1 Change log file for Exim from version 4.21
2 -------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
31 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
32 supports it and a size is available (ie. the sending peer gave us one).
34 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
35 years of being deprecated). The replacements are the ACLs
36 acl_smtp_mime and acl_not_smtp_mime.
38 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
39 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
40 or even an in-clear connection were permitted. Now, if the host lookup
41 was dnssec and dane was requested then the host is only used if the
42 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
43 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
44 if one fails this test.
45 This means that a poorly-configured remote DNS will make it incommunicado;
46 but it protects against a DNS-interception attack on it.
48 JH/04 Bug 1810: make continued-use of an open smtp transport connection
49 non-noisy when a race steals the message being considered.
51 JH/05 If main configuration option tls_certificate is unset, generate a
52 self-signed certificate for inbound TLS connections.
54 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
55 in rewrites and routers.
57 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
58 and logged a warning sing 4.83; now they are a configuration file error.
60 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
61 (lacking @domain). Apply the same qualification processing as RCPT.
63 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
65 JH/10 Support ${sha256:} applied to a string (as well as the previous
68 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
69 a cutthrough deliver is pending, as we always want to make a connection.
70 This also avoids re-routing the message when later placing the cutthrough
71 connection after a verify cache hit.
72 Do not update it with the verify result either.
74 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
75 when routing results in more than one destination address.
77 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
78 signing (which inhibits the cutthrough capability). Previously only
79 the presence of an option was tested; now an expansion evaluating as
80 empty is permissible (obviously it should depend only on data available
81 when the cutthrough connection is made).
83 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
84 the relevant preceding SMTP command did not note the pipelining mode.
86 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
87 Previously they were not counted.
89 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
90 as one having no matching records. Previously we deferred the message
91 that needed the lookup.
93 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
94 distinguished as "(=".
96 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
97 for missing MX records. Previously it only worked for missing A records.
99 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
101 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
102 after the data-go-ahead and data-ack. Patch from Jason Betts.
104 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
105 even for a "none" policy. Patch from Tony Meyer.
107 JH/22 Fix continued use of a connection for further deliveries. If a port was
108 specified by a router, it must also match for the delivery to be
111 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
112 When one of the recipients of a message was unsuitable for the connection
113 (has no matching addresses), we lost track of needing to mark it
114 deferred. As a result mail would be lost.
116 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
118 JH/25 Decoding ACL controls is now done using a binary search; the source code
119 takes up less space and should be simpler to maintain. Merge the ACL
120 condition decode tables also, with similar effect.
122 JH/26 Fix problem with one_time used on a redirect router which returned the
123 parent address unchanged. A retry would see the parent address marked as
124 delivered, so not attempt the (identical) child. As a result mail would
127 JH/27 Fix a possible security hole, wherein a process operating with the Exim
128 UID can gain a root shell. Credit to http://www.halfdog.net/ for
129 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
132 JH/28 Enable {spool,log} filesystem space and inode checks as default.
133 Main config options check_{log,spool}_{inodes,space} are now
134 100 inodes, 10MB unless set otherwise in the configuration.
136 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
137 Previously it only applied to the main-section connection policy
140 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
142 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
143 by me. Added RFC7919 DH primes as an alternative.
145 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
146 are not in the system include path.
148 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
149 GnuTLS, when a session startup failed (eg because the client disconnected)
150 Exim did stdio operations after fclose. This was exposed by a recent
151 change which nulled out the file handle after the fclose.
153 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
154 signed directly by the cert-signing cert, rather than an intermediate
155 OCSP-signing cert. This is the model used by LetsEncrypt.
157 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
159 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
160 an incoming connection.
162 HS/02 Bug 1802: Do not half-close the connection after sending a request
165 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
166 fallback to "prime256v1".
168 JH/34 SECURITY: Use proper copy of DATA command in error message.
169 Could leak key material. Remotely exploitable. CVE-2016-9963.
175 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
176 and 3.4.4 - once the server is enabled to respond to an OCSP request
177 it does even when not requested, resulting in a stapling non-aware
178 client dropping the TLS connection.
180 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
181 support variable-length bit vectors. No functional change.
183 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
184 The I= interface field on outgoing lines is now after the H= remote
185 host field, same as incoming lines. There is a separate
186 outgoing_interface log selector which allows you to disable the
189 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
190 If not running log_selector +smtp_connection the mainlog would be held
191 open indefinitely after a "too many connections" event, including to a
192 deleted file after a log rotate. Leave the per net connection logging
193 leaving it open for efficiency as that will be quickly detected by the
194 check on the next write.
196 HS/01 Bug 1671: Fix post transport crash.
197 Processing the wait-<transport> messages could crash the delivery
198 process if the message IDs didn't exist for some reason. When
199 using 'split_spool_directory=yes' the construction of the spool
200 file name failed already, exposing the same netto behaviour.
202 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
203 mime_regex ACL conditions.
205 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
206 to DSN fail messages (bounces): remote IP, remote greeting, remote response
207 to HELO, local diagnostic string.
209 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
210 log line to debug. Even when configured with a tls authenticator many
211 client connections are expected to not authenticate in this way, so
212 an authenticate fail is not an error.
214 HS/02 Add the Exim version string to the process info. This way exiwhat
215 gives some more detail about the running daemon.
217 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
218 matter for fast-change records such as DNSBLs.
220 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
221 retry record, even if constant. There may be multiple transports with
222 different interface settings and the retry behaviour needs to be kept
225 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
227 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
229 JH/10 Bug 840: fix log_defer_output option of pipe transport
231 JH/11 Bug 830: use same host for all RCPTS of a message, even under
232 hosts_randomize. This matters a lot when combined with mua_wrapper.
234 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
235 ${quote_pgsql:<string>} operator.
237 JH/13 Bug 1708: avoid misaligned access in cached lookup.
239 JH/14 Change header file name for freeradius-client. Relevant if compiling
240 with Radius support; from the Gentoo tree and checked under Fedora.
242 JH/15 Bug 1712: Introduce $prdr_requested flag variable
244 JH/16 Bug 1714: Permit an empty string as expansion result for transport
245 option transport_filter, meaning no filtering.
247 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
249 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
250 defaults to "*" (all hosts). The variable is now available when not built
251 with TLS, default unset, mainly to enable keeping the testsuite sane.
252 If a server certificate is not supplied (via tls_certificate) an error is
253 logged, and clients will find TLS connections fail on startup. Presumably
254 they will retry in-clear.
255 Packagers of Exim are strongly encouraged to create a server certificate
256 at installation time.
258 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
259 with the $config_file variable.
261 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
262 in transport context, after the attempt, and per-recipient. The latter type
263 is per host attempted. The event data is the error message, and the errno
264 information encodes the lookup type (A vs. MX) used for the (first) host,
265 and the trailing two digits of the smtp 4xx response.
267 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
268 to write to mainlog (or rejectlog, paniclog) in the window between file
269 creation and permissions/ownership being changed. Particularly affects
270 installations where exicyclog is run as root, rather than exim user;
271 result is that the running daemon panics and dies.
273 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
275 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
276 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
277 "pri" and "weight". Note that the previous implicit priority given by the
278 list order is no longer honoured.
280 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
283 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
284 by defining SUPPORT_SOCKS.
286 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
287 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
288 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
289 variables are renamed to proxy_{local,external}_{address,port}.
291 JH/27 Move Internationalisation support from Experimental to mainline, enabled
292 for a build by defining SUPPORT_I18N
294 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
295 of the query string, and make ${quote_redis:} do that quoting.
297 JH/29 Move Events support from Experimental to mainline, enabled by default
298 and removable for a build by defining DISABLE_EVENT.
300 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
302 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
303 cached by the daemon.
305 JH/32 Move Redis support from Experimental to mainline, enabled for a build
306 by defining LOOKUP_REDIS. The libhiredis library is required.
308 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
309 keys are given for lookup.
311 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
312 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
313 only supported when built with TLS support. The PolarSSL SHA routines
314 are still used when the TLS library is too old for convenient support.
316 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
317 openssl_options), for security. OpenSSL forces this from version 1.1.0
318 server-side so match that on older versions.
320 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
321 allocation for $value could be released as the expansion processing
322 concluded, but leaving the global pointer active for it.
324 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
325 and to use the domains and local_parts ACL conditions.
327 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
328 incorrectly not doubled on cutthrough transmission, hence seen as a
329 body-termination at the receiving system - resulting in truncated mails.
330 Commonly the sender saw a TCP-level error, and retransmitted the message
331 via the normal store-and-forward channel. This could result in duplicates
332 received - but deduplicating mailstores were liable to retain only the
333 initial truncated version.
335 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
337 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
339 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
340 we're in there, support oversigning also; bug 1309.
342 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
344 HS/04 Add support for keep_environment and add_environment options.
346 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
347 either intentional arithmetic overflow during PRNG, or testing config-
350 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
351 delivery resulted in actual delivery. Cancel cutthrough before DATA
354 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
355 rejects a recipient: pass the reject to the originator.
357 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
358 Many were false-positives and ignorable, but it's worth fixing the
361 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
362 for the new environment-manipulation done at startup. Move the routines
363 from being local to tls.c to being global via the os.c file.
365 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
366 an extract embedded as result-arg for a map, the first arg for extract
367 is unavailable so we cannot tell if this is a numbered or keyed
368 extraction. Accept either.
374 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
377 JH/02 The smtp transport option "multi_domain" is now expanded.
379 JH/03 The smtp transport now requests PRDR by default, if the server offers
382 JH/04 Certificate name checking on server certificates, when exim is a client,
383 is now done by default. The transport option tls_verify_cert_hostnames
384 can be used to disable this per-host. The build option
385 EXPERIMENTAL_CERTNAMES is withdrawn.
387 JH/05 The value of the tls_verify_certificates smtp transport and main options
388 default to the word "system" to access the system default CA bundle.
389 For GnuTLS, only version 3.0.20 or later.
391 JH/06 Verification of the server certificate for a TLS connection is now tried
392 (but not required) by default. The verification status is now logged by
393 default, for both outbound TLS and client-certificate supplying inbound
396 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
399 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
400 Status Notification (bounce) messages are now MIME format per RFC 3464.
401 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
402 under the control of the dsn_advertise_hosts option, and routers may
403 have a dsn_lasthop option.
405 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
406 default, modifiable by a malware= option. The list separator for
407 the options can now be changed in the usual way. Bug 68.
409 JH/10 The smtp_receive_timeout main option is now expanded before use.
411 JH/11 The incoming_interface log option now also enables logging of the
412 local interface on delivery outgoing connections.
414 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
415 if the interface and destination host and port all match.
417 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
420 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
421 Patch from Andrew Lewis.
423 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
424 now supports optional time-restrictions, weighting, and priority
425 modifiers per server. Patch originally by <rommer@active.by>.
427 JH/16 The spamd_address main option now supports a mixed list of local
428 and remote servers. Remote servers can be IPv6 addresses, and
429 specify a port-range.
431 JH/17 Bug 68: The spamd_address main option now supports an optional
432 timeout value per server.
434 JH/18 Bug 1581: Router and transport options headers_add/remove can
435 now have the list separator specified.
437 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
440 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
443 JH/21 Support for the A6 type of dns record is withdrawn.
445 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
446 rather than the verbs used.
448 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
449 from 255 to 1024 chars.
451 JH/24 Verification callouts now attempt to use TLS by default.
453 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
454 are generic router options now. The defaults didn't change.
456 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
457 Original patch from Alexander Shikoff, worked over by JH.
459 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
460 files if ZCAT_COMMAND is not executable.
462 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
464 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
466 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
467 Normally benign, it bites when the pair was led to by a CNAME;
468 modern usage is to not canonicalize the domain to a CNAME target
469 (and we were inconsistent anyway for A-only vs AAAA+A).
471 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
473 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
474 when evaluating $sender_host_dnssec.
476 JH/31 Check the HELO verification lookup for DNSSEC, adding new
477 $sender_helo_dnssec variable.
479 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
481 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
483 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
485 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
486 documented as working, but never had. Support all but $spam_report.
488 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
489 added for tls authenticator.
491 HS/03 Add perl_taintmode main config option
497 TL/01 When running the test suite, the README says that variables such as
498 no_msglog_check are global and can be placed anywhere in a specific
499 test's script, however it was observed that placement needed to be near
500 the beginning for it to behave that way. Changed the runtest perl
501 script to read through the entire script once to detect and set these
502 variables, reset to the beginning of the script, and then run through
503 the script parsing/test process like normal.
505 TL/02 The BSD's have an arc4random API. One of the functions to induce
506 adding randomness was arc4random_stir(), but it has been removed in
507 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
508 function when detected.
510 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
511 cause callback expansion.
513 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
514 syntax errors in an expansion can be treated as a string instead of
515 logging or causing an error, due to the internal use of bool_lax
516 instead of bool when processing it.
518 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
519 server certificates when making smtp deliveries.
521 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
523 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
525 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
527 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
528 Merged patch from Sebastian Wiedenroth.
530 JH/05 Fix results-pipe from transport process. Several recipients, combined
531 with certificate use, exposed issues where response data items split
532 over buffer boundaries were not parsed properly. This eventually
533 resulted in duplicates being sent. This issue only became common enough
534 to notice due to the introduction of connection certificate information,
535 the item size being so much larger. Found and fixed by Wolfgang Breyha.
537 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
538 size buffer was used, resulting in syntax errors when an expansion
541 JH/07 Add support for directories of certificates when compiled with a GnuTLS
542 version 3.3.6 or later.
544 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
545 is EXPERIMENTAL_EVENT, the main-configuration and transport options
546 both become "event_action", the variables become $event_name, $event_data
547 and $event_defer_errno. There is a new variable $verify_mode, usable in
548 routers, transports and related events. The tls:cert event is now also
549 raised for inbound connections, if the main configuration event_action
552 TL/06 In test suite, disable OCSP for old versions of openssl which contained
553 early OCSP support, but no stapling (appears to be less than 1.0.0).
555 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
556 server certificate names available under the smtp transport option
557 "tls_verify_cert_hostname" now do not permit multi-component wildcard
560 JH/10 Time-related extraction expansions from certificates now use the main
561 option "timezone" setting for output formatting, and are consistent
562 between OpenSSL and GnuTLS compilations. Bug 1541.
564 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
565 encoded parameter in the incoming message. Bug 1558.
567 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
568 include certificate info, eximon was claiming there were spoolfile
571 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
573 JH/14 Log delivery-related information more consistently, using the sequence
574 "H=<name> [<ip>]" wherever possible.
576 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
577 are problematic for Debian distribution, omit them from the release
580 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
582 JH/16 Fix string representation of time values on 64bit time_t architectures.
585 JH/17 Fix a null-indirection in certextract expansions when a nondefault
586 output list separator was used.
591 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
592 checkers that were complaining about end of non-void function with no
595 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
596 This was a regression introduced in 4.83 by another bugfix.
598 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
600 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
601 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
607 TF/01 Correctly close the server side of TLS when forking for delivery.
609 When a message was received over SMTP with TLS, Exim failed to clear up
610 the incoming connection properly after forking off the child process to
611 deliver the message. In some situations the subsequent outgoing
612 delivery connection happened to have the same fd number as the incoming
613 connection previously had. Exim would try to use TLS and fail, logging
614 a "Bad file descriptor" error.
616 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
617 utilities have not been installed.
619 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
620 temporary space as the ACL may create new global variables.
622 TL/01 LDAP support uses per connection or global context settings, depending
623 upon the detected version of the libraries at build time.
625 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
626 to extract and use the src ip:port in logging and expansions as if it
627 were a direct connection from the outside internet. PPv2 support was
628 updated based on HAProxy spec change in May 2014.
630 JH/02 Add ${listextract {number}{list}{success}{fail}}.
632 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
633 Properly escape header and check for NULL return.
635 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
638 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
640 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
641 characters in header names, implemented as a verify condition.
642 Contributed by Michael Fischer v. Mollard.
644 TL/05 Rename SPF condition results err_perm and err_temp to standardized
645 results permerror and temperror. Previous values are deprecated but
646 still accepted. In a future release, err_perm and err_temp will be
647 completely removed, which will be a backward incompatibility if the
648 ACL tests for either of these two old results. Patch contributed by
649 user bes-internal on the mailing list.
651 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
653 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
654 selectors, in both main and reject logs.
656 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
659 JH/07 Add malware type "sock" for talking to simple daemon.
661 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
663 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
664 routers/transports under cutthrough routing.
666 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
667 numbers. Touch up "bool" conditional to keep the same definition.
669 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
671 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
673 JH/12 Expand items in router/transport headers_add or headers_remove lists
674 individually rather than the list as a whole. Bug 1452.
676 Required for reasonable handling of multiple headers_ options when
677 they may be empty; requires that headers_remove items with embedded
678 colons must have them doubled (or the list-separator changed).
680 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
681 view the policy declared in the DMARC record. Currently, $dmarc_status
682 is a combined value of both the record presence and the result of the
685 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
687 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
688 dnslookup router and the smtp transport (applying to the forward
691 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
692 of ldap servers used for a specific lookup. Patch provided by Heiko
695 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
696 New variable $lookup_dnssec_authenticated for observability.
698 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
699 Patch submitted by Lars Timman.
701 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
703 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
704 Requires trusted mode and valid format message id, aborts otherwise.
705 Patch contributed by Heiko Schlichting.
707 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
708 certextract with support for various fields. Bug 1358.
710 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
711 is requested by default, modifiable by smtp transport option
714 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
715 operate on certificate variables to give certificate fingerprints
716 Also new ${sha256:cert_variable}.
718 JH/23 The PRDR feature is moved from being Experimental into the mainline.
720 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
721 Christian Aistleitner.
723 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
725 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
726 file. Patch from Wolfgang Breyha.
728 JH/25 Expand the coverage of the delivery $host and $host_address to
729 client authenticators run in verify callout. Bug 1476.
731 JH/26 Port service names are now accepted for tls_on_connect_ports, to
732 align with daemon_smtp_ports. Bug 72.
734 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
735 support and error reporting did not work properly.
737 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
738 and is readable. Patch from Andrew Colin Kissa.
740 TL/14 Enhance documentation of ${run expansion and how it parses the
741 commandline after expansion, particularly in the case when an
742 unquoted variable expansion results in an empty value.
744 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
746 PP/02 Fix internal collision of T_APL on systems which support RFC3123
747 by renaming away from it. Addresses GH issue 15, reported by
750 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
752 TL/15 SECURITY: prevent double expansion in math comparison functions
753 (can expand unsanitized data). Not remotely exploitable.
760 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
762 PP/02 Make -n do something, by making it not do something.
763 When combined with -bP, the name of an option is not output.
765 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
768 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
769 $sender_host_name and config options to manage this, and basic check
772 PP/05 DSCP support for outbound connections and control modifier for inbound.
774 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
775 (Only plugin which currently uses this is kerberos4, which nobody should
776 be using, but we should make it available and other future plugins might
777 conceivably use it, even though it would break NAT; stuff *should* be
778 using channel bindings instead).
780 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
781 name; added for Sendmail compatibility; requires admin caller.
782 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
783 just ignore it); requires trusted caller.
784 Also parse but ignore: -Ac -Am -X<logfile>
787 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
789 TL/02 Add +smtp_confirmation as a default logging option.
791 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
792 Patch by Magnus Holmgren from 2007-02-20.
794 TL/04 Bugzilla 1281 - Spec typo.
795 Bugzilla 1283 - Spec typo.
796 Bugzilla 1290 - Spec grammar fixes.
798 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
800 TL/06 Add Experimental DMARC support using libopendmarc libraries.
802 TL/07 Fix an out of order global option causing a segfault. Reported to dev
803 mailing list by by Dmitry Isaikin.
805 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
807 JH/02 Support "G" suffix to numbers in ${if comparisons.
809 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
811 NM/01 Bugzilla 1197 - Spec typo
812 Bugzilla 1196 - Spec examples corrections
814 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
816 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
817 gnutls_enable_pkcs11, but renamed to more accurately indicate its
820 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
821 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
823 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
824 "acl {{name}{arg}...}", and optional args on acl condition
827 JH/05 Permit multiple router/transport headers_add/remove lines.
829 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
831 JH/07 Avoid using a waiting database for a single-message-only transport.
832 Performance patch from Paul Fisher. Bugzilla 1262.
834 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
837 JH/09 Add $headers_added variable, with content from use of ACL modifier
838 add_header (but not yet added to the message). Bugzilla 199.
840 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
841 Pulled from Bugzilla 817 by Wolfgang Breyha.
843 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
845 (nb: this is the same fix as in Exim 4.80.1)
847 JH/11 Add A= logging on delivery lines, and a client_set_id option on
850 JH/12 Add optional authenticated_sender logging to A= and a log_selector
853 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
855 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
856 advertise SMTP AUTH mechanism to us, instead of a generic
857 protocol violation error. Also, make Exim more robust to bad
858 data from the Dovecot auth socket.
860 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
862 When a queue runner is handling a message, Exim first routes the
863 recipient addresses, during which it prunes them based on the retry
864 hints database. After that it attempts to deliver the message to
865 any remaining recipients. It then updates the hints database using
868 So if a recipient address works intermittently, it can get repeatedly
869 deferred at routing time. The retry hints record remains fresh so the
870 address never reaches the final cutoff time.
872 This is a fairly common occurrence when a user is bumping up against
873 their storage quota. Exim had some logic in its local delivery code
874 to deal with this. However it did not apply to per-recipient defers
875 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
877 This change adds a proper retry rule check during routing so that the
878 final cutoff time is checked against the message's age. We only do
879 this check if there is an address retry record and there is not a
880 domain retry record; this implies that previous attempts to handle
881 the address had the retry_use_local_parts option turned on. We use
882 this as an approximation for the destination being like a local
883 delivery, as in LMTP.
885 I suspect this new check makes the old local delivery cutoff check
886 redundant, but I have not verified this so I left the code in place.
888 TF/02 Correct gecos expansion when From: is a prefix of the username.
890 Test 0254 submits a message to Exim with the header
894 When I ran the test suite under the user fanf2, Exim expanded
895 the header to contain my full name, whereas it should have added
896 a Resent-Sender: header. It erroneously treats any prefix of the
897 username as equal to the username.
899 This change corrects that bug.
901 GF/01 DCC debug and logging tidyup
902 Error conditions log to paniclog rather than rejectlog.
903 Debug lines prefixed by "DCC: " to remove any ambiguity.
905 TF/03 Avoid unnecessary rebuilds of lookup-related code.
907 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
908 Bug spotted by Jeremy Harris; was flawed since initial commit.
909 Would have resulted in OCSP responses post-SNI triggering an Exim
910 NULL dereference and crash.
912 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
914 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
915 Bug detection, analysis and fix by Samuel Thibault.
916 Bugzilla 1331, Debian bug #698092.
918 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
920 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
921 Server implementation by Todd Lyons, client by JH.
922 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
923 config variable "prdr_enable" controls whether the server
924 advertises the facility. If the client requests PRDR a new
925 acl_data_smtp_prdr ACL is called once for each recipient, after
926 the body content is received and before the acl_smtp_data ACL.
927 The client is controlled by both of: a hosts_try_prdr option
928 on the smtp transport, and the server advertisement.
929 Default client logging of deliveries and rejections involving
930 PRDR are flagged with the string "PRDR".
932 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
933 fclose(). Diagnosis by Todd Lyons.
935 PP/17 Update configure.default to handle IPv6 localhost better.
936 Patch by Alain Williams (plus minor tweaks).
939 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
940 This is now consistent with GnuTLS, and is now documented: the
941 previous undocumented portable approach to treating the option as
942 unset was to force an expansion failure. That still works, and
943 an empty string is now equivalent.
945 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
946 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
947 not performing validation itself.
949 PP/20 Added force_command boolean option to pipe transport.
950 Patch from Nick Koston, of cPanel Inc.
952 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
955 TF/04 Added udpsend ACL modifier and hexquote expansion operator
957 PP/21 Fix eximon continuous updating with timestamped log-files.
958 Broken in a format-string cleanup in 4.80, missed when I repaired the
959 other false fix of the same issue.
960 Report and fix from Heiko Schlichting.
963 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
964 Report from Prashanth Katuri.
966 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
967 It's SecureTransport, so affects any MacOS clients which use the
968 system-integrated TLS libraries, including email clients.
970 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
971 using a MIME ACL for non-SMTP local injection.
972 Report and assistance in diagnosis by Warren Baker.
974 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
976 JH/16 Fix comparisons for 64b. Bugzilla 1385.
978 TL/09 Add expansion variable $authenticated_fail_id to keep track of
979 last id that failed so it may be referenced in subsequent ACL's.
981 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
984 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
985 ldap library initialization, allowing self-signed CA's to be
986 used. Also properly sets require_cert option later in code by
987 using NULL (global ldap config) instead of ldap handle (per
988 session). Bug diagnosis and testing by alxgomz.
990 TL/12 Enhanced documentation in the ratelimit.pl script provided in
991 the src/util/ subdirectory.
993 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
994 renamed to Transport Post Delivery Action by Jeremy Harris, as
997 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
998 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
999 redis_servers = needs to be configured which will be used by the redis
1000 lookup. Patch from Warren Baker, of The Packet Hub.
1002 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1004 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1005 hostname or reverse DNS when processing a host list. Used suggestions
1006 from multiple comments on this bug.
1008 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1010 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1011 Missed a few lines, added it to make the runtest require no keyboard
1014 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1015 contains upper case chars. Make router use caseful_local_part.
1017 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1018 support when GnuTLS has been built with p11-kit.
1024 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1026 This, or similar/improved, will also be change PP/11 of 4.82.
1032 PP/01 Handle short writes when writing local log-files.
1033 In practice, only affects FreeBSD (8 onwards).
1034 Bugzilla 1053, with thanks to Dmitry Isaikin.
1036 NM/01 Bugzilla 949 - Documentation tweak
1038 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1041 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1043 PP/02 Implemented gsasl authenticator.
1045 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1047 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1048 `pkg-config foo` for cflags/libs.
1050 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1051 with rest of GSASL and with heimdal_gssapi.
1053 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1054 `pkg-config foo` for cflags/libs for the TLS implementation.
1056 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1057 properties get this fed in as external SSF. A number of robustness
1058 and debugging improvements to the cyrus_sasl authenticator.
1060 PP/08 cyrus_sasl server now expands the server_realm option.
1062 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1063 Patch by Jeremy Harris.
1065 PP/10 Added dbmjz lookup type.
1067 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1069 PP/12 MAIL args handles TAB as well as SP, for better interop with
1070 non-compliant senders.
1071 Analysis and variant patch by Todd Lyons.
1073 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1074 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1075 Patch from Dirk Mueller <dmueller@suse.com>
1077 PP/13 tls_peerdn now print-escaped for spool files.
1078 Observed some $tls_peerdn in wild which contained \n, which resulted
1079 in spool file corruption.
1081 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1082 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1083 or write after TLS renegotiation, which otherwise led to messages
1086 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1087 as a tracking header (ie: a signed header comes before the signature).
1088 Patch from Wolfgang Breyha.
1090 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1091 comma-sep list; embedded commas doubled.
1093 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1095 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1097 Report and patch from Dmitry Banschikov.
1099 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1100 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1101 failures. We appear to now support TLS1.1+ with Exim.
1103 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1104 lets Exim select keys and certificates based upon TLS SNI from client.
1105 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1106 before an outbound SMTP session. New log_selector, +tls_sni.
1108 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1109 NULL dereference. Report and patch from Alun Jones.
1111 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1112 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1113 Not seeing resolver debug output on NetBSD, but suspect this is a
1114 resolver implementation change.
1116 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1117 Left warnings. Added "eximon gdb" invocation mode.
1119 PP/21 Defaulting "accept_8bitmime" to true, not false.
1121 PP/22 Added -bw for inetd wait mode support.
1123 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1124 locate the relevant includes and libraries. Made this the default.
1126 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1127 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1129 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1130 This may cause build issues on older platforms.
1132 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1133 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1134 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1135 Added SNI support via GnuTLS too.
1136 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1138 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1140 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1141 Applied second patch from Janne, implementing suggestion to default
1142 multiple-strings-in-record handling to match SPF spec.
1144 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1146 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1147 read-only, out of scope).
1148 Patch from Wolfgang Breyha, report from Stuart Northfield.
1150 PP/29 Fix three issues highlighted by clang analyser static analysis.
1151 Only crash-plausible issue would require the Cambridge-specific
1152 iplookup router and a misconfiguration.
1153 Report from Marcin Mirosław.
1155 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1157 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1158 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1159 As part of this, removing so much warning spew let me fix some minor
1160 real issues in debug logging.
1162 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1163 assignment on my part. Fixed.
1165 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1166 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1167 Janne Snabb (who went above and beyond: thank you).
1169 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1170 string otherwise requires a connection and a bunch more work and it's
1171 relatively easy to get wrong. Should also expose TLS library linkage
1174 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1175 64-bit ${eval} (JH/03).
1177 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1178 GNU libc to support some of the 64-bit stuff, should not lead to
1179 conflicts. Defined before os.h is pulled in, so if a given platform
1180 needs to override this, it can.
1182 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1183 protection layer was required, which is not implemented.
1184 Bugzilla 1254, patch from Wolfgang Breyha.
1186 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1187 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1188 tls_dhparam take prime identifiers. Also unbreak combination of
1189 OpenSSL+DH_params+TLSSNI.
1191 PP/39 Disable SSLv2 by default in OpenSSL support.
1197 PP/01 Solaris build fix for Oracle's LDAP libraries.
1198 Bugzilla 1109, patch from Stephen Usher.
1200 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1202 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1205 TF/02 Fix a couple more cases where we did not log the error message
1206 when unlink() failed. See also change 4.74-TF/03.
1208 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1209 lock up or crash if it happened to be inside a call to libc when it
1210 got a SIGUSR1 from exiwhat.
1212 The SIGUSR1 handler appends the current process status to the process
1213 log which is later printed by exiwhat. It used to use the general
1214 purpose logging code to do this, but several functions it calls are
1215 not safe for signals.
1217 The new output code in the SIGUSR1 handler is specific to the process
1218 log, and simple enough that it's easy to inspect for signal safety.
1219 Removing some special cases also simplifies the general logging code.
1220 Removing the spurious timestamps from the process log simplifies
1223 TF/04 Improved ratelimit ACL condition.
1225 The /noupdate option has been deprecated in favour of /readonly which
1226 has clearer semantics. The /leaky, /strict, and /readonly update modes
1227 are mutually exclusive. The update mode is no longer included in the
1228 database key; it just determines when the database is updated. (This
1229 means that when you upgrade Exim will forget old rate measurements.)
1231 Exim now checks that the per_* options are used with an update mode that
1232 makes sense for the current ACL. For example, when Exim is processing a
1233 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1234 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1235 must specify per_mail/readonly. If you omit the update mode it defaults to
1236 /leaky where that makes sense (as before) or /readonly where required.
1238 The /noupdate option is now undocumented but still supported for
1239 backwards compatibility. It is equivalent to /readonly except that in
1240 ACLs where /readonly is required you may specify /leaky/noupdate or
1241 /strict/noupdate which are treated the same as /readonly.
1243 A useful new feature is the /count= option. This is a generalization
1244 of the per_byte option, so that you can measure the throughput of other
1245 aggregate values. For example, the per_byte option is now equivalent
1246 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1248 The per_rcpt option has been generalized using the /count= mechanism
1249 (though it's more complicated than the per_byte equivalence). When it is
1250 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1251 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1252 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1253 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1254 Note that using per_rcpt with a non-readonly update mode in more than
1255 one ACL will cause the recipients to be double-counted. (The per_mail
1256 and per_byte options don't have this problem.)
1258 The handling of very low rates has changed slightly. If the computed rate
1259 is less than the event's count (usually one) then this event is the first
1260 after a long gap. In this case the rate is set to the same as this event's
1261 count, so that the first message of a spam run is counted properly.
1263 The major new feature is a mechanism for counting the rate of unique
1264 events. The new per_addr option counts the number of different
1265 recipients that someone has sent messages to in the last time period. It
1266 behaves like per_rcpt if all the recipient addresses are different, but
1267 duplicate recipient addresses do not increase the measured rate. Like
1268 the /count= option this is a general mechanism, so the per_addr option
1269 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1270 example, measure the rate that a client uses different sender addresses
1271 with the options per_mail/unique=$sender_address. There are further
1272 details in the main documentation.
1274 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1276 TF/06 Removed a few PCRE remnants.
1278 TF/07 Automatically extract Exim's version number from tags in the git
1279 repository when doing development or release builds.
1281 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1282 Bugzilla 879. Patch from Paul Fisher.
1284 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1285 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1288 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1290 PP/05 Variable $av_failed, true if the AV scanner deferred.
1291 Bugzilla 1078. Patch from John Horne.
1293 PP/06 Stop make process more reliably on build failure.
1294 Bugzilla 1087. Patch from Heiko Schlittermann.
1296 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1297 Bugzilla 1089. Patch from Heiko Schlittermann.
1299 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1300 Bugzilla 1131. Patch from Holger Weiß.
1302 PP/09 Handle IPv6 addresses with SPF.
1303 Bugzilla 860. Patch from Wolfgang Breyha.
1305 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1307 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1310 PP/11 match_* no longer expand right-hand-side by default.
1311 New compile-time build option, EXPAND_LISTMATCH_RHS.
1312 New expansion conditions, "inlist", "inlisti".
1314 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1316 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1318 PP/14 fix log_write() format string regression from TF/03.
1319 Bugzilla 1152. Patch from Dmitry Isaikin.
1325 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1327 PP/02 Harmonised TLS library version reporting; only show if debugging.
1328 Layout now matches that introduced for other libraries in 4.74 PP/03.
1330 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1332 PP/04 New "dns_use_edns0" global option.
1334 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1337 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1338 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1340 TK/01 Updated PolarSSL code to 0.14.2.
1341 Bugzilla 1097. Patch from Andreas Metzler.
1343 PP/07 Catch divide-by-zero in ${eval:...}.
1344 Fixes bugzilla 1102.
1346 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1349 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1350 format-string attack -- SECURITY: remote arbitrary code execution.
1352 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1353 time unintentionally subject to list matching rules, letting the header
1354 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1355 arbitrary string expansion). This allowed for information disclosure.
1357 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1358 INT_MIN/-1 -- value coerced to INT_MAX.
1364 NM/01 Workaround for PCRE version dependency in version reporting
1367 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1368 This fixes portability to compilers other than gcc, notably
1369 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1371 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1372 makefiles for portability to HP-UX and POSIX correctness.
1374 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1375 Also via indented variable definition in the Makefile.
1376 (Debugging by Oliver Heesakkers).
1378 PP/02 Restore caching of spamd results with expanded spamd_address.
1379 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1381 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1382 Improves build reliability. Fix from: Frank Elsner
1384 NM/02 Fix wide character breakage in the rfc2047 coding
1385 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1387 NM/03 Allow underscore in dnslist lookups
1388 Fixes bug 1026. Patch from Graeme Fowler
1390 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1391 Code patches from Adam Ciarcinski of NetBSD.
1393 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1396 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1397 is logged, to avoid truncation. Patch from John Horne.
1399 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1400 Patch from Jakob Hirsch.
1402 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1403 SQL string expansion failure details.
1404 Patch from Andrey Oktyabrski.
1406 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1407 Patch from Simon Arlott.
1409 PP/09 New lookups functionality failed to compile on old gcc which rejects
1410 extern declarations in function scope.
1411 Patch from Oliver Fleischmann
1413 PP/10 Use sig_atomic_t for flags set from signal handlers.
1414 Check getgroups() return and improve debugging.
1415 Fixed developed for diagnosis in bug 927 (which turned out to be
1418 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1419 Patch from Mark Zealey.
1421 PP/12 Bugzilla 1056: Improved spamd server selection.
1422 Patch from Mark Zealey.
1424 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1425 Based on patch from Heiko Schlittermann.
1427 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1428 Patch from Uwe Doering, sign-off by Michael Haardt.
1430 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1431 variable declaration deep within a block. Bug and patch from
1434 PP/15 lookups-Makefile IRIX compatibility coercion.
1436 PP/16 Make DISABLE_DKIM build knob functional.
1438 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1439 Patch by Simon Arlott
1441 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1442 variable argument macros. Our copy now differs from upstream.
1448 TF/01 Failure to get a lock on a hints database can have serious
1449 consequences so log it to the panic log.
1451 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1452 controlled using the smtp_confirmation log selector.
1454 TF/03 Include the error message when we fail to unlink a spool file.
1456 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1457 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1458 for maintaining out-of-tree patches for some time.
1460 PP/01 Bugzilla 139: Documentation and portability issues.
1461 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1462 Handle per-OS dynamic-module compilation flags.
1464 PP/02 Let /dev/null have normal permissions.
1465 The 4.73 fixes were a little too stringent and complained about the
1466 permissions on /dev/null. Exempt it from some checks.
1467 Reported by Andreas M. Kirchwitz.
1469 PP/03 Report version information for many libraries, including
1470 Exim version information for dynamically loaded libraries. Created
1471 version.h, now support a version extension string for distributors
1472 who patch heavily. Dynamic module ABI change.
1474 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1475 privilege escalation vulnerability whereby the Exim run-time user
1476 can cause root to append content of the attacker's choosing to
1479 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1482 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1483 If dropping privileges for untrusted macros, we disabled normal logging
1484 on the basis that it would fail; for the Exim run-time user, this is not
1485 the case, and it resulted in successful deliveries going unlogged.
1486 Fixed. Reported by Andreas Metzler.
1492 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1493 only prepend for the Resent-* case. Fixes regression introduced in
1494 Exim 4.70 by NM/22 for Bugzilla 607.
1496 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1497 increasing numbers of administrators be bitten by this.
1499 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1501 PP/03 Bugzilla 994: added openssl_options main configuration option.
1503 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1505 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1507 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1509 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1510 without a peer certificate, leading to a segfault because of an
1511 assumption that peers always have certificates. Be a little more
1512 paranoid. Problem reported by Martin Tscholak.
1514 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1515 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1516 NB: ClamAV planning to remove STREAM in "middle of 2010".
1517 CL also introduces -bmalware, various -d+acl logging additions and
1518 more caution in buffer sizes.
1520 PP/09 Implemented reverse_ip expansion operator.
1522 PP/10 Bugzilla 937: provide a "debug" ACL control.
1524 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1526 PP/12 Bugzilla 973: Implement --version.
1528 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1530 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1532 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1534 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1535 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1536 ignore trailing whitespace.
1538 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1540 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1543 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1544 Notification from Dr Andrew Aitchison.
1546 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1547 ExtendedDetectionInfo response format.
1548 Notification from John Horne.
1550 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1553 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1554 XSL and documented dependency on system catalogs, with examples of how
1557 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1560 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1561 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1562 configuration file which is writeable by the Exim user or group.
1564 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1565 of configuration files to cover files specified with the -C option if
1566 they are going to be used with root privileges, not just the default
1569 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1570 option (effectively making it always true).
1572 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1573 files to be used while preserving root privileges.
1575 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1576 that rogue child processes cannot use them.
1578 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1579 run-time user, instead of root.
1581 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1582 Exim run-time user without dropping privileges.
1584 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1585 result string, instead of calling string_vformat() twice with the same
1588 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1589 for other users. Others should always drop root privileges if they use
1590 -C on the command line, even for a whitelisted configure file.
1592 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1594 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1600 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1601 $data_path, and $header_path variables; fixed documentation bugs and
1604 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1605 exipick to access non-standard spools, including the "frozen" queue
1608 NM/01 Bugzilla 965: Support mysql stored procedures.
1609 Patch from Alain Williams
1611 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1613 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1614 Patch from Andreas Metzler
1616 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1617 Patch from Kirill Miazine
1619 NM/05 Bugzilla 671: Added umask to procmail example.
1621 JJ/03 installed exipick 20100323.0, fixing doc bug
1623 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1624 directory. Notification and patch from Dan Rosenberg.
1626 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1628 TK/02 Improve log output when DKIM signing operation fails.
1630 MH/01 Treat the transport option dkim_domain as a colon separated
1631 list, not as a single string, and sign the message with each element,
1632 omitting multiple occurences of the same signer.
1634 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1635 Bugzilla 985, 986. Patch by Simon Arlott
1637 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1638 Patch by Simon Arlott
1640 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1641 MBX locking. Notification from Dan Rosenberg.
1647 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1649 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1651 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1653 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1655 NM/04 Bugzilla 915: Flag broken perl installation during build.
1661 TK/01 Added patch by Johannes Berg that expands the main option
1662 "spamd_address" if it starts with a dollar sign.
1664 TK/02 Write list of recipients to X-Envelope-Sender header when building
1665 the mbox-format spool file for content scanning (suggested by Jakob
1668 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1669 (http://www.dcc-servers.net/) support via dccifd. Activated by
1670 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1672 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1673 by Mark Daniel Reidel <mr@df.eu>.
1675 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1676 When building exim an external PCRE library is now needed -
1677 PCRE is a system library on the majority of modern systems.
1678 See entry on PCRE_LIBS in EDITME file.
1680 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1681 conversation. Added nologin parameter to request.
1682 Patch contributed by Kirill Miazine.
1684 TF/01 Do not log submission mode rewrites if they do not change the address.
1686 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1688 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1689 log files in place. Contributed by Roberto Lima.
1691 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1693 TF/03 Bugzilla 615: When checking the local_parts router precondition
1694 after a local_part_suffix or local_part_prefix option, Exim now
1695 does not use the address's named list lookup cache, since this
1696 contains cached lookups for the whole local part.
1698 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1699 Robert Millan. Documentation is in experimental-spec.txt.
1701 TF/04 Bugzilla 668: Fix parallel build (make -j).
1703 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1705 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1706 Patch provided by Jan Srzednicki.
1708 TF/05 Leading white space used to be stripped from $spam_report which
1709 wrecked the formatting. Now it is preserved.
1711 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1712 that they are available at delivery time.
1714 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1716 TF/08 TLS error reporting now respects the incoming_interface and
1717 incoming_port log selectors.
1719 TF/09 Produce a more useful error message if an SMTP transport's hosts
1720 setting expands to an empty string.
1722 NM/06 Bugzilla 744: EXPN did not work under TLS.
1723 Patch provided by Phil Pennock.
1725 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1726 Patch provided by Richard Godbee.
1728 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1729 acl_smtp_notquit, added index entry.
1731 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1732 Patch provided by Eugene Bujak.
1734 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1735 accept(). Patch provided by Maxim Dounin.
1737 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1738 Patch provided by Phil Pennock.
1740 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1742 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1743 Patch provided by Brad "anomie" Jorsch.
1745 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1746 Patch provided by Dean Brooks.
1748 TK/05 Add native DKIM support (does not depend on external libraries).
1750 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1751 Patch provided by Graeme Fowler.
1753 NM/16 Bugzilla 851: Documentation example syntax fix.
1755 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1757 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1760 NM/19 Bugzilla 745: TLS version reporting.
1761 Patch provided by Phil Pennock.
1763 NM/20 Bugzilla 167: bool: condition support.
1764 Patch provided by Phil Pennock.
1766 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1767 clients. Patch provided by Phil Pennock.
1769 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1770 Patch provided by Brad "anomie" Jorsch.
1772 NM/23 Bugzilla 687: Fix misparses in eximstats.
1773 Patch provided by Heiko Schlittermann.
1775 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1776 Patch provided by Heiko Schlittermann.
1778 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1779 plus update to original patch.
1781 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1783 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1784 Patch provided by David Brownlee.
1786 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1788 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1790 NM/30 Bugzilla 888: TLS documentation bugfixes.
1792 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1794 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1795 Unlike the original bugzilla I have changed all shell scripts in src tree.
1797 NM/33 Bugzilla 898: Transport filter timeout fix.
1798 Patch by Todd Rinaldo.
1800 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1801 Patch by Serge Demonchaux.
1803 NM/35 Bugzilla 39: Base64 decode bug fixes.
1804 Patch by Jakob Hirsch.
1806 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1808 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1810 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1812 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1818 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1819 ALT-N's libdkim that I have put here:
1820 http://duncanthrax.net/exim-experimental/
1822 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1823 were called 'true' and it seems that C99 defines that as a reserved
1824 keyword to be used with 'bool' variable types. That means you could
1825 not include C99-style headers which use bools without triggering
1826 build errors in sieve.c.
1828 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1829 as mailq or other aliases. Changed the --help handling significantly
1830 to do whats expected. exim_usage() emits usage/help information.
1832 SC/01 Added the -bylocaldomain option to eximstats.
1834 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1836 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1838 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1844 PH/01 Another patch from the Sieve maintainer.
1846 PH/02 When an IPv6 address is converted to a string for single-key lookup
1847 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1848 dots are used instead of colons so that keys in lsearch files need not
1849 contain colons. This was done some time before quoting was made available
1850 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1851 (notated using the quote facility) so as to distinguish them from IPv4
1852 keys. This meant that lookups for IP addresses in host lists did not work
1853 for iplsearch lookups.
1855 This has been fixed by arranging for IPv6 addresses to be expressed with
1856 colons if the lookup type is iplsearch. This is not incompatible, because
1857 previously such lookups could never work.
1859 The situation is now rather anomalous, since one *can* have colons in
1860 ordinary lsearch keys. However, making the change in all cases is
1861 incompatible and would probably break a number of configurations.
1863 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1866 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1867 conversion specification without a maximum field width, thereby enabling
1868 a rogue spamd server to cause a buffer overflow. While nobody in their
1869 right mind would setup Exim to query an untrusted spamd server, an
1870 attacker that gains access to a server running spamd could potentially
1871 exploit this vulnerability to run arbitrary code as the Exim user.
1873 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1874 $primary_hostname instead of what libspf2 thinks the hosts name is.
1876 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1877 a directory entry by the name of the lookup key. Previously, if a
1878 symlink pointed to a non-existing file or a file in a directory that
1879 Exim lacked permissions to read, a lookup for a key matching that
1880 symlink would fail. Now it is enough that a matching directory entry
1881 exists, symlink or not. (Bugzilla 503.)
1883 PH/03 The body_linecount and body_zerocount variables are now exported in the
1886 PH/04 Added the $dnslist_matched variable.
1888 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1889 This means they are set thereafter only if the connection becomes
1892 PH/06 Added the client_condition to authenticators so that some can be skipped
1893 by clients under certain conditions.
1895 PH/07 The error message for a badly-placed control=no_multiline_responses left
1896 "_responses" off the end of the name.
1898 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1900 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1901 (without spaces) instead of just copying the configuration text.
1903 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1905 PH/11 Added $max_received_linelength.
1907 PH/12 Added +ignore_defer and +include_defer to host lists.
1909 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1910 way in which PCRE > 7.0 is built.
1912 PH/14 Implemented queue_only_load_latch.
1914 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1915 MAIL command. The effect was to mangle the value on 64-bit systems.
1917 PH/16 Another patch from the Sieve maintainer.
1919 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1921 PH/18 If a system quota error occurred while trying to create the file for
1922 a maildir delivery, the message "Mailbox is full" was not appended to the
1923 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1924 only to a quota excession during the actual writing of the file.
1926 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1927 characters?) which causes problems in log lines. The DN values are now
1928 passed through string_printing() before being added to log lines.
1930 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1931 and InterBase are left for another time.)
1933 PH/21 Added message_body_newlines option.
1935 PH/22 Guard against possible overflow in moan_check_errorcopy().
1937 PH/23 POSIX allows open() to be a macro; guard against that.
1939 PH/24 If the recipient of an error message contained an @ in the local part
1940 (suitably quoted, of course), incorrect values were put in $domain and
1941 $local_part during the evaluation of errors_copy.
1947 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1948 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1951 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1952 issue a MAIL command.
1954 PH/02 In an ACL statement such as
1956 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1958 if a client was not listed at all, or was listed with a value other than
1959 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1960 the condition was not true (as it should be), so access was not denied.
1961 The bug was that the ! inversion was incorrectly passed on to the second
1962 item. This has been fixed.
1964 PH/03 Added additional dnslists conditions == and =& which are different from
1965 = and & when the dns lookup returns more than one IP address.
1967 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1968 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1970 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1971 FSYNC, which compiles an option called disable_fsync that allows for
1972 bypassing fsync(). The documentation is heavily laced with warnings.
1974 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1976 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1977 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1978 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1979 including adding "make clean"; (3) Added -fPIC when compiling the test
1980 dynamically loaded module, to get rid of a warning.
1982 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1983 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1984 = 0s. The bug is otherwise harmless.
1986 PH/07 There was a bug in the dovecot authenticator such that the value of
1987 $auth1 could be overwritten, and so not correctly preserved, after a
1988 successful authentication. This usually meant that the value preserved by
1989 the server_setid option was incorrect.
1991 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1993 PH/09 Installed PCRE release 7.0.
1995 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1996 run for batched SMTP input. It is now run at the start of every message
1997 in the batch. While fixing this I discovered that the process information
1998 (output by running exiwhat) was not always getting set for -bs and -bS
1999 input. This is fixed, and it now also says "batched" for BSMTP.
2001 PH/11 Added control=no_pipelining.
2003 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2004 patch, slightly modified), and move the expansion of helo_data till after
2005 the connection is made in the smtp transport (so it can use these
2008 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2010 PH/14 Added log_selector = +pid.
2012 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2014 PH/16 Add ${if forany and ${if forall.
2016 PH/17 Added dsn_from option to vary the From: line in DSNs.
2018 PH/18 Flush SMTP output before performing a callout, unless control =
2019 no_callout_flush is set.
2021 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2022 was true (the default) a successful delivery failed to delete the retry
2023 item, thus causing premature timeout of the address. The bug is now
2026 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2028 PH/21 Long custom messages for fakedefer and fakereject are now split up
2029 into multiline responses in the same way that messages for "deny" and
2030 other ACL rejections are.
2032 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2033 with slight modification.
2035 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2036 draft, changing the syntax and factoring some duplicate code".
2038 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2039 for deliveries of the second and subsequent messages over the same SMTP
2042 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2043 ${reduce, with only minor "tidies".
2045 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2047 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2048 expansion side effects.
2050 PH/27 When a message times out after an over-quota error from an Exim-imposed
2051 quota, the bounce message says "mailbox is full". This message was not
2052 being given when it was a system quota that was exceeded. It now should
2055 MH/03 Made $recipients available in local_scan(). local_scan() already has
2056 better access to the recipient list through recipients_list[], but
2057 $recipients can be useful in postmaster-provided expansion strings.
2059 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2060 in the case of a MAIL command with additional options following the
2061 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2062 were accidentally chopped off.
2064 PH/29 SMTP synchronization checks are implemented when a command is read -
2065 there is a check that no more input is waiting when there shouldn't be
2066 any. However, for some commands, a delay in an ACL can mean that it is
2067 some time before the response is written. In this time, more input might
2068 arrive, invalidly. So now there are extra checks after an ACL has run for
2069 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2070 pipelining has not been advertised.
2072 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2074 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2075 returned in the SMTP response when smtp_return_error_details was false.
2076 This has been fixed.
2078 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2079 instead of the C I/O that was originally supplied, because problems were
2080 reported on Solaris.
2082 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2083 Exim which did not show up earlier: it was assuming that a call to
2084 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2085 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2086 was a macro that became an assignment, so it seemed to work. This has
2087 changed to a proper function call with a void return, hence the compile
2088 error. Exim's code has been fixed.
2090 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2093 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2095 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2097 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2098 because it turns out to be too big or there is a timeout) there is no
2099 "Completed" line in the log. When some messages of this type were
2100 selected by exigrep, they were listed as "not completed". Others were
2101 picked up by some special patterns. I have improved the selection
2102 criteria to be more general.
2104 PH/38 The host_find_failed option in the manualroute router can now be set
2105 to "ignore", to completely ignore a host whose IP address cannot be
2106 found. If all hosts are ignored, the behaviour is controlled by the new
2107 host_all_ignored option.
2109 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2110 homing or because of multiple MX records with /mx) generated more than
2111 one IP address, and the following item turned out to be the local host,
2112 all the secondary addresses of the first item were incorrectly removed
2113 from the list, along with the local host and any following hosts (which
2114 is what is supposed to happen).
2116 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2117 whoever called Exim into the -H file. In the case of the daemon it was
2118 behaving confusingly. When first started, it used values for whoever
2119 started the daemon, but after a SIGHUP it used the Exim user (because it
2120 calls itself on a restart). I have changed the code so that it now always
2123 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2124 message are rejected with the same error (e.g. no authentication or bad
2125 sender address), and a DATA command is nevertheless sent (as can happen
2126 with PIPELINING or a stupid MUA), the error message that was given to the
2127 RCPT commands is included in the rejection of the DATA command. This is
2128 intended to be helpful for MUAs that show only the final error to their
2131 PH/42 Another patch from the Sieve maintainer.
2133 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2134 Eximstats - Fixed some broken HTML links and added missing column headers
2136 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2137 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2139 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2141 PH/43 Yet another patch from the Sieve maintainer.
2143 PH/44 I found a way to check for a TCP/IP connection going away before sending
2144 the response to the final '.' that terminates a message, but only in the
2145 case where the client has not sent further data following the '.'
2146 (unfortunately, this is allowed). However, in many cases there won't be
2147 any further data because there won't be any more messages to send. A call
2148 to select() can be used: if it shows that the input is "ready", there is
2149 either input waiting, or the socket has been closed. An attempt to read
2150 the next input character can distinguish the two cases. Previously, Exim
2151 would have sent an OK response which the client would never have see.
2152 This could lead to message repetition. This fix should cure that, at
2153 least in a lot of common cases.
2155 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2156 advertised in response to EHLO.
2162 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2163 fixed by 4.65/MH/01 (is this a record?) are fixed:
2165 (i) An empty string was always treated as zero by the numeric comparison
2166 operators. This behaviour has been restored.
2168 (ii) It is documented that the numeric comparison operators always treat
2169 their arguments as decimal numbers. This was broken in that numbers
2170 starting with 0 were being interpreted as octal.
2172 While fixing these problems I realized that there was another issue that
2173 hadn't been noticed. Values of message_size_limit (both the global option
2174 and the transport option) were treated as octal if they started with 0.
2175 The documentation was vague. These values are now always treated as
2176 decimal, and I will make that clear in the documentation.
2182 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2183 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2186 MH/01 Don't check that the operands of numeric comparison operators are
2187 integers when their expansion is in "skipping" mode (fixes bug
2188 introduced by 4.64-PH/07).
2190 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2191 child addresses, Exim now panics and dies. Previously, because the count
2192 is held in a short int, deliveries were likely to be lost. As such a
2193 large number of recipients for a single message is ridiculous
2194 (performance will be very, very poor), I have chosen to impose a limit
2195 rather than extend the field.
2201 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2202 leftover -K file (the existence of which was triggered by #402).
2203 While we were at it, introduced process PID as part of the -K
2204 filename. This should rule out race conditions when creating
2207 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2208 processing considerably. Previous code took too long for large mails,
2209 triggering a timeout which in turn triggers #401.
2211 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2212 in the DK code in transports.c. sendfile() is not really portable,
2213 hence the _LINUX specificness.
2215 TF/01 In the add_headers option to the mail command in an Exim filter,
2216 there was a bug that Exim would claim a syntax error in any
2217 header after the first one which had an odd number of characters
2220 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2221 callout verification, Exim cached a "reject" for the entire domain. This
2222 is correct for most verifications, but it is not correct for a recipient
2223 verification with use_sender or use_postmaster set, because in that case
2224 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2225 case of MAIL FROM:<> rejection from other early rejections (e.g.
2226 rejection of HELO). When verifying a recipient using a non-null MAIL
2227 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2228 Whatever the result of the callout, the value of the domain cache is
2229 left unchanged (for any other kind of callout, getting as far as trying
2230 RCPT means that the domain itself is ok).
2232 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2235 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2236 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2237 session, the daemon ignores ECONNECT errors and logs others; it now
2238 ignores EPIPE as well.
2240 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2241 (quoted-printable decoding).
2243 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2244 later the small subsequent patch to fix an introduced bug.
2246 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2248 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2250 PH/08 An error is now given if message_size_limit is specified negative.
2252 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2253 to be given (somewhat) arbitrary names.
2255 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2258 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2259 miscellaneous code fixes
2261 PH/10 Added the log_reject_target ACL modifier to specify where to log
2264 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2265 hostname. This is wrong, because it relates to the incoming message (and
2266 probably the interface on which it is arriving) and not to the outgoing
2267 callout (which could be using a different interface). This has been
2268 changed to use the value of the helo_data option from the smtp transport
2269 instead - this is what is used when a message is actually being sent. If
2270 there is no remote transport (possible with a router that sets up host
2271 addresses), $smtp_active_hostname is used.
2273 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2274 tweaks were necessary in order to get it to work (see also 21 below):
2275 (a) The code assumed that strncpy() returns a negative number on buffer
2276 overflow, which isn't the case. Replaced with Exim's string_format()
2278 (b) There were several signed/unsigned issues. I just did the minimum
2279 hacking in of casts. There is scope for a larger refactoring.
2280 (c) The code used strcasecmp() which is not a standard C function.
2281 Replaced with Exim's strcmpic() function.
2282 (d) The code set only $1; it now sets $auth1 as well.
2283 (e) A simple test gave the error "authentication client didn't specify
2284 service in request". It would seem that Dovecot has changed its
2285 interface. Fortunately there's a specification; I followed it and
2286 changed what the client sends and it appears to be working now.
2288 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2291 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2292 address A is aliased to B and C, where B exists and C does not. Without
2293 -v the output is "A verified" because verification stops after a
2294 successful redirection if more than one address is generated. However,
2295 with -v the child addresses are also verified. Exim was outputting "A
2296 failed to verify" and then showing the successful verification for C,
2297 with its parentage. It now outputs "B failed to verify", showing B's
2298 parentage before showing the successful verification of C.
2300 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2301 look up a TXT record in a specific list after matching in a combined
2304 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2305 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2306 they consult the DNS. I had assumed they would set it the way they
2307 wanted; and indeed my experiments on Linux seem to show that in some
2308 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2309 To be on the safe side, however, I have now made the interface to
2310 host_find_byname() similar to host_find_bydns(), with an argument
2311 containing the DNS resolver options. The host_find_byname() function now
2312 sets these options at its start, just as host_find_bydns() does. The smtp
2313 transport options dns_qualify_single and dns_search_parents are passed to
2314 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2315 of host_find_byname() use the default settings of RES_DEFNAMES
2316 (qualify_single) but not RES_DNSRCH (search_parents).
2318 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2319 spool_read_header() do less string testing, by means of a preliminary
2320 switch on the second character of optional "-foo" lines. (This is
2321 overdue, caused by the large number of possibilities that now exist.
2322 Originally there were few.) While I was there, I also converted the
2323 str(n)cmp tests so they don't re-test the leading "-" and the first
2324 character, in the hope this might squeeze out yet more improvement.
2326 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2327 flag allowing group syntax was set by the header_syntax check but not
2328 turned off, possible causing trouble later; (2) The flag was not being
2329 set at all for the header_verify test, causing "group"-style headers to
2330 be rejected. I have now set it in this case, and also caused header_
2331 verify to ignore an empty address taken from a group. While doing this, I
2332 came across some other cases where the code for allowing group syntax
2333 while scanning a header line wasn't quite right (mostly, not resetting
2334 the flag correctly in the right place). These bugs could have caused
2335 trouble for malformed header lines. I hope it is now all correct.
2337 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2338 with the "reply" argument non-NULL. The code, however (which originally
2339 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2340 but it didn't always do it. This confused somebody who was copying the
2341 code for some other use. I have removed all the tests.
2343 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2344 feature that was used to support insecure browsers during the U.S. crypto
2345 embargo. It requires special client support, and Exim is probably the
2346 only MTA that supported it -- and would never use it because real RSA is
2347 always available. This code has been removed, because it had the bad
2348 effect of slowing Exim down by computing (never used) parameters for the
2349 RSA_EXPORT functionality.
2351 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2352 authenticator to fail if there's a tab character in the incoming data
2353 (there should never be unless someone is messing about, as it's supposed
2354 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2355 if the connection is using TLS or if the remote IP is the same as the
2356 local IP, and the "valid-client-cert option" if a client certificate has
2359 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2360 authenticators. This can be used for authorization after authentication
2361 succeeds. (In the case of plaintext, it servers for both authentication
2364 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2365 if any retry times were supplied.
2367 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2368 connection, where there is no client IP address to check. In this
2369 situation, the verify now always succeeds.
2371 PH/25 Applied John Jetmore's -Mset patch.
2373 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2375 PH/27 In a string expansion for a processed (not raw) header when multiple
2376 headers of the same name were present, leading whitespace was being
2377 removed from all of them, but trailing whitespace was being removed only
2378 from the last one. Now trailing whitespace is removed from each header
2379 before concatenation. Completely empty headers in a concatenation (as
2380 before) are ignored.
2382 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2383 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2385 PH/29 [Removed. This was a change that I later backed out, and forgot to
2386 correct the ChangeLog entry (that I had efficiently created) before
2387 committing the later change.]
2389 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2390 address errors (4xx response to RCPT) over the same connection as other
2391 messages routed to the same hosts. Such deliveries are always "forced",
2392 so retry times are not inspected. This resulted in far too many retries
2393 for the affected addresses. The effect occurred only when there were more
2394 hosts than the hosts_max_try setting in the smtp transport when it had
2395 the 4xx errors. Those hosts that it had tried were not added to the list
2396 of hosts for which the message was waiting, so if all were tried, there
2397 was no problem. Two fixes have been applied:
2399 (i) If there are any address or message errors in an SMTP delivery, none
2400 of the hosts (tried or untried) are now added to the list of hosts
2401 for which the message is waiting, so the message should not be a
2402 candidate for sending over the same connection that was used for a
2403 successful delivery of some other message. This seems entirely
2404 reasonable: after all the message is NOT "waiting for some host".
2405 This is so "obvious" that I'm not sure why it wasn't done
2406 previously. Hope I haven't missed anything, but it can't do any
2407 harm, as the worst effect is to miss an optimization.
2409 (ii) If, despite (i), such a delivery is accidentally attempted, the
2410 routing retry time is respected, so at least it doesn't keep
2411 hammering the server.
2413 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2414 in ${readsocket because some servers need this prod.
2416 PH/32 Added some extra debug output when updating a wait-xxx database.
2418 PH/33 The hint "could be header name not terminated by colon", which has been
2419 given for certain expansion errors for a long time, was not being given
2420 for the ${if def:h_colon_omitted{... case.
2422 PH/34 The spec says: "With one important exception, whenever a domain list is
2423 being scanned, $domain contains the subject domain." There was at least
2424 one case where this was not true.
2426 PH/35 The error "getsockname() failed: connection reset by peer" was being
2427 written to the panic log as well as the main log, but it isn't really
2428 panic-worthy as it just means the connection died rather early on. I have
2429 removed the panic log writing for the ECONNRESET error when getsockname()
2432 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2433 runs only) independently of the message's sender address. This meant
2434 that, if the 4xx error was in fact related to the sender, a different
2435 message to the same recipient with a different sender could confuse
2436 things. In particular, this can happen when sending to a greylisting
2437 server, but other circumstances could also provoke similar problems.
2438 I have changed the default so that the retry time for these errors is now
2439 based a combination of the sender and recipient addresses. This change
2440 can be overridden by setting address_retry_include_sender=false in the
2443 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2444 remote server are returned as part of bounce messages. This was not
2445 happening for LMTP over a pipe (the lmtp transport), but now it is the
2446 same for both kinds of LMTP.
2448 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2449 in header lines that were in fact CNAMEs. This is no longer the case.
2451 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2452 and queue runs started by the daemon processed all messages. This has
2453 been fixed so that -R and -S can now usefully be given with -q<time>.
2455 PH/40 Import PCRE release 6.7 (fixes some bugs).
2457 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2459 PH/42 Give an error if -q is specified more than once.
2461 PH/43 Renamed the variables $interface_address and $interface_port as
2462 $received_ip_address and $received_port, to make it clear that these
2463 values apply to message reception, and not to the outgoing interface when
2464 a message is delivered. (The old names remain recognized, of course.)
2466 PH/44 There was no timeout on the connect() call when using a Unix domain
2467 socket in the ${readsocket expansion. There now is.
2469 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2470 be meaningful with "accept".
2472 SC/01 Eximstats V1.43
2473 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2475 SC/02 Eximstats V1.44
2476 Use a glob alias rather than an array ref in the generated
2477 parser. This improves both readability and performance.
2479 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2480 Collect SpamAssassin and rejection statistics.
2481 Don't display local sender or destination tables unless
2482 there is data to show.
2483 Added average volumes into the top table text output.
2485 SC/04 Eximstats V1.46
2486 Collect data on the number of addresses (recipients)
2487 as well as the number of messages.
2489 SC/05 Eximstats V1.47
2490 Added 'Message too big' to the list of mail rejection
2491 reasons (thanks to Marco Gaiarin).
2493 SC/06 Eximstats V1.48
2494 Mainlog lines which have GMT offsets and are too short to
2495 have a flag are now skipped.
2497 SC/07 Eximstats V1.49 (Alain Williams)
2498 Added the -emptyok flag.
2500 SC/08 Eximstats V1.50
2501 Fixes for obtaining the IP address from reject messages.
2503 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2504 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2505 whitespace changes from 4.64-PH/27
2507 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2510 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2511 are found, allow negative numbers in numeric criteria)
2513 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2515 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2516 to match changes made in 4.64-PH/43
2518 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2520 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2521 "domain" is actually an IP address, and force a failure. This locks out
2522 those revolvers/nameservers that support "A-for-A" lookups, in
2523 contravention of the specifications.
2525 PH/48 When a host name was looked up from an IP address, and the subsequent
2526 forward lookup of the name timed out, the host name was left in
2527 $sender_host_name, contrary to the specification.
2529 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2530 restricted to single-key lookups, Exim was not diagnosing an error if
2531 * or *@ was used with a query-style lookup.
2533 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2535 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2536 long ago, but noone interested enough thought of it. Let's just say that
2537 the "1.1" means that there are some new functions that weren't there at
2538 some point in the past.
2540 PH/51 Error processing for expansion failure of helo_data from an smtp
2541 transport during callout processing was broken.
2543 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2544 tested/used via the -bh/-bhc/-bs options.
2546 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2547 bug, fixed in subsequent PCRE releases).
2549 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2550 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2552 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2558 SC/01 Use a glob alias rather than an array ref in eximstats generated
2559 parser. This improves both readability and performance.
2561 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2562 Don't display local sender or destination tables in eximstats unless
2563 there is data to show.
2564 Added average volumes into the eximstats top table text output.
2566 SC/03 Collect data on the number of addresses (recipients) as well
2567 as the number of messages in eximstats.
2569 TF/01 Correct an error in the documentation for the redirect router. Exim
2570 does (usually) call initgroups() when daemonizing.
2572 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2573 with consistent privilege compared to when running as a daemon.
2575 TF/03 Note in the spec that $authenticated_id is not set for local
2576 submissions from trusted users.
2578 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2579 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2581 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2582 by adding some example configuration directives to the default
2583 configuration file. A little bit of work is required to uncomment the
2584 directives and define how usernames and passwords are checked, but
2585 there is now a framework to start from.
2587 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2588 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2589 without this. I don't know how relevant this is to other LDAP libraries.
2591 PH/02 Add the verb name to the "unknown ACL verb" error.
2593 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2595 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2597 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2598 directory not expanded when it should be if an expanded home directory
2599 was set for the address (which is overridden by the transport).
2601 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2604 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2605 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2606 because it is too late at that time, and has no effect.
2608 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2609 security issue with \' (bugzilla #107). I could not use the
2610 PQescapeStringConn() function, because it needs a PGconn value as one of
2613 PH/08 When testing addresses using -bt, indicate those final addresses that
2614 are duplicates that would not cause an additional delivery. At least one
2615 person was confused, thinking that -bt output corresponded to deliveries.
2616 (Suppressing duplicates isn't a good idea as you lose the information
2617 about possibly different redirections that led to the duplicates.)
2619 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2620 systems where poll() doesn't work, in particular OS X.
2622 PH/10 Added more information to debugging output for retry time not reached.
2624 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2625 operations in malware.c.
2627 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2630 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2631 syslog_duplication set false, log lines that would normally be written
2632 both the the main log and to the reject log were not written to syslog at
2635 PH/14 In the default configuration, change the use of "message" in ACL warn
2636 statements to "add_header".
2638 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2639 not followed by a command (e.g. "seen endif").
2641 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2642 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2645 PH/17 Added extra conditions to the default value of delay_warning_condition
2649 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2650 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2651 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2654 The Auto-Submitted: and various List- headers are standardised, whereas I
2655 don't think Precedence: ever was.
2657 PH/18 Refactored debugging code in route_finduser() to show more information,
2658 in particular, the error code if getpwnam() issues one.
2660 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2661 This is apparently needed in addition to the PH/07 change above to avoid
2662 any possible encoding problems.
2664 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2665 but not after initializing Perl.
2667 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2668 output them only if debugging. By default they are written stderr,
2669 apparently, which is not desirable.
2671 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2674 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2677 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2679 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2680 authenticated or an ident call has been made. Suppress the default
2681 values for $authenticated_id and $authenticated_sender (but permit -oMai
2682 and -oMas) when testing with -bh.
2684 PH/24 Re-jigged the order of the tests in the default configuration so that the
2685 tests for valid domains and recipients precede the DNS black list and CSA
2686 tests, on the grounds that those ones are more expensive.
2688 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2689 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2690 This bug exists in every version of Exim that I still have, right back to
2693 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2694 However, an attempt to turn on case-sensitivity in a regex key by
2695 including (?-i) didn't work because the subject string was already
2696 lowercased, and the effects were non-intuitive. It turns out that a
2697 one-line patch can be used to allow (?-i) to work as expected.
2703 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2704 other effects) broke the use of negated acl sub-conditions.
2706 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2709 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2710 "Deny" causes Exim to reject the incoming connection with a 554 error.
2711 Unfortunately, if there is a major crisis, such as a disk failure,
2712 tcp-wrappers gives "deny", whereas what one would like would be some
2713 kind of temporary error. A kludge has been added to help with this.
2714 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2715 554 error is used if errno is still zero or contains ENOENT (which occurs
2716 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2719 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2721 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2722 errors. Otherwise a message that provokes a temporary error (when other
2723 messages do not) can cause a whole host to time out.
2725 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2726 addresses were routed directly to files or pipes from a redirect router.
2727 File deliveries just didn't batch; pipe deliveries might have suffered
2730 PH/06 A failure to get a lock for a hints database would erroneously always say
2731 "Failed to get write lock", even when it was really a read lock.
2733 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2734 of 0600. This has been changed to use the value of the lockfile_mode
2735 option (which defaults to 0600).
2737 PH/08 Applied small patch from the Sieve maintainer.
2739 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2740 folder from quota calculations, a direct delivery into this folder messed
2741 up the contents of the maildirsize file. This was because the regex was
2742 used only to exclude .Trash (or whatever) when the size of the mailbox
2743 was calculated. There was no check that a delivery was happening into an
2744 excluded directory. This bug has been fixed by ignoring all quota
2745 processing for deliveries into excluded directories.
2747 PH/10 Added the maildirfolder_create_regex option to appendfile.
2753 PH/01 The code for finding all the local interface addresses on a FreeBSD
2754 system running IPv6 was broken. This may well have applied to all BSD
2755 systems, as well as to others that have similar system calls. The broken
2756 code found IPv4 interfaces correctly, but gave incorrect values for the
2757 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2758 that it would not match correctly against @[] and not recognize the IPv6
2761 PH/02 The ipliteral router was not recognizing addresses of the form user@
2762 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2764 PH/03 Added disable_ipv6.
2766 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2767 decoded form, because it is most often used to construct To: headers
2768 lines in autoreplies, and the decoded form may well be syntactically
2769 invalid. However, $reply_address has leading white space removed, and all
2770 newlines turned into spaces so that the autoreply transport does not
2773 PH/05 If group was specified without a user on a router, and no group or user
2774 was specified on a transport, the group from the router was ignored.
2776 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2777 for visible compile-time settings that can be used to change these
2778 numbers, for those that want even more. Backwards compatibility with old
2779 spool files has been maintained. However, going back to a previous Exim
2780 release will lost any variables that are in spool files.
2782 PH/07 Two small changes when running in the test harness: increase delay when
2783 passing a TCP/IP connection to a new process, in case the original
2784 process has to generate a bounce, and remove special handling of
2785 127.0.0.2 (sic), which is no longer necessary.
2787 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2788 be the same on different OS.
2790 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2793 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2794 whether --show-vars was specified or not
2796 JJ/02 exipick: Added support for new ACL variable spool format introduced
2799 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2800 syntactically invalid From: or Reply-to: line, and a filter used this to
2801 generate an autoreply, and therefore failed to obtain an address for the
2802 autoreply, Exim could try to deliver to a non-existent relative file
2803 name, causing unrelated and misleading errors. What now happens is that
2804 it logs this as a hard delivery error, but does not attempt to create a
2807 PH/11 The exinext utility has a -C option for testing purposes, but although
2808 the given file was scanned by exinext itself; it wasn't being passed on
2809 when Exim was called.
2811 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2812 an end-of-file indication when reading a command response.
2814 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2815 compiled. In many other places in Exim, IPv6 addresses are always
2816 recognized, so I have changed this. It also means that IPv4 domain
2817 literals of the form [IPV4:n.n.n.n] are now always recognized.
2819 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2820 used if the router is not running as root, for example, when verifying at
2821 ACL time, or when using -bh. The debugging output from this situation was
2822 non-existent - all you got was a failure to exec. I have made two
2825 (a) Failures to set uid/gid, the current directory, or a process leader
2826 in a subprocess such as that created by queryprogram now generate
2827 suitable debugging output when -d is set.
2829 (b) The queryprogram router detects when it is not running as root,
2830 outputs suitable debugging information if -d is set, and then runs
2831 the subprocess without attempting to change uid/gid.
2833 PH/15 Minor change to Makefile for building test_host (undocumented testing
2836 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2837 additional section of a DNS packet that returns MX or SRV records.
2838 Instead, it always explicitly searches for A/AAAA records. This avoids
2839 major problems that occur when a DNS server includes only records of one
2840 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2841 fixed another bug: if SRV records were looked up and the corresponding
2842 address records were *not* found in the additional section, the port
2843 values from the SRV records were lost.
2845 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2846 using the correct key (the original address) when searching the retry
2847 rules in order to find which one to use for generating the retry hint.
2849 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2850 adding the default one. Similarly, if it contains a Reply-To: header, the
2851 errors_reply_to option, if set, is not used.
2853 PH/19 When calculating a retry time, Exim used to measure the "time since
2854 failure" by looking at the "first failed" field in the retry record. Now
2855 it does not use this if it is later than than the arrival time of the
2856 message. Instead it uses the arrival time. This makes for better
2857 behaviour in cases where some deliveries succeed, thus re-setting the
2858 "first failed" field. An example is a quota failure for a huge message
2859 when small messages continue to be delivered. Without this change, the
2860 "time since failure" will always be short, possible causing more frequent
2861 delivery attempts for the huge message than are intended.
2862 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2864 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2865 $1, $2, $3) because the numerical variables can be reset during some
2866 expansion items (e.g. "match"), thereby losing the authentication data.
2868 PH/21 Make -bV show the size of off_t variables so that the test suite can
2869 decide whether to run tests for quotas > 2G.
2871 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2872 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2873 filecount value is greater than 2G or if a quota value is greater than 2G
2874 on a system where the size of off_t is not greater than 4, a panic error
2877 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2878 never match. The debug and -bh output now contains an explicit error
2879 message indicating a malformed IPv4 address or mask.
2881 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2882 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2883 PH/23 above applies.
2885 PH/25 Do not write to syslog when running in the test harness. The only
2886 occasion when this arises is a failure to open the main or panic logs
2887 (for which there is an explicit test).
2889 PH/26 Added the /no_tell option to "control=freeze".
2891 PH/27 If a host name lookup failed very early in a connection, for example, if
2892 the IP address matched host_lookup and the reverse lookup yielded a name
2893 that did not have a forward lookup, an error message of the form "no IP
2894 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2895 could be logged. Now it outputs the IP address instead of "NULL".
2897 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2898 allows the sender and the authenticated sender to be set when
2899 submitting a message from within Exim. Since child_open_exim() is
2900 documented for local_scan(), the new function should be too.
2902 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2903 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2904 results in an empty string is now treated as unset.
2906 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2908 PH/31 Added sender_verify_fail logging option.
2910 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2911 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2912 tidied the source and removed it altogether.
2914 PH/33 When a queue run was abandoned because the load average was too high, a
2915 log line was always written; now it is written only if the queue_run log
2916 selector is set. In addition, the log line for abandonment now contains
2917 information about the queue run such as the pid. This is always present
2918 in "start" and "stop" lines but was omitted from the "abandon" line.
2920 PH/34 Omit spaces between a header name and the colon in the error message that
2921 is given when verify = headers_syntax fails (if there are lots of them,
2922 the message gets confusing).
2924 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2925 names, as there are now some PTR records that contain slashes. This check
2926 is only to protect against broken name servers that fall over on strange
2927 characters, so the fact that it applies to all lookups doesn't matter.
2929 PH/36 Now that the new test suite is complete, we can remove some of the
2930 special code in Exim that was needed for the old test suite. For example,
2931 sorting DNS records because real resolvers return them in an arbitrary
2932 order. The new test suite's fake resolver always returns records in the
2935 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2936 bounces) except when queue_only is set, to avoid logging races between
2937 the different processes.
2939 PH/38 Panic-die if .include specifies a non-absolute path.
2941 PH/39 A tweak to the "H" retry rule from its user.
2943 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2944 a label. They prevented compilation on older perls.
2946 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2947 a warning to be raised on newish perls.
2949 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2950 on queue. Changes to match documented behaviour of showing count of
2951 messages matching specified criteria.
2953 PH/40 Changed the default ident timeout from 30s to 5s.
2955 PH/41 Added support for the use of login_cap features, on those BSD systems
2956 that have them, for controlling the resources used by pipe deliveries.
2958 PH/42 The content-scanning code uses fopen() to create files in which to put
2959 message data. Previously it was not paying any attention to the mode of
2960 the files. Exim runs with umask(0) because the rest of the code creates
2961 files with open(), and sets the required mode explicitly. Thus, these
2962 files were ending up world-writeable. This was not a big issue, because,
2963 being within the spool directory, they were not world-accessible. I have
2964 created a function called modefopen, which takes an additional mode
2965 argument. It sets umask(777), creates the file, chmods it to the required
2966 mode, then resets the umask. All the relevant calls to fopen() in the
2967 content scanning code have been changed to use this function.
2969 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2970 to 24 hours. This avoids potential overflow problems when processing G
2971 and H retry rules. I suspect nobody ever tinkers with this value.
2973 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2975 PH/45 When the plaintext authenticator is running as a client, the server's
2976 challenges are checked to ensure they are valid base64 strings. By
2977 default, the authentication attempt is cancelled if an invalid string is
2978 received. Setting client_ignore_invalid_base64 true ignores these errors.
2979 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2980 they are received. Thus, the responses can be made to depend on the
2981 challenges. If an invalid string is ignored, an empty string is placed in
2984 PH/46 Messages that are created by the autoreply transport now contains a
2985 References: header, in accordance with RFCs 2822 and 3834.
2987 PH/47 Added authenticated_sender_force to the smtp transport.
2989 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2991 PH/49 Installed latest patch from the Sieve maintainer.
2993 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2994 was also set, the appendfile transport was unnecessarily scanning a
2995 directory of message files (e.g. for maildir delivery) to find the count
2996 of files (along with the size), even though it did not need this
2997 information. It now does the scan only if it needs to find either the
2998 size of the count of files.
3000 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3002 PH/52 Two bugs concerned with error handling when the smtp transport is
3005 (i) Exim was not creating retry information for temporary errors given
3006 for individual recipients after the DATA command when the smtp transport
3007 was used in LMTP mode. This meant that they could be retried too
3008 frequently, and not timed out correctly.
3010 (ii) Exim was setting the flag that allows error details to be returned
3011 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3012 recipients that were returned after the DATA command.
3014 PH/53 This is related to PH/52, but is more general: for any failing address,
3015 when detailed error information was permitted to be returned to the
3016 sender, but the error was temporary, then after the final timeout, only
3017 "retry timeout exceeded" was returned. Now it returns the full error as
3018 well as "retry timeout exceeded".
3020 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3021 do this, and (what is worse) MTAs that accept it.
3023 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3024 will now be deprecated.
3026 PH/56 New os.c-cygwin from the Cygwin maintainer.
3028 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3029 formats (previously only available in exim formats via -bpr, -bpru,
3030 and -bpra. Now also available in native and exiqgrep formats)
3032 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3033 with very large, slow to parse queues
3035 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3037 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3039 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3040 responses to authentication challenges, though it was showing the
3041 challenges; (ii) I've removed the CR characters from the debug output for
3044 PH/58 Allow for the insertion of a newline as well as a space when a string
3045 is turned into more than one encoded-word during RFC 2047 encoding. The
3046 Sieve code now uses this.
3048 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3049 data_4xx, lost_connection, tls_required.
3051 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3052 message was being sent as an SMTP response.
3054 PH/61 Add -l and -k options to exicyclog.
3056 PH/62 When verifying, if an address was redirected to one new address, so that
3057 verification continued, and the new address failed or deferred after
3058 having set something in $address_data, the value of $address_data was not
3059 passed back to the ACL. This was different to the case when no
3060 redirection occurred. The value is now passed back in both cases.
3062 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3063 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3064 use login_cap.h, so on its own it isn't the distinguishing feature. The
3065 new name refers directly to the setclassresources() function.
3067 PH/65 Added configuration files for NetBSD3.
3069 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3071 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3072 is preferred over IPv4.
3074 PH/68 The bounce_return_message and bounce_return_body options were not being
3075 honoured for bounces generated during the reception of non-SMTP messages.
3076 In particular, this applied to messages rejected by the ACL. This bug has
3077 been fixed. However, if bounce_return_message is true and bounce_return_
3078 body is false, the headers that are returned for a non-SMTP message
3079 include only those that have been read before the error was detected.
3080 (In the case of an ACL rejection, they have all been read.)
3082 PH/69 The HTML version of the specification is now built in a directory called
3083 spec_html instead of spec.html, because the latter looks like a path with
3084 a MIME-type, and this confuses some software.
3086 PH/70 Catch two compiler warnings in sieve.c.
3088 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3089 function verify_get_ident() calls ip_connect() to connect a socket, but
3090 if the "connect()" function timed out, ip_connect() used to close the
3091 socket. However, verify_get_ident() also closes the socket later, and in
3092 between Exim writes to the log, which may get opened at this point. When
3093 the socket was closed in ip_connect(), the log could get the same file
3094 descriptor number as the socket. This naturally causes chaos. The fix is
3095 not to close the socket in ip_connect(); the socket should be closed by
3096 the function that creates it. There was only one place in the code where
3097 this was missing, in the iplookup router, which I don't think anybody now
3098 uses, but I've fixed it anyway.
3100 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3101 well as to direct DNS lookups. Otherwise the handling of names in host
3102 lists is inconsistent and therefore confusing.
3108 PH/01 Two changes to the default runtime configuration:
3110 (1) Move the checks for relay_from_hosts and authenticated clients from
3111 after to before the (commented out) DNS black list checks.
3113 (2) Add control=submission to the relay_from_hosts and authenticated
3114 clients checks, on the grounds that messages accepted by these
3115 statements are most likely to be submissions.
3117 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3119 (1) Generate an error if the third argument for the ${prvs expansion is
3122 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3125 (3) Reset the variables that are obtained from the first argument of
3126 ${prvscheck and used in the second argument before leaving the code,
3127 because their memory is reclaimed, so using them afterwards may do
3130 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3131 one (it's much easier than Tom thought :-).
3133 (5) Because of (4), we can now allow for the use of $prvscheck_result
3134 inside the third argument.
3136 PH/03 For some reason, the default setting of PATH when running a command from
3137 a pipe transport was just "/usr/bin". I have changed it to
3140 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3141 anything to be listed in the output from -bV.
3143 PH/05 When a filter generated an autoreply, the entire To: header line was
3144 quoted in the delivery log line, like this:
3146 => >A.N.Other <ano@some.domain> <original@ddress> ...
3148 This has been changed so that it extracts the operative address. There
3149 may be more than one such address. If so, they are comma-separated, like
3152 => >ano@some.domain,ona@other.domain <original@ddress> ...
3154 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3155 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3156 not being looked up in the rDNS to get a host name, Exim was showing the
3157 IP address twice in Received: lines, even though the IP addresses were
3158 identical. For example:
3160 Received: from [1.2.3.4] (helo=[1.2.3.4])
3162 However, if the real host name was known, it was omitting the HELO data
3163 if it matched the actual IP address. This has been tidied up so that it
3164 doesn't show the same IP address twice.
3166 PH/07 When both +timestamp and +memory debugging was on, the value given by
3167 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3168 called by the debug printing, thereby overwriting the timestamp buffer.
3169 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3171 PH/08 When the original message was included in an autoreply transport, it
3172 always said "this is a copy of the message, including all the headers",
3173 even if body_only or headers_only was set. It now gives an appropriate
3176 PH/09 Applied a patch from the Sieve maintainer which:
3178 o fixes some comments
3179 o adds the (disabled) notify extension core
3180 o adds some debug output for the result of if/elsif tests
3181 o points to the current vacation draft in the documentation
3182 and documents the missing references header update
3186 o fixes a bug in processing the envelope test (when testing
3187 multiple envelope elements, the last element determined the
3190 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3191 Electronic Mail") by including:
3193 Auto-submitted: auto-generated
3195 in the messages that it generates (bounce messages and others, such as
3196 warnings). In the case of bounce messages for non-SMTP messages, there was
3197 also a typo: it was using "Auto_submitted" (underscore instead of
3198 hyphen). Since every message generated by Exim is necessarily in response
3199 to another message, thes have all been changed to:
3201 Auto-Submitted: auto-replied
3203 in accordance with these statements in the RFC:
3205 The auto-replied keyword:
3207 - SHOULD be used on messages sent in direct response to another
3208 message by an automatic process,
3210 - MUST NOT be used on manually-generated messages,
3212 - MAY be used on Delivery Status Notifications (DSNs) and Message
3213 Disposition Notifications (MDNs),
3215 - MUST NOT be used on messages generated by automatic or periodic
3216 processes, except for messages which are automatic responses to
3219 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3220 to the default Received: header definition.
3222 PH/12 Added log selector acl_warn_skipped (default on).
3224 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3225 variables because (a) they are in the wrong storage pool and (b) even if
3226 they were copied, it wouldn't work properly because of the caching.
3228 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3229 checking when decoding. Apparently there are clients that generate
3230 overlong encoded strings. Why am I not surprised?
3232 PH/15 If the first argument of "${if match_address" was not empty, but did not
3233 contain an "@" character, Exim crashed. Now it writes a panic log message
3234 and treats the condition as false.
3236 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3238 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3239 extension "envelope-auth". The code is finished and in agreement with
3240 other implementations, but there is no documentation so far and in fact,
3241 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3242 not changing the active code.
3244 Print executed "if" and "elsif" statements when debugging is used. This
3245 helps a great deal to understand what a filter does.
3247 Document more things not specified clearly in RFC3028. I had all this
3248 sorted out, when out of a sudden new issues came to my mind. Oops."
3250 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3253 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3254 canonical form (as documented). However, after a host name lookup from
3255 the IP address, check_host() was doing a simple string comparison with
3256 addresses acquired from the DNS when checking that the found name did
3257 have the original IP as one of its addresses. Since any found IPv6
3258 addresses are likely to be in abbreviated form, the comparison could
3259 fail. Luckily, there already exists a function for doing the comparison
3260 by converting both addresses to binary, so now that is used instead of
3261 the text comparison.
3263 PH/20 There was another similar case to PH/19, when a complete host name was
3264 given in a host list; looking up its IP address could give an abbreviated
3265 form, whereas the current host's name might or might not be abbreviated.
3266 The same fix has been applied.
3272 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3273 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3276 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3277 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3279 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3281 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3282 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3283 Also, for "Auto-Submitted", treat anything other than "no" as
3284 non-personal, in accordance with RFC 3834. (Previously it treated
3285 anything starting "auto-" as non-personal.)
3287 TF/01 The control=submission/name=... option had a problem with syntax
3288 errors if the name included a slash character. The /name= option
3289 now slurps the rest of the string, so it can include any characters
3290 but it must come last in the list of options (after /sender_retain
3293 PH/05 Some modifications to the interface to the fake nameserver for the new
3301 TK/01 Added the "success_on_redirect" address verification option. See
3302 NewStuff for rationale and an example.
3304 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3306 PH/02 Patch to exigrep to allow it to work on syslog lines.
3308 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3309 fread() to skip over the body file's header line, because in Cygwin the
3310 header line is locked and is inaccessible.
3312 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3313 co-exist for some time) to make it clear that it is the Exim ID that is
3314 referenced, not the Message-ID: header line.
3316 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3317 string_format() function, because snprintf() does not exist on all
3320 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3321 ${stat: expansion item.
3323 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3324 protocol synchronization error", to keep the pedants happy.
3326 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3327 well as for IRIX systems, when gcc is being used. See the host.c source
3330 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3332 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3335 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3336 transport and to the smtp transport in LMTP mode.
3338 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3340 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3341 run a filter in a subprocess. This could lead to confusion in subsequent
3342 lookups in the parent process. There should also be a search_tidyup() at
3343 the end of the subprocess.
3345 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3346 only if the host matched helo_try_verify_hosts, which caused the
3347 verification to occur when the EHLO/HELO command was issued. The ACL just
3348 tested the remembered result. Now, if a previous verification attempt has
3349 not happened, "verify = helo" does it there and then.
3351 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3353 TK/03 Fix log output including CR from clamd.
3355 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3356 exist provoked a memory error which could cause a segfault.
3358 PH/15 Installed PCRE 6.2
3360 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3362 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3363 of the problem. Specifically, suggested +O2 rather than +O1 for the
3366 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3368 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3369 record, the port was not correctly logged when the outgoing_port log
3370 selector was set (it logged the transort's default port).
3372 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3373 fallback_hosts, and "hosts" in the smtp transport.
3375 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3376 host errors such as "Connection refused".
3378 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3379 authentication with radiusclient 0.4.9:
3381 - Error returned from rc_read_config was caught wrongly
3382 - Username/password not passed on to radius server due to wrong length.
3384 The presumption is that some radiusclient API changes for 4.51/PH/17
3385 were not taken care of correctly. The code is still untested by me (my
3386 Linux distribution still has 0.3.2 of radiusclient), but it was
3387 contributed by a Radius user.
3389 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3390 expanding the "port" option of the smtp transport.
3392 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3393 while reading a MIME header. Thanks to Tom Hughes for a patch.
3395 PH/24 Include config.h inside local_scan.h so that configuration settings are
3398 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3399 that in an ACL for RCPT (for example), you can examine exactly what was
3402 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3403 commands, but it was not correctly comparing the address with the actual
3404 client host address. Thus, it would show the EHLO address in Received:
3405 header lines when this was not necessary.
3407 PH/27 Added the % operator to ${eval:}.
3409 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3410 it should be ignoring failures (because with -C, for example, it has lost
3411 privilege). It wasn't ignoring creation failures other than "already
3414 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3417 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3418 because an input line was too long, either on its own, or by virtue of
3419 too many continuations, the temporary file was not being removed, and the
3420 return code was incorrect.
3422 PH/31 Missing "BOOL" in function definition in filtertest.c.
3424 PH/32 Applied Sieve patches from the maintainer.
3426 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3428 PH/33 Added "verify = not_blind".
3430 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3431 Local/Makefile (with some defaults set). These are used in built scripts
3432 such as exicyclog, but they have never been used in the exim_install
3433 script (though there are many overriding facilities there). I have
3434 arranged that the exim_install script now takes note of these two
3437 PH/35 Installed configuration files for Dragonfly.
3439 PH/36 When a locally submitted message by a trusted user did not contain a
3440 From: header, and the sender address was obtained from -f or from an SMTP
3441 MAIL command, and the trusted user did not use -F to supply a sender
3442 name, $originator_name was incorrectly used when constructing a From:
3443 header. Furthermore, $originator_name was used for submission mode
3444 messages from external hosts without From: headers in a similar way,
3445 which is clearly wrong.
3447 PH/37 Added control=suppress_local_fixups.
3449 PH/38 When log_selector = +received_sender was set, and the addition of the
3450 sender made the log line's construction buffer exactly full, or one byte
3451 less than full, an overflow happened when the terminating "\n" was
3454 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3455 when the result of a list match is failure because a DNS lookup failed.
3457 PH/40 RM_COMMAND is now used in the building process.
3459 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3460 the "build-* directories that it finds.
3462 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3463 address in a domain literal was a prefix of an interface address.
3465 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3466 when verifying a sender address, unless rewrite_headers is false.
3468 PH/44 Wrote a long comment about why errors_to addresses are verified as
3469 recipients, not senders.
3471 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3472 the ratelimit ACL was added.
3474 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3476 PH/46 Added extra information about PostgreSQL errors to the error string.
3478 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3479 avoiding the need to install special zones in a real server. This is
3480 backwards compatible; if it can't find the fake resolver, it drops back.
3481 Thus, both old and new test suites can be run.
3483 TF/02 Added util/ratelimit.pl
3485 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3486 clock is set back in time.
3488 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3489 Candler <B.Candler@pobox.com>.
3491 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3492 OK for addresses that are the result of redirections.
3494 PH/48 A number of further additions for the benefit of the new test suite,
3495 including a fake gethostbyname() that interfaces to the fake DNS resolver
3498 TF/06 The fix for widen_domains has also been applied to qualify_single and
3499 search_parents which are the other dnslookup options that can cause
3502 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3505 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3507 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3508 service" if it encountered an MX record with an empty target hostname.
3509 The message is now "an MX or SRV record indicated no SMTP service".
3511 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3512 if the DNS of the sending site is misconfigured. This is quite a
3513 common situation. This change restores the behaviour of treating a
3514 helo verification defer as a failure.
3516 PH/51 If self=fail was set on a router, the bounce message did not include the
3517 actual error message.
3523 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3525 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3526 command itself ended in error, the underlying message about the transport
3527 filter timeout was being overwritten with the pipe command error. Now the
3528 underlying error message should be appended to the second error message.
3530 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3532 PH/02 Reduce the amount of output that "make" produces by default. Full output
3533 can still be requested.
3535 PH/03 The warning log line about a condition test deferring for a "warn" verb
3536 was being output only once per connection, rather than after each
3537 occurrence (because it was using the same function as for successful
3538 "warn" verbs). This seems wrong, so I have changed it.
3540 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3541 it should not have, which might have caused a crash in the right
3542 circumstances, but probably never did.
3544 PH/04 Installed a modified version of Tony Finch's patch to make submission
3545 mode fix the return path as well as the Sender: header line, and to
3546 add a /name= option so that you can make the user's friendly name appear
3549 TF/03 Added the control = fakedefer ACL modifier.
3551 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3552 Mark Lowes for thorough testing.
3554 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3556 TK/03 Merged latest SRS patch from Miles Wilton.
3558 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3559 with the definition in sysexits.h (which is #included earlier).
3560 Fortunately, Exim does not actually use EX_OK. The code used to try to
3561 preserve the sysexits.h value, by assuming that macro definitions were
3562 scanned for macro replacements. I have been disabused of this notion,
3563 so now the code just undefines EX_OK before #including unistd.h.
3565 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3566 in the smtp transport. When a block could not be written in a single
3567 write() function, the timeout was being re-applied to each part-write.
3568 This seems wrong - if the receiver was accepting one byte at a time it
3569 would take for ever. The timeout is now adjusted when this happens. It
3570 doesn't have to be particularly precise.
3572 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3573 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3575 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3576 without a domain if the check to <postmaster@domain> fails.
3578 SC/01 Eximstats: added -xls and the ability to specify output files
3579 (patch written by Frank Heydlauf).
3581 SC/02 Eximstats: use FileHandles for outputting results.
3583 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3585 SC/04 Eximstats: fixed display of large numbers with -nvr option
3587 SC/05 Eximstats: fixed merging of reports with empty tables.
3589 SC/06 Eximstats: added the -include_original_destination flag
3591 SC/07 Eximstats: removed tabs and trailing whitespace.
3593 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3595 TK/06 MBOX spool code: Add real "From " MBOX separator line
3596 so the .eml file is really in mbox format (even though
3597 most programs do not really care). Patch from Alex Miller.
3599 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3600 The latter is generated from $received_to and is only set if the
3601 message has one envelope recipient. SA can use these headers,
3602 obviously out-of-the-box. Patch from Alex Miller.
3604 PH/08 The ${def test on a variable was returning false if the variable's
3605 value was "0", contrary to what the specification has always said!
3606 The result should be true unless the variable is empty.
3608 PH/09 The syntax error of a character other than { following "${if
3609 def:variable_name" (after optional whitespace) was not being diagnosed.
3610 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3611 accidental colon was present, for example, could give incorrect results.
3613 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3614 result is used (not including appendfile, where other changes are about
3617 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3618 This involved changing a lot of size variables from int to off_t. It
3619 should work with maildirs and everything.
3621 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3622 spamd dying while we are connected to it.
3624 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3627 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3628 function for BDB 4.3.
3630 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3632 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3633 the sources for PCRE have been re-arranged and more files are now
3636 PH/15 The code I had for printing potentially long long variables in PH/11
3637 above was not the best (it lost precision). The length of off_t variables
3638 is now inspected at build time, and an appropriate printing format (%ld
3639 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3640 to be "long long int" or "long int". This is needed for the internal
3641 formatting function string_vformat().
3643 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3644 the configuration file to be ":syslog", then the script "guesses" where
3645 the logs files are, rather than using the compiled in default. In our
3646 case the guess is not the same as the compiled default, so the script
3647 suddenly stopped working when I started to use syslog. The patch checks
3648 to see if log_file_path is "". If so, it attempts to read it from exim
3649 with no configuration file to get the compiled in version, before it
3650 falls back to the previous guessing code."
3652 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3653 implementing BATV in an Exim configuration. See NewStuff for the gory
3656 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3657 Makefile that are specific to HP-UX.
3659 PH/18 If the "use_postmaster" option was set for a recipient callout together
3660 with the "random" option, the postmaster address was used as the MAIL
3661 FROM address for the random test, but not for the subsequent recipient
3662 test. It is now used for both.
3664 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3665 patch removes a few documentation additions to RFC 3028, because the
3666 latest draft now contains them. It adds the new en;ascii-case comparator
3667 and a new error check for 8bit text in MIME parts. Comparator and
3668 require names are now matched exactly. I enabled the subaddress
3669 extension, but it is not well tested yet (read: it works for me)."
3671 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3672 rework some of the code of TK/09 above to avoid the hardwired use of
3673 "%lld" and "long long". Replaced the call to snprintf() with a call to
3676 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3677 records point to non-existent hosts", "retry timeout exceeded", and
3678 "retry time not reached for any host after a long failure period".
3680 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3681 experimental DomainKeys support:
3683 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3684 (2) On an error such as an illegally used "control", the wrong name for
3685 the control was given.
3687 These problems did NOT occur unless DomainKeys support was compiled.
3689 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3691 PH/24 Added ${if match_ip condition.
3693 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3694 databases so that it will be absolutely obvious if a crash occurs in the
3695 DB library. This is a regular occurrence (often caused by mis-matched
3698 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3699 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3700 picked up on a user's system that detects such things. There doesn't seem
3701 to be a gcc warning option for this - only an attribute that has to be
3702 put on the function's prototype. It seems that in Fedora Core 4 they have
3703 set this on a number of new functions. No doubt there will be more in due
3706 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3707 specify a transport. However, if an address that was verified by such a
3708 router was the subject of a callout, Exim crashed because it tried to
3709 read the rcpt_include_affixes from the non-existent transport. Now it
3710 just assumes that the setting of that option is false. This bug was
3711 introduced by 4.51/PH/31.
3713 PH/28 Changed -d+all to exclude +memory, because that information is very
3714 rarely of interest, but it makes the output a lot bigger. People tend to
3715 do -d+all out of habit.
3717 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3718 code in os-type was giving problems when libc.so lives in lib64, like on
3721 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3722 aren't the modern standard, and it seems that some systems' include files
3723 don't always have them. Exim was already checking for some of the newer
3724 ones like T_AAAA, and defining it itself. I've added checks for all the
3725 record types that Exim uses.
3727 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3728 not automatically generating a new one, as it is supposed to. This
3729 prevented TLS from working. If the file did exist, but contained invalid
3730 data, a new version was generated, as expected. It was only the case of a
3731 non-existent file that was broken.
3733 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3734 with a change in libdomainkeys > 0.64.
3736 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3737 from DNS. If the selector record carries the flag, it now has
3738 precedence over the domain-wide flag.
3740 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3742 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3743 the use of a port name that isn't defined in /etc/services) occurred, the
3744 message was deferred as in a normal delivery, and thus remained on the
3745 spool, instead of being failed because of the mua_wrapper setting. This
3746 is now fixed, and I tidied up some of the mua_wrapper messages at the
3749 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3750 the messages in a hash of arrays rather than using individual hashes.
3751 This is a bit cleaner and results in dramatic memory savings, albeit
3752 at a slight CPU cost.
3754 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3755 as requested by Marc Sherman.
3757 SC/10 Eximstats: added histograms for user specified patterns as requested
3760 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3762 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3763 fopen() in the content-scanning modules that did not already have it.
3769 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3770 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3772 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3774 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3776 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3777 to test Sieve filters that use "vacation".
3779 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3780 that changes the way the GnuTLS parameters are stored in the cache file.
3781 The new format can be generated externally. For backward compatibility,
3782 if the data in the cache doesn't make sense, Exim assumes it has read an
3783 old-format file, and it generates new data and writes a new file. This
3784 means that you can't go back to an older release without removing the
3787 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3788 work if there are any delivery delays because "one_time" forces the
3789 parent to be marked "delivered", so its unseen clone is never tried
3790 again. For this reason, Exim now forbids the simultaneous setting of
3793 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3794 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3795 entry if you want to know the details. The fix, however, neglected to
3796 consider the case when local delivery batching is involved. The test for
3797 "previously delivered" was not happening when checking to see if an
3798 address could be batched with a previous (undelivered) one; under
3799 certain circumstances this could lead to multiple deliveries to the same
3802 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3803 in its include files, and this causes problems building Exim.
3805 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3806 header_syntax) but Exim was just ignoring anything given after a slash.
3807 In particular, this caused confusion with an attempt to use "verify =
3808 reverse_host_lookup/defer_ok". An error is now given when options are
3809 supplied for verify items that do not have them. (Maybe reverse_host_
3810 lookup should have a defer_ok option, but that's a different point.)
3812 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3813 defined by RFC 821) to 2048, because there were problems with some AUTH
3814 commands, and RFC 1869 says the size should be increased for extended
3815 SMTP commands that take arguments.
3817 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3820 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3821 "unknown" error; now it says that the functionality isn't in the binary.
3823 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3824 an address' error message when a string expansion fails (syntax or
3825 whatever). Otherwise the password may appear in the log. Following change
3826 PH/42 below, there is no longer a chance of it appearing in a bounce
3829 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3831 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3832 of its addresses was ever tried. (Bugzilla bug #2.)
3834 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3835 the result incorrectly in the debug output. (It correctly added a newline
3836 to what was transported.)
3838 TF/01 Added $received_time.
3840 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3841 commented out examples of how to interface to a virus scanner and to
3842 SpamAssassin. Also added commented examples of av_scanner and
3843 spamd_address settings.
3845 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3846 and controls are allowed in which ACLs. There were a couple of minor
3847 errors. Some of the entries in the conditions table (which is a table of
3848 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3849 negation of where the condition IS allowed.
3851 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3853 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3854 header file does not have a version number, so I've had to invent a new
3855 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3856 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3857 radiusclient), but it was contributed by a Radius user.
3859 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3860 files or directories, for OpenSSL.
3862 PH/19 When an Exim process that is running as root has to create an Exim log
3863 file, it does so in a subprocess that runs as exim:exim so as to get the
3864 ownership right at creation (otherwise, other Exim processes might see
3865 the file with the wrong ownership). There was no test for failure of this
3866 fork() call, which would lead to the process getting stuck as it waited
3867 for a non-existent subprocess. Forks do occasionally fail when resources
3868 run out. I reviewed all the other calls to fork(); they all seem to check
3871 PH/20 When checking for unexpected SMTP input at connect time (before writing
3872 the banner), Exim was not dealing correctly with a non-positive return
3873 from the read() function. If the client had disconnected by this time,
3874 the result was a log entry for a synchronization error with an empty
3875 string after "input=" when read() returned zero. If read() returned -1
3876 (an event I could not check), uninitialized data bytes were printed.
3877 There were reports of junk text (parts of files, etc) appearing after
3880 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3882 PH/22 Added support for macro redefinition, and (re)definition in between
3883 driver and ACL definitions.
3885 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3886 forgetting to use the resulting value; it was using the unexpanded value.
3888 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3889 hadn't been configured. The fix is from Juergen Kreileder, who
3890 understands it better than I do:
3892 "Here's what I see happening with three configured cyrus_sasl
3893 authenticators configured (plain, login, cram-md5):
3895 On startup auth_cyrus_sasl_init() gets called for each of these.
3896 This means three calls to sasl_listmech() without a specified mech_list.
3897 => SASL tests which mechs of all available mechs actually work
3898 => three warnings about OTP not working
3899 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3901 With the patch, sasl_listmech() also gets called three times. But now
3902 SASL's mech_list option is set to the server_mech specified in the the
3903 authenticator. Or in other words, the answer from sasl_listmech()
3904 gets limited to just the mech you're testing for (which is different
3906 => the return list contains just 'plain' or 'login', 'cram-md5' or
3907 nothing depending on the value of ob->server_mech.
3909 I've just tested the patch: Authentication still works fine,
3910 unavailable mechs specified in the exim configuration are still
3911 caught, and the auth.log warnings about OTP are gone."
3913 PH/25 When debugging is enabled, the contents of the command line are added
3914 to the debugging output, even when log_selector=+arguments is not
3917 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3918 answer is "GNU", and only if the return is "GNU/something" is the answer
3921 PH/27 $acl_verify_message is now set immediately after the failure of a
3922 verification in an ACL, and so is available in subsequent modifiers. In
3923 particular, the message can be preserved by coding like this:
3925 warn !verify = sender
3926 set acl_m0 = $acl_verify_message
3928 Previously, $acl_verify_message was set only while expanding "message"
3929 and "log_message" when a very denied access.
3931 PH/28 Modified OS/os.c-Linux with
3933 -#ifndef OS_LOAD_AVERAGE
3934 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3936 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3937 nomenclature these days.)
3939 PH/29 Installed patch from the Sieve maintainer that adds the options
3940 sieve_useraddress and sieve_subaddress to the redirect router.
3942 PH/30 In these circumstances:
3943 . Two addresses routed to the same list of hosts;
3944 . First host does not offer TLS;
3945 . First host accepts first address;
3946 . First host gives temporary error to second address;
3947 . Second host offers TLS and a TLS session is established;
3948 . Second host accepts second address.
3949 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3950 and peerdn, if requested) that were in fact used only for the second
3953 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3954 attention to any local part prefix or suffix that was matched by the
3955 router that accepted the address. It now behaves in the same way as it
3956 does for delivery: the affixes are removed from the local part unless
3957 rcpt_include_affixes is set on the transport.
3959 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3960 timeout during the DATA phase of an incoming message.
3962 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3963 applied a patch sanctioned by the Sieve maintainer.
3965 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3966 the uid or gid is negative. A case of a negative gid caused this to be
3967 noticed. The fix allows for either to be negative.
3969 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3970 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3973 PH/36 The change PH/12 above was broken. Fixed it.
3975 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3976 the grounds that routing the same address twice would always produce the
3977 same answer. This might have been true once, but it is certainly no
3978 longer true now. Routing a child address may depend on the previous
3979 routing that produced that child. Some complicated redirection strategies
3980 went wrong when messages had multiple recipients, and made Exim's
3981 behaviour dependent on the order in which the addresses were given.
3983 I have moved the duplicate checking until after the routing is complete.
3984 Exim scans the addresses that are assigned to local and remote
3985 transports, and removes any duplicates. This means that more work will be
3986 done, as duplicates will always all be routed, but duplicates are
3987 presumably rare, so I don't expect this is of any significance.
3989 For deliveries to pipes, files, and autoreplies, the duplicate checking
3990 still happens during the routing process, since they are not going to be
3993 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3994 It corrects a timeout issue with spamd. This is Ian's comment: "The
3995 background is that sometimes spamd either never reads data from a
3996 connection it has accepted, or it never writes response data. The exiscan
3997 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3998 blindly assumes that writes won't block so it may never time out."
4000 PH/39 Allow G after quota size as well as K and M.
4002 PH/40 The value set for $authenticated_id in an authenticator may not contain
4003 binary zeroes or newlines because the value is written to log lines and
4004 to spool files. There was no check on this. Now the value is run through
4005 the string_printing() function so that such characters are converted to
4006 printable escape sequences.
4008 PH/41 $message_linecount is a new variable that contains the total number of
4009 lines in the message. Compare $body_linecount, which is the count for the
4012 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4013 bounce and delay warning messages, except in certain special cases, which
4016 (a) An SMTP error message from a remote host;
4017 (b) A message specified in a :fail: redirection;
4018 (c) A message specified in a "fail" command in a system filter;
4019 (d) A message specified in a FAIL return from the queryprogram router;
4020 (e) A message specified by the cannot_route_message router option.
4022 In these cases only, Exim does include the error details in bounce and
4023 warning messages. There are also a few cases where bland messages such
4024 as "unrouteable address" or "local delivery error" are given.
4026 PH/43 $value is now also set for the "else" part of a ${run expansion.
4028 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4029 being worked on, but at least Exim now implements the latest version to
4032 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4033 process to complete was treated as a delivery failure, a timeout while
4034 writing the message to the pipe was logged, but erroneously treated as a
4035 successful delivery. Such timeouts include transport filter timeouts. For
4036 consistency with the overall process timeout, these timeouts are now
4037 treated as errors, giving rise to delivery failures by default. However,
4038 there is now a new Boolean option for the pipe transport called
4039 timeout_defer, which, if set TRUE, converts the failures into defers for
4040 both kinds of timeout. A transport filter timeout is now identified in
4043 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4044 systems where "make" and "gmake" are different, calling "gmake" at top
4045 level broke things. I've arranged for the value of $(MAKE) to be passed
4046 from the Makefile to this script so that it can call the same version of
4050 A note about Exim versions 4.44 and 4.50
4051 ----------------------------------------
4053 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4054 changes of various kinds. As a consequence, a big documentation update was
4055 needed. This delayed the release for rather longer than seemed good, especially
4056 in the light of a couple of (minor) security issues. Therefore, the changes
4057 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4058 release. So 4.44 and 4.50 are in effect two different branches that both start
4061 I have left the 4.50 change log unchanged; it contains all the changes since
4062 4.43. The change log for 4.44 is below; many of its items are identical to
4063 those for 4.50. This seems to be the most sensible way to preserve the
4064 historical information.
4070 1. Minor wording change to the doc/README.SIEVE file.
4072 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4073 computation of the current number of files was incorrect.
4075 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4076 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4079 4. Give more explanation in the error message when the command for a transport
4080 filter fails to execute.
4082 5. There are several places where Exim runs a non-Exim command in a
4083 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4084 was being done only for the command run by the queryprogram router. It is
4085 now done for all such subprocesses. The other cases are: ${run, transport
4086 filters, and the commands run by the lmtp and pipe transports.
4088 6. Added CONFIGURE_GROUP build-time option.
4090 7. Some older OS have a limit of 256 on the maximum number of file
4091 descriptors. Exim was using setrlimit() to set 1000 as a large value
4092 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4093 systems. I've change it so that if it can't get 1000, it tries for 256.
4095 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4096 was an oversight, and furthermore, ever since the addition of extra
4097 controls (e.g. 4.43/32), the checks on when to allow different forms of
4098 "control" were broken. There should now be diagnostics for all cases when a
4099 control that does not make sense is encountered.
4101 9. Added the /retain_sender option to "control=submission".
4103 10. $recipients is now available in the predata ACL (oversight).
4105 11. Tidy the search cache before the fork to do a delivery from a message
4106 received from the command line. Otherwise the child will trigger a lookup
4107 failure and thereby defer the delivery if it tries to use (for example) a
4108 cached ldap connection that the parent has called unbind on.
4110 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4111 of $address_data from the recipient verification was clobbered by the
4112 sender verification.
4114 13. The value of address_data from a sender verification is now available in
4115 $sender_address_data in subsequent conditions in the ACL statement.
4117 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4119 15. Added a new option "connect=<time>" to callout options, to set a different
4122 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4123 was its contents. (It was OK if the option was not defined at all.)
4125 17. A "Completed" log line is now written for messages that are removed from
4126 the spool by the -Mrm option.
4128 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4129 information about exactly what failed.
4131 19. Added -dd to debug only the daemon process.
4133 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4134 handles timeouts, both on the server side and network timeouts. Renamed the
4135 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4137 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4138 It is now set to "smtps".
4140 22. $host_address is now set to the target address during the checking of
4141 ignore_target_hosts.
4143 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4144 being passed; this would have caused $sender_host_name to have been used if
4145 matching the list had actually called for a host name (not very likely,
4146 since this list is usually IP addresses). A host name is now passed as
4149 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4150 code that specifies a non-restarting handler (typically sigaction() in
4151 modern systems) in an attempt to fix a rare and obscure crash bug.
4153 25. Narrowed the window for a race in the daemon that could cause it to ignore
4154 SIGCHLD signals. This is not a major problem, because they are used only to
4155 wake it up if nothing else does.
4157 26. A malformed maildirsize file could cause Exim to calculate negative values
4158 for the mailbox size or file count. Odd effects could occur as a result.
4159 The maildirsize information is now recalculated if the size or filecount
4162 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4163 support for a long time. Removed HAVE_SYS_VFS_H.
4165 28. Installed the latest version of exipick from John Jetmore.
4167 29. In an address list, if the pattern was not a regular expression, an empty
4168 subject address (from a bounce message) matched only if the pattern was an
4169 empty string. Non-empty patterns were not even tested. This was the wrong
4170 because it is perfectly reasonable to use an empty address as part of a
4171 database query. An empty address is now tested by patterns that are
4172 lookups. However, all the other forms of pattern expect the subject to
4173 contain a local part and a domain, and therefore, for them, an empty
4174 address still always fails if the pattern is not itself empty.
4176 30. Exim went into a mad DNS loop when attempting to do a callout where the
4177 host was specified on an smtp transport, and looking it up yielded more
4178 than one IP address.
4180 31. Re-factored the code for checking spool and log partition space into a
4181 function that finds that data and another that does the check. The former
4182 is then used to implement four new variables: $spool_space, $log_space,
4183 $spool_inodes, and $log_inodes.
4185 32. The RFC2047 encoding function was originally intended for short strings
4186 such as real names; it was not keeping to the 75-character limit for
4187 encoded words that the RFC imposes. It now respects the limit, and
4188 generates multiple encoded words if necessary. To be on the safe side, I
4189 have increased the buffer size for the ${rfc2047: expansion operator from
4192 33. It is now permitted to omit both strings after an "if" condition; if the
4193 condition is true, the result is "true". As before, when the second string
4194 is omitted, a false condition yields an empty string. This makes it less
4195 cumbersome to write custom ACL and router conditions.
4197 34. Failure to deliver a bounce message always caused it to be frozen, even if
4198 there was an errors_to setting on the router. The errors_to setting is now
4201 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4202 canonical form (fully expanded) before being placed in
4203 $sender_host_address.
4205 36. The table in the code that translates DNS record types into text (T_A to
4206 "A" for instance) was missing entries for NS and CNAME. It is just possible
4207 that this could have caused confusion if both these types were looked up
4208 for the same domain, because the text type is used as part of Exim's
4209 per-process caching. But the chance of anyone hitting this buglet seems
4212 37. The dnsdb lookup has been extended in a number of ways.
4214 (1) There is a new type, "zns", which walks up the domain tree until it
4215 finds some nameserver records. It should be used with care.
4217 (2) There is a new type, "mxh", which is like "mx" except that it returns
4218 just the host names, not the priorities.
4220 (3) It is now possible to give a list of domains (or IP addresses) to be
4221 looked up. The behaviour when one of the lookups defers can be
4222 controlled by a keyword.
4224 (4) It is now possible to specify the separator character for use when
4225 multiple records are returned.
4227 38. The dnslists ACL condition has been extended: it is now possible to supply
4228 a list of IP addresses and/or domains to be looked up in a particular DNS
4231 39. Added log_selector=+queue_time_overall.
4233 40. When running the queue in the test harness, wait just a tad after forking a
4234 delivery process, to get repeatability of debugging output.
4236 41. Include certificate and key file names in error message when GnuTLS fails
4237 to set them up, because the GnuTLS error message doesn't include the name
4238 of the failing file when there is a problem reading it.
4240 42. Allow both -bf and -bF in the same test run.
4242 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4244 44. The "Exiscan patch" is now merged into the mainline Exim source.
4246 45. Sometimes the final signoff response after QUIT could fail to get
4247 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4248 < 0 before doing a fflush(). This bug looks as though it goes back to the
4249 introduction of TLS in release 3.20, but "sometimes" must have been rare
4250 because the tests only now provoked it.
4252 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4253 (this can affect the format of dates).
4255 47. exim_tidydb, when checking for the continued existence of a message for
4256 which it has found a message-specific retry record, was not finding
4257 messages that were in split spool directories. Consequently, it was
4258 deleting retry records that should have stayed in existence.
4260 48. Steve fixed some bugs in eximstats.
4262 49. The SPA authentication driver was not abandoning authentication and moving
4263 on to the next authenticator when an expansion was forced to fail,
4264 contradicting the general specification for all authenticators. Instead it
4265 was generating a temporary error. It now behaves as specified.
4267 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4268 (the order specifies the preference for clients). The order is now AES256,
4269 AES128, 3DES, ARCFOUR128.
4271 51. Small patch to Sieve code - explicitly set From: when generating an
4274 52. Exim crashed if a remote delivery caused a very long error message to be
4275 recorded - for instance if somebody sent an entire SpamAssassin report back
4276 as a large number of 550 error lines. This bug was coincidentally fixed by
4277 increasing the size of one of Exim's internal buffers (big_buffer) that
4278 happened as part of the Exiscan merge. However, to be on the safe side, I
4279 have made the code more robust (and fixed the comments that describe what
4282 53. Now that there can be additional text after "Completed" in log lines (if
4283 the queue_time_overall log selector is set), a one-byte patch to exigrep
4284 was needed to allow it to recognize "Completed" as not the last thing in
4287 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4288 patch that reportedly fixes this has been added. I am not expert enough to
4289 create a test for it. This is what the patch creator wrote:
4291 "I found a little strange behaviour of ldap code when working with
4292 Windows 2003 AD Domain, where users was placed in more than one
4293 Organization Units. When I tried to give exim partial DN, the exit code
4294 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4295 But simultaneously result of request was absolutely normal ldap result,
4296 so I produce this patch..."
4298 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4299 REFERENCE, so I have modified the code to exclude the patch when that macro
4302 55. Some experimental protocols are using DNS PTR records for new purposes. The
4303 keys for these records are domain names, not reversed IP addresses. The
4304 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4305 leaves it alone. Component reversal etc. now happens only for IP addresses.
4308 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4310 57. Double the size of the debug message buffer (to 2048) so that more of very
4311 long debug lines gets shown.
4313 58. The exicyclog utility now does better if the number of log files to keep
4314 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4316 59. Two changes related to the smtp_active_hostname option:
4318 (1) $smtp_active_hostname is now available as a variable.
4319 (2) The default for smtp_banner uses $smtp_active_hostname instead
4320 of $primary_hostname.
4322 60. The host_aton() function is supposed to be passed a string that is known
4323 to be a valid IP address. However, in the case of IPv6 addresses, it was
4324 not checking this. This is a hostage to fortune. Exim now panics and dies
4325 if the condition is not met. A case was found where this could be provoked
4326 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4327 components; fortuitously, this particular loophole had already been fixed
4328 by change 4.50/55 above.
4330 If there are any other similar loopholes, the new check in host_aton()
4331 itself should stop them being exploited. The report I received stated that
4332 data on the command line could provoke the exploit when Exim was running as
4333 exim, but did not say which command line option was involved. All I could
4334 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4335 running as the user.
4338 61. There was a buffer overflow vulnerability in the SPA authentication code
4339 (which came originally from the Samba project). I have added a test to the
4340 spa_base64_to_bits() function which I hope fixes it.
4343 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4344 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4346 63. The daemon start-up calls getloadavg() while still root for those OS that
4347 need the first call to be done as root, but it missed one case: when
4348 deliver_queue_load_max is set with deliver_drop_privilege. This is
4349 necessary for the benefit of the queue runner, because there is no re-exec
4350 when deliver_drop_privilege is set.
4352 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4353 This has been fixed.
4355 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4356 was in use, was not putting the data itself into the right store pool;
4357 consequently, it could be overwritten for a subsequent message in the same
4358 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4361 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4363 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4364 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4365 the function were treating the return as a boolean value, which happened to
4366 work because 0=false and not-0=true, but is not correct code.
4368 68. The host_aton() function was not handling scoped IPv6 addresses (those
4369 with, for example, "%eth0" on the end) correctly.
4371 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4372 negated items (that is, ~something) in unsigned ints. Some compilers
4373 apparently mutter when there is no cast.
4375 70. If an address verification called from an ACL failed, and did not produce a
4376 user-specific message (i.e. there was only a "system" message), nothing was
4377 put in $acl_verify_message. In this situation, it now puts the system
4380 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4381 session; change 4.31/43 added the unwanted input to the log line - except
4382 that it did not do this in the start of session case. It now does.
4384 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4385 This is wrong and can cause the other end to generate a synchronization
4386 error if it is another Exim or anything else that does the synchronization
4387 check. A QUIT command is no longer sent after a timeout.
4389 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4390 during host lookups.
4392 74. The defer_ok option of callout verification was not working if it was used
4393 when verifying addresses in header lines, that is, for this case:
4395 verify = header_sender/callout=defer_ok
4397 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4398 those file descriptors could be used for SMTP connections. If anything
4399 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4400 could be sent to the SMTP client, causing chaos. The daemon now opens
4401 stdin, stdout, and stderr to /dev/null when it puts itself into the
4404 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4405 log by default. The user can override this with suitable Perl magic.
4407 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4408 the log message when discard triggers, was not working for the DATA ACL or
4409 for the non-SMTP ACL.
4411 78. Error message wording change in sieve.c.
4413 79. If smtp_accept_max_per_host was set, the number of connections could be
4414 restricted to fewer than expected, because the daemon was trying to set up
4415 a new connection before checking whether the processes handling previous
4416 connections had finished. The check for completed processes is now done
4417 earlier. On busy systems, this bug wouldn't be noticed because something
4418 else would have woken the daemon, and it would have reaped the completed
4421 80. If a message was submitted locally by a user whose login name contained one
4422 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4423 It caused a spool format error. I have fixed the spool reading code. A
4424 related problem was that the "from" clause in the Received: line became
4425 illegal because of the space(s). It is now covered by ${quote_local_part.
4427 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4430 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4432 83. Patch from Sieve maintainer to fix unterminated string problem in
4433 "vacation" handling.
4435 84. Some minor changes to the Linux configuration files to help with other
4436 OS variants using glibc.
4438 85. One more patch for Sieve to update vacation handling to latest spec.
4441 ----------------------------------------------------
4442 See the note above about the 4.44 and 4.50 releases.
4443 ----------------------------------------------------
4449 1. Change 4.43/35 introduced a bug that caused file counts to be
4450 incorrectly computed when quota_filecount was set in an appendfile
4453 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4454 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4457 3. Give more explanation in the error message when the command for a transport
4458 filter fails to execute.
4460 4. There are several places where Exim runs a non-Exim command in a
4461 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4462 was being done only for the command run by the queryprogram router. It is
4463 now done for all such subprocesses. The other cases are: ${run, transport
4464 filters, and the commands run by the lmtp and pipe transports.
4466 5. Some older OS have a limit of 256 on the maximum number of file
4467 descriptors. Exim was using setrlimit() to set 1000 as a large value
4468 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4469 systems. I've change it so that if it can't get 1000, it tries for 256.
4471 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4472 was an oversight, and furthermore, ever since the addition of extra
4473 controls (e.g. 4.43/32), the checks on when to allow different forms of
4474 "control" were broken. There should now be diagnostics for all cases when a
4475 control that does not make sense is encountered.
4477 7. $recipients is now available in the predata ACL (oversight).
4479 8. Tidy the search cache before the fork to do a delivery from a message
4480 received from the command line. Otherwise the child will trigger a lookup
4481 failure and thereby defer the delivery if it tries to use (for example) a
4482 cached ldap connection that the parent has called unbind on.
4484 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4485 of $address_data from the recipient verification was clobbered by the
4486 sender verification.
4488 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4489 was its contents. (It was OK if the option was not defined at all.)
4491 11. A "Completed" log line is now written for messages that are removed from
4492 the spool by the -Mrm option.
4494 12. $host_address is now set to the target address during the checking of
4495 ignore_target_hosts.
4497 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4498 being passed; this would have caused $sender_host_name to have been used if
4499 matching the list had actually called for a host name (not very likely,
4500 since this list is usually IP addresses). A host name is now passed as
4503 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4504 code that specifies a non-restarting handler (typically sigaction() in
4505 modern systems) in an attempt to fix a rare and obscure crash bug.
4507 15. Narrowed the window for a race in the daemon that could cause it to ignore
4508 SIGCHLD signals. This is not a major problem, because they are used only to
4509 wake it up if nothing else does.
4511 16. A malformed maildirsize file could cause Exim to calculate negative values
4512 for the mailbox size or file count. Odd effects could occur as a result.
4513 The maildirsize information is now recalculated if the size or filecount
4516 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4517 support for a long time. Removed HAVE_SYS_VFS_H.
4519 18. Updated exipick to current release from John Jetmore.
4521 19. Allow an empty sender to be matched against a lookup in an address list.
4522 Previously the only cases considered were a regular expression, or an
4525 20. Exim went into a mad DNS lookup loop when doing a callout where the
4526 host was specified on the transport, if the DNS lookup yielded more than
4529 21. The RFC2047 encoding function was originally intended for short strings
4530 such as real names; it was not keeping to the 75-character limit for
4531 encoded words that the RFC imposes. It now respects the limit, and
4532 generates multiple encoded words if necessary. To be on the safe side, I
4533 have increased the buffer size for the ${rfc2047: expansion operator from
4536 22. Failure to deliver a bounce message always caused it to be frozen, even if
4537 there was an errors_to setting on the router. The errors_to setting is now
4540 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4541 canonical form (fully expanded) before being placed in
4542 $sender_host_address.
4544 24. Updated eximstats to version 1.33
4546 25. Include certificate and key file names in error message when GnuTLS fails
4547 to set them up, because the GnuTLS error message doesn't include the name
4548 of the failing file when there is a problem reading it.
4550 26. Expand error message when OpenSSL has problems setting up cert/key files.
4553 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4554 (this can affect the format of dates).
4556 28. exim_tidydb, when checking for the continued existence of a message for
4557 which it has found a message-specific retry record, was not finding
4558 messages that were in split spool directories. Consequently, it was
4559 deleting retry records that should have stayed in existence.
4561 29. eximstats updated to version 1.35
4562 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4563 1.35 - bugfix such that pie charts by volume are generated correctly
4565 30. The SPA authentication driver was not abandoning authentication and moving
4566 on to the next authenticator when an expansion was forced to fail,
4567 contradicting the general specification for all authenticators. Instead it
4568 was generating a temporary error. It now behaves as specified.
4570 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4571 (the order specifies the preference for clients). The order is now AES256,
4572 AES128, 3DES, ARCFOUR128.
4574 31. Small patch to Sieve code - explicitly set From: when generating an
4577 32. Exim crashed if a remote delivery caused a very long error message to be
4578 recorded - for instance if somebody sent an entire SpamAssassin report back
4579 as a large number of 550 error lines. This bug was coincidentally fixed by
4580 increasing the size of one of Exim's internal buffers (big_buffer) that
4581 happened as part of the Exiscan merge. However, to be on the safe side, I
4582 have made the code more robust (and fixed the comments that describe what
4585 33. Some experimental protocols are using DNS PTR records for new purposes. The
4586 keys for these records are domain names, not reversed IP addresses. The
4587 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4588 leaves it alone. Component reversal etc. now happens only for IP addresses.
4591 34. The host_aton() function is supposed to be passed a string that is known
4592 to be a valid IP address. However, in the case of IPv6 addresses, it was
4593 not checking this. This is a hostage to fortune. Exim now panics and dies
4594 if the condition is not met. A case was found where this could be provoked
4595 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4596 components; fortuitously, this particular loophole had already been fixed
4597 by change 4.50/55 or 4.44/33 above.
4599 If there are any other similar loopholes, the new check in host_aton()
4600 itself should stop them being exploited. The report I received stated that
4601 data on the command line could provoke the exploit when Exim was running as
4602 exim, but did not say which command line option was involved. All I could
4603 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4604 running as the user.
4607 35. There was a buffer overflow vulnerability in the SPA authentication code
4608 (which came originally from the Samba project). I have added a test to the
4609 spa_base64_to_bits() function which I hope fixes it.
4612 36. The daemon start-up calls getloadavg() while still root for those OS that
4613 need the first call to be done as root, but it missed one case: when
4614 deliver_queue_load_max is set with deliver_drop_privilege. This is
4615 necessary for the benefit of the queue runner, because there is no re-exec
4616 when deliver_drop_privilege is set.
4618 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4619 was in use, was not putting the data itself into the right store pool;
4620 consequently, it could be overwritten for a subsequent message in the same
4621 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4624 38. Sometimes the final signoff response after QUIT could fail to get
4625 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4626 < 0 before doing a fflush(). This bug looks as though it goes back to the
4627 introduction of TLS in release 3.20, but "sometimes" must have been rare
4628 because the tests only now provoked it.
4634 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4635 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4636 Now it returns an integer. A number of calls were still expecting a T/F
4637 return. Fortuitously, in all cases, the tests worked in OK situations,
4638 which is the norm. However, things would have gone wrong on any write
4639 failures on the smtp file descriptor. This function is used when sending
4640 messages over SMTP and also when doing verify callouts.
4642 2. When Exim is called to do synchronous delivery of a locally submitted
4643 message (the -odf or -odi options), it no longer closes stderr before doing
4646 3. Implemented the mua_wrapper option.
4648 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4650 5. Implemented the functions header_remove(), header_testname(),
4651 header_add_at_position(), and receive_remove_recipient(), and exported them
4654 6. If an ACL "warn" statement specified the addition of headers, Exim already
4655 inserted X-ACL-Warn: at the start if there was no header name. However, it
4656 was not making this test for the second and subsequent header lines if
4657 there were newlines in the string. This meant that an invalid header could
4658 be inserted if Exim was badly configured.
4660 7. Allow an ACL "warn" statement to add header lines at the start or after all
4661 the Received: headers, as well as at the end.
4663 8. Added the rcpt_4xx retry error code.
4665 9. Added postmaster_mailfrom=xxx to callout verification option.
4667 10. Added mailfrom=xxxx to the callout verification option, for verify=
4670 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4671 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4673 12. Inserted some casts to stop certain compilers warning when using pointer
4674 differences as field lengths or precisions in printf-type calls (mostly
4675 affecting debugging statements).
4677 13. Added optional readline() support for -be (dynamically loaded).
4679 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4680 same clock tick as a message's arrival, so that its received time was the
4681 same as the "first fail" time on the retry record, and that message
4682 remained on the queue past the ultimate address timeout, every queue runner
4683 would try a delivery (because it was past the ultimate address timeout) but
4684 after another failure, the ultimate address timeout, which should have then
4685 bounced the address, did not kick in. This was a "< instead of <=" error;
4686 in most cases the first failure would have been in the next clock tick
4687 after the received time, and all would be well.
4689 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4690 being recognized when the domain list was tested by the match_domain
4691 condition in an expansion string.
4693 16. Added the ${str2b64: operator.
4695 17. Exim was always calling setrlimit() to set a large limit for the number of
4696 processes, without checking whether the existing limit was already
4697 adequate. (It did check for the limit on file descriptors.) Furthermore,
4698 errors from getrlimit() and setrlimit() were being ignored. Now they are
4699 logged to the main and panic logs, but Exim does carry on, to try to do its
4700 job under whatever limits there are.
4702 18. Imported PCRE 5.0.
4704 19. Trivial typo in log message " temporarily refused connection" (the leading
4707 20. If the log selector return_path_on_delivery was set and an address was
4708 redirected to /dev/null, the delivery process crashed because it assumed
4709 that a return path would always be set for a "successful" delivery. In this
4710 case, the whole delivery is bypassed as an optimization, and therefore no
4713 21. Internal re-arrangement: the function for sending a challenge and reading
4714 a response while authentication was assuming a zero-terminated challenge
4715 string. It's now changed to take a pointer and a length, to allow for
4716 binary data in such strings.
4718 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4720 23. Exim was not respecting finduser_retries when seeking the login of the
4721 uid under which it was called; it was always trying 10 times. (The default
4722 setting of finduser_retries is zero.) Also, it was sleeping after the final
4723 failure, which is pointless.
4725 24. Implemented tls_on_connect_ports.
4727 25. Implemented acl_smtp_predata.
4729 26. If the domain in control=submission is set empty, Exim assumes that the
4730 authenticated id is a complete email address when it generates From: or
4731 Sender: header lines.
4733 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4734 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4735 chown and chgrp in /bin and hostname in /usr/bin.
4737 28. Exim was keeping the "process log" file open after each use, just as it
4738 does for the main log. This opens the possibility of it remaining open for
4739 long periods when the USR1 signal hits a daemon. Occasional processlog
4740 errors were reported, that could have been caused by this. Anyway, it seems
4741 much more sensible not to leave this file open at all, so that is what now
4744 29. The long-running daemon process does not normally write to the log once it
4745 has entered its main loop, and it closes the log before doing so. This is
4746 so that log files can straightforwardly be renamed and moved. However,
4747 there are a couple of unusual error situations where the daemon does write
4748 log entries, and I had neglected to close the log afterwards.
4750 30. The text of an SMTP error response that was received during a remote
4751 delivery was being truncated at 512 bytes. This is too short for some of
4752 the long messages that one sometimes sees. I've increased the limit to
4755 31. It is now possible to make retry rules that apply only when a message has a
4756 specific sender, in particular, an empty sender.
4758 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4759 it possible to be selective about when SMTP synchronization is enforced.
4761 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4763 32. Implemented hosts_connection_nolog.
4765 33. Added an ACL for QUIT.
4767 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4770 35. Added mailbox_size and mailbox_filecount to appendfile.
4772 36. Added control = no_multiline_responses to ACLs.
4774 37. There was a bug in the logic of the code that waits for the clock to tick
4775 in the case where the clock went backwards by a substantial amount such
4776 that the microsecond fraction of "now" was more than the microsecond
4777 fraction of "then" (but the whole seconds number was less).
4779 38. Added support for the libradius Radius client library this is found on
4780 FreeBSD (previously only the radiusclient library was supported).
4786 1. When certain lookups returned multiple values in the form name=value, the
4787 quoting of the values was not always being done properly. Specifically:
4788 (a) If the value started with a double quote, but contained no whitespace,
4790 (b) If the value contained whitespace other than a space character (i.e.
4791 tabs or newlines or carriage returns) it was not quoted.
4792 This fix has been applied to the mysql and pgsql lookups by writing a
4793 separate quoting function and calling it from the lookup code. The fix
4794 should probably also be applied to nisplus, ibase and oracle lookups, but
4795 since I cannot test any of those, I have not disturbed their existing code.
4797 2. A hit in the callout cache for a specific address caused a log line with no
4798 reason for rejecting RCPT. Now it says "Previous (cached) callout
4799 verification failure".
4801 3. There was an off-by-one bug in the queryprogram router. An over-long
4802 return line was truncated at 256 instead of 255 characters, thereby
4803 overflowing its buffer with the terminating zero. As well as fixing this, I
4804 have increased the buffer size to 1024 (and made a note to document this).
4806 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4807 when Exim is waiting for an SMTP response from a remote server, Exim
4808 restarts its select() call on the socket, thereby resetting its timeout.
4809 This is not a problem when such interrupts are rare. Somebody set up a cron
4810 job to run exiwhat every 2 minutes, which is less than the normal select()
4811 timeout (5 or 10 minutes). This meant that the select() timeout never
4812 kicked in because it was always reset. I have fixed this by comparing the
4813 time when an interrupt arrives with the time at the start of the first call
4814 to select(). If more time than the timeout has elapsed, the interrupt is
4815 treated as a timeout.
4817 5. Some internal re-factoring in preparation for the addition of Sieve
4818 extensions (by MH). In particular, the "personal" test is moved to a
4819 separate function, and given an option for scanning Cc: and Bcc: (which is
4820 not set for Exim filters).
4822 6. When Exim created an email address using the login of the caller as the
4823 local part (e.g. when creating a From: or Sender: header line), it was not
4824 quoting the local part when it contained special characters such as @.
4826 7. Installed new OpenBSD configuration files.
4828 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4829 try to make them clearer.
4831 9. Callout options, other than the timeout value, were being ignored when
4832 verifying sender addresses in header lines. For example, when using
4834 verify = header_sender/callout=no_cache
4836 the cache was (incorrectly) being used.
4838 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4839 only the Cygwin environment.
4841 11. When return_path_on_delivery was set as a log selector, if different remote
4842 addresses in the same message used different return paths and parallel
4843 remote delivery occurred, the wrong values would sometimes be logged.
4844 (Whenever a remote delivery process finished, the return path value from
4845 the most recently started remote delivery process was logged.)
4847 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4848 lines when AUTH and/or TLS are in use. This is the "received protocol"
4849 field. Exim used to use "asmtp" for authenticated SMTP, without any
4850 indication (in the protocol name) for TLS use. Now it follows the RFC and
4851 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4852 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4853 names appear in log lines as well as in Received: header lines.
4855 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4856 extensions, and comparison tests, and to fix some bugs.
4858 14. Changes to the "personal" filter test:
4860 (1) The test was buggy in that it was just doing the equivalent of
4861 "contains" tests on header lines. For example, if a user's address was
4862 anne@some.where, the "personal" test would incorrectly be true for
4864 To: susanne@some.where
4866 This test is now done by extracting each address from the header in turn,
4867 and checking the entire address. Other tests that are part of "personal"
4868 are now done using regular expressions (for example, to check local parts
4869 of addresses in From: header lines).
4871 (2) The list of non-personal local parts in From: addresses has been
4872 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4873 taken from the Sieve specification recommendations.
4875 (3) If the message contains any header line starting with "List-" it is
4876 treated as non-personal.
4878 (4) The test for "circular" in the Subject: header line has been removed
4879 because it now seems ill-conceived.
4881 15. Minor typos in src/EDITME comments corrected.
4883 16. Installed latest exipick from John Jetmore.
4885 17. If headers_add on a router specified a text string that was too long for
4886 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4887 of string_sprintf() is now avoided.
4889 18. $message_body_size was not set (it was always zero) when running the DATA
4890 ACL and the local_scan() function.
4892 19. For the "mail" command in an Exim filter, no default was being set for
4893 the once_repeat time, causing a random time value to be used if "once" was
4894 specified. (If the value happened to be <= 0, no repeat happened.) The
4895 default is now 0s, meaning "never repeat". The "vacation" command was OK
4896 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4897 (I found it when inspecting the code).
4899 20. There is now an overall timeout for performing a callout verification. It
4900 defaults to 4 times the callout timeout, which applies to individual SMTP
4901 commands during the callout. The overall timeout applies when there is more
4902 than one host that can be tried. The timeout is checked before trying the
4903 next host. This prevents very long delays if there are a large number of
4904 hosts and all are timing out (e.g. when the network connections are timing
4905 out). The value of the overall timeout can be changed by specifying an
4906 additional sub-option for "callout", called "maxwait". For example:
4908 verify = sender/callout=5s,maxwait=20s
4910 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4911 to the end before writing, but this should make it even safer).
4913 22. Exim was forgetting that it had advertised PIPELINING for the second and
4914 subsequent messages on an SMTP connection. It was also not resetting its
4915 memory on STARTTLS and an internal HELO.
4917 23. When Exim logs an SMTP synchronization error within a session, it now
4918 records whether PIPELINING has been advertised or not.
4920 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4921 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4922 rather than long int.
4924 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4926 26. Added the never_mail option to autoreply.
4932 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4933 crash if the getsockname() call failed; this can happen if a connection is
4934 closed very soon after it is established. The problem was simply in the
4935 order in which certain operations were done, causing Exim to try to write
4936 to the SMTP stream before it had set up the file descriptor. The bug has
4937 been fixed by making things happen in the correct order.
4943 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4944 before the connection was closed, thus losing the rejection response.
4946 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4947 some early Solaris releases, but causes trouble in current releases where
4948 socklen_t is defined.
4950 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4953 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4956 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4957 that releasing the top of it at the end releases what was used for sub-
4958 expansions (unless the block got too big). However, discard this block if
4959 the first thing is a variable or header, so that we can use its block when
4960 it is dynamic (useful for very large $message_headers, for example).
4962 6. Lookups now cache *every* query, not just the most recent. A new, separate
4963 store pool is used for this. It can be recovered when all lookup caches are
4964 flushed. Lookups now release memory at the end of their result strings.
4965 This has involved some general refactoring of the lookup sources.
4967 7. Some code has been added to the store_xxx() functions to reduce the amount
4968 of flapping under certain conditions.
4970 8. log_incoming_interface used to affect only the <= reception log lines. Now
4971 it causes the local interface and port to be added to several more SMTP log
4972 lines, for example "SMTP connection from", and rejection lines.
4974 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4976 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4978 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4979 could be overwritten at the end of the current message (or the start of a
4980 new message if it was set in a HELO ACL). The value is now preserved for
4981 the duration of the SMTP connection.
4983 12. If a transport had a headers_rewrite setting, and a matching header line
4984 contained an unqualified address, that address was qualified, even if it
4985 did not match any rewriting rules. The underlying bug was that the values
4986 of the flags that permit the existence of unqualified sender and recipient
4987 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4988 non-local messages, and by -bnq for local messages) were not being
4989 preserved with the message after it was received.
4991 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4992 "next input=" as part of the text comprising the host identity instead of
4993 the correct text. The code was using the same buffer for two different
4994 strings. However, depending on which order the printing function evaluated
4995 its arguments, the bug did not always show up. Under Linux, for example, my
4996 test suite worked just fine.
4998 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4999 doesn't work with some older versions of Perl. It has been changed to "my",
5000 which in any case is probably the better facility to use.
5002 15. A really picky compiler found some instances of statements for creating
5003 error messages that either had too many or two few arguments for the format
5006 16. The size of the buffer for calls to the DNS resolver has been increased
5007 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5008 for addresses that have a lot of PTR records. This alleviates a problem; it
5009 does not fully solve it.
5011 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5012 buffer now truncates the list and logs the incident, which is the same
5013 action as happens when Exim is looking up a host name and its aliases.
5014 Previously in this situation something unpredictable would happen;
5015 sometimes it was "internal error: store_reset failed".
5017 18. If a server dropped the connection unexpectedly when an Exim client was
5018 using GnuTLS and trying to read a response, the client delivery process
5019 crashed while trying to generate an error log message.
5021 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5022 string, for example:
5024 warn message = H1: something\nH2: something
5026 the text was added as a single header line from Exim's point of view
5027 though it ended up OK in the delivered message. However, searching for the
5028 second and subsequent header lines using $h_h2: did not work. This has been
5029 fixed. Similarly, if a system filter added multiple headers in this way,
5030 the routers could not see them.
5032 20. Expanded the error message when iplsearch is called with an invalid key to
5033 suggest using net-iplsearch in a host list.
5035 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5036 ACLs are no longer actually imposed (and a message to that effect is
5039 22. If a "gecos" field in a passwd entry contained escaped characters, in
5040 particular, if it contained a \" sequence, Exim got it wrong when building
5041 a From: or a Sender: header from that name. A second bug also caused
5042 incorrect handling when an unquoted " was present following a character
5043 that needed quoting.
5045 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5046 was not being matched caselessly.
5048 24. Arranged for all hyphens in the exim.8 source to be escaped with
5051 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5052 properly. Recipient callout cache records were still being keyed to include
5053 the sender, even when use_sender was set false. This led to far more
5054 callouts that were necessary. The sender is no longer included in the key
5055 when use_sender is false.
5057 26. Added "control = submission" modifier to ACLs.
5059 27. Added the ${base62d: operator to decode base 62 numbers.
5061 28. dnsdb lookups can now access SRV records.
5063 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5064 the configuration file.
5066 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5067 (-v) mode. This makes the output for a verbose queue run more intelligible.
5069 31. Added a use_postmaster feature to recipient callouts.
5071 32. Added the $body_zerocount variable, containing the number of binary zero
5072 bytes in the message body.
5074 33. The time of last modification of the "new" subdirectory is now used as the
5075 "mailbox time last read" when there is a quota error for a maildir
5078 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5080 35. Added +ignore_unknown as a special item in host lists.
5082 36. Code for decoding IPv6 addresses in host lists is now included, even if
5083 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5084 address was recognized as an IP address, but was then not correctly decoded
5085 into binary, causing unexpected and incorrect effects when compared with
5092 1. Very minor rewording of debugging text in manualroute to say "list of
5093 hosts" instead of "hostlist".
5095 2. If verify=header_syntax was set, and a header line with an unqualified
5096 address (no domain) and a large number of spaces between the end of the
5097 name and the colon was received, the reception process suffered a buffer
5098 overflow, and (when I tested it) crashed. This was caused by some obsolete
5099 code that should have been removed. The fix is to remove it!
5101 3. When running in the test harness, delay a bit after writing a bounce
5102 message to get a bit more predictability in the log output.
5104 4. Added a call to search_tidyup() just before forking a reception process. In
5105 theory, someone could use a lookup in the expansion of smtp_accept_max_
5106 per_host which, without the tidyup, could leave open a database connection.
5108 5. Added the variables $recipient_data and $sender_data which get set from a
5109 lookup success in an ACL "recipients" or "senders" condition, or a router
5110 "senders" option, similar to $domain_data and $local_part_data.
5112 6. Moved the writing of debug_print from before to after the "senders" test
5115 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5116 problems for message scanning, either using a data ACL, or using
5117 local_scan() because the Received: header was not generated till after they
5118 were called (in order to set the time as the time of reception completion).
5119 I have revised the way this works. The header is now generated after the
5120 body is received, but before the ACL or local_scan() are called. After they
5121 are run, the timestamp in the header is updated.
5127 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5128 before starting a queue runner without re-exec. This happened only when
5129 deliver_drop_privilege was set or when the Exim user was set to root. The
5130 effect of the bug was that timeouts during subsequent deliveries caused
5131 crashes instead of being properly handled. The handler is now left at its
5132 default (and expected) setting.
5134 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5135 message, again when deliver_drop_privilege is set or Exim is run as root.
5136 The bug described in (1) was not present in this case, but the tidying up
5137 of the other signals was missing. I have made the two cases consistent.
5139 3. The ignore_target_hosts setting on a manualroute router was being ignored
5140 for hosts that were looked up using the /MX notation.
5142 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5145 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5146 operated on the sender address. After changing the $sender_address to <>
5147 for the sender address verify, Exim was re-instated it as the original
5148 (before rewriting) address, but remembering that it had rewritten it, so it
5149 wasn't rewriting it again. This bug also had the effect of breaking the
5150 sender address verification caching when the sender address was rewritten.
5152 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5153 This has been changed so that if the ip literal address matches
5154 ignore_target_hosts, the router declines.
5156 7. Added expansion conditions match_domain, match_address, and match_local_
5157 part (NOT match_host).
5159 8. The placeholder for the Received: header didn't have a length field set.
5161 9. Added code to Exim itself and to exim_lock to test for a specific race
5162 condition that could lead to file corruption when using MBX delivery. The
5163 issue is with the lockfile that is created in /tmp. If this file is removed
5164 after a process has opened it but before that process has acquired a lock,
5165 there is the potential for a second process to recreate the file and also
5166 acquire a lock. This could lead to two Exim processes writing to the file
5167 at the same time. The added code performs the same test as UW imapd; it
5168 checks after acquiring the lock that its file descriptor still refers to
5169 the same named file.
5171 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5172 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5175 11. Added the smtp_active_hostname option. If used, this will typically be made
5176 to depend on the incoming interface address. Because $interface_address is
5177 not set up until the daemon has forked a reception process, error responses
5178 that can happen earlier (such as "too many connections") no longer contain
5181 12. If an expansion in a condition on a "warn" statement fails because a lookup
5182 defers, the "warn" statement is abandoned, and the next ACL statement is
5183 processed. Previously this caused the whole ACL to be aborted.
5185 13. Added the iplsearch lookup type.
5187 14. Added ident_timeout as a log selector.
5189 15. Added tls_certificate_verified as a log selector.
5191 16. Added a global option tls_require_ciphers (compare the smtp transport
5192 option of the same name). This controls incoming TLS connections.
5194 17. I finally figured out how to make tls_require_ciphers do a similar thing
5195 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5196 before starting the TLS session.
5198 18. Tabs are now shown as \t in -bP output.
5200 19. If the log selector return_path_on_delivery was set, Exim crashed when
5201 bouncing a message because it had too many Received: header lines.
5203 20. If two routers both had headers_remove settings, and the first one included
5204 a superfluous trailing colon, the final name in the first list and the
5205 first name in the second list were incorrectly joined into one item (with a
5206 colon in the middle).
5212 1. Added -C and -D options to the exinext utility, mainly to make it easier
5213 to include in the automated testing, but these could be helpful when
5214 multiple configurations are in use.
5216 2. The exinext utility was not formatting the output nicely when there was
5217 an alternate port involved in the retry record key, nor when there was a
5218 message id as well (for retries that were specific to a specific message
5219 and a specific host). It was also confused by IPv6 addresses, because of
5220 the additional colons they contain. I have fixed the IPv4 problem, and
5221 patched it up to do a reasonable job for IPv6.
5223 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5224 delivery, the log line now contains "pipelined" if PIPELINING was used.
5226 4. An SMTP transport process used to panic and die if the bind() call to set
5227 an explicit outgoing interface failed. This has been changed; it is now
5228 treated in the same way as a connect() failure.
5230 5. A reference to $sender_host_name in the part of a conditional expansion
5231 that was being skipped was still causing a DNS lookup. This no longer
5234 6. The def: expansion condition was not recognizing references to header lines
5235 that used bh_ and bheader_.
5237 7. Added the _cache feature to named lists.
5239 8. The code for checking quota_filecount in the appendfile transport was
5240 allowing one more file than it should have been.
5242 9. For compatibility with Sendmail, the command line option
5250 and sets the incoming protocol and host name (for trusted callers). The
5251 host name and its colon can be omitted when only the protocol is to be set.
5252 Note the Exim already has two private options, -pd and -ps, that refer to
5253 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5254 "s", but I don't think that's a major issue.
5256 10. A number of refactoring changes to the code, none of which should affect
5259 (a) The number of logging options was getting close to filling up the
5260 32-bit word that was used as a bit map. I have split them into two classes:
5261 those that are passed in the argument to log_write(), and those that are
5262 only ever tested independently outside of that function. These are now in
5263 separate 32-bit words, so there is plenty of room for expansion again.
5264 There is no change in the user interface or the logging behaviour.
5266 (b) When building, for example, log lines, the code previously used a
5267 macro that called string_cat() twice, in order to add two strings. This is
5268 not really sufficiently general. Furthermore, there was one instance where
5269 it was actually wrong because one of the argument was used twice, and in
5270 one call a function was used. (As it happened, calling the function twice
5271 did not affect the overall behaviour.) The macro has been replaced by a
5272 function that can join an arbitrary number of extra strings onto a growing
5275 (c) The code for expansion conditions now uses a table and a binary chop
5276 instead of a serial search (which was left over from when there were very
5277 few conditions). Also, it now recognizes conditions like "pam" even when
5278 the relevant support is not compiled in: a suitably worded error message is
5279 given if an attempt is made to use such a condition.
5281 11. Added ${time_interval:xxxxx}.
5283 12. A bug was causing one of the ddress fields not to be passed back correctly
5284 from remote delivery subprocesses. The field in question was not being
5285 subsequently used, so this caused to problems in practice.
5287 13. Added new log selectors queue_time and deliver_time.
5289 14. Might have fixed a bug in maildirsizefile handling that threw up
5290 "unexpected character" debug warnings, and recalculated the data
5291 unnecessarily. In any case, I expanded the warning message to give more
5294 15. Added the message "Restricted characters in address" to the statements in
5295 the default ACL that block characters like @ and % in local parts.
5297 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5298 Three changes have been made:
5300 (a) There was a serious bug; a negative response to MAIL caused the whole
5301 recipient domain to be cached as invalid, thereby blocking all messages
5302 to all local parts at the same domain, from all senders. This bug has
5303 been fixed. The domain is no longer cached after a negative response to
5304 MAIL if the sender used is not empty.
5306 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5309 (c) A new callout option, "use_sender" has been added for people who want
5310 the modified behaviour.
5316 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5319 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5320 indeed breaks things for older releases.
5322 3. Added additional logging to the case where there is a problem reading data
5323 from a filter that is running in a subprocess using a pipe, in order to
5324 try to track down a specific problem.
5326 4. Testing facility fudge: when running in the test harness and attempting
5327 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5328 getting "No route to host". Convert this to a timeout.
5330 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5333 6. Some OS don't have socklen_t but use size_t instead. This affects the
5334 fifth argument of getsockopt() amongst other things. This is now
5335 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5336 can be set for individual OS. I have set it for SunOS5, OSF1, and
5337 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5338 some earlier ones do not.
5340 7. Change 4.30/15 was not doing the test caselessly.
5342 8. The standard form for an IPv6 address literal was being rejected by address
5343 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5344 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5345 this, as well as the form without the "IPv6" on the front (but only when
5346 address literals are enabled, of course).
5348 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5350 10. Exim crashed if a message with an empty sender address specified by -f
5351 encountered a router with an errors_to setting. This could be provoked only
5352 by a command such as
5356 where an empty string was supplied; "<>" did not hit this bug.
5358 11. Installed PCRE release 4.5.
5360 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5361 remained set. It is now erased.
5363 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5364 times from message ids (which are base 36 rather than the normal 62).
5366 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5367 were being counted as actual protocol errors, and logged if the log
5368 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5369 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5370 and DATA following a set of rejected RCPTs do not count as protocol errors.
5371 In other words, Exim assumes they were pipelined, though this may not
5372 actually be the case. Of course, in all cases the client gets an
5373 appropriate error code.
5375 15. If a lookup fails in an ACL condition, a message about the failure may
5376 be available; it is used if testing the ACL cannot continue, because most
5377 such messages specify what the cause of the deferral is. However, some
5378 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5379 that caused an old message to be retained and used if a later statement
5380 caused a defer, replacing the real cause of the deferral.
5382 16. If an IP address had so many PTR records that the DNS lookup buffer
5383 was not large enough to hold them, Exim could crash while trying to process
5384 the truncated data. It now detects and logs this case.
5386 17. Further to 4.21/58, another change has been made: if (and only if) the
5387 first line of a message (the first header line) ends with CRLF, a bare LF
5388 in a subsequent header line has a space inserted after it, so as not to
5389 terminate the header.
5391 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5392 unnecessarily, used atoi() instead of strtol(), and didn't check the
5393 termination when getting file sizes from file names by regex.
5395 19. Completely re-implemented the support for maildirsize files, in the light
5396 of a number of problems with the previous contributed implementation
5397 (4.30/29). In particular:
5399 . If the quota is zero, the maildirsize file is maintained, but no quota is
5402 . If the maildir directory does not exist, it is created before any attempt
5403 to write a maildirsize file.
5405 . The quota value in the file is just a cache; if the quota is changed in
5406 the transport, the new value overrides.
5408 . A regular expression is available for excluding directories from the
5411 20. The autoreply transport checks the characters in options that define the
5412 message's headers; it allows continued headers, but it was checking with
5413 isspace() after an embedded newline instead of explicitly looking for a
5416 21. If all the "regular" hosts to which an address was routed had passed their
5417 expiry times, and had not reached their retry times, the address was
5418 bounced, even if fallback hosts were defined. Now Exim should go on to try
5421 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5422 equivalent code in the SMTP transport. Some hosts send humungous responses
5423 to HELO/EHLO, more than 1024 it seems.
5425 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5426 gives compiler warnings in some environments. I've now done it "properly",
5429 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5430 (because of problems with the built-in one) was declared to return uschar *
5431 instead of char *, causing compiler failure.
5433 25. Fixed a file descriptor leak when processing alias/forward files.
5435 26. Fixed a minor format string issue in dbfn.c.
5437 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5439 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5440 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5443 29. When a sender address is verified, it is cached, to save repeating the test
5444 when there is more than one recipient in a message. However, when the
5445 verification involves a callout, it is possible for different callout
5446 options to be set for different recipients. It is too complicated to keep
5447 track of this in the cache, so now Exim always runs a verification when a
5448 callout is required, relying on the callout cache for the optimization.
5449 The overhead is duplication of the address routing, but this should not be
5452 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5453 to be verified with callout=postmaster, and the main callout worked but the
5454 postmaster check failed, the verification correctly failed. However, if a
5455 subsequent RCPT command asked for sender verification *without* the
5456 postmaster check, incorrect caching caused this verification also to fail,
5459 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5460 it was not caching the DNS options (qualify_single, search_parents) that
5461 were used when the lookup failed. A subsequent lookup with different
5462 options therefore always gave the same answer, though there were cases
5463 where it should not have. (Example: a "domains = !$mx_any" option on a
5464 dnslookup router: the "domains" option is always processed without any
5465 widening, but the router might have qualify_single set.) Now Exim uses the
5466 cached value only when the same options are set.
5468 32. Added John Jetmore's "exipick" utility to the distribution.
5470 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5471 than a timeout (e.g. a certificate is required, and is not provided), an
5472 Exim server now closes the connection immediately. Previously it waited for
5473 the client to close - but if the client is SSL, it seems that they each
5474 wait for each other, leading to a delay before one of them times out.
5476 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5477 maintained 0.8.x compatibility because I don't think many are using it, and
5478 it is clearly obsolete.
5480 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5483 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5484 client certificate was expired. A simple patch fixes this, though I don't
5485 understand the full logic of why the verify callback is called multiple
5488 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5489 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5490 which causes problems with some clients (such as the Certicom SSL Plus
5491 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5492 disables the coutermeasure allowing Eudora to connect."
5494 38. Exim was not checking that a write() to a log file succeeded. This could
5495 lead to Bad Things if a log got too big, in particular if it hit a file
5496 size limit. Exim now panics and dies if it cannot write to a log file, just
5497 as it does if it cannot open a log file.
5499 39. Modified OS/Makefile-Linux so that it now contains
5501 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5503 The two -D definitions ensure that Exim is compiled with large file
5504 support, which makes it possible to handle log files that are bigger than
5507 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5508 instance) a domain was checked against a named list that involved a lookup,
5509 causing $domain_data to be set, then another domain was checked against the
5510 same list, then the first domain was re-checked, the value of $domain_data
5511 after the final check could be wrong. In particular, if the second check
5512 failed, it could be set empty. This bug probably also applied to
5515 41. The strip_trailing_dot option was not being applied to the address given
5516 with the -f command-line option.
5518 42. The code for reading a message's header from the spool was incrementing
5519 $received_count, but never initializing it. This meant that the value was
5520 incorrect (doubled) while delivering a message in the same process in which
5521 it was received. In the most common configuration of Exim, this never
5522 happens - a fresh exec is done - but it can happen when
5523 deliver_drop_privilege is set.
5525 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5526 it now includes up to 150 characters of the unexpected data in the log
5529 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5530 and building data strings. The size of both of these buffers was 10 000
5531 bytes - far larger than anybody would *ever* want, thought I. Needless to
5532 say, somebody hit the limit. I have increased the maximum line length to
5533 20 000 and the maximum data length of concatenated lines to 100 000. I have
5534 also fixed two bugs, because there was no checking on these buffers. Tsk,
5535 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5536 buffer is too small.
5538 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5539 lsearch lookups. Now it does.
5541 46. When parsing a route_list item in a manualroute router, a fixed-length
5542 buffer was used for the list of hosts. I made this 1024 bytes long,
5543 thinking that nobody would ever have a list of hosts that long. Wrong.
5544 Somebody had a whole pile of complicated expansion conditions, and the
5545 string was silently truncated, leading to an expansion error. It turns out
5546 that it is easier to change to an unlimited length (owing to other changes
5547 that have happened since this code was originally written) than to build
5548 structure for giving a limitation error. The length of the item that
5549 expands into the list of hosts is now unlimited.
5551 47. The lsearch lookup could not handle data where the length of text line was
5552 more than 4095 characters. Such lines were truncated, leading to shortened
5553 data being returned. It should now handle lines of any length.
5555 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5556 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5559 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5560 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5561 no longer happens in comment lines. A list of replacements is now placed
5562 at the head of all of the source files, except those whose only change is
5563 to replace PERL_COMMAND in the very first #! line.
5565 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5566 messages on the queue, with a bottom-up merge sort, using code contributed
5567 by Michael Haardt. This should make operations like -bp somewhat faster on
5568 large queues. It won't affect queue runners, except when queue_run_in_order
5571 51. Installed eximstats 1.31 in the distribution.
5573 52. Added support for SRV lookups to the dnslookup router.
5575 53. If an ACL referred to $message_body or $message_body_end, the value was not
5576 reset for any messages that followed in the same SMTP session.
5578 54. The store-handling optimization for building very long strings was not
5579 differentiating between the different store pools. I don't think this
5580 actually made any difference in practice, but I've tidied it.
5582 55. While running the routers to verify a sender address, $sender_address
5583 was still set to the sender address. This is wrong, because when routing to
5584 send a bounce to the sender, it would be empty. Therefore, I have changed
5585 it so that, while verifying a sender address, $sender_address is set to <>.
5586 (There is no change to what happens when verifying a recipient address.)
5588 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5589 target A or AAAA records (if not already returned) without resetting the
5590 qualify_single or search_parents options of the DNS resolver. These are
5591 inappropriate in this case because the targets of MX and SRV records must
5592 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5593 target that, when qualified, matched something in the local domain. These
5594 two options are now turned off when doing these lookups.
5596 57. It seems that at least some releases of Reiserfs (which does not have the
5597 concept of a fixed number of inodes) returns zero and not -1 for the
5598 number of available inodes. This interacted badly with check_spool_inodes,
5599 which assumed that -1 was the "no such thing" setting. What I have done is
5600 to check that the total number of inodes is greater than zero before doing
5601 the test of how many are available.
5603 58. When a "warn" ACL statement has a log_message modifier, the message is
5604 remembered, and not repeated. This is to avoid a lot of repetition when a
5605 message has many recipients that cause the same warning to be written.
5606 However, Exim was preserving the list of already written lines for an
5607 entire SMTP session, which doesn't seem right. The memory is now reset if a
5608 new message is started.
5610 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5611 debugging output unless log_rewrite was also set.
5613 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5614 of (int)(handle) when we know that handle contains (void *)(-1).
5616 61. The Exim daemon panic-logs an error return when it closes the incoming
5617 connection. However "connection reset by peer" seems to be common, and
5618 isn't really an error worthy of noting specially, so that particular error
5621 62. When Exim is trying to find all the local interfaces, it used to panic and
5622 die if the ioctl to get the interface flags failed. However, it seems that
5623 on at least one OS (Solaris 9) it is possible to have an interface that is
5624 included in the list of interfaces, but for which you get a failure error
5625 for this call. This happens when the interface is not "plumbed" into a
5626 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5627 failure of the "get flags" call assumes that the interface is down.
5629 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5630 makes life easier for people who are doing arithmetic on fields extracted
5631 from dates, where you often get leading zeros that should not be
5632 interpreted as octal.
5634 64. Added qualify_domain to the redirect router, to override the global
5637 65. If a pathologically long header line contained very many addresses (the
5638 report of this problem mentioned 10 000) and each of them was rewritten,
5639 Exim could use up a very large amount of memory. (It kept on making new
5640 copies of the header line as it rewrote, and never released the old ones.)
5641 At the expense of a bit more processing, the header rewriting function has
5642 been changed so that it no longer eats memory in this way.
5644 66. The generation of the Received: header has been moved from the time that a
5645 message starts to be received, to the time that it finishes. The timestamp
5646 in the Received: header should now be very close to that of the <= log
5647 line. There are two side-effects of this change:
5649 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5650 logged header lines no longer include the local Received: line, because
5651 it has not yet been created. The same applies to a copy of the message
5652 that is returned to a non-SMTP sender when a message is rejected.
5654 (b) When a filter file is tested using -bf, no additional Received: header
5655 is added to the test message. After some thought, I decided that this
5658 This change does not affect the value of $received_for. It is still set
5659 after address rewriting, but before local_scan() is called.
5661 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5663 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5664 gave an unhelpful panic error message, and a defer error. I have managed to
5665 change this behaviour so that it now rejects any supplied certificate,
5666 which seems right, as the list of acceptable certificates is empty.
5668 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5669 gave an unhelpful defer error. I have not managed to make this reject any
5670 supplied certificates, but the error message it gives is "no certificate
5671 supplied", which is not helpful.
5673 70. exigrep's output now also includes lines that are not associated with any
5674 message, but which match the given pattern. Implemented by a patch from
5675 Martin Sluka, which also tidied up the Perl a bit.
5677 71. Recipient callout verification, like sender verification, was using <> in
5678 the MAIL FROM command. This isn't really the right thing, since the actual
5679 sender may affect whether the remote host accepts the recipient or not. I
5680 have changed it to use the actual sender in the callout; this means that
5681 the cache record is now keyed on a recipient/sender pair, not just the
5682 recipient address. There doesn't seem to be a real danger of callout loops,
5683 since a callout by the remote host to check the sender would use <>.
5684 [SEE ABOVE: changed after hitting problems.]
5686 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5687 temporary errors. However, in the case of such a code being given after
5688 the end of a data transmission (i.e. after ".") Exim was failing to write
5689 a retry record for the message. (Yes, there was some broken host that was
5690 actually sending 8xx at this point.)
5692 73. An unknown lookup type in a host list could cause Exim to panic-die when
5693 the list was checked. (An example that provoked this was putting <; in the
5694 middle of a list instead of at the start.) If this happened during a DATA
5695 ACL check, a -D file could be left lying around. This kind of configuration
5696 error no longer causes Exim to die; instead it causes a defer error. The
5697 incident is still logged to the main and panic logs.
5699 74. Buglet left over from Exim 3 conversion. The message "too many messages
5700 in one connection" was written to the rejectlog but not the mainlog, except
5701 when address rewriting (yes!) was being logged.
5703 75. Added write_rejectlog option.
5705 76. When a system filter was run not as root (that is, when system_filter_user
5706 was set), the values of the $n variables were not being returned to the
5707 main process; thus, they were not subsequently available in the $sn
5710 77. Added +return_path_on_delivery log selector.
5712 78. A connection timeout was being treated differently from recipients deferred
5713 when testing hosts_max_try with a message that was older than the host's
5714 retry timeout. (The host should not be counted, thus allowing all hosts to
5715 be tried at least once before bouncing.) This may have been the cause of an
5716 occasionally reported bug whereby a message would remain on the queue
5717 longer than the retry timeout, but would be bounced if a delivery was
5718 forced. I say "may" because I never totally pinned down the problem;
5719 setting up timeout/retry tests is difficult. See also the next item.
5721 79. The ultimate address timeout was not being applied to errors that involved
5722 a combination of host plus message (for example, a timeout on a MAIL
5723 command). When an address resolved to a number of possible hosts, and they
5724 were not all tried for each delivery (e.g. because of hosts_max_try), a
5725 message could remain on the queue longer than the retry timeout.
5727 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5730 81. Fixed an obscure SMTP outgoing bug which required at least the following
5731 conditions: (a) there was another message waiting for the same server;
5732 (b) the server returned 5xx to all RCPT commands in the first message so
5733 that the message was not completed; (c) the server dropped the connection
5734 or gave a negative response to the RSET that Exim sends to abort the
5735 transaction. The observed case was a dropped connection after DATA that had
5736 been sent in pipelining mode. That is, the server had advertised PIPELINING
5737 but was not implementing it correctly. The effect of the bug was incorrect
5738 behaviour, such as trying another host, and this could lead to a crash.
5744 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5745 and daemon.c were passed as pointers to ints; they should have been
5746 pointers to socklen_t variables (which are typically unsigned ints).
5748 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5751 3. Fixed a really odd bug that affected only the testing scheme; patching a
5752 certain fixed string in the binary changed the value of another string that
5753 happened to be identical to the end of the original first string.
5755 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5756 name", it returns that address as the IP address. On some operating
5757 systems (e.g. Solaris), it also passes back the IP address string as the
5758 "host name". However, on others (e.g. Linux), it passes back an empty
5759 string. Exim wasn't checking for this, and was changing the host name to an
5760 empty string, assuming it had been canonicalized.
5762 5. Although rare, it is permitted to have more than one PTR record for a given
5763 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5764 all the names associated with an address, because they do in Solaris.
5765 However, it seems that they do not in Linux for data that comes from the
5766 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5767 I found this out when I moved to a new Linux workstation and tried to run
5768 the Exim test suite.
5770 To get round this problem I have changed the code so that it now does its
5771 own call to the DNS to look up PTR records when searching for a host name.
5772 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5773 addresses that are only in /etc/hosts are still found.
5775 This behaviour is, however, controlled by an option called host_lookup_
5776 order, which defaults to "bydns:byaddr". If people want to use the other
5777 order, or indeed, just use one or the other means of lookup, they can
5778 specify it in this variable.
5780 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5781 some operating systems, this comes back from gethostbyaddr() as an empty
5782 string, and this is what Exim used to test for. However, it seems that in
5783 other systems, "." is yielded. Exim now tests for this case too.
5785 7. The values of check_spool_space and check_log_space are now held internally
5786 as a number of kilobytes instead of an absolute number of bytes. If a
5787 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5788 kilobyte. This means that much larger values can be stored.
5790 8. Exim monitor: an attempt to get the action menu when not actually pointing
5791 at a message produces an empty menu entitled "No message selected". This
5792 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5793 no entries in it ("Shell widget menu has zero width and/or height"). So I
5794 have added a single, blank menu entry in this case.
5796 9. Added ${quote_local_part.
5798 10. MIME decoding is now applied to the contents of Subject: header lines when
5801 11. Now that a reference to $sender_host_address automatically causes a reverse
5802 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5803 host lookup before query-style lookups in lists that might use this
5804 variable. This has therefore been abolished, and the "net-" prefix is no
5805 longer necessary for query-style lookups.
5807 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5808 have been a typo for LFLAGS, so it has been changed.
5810 13. The install script calls Exim with "-C /dev/null" in order to find the
5811 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5812 to be output. However, since Exim outputs its version number before the
5813 error, it didn't break the script. It just looked ugly. I fixed this by
5814 always allowing "-C /dev/null" if the caller is root.
5816 14. Ignore overlarge ACL variable number when reading spool file - insurance
5817 against a later release with more variables having written the file.
5819 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5820 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5821 this, as well as the form without the "IPv6" on the front.
5823 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5824 OS/Makefile-Darwin file.
5826 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5827 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5829 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5830 variable. However, some OS use ints for inodes. I've added cast to long int
5831 to get rid of the compiler warning.
5833 19. I had forgotten to lock out "/../" in configuration file names when
5834 ALT_CONFIG_PREFIX was set.
5836 20. Routers used for verification do not need to specify transports. However,
5837 if such a router generated a host list, and callout was configured, Exim
5838 crashed, because it could not find a port number from the (non-existent)
5839 transport. It now assumes port 25 in this circumstance.
5841 21. Added the -t option to exigrep.
5843 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5844 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5845 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5846 them is set, LOOKUP_LSEARCH is forced.
5848 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5850 24. Added sender and host information to the "rejected by local_scan()" log
5851 line; previously there was no indication of these.
5853 25. Added .include_if_exists.
5855 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5856 Linux. It turns out that not all file systems support this. Apparently some
5857 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5858 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5859 not supported on the file descriptor, is now ignored when Exim is trying to
5860 sync a directory. This applies only to Linux.
5862 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5864 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5865 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5868 29. Added support for maildirsize files from supplied patch (modified a bit).
5870 30. The use of :fail: followed by an empty string could lead Exim to respond to
5871 sender verification failures with (e.g.):
5873 550 Verification failed for <xxx>
5874 550 Sender verify failed
5876 where the first response line was missing the '-' that indicates it is not
5877 the final line of the response.
5879 31. The loop for finding the name of the user that called Exim had a hardwired
5880 limit of 10; it now uses the value of finduser_retries, which is used for
5881 all other user lookups.
5883 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5886 33. Exim was neglecting to zero errno before one call of strtol() when
5887 expanding a string and expecting an integer value. On some systems this
5888 resulted in spurious "integer overflow" errors. Also, it was casting the
5889 result into an int without checking.
5891 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5892 did not work. The code looks as if it has *never* worked, though it appears
5893 to have been documented since at least release 1.62. I have made it work.
5895 35. The "timeout_DNS" error in retry rules, also documented since at least
5896 1.62, also never worked. As it isn't clear exactly what this means, and
5897 clearly it isn't a major issue, I have abolished the feature by treating it
5898 as "timeout", and writing a warning to the main and panic logs.
5900 36. The display of retry rules for -brt wasn't always showing the error code
5903 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5904 timeout_connect_A, timeout_connect_MX.
5906 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5907 to the empty sender.
5909 39. The daemon was not analysing the content of -oX till after it had closed
5910 stderr and disconnected from the controlling terminal. This meant that any
5911 syntax errors were only noted on the panic log, and the return code from
5912 the command was 0. By re-arranging the code a little, I've made the
5913 decoding happen first, so such errors now appear on stderr, and the return
5914 code is 1. However, the actual setting up of the sockets still happens in
5915 the disconnected process, so errors there are still only recorded on the
5918 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5919 connections (as happens on some IP stacks) was logged at start up time as
5920 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5921 it from "IPv6 and IPv4", which means that two separate sockets are being
5924 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5925 says whether AF_INET or AF_INET6 was passed as an argument.
5927 42. Exiwhat output was messed up when time zones were included in log
5930 43. Exiwhat now gives more information about the daemon's listening ports,
5931 and whether -tls-on-connect was used.
5933 44. The "port" option of the smtp transport is now expanded.
5935 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5936 silently ignored. Now an error message is written to the main and panic
5939 46. There's a new ACL modifier called "logwrite" which writes to a log file
5940 as soon as it is encountered.
5942 47. Added $local_user_uid and $local_user_gid at routing time.
5944 48. Exim crashed when trying to verify a sender address that was being
5947 49. Exim was recognizing only a space character after ".include". It now also
5948 recognizes a tab character.
5950 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5951 extracting the relevant information from the specification. The man page no
5952 longer contains scrambled data for the -d option, and I've added a section
5953 at the front about calling Exim under different names.
5955 51. Added "extra_headers" argument to the "mail" command in filter files.
5957 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5960 53. Installed eximstats 1.29.
5962 54. Added transport_filter_timeout as a generic transport option.
5964 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5965 Cc: header lines. This was required by RFC 822, but it not required by RFC
5968 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5969 incoming messages that did not have them. Now it does so only if the
5970 message originates locally, that is, if there is no associated remote host
5971 address. When Resent- header lines are present, this applies to the Resent-
5972 lines rather than the non-Resent- lines.
5974 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5975 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5977 58. Messages for configuration errors now include the name of the main
5978 configuration file - useful now that there may be more than one file in a
5979 list (.included file names were always shown).
5981 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5982 for those rare installations that do not start the daemon as root or run it
5983 setuid root. I've cut out the call to initgroups() if the daemon is not
5986 60. The Exim user and group can now be bound into the binary as text strings
5987 that are looked up at the start of Exim's processing.
5989 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5991 62. Added $mailstore_basename variable.
5993 63. Installed patch to sieve.c from Michael Haardt.
5995 64. When Exim failed to open the panic log after failing to open the main log,
5996 the original message it was trying to log was written to stderr and debug
5997 output, but if they were not available (the usual case in production), it
5998 was lost. Now it is written to syslog before the two lines that record the
5999 failures to open the logs.
6001 65. Users' Exim filters run in subprocesses under the user's uid. It is
6002 possible for a "deliver" command or an alias in a "personal" command to
6003 provoke an address rewrite. If logging of address rewriting is configured,
6004 this fails because the process is not running as root or exim. There may be
6005 a better way of dealing with this, but for the moment (because 4.30 needs
6006 to be released), I have disabled address rewrite logging when running a
6007 filter in a non-root, non-exim process.
6013 1. The buildconfig auxiliary program wasn't quoting the value set for
6014 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6015 not defined. This bug was masked in 4.22 by the effect that was fixed in
6018 2. Some messages that were rejected after a message id was allocated were
6019 shown as "incomplete" by exigrep. It no longer does this for messages that
6020 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6022 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6023 have allow_domain_literals set, the ID did not get logged in the <= line.
6024 Domain literals are now always recognized in Message-ID: header lines.
6026 4. The first argument for a ${extract expansion item is the key name or field
6027 number. Leading and trailing spaces in this item were not being ignored,
6028 causing some misleading effects.
6030 5. When deliver_drop_privilege was set, single queue runner processes started
6031 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6032 same command in the process it spins off) were not dropping privilege.
6034 6. When the daemon running as "exim" started a queue runner, it always
6035 re-executed Exim in the spun-off process. This is a waste of effort when
6036 deliver_drop_privilege is set. The new process now just calls the
6037 queue-runner function directly.
6043 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6046 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6047 ignored. Though the use of -oP was forcing the writing of a pid file, it
6048 was always written to the default place.
6050 3. If the message "no IP address found for host xxxx" is generated during
6051 incoming verification, it is now followed by identification of the incoming
6052 connection (so you can more easily find what provoked it).
6054 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6056 5. Added some features to "harden" Exim a bit more against certain attacks:
6058 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6059 be put in Local/Makefile. This is like the never_users runtime option,
6060 but it cannot be overridden. The default setting is "root".
6062 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6063 prefix string with which any file named in a -C command line option
6066 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6067 is retained for -C and -D only if the caller of Exim is root. Without
6068 it, the exim user may also use -C and -D and retain privilege.
6070 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6071 command line option is disabled.
6073 6. Macro names set by the -D option must start with an upper case letter, just
6074 like macro names defined in the configuration file.
6076 7. Added "dereference=" facility to LDAP.
6078 8. Two instances of the typo "uknown" in the source files are fixed.
6080 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6081 the Configure-Makefile script screwed up while processing it.
6083 10. Incorporated PCRE 4.4.
6085 11. The SMTP synchronization check was not operating right at the start of an
6086 SMTP session. For example, it could not catch a HELO sent before the client
6087 waited for the greeting. There is now a check for outstanding input at the
6088 point when the greeting is written. Because of the duplex, asynchronous
6089 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6090 way, but not yet received, when the check is performed.
6092 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6093 on TCP/IP sockets, because this apparently causes some broken clients to
6096 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6097 unchanged) from the Cygwin maintainer.
6099 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6100 was supported instead of testing for mbx. Effectively a typo.
6102 15. The spa authenticator server code was not checking that the input it
6103 received was valid base64.
6105 16. The debug output line for the "set" modifier in ACLs was not showing the
6106 name of the variable that was being set.
6108 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6110 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6111 lookup on its own. Something else was needed to trigger the lookup. For
6112 example, a match in host_lookup or the need for a host name in a host list.
6113 Now, if $sender_host_name is referenced and the host name has not yet been
6114 looked up, a lookup is performed. If the lookup fails, the variable remains
6115 empty, and $host_lookup_failed is set to "1".
6117 19. Added "eqi" as a case-independent comparison operator.
6119 20. The saslauthd authentication condition could segfault if neither service
6120 nor realm was specified.
6122 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6123 error message that was logged was misleading, and incoming SMTP
6124 connections were dropped. The message is now more accurate, and temporary
6125 errors are given to SMTP connections.
6127 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6128 (except RSET) with a 421 error, until QUIT is received. However, it was
6129 failing to send a response to QUIT.
6131 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6132 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6133 ACL is now run after the helo_try_verify_hosts code.
6135 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6136 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6137 case-independent, so other case variants are also recognized). Apparently
6138 some systems use these upper case variants.
6140 25. If more than two messages were waiting for the same host, and a transport
6141 filter was specified for the transport, Exim sent two messages over the
6142 same TCP/IP connection, and then failed with "socket operation on non-
6143 socket" when it tried to send the third.
6145 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6147 27. The extern definition of crypt16() in expand.c was not being excluded when
6148 the OS had its own crypt16() function.
6150 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6151 as a preferred synonym for return_size_limit, both as an option and as an
6154 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6156 30. Changed the default configuration ACL to relax the local part checking rule
6157 for addresses that are not in any local domains. For these addresses,
6158 slashes and pipe symbols are allowed within local parts, but the sequence
6159 /../ is explicitly forbidden.
6161 31. SPA server authentication was not clearing the challenge buffer before
6164 32. log_message in a "warn" ACL statement was writing to the reject log as
6165 well as to the main log, which contradicts the documentation and doesn't
6166 seem right (because no rejection is happening). So I have stopped it.
6168 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6169 However, I am unable to do any testing of this.
6171 34. Fixed an infelicity in the appendfile transport. When checking directories
6172 for a mailbox, to see if any needed to be created, it was accidentally
6173 using path names with one or more superfluous leading slashes; tracing
6174 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6176 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6177 contents are added to the log line that is written for every discarded
6178 recipient. (Previously a log_message setting was ignored.)
6180 36. The ${quote: operator now quotes the string if it is empty.
6182 37. The install script runs exim in order to find its version number. If for
6183 some reason other than non-existence or emptiness, which it checks, it
6184 could not run './exim', it was installing it with an empty version number,
6185 i.e. as "exim-". This error state is now caught, and the installation is
6188 38. An argument was missing from the function that creates an error message
6189 when Exim fails to connect to the socket for saslauthd authentication.
6190 This could cause Exim to crash, or give a corrupted message.
6192 39. Added isip, isip4, and isip6 to ${if conditions.
6194 40. The ACL variables $acl_xx are now saved with the message, and can be
6195 accessed later in routers, transports, and filters.
6197 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6198 strings in the file are not string-expanded.
6200 42. If a MAIL command specified a SIZE value that was too large to fit into an
6201 int variable, the check against message_size_limit failed. Such values are
6202 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6203 day this will have to be increased, but I don't think I want to be around
6204 when emails are that large.
6211 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6212 iconv() is not standard in FreeBSD.
6214 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6215 IPv6 enabled. The observed symptom was a segmentation fault on return from
6216 the function os_common_find_running_interfaces() in src/os.c.
6218 3. In the check_special_case() function in daemon.c I had used "errno" as an
6219 argument name, which causes warnings on some systems. This was basically a
6220 typo, since it was named "eno" in the comments!
6222 4. The code that waits for the clock to tick (at a resolution of some fraction
6223 of a second) so as to ensure message-id uniqueness was always waiting for
6224 at least one whole tick, when it could have waited for less. [This is
6225 almost certainly not relevant at current processor speeds, where it is
6226 unlikely to ever wait at all. But we try to future-proof.]
6228 5. The function that sleeps for a time interval that includes fractions of a
6229 second contained a race. It did not block SIGALRM between setting the
6230 timer, and suspending (a couple of lines later). If the interval was short
6231 and the sigsuspend() was delayed until after it had expired, the suspension
6232 never ended. On busy systems this could lead to processes getting stuck for
6235 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6236 process, before it forks any delivery processes. The open lookup caching
6237 mechanism meant that the open file or database connection was passed into
6238 the delivery process. The problem was that delivery processes always tidy
6239 up cached lookup data. This could cause a problem for the next delivery
6240 process started by the queue runner, because the external queue runner
6241 process does not know about the closure. So the next delivery process
6242 still has data in the lookup cache. In the case of a file lookup, there was
6243 no problem because closing a file descriptor in a subprocess doesn't affect
6244 the parent. However, if the lookup was caching a connection to a database,
6245 the connection was closed, and the second delivery process was likely to
6246 see errors such as "PGSQL: query failed: server closed the connection
6247 unexpectedly". The problem has been fixed by closing all cached lookups
6248 in a queue runner before running a delivery process.
6250 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6251 seem to have the "const" qualifier which it has on other OS. I've
6254 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6255 *of the same type* (client or server) with the same public name that an
6256 error should be diagnosed.
6258 9. When Exim looked up a host name for an IP address, but failed to find the
6259 original IP address when looking up the host name (a safety check), it
6260 output the message "<ip address> does not match any IP for NULL", which was
6261 confusing, to say the least. The bug was that the host name should have
6262 appeared instead of "NULL".
6264 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6265 user that is built into the binary, and the -C or -D options is used, root
6266 privilege is dropped before the configuration file is read. In addition,
6267 logging is switched to stderr instead of the normal log files. If the
6268 configuration then re-defines the Exim user, the unprivileged environment
6269 is probably not what is expected, so Exim logs a panic warning message (but
6272 However, if deliver_drop_privilege is set, the unprivileged state may well
6273 be exactly what is intended, so the warning has been cut out in that case,
6274 and Exim is allowed to try to write to its normal log files.
6280 1. smtp_return_error_details was not giving details for temporary sender
6281 or receiver verification errors.
6283 2. Diagnose a configuration error if two authenticators have the same public
6286 3. Exim used not to create the message log file for a message until the first
6287 delivery attempt. This could be confusing when incoming messages were held
6288 for policy or load reasons. The message log file is now created at the time
6289 the message is received, and an initial "Received" line is written to it.
6291 4. The automatically generated man page for command line options had a minor
6292 bug that caused no ill effects; however, a more serious problem was that
6293 the procedure for building the man page automatically didn't always
6294 operate. Consequently, release 4.20 contains an out-of-date version. This
6295 shouldn't happen again.
6297 5. When building Exim with embedded Perl support, the script that builds the
6298 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6299 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6301 6. The freeze_tell option was not being used for messages that were frozen on
6302 arrival, either by an ACL or by local_scan().
6304 7. Added the smtp_incomplete_transaction log selector.
6306 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6307 was accepting AUTH without a new EHLO.
6309 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6310 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6313 10. Logging of TCP/IP connections (when configured) now happens in the main
6314 daemon process instead of the child process, so that the TCP/IP connection
6315 count is more accurate (but it can never be perfect).
6317 11. The use of "drop" in a nested ACL was not being handled correctly in the
6318 outer ACL. Now, if condition failure induced by the nested "drop" causes
6319 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6320 or "require"), the connection is dropped.
6322 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6323 that yield "discard" can now be used with an "accept" or a "discard" verb,
6324 but an error is generated for any others (because I can't see a useful way
6325 to define what should happen).
6327 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6328 are now processed in the same way as lines in the Exim configuration file.
6329 In particular, continuation lines are supported.
6331 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6333 15. Added -ti meaning -t -i.
6335 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6336 domains, and warn by logging if others are found.
6338 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6339 returned from GIFCONF when Exim is trying to find the list of interfaces on
6340 a host. The code in os.c has been modified to copy each ifreq to an aligned
6341 structure in all cases.
6343 Also, in some cases, the returned ifreq's were being copied to a 'struct
6344 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6345 means the last couple of bytes of an IPv6 address could be chopped if the
6346 ifreq contained only a normal sockaddr (14 bytes storage).
6348 18. Named domain lists were not supported in the hosts_treat_as_local option.
6349 An entry such as +xxxx was not recognized, and was treated as a literal
6352 19. Ensure that header lines added by a DATA ACL are included in the reject log
6353 if the ACL subsequently rejects the message.
6355 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6356 MD5 (which is deprecated).
6358 21. When testing a filter file using -bf, Exim was writing a message when it
6359 took the sender from a "From " line in the message, but it was not doing so
6360 when it took $return_path from a Return-Path: header line. It now does.
6362 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6363 with a valid header line field name (a series of printing characters
6364 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6366 23. Changed "disc" in the source to "disk" to conform to the documentation and
6367 the book and for uniformity.
6369 24. Ignore Sendmail's -Ooption=value command line item.
6371 25. When execve() failed while trying to run a command in a pipe transport,
6372 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6373 could be confused with a return value of 69 from the command itself. This
6374 has been changed to 127, the value the shell returns if it is asked to run
6375 a non-existent command. The wording for the related log line suggests a
6376 non-existent command as the problem.
6378 26. If received_header_text expands to an empty string, do not add a Received:
6379 header line to the message. (Well, it adds a token one on the spool, but
6380 marks it "old" so that it doesn't get used or transmitted.)
6382 27. Installed eximstats 1.28 (addition of -nt option).
6384 28. There was no check for failure on the call to getsockname() in the daemon
6385 code. This can fail if there is a shortage of resources on the system, with
6386 ENOMEM, for example. A temporary error is now given on failure.
6388 29. Contrary to the C standard, it seems that in some environments, the
6389 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6390 program. Exim now does this explicitly; it affects the formatting of
6391 timestamps using strftime().
6393 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6394 complaints. I've now initialized all the variables, to avoid this.
6396 32. Header lines added by a system filter were not being "seen" during
6397 transport-time rewrites.
6399 33. The info_callback() function passed to OpenSSL is set up with type void
6400 (*)(SSL *, int, int), as described somewhere. However, when calling the
6401 function (actually a macro) that sets it up, the type void(*)() is
6402 expected. I've put in a cast to prevent warnings from picky compilers.
6404 34. If a DNS black list lookup found a CNAME record, but there were no A
6405 records associated with the domain it pointed at, Exim crashed.
6407 35. If a DNS black list lookup returned more than one A record, Exim ignored
6408 all but the first. It now scans all returned addresses if a particular IP
6409 value is being sought. In this situation, the contents of the
6410 $dnslist_value variable are a list of all the addresses, separated by a
6413 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6414 to accept a host name and all its aliases if the forward lookup for any of
6415 them yielded the IP address of the incoming connection. Now it accepts only
6416 those names whose forward lookup yields the correct IP address. Any other
6417 names are discarded. This closes a loophole whereby a rogue DNS
6418 administrator could create reverse DNS records to break through a
6419 wildcarded host restriction in an ACL.
6421 37. If a user filter or a system filter that ran in a subprocess used any of
6422 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6423 the wrong values were passed to the pipe command ($thisaddress had the
6424 value of $0, $0 had the value of $1, etc). This bug was introduced by
6425 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6427 38. Improved the line breaking for long SMTP error messages from ACLs.
6428 Previously, if there was no break point between 40 and 75 characters, Exim
6429 left the rest of the message alone. Two changes have been made: (a) I've
6430 reduced the minimum length to 35 characters; (b) if it can't find a break
6431 point between 35 and 75 characters, it looks ahead and uses the first one
6432 that it finds. This may give the occasional overlong line, but at least the
6433 remaining text gets split now.
6435 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6436 file descriptors might be low, and that setting 1000 would always raise it.
6437 It turns out that in some environments, the limit is already over 1000 and
6438 that lowering it causes trouble. So now Exim takes care not to decrease it.
6440 40. When delivering a message, the value of $return_path is set to $sender_
6441 address at the start of routing (routers may change the value). By an
6442 oversight, this default was not being set up when an address was tested by
6443 -bt or -bv, which affected the outcome if any router or filter referred to
6446 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6447 log only when "message" or "log_message" are set. However, if one of the
6448 conditions was an address verification, or a call to a nested ACL, the
6449 messages generated by the underlying test were being passed through. This
6450 no longer happens. The underlying message is available in $acl_verify_
6451 message for both "message" and "log_message" expansions, so it can be
6452 passed through if needed.
6454 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6455 new expansion $bh_ to give the encoded byte string without charset
6456 translation. Translation happens only if iconv() is available; HAVE_ICONV
6457 indicates this at build time. HEADERS_CHARSET gives the charset to
6458 translate to; headers_charset can change it in the configuration, and
6459 "headers charset" can change it in an individual filter file.
6461 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6462 that creates RFC 2047 encoded "words" labels them as that charset instead
6463 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6464 expansion operator; (ii) when Exim creates a From: line for a local
6465 message; (iii) when a header line is rewritten to include a "phrase" part.
6467 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6468 buggy, causing it to skip the first lines of messages whose message ID
6469 ended in 'D'. This would not have bitten before Exim release 4.14, because
6470 message IDs were unlikely to end in 'D' before then. The effect was to have
6471 incorrect size information for certain domains.
6473 45. #include "config.h" was missing at the start of the crypt16.c module. This
6474 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6477 46. If there was a timeout during a "random" callout check, Exim treated it as
6478 a failure of the random address, and carried on sending RSET and the real
6479 address. If the delay was just some slowness somewhere, the response to the
6480 original RCPT would be taken as a response to RSET and so on, causing
6481 mayhem of various kinds.
6483 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6484 when I implemented it. It didn't allow for the fact that some option values
6485 may legitimately be negative (e.g. size_addition), and it didn't even do
6486 the right test for positive values.
6488 48. Domain names in DNS records are case-independent. Exim always looks them up
6489 in lower case. Some resolvers return domain names in exactly the case they
6490 appear in the zone file, that is, they may contain uppercase letters. Not
6491 all resolvers do this - some return always lower case. Exim was treating a
6492 change of case by a resolver as a change of domain, similar to a widening
6493 of a domain abbreviation. This triggered its re-routing code and so it was
6494 trying to route what was effectively the same domain again. This normally
6495 caused routing to fail (because the router wouldn't handle the domain
6496 twice). Now Exim checks for this case specially, and just changes the
6497 casing of the domain that it ultimately uses when it transmits the message
6500 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6503 50. If a filter generated a file delivery with a non-absolute name (possible if
6504 no home directory exists for the router), the forbid_file option was not
6507 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6508 the existing equality matching.
6510 52. Exim was using ints instead of ino_t variables in some places where it was
6511 dealing with inode numbers.
6513 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6514 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6515 called TMPDIR, and if it finds it is different, it changes its value.
6517 54. The smtp_printf() function is now made available to local_scan() so
6518 additional output lines can be written before returning. There is also an
6519 smtp_fflush() function to enable the detection of a dropped connection.
6520 The variables smtp_input and smtp_batched_input are exported to
6523 55. Changed the default runtime configuration: the message "Unknown user"
6524 has been removed from the ACL, and instead placed on the localuser router,
6525 using the cannot_route_message feature. This means that any verification
6526 failures that generate their own messages won't get overridden. Similarly,
6527 the "Unrouteable address" message that was in the ACL for unverifiable
6528 relay addresses has also been removed.
6530 56. Added hosts_avoid_esmtp to the smtp transport.
6532 57. The exicyclog script was not checking for the esoteric option
6533 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6534 will work only if exicyclog is run under the appropriate euid.
6536 58. Following a discussion on the list, the rules by which Exim recognises line
6537 endings on incoming messages have been changed. The -dropcr and drop_cr
6538 options are now no-ops, retained only for backwards compatibility. The
6539 following line terminators are recognized: LF CRLF CR. However, special
6540 processing applies to CR:
6542 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6543 nor a local message in the state where . is a terminator.
6545 (ii) If a bare CR is encountered in a header line, an extra space is added
6546 after the line terminator so as not to end the header. The reasoning
6547 behind this is that bare CRs in header lines are most likely either
6548 to be mistakes, or people trying to play silly games.
6550 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6551 was being incorrectly given as 18 bytes larger than it should have been.
6552 This is a VOB (very old bug).
6554 60. This may never have affected anything current, but just in case it has:
6555 When the local host is found other than at the start of a list of hosts,
6556 the local host, those with the same MX, and any that follow, are discarded.
6557 When the list in question was part of a longer list of hosts, the following
6558 hosts (not currently being processed) were also being discarded. This no
6559 longer happens. I'm not sure if this situation could ever has previously
6562 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6565 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6566 header. This is something that is recommended in a new Internet Draft, and
6567 is something that is documented as being done by Sendmail. There are two
6568 possible values. For messages generated by the autoreply transport, Exim
6571 Auto-Submitted: auto-replied
6573 whereas for all other generated messages (e.g. bounces) it adds
6575 Auto-Submitted: auto-generated
6577 63. The "personal" condition in filters now includes a test for the
6578 Auto-Submitted: header. If it contains the string "auto-" the message it
6579 not considered personal.
6581 64. Added rcpt_include_affixes as a generic transport option.
6583 65. Added queue_only_override (default true).
6585 66. Added the syslog_duplication option.
6587 67. If what should have been the first header line of a message consisted of
6588 a space followed by a colon, Exim was mis-interpreting it as a header line.
6589 It isn't of course - it is syntactically invalid and should therefore be
6590 treated as the start of the message body. The misbehaviour could have
6591 caused a number of strange effects, including loss of data in subsequent
6592 header lines, and spool format errors.
6594 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6595 client host had authenticated. This control can now be exercised by an ACL
6596 for more flexibility.
6598 69. By default, callouts do not happen when testing with -bh. There is now a
6599 variant, -bhc, which does actually run the callout code, including
6600 consulting and updating the callout cache.
6602 70. Added support for saslauthd authentication, courtesy of Alexander
6605 71. If statvfs() failed on the spool or log directories while checking their
6606 size for availability, Exim confusingly gave the error "space shortage".
6607 Furthermore, in debugging mode it crashed with a floating point exception.
6608 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6609 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6610 serious problem, Exim now writes to the main and panic logs when this
6611 happens, with details of the failure. It then refuses to accept the
6612 incoming message, giving the message "spool directory problem" or "log
6613 directory problem" with a 421 code for SMTP messages.
6615 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6616 0, 1, and 2 exist, because some OS complain for execs without them (see
6617 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6618 descriptors. However, the code omitted to check that the open succeeded,
6619 causing mysterious errors if for some reason the permissions on /dev/null
6620 got screwed. Now Exim writes a message to the main and panic logs, and
6621 bombs out if it can't open /dev/null.
6623 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6624 interact so that it is all more flexible. It is supposed to remain
6625 backwards compatible. Also added extra_local_interfaces.
6627 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6628 to bomb out with an assertion failure - to the client this appears as a
6629 connection drop. This problem occurs in the part of the code that was taken
6630 from the Samba project. Fortunately, the assertion is in a very simple
6631 function, so I have fixed this by reproducing the function inline in the
6632 one place where it is called, and arranging for authentication to fail
6633 instead of killing the process with assert().
6635 75. The SPA client code was not working when the server requested OEM rather
6636 than Unicode encoding.
6638 76. Added code to make require_files with a specific uid setting more usable in
6639 the case where statting the file as root fails - usually a non-root-mounted
6640 NFS file system. When this happens and the failure is EACCES, Exim now
6641 forks a subprocess and does the per-uid checking as the relevant uid.
6643 77. Added process_log_path.
6645 78. If log_file_path was not explicitly set, a setting of check_log_space or
6646 check_log_inodes was ignored.
6648 79. If a space check for the spool or log partitions fails, the incident is now
6649 logged. Of course, in the latter case the data may get lost...
6651 80. Added the %p formatting code to string_format() so that it can be used to
6652 print addresses in debug_print(). Adjusted all the address printing in the
6653 debugging in store.c to use %p rather than %d.
6655 81. There was a concern that a line of code in smtp_in.c could overflow a
6656 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6657 initially expressed, the concern was not well-founded, because trailing
6658 spaces are removed early. However, if the trailing spaces were followed by
6659 a NULL, they did not get removed, so the overflow was possible. Two fixes
6662 (a) I re-wrote the offending code in a cleaner fashion.
6663 (b) If an incoming SMTP command contains a NULL character, it is rejected
6666 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6667 runs initgroups(), so that if the Exim user is in any additional groups,
6668 they will be used during message reception.
6674 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob