, but not yet received, when the check is performed.
6015 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6016 on TCP/IP sockets, because this apparently causes some broken clients to
6019 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6020 unchanged) from the Cygwin maintainer.
6022 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6023 was supported instead of testing for mbx. Effectively a typo.
6025 15. The spa authenticator server code was not checking that the input it
6026 received was valid base64.
6028 16. The debug output line for the "set" modifier in ACLs was not showing the
6029 name of the variable that was being set.
6031 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6033 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6034 lookup on its own. Something else was needed to trigger the lookup. For
6035 example, a match in host_lookup or the need for a host name in a host list.
6036 Now, if $sender_host_name is referenced and the host name has not yet been
6037 looked up, a lookup is performed. If the lookup fails, the variable remains
6038 empty, and $host_lookup_failed is set to "1".
6040 19. Added "eqi" as a case-independent comparison operator.
6042 20. The saslauthd authentication condition could segfault if neither service
6043 nor realm was specified.
6045 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6046 error message that was logged was misleading, and incoming SMTP
6047 connections were dropped. The message is now more accurate, and temporary
6048 errors are given to SMTP connections.
6050 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6051 (except RSET) with a 421 error, until QUIT is received. However, it was
6052 failing to send a response to QUIT.
6054 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6055 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6056 ACL is now run after the helo_try_verify_hosts code.
6058 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6059 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6060 case-independent, so other case variants are also recognized). Apparently
6061 some systems use these upper case variants.
6063 25. If more than two messages were waiting for the same host, and a transport
6064 filter was specified for the transport, Exim sent two messages over the
6065 same TCP/IP connection, and then failed with "socket operation on non-
6066 socket" when it tried to send the third.
6068 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6070 27. The extern definition of crypt16() in expand.c was not being excluded when
6071 the OS had its own crypt16() function.
6073 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6074 as a preferred synonym for return_size_limit, both as an option and as an
6077 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6079 30. Changed the default configuration ACL to relax the local part checking rule
6080 for addresses that are not in any local domains. For these addresses,
6081 slashes and pipe symbols are allowed within local parts, but the sequence
6082 /../ is explicitly forbidden.
6084 31. SPA server authentication was not clearing the challenge buffer before
6087 32. log_message in a "warn" ACL statement was writing to the reject log as
6088 well as to the main log, which contradicts the documentation and doesn't
6089 seem right (because no rejection is happening). So I have stopped it.
6091 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6092 However, I am unable to do any testing of this.
6094 34. Fixed an infelicity in the appendfile transport. When checking directories
6095 for a mailbox, to see if any needed to be created, it was accidentally
6096 using path names with one or more superfluous leading slashes; tracing
6097 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6099 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6100 contents are added to the log line that is written for every discarded
6101 recipient. (Previously a log_message setting was ignored.)
6103 36. The ${quote: operator now quotes the string if it is empty.
6105 37. The install script runs exim in order to find its version number. If for
6106 some reason other than non-existence or emptiness, which it checks, it
6107 could not run './exim', it was installing it with an empty version number,
6108 i.e. as "exim-". This error state is now caught, and the installation is
6111 38. An argument was missing from the function that creates an error message
6112 when Exim fails to connect to the socket for saslauthd authentication.
6113 This could cause Exim to crash, or give a corrupted message.
6115 39. Added isip, isip4, and isip6 to ${if conditions.
6117 40. The ACL variables $acl_xx are now saved with the message, and can be
6118 accessed later in routers, transports, and filters.
6120 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6121 strings in the file are not string-expanded.
6123 42. If a MAIL command specified a SIZE value that was too large to fit into an
6124 int variable, the check against message_size_limit failed. Such values are
6125 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6126 day this will have to be increased, but I don't think I want to be around
6127 when emails are that large.
6134 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6135 iconv() is not standard in FreeBSD.
6137 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6138 IPv6 enabled. The observed symptom was a segmentation fault on return from
6139 the function os_common_find_running_interfaces() in src/os.c.
6141 3. In the check_special_case() function in daemon.c I had used "errno" as an
6142 argument name, which causes warnings on some systems. This was basically a
6143 typo, since it was named "eno" in the comments!
6145 4. The code that waits for the clock to tick (at a resolution of some fraction
6146 of a second) so as to ensure message-id uniqueness was always waiting for
6147 at least one whole tick, when it could have waited for less. [This is
6148 almost certainly not relevant at current processor speeds, where it is
6149 unlikely to ever wait at all. But we try to future-proof.]
6151 5. The function that sleeps for a time interval that includes fractions of a
6152 second contained a race. It did not block SIGALRM between setting the
6153 timer, and suspending (a couple of lines later). If the interval was short
6154 and the sigsuspend() was delayed until after it had expired, the suspension
6155 never ended. On busy systems this could lead to processes getting stuck for
6158 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6159 process, before it forks any delivery processes. The open lookup caching
6160 mechanism meant that the open file or database connection was passed into
6161 the delivery process. The problem was that delivery processes always tidy
6162 up cached lookup data. This could cause a problem for the next delivery
6163 process started by the queue runner, because the external queue runner
6164 process does not know about the closure. So the next delivery process
6165 still has data in the lookup cache. In the case of a file lookup, there was
6166 no problem because closing a file descriptor in a subprocess doesn't affect
6167 the parent. However, if the lookup was caching a connection to a database,
6168 the connection was closed, and the second delivery process was likely to
6169 see errors such as "PGSQL: query failed: server closed the connection
6170 unexpectedly". The problem has been fixed by closing all cached lookups
6171 in a queue runner before running a delivery process.
6173 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6174 seem to have the "const" qualifier which it has on other OS. I've
6177 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6178 *of the same type* (client or server) with the same public name that an
6179 error should be diagnosed.
6181 9. When Exim looked up a host name for an IP address, but failed to find the
6182 original IP address when looking up the host name (a safety check), it
6183 output the message "<ip address> does not match any IP for NULL", which was
6184 confusing, to say the least. The bug was that the host name should have
6185 appeared instead of "NULL".
6187 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6188 user that is built into the binary, and the -C or -D options is used, root
6189 privilege is dropped before the configuration file is read. In addition,
6190 logging is switched to stderr instead of the normal log files. If the
6191 configuration then re-defines the Exim user, the unprivileged environment
6192 is probably not what is expected, so Exim logs a panic warning message (but
6195 However, if deliver_drop_privilege is set, the unprivileged state may well
6196 be exactly what is intended, so the warning has been cut out in that case,
6197 and Exim is allowed to try to write to its normal log files.
6203 1. smtp_return_error_details was not giving details for temporary sender
6204 or receiver verification errors.
6206 2. Diagnose a configuration error if two authenticators have the same public
6209 3. Exim used not to create the message log file for a message until the first
6210 delivery attempt. This could be confusing when incoming messages were held
6211 for policy or load reasons. The message log file is now created at the time
6212 the message is received, and an initial "Received" line is written to it.
6214 4. The automatically generated man page for command line options had a minor
6215 bug that caused no ill effects; however, a more serious problem was that
6216 the procedure for building the man page automatically didn't always
6217 operate. Consequently, release 4.20 contains an out-of-date version. This
6218 shouldn't happen again.
6220 5. When building Exim with embedded Perl support, the script that builds the
6221 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6222 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6224 6. The freeze_tell option was not being used for messages that were frozen on
6225 arrival, either by an ACL or by local_scan().
6227 7. Added the smtp_incomplete_transaction log selector.
6229 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6230 was accepting AUTH without a new EHLO.
6232 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6233 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6236 10. Logging of TCP/IP connections (when configured) now happens in the main
6237 daemon process instead of the child process, so that the TCP/IP connection
6238 count is more accurate (but it can never be perfect).
6240 11. The use of "drop" in a nested ACL was not being handled correctly in the
6241 outer ACL. Now, if condition failure induced by the nested "drop" causes
6242 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6243 or "require"), the connection is dropped.
6245 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6246 that yield "discard" can now be used with an "accept" or a "discard" verb,
6247 but an error is generated for any others (because I can't see a useful way
6248 to define what should happen).
6250 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6251 are now processed in the same way as lines in the Exim configuration file.
6252 In particular, continuation lines are supported.
6254 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6256 15. Added -ti meaning -t -i.
6258 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6259 domains, and warn by logging if others are found.
6261 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6262 returned from GIFCONF when Exim is trying to find the list of interfaces on
6263 a host. The code in os.c has been modified to copy each ifreq to an aligned
6264 structure in all cases.
6266 Also, in some cases, the returned ifreq's were being copied to a 'struct
6267 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6268 means the last couple of bytes of an IPv6 address could be chopped if the
6269 ifreq contained only a normal sockaddr (14 bytes storage).
6271 18. Named domain lists were not supported in the hosts_treat_as_local option.
6272 An entry such as +xxxx was not recognized, and was treated as a literal
6275 19. Ensure that header lines added by a DATA ACL are included in the reject log
6276 if the ACL subsequently rejects the message.
6278 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6279 MD5 (which is deprecated).
6281 21. When testing a filter file using -bf, Exim was writing a message when it
6282 took the sender from a "From " line in the message, but it was not doing so
6283 when it took $return_path from a Return-Path: header line. It now does.
6285 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6286 with a valid header line field name (a series of printing characters
6287 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6289 23. Changed "disc" in the source to "disk" to conform to the documentation and
6290 the book and for uniformity.
6292 24. Ignore Sendmail's -Ooption=value command line item.
6294 25. When execve() failed while trying to run a command in a pipe transport,
6295 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6296 could be confused with a return value of 69 from the command itself. This
6297 has been changed to 127, the value the shell returns if it is asked to run
6298 a non-existent command. The wording for the related log line suggests a
6299 non-existent command as the problem.
6301 26. If received_header_text expands to an empty string, do not add a Received:
6302 header line to the message. (Well, it adds a token one on the spool, but
6303 marks it "old" so that it doesn't get used or transmitted.)
6305 27. Installed eximstats 1.28 (addition of -nt option).
6307 28. There was no check for failure on the call to getsockname() in the daemon
6308 code. This can fail if there is a shortage of resources on the system, with
6309 ENOMEM, for example. A temporary error is now given on failure.
6311 29. Contrary to the C standard, it seems that in some environments, the
6312 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6313 program. Exim now does this explicitly; it affects the formatting of
6314 timestamps using strftime().
6316 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6317 complaints. I've now initialized all the variables, to avoid this.
6319 32. Header lines added by a system filter were not being "seen" during
6320 transport-time rewrites.
6322 33. The info_callback() function passed to OpenSSL is set up with type void
6323 (*)(SSL *, int, int), as described somewhere. However, when calling the
6324 function (actually a macro) that sets it up, the type void(*)() is
6325 expected. I've put in a cast to prevent warnings from picky compilers.
6327 34. If a DNS black list lookup found a CNAME record, but there were no A
6328 records associated with the domain it pointed at, Exim crashed.
6330 35. If a DNS black list lookup returned more than one A record, Exim ignored
6331 all but the first. It now scans all returned addresses if a particular IP
6332 value is being sought. In this situation, the contents of the
6333 $dnslist_value variable are a list of all the addresses, separated by a
6336 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6337 to accept a host name and all its aliases if the forward lookup for any of
6338 them yielded the IP address of the incoming connection. Now it accepts only
6339 those names whose forward lookup yields the correct IP address. Any other
6340 names are discarded. This closes a loophole whereby a rogue DNS
6341 administrator could create reverse DNS records to break through a
6342 wildcarded host restriction in an ACL.
6344 37. If a user filter or a system filter that ran in a subprocess used any of
6345 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6346 the wrong values were passed to the pipe command ($thisaddress had the
6347 value of $0, $0 had the value of $1, etc). This bug was introduced by
6348 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6350 38. Improved the line breaking for long SMTP error messages from ACLs.
6351 Previously, if there was no break point between 40 and 75 characters, Exim
6352 left the rest of the message alone. Two changes have been made: (a) I've
6353 reduced the minimum length to 35 characters; (b) if it can't find a break
6354 point between 35 and 75 characters, it looks ahead and uses the first one
6355 that it finds. This may give the occasional overlong line, but at least the
6356 remaining text gets split now.
6358 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6359 file descriptors might be low, and that setting 1000 would always raise it.
6360 It turns out that in some environments, the limit is already over 1000 and
6361 that lowering it causes trouble. So now Exim takes care not to decrease it.
6363 40. When delivering a message, the value of $return_path is set to $sender_
6364 address at the start of routing (routers may change the value). By an
6365 oversight, this default was not being set up when an address was tested by
6366 -bt or -bv, which affected the outcome if any router or filter referred to
6369 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6370 log only when "message" or "log_message" are set. However, if one of the
6371 conditions was an address verification, or a call to a nested ACL, the
6372 messages generated by the underlying test were being passed through. This
6373 no longer happens. The underlying message is available in $acl_verify_
6374 message for both "message" and "log_message" expansions, so it can be
6375 passed through if needed.
6377 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6378 new expansion $bh_ to give the encoded byte string without charset
6379 translation. Translation happens only if iconv() is available; HAVE_ICONV
6380 indicates this at build time. HEADERS_CHARSET gives the charset to
6381 translate to; headers_charset can change it in the configuration, and
6382 "headers charset" can change it in an individual filter file.
6384 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6385 that creates RFC 2047 encoded "words" labels them as that charset instead
6386 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6387 expansion operator; (ii) when Exim creates a From: line for a local
6388 message; (iii) when a header line is rewritten to include a "phrase" part.
6390 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6391 buggy, causing it to skip the first lines of messages whose message ID
6392 ended in 'D'. This would not have bitten before Exim release 4.14, because
6393 message IDs were unlikely to end in 'D' before then. The effect was to have
6394 incorrect size information for certain domains.
6396 45. #include "config.h" was missing at the start of the crypt16.c module. This
6397 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6400 46. If there was a timeout during a "random" callout check, Exim treated it as
6401 a failure of the random address, and carried on sending RSET and the real
6402 address. If the delay was just some slowness somewhere, the response to the
6403 original RCPT would be taken as a response to RSET and so on, causing
6404 mayhem of various kinds.
6406 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6407 when I implemented it. It didn't allow for the fact that some option values
6408 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6409 the right test for positive values.
6411 48. Domain names in DNS records are case-independent. Exim always looks them up
6412 in lower case. Some resolvers return domain names in exactly the case they
6413 appear in the zone file, that is, they may contain uppercase letters. Not
6414 all resolvers do this - some return always lower case. Exim was treating a
6415 change of case by a resolver as a change of domain, similar to a widening
6416 of a domain abbreviation. This triggered its re-routing code and so it was
6417 trying to route what was effectively the same domain again. This normally
6418 caused routing to fail (because the router wouldn't handle the domain
6419 twice). Now Exim checks for this case specially, and just changes the
6420 casing of the domain that it ultimately uses when it transmits the message
6423 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6426 50. If a filter generated a file delivery with a non-absolute name (possible if
6427 no home directory exists for the router), the forbid_file option was not
6430 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6431 the existing equality matching.
6433 52. Exim was using ints instead of ino_t variables in some places where it was
6434 dealing with inode numbers.
6436 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6437 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6438 called TMPDIR, and if it finds it is different, it changes its value.
6440 54. The smtp_printf() function is now made available to local_scan() so
6441 additional output lines can be written before returning. There is also an
6442 smtp_fflush() function to enable the detection of a dropped connection.
6443 The variables smtp_input and smtp_batched_input are exported to
6446 55. Changed the default runtime configuration: the message "Unknown user"
6447 has been removed from the ACL, and instead placed on the localuser router,
6448 using the cannot_route_message feature. This means that any verification
6449 failures that generate their own messages won't get overridden. Similarly,
6450 the "Unrouteable address" message that was in the ACL for unverifiable
6451 relay addresses has also been removed.
6453 56. Added hosts_avoid_esmtp to the smtp transport.
6455 57. The exicyclog script was not checking for the esoteric option
6456 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6457 will work only if exicyclog is run under the appropriate euid.
6459 58. Following a discussion on the list, the rules by which Exim recognises line
6460 endings on incoming messages have been changed. The -dropcr and drop_cr
6461 options are now no-ops, retained only for backwards compatibility. The
6462 following line terminators are recognized: LF CRLF CR. However, special
6463 processing applies to CR:
6465 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6466 nor a local message in the state where . is a terminator.
6468 (ii) If a bare CR is encountered in a header line, an extra space is added
6469 after the line terminator so as not to end the header. The reasoning
6470 behind this is that bare CRs in header lines are most likely either
6471 to be mistakes, or people trying to play silly games.
6473 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6474 was being incorrectly given as 18 bytes larger than it should have been.
6475 This is a VOB (very old bug).
6477 60. This may never have affected anything current, but just in case it has:
6478 When the local host is found other than at the start of a list of hosts,
6479 the local host, those with the same MX, and any that follow, are discarded.
6480 When the list in question was part of a longer list of hosts, the following
6481 hosts (not currently being processed) were also being discarded. This no
6482 longer happens. I'm not sure if this situation could ever has previously
6485 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6488 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6489 header. This is something that is recommended in a new Internet Draft, and
6490 is something that is documented as being done by Sendmail. There are two
6491 possible values. For messages generated by the autoreply transport, Exim
6494 Auto-Submitted: auto-replied
6496 whereas for all other generated messages (e.g. bounces) it adds
6498 Auto-Submitted: auto-generated
6500 63. The "personal" condition in filters now includes a test for the
6501 Auto-Submitted: header. If it contains the string "auto-" the message it
6502 not considered personal.
6504 64. Added rcpt_include_affixes as a generic transport option.
6506 65. Added queue_only_override (default true).
6508 66. Added the syslog_duplication option.
6510 67. If what should have been the first header line of a message consisted of
6511 a space followed by a colon, Exim was mis-interpreting it as a header line.
6512 It isn't of course - it is syntactically invalid and should therefore be
6513 treated as the start of the message body. The misbehaviour could have
6514 caused a number of strange effects, including loss of data in subsequent
6515 header lines, and spool format errors.
6517 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6518 client host had authenticated. This control can now be exercised by an ACL
6519 for more flexibility.
6521 69. By default, callouts do not happen when testing with -bh. There is now a
6522 variant, -bhc, which does actually run the callout code, including
6523 consulting and updating the callout cache.
6525 70. Added support for saslauthd authentication, courtesy of Alexander
6528 71. If statvfs() failed on the spool or log directories while checking their
6529 size for availability, Exim confusingly gave the error "space shortage".
6530 Furthermore, in debugging mode it crashed with a floating point exception.
6531 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6532 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6533 serious problem, Exim now writes to the main and panic logs when this
6534 happens, with details of the failure. It then refuses to accept the
6535 incoming message, giving the message "spool directory problem" or "log
6536 directory problem" with a 421 code for SMTP messages.
6538 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6539 0, 1, and 2 exist, because some OS complain for execs without them (see
6540 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6541 descriptors. However, the code omitted to check that the open succeeded,
6542 causing mysterious errors if for some reason the permissions on /dev/null
6543 got screwed. Now Exim writes a message to the main and panic logs, and
6544 bombs out if it can't open /dev/null.
6546 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6547 interact so that it is all more flexible. It is supposed to remain
6548 backwards compatible. Also added extra_local_interfaces.
6550 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6551 to bomb out with an assertion failure - to the client this appears as a
6552 connection drop. This problem occurs in the part of the code that was taken
6553 from the Samba project. Fortunately, the assertion is in a very simple
6554 function, so I have fixed this by reproducing the function inline in the
6555 one place where it is called, and arranging for authentication to fail
6556 instead of killing the process with assert().
6558 75. The SPA client code was not working when the server requested OEM rather
6559 than Unicode encoding.
6561 76. Added code to make require_files with a specific uid setting more usable in
6562 the case where statting the file as root fails - usually a non-root-mounted
6563 NFS file system. When this happens and the failure is EACCES, Exim now
6564 forks a subprocess and does the per-uid checking as the relevant uid.
6566 77. Added process_log_path.
6568 78. If log_file_path was not explicitly set, a setting of check_log_space or
6569 check_log_inodes was ignored.
6571 79. If a space check for the spool or log partitions fails, the incident is now
6572 logged. Of course, in the latter case the data may get lost...
6574 80. Added the %p formatting code to string_format() so that it can be used to
6575 print addresses in debug_print(). Adjusted all the address printing in the
6576 debugging in store.c to use %p rather than %d.
6578 81. There was a concern that a line of code in smtp_in.c could overflow a
6579 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6580 initially expressed, the concern was not well-founded, because trailing
6581 spaces are removed early. However, if the trailing spaces were followed by
6582 a NULL, they did not get removed, so the overflow was possible. Two fixes
6585 (a) I re-wrote the offending code in a cleaner fashion.
6586 (b) If an incoming SMTP command contains a NULL character, it is rejected
6589 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6590 runs initgroups(), so that if the Exim user is in any additional groups,
6591 they will be used during message reception.
6597 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob