1 Change log file for Exim from version 4.21
2 -------------------------------------------
6 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
7 supports it and a size is available (ie. the sending peer gave us one).
9 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
10 years of being deprecated). The replacements are the ACLs
11 acl_smtp_mime and acl_not_smtp_mime.
13 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
14 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
15 or even an in-clear connection were permitted. Now, if the host lookup
16 was dnssec and dane was requested then the host is only used if the
17 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
18 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
19 if one fails this test.
20 This means that a poorly-configured remote DNS will make it incommunicado;
21 but it protects against a DNS-interception attack on it.
23 JH/04 Bug 1810: make continued-use of an open smtp transport connection
24 non-noisy when a race steals the message being considered.
26 JH/05 If main configuration option tls_certificate is unset, generate a
27 self-signed certificate for inbound TLS connections.
29 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
30 in rewrites and routers.
32 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
33 and logged a warning sing 4.83; now they are a configuration file error.
35 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
36 (lacking @domain). Apply the same qualification processing as RCPT.
38 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
40 JH/10 Support ${sha256:} applied to a string (as well as the previous
43 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
44 a cutthrough deliver is pending, as we always want to make a connection.
45 This also avoids re-routing the message when later placing the cutthrough
46 connection after a verify cache hit.
47 Do not update it with the verify result either.
49 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
50 when routing results in more than one destination address.
52 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
53 signing (which inhibits the cutthrough capability). Previously only
54 the presence of an option was tested; now an expansion evaluating as
55 empty is permissible (obviously it should depend only on data available
56 when the cutthrough connection is made).
58 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
59 the relevant preceding SMTP command did not note the pipelining mode.
61 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
62 Previously they were not counted.
64 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
65 as one having no matching records. Previously we deferred the message
66 that needed the lookup.
68 JH/17 Fakereject: previously logged as a norml message arrival "<="; now
69 distinguished as "(=".
71 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
72 for missing MX records. Previously it only worked for missing A records.
74 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
76 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
77 after the data-go-ahead and data-ack. Patch from Jason Betts.
79 JH/21 Bug 1846: Send DMARC forensic reports for reject and quaratine results,
80 even for a "none" policy. Patch from Tony Meyer.
82 JH/22 Fix continued use of a connection for further deliveries. If a port was
83 specified by a router, it must also match for the delivery to be
86 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
87 When one of the recipients of a message was unsuitable for the connection
88 (has no matching addresses), we lost track of needing to mark it
89 deferred. As a result mail would be lost.
91 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
93 JH/25 Decoding ACL controls is now done using a binary search; the source code
94 takes up less space and should be simpler to maintain. Merge the ACL
95 condition decode tables also, with similar effect.
97 JH/26 Fix problem with one_time used on a redirect router which returned the
98 parent address unchanged. A retry would see the parent address marked as
99 delivered, so not attempt the (identical) child. As a result mail would
102 JH/27 Fix a possible security hole, wherein a process operating with the Exim
103 UID can gain a root shell. Credit to http://www.halfdog.net/ for
104 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
107 JH/28 Enable {spool,log} filesystem space and inode checks as default.
108 Main config options check_{log,spool}_{inodes,space} are now
109 100 inodes, 10MB unless set otherwise in the configuration.
111 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
112 Previously it only applied to the main-section connection policy
115 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
117 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
118 by me. Added RFC7919 DH primes as an alternative.
120 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
121 are not in the system include path.
123 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
124 GnuTLS, when a session startup failed (eg because the client disconnected)
125 Exim did stdio operations after fclose. This was exposed by a recent
126 change which nulled out the file handle after the fclose.
128 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
129 signed directly by the cert-signing cert, rather than an intermediate
130 OCSP-signing cert. This is the model used by LetsEncrypt.
135 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
136 and 3.4.4 - once the server is enabled to respond to an OCSP request
137 it does even when not requested, resulting in a stapling non-aware
138 client dropping the TLS connection.
140 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
141 support variable-length bit vectors. No functional change.
143 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
144 The I= interface field on outgoing lines is now after the H= remote
145 host field, same as incoming lines. There is a separate
146 outgoing_interface log selector which allows you to disable the
149 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
150 If not running log_selector +smtp_connection the mainlog would be held
151 open indefinitely after a "too many connections" event, including to a
152 deleted file after a log rotate. Leave the per net connection logging
153 leaving it open for efficiency as that will be quickly detected by the
154 check on the next write.
156 HS/01 Bug 1671: Fix post transport crash.
157 Processing the wait-<transport> messages could crash the delivery
158 process if the message IDs didn't exist for some reason. When
159 using 'split_spool_directory=yes' the construction of the spool
160 file name failed already, exposing the same netto behaviour.
162 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
163 mime_regex ACL conditions.
165 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
166 to DSN fail messages (bounces): remote IP, remote greeting, remote response
167 to HELO, local diagnostic string.
169 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
170 log line to debug. Even when configured with a tls authenticator many
171 client connections are expected to not authenticate in this way, so
172 an authenticate fail is not an error.
174 HS/02 Add the Exim version string to the process info. This way exiwhat
175 gives some more detail about the running daemon.
177 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
178 matter for fast-change records such as DNSBLs.
180 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
181 retry record, even if constant. There may be multiple transports with
182 different interface settings and the retry behaviour needs to be kept
185 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
187 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
189 JH/10 Bug 840: fix log_defer_output option of pipe transport
191 JH/11 Bug 830: use same host for all RCPTS of a message, even under
192 hosts_randomize. This matters a lot when combined with mua_wrapper.
194 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
195 ${quote_pgsql:<string>} operator.
197 JH/13 Bug 1708: avoid misaligned access in cached lookup.
199 JH/14 Change header file name for freeradius-client. Relevant if compiling
200 with Radius support; from the Gentoo tree and checked under Fedora.
202 JH/15 Bug 1712: Introduce $prdr_requested flag variable
204 JH/16 Bug 1714: Permit an empty string as expansion result for transport
205 option transport_filter, meaning no filtering.
207 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
209 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
210 defaults to "*" (all hosts). The variable is now available when not built
211 with TLS, default unset, mainly to enable keeping the testuite sane.
212 If a server certificate is not supplied (via tls_certificate) an error is
213 logged, and clients will find TLS connections fail on startup. Presumably
214 they will retry in-clear.
215 Packagers of Exim are strongly encouraged to create a server certificate
216 at installation time.
218 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
219 with the $config_file variable.
221 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
222 in transport context, after the attempt, and per-recipient. The latter type
223 is per host attempted. The event data is the error message, and the errno
224 information encodes the lookup type (A vs. MX) used for the (first) host,
225 and the trailing two digits of the smtp 4xx reponse.
227 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
228 to write to mainlog (or rejectlog, paniclog) in the window between file
229 creation and permissions/ownership being changed. Particularly affects
230 installations where exicyclog is run as root, rather than exim user;
231 result is that the running daemon panics and dies.
233 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
235 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
236 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
237 "pri" and "weight". Note that the previous implicit priority given by the
238 list order is no longer honoured.
240 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
243 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
244 by defining SUPPORT_SOCKS.
246 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
247 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
248 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
249 variables are renamed to proxy_{local,external}_{address,port}.
251 JH/27 Move Internationalisation support from Experimental to mainline, enabled
252 for a build by defining SUPPORT_I18N
254 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
255 of the query string, and make ${quote_redis:} do that quoting.
257 JH/29 Move Events support from Experimental to mainline, enabled by default
258 and removable for a build by defining DISABLE_EVENT.
260 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
262 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
263 cached by the daemon.
265 JH/32 Move Redis support from Experimental to mainline, enabled for a build
266 by defining LOOKUP_REDIS. The libhiredis library is required.
268 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
269 keys are given for lookup.
271 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
272 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
273 only supported when built with TLS support. The PolarSSL SHA routines
274 are still used when the TLS library is too old for convenient support.
276 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
277 openssl_options), for security. OpenSSL forces this from version 1.1.0
278 server-side so match that on older versions.
280 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
281 allocation for $value could be released as the expansion processing
282 concluded, but leaving the global pointer active for it.
284 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
285 and to use the domains and local_parts ACL conditions.
287 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
288 incorrectly not doubled on cutthrough transmission, hence seen as a
289 body-termination at the receiving system - resulting in truncated mails.
290 Commonly the sender saw a TCP-level error, and retransmitted the message
291 via the normal store-and-forward channel. This could result in duplicates
292 received - but deduplicating mailstores were liable to retain only the
293 initial truncated version.
295 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
297 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
299 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
300 we're in there, support oversigning also; bug 1309.
302 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
304 HS/04 Add support for keep_environment and add_environment options.
306 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
307 either intentional arithmetic overflow during PRNG, or testing config-
310 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
311 delivery resulted in actual delivery. Cancel cutthrough before DATA
314 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
315 rejects a recipient: pass the reject to the originator.
317 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
318 Many were false-positives and ignorable, but it's worth fixing the
321 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
322 for the new environment-manipulation done at startup. Move the routines
323 from being local to tls.c to being global via the os.c file.
325 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
326 an extract embedded as result-arg for a map, the first arg for extract
327 is unavailable so we cannot tell if this is a numbered or keyed
328 extraction. Accept either.
334 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
337 JH/02 The smtp transport option "multi_domain" is now expanded.
339 JH/03 The smtp transport now requests PRDR by default, if the server offers
342 JH/04 Certificate name checking on server certificates, when exim is a client,
343 is now done by default. The transport option tls_verify_cert_hostnames
344 can be used to disable this per-host. The build option
345 EXPERIMENTAL_CERTNAMES is withdrawn.
347 JH/05 The value of the tls_verify_certificates smtp transport and main options
348 default to the word "system" to access the system default CA bundle.
349 For GnuTLS, only version 3.0.20 or later.
351 JH/06 Verification of the server certificate for a TLS connection is now tried
352 (but not required) by default. The verification status is now logged by
353 default, for both outbound TLS and client-certificate supplying inbound
356 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
359 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
360 Status Notification (bounce) messages are now MIME format per RFC 3464.
361 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
362 under the control of the dsn_advertise_hosts option, and routers may
363 have a dsn_lasthop option.
365 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
366 default, modifiable by a malware= option. The list separator for
367 the options can now be changed in the usual way. Bug 68.
369 JH/10 The smtp_receive_timeout main option is now expanded before use.
371 JH/11 The incoming_interface log option now also enables logging of the
372 local interface on delivery outgoing connections.
374 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
375 if the interface and destination host and port all match.
377 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
380 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
381 Patch from Andrew Lewis.
383 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
384 now supports optional time-restrictions, weighting, and priority
385 modifiers per server. Patch originally by <rommer@active.by>.
387 JH/16 The spamd_address main option now supports a mixed list of local
388 and remote servers. Remote servers can be IPv6 addresses, and
389 specify a port-range.
391 JH/17 Bug 68: The spamd_address main option now supports an optional
392 timeout value per server.
394 JH/18 Bug 1581: Router and transport options headers_add/remove can
395 now have the list separator specified.
397 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
400 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
403 JH/21 Support for the A6 type of dns record is withdrawn.
405 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
406 rather than the verbs used.
408 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
409 from 255 to 1024 chars.
411 JH/24 Verification callouts now attempt to use TLS by default.
413 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
414 are generic router options now. The defaults didn't change.
416 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
417 Original patch from Alexander Shikoff, worked over by JH.
419 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
420 files if ZCAT_COMMAND is not executable.
422 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
424 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
426 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
427 Normally benign, it bites when the pair was led to by a CNAME;
428 modern usage is to not canoicalize the domain to a CNAME target
429 (and we were inconsistent anyway for A-only vs AAAA+A).
431 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
433 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
434 when evaluating $sender_host_dnssec.
436 JH/31 Check the HELO verification lookup for DNSSEC, adding new
437 $sender_helo_dnssec variable.
439 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
441 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
443 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
445 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
446 documented as working, but never had. Support all but $spam_report.
448 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
449 added for tls authenticator.
451 HS/03 Add perl_taintmode main config option
456 TL/01 When running the test suite, the README says that variables such as
457 no_msglog_check are global and can be placed anywhere in a specific
458 test's script, however it was observed that placement needed to be near
459 the beginning for it to behave that way. Changed the runtest perl
460 script to read through the entire script once to detect and set these
461 variables, reset to the beginning of the script, and then run through
462 the script parsing/test process like normal.
464 TL/02 The BSD's have an arc4random API. One of the functions to induce
465 adding randomness was arc4random_stir(), but it has been removed in
466 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
467 function when detected.
469 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
470 cause callback expansion.
472 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
473 syntax errors in an expansion can be treated as a string instead of
474 logging or causing an error, due to the internal use of bool_lax
475 instead of bool when processing it.
477 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
478 server certificates when making smtp deliveries.
480 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
482 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
484 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
486 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
487 Merged patch from Sebastian Wiedenroth.
489 JH/05 Fix results-pipe from transport process. Several recipients, combined
490 with certificate use, exposed issues where response data items split
491 over buffer boundaries were not parsed properly. This eventually
492 resulted in duplicates being sent. This issue only became common enough
493 to notice due to the introduction of conection certificate information,
494 the item size being so much larger. Found and fixed by Wolfgang Breyha.
496 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
497 size buffer was used, resulting in syntax errors when an expansion
500 JH/07 Add support for directories of certificates when compiled with a GnuTLS
501 version 3.3.6 or later.
503 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
504 is EXPERIMENTAL_EVENT, the main-configuration and transport options
505 both become "event_action", the variables become $event_name, $event_data
506 and $event_defer_errno. There is a new variable $verify_mode, usable in
507 routers, transports and related events. The tls:cert event is now also
508 raised for inbound connections, if the main configuration event_action
511 TL/06 In test suite, disable OCSP for old versions of openssl which contained
512 early OCSP support, but no stapling (appears to be less than 1.0.0).
514 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
515 server certificate names available under the smtp transport option
516 "tls_verify_cert_hostname" now do not permit multi-component wildcard
519 JH/10 Time-related extraction expansions from certificates now use the main
520 option "timezone" setting for output formatting, and are consistent
521 between OpenSSL and GnuTLS compilations. Bug 1541.
523 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
524 encoded parameter in the incoming message. Bug 1558.
526 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
527 include certificate info, eximon was claiming there were spoolfile
530 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
532 JH/14 Log delivery-related information more consistently, using the sequence
533 "H=<name> [<ip>]" wherever possible.
535 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
536 are problematic for Debian distribution, omit them from the release
539 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
541 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
544 JH/17 Fix a null-indirection in certextract expansions when a nondefault
545 output list separator was used.
550 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
551 checkers that were complaining about end of non-void function with no
554 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
555 This was a regression intruduced in 4.83 by another bugfix.
557 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
559 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
560 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
566 TF/01 Correctly close the server side of TLS when forking for delivery.
568 When a message was received over SMTP with TLS, Exim failed to clear up
569 the incoming connection properly after forking off the child process to
570 deliver the message. In some situations the subsequent outgoing
571 delivery connection happened to have the same fd number as the incoming
572 connection previously had. Exim would try to use TLS and fail, logging
573 a "Bad file descriptor" error.
575 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
576 utilities have not been installed.
578 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
579 temporary space as the ACL may create new global variables.
581 TL/01 LDAP support uses per connection or global context settings, depending
582 upon the detected version of the libraries at build time.
584 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
585 to extract and use the src ip:port in logging and expansions as if it
586 were a direct connection from the outside internet. PPv2 support was
587 updated based on HAProxy spec change in May 2014.
589 JH/02 Add ${listextract {number}{list}{success}{fail}}.
591 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
592 Properly escape header and check for NULL return.
594 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
597 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
599 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
600 characters in header names, implemented as a verify condition.
601 Contributed by Michael Fischer v. Mollard.
603 TL/05 Rename SPF condition results err_perm and err_temp to standardized
604 results permerror and temperror. Previous values are deprecated but
605 still accepted. In a future release, err_perm and err_temp will be
606 completely removed, which will be a backward incompatibility if the
607 ACL tests for either of these two old results. Patch contributed by
608 user bes-internal on the mailing list.
610 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
612 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
613 selectors, in both main and reject logs.
615 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
618 JH/07 Add malware type "sock" for talking to simple daemon.
620 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
622 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
623 routers/transports under cutthrough routing.
625 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
626 numbers. Touch up "bool" conditional to keep the same definition.
628 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
630 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
632 JH/12 Expand items in router/transport headers_add or headers_remove lists
633 individually rather than the list as a whole. Bug 1452.
635 Required for reasonable handling of multiple headers_ options when
636 they may be empty; requires that headers_remove items with embedded
637 colons must have them doubled (or the list-separator changed).
639 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
640 view the policy declared in the DMARC record. Currently, $dmarc_status
641 is a combined value of both the record presence and the result of the
644 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
646 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
647 dnslookup router and the smtp transport (applying to the forward
650 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
651 of ldap servers used for a specific lookup. Patch provided by Heiko
654 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
655 New variable $lookup_dnssec_authenticated for observability.
657 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
658 Patch submitted by Lars Timman.
660 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
662 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
663 Requires trusted mode and valid format message id, aborts otherwise.
664 Patch contributed by Heiko Schlichting.
666 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
667 certextract with support for various fields. Bug 1358.
669 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
670 is requested by default, modifiable by smtp transport option
673 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
674 operate on certificate variables to give certificate fingerprints
675 Also new ${sha256:cert_variable}.
677 JH/23 The PRDR feature is moved from being Experimental into the mainline.
679 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
680 Christian Aistleitner.
682 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
684 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
685 file. Patch from Wolfgang Breyha.
687 JH/25 Expand the coverage of the delivery $host and $host_address to
688 client authenticators run in verify callout. Bug 1476.
690 JH/26 Port service names are now accepted for tls_on_connect_ports, to
691 align with daemon_smtp_ports. Bug 72.
693 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
694 support and error reporting did not work properly.
696 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
697 and is readable. Patch from Andrew Colin Kissa.
699 TL/14 Enhance documentation of ${run expansion and how it parses the
700 commandline after expansion, particularly in the case when an
701 unquoted variable expansion results in an empty value.
703 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
705 PP/02 Fix internal collision of T_APL on systems which support RFC3123
706 by renaming away from it. Addresses GH issue 15, reported by
709 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
711 TL/15 SECURITY: prevent double expansion in math comparison functions
712 (can expand unsanitized data). Not remotely exploitable.
719 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
721 PP/02 Make -n do something, by making it not do something.
722 When combined with -bP, the name of an option is not output.
724 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
727 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
728 $sender_host_name and config options to manage this, and basic check
731 PP/05 DSCP support for outbound connections and control modifier for inbound.
733 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
734 (Only plugin which currently uses this is kerberos4, which nobody should
735 be using, but we should make it available and other future plugins might
736 conceivably use it, even though it would break NAT; stuff *should* be
737 using channel bindings instead).
739 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
740 name; added for Sendmail compatibility; requires admin caller.
741 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
742 just ignore it); requires trusted caller.
743 Also parse but ignore: -Ac -Am -X<logfile>
746 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
748 TL/02 Add +smtp_confirmation as a default logging option.
750 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
751 Patch by Magnus Holmgren from 2007-02-20.
753 TL/04 Bugzilla 1281 - Spec typo.
754 Bugzilla 1283 - Spec typo.
755 Bugzilla 1290 - Spec grammar fixes.
757 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
759 TL/06 Add Experimental DMARC support using libopendmarc libraries.
761 TL/07 Fix an out of order global option causing a segfault. Reported to dev
762 mailing list by by Dmitry Isaikin.
764 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
766 JH/02 Support "G" suffix to numbers in ${if comparisons.
768 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
770 NM/01 Bugzilla 1197 - Spec typo
771 Bugzilla 1196 - Spec examples corrections
773 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
775 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
776 gnutls_enable_pkcs11, but renamed to more accurately indicate its
779 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
780 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
782 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
783 "acl {{name}{arg}...}", and optional args on acl condition
786 JH/05 Permit multiple router/transport headers_add/remove lines.
788 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
790 JH/07 Avoid using a waiting database for a single-message-only transport.
791 Performance patch from Paul Fisher. Bugzilla 1262.
793 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
796 JH/09 Add $headers_added variable, with content from use of ACL modifier
797 add_header (but not yet added to the message). Bugzilla 199.
799 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
800 Pulled from Bugzilla 817 by Wolfgang Breyha.
802 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
804 (nb: this is the same fix as in Exim 4.80.1)
806 JH/11 Add A= logging on delivery lines, and a client_set_id option on
809 JH/12 Add optional authenticated_sender logging to A= and a log_selector
812 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
814 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
815 advertise SMTP AUTH mechanism to us, instead of a generic
816 protocol violation error. Also, make Exim more robust to bad
817 data from the Dovecot auth socket.
819 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
821 When a queue runner is handling a message, Exim first routes the
822 recipient addresses, during which it prunes them based on the retry
823 hints database. After that it attempts to deliver the message to
824 any remaining recipients. It then updates the hints database using
827 So if a recipient address works intermittently, it can get repeatedly
828 deferred at routing time. The retry hints record remains fresh so the
829 address never reaches the final cutoff time.
831 This is a fairly common occurrence when a user is bumping up against
832 their storage quota. Exim had some logic in its local delivery code
833 to deal with this. However it did not apply to per-recipient defers
834 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
836 This change adds a proper retry rule check during routing so that the
837 final cutoff time is checked against the message's age. We only do
838 this check if there is an address retry record and there is not a
839 domain retry record; this implies that previous attempts to handle
840 the address had the retry_use_local_parts option turned on. We use
841 this as an approximation for the destination being like a local
842 delivery, as in LMTP.
844 I suspect this new check makes the old local delivery cutoff check
845 redundant, but I have not verified this so I left the code in place.
847 TF/02 Correct gecos expansion when From: is a prefix of the username.
849 Test 0254 submits a message to Exim with the header
853 When I ran the test suite under the user fanf2, Exim expanded
854 the header to contain my full name, whereas it should have added
855 a Resent-Sender: header. It erroneously treats any prefix of the
856 username as equal to the username.
858 This change corrects that bug.
860 GF/01 DCC debug and logging tidyup
861 Error conditions log to paniclog rather than rejectlog.
862 Debug lines prefixed by "DCC: " to remove any ambiguity.
864 TF/03 Avoid unnecessary rebuilds of lookup-related code.
866 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
867 Bug spotted by Jeremy Harris; was flawed since initial commit.
868 Would have resulted in OCSP responses post-SNI triggering an Exim
869 NULL dereference and crash.
871 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
873 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
874 Bug detection, analysis and fix by Samuel Thibault.
875 Bugzilla 1331, Debian bug #698092.
877 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
879 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
880 Server implementation by Todd Lyons, client by JH.
881 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
882 config variable "prdr_enable" controls whether the server
883 advertises the facility. If the client requests PRDR a new
884 acl_data_smtp_prdr ACL is called once for each recipient, after
885 the body content is received and before the acl_smtp_data ACL.
886 The client is controlled by bolth of: a hosts_try_prdr option
887 on the smtp transport, and the server advertisement.
888 Default client logging of deliveries and rejections involving
889 PRDR are flagged with the string "PRDR".
891 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
892 fclose(). Diagnosis by Todd Lyons.
894 PP/17 Update configure.default to handle IPv6 localhost better.
895 Patch by Alain Williams (plus minor tweaks).
898 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
899 This is now consistent with GnuTLS, and is now documented: the
900 previous undocumented portable approach to treating the option as
901 unset was to force an expansion failure. That still works, and
902 an empty string is now equivalent.
904 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
905 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
906 not performing validation itself.
908 PP/20 Added force_command boolean option to pipe transport.
909 Patch from Nick Koston, of cPanel Inc.
911 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
914 TF/04 Added udpsend ACL modifer and hexquote expansion operator
916 PP/21 Fix eximon continuous updating with timestamped log-files.
917 Broken in a format-string cleanup in 4.80, missed when I repaired the
918 other false fix of the same issue.
919 Report and fix from Heiko Schlichting.
922 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
923 Report from Prashanth Katuri.
925 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
926 It's SecureTransport, so affects any MacOS clients which use the
927 system-integrated TLS libraries, including email clients.
929 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
930 using a MIME ACL for non-SMTP local injection.
931 Report and assistance in diagnosis by Warren Baker.
933 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
935 JH/16 Fix comparisons for 64b. Bugzilla 1385.
937 TL/09 Add expansion variable $authenticated_fail_id to keep track of
938 last id that failed so it may be referenced in subsequent ACL's.
940 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
943 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
944 ldap library initialization, allowing self-signed CA's to be
945 used. Also properly sets require_cert option later in code by
946 using NULL (global ldap config) instead of ldap handle (per
947 session). Bug diagnosis and testing by alxgomz.
949 TL/12 Enhanced documentation in the ratelimit.pl script provided in
950 the src/util/ subdirectory.
952 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
953 renamed to Transport Post Delivery Action by Jeremy Harris, as
956 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
957 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
958 redis_servers = needs to be configured which will be used by the redis
959 lookup. Patch from Warren Baker, of The Packet Hub.
961 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
963 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
964 hostname or reverse DNS when processing a host list. Used suggestions
965 from multiple comments on this bug.
967 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
969 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
970 Missed a few lines, added it to make the runtest require no keyboard
973 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
974 contains upper case chars. Make router use caseful_local_part.
976 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
977 support when GnuTLS has been built with p11-kit.
983 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
985 This, or similar/improved, will also be change PP/11 of 4.82.
991 PP/01 Handle short writes when writing local log-files.
992 In practice, only affects FreeBSD (8 onwards).
993 Bugzilla 1053, with thanks to Dmitry Isaikin.
995 NM/01 Bugzilla 949 - Documentation tweak
997 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1000 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1002 PP/02 Implemented gsasl authenticator.
1004 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1006 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1007 `pkg-config foo` for cflags/libs.
1009 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1010 with rest of GSASL and with heimdal_gssapi.
1012 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1013 `pkg-config foo` for cflags/libs for the TLS implementation.
1015 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1016 properties get this fed in as external SSF. A number of robustness
1017 and debugging improvements to the cyrus_sasl authenticator.
1019 PP/08 cyrus_sasl server now expands the server_realm option.
1021 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1022 Patch by Jeremy Harris.
1024 PP/10 Added dbmjz lookup type.
1026 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1028 PP/12 MAIL args handles TAB as well as SP, for better interop with
1029 non-compliant senders.
1030 Analysis and variant patch by Todd Lyons.
1032 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1033 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1034 Patch from Dirk Mueller <dmueller@suse.com>
1036 PP/13 tls_peerdn now print-escaped for spool files.
1037 Observed some $tls_peerdn in wild which contained \n, which resulted
1038 in spool file corruption.
1040 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1041 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1042 or write after TLS renegotiation, which otherwise led to messages
1045 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1046 as a tracking header (ie: a signed header comes before the signature).
1047 Patch from Wolfgang Breyha.
1049 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1050 comma-sep list; embedded commas doubled.
1052 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1054 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1056 Report and patch from Dmitry Banschikov.
1058 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1059 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1060 failures. We appear to now support TLS1.1+ with Exim.
1062 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1063 lets Exim select keys and certificates based upon TLS SNI from client.
1064 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1065 before an outbound SMTP session. New log_selector, +tls_sni.
1067 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1068 NULL dereference. Report and patch from Alun Jones.
1070 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1071 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1072 Not seeing resolver debug output on NetBSD, but suspect this is a
1073 resolver implementation change.
1075 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1076 Left warnings. Added "eximon gdb" invocation mode.
1078 PP/21 Defaulting "accept_8bitmime" to true, not false.
1080 PP/22 Added -bw for inetd wait mode support.
1082 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1083 locate the relevant includes and libraries. Made this the default.
1085 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1086 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1088 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1089 This may cause build issues on older platforms.
1091 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1092 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1093 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1094 Added SNI support via GnuTLS too.
1095 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1097 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1099 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1100 Applied second patch from Janne, implementing suggestion to default
1101 multiple-strings-in-record handling to match SPF spec.
1103 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1105 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1106 read-only, out of scope).
1107 Patch from Wolfgang Breyha, report from Stuart Northfield.
1109 PP/29 Fix three issues highlighted by clang analyser static analysis.
1110 Only crash-plausible issue would require the Cambridge-specific
1111 iplookup router and a misconfiguration.
1112 Report from Marcin Mirosław.
1114 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1116 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1117 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1118 As part of this, removing so much warning spew let me fix some minor
1119 real issues in debug logging.
1121 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1122 assignment on my part. Fixed.
1124 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1125 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1126 Janne Snabb (who went above and beyond: thank you).
1128 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1129 string otherwise requires a connection and a bunch more work and it's
1130 relatively easy to get wrong. Should also expose TLS library linkage
1133 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1134 64-bit ${eval} (JH/03).
1136 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1137 GNU libc to support some of the 64-bit stuff, should not lead to
1138 conflicts. Defined before os.h is pulled in, so if a given platform
1139 needs to override this, it can.
1141 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1142 protection layer was required, which is not implemented.
1143 Bugzilla 1254, patch from Wolfgang Breyha.
1145 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1146 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1147 tls_dhparam take prime identifiers. Also unbreak combination of
1148 OpenSSL+DH_params+TLSSNI.
1150 PP/39 Disable SSLv2 by default in OpenSSL support.
1156 PP/01 Solaris build fix for Oracle's LDAP libraries.
1157 Bugzilla 1109, patch from Stephen Usher.
1159 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1161 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1164 TF/02 Fix a couple more cases where we did not log the error message
1165 when unlink() failed. See also change 4.74-TF/03.
1167 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1168 lock up or crash if it happened to be inside a call to libc when it
1169 got a SIGUSR1 from exiwhat.
1171 The SIGUSR1 handler appends the current process status to the process
1172 log which is later printed by exiwhat. It used to use the general
1173 purpose logging code to do this, but several functions it calls are
1174 not safe for signals.
1176 The new output code in the SIGUSR1 handler is specific to the process
1177 log, and simple enough that it's easy to inspect for signal safety.
1178 Removing some special cases also simplifies the general logging code.
1179 Removing the spurious timestamps from the process log simplifies
1182 TF/04 Improved ratelimit ACL condition.
1184 The /noupdate option has been deprecated in favour of /readonly which
1185 has clearer semantics. The /leaky, /strict, and /readonly update modes
1186 are mutually exclusive. The update mode is no longer included in the
1187 database key; it just determines when the database is updated. (This
1188 means that when you upgrde Exim will forget old rate measurements.)
1190 Exim now checks that the per_* options are used with an update mode that
1191 makes sense for the current ACL. For example, when Exim is processing a
1192 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1193 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1194 must specify per_mail/readonly. If you omit the update mode it defaults to
1195 /leaky where that makes sense (as before) or /readonly where required.
1197 The /noupdate option is now undocumented but still supported for
1198 backwards compatibility. It is equivalent to /readonly except that in
1199 ACLs where /readonly is required you may specify /leaky/noupdate or
1200 /strict/noupdate which are treated the same as /readonly.
1202 A useful new feature is the /count= option. This is a generalization
1203 of the per_byte option, so that you can measure the throughput of other
1204 aggregate values. For example, the per_byte option is now equivalent
1205 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1207 The per_rcpt option has been generalized using the /count= mechanism
1208 (though it's more complicated than the per_byte equivalence). When it is
1209 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1210 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1211 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1212 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1213 Note that using per_rcpt with a non-readonly update mode in more than
1214 one ACL will cause the recipients to be double-counted. (The per_mail
1215 and per_byte options don't have this problem.)
1217 The handling of very low rates has changed slightly. If the computed rate
1218 is less than the event's count (usually one) then this event is the first
1219 after a long gap. In this case the rate is set to the same as this event's
1220 count, so that the first message of a spam run is counted properly.
1222 The major new feature is a mechanism for counting the rate of unique
1223 events. The new per_addr option counts the number of different
1224 recipients that someone has sent messages to in the last time period. It
1225 behaves like per_rcpt if all the recipient addresses are different, but
1226 duplicate recipient addresses do not increase the measured rate. Like
1227 the /count= option this is a general mechanism, so the per_addr option
1228 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1229 example, measure the rate that a client uses different sender addresses
1230 with the options per_mail/unique=$sender_address. There are further
1231 details in the main documentation.
1233 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1235 TF/06 Removed a few PCRE remnants.
1237 TF/07 Automatically extract Exim's version number from tags in the git
1238 repository when doing development or release builds.
1240 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1241 Bugzilla 879. Patch from Paul Fisher.
1243 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1244 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1247 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1249 PP/05 Variable $av_failed, true if the AV scanner deferred.
1250 Bugzilla 1078. Patch from John Horne.
1252 PP/06 Stop make process more reliably on build failure.
1253 Bugzilla 1087. Patch from Heiko Schlittermann.
1255 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1256 Bugzilla 1089. Patch from Heiko Schlittermann.
1258 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1259 Bugzilla 1131. Patch from Holger Weiß.
1261 PP/09 Handle IPv6 addresses with SPF.
1262 Bugzilla 860. Patch from Wolfgang Breyha.
1264 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1266 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1269 PP/11 match_* no longer expand right-hand-side by default.
1270 New compile-time build option, EXPAND_LISTMATCH_RHS.
1271 New expansion conditions, "inlist", "inlisti".
1273 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1275 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1277 PP/14 fix log_write() format string regression from TF/03.
1278 Bugzilla 1152. Patch from Dmitry Isaikin.
1284 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1286 PP/02 Harmonised TLS library version reporting; only show if debugging.
1287 Layout now matches that introduced for other libraries in 4.74 PP/03.
1289 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1291 PP/04 New "dns_use_edns0" global option.
1293 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1296 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1297 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1299 TK/01 Updated PolarSSL code to 0.14.2.
1300 Bugzilla 1097. Patch from Andreas Metzler.
1302 PP/07 Catch divide-by-zero in ${eval:...}.
1303 Fixes bugzilla 1102.
1305 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1308 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1309 format-string attack -- SECURITY: remote arbitrary code execution.
1311 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1312 time unintentionally subject to list matching rules, letting the header
1313 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1314 arbitrary string expansion). This allowed for information disclosure.
1316 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1317 INT_MIN/-1 -- value coerced to INT_MAX.
1323 NM/01 Workround for PCRE version dependency in version reporting
1326 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1327 This fixes portability to compilers other than gcc, notably
1328 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1330 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1331 makefiles for portability to HP-UX and POSIX correctness.
1333 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1334 Also via indented variable definition in the Makefile.
1335 (Debugging by Oliver Heesakkers).
1337 PP/02 Restore caching of spamd results with expanded spamd_address.
1338 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1340 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1341 Improves build reliability. Fix from: Frank Elsner
1343 NM/02 Fix wide character breakage in the rfc2047 coding
1344 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1346 NM/03 Allow underscore in dnslist lookups
1347 Fixes bug 1026. Patch from Graeme Fowler
1349 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1350 Code patches from Adam Ciarcinski of NetBSD.
1352 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1355 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1356 is logged, to avoid truncation. Patch from John Horne.
1358 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1359 Patch from Jakob Hirsch.
1361 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1362 SQL string expansion failure details.
1363 Patch from Andrey Oktyabrski.
1365 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1366 Patch from Simon Arlott.
1368 PP/09 New lookups functionality failed to compile on old gcc which rejects
1369 extern declarations in function scope.
1370 Patch from Oliver Fleischmann
1372 PP/10 Use sig_atomic_t for flags set from signal handlers.
1373 Check getgroups() return and improve debugging.
1374 Fixed developed for diagnosis in bug 927 (which turned out to be
1377 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1378 Patch from Mark Zealey.
1380 PP/12 Bugzilla 1056: Improved spamd server selection.
1381 Patch from Mark Zealey.
1383 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1384 Based on patch from Heiko Schlittermann.
1386 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1387 Patch from Uwe Doering, sign-off by Michael Haardt.
1389 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1390 variable declaration deep within a block. Bug and patch from
1393 PP/15 lookups-Makefile IRIX compatibilty coercion.
1395 PP/16 Make DISABLE_DKIM build knob functional.
1397 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1398 Patch by Simon Arlott
1400 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1401 variable argument macros. Our copy now differs from upstream.
1407 TF/01 Failure to get a lock on a hints database can have serious
1408 consequences so log it to the panic log.
1410 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1411 controlled using the smtp_confirmation log selector.
1413 TF/03 Include the error message when we fail to unlink a spool file.
1415 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1416 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1417 for maintaining out-of-tree patches for some time.
1419 PP/01 Bugzilla 139: Documentation and portability issues.
1420 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1421 Handle per-OS dynamic-module compilation flags.
1423 PP/02 Let /dev/null have normal permissions.
1424 The 4.73 fixes were a little too stringent and complained about the
1425 permissions on /dev/null. Exempt it from some checks.
1426 Reported by Andreas M. Kirchwitz.
1428 PP/03 Report version information for many libraries, including
1429 Exim version information for dynamically loaded libraries. Created
1430 version.h, now support a version extension string for distributors
1431 who patch heavily. Dynamic module ABI change.
1433 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1434 privilege escalation vulnerability whereby the Exim run-time user
1435 can cause root to append content of the attacker's choosing to
1438 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1441 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1442 If dropping privileges for untrusted macros, we disabled normal logging
1443 on the basis that it would fail; for the Exim run-time user, this is not
1444 the case, and it resulted in successful deliveries going unlogged.
1445 Fixed. Reported by Andreas Metzler.
1451 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1452 only prepend for the Resent-* case. Fixes regression introduced in
1453 Exim 4.70 by NM/22 for Bugzilla 607.
1455 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1456 increasing numbers of administrators be bitten by this.
1458 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1460 PP/03 Bugzilla 994: added openssl_options main configuration option.
1462 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1464 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1466 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1468 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1469 without a peer certificate, leading to a segfault because of an
1470 assumption that peers always have certificates. Be a little more
1471 paranoid. Problem reported by Martin Tscholak.
1473 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1474 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1475 NB: ClamAV planning to remove STREAM in "middle of 2010".
1476 CL also introduces -bmalware, various -d+acl logging additions and
1477 more caution in buffer sizes.
1479 PP/09 Implemented reverse_ip expansion operator.
1481 PP/10 Bugzilla 937: provide a "debug" ACL control.
1483 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1485 PP/12 Bugzilla 973: Implement --version.
1487 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1489 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1491 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1493 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1494 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1495 ignore trailing whitespace.
1497 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1499 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1502 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1503 Notification from Dr Andrew Aitchison.
1505 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1506 ExtendedDetectionInfo response format.
1507 Notification from John Horne.
1509 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1512 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1513 XSL and documented dependency on system catalogs, with examples of how
1516 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1519 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1520 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1521 configuration file which is writeable by the Exim user or group.
1523 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1524 of configuration files to cover files specified with the -C option if
1525 they are going to be used with root privileges, not just the default
1528 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1529 option (effectively making it always true).
1531 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1532 files to be used while preserving root privileges.
1534 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1535 that rogue child processes cannot use them.
1537 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1538 run-time user, instead of root.
1540 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1541 Exim run-time user without dropping privileges.
1543 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1544 result string, instead of calling string_vformat() twice with the same
1547 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1548 for other users. Others should always drop root privileges if they use
1549 -C on the command line, even for a whitelisted configure file.
1551 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1553 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1559 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1560 $data_path, and $header_path variables; fixed documentation bugs and
1563 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1564 exipick to access non-standard spools, including the "frozen" queue
1567 NM/01 Bugzilla 965: Support mysql stored procedures.
1568 Patch from Alain Williams
1570 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1572 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1573 Patch from Andreas Metzler
1575 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1576 Patch from Kirill Miazine
1578 NM/05 Bugzilla 671: Added umask to procmail example.
1580 JJ/03 installed exipick 20100323.0, fixing doc bug
1582 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1583 directory. Notification and patch from Dan Rosenberg.
1585 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1587 TK/02 Improve log output when DKIM signing operation fails.
1589 MH/01 Treat the transport option dkim_domain as a colon separated
1590 list, not as a single string, and sign the message with each element,
1591 omitting multiple occurences of the same signer.
1593 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1594 Bugzilla 985, 986. Patch by Simon Arlott
1596 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1597 Patch by Simon Arlott
1599 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1600 MBX locking. Notification from Dan Rosenberg.
1606 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1608 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1610 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1612 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1614 NM/04 Bugzilla 915: Flag broken perl installation during build.
1620 TK/01 Added patch by Johannes Berg that expands the main option
1621 "spamd_address" if it starts with a dollar sign.
1623 TK/02 Write list of recipients to X-Envelope-Sender header when building
1624 the mbox-format spool file for content scanning (suggested by Jakob
1627 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1628 (http://www.dcc-servers.net/) support via dccifd. Activated by
1629 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1631 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1632 by Mark Daniel Reidel <mr@df.eu>.
1634 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1635 When building exim an external PCRE library is now needed -
1636 PCRE is a system library on the majority of modern systems.
1637 See entry on PCRE_LIBS in EDITME file.
1639 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1640 conversation. Added nologin parameter to request.
1641 Patch contributed by Kirill Miazine.
1643 TF/01 Do not log submission mode rewrites if they do not change the address.
1645 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1647 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1648 log files in place. Contributed by Roberto Lima.
1650 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1652 TF/03 Bugzilla 615: When checking the local_parts router precondition
1653 after a local_part_suffix or local_part_prefix option, Exim now
1654 does not use the address's named list lookup cache, since this
1655 contains cached lookups for the whole local part.
1657 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1658 Robert Millan. Documentation is in experimental-spec.txt.
1660 TF/04 Bugzilla 668: Fix parallel build (make -j).
1662 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1664 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1665 Patch provided by Jan Srzednicki.
1667 TF/05 Leading white space used to be stripped from $spam_report which
1668 wrecked the formatting. Now it is preserved.
1670 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1671 that they are available at delivery time.
1673 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1675 TF/08 TLS error reporting now respects the incoming_interface and
1676 incoming_port log selectors.
1678 TF/09 Produce a more useful error message if an SMTP transport's hosts
1679 setting expands to an empty string.
1681 NM/06 Bugzilla 744: EXPN did not work under TLS.
1682 Patch provided by Phil Pennock.
1684 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1685 Patch provided by Richard Godbee.
1687 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1688 acl_smtp_notquit, added index entry.
1690 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1691 Patch provided by Eugene Bujak.
1693 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1694 accept(). Patch provided by Maxim Dounin.
1696 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1697 Patch provided by Phil Pennock.
1699 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1701 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1702 Patch provided by Brad "anomie" Jorsch.
1704 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1705 Patch provided by Dean Brooks.
1707 TK/05 Add native DKIM support (does not depend on external libraries).
1709 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1710 Patch provided by Graeme Fowler.
1712 NM/16 Bugzilla 851: Documentation example syntax fix.
1714 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1716 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1719 NM/19 Bugzilla 745: TLS version reporting.
1720 Patch provided by Phil Pennock.
1722 NM/20 Bugzilla 167: bool: condition support.
1723 Patch provided by Phil Pennock.
1725 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1726 clients. Patch provided by Phil Pennock.
1728 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1729 Patch provided by Brad "anomie" Jorsch.
1731 NM/23 Bugzilla 687: Fix misparses in eximstats.
1732 Patch provided by Heiko Schlittermann.
1734 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1735 Patch provided by Heiko Schlittermann.
1737 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1738 plus update to original patch.
1740 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1742 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1743 Patch provided by David Brownlee.
1745 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1747 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1749 NM/30 Bugzilla 888: TLS documentation bugfixes.
1751 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1753 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1754 Unlike the original bugzilla I have changed all shell scripts in src tree.
1756 NM/33 Bugzilla 898: Transport filter timeout fix.
1757 Patch by Todd Rinaldo.
1759 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1760 Patch by Serge Demonchaux.
1762 NM/35 Bugzilla 39: Base64 decode bug fixes.
1763 Patch by Jakob Hirsch.
1765 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1767 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1769 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1771 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1777 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1778 ALT-N's libdkim that I have put here:
1779 http://duncanthrax.net/exim-experimental/
1781 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1782 were called 'true' and it seems that C99 defines that as a reserved
1783 keyword to be used with 'bool' variable types. That means you could
1784 not include C99-style headers which use bools without triggering
1785 build errors in sieve.c.
1787 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1788 as mailq or other aliases. Changed the --help handling significantly
1789 to do whats expected. exim_usage() emits usage/help information.
1791 SC/01 Added the -bylocaldomain option to eximstats.
1793 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1795 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1797 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1803 PH/01 Another patch from the Sieve maintainer.
1805 PH/02 When an IPv6 address is converted to a string for single-key lookup
1806 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1807 dots are used instead of colons so that keys in lsearch files need not
1808 contain colons. This was done some time before quoting was made available
1809 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1810 (notated using the quote facility) so as to distinguish them from IPv4
1811 keys. This meant that lookups for IP addresses in host lists did not work
1812 for iplsearch lookups.
1814 This has been fixed by arranging for IPv6 addresses to be expressed with
1815 colons if the lookup type is iplsearch. This is not incompatible, because
1816 previously such lookups could never work.
1818 The situation is now rather anomolous, since one *can* have colons in
1819 ordinary lsearch keys. However, making the change in all cases is
1820 incompatible and would probably break a number of configurations.
1822 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1825 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1826 conversion specification without a maximum field width, thereby enabling
1827 a rogue spamd server to cause a buffer overflow. While nobody in their
1828 right mind would setup Exim to query an untrusted spamd server, an
1829 attacker that gains access to a server running spamd could potentially
1830 exploit this vulnerability to run arbitrary code as the Exim user.
1832 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1833 $primary_hostname instead of what libspf2 thinks the hosts name is.
1835 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1836 a directory entry by the name of the lookup key. Previously, if a
1837 symlink pointed to a non-existing file or a file in a directory that
1838 Exim lacked permissions to read, a lookup for a key matching that
1839 symlink would fail. Now it is enough that a matching directory entry
1840 exists, symlink or not. (Bugzilla 503.)
1842 PH/03 The body_linecount and body_zerocount variables are now exported in the
1845 PH/04 Added the $dnslist_matched variable.
1847 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1848 This means they are set thereafter only if the connection becomes
1851 PH/06 Added the client_condition to authenticators so that some can be skipped
1852 by clients under certain conditions.
1854 PH/07 The error message for a badly-placed control=no_multiline_responses left
1855 "_responses" off the end of the name.
1857 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1859 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1860 (without spaces) instead of just copying the configuration text.
1862 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1864 PH/11 Added $max_received_linelength.
1866 PH/12 Added +ignore_defer and +include_defer to host lists.
1868 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1869 way in which PCRE > 7.0 is built.
1871 PH/14 Implemented queue_only_load_latch.
1873 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1874 MAIL command. The effect was to mangle the value on 64-bit systems.
1876 PH/16 Another patch from the Sieve maintainer.
1878 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1880 PH/18 If a system quota error occurred while trying to create the file for
1881 a maildir delivery, the message "Mailbox is full" was not appended to the
1882 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1883 only to a quota excession during the actual writing of the file.
1885 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1886 characters?) which causes problems in log lines. The DN values are now
1887 passed through string_printing() before being added to log lines.
1889 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1890 and InterBase are left for another time.)
1892 PH/21 Added message_body_newlines option.
1894 PH/22 Guard against possible overflow in moan_check_errorcopy().
1896 PH/23 POSIX allows open() to be a macro; guard against that.
1898 PH/24 If the recipient of an error message contained an @ in the local part
1899 (suitably quoted, of course), incorrect values were put in $domain and
1900 $local_part during the evaluation of errors_copy.
1906 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1907 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1910 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1911 issue a MAIL command.
1913 PH/02 In an ACL statement such as
1915 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1917 if a client was not listed at all, or was listed with a value other than
1918 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1919 the condition was not true (as it should be), so access was not denied.
1920 The bug was that the ! inversion was incorrectly passed on to the second
1921 item. This has been fixed.
1923 PH/03 Added additional dnslists conditions == and =& which are different from
1924 = and & when the dns lookup returns more than one IP address.
1926 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1927 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1929 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1930 FSYNC, which compiles an option called disable_fsync that allows for
1931 bypassing fsync(). The documentation is heavily laced with warnings.
1933 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1935 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1936 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1937 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1938 including adding "make clean"; (3) Added -fPIC when compiling the test
1939 dynamically loaded module, to get rid of a warning.
1941 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1942 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1943 = 0s. The bug is otherwise harmless.
1945 PH/07 There was a bug in the dovecot authenticator such that the value of
1946 $auth1 could be overwritten, and so not correctly preserved, after a
1947 successful authentication. This usually meant that the value preserved by
1948 the server_setid option was incorrect.
1950 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1952 PH/09 Installed PCRE release 7.0.
1954 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1955 run for batched SMTP input. It is now run at the start of every message
1956 in the batch. While fixing this I discovered that the process information
1957 (output by running exiwhat) was not always getting set for -bs and -bS
1958 input. This is fixed, and it now also says "batched" for BSMTP.
1960 PH/11 Added control=no_pipelining.
1962 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1963 patch, slightly modified), and move the expansion of helo_data till after
1964 the connection is made in the smtp transport (so it can use these
1967 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1969 PH/14 Added log_selector = +pid.
1971 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1973 PH/16 Add ${if forany and ${if forall.
1975 PH/17 Added dsn_from option to vary the From: line in DSNs.
1977 PH/18 Flush SMTP output before performing a callout, unless control =
1978 no_callout_flush is set.
1980 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1981 was true (the default) a successful delivery failed to delete the retry
1982 item, thus causing premature timeout of the address. The bug is now
1985 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1987 PH/21 Long custom messages for fakedefer and fakereject are now split up
1988 into multiline reponses in the same way that messages for "deny" and
1989 other ACL rejections are.
1991 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1992 with slight modification.
1994 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1995 draft, changing the syntax and factoring some duplicate code".
1997 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1998 for deliveries of the second and subsequent messages over the same SMTP
2001 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2002 ${reduce, with only minor "tidies".
2004 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2006 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2007 expansion side effects.
2009 PH/27 When a message times out after an over-quota error from an Exim-imposed
2010 quota, the bounce message says "mailbox is full". This message was not
2011 being given when it was a system quota that was exceeded. It now should
2014 MH/03 Made $recipients available in local_scan(). local_scan() already has
2015 better access to the recipient list through recipients_list[], but
2016 $recipients can be useful in postmaster-provided expansion strings.
2018 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2019 in the case of a MAIL command with additional options following the
2020 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2021 were accidentally chopped off.
2023 PH/29 SMTP synchronization checks are implemented when a command is read -
2024 there is a check that no more input is waiting when there shouldn't be
2025 any. However, for some commands, a delay in an ACL can mean that it is
2026 some time before the response is written. In this time, more input might
2027 arrive, invalidly. So now there are extra checks after an ACL has run for
2028 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2029 pipelining has not been advertised.
2031 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2033 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2034 returned in the SMTP response when smtp_return_error_details was false.
2035 This has been fixed.
2037 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2038 instead of the C I/O that was originally supplied, because problems were
2039 reported on Solaris.
2041 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2042 Exim which did not show up earlier: it was assuming that a call to
2043 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2044 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2045 was a macro that became an assignment, so it seemed to work. This has
2046 changed to a proper function call with a void return, hence the compile
2047 error. Exim's code has been fixed.
2049 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2052 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2054 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2056 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2057 because it turns out to be too big or there is a timeout) there is no
2058 "Completed" line in the log. When some messages of this type were
2059 selected by exigrep, they were listed as "not completed". Others were
2060 picked up by some special patterns. I have improved the selection
2061 criteria to be more general.
2063 PH/38 The host_find_failed option in the manualroute router can now be set
2064 to "ignore", to completely ignore a host whose IP address cannot be
2065 found. If all hosts are ignored, the behaviour is controlled by the new
2066 host_all_ignored option.
2068 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2069 homing or because of multiple MX records with /mx) generated more than
2070 one IP address, and the following item turned out to be the local host,
2071 all the secondary addresses of the first item were incorrectly removed
2072 from the list, along with the local host and any following hosts (which
2073 is what is supposed to happen).
2075 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2076 whoever called Exim into the -H file. In the case of the daemon it was
2077 behaving confusingly. When first started, it used values for whoever
2078 started the daemon, but after a SIGHUP it used the Exim user (because it
2079 calls itself on a restart). I have changed the code so that it now always
2082 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2083 message are rejected with the same error (e.g. no authentication or bad
2084 sender address), and a DATA command is nevertheless sent (as can happen
2085 with PIPELINING or a stupid MUA), the error message that was given to the
2086 RCPT commands is included in the rejection of the DATA command. This is
2087 intended to be helpful for MUAs that show only the final error to their
2090 PH/42 Another patch from the Sieve maintainer.
2092 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2093 Eximstats - Fixed some broken HTML links and added missing column headers
2095 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2096 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2098 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2100 PH/43 Yet another patch from the Sieve maintainer.
2102 PH/44 I found a way to check for a TCP/IP connection going away before sending
2103 the response to the final '.' that terminates a message, but only in the
2104 case where the client has not sent further data following the '.'
2105 (unfortunately, this is allowed). However, in many cases there won't be
2106 any further data because there won't be any more messages to send. A call
2107 to select() can be used: if it shows that the input is "ready", there is
2108 either input waiting, or the socket has been closed. An attempt to read
2109 the next input character can distinguish the two cases. Previously, Exim
2110 would have sent an OK response which the client would never have see.
2111 This could lead to message repetition. This fix should cure that, at
2112 least in a lot of common cases.
2114 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2115 advertised in response to EHLO.
2121 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2122 fixed by 4.65/MH/01 (is this a record?) are fixed:
2124 (i) An empty string was always treated as zero by the numeric comparison
2125 operators. This behaviour has been restored.
2127 (ii) It is documented that the numeric comparison operators always treat
2128 their arguments as decimal numbers. This was broken in that numbers
2129 starting with 0 were being interpreted as octal.
2131 While fixing these problems I realized that there was another issue that
2132 hadn't been noticed. Values of message_size_limit (both the global option
2133 and the transport option) were treated as octal if they started with 0.
2134 The documentation was vague. These values are now always treated as
2135 decimal, and I will make that clear in the documentation.
2141 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2142 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2145 MH/01 Don't check that the operands of numeric comparison operators are
2146 integers when their expansion is in "skipping" mode (fixes bug
2147 introduced by 4.64-PH/07).
2149 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2150 child addresses, Exim now panics and dies. Previously, because the count
2151 is held in a short int, deliveries were likely to be lost. As such a
2152 large number of recipients for a single message is ridiculous
2153 (performance will be very, very poor), I have chosen to impose a limit
2154 rather than extend the field.
2160 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2161 leftover -K file (the existence of which was triggered by #402).
2162 While we were at it, introduced process PID as part of the -K
2163 filename. This should rule out race conditions when creating
2166 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2167 processing considerably. Previous code took too long for large mails,
2168 triggering a timeout which in turn triggers #401.
2170 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2171 in the DK code in transports.c. sendfile() is not really portable,
2172 hence the _LINUX specificness.
2174 TF/01 In the add_headers option to the mail command in an Exim filter,
2175 there was a bug that Exim would claim a syntax error in any
2176 header after the first one which had an odd number of characters
2179 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2180 callout verification, Exim cached a "reject" for the entire domain. This
2181 is correct for most verifications, but it is not correct for a recipient
2182 verification with use_sender or use_postmaster set, because in that case
2183 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2184 case of MAIL FROM:<> rejection from other early rejections (e.g.
2185 rejection of HELO). When verifying a recipient using a non-null MAIL
2186 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2187 Whatever the result of the callout, the value of the domain cache is
2188 left unchanged (for any other kind of callout, getting as far as trying
2189 RCPT means that the domain itself is ok).
2191 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2194 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2195 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2196 session, the daemon ignores ECONNECT errors and logs others; it now
2197 ignores EPIPE as well.
2199 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2200 (quoted-printable decoding).
2202 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2203 later the small subsequent patch to fix an introduced bug.
2205 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2207 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2209 PH/08 An error is now given if message_size_limit is specified negative.
2211 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2212 to be given (somewhat) arbitrary names.
2214 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2217 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2218 miscellaneous code fixes
2220 PH/10 Added the log_reject_target ACL modifier to specify where to log
2223 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2224 hostname. This is wrong, because it relates to the incoming message (and
2225 probably the interface on which it is arriving) and not to the outgoing
2226 callout (which could be using a different interface). This has been
2227 changed to use the value of the helo_data option from the smtp transport
2228 instead - this is what is used when a message is actually being sent. If
2229 there is no remote transport (possible with a router that sets up host
2230 addresses), $smtp_active_hostname is used.
2232 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2233 tweaks were necessary in order to get it to work (see also 21 below):
2234 (a) The code assumed that strncpy() returns a negative number on buffer
2235 overflow, which isn't the case. Replaced with Exim's string_format()
2237 (b) There were several signed/unsigned issues. I just did the minimum
2238 hacking in of casts. There is scope for a larger refactoring.
2239 (c) The code used strcasecmp() which is not a standard C function.
2240 Replaced with Exim's strcmpic() function.
2241 (d) The code set only $1; it now sets $auth1 as well.
2242 (e) A simple test gave the error "authentication client didn't specify
2243 service in request". It would seem that Dovecot has changed its
2244 interface. Fortunately there's a specification; I followed it and
2245 changed what the client sends and it appears to be working now.
2247 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2250 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2251 address A is aliased to B and C, where B exists and C does not. Without
2252 -v the output is "A verified" because verification stops after a
2253 successful redirection if more than one address is generated. However,
2254 with -v the child addresses are also verified. Exim was outputting "A
2255 failed to verify" and then showing the successful verification for C,
2256 with its parentage. It now outputs "B failed to verify", showing B's
2257 parentage before showing the successful verification of C.
2259 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2260 look up a TXT record in a specific list after matching in a combined
2263 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2264 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2265 they consult the DNS. I had assumed they would set it the way they
2266 wanted; and indeed my experiments on Linux seem to show that in some
2267 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2268 To be on the safe side, however, I have now made the interface to
2269 host_find_byname() similar to host_find_bydns(), with an argument
2270 containing the DNS resolver options. The host_find_byname() function now
2271 sets these options at its start, just as host_find_bydns() does. The smtp
2272 transport options dns_qualify_single and dns_search_parents are passed to
2273 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2274 of host_find_byname() use the default settings of RES_DEFNAMES
2275 (qualify_single) but not RES_DNSRCH (search_parents).
2277 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2278 spool_read_header() do less string testing, by means of a preliminary
2279 switch on the second character of optional "-foo" lines. (This is
2280 overdue, caused by the large number of possibilities that now exist.
2281 Originally there were few.) While I was there, I also converted the
2282 str(n)cmp tests so they don't re-test the leading "-" and the first
2283 character, in the hope this might squeeze out yet more improvement.
2285 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2286 flag allowing group syntax was set by the header_syntax check but not
2287 turned off, possible causing trouble later; (2) The flag was not being
2288 set at all for the header_verify test, causing "group"-style headers to
2289 be rejected. I have now set it in this case, and also caused header_
2290 verify to ignore an empty address taken from a group. While doing this, I
2291 came across some other cases where the code for allowing group syntax
2292 while scanning a header line wasn't quite right (mostly, not resetting
2293 the flag correctly in the right place). These bugs could have caused
2294 trouble for malformed header lines. I hope it is now all correct.
2296 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2297 with the "reply" argument non-NULL. The code, however (which originally
2298 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2299 but it didn't always do it. This confused somebody who was copying the
2300 code for some other use. I have removed all the tests.
2302 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2303 feature that was used to support insecure browsers during the U.S. crypto
2304 embargo. It requires special client support, and Exim is probably the
2305 only MTA that supported it -- and would never use it because real RSA is
2306 always available. This code has been removed, because it had the bad
2307 effect of slowing Exim down by computing (never used) parameters for the
2308 RSA_EXPORT functionality.
2310 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2311 authenticator to fail if there's a tab character in the incoming data
2312 (there should never be unless someone is messing about, as it's supposed
2313 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2314 if the connection is using TLS or if the remote IP is the same as the
2315 local IP, and the "valid-client-cert option" if a client certificate has
2318 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2319 authenticators. This can be used for authorization after authentication
2320 succeeds. (In the case of plaintext, it servers for both authentication
2323 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2324 if any retry times were supplied.
2326 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2327 connection, where there is no client IP address to check. In this
2328 situation, the verify now always succeeds.
2330 PH/25 Applied John Jetmore's -Mset patch.
2332 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2334 PH/27 In a string expansion for a processed (not raw) header when multiple
2335 headers of the same name were present, leading whitespace was being
2336 removed from all of them, but trailing whitespace was being removed only
2337 from the last one. Now trailing whitespace is removed from each header
2338 before concatenation. Completely empty headers in a concatenation (as
2339 before) are ignored.
2341 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2342 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2344 PH/29 [Removed. This was a change that I later backed out, and forgot to
2345 correct the ChangeLog entry (that I had efficiently created) before
2346 committing the later change.]
2348 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2349 address errors (4xx response to RCPT) over the same connection as other
2350 messages routed to the same hosts. Such deliveries are always "forced",
2351 so retry times are not inspected. This resulted in far too many retries
2352 for the affected addresses. The effect occurred only when there were more
2353 hosts than the hosts_max_try setting in the smtp transport when it had
2354 the 4xx errors. Those hosts that it had tried were not added to the list
2355 of hosts for which the message was waiting, so if all were tried, there
2356 was no problem. Two fixes have been applied:
2358 (i) If there are any address or message errors in an SMTP delivery, none
2359 of the hosts (tried or untried) are now added to the list of hosts
2360 for which the message is waiting, so the message should not be a
2361 candidate for sending over the same connection that was used for a
2362 successful delivery of some other message. This seems entirely
2363 reasonable: after all the message is NOT "waiting for some host".
2364 This is so "obvious" that I'm not sure why it wasn't done
2365 previously. Hope I haven't missed anything, but it can't do any
2366 harm, as the worst effect is to miss an optimization.
2368 (ii) If, despite (i), such a delivery is accidentally attempted, the
2369 routing retry time is respected, so at least it doesn't keep
2370 hammering the server.
2372 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2373 in ${readsocket because some servers need this prod.
2375 PH/32 Added some extra debug output when updating a wait-xxx database.
2377 PH/33 The hint "could be header name not terminated by colon", which has been
2378 given for certain expansion errors for a long time, was not being given
2379 for the ${if def:h_colon_omitted{... case.
2381 PH/34 The spec says: "With one important exception, whenever a domain list is
2382 being scanned, $domain contains the subject domain." There was at least
2383 one case where this was not true.
2385 PH/35 The error "getsockname() failed: connection reset by peer" was being
2386 written to the panic log as well as the main log, but it isn't really
2387 panic-worthy as it just means the connection died rather early on. I have
2388 removed the panic log writing for the ECONNRESET error when getsockname()
2391 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2392 runs only) independently of the message's sender address. This meant
2393 that, if the 4xx error was in fact related to the sender, a different
2394 message to the same recipient with a different sender could confuse
2395 things. In particualar, this can happen when sending to a greylisting
2396 server, but other circumstances could also provoke similar problems.
2397 I have changed the default so that the retry time for these errors is now
2398 based a combination of the sender and recipient addresses. This change
2399 can be overridden by setting address_retry_include_sender=false in the
2402 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2403 remote server are returned as part of bounce messages. This was not
2404 happening for LMTP over a pipe (the lmtp transport), but now it is the
2405 same for both kinds of LMTP.
2407 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2408 in header lines that were in fact CNAMEs. This is no longer the case.
2410 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2411 and queue runs started by the daemon processed all messages. This has
2412 been fixed so that -R and -S can now usefully be given with -q<time>.
2414 PH/40 Import PCRE release 6.7 (fixes some bugs).
2416 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2418 PH/42 Give an error if -q is specified more than once.
2420 PH/43 Renamed the variables $interface_address and $interface_port as
2421 $received_ip_address and $received_port, to make it clear that these
2422 values apply to message reception, and not to the outgoing interface when
2423 a message is delivered. (The old names remain recognized, of course.)
2425 PH/44 There was no timeout on the connect() call when using a Unix domain
2426 socket in the ${readsocket expansion. There now is.
2428 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2429 be meaningful with "accept".
2431 SC/01 Eximstats V1.43
2432 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2434 SC/02 Eximstats V1.44
2435 Use a glob alias rather than an array ref in the generated
2436 parser. This improves both readability and performance.
2438 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2439 Collect SpamAssassin and rejection statistics.
2440 Don't display local sender or destination tables unless
2441 there is data to show.
2442 Added average volumes into the top table text output.
2444 SC/04 Eximstats V1.46
2445 Collect data on the number of addresses (recipients)
2446 as well as the number of messages.
2448 SC/05 Eximstats V1.47
2449 Added 'Message too big' to the list of mail rejection
2450 reasons (thanks to Marco Gaiarin).
2452 SC/06 Eximstats V1.48
2453 Mainlog lines which have GMT offsets and are too short to
2454 have a flag are now skipped.
2456 SC/07 Eximstats V1.49 (Alain Williams)
2457 Added the -emptyok flag.
2459 SC/08 Eximstats V1.50
2460 Fixes for obtaining the IP address from reject messages.
2462 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2463 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2464 whitesspace changes from 4.64-PH/27
2466 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2469 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2470 are found, allow negative numbers in numeric criteria)
2472 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2474 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2475 to match changes made in 4.64-PH/43
2477 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2479 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2480 "domain" is actually an IP address, and force a failure. This locks out
2481 those revolvers/nameservers that support "A-for-A" lookups, in
2482 contravention of the specifications.
2484 PH/48 When a host name was looked up from an IP address, and the subsequent
2485 forward lookup of the name timed out, the host name was left in
2486 $sender_host_name, contrary to the specification.
2488 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2489 restricted to single-key lookups, Exim was not diagnosing an error if
2490 * or *@ was used with a query-style lookup.
2492 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2494 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2495 long ago, but noone interested enough thought of it. Let's just say that
2496 the "1.1" means that there are some new functions that weren't there at
2497 some point in the past.
2499 PH/51 Error processing for expansion failure of helo_data from an smtp
2500 transport during callout processing was broken.
2502 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2503 tested/used via the -bh/-bhc/-bs options.
2505 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2506 bug, fixed in subsequent PCRE releases).
2508 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2509 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2511 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2517 SC/01 Use a glob alias rather than an array ref in eximstats generated
2518 parser. This improves both readability and performance.
2520 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2521 Don't display local sender or destination tables in eximstats unless
2522 there is data to show.
2523 Added average volumes into the eximstats top table text output.
2525 SC/03 Collect data on the number of addresses (recipients) as well
2526 as the number of messages in eximstats.
2528 TF/01 Correct an error in the documentation for the redirect router. Exim
2529 does (usually) call initgroups() when daemonizing.
2531 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2532 with consistent privilege compared to when running as a daemon.
2534 TF/03 Note in the spec that $authenticated_id is not set for local
2535 submissions from trusted users.
2537 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2538 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2540 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2541 by adding some example configuration directives to the default
2542 configuration file. A little bit of work is required to uncomment the
2543 directives and define how usernames and passwords are checked, but
2544 there is now a framework to start from.
2546 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2547 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2548 without this. I don't know how relevant this is to other LDAP libraries.
2550 PH/02 Add the verb name to the "unknown ACL verb" error.
2552 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2554 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2556 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2557 directory not expanded when it should be if an expanded home directory
2558 was set for the address (which is overridden by the transport).
2560 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2563 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2564 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2565 because it is too late at that time, and has no effect.
2567 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2568 security issue with \' (bugzilla #107). I could not use the
2569 PQescapeStringConn() function, because it needs a PGconn value as one of
2572 PH/08 When testing addresses using -bt, indicate those final addresses that
2573 are duplicates that would not cause an additional delivery. At least one
2574 person was confused, thinking that -bt output corresponded to deliveries.
2575 (Suppressing duplicates isn't a good idea as you lose the information
2576 about possibly different redirections that led to the duplicates.)
2578 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2579 systems where poll() doesn't work, in particular OS X.
2581 PH/10 Added more information to debugging output for retry time not reached.
2583 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2584 operations in malware.c.
2586 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2589 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2590 syslog_duplication set false, log lines that would normally be written
2591 both the the main log and to the reject log were not written to syslog at
2594 PH/14 In the default configuration, change the use of "message" in ACL warn
2595 statements to "add_header".
2597 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2598 not followed by a command (e.g. "seen endif").
2600 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2601 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2604 PH/17 Added extra conditions to the default value of delay_warning_condition
2608 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2609 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2610 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2613 The Auto-Submitted: and various List- headers are standardised, whereas I
2614 don't think Precedence: ever was.
2616 PH/18 Refactored debugging code in route_finduser() to show more information,
2617 in particular, the error code if getpwnam() issues one.
2619 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2620 This is apparently needed in addition to the PH/07 change above to avoid
2621 any possible encoding problems.
2623 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2624 but not after initializing Perl.
2626 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2627 output them only if debugging. By default they are written stderr,
2628 apparently, which is not desirable.
2630 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2633 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2636 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2638 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2639 authenticated or an ident call has been made. Suppress the default
2640 values for $authenticated_id and $authenticated_sender (but permit -oMai
2641 and -oMas) when testing with -bh.
2643 PH/24 Re-jigged the order of the tests in the default configuration so that the
2644 tests for valid domains and recipients precede the DNS black list and CSA
2645 tests, on the grounds that those ones are more expensive.
2647 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2648 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2649 This bug exists in every version of Exim that I still have, right back to
2652 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2653 However, an attempt to turn on case-sensitivity in a regex key by
2654 including (?-i) didn't work because the subject string was already
2655 lowercased, and the effects were non-intuitive. It turns out that a
2656 one-line patch can be used to allow (?-i) to work as expected.
2662 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2663 other effects) broke the use of negated acl sub-conditions.
2665 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2668 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2669 "Deny" causes Exim to reject the incoming connection with a 554 error.
2670 Unfortunately, if there is a major crisis, such as a disk failure,
2671 tcp-wrappers gives "deny", whereas what one would like would be some
2672 kind of temporary error. A kludge has been added to help with this.
2673 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2674 554 error is used if errno is still zero or contains ENOENT (which occurs
2675 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2678 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2680 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2681 errors. Otherwise a message that provokes a temporary error (when other
2682 messages do not) can cause a whole host to time out.
2684 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2685 addresses were routed directly to files or pipes from a redirect router.
2686 File deliveries just didn't batch; pipe deliveries might have suffered
2689 PH/06 A failure to get a lock for a hints database would erroneously always say
2690 "Failed to get write lock", even when it was really a read lock.
2692 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2693 of 0600. This has been changed to use the value of the lockfile_mode
2694 option (which defaults to 0600).
2696 PH/08 Applied small patch from the Sieve maintainer.
2698 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2699 folder from quota calculations, a direct delivery into this folder messed
2700 up the contents of the maildirsize file. This was because the regex was
2701 used only to exclude .Trash (or whatever) when the size of the mailbox
2702 was calculated. There was no check that a delivery was happening into an
2703 excluded directory. This bug has been fixed by ignoring all quota
2704 processing for deliveries into excluded directories.
2706 PH/10 Added the maildirfolder_create_regex option to appendfile.
2712 PH/01 The code for finding all the local interface addresses on a FreeBSD
2713 system running IPv6 was broken. This may well have applied to all BSD
2714 systems, as well as to others that have similar system calls. The broken
2715 code found IPv4 interfaces correctly, but gave incorrect values for the
2716 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2717 that it would not match correctly against @[] and not recognize the IPv6
2720 PH/02 The ipliteral router was not recognizing addresses of the form user@
2721 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2723 PH/03 Added disable_ipv6.
2725 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2726 decoded form, because it is most often used to construct To: headers
2727 lines in autoreplies, and the decoded form may well be syntactically
2728 invalid. However, $reply_address has leading white space removed, and all
2729 newlines turned into spaces so that the autoreply transport does not
2732 PH/05 If group was specified without a user on a router, and no group or user
2733 was specified on a transport, the group from the router was ignored.
2735 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2736 for visible compile-time settings that can be used to change these
2737 numbers, for those that want even more. Backwards compatibility with old
2738 spool files has been maintained. However, going back to a previous Exim
2739 release will lost any variables that are in spool files.
2741 PH/07 Two small changes when running in the test harness: increase delay when
2742 passing a TCP/IP connection to a new process, in case the original
2743 process has to generate a bounce, and remove special handling of
2744 127.0.0.2 (sic), which is no longer necessary.
2746 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2747 be the same on different OS.
2749 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2752 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2753 whether --show-vars was specified or not
2755 JJ/02 exipick: Added support for new ACL variable spool format introduced
2758 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2759 syntactically invalid From: or Reply-to: line, and a filter used this to
2760 generate an autoreply, and therefore failed to obtain an address for the
2761 autoreply, Exim could try to deliver to a non-existent relative file
2762 name, causing unrelated and misleading errors. What now happens is that
2763 it logs this as a hard delivery error, but does not attempt to create a
2766 PH/11 The exinext utility has a -C option for testing purposes, but although
2767 the given file was scanned by exinext itself; it wasn't being passed on
2768 when Exim was called.
2770 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2771 an end-of-file indication when reading a command response.
2773 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2774 compiled. In many other places in Exim, IPv6 addresses are always
2775 recognized, so I have changed this. It also means that IPv4 domain
2776 literals of the form [IPV4:n.n.n.n] are now always recognized.
2778 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2779 used if the router is not running as root, for example, when verifying at
2780 ACL time, or when using -bh. The debugging output from this situation was
2781 non-existent - all you got was a failure to exec. I have made two
2784 (a) Failures to set uid/gid, the current directory, or a process leader
2785 in a subprocess such as that created by queryprogram now generate
2786 suitable debugging ouput when -d is set.
2788 (b) The queryprogram router detects when it is not running as root,
2789 outputs suitable debugging information if -d is set, and then runs
2790 the subprocess without attempting to change uid/gid.
2792 PH/15 Minor change to Makefile for building test_host (undocumented testing
2795 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2796 additional section of a DNS packet that returns MX or SRV records.
2797 Instead, it always explicitly searches for A/AAAA records. This avoids
2798 major problems that occur when a DNS server includes only records of one
2799 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2800 fixed another bug: if SRV records were looked up and the corresponding
2801 address records were *not* found in the additional section, the port
2802 values from the SRV records were lost.
2804 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2805 using the correct key (the original address) when searching the retry
2806 rules in order to find which one to use for generating the retry hint.
2808 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2809 adding the default one. Similarly, if it contains a Reply-To: header, the
2810 errors_reply_to option, if set, is not used.
2812 PH/19 When calculating a retry time, Exim used to measure the "time since
2813 failure" by looking at the "first failed" field in the retry record. Now
2814 it does not use this if it is later than than the arrival time of the
2815 message. Instead it uses the arrival time. This makes for better
2816 behaviour in cases where some deliveries succeed, thus re-setting the
2817 "first failed" field. An example is a quota failure for a huge message
2818 when small messages continue to be delivered. Without this change, the
2819 "time since failure" will always be short, possible causing more frequent
2820 delivery attempts for the huge message than are intended.
2821 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2823 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2824 $1, $2, $3) because the numerical variables can be reset during some
2825 expansion items (e.g. "match"), thereby losing the authentication data.
2827 PH/21 Make -bV show the size of off_t variables so that the test suite can
2828 decide whether to run tests for quotas > 2G.
2830 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2831 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2832 filecount value is greater than 2G or if a quota value is greater than 2G
2833 on a system where the size of off_t is not greater than 4, a panic error
2836 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2837 never match. The debug and -bh output now contains an explicit error
2838 message indicating a malformed IPv4 address or mask.
2840 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2841 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2842 PH/23 above applies.
2844 PH/25 Do not write to syslog when running in the test harness. The only
2845 occasion when this arises is a failure to open the main or panic logs
2846 (for which there is an explicit test).
2848 PH/26 Added the /no_tell option to "control=freeze".
2850 PH/27 If a host name lookup failed very early in a connection, for example, if
2851 the IP address matched host_lookup and the reverse lookup yielded a name
2852 that did not have a forward lookup, an error message of the form "no IP
2853 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2854 could be logged. Now it outputs the IP address instead of "NULL".
2856 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2857 allows the sender and the authenticated sender to be set when
2858 submitting a message from within Exim. Since child_open_exim() is
2859 documented for local_scan(), the new function should be too.
2861 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2862 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2863 results in an empty string is now treated as unset.
2865 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2867 PH/31 Added sender_verify_fail logging option.
2869 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2870 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2871 tidied the source and removed it altogether.
2873 PH/33 When a queue run was abandoned because the load average was too high, a
2874 log line was always written; now it is written only if the queue_run log
2875 selector is set. In addition, the log line for abandonment now contains
2876 information about the queue run such as the pid. This is always present
2877 in "start" and "stop" lines but was omitted from the "abandon" line.
2879 PH/34 Omit spaces between a header name and the colon in the error message that
2880 is given when verify = headers_syntax fails (if there are lots of them,
2881 the message gets confusing).
2883 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2884 names, as there are now some PTR records that contain slashes. This check
2885 is only to protect against broken name servers that fall over on strange
2886 characters, so the fact that it applies to all lookups doesn't matter.
2888 PH/36 Now that the new test suite is complete, we can remove some of the
2889 special code in Exim that was needed for the old test suite. For example,
2890 sorting DNS records because real resolvers return them in an arbitrary
2891 order. The new test suite's fake resolver always returns records in the
2894 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2895 bounces) except when queue_only is set, to avoid logging races between
2896 the different processes.
2898 PH/38 Panic-die if .include specifies a non-absolute path.
2900 PH/39 A tweak to the "H" retry rule from its user.
2902 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2903 a label. They prevented compilation on older perls.
2905 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2906 a warning to be raised on newish perls.
2908 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2909 on queue. Changes to match documented behaviour of showing count of
2910 messages matching specified criteria.
2912 PH/40 Changed the default ident timeout from 30s to 5s.
2914 PH/41 Added support for the use of login_cap features, on those BSD systems
2915 that have them, for controlling the resources used by pipe deliveries.
2917 PH/42 The content-scanning code uses fopen() to create files in which to put
2918 message data. Previously it was not paying any attention to the mode of
2919 the files. Exim runs with umask(0) because the rest of the code creates
2920 files with open(), and sets the required mode explicitly. Thus, these
2921 files were ending up world-writeable. This was not a big issue, because,
2922 being within the spool directory, they were not world-accessible. I have
2923 created a function called modefopen, which takes an additional mode
2924 argument. It sets umask(777), creates the file, chmods it to the required
2925 mode, then resets the umask. All the relevant calls to fopen() in the
2926 content scanning code have been changed to use this function.
2928 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2929 to 24 hours. This avoids potential overflow problems when processing G
2930 and H retry rules. I suspect nobody ever tinkers with this value.
2932 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2934 PH/45 When the plaintext authenticator is running as a client, the server's
2935 challenges are checked to ensure they are valid base64 strings. By
2936 default, the authentication attempt is cancelled if an invalid string is
2937 received. Setting client_ignore_invalid_base64 true ignores these errors.
2938 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2939 they are received. Thus, the responses can be made to depend on the
2940 challenges. If an invalid string is ignored, an empty string is placed in
2943 PH/46 Messages that are created by the autoreply transport now contains a
2944 References: header, in accordance with RFCs 2822 and 3834.
2946 PH/47 Added authenticated_sender_force to the smtp transport.
2948 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2950 PH/49 Installed latest patch from the Sieve maintainer.
2952 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2953 was also set, the appendfile transport was unnecessarily scanning a
2954 directory of message files (e.g. for maildir delivery) to find the count
2955 of files (along with the size), even though it did not need this
2956 information. It now does the scan only if it needs to find either the
2957 size of the count of files.
2959 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2961 PH/52 Two bugs concerned with error handling when the smtp transport is
2964 (i) Exim was not creating retry information for temporary errors given
2965 for individual recipients after the DATA command when the smtp transport
2966 was used in LMTP mode. This meant that they could be retried too
2967 frequently, and not timed out correctly.
2969 (ii) Exim was setting the flag that allows error details to be returned
2970 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2971 recipients that were returned after the DATA command.
2973 PH/53 This is related to PH/52, but is more general: for any failing address,
2974 when detailed error information was permitted to be returned to the
2975 sender, but the error was temporary, then after the final timeout, only
2976 "retry timeout exceeded" was returned. Now it returns the full error as
2977 well as "retry timeout exceeded".
2979 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2980 do this, and (what is worse) MTAs that accept it.
2982 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2983 will now be deprecated.
2985 PH/56 New os.c-cygwin from the Cygwin maintainer.
2987 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2988 formats (previously only available in exim formats via -bpr, -bpru,
2989 and -bpra. Now also available in native and exiqgrep formats)
2991 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2992 with very large, slow to parse queues
2994 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2996 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2998 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2999 responses to authentication challenges, though it was showing the
3000 challenges; (ii) I've removed the CR characters from the debug output for
3003 PH/58 Allow for the insertion of a newline as well as a space when a string
3004 is turned into more than one encoded-word during RFC 2047 encoding. The
3005 Sieve code now uses this.
3007 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3008 data_4xx, lost_connection, tls_required.
3010 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3011 message was being sent as an SMTP response.
3013 PH/61 Add -l and -k options to exicyclog.
3015 PH/62 When verifying, if an address was redirected to one new address, so that
3016 verification continued, and the new address failed or deferred after
3017 having set something in $address_data, the value of $address_data was not
3018 passed back to the ACL. This was different to the case when no
3019 redirection occurred. The value is now passed back in both cases.
3021 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3022 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3023 use login_cap.h, so on its own it isn't the distinguishing feature. The
3024 new name refers directly to the setclassresources() function.
3026 PH/65 Added configuration files for NetBSD3.
3028 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3030 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3031 is preferred over IPv4.
3033 PH/68 The bounce_return_message and bounce_return_body options were not being
3034 honoured for bounces generated during the reception of non-SMTP messages.
3035 In particular, this applied to messages rejected by the ACL. This bug has
3036 been fixed. However, if bounce_return_message is true and bounce_return_
3037 body is false, the headers that are returned for a non-SMTP message
3038 include only those that have been read before the error was detected.
3039 (In the case of an ACL rejection, they have all been read.)
3041 PH/69 The HTML version of the specification is now built in a directory called
3042 spec_html instead of spec.html, because the latter looks like a path with
3043 a MIME-type, and this confuses some software.
3045 PH/70 Catch two compiler warnings in sieve.c.
3047 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3048 function verify_get_ident() calls ip_connect() to connect a socket, but
3049 if the "connect()" function timed out, ip_connect() used to close the
3050 socket. However, verify_get_ident() also closes the socket later, and in
3051 between Exim writes to the log, which may get opened at this point. When
3052 the socket was closed in ip_connect(), the log could get the same file
3053 descriptor number as the socket. This naturally causes chaos. The fix is
3054 not to close the socket in ip_connect(); the socket should be closed by
3055 the function that creates it. There was only one place in the code where
3056 this was missing, in the iplookup router, which I don't think anybody now
3057 uses, but I've fixed it anyway.
3059 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3060 well as to direct DNS lookups. Otherwise the handling of names in host
3061 lists is inconsistent and therefore confusing.
3067 PH/01 Two changes to the default runtime configuration:
3069 (1) Move the checks for relay_from_hosts and authenticated clients from
3070 after to before the (commented out) DNS black list checks.
3072 (2) Add control=submission to the relay_from_hosts and authenticated
3073 clients checks, on the grounds that messages accepted by these
3074 statements are most likely to be submissions.
3076 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3078 (1) Generate an error if the third argument for the ${prvs expansion is
3081 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3084 (3) Reset the variables that are obtained from the first argument of
3085 ${prvscheck and used in the second argument before leaving the code,
3086 because their memory is reclaimed, so using them afterwards may do
3089 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3090 one (it's much easier than Tom thought :-).
3092 (5) Because of (4), we can now allow for the use of $prvscheck_result
3093 inside the third argument.
3095 PH/03 For some reason, the default setting of PATH when running a command from
3096 a pipe transport was just "/usr/bin". I have changed it to
3099 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3100 anything to be listed in the output from -bV.
3102 PH/05 When a filter generated an autoreply, the entire To: header line was
3103 quoted in the delivery log line, like this:
3105 => >A.N.Other <ano@some.domain> <original@ddress> ...
3107 This has been changed so that it extracts the operative address. There
3108 may be more than one such address. If so, they are comma-separated, like
3111 => >ano@some.domain,ona@other.domain <original@ddress> ...
3113 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3114 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3115 not being looked up in the rDNS to get a host name, Exim was showing the
3116 IP address twice in Received: lines, even though the IP addresses were
3117 identical. For example:
3119 Received: from [1.2.3.4] (helo=[1.2.3.4])
3121 However, if the real host name was known, it was omitting the HELO data
3122 if it matched the actual IP address. This has been tidied up so that it
3123 doesn't show the same IP address twice.
3125 PH/07 When both +timestamp and +memory debugging was on, the value given by
3126 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3127 called by the debug printing, thereby overwriting the timestamp buffer.
3128 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3130 PH/08 When the original message was included in an autoreply transport, it
3131 always said "this is a copy of the message, including all the headers",
3132 even if body_only or headers_only was set. It now gives an appropriate
3135 PH/09 Applied a patch from the Sieve maintainer which:
3137 o fixes some comments
3138 o adds the (disabled) notify extension core
3139 o adds some debug output for the result of if/elsif tests
3140 o points to the current vacation draft in the documentation
3141 and documents the missing references header update
3145 o fixes a bug in processing the envelope test (when testing
3146 multiple envelope elements, the last element determinted the
3149 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3150 Electronic Mail") by including:
3152 Auto-submitted: auto-generated
3154 in the messages that it generates (bounce messages and others, such as
3155 warnings). In the case of bounce messages for non-SMTP mesages, there was
3156 also a typo: it was using "Auto_submitted" (underscore instead of
3157 hyphen). Since every message generated by Exim is necessarily in response
3158 to another message, thes have all been changed to:
3160 Auto-Submitted: auto-replied
3162 in accordance with these statements in the RFC:
3164 The auto-replied keyword:
3166 - SHOULD be used on messages sent in direct response to another
3167 message by an automatic process,
3169 - MUST NOT be used on manually-generated messages,
3171 - MAY be used on Delivery Status Notifications (DSNs) and Message
3172 Disposition Notifications (MDNs),
3174 - MUST NOT be used on messages generated by automatic or periodic
3175 processes, except for messages which are automatic responses to
3178 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3179 to the default Received: header definition.
3181 PH/12 Added log selector acl_warn_skipped (default on).
3183 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3184 variables because (a) they are in the wrong storage pool and (b) even if
3185 they were copied, it wouldn't work properly because of the caching.
3187 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3188 checking when decoding. Apparently there are clients that generate
3189 overlong encoded strings. Why am I not surprised?
3191 PH/15 If the first argument of "${if match_address" was not empty, but did not
3192 contain an "@" character, Exim crashed. Now it writes a panic log message
3193 and treats the condition as false.
3195 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3197 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3198 extension "envelope-auth". The code is finished and in agreement with
3199 other implementations, but there is no documentation so far and in fact,
3200 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3201 not changing the active code.
3203 Print executed "if" and "elsif" statements when debugging is used. This
3204 helps a great deal to understand what a filter does.
3206 Document more things not specified clearly in RFC3028. I had all this
3207 sorted out, when out of a sudden new issues came to my mind. Oops."
3209 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3212 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3213 canonical form (as documented). However, after a host name lookup from
3214 the IP address, check_host() was doing a simple string comparison with
3215 addresses acquired from the DNS when checking that the found name did
3216 have the original IP as one of its addresses. Since any found IPv6
3217 addresses are likely to be in abbreviated form, the comparison could
3218 fail. Luckily, there already exists a function for doing the comparison
3219 by converting both addresses to binary, so now that is used instead of
3220 the text comparison.
3222 PH/20 There was another similar case to PH/19, when a complete host name was
3223 given in a host list; looking up its IP address could give an abbreviated
3224 form, whereas the current host's name might or might not be abbreviated.
3225 The same fix has been applied.
3231 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3232 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3235 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3236 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3238 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3240 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3241 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3242 Also, for "Auto-Submitted", treat anything other than "no" as
3243 non-personal, in accordance with RFC 3834. (Previously it treated
3244 anything starting "auto-" as non-personal.)
3246 TF/01 The control=submission/name=... option had a problem with syntax
3247 errors if the name included a slash character. The /name= option
3248 now slurps the rest of the string, so it can include any characters
3249 but it must come last in the list of options (after /sender_retain
3252 PH/05 Some modifications to the interface to the fake nameserver for the new
3260 TK/01 Added the "success_on_redirect" address verification option. See
3261 NewStuff for rationale and an example.
3263 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3265 PH/02 Patch to exigrep to allow it to work on syslog lines.
3267 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3268 fread() to skip over the body file's header line, because in Cygwin the
3269 header line is locked and is inaccessible.
3271 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3272 co-exist for some time) to make it clear that it is the Exim ID that is
3273 referenced, not the Message-ID: header line.
3275 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3276 string_format() function, because snprintf() does not exist on all
3279 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3280 ${stat: expansion item.
3282 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3283 protocol synchronization error", to keep the pedants happy.
3285 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3286 well as for IRIX systems, when gcc is being used. See the host.c source
3289 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3291 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3294 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3295 transport and to the smtp transport in LMTP mode.
3297 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3299 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3300 run a filter in a subprocess. This could lead to confusion in subsequent
3301 lookups in the parent process. There should also be a search_tidyup() at
3302 the end of the subprocess.
3304 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3305 only if the host matched helo_try_verify_hosts, which caused the
3306 verification to occur when the EHLO/HELO command was issued. The ACL just
3307 tested the remembered result. Now, if a previous verification attempt has
3308 not happened, "verify = helo" does it there and then.
3310 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3312 TK/03 Fix log output including CR from clamd.
3314 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3315 exist provoked a memory error which could cause a segfault.
3317 PH/15 Installed PCRE 6.2
3319 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3321 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3322 of the problem. Specifically, suggested +O2 rather than +O1 for the
3325 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3327 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3328 record, the port was not correctly logged when the outgoing_port log
3329 selector was set (it logged the transort's default port).
3331 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3332 fallback_hosts, and "hosts" in the smtp transport.
3334 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3335 host errors such as "Connection refused".
3337 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3338 authentication with radiusclient 0.4.9:
3340 - Error returned from rc_read_config was caught wrongly
3341 - Username/password not passed on to radius server due to wrong length.
3343 The presumption is that some radiusclient API changes for 4.51/PH/17
3344 were not taken care of correctly. The code is still untested by me (my
3345 Linux distribution still has 0.3.2 of radiusclient), but it was
3346 contributed by a Radius user.
3348 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3349 expanding the "port" option of the smtp transport.
3351 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3352 while reading a MIME header. Thanks to Tom Hughes for a patch.
3354 PH/24 Include config.h inside local_scan.h so that configuration settings are
3357 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3358 that in an ACL for RCPT (for example), you can examine exactly what was
3361 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3362 commands, but it was not correctly comparing the address with the actual
3363 client host address. Thus, it would show the EHLO address in Received:
3364 header lines when this was not necessary.
3366 PH/27 Added the % operator to ${eval:}.
3368 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3369 it should be ignoring failures (because with -C, for example, it has lost
3370 privilege). It wasn't ignoring creation failures other than "already
3373 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3376 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3377 because an input line was too long, either on its own, or by virtue of
3378 too many continuations, the temporary file was not being removed, and the
3379 return code was incorrect.
3381 PH/31 Missing "BOOL" in function definition in filtertest.c.
3383 PH/32 Applied Sieve patches from the maintainer.
3385 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3387 PH/33 Added "verify = not_blind".
3389 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3390 Local/Makefile (with some defaults set). These are used in built scripts
3391 such as exicyclog, but they have never been used in the exim_install
3392 script (though there are many overriding facilities there). I have
3393 arranged that the exim_install script now takes note of these two
3396 PH/35 Installed configuration files for Dragonfly.
3398 PH/36 When a locally submitted message by a trusted user did not contain a
3399 From: header, and the sender address was obtained from -f or from an SMTP
3400 MAIL command, and the trusted user did not use -F to supply a sender
3401 name, $originator_name was incorrectly used when constructing a From:
3402 header. Furthermore, $originator_name was used for submission mode
3403 messages from external hosts without From: headers in a similar way,
3404 which is clearly wrong.
3406 PH/37 Added control=suppress_local_fixups.
3408 PH/38 When log_selector = +received_sender was set, and the addition of the
3409 sender made the log line's construction buffer exactly full, or one byte
3410 less than full, an overflow happened when the terminating "\n" was
3413 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3414 when the result of a list match is failure because a DNS lookup failed.
3416 PH/40 RM_COMMAND is now used in the building process.
3418 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3419 the "build-* directories that it finds.
3421 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3422 address in a domain literal was a prefix of an interface address.
3424 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3425 when verifying a sender address, unless rewrite_headers is false.
3427 PH/44 Wrote a long comment about why errors_to addresses are verified as
3428 recipients, not senders.
3430 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3431 the ratelimit ACL was added.
3433 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3435 PH/46 Added extra information about PostgreSQL errors to the error string.
3437 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3438 avoiding the need to install special zones in a real server. This is
3439 backwards compatible; if it can't find the fake resolver, it drops back.
3440 Thus, both old and new test suites can be run.
3442 TF/02 Added util/ratelimit.pl
3444 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3445 clock is set back in time.
3447 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3448 Candler <B.Candler@pobox.com>.
3450 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3451 OK for addresses that are the result of redirections.
3453 PH/48 A number of further additions for the benefit of the new test suite,
3454 including a fake gethostbyname() that interfaces to the fake DNS resolver
3457 TF/06 The fix for widen_domains has also been applied to qualify_single and
3458 search_parents which are the other dnslookup options that can cause
3461 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3464 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3466 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3467 service" if it encountered an MX record with an empty target hostname.
3468 The message is now "an MX or SRV record indicated no SMTP service".
3470 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3471 if the DNS of the sending site is misconfigured. This is quite a
3472 common situation. This change restores the behaviour of treating a
3473 helo verification defer as a failure.
3475 PH/51 If self=fail was set on a router, the bounce message did not include the
3476 actual error message.
3482 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3484 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3485 command itself ended in error, the underlying message about the transport
3486 filter timeout was being overwritten with the pipe command error. Now the
3487 underlying error message should be appended to the second error message.
3489 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3491 PH/02 Reduce the amount of output that "make" produces by default. Full output
3492 can still be requested.
3494 PH/03 The warning log line about a condition test deferring for a "warn" verb
3495 was being output only once per connection, rather than after each
3496 occurrence (because it was using the same function as for successful
3497 "warn" verbs). This seems wrong, so I have changed it.
3499 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3500 it should not have, which might have caused a crash in the right
3501 circumstances, but probably never did.
3503 PH/04 Installed a modified version of Tony Finch's patch to make submission
3504 mode fix the return path as well as the Sender: header line, and to
3505 add a /name= option so that you can make the user's friendly name appear
3508 TF/03 Added the control = fakedefer ACL modifier.
3510 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3511 Mark Lowes for thorough testing.
3513 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3515 TK/03 Merged latest SRS patch from Miles Wilton.
3517 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3518 with the definition in sysexits.h (which is #included earlier).
3519 Fortunately, Exim does not actually use EX_OK. The code used to try to
3520 preserve the sysexits.h value, by assumimg that macro definitions were
3521 scanned for macro replacements. I have been disabused of this notion,
3522 so now the code just undefines EX_OK before #including unistd.h.
3524 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3525 in the smtp transport. When a block could not be written in a single
3526 write() function, the timeout was being re-applied to each part-write.
3527 This seems wrong - if the receiver was accepting one byte at a time it
3528 would take for ever. The timeout is now adjusted when this happens. It
3529 doesn't have to be particularly precise.
3531 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3532 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3534 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3535 without a domain if the check to <postmaster@domain> fails.
3537 SC/01 Eximstats: added -xls and the ability to specify output files
3538 (patch written by Frank Heydlauf).
3540 SC/02 Eximstats: use FileHandles for outputing results.
3542 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3544 SC/04 Eximstats: fixed display of large numbers with -nvr option
3546 SC/05 Eximstats: fixed merging of reports with empty tables.
3548 SC/06 Eximstats: added the -include_original_destination flag
3550 SC/07 Eximstats: removed tabs and trailing whitespace.
3552 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3554 TK/06 MBOX spool code: Add real "From " MBOX separator line
3555 so the .eml file is really in mbox format (even though
3556 most programs do not really care). Patch from Alex Miller.
3558 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3559 The latter is generated from $received_to and is only set if the
3560 message has one envelope recipient. SA can use these headers,
3561 obviously out-of-the-box. Patch from Alex Miller.
3563 PH/08 The ${def test on a variable was returning false if the variable's
3564 value was "0", contrary to what the specification has always said!
3565 The result should be true unless the variable is empty.
3567 PH/09 The syntax error of a character other than { following "${if
3568 def:variable_name" (after optional whitespace) was not being diagnosed.
3569 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3570 accidental colon was present, for example, could give incorrect results.
3572 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3573 result is used (not including appendfile, where other changes are about
3576 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3577 This involved changing a lot of size variables from int to off_t. It
3578 should work with maildirs and everything.
3580 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3581 spamd dying while we are connected to it.
3583 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3586 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3587 function for BDB 4.3.
3589 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3591 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3592 the sources for PCRE have been re-arranged and more files are now
3595 PH/15 The code I had for printing potentially long long variables in PH/11
3596 above was not the best (it lost precision). The length of off_t variables
3597 is now inspected at build time, and an appropriate printing format (%ld
3598 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3599 to be "long long int" or "long int". This is needed for the internal
3600 formatting function string_vformat().
3602 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3603 the configuration file to be ":syslog", then the script "guesses" where
3604 the logs files are, rather than using the compiled in default. In our
3605 case the guess is not the same as the compiled default, so the script
3606 suddenly stopped working when I started to use syslog. The patch checks
3607 to see if log_file_path is "". If so, it attempts to read it from exim
3608 with no configuration file to get the compiled in version, before it
3609 falls back to the previous guessing code."
3611 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3612 implementing BATV in an Exim configuration. See NewStuff for the gory
3615 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3616 Makefile that are specific to HP-UX.
3618 PH/18 If the "use_postmaster" option was set for a recipient callout together
3619 with the "random" option, the postmaster address was used as the MAIL
3620 FROM address for the random test, but not for the subsequent recipient
3621 test. It is now used for both.
3623 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3624 patch removes a few documentation additions to RFC 3028, because the
3625 latest draft now contains them. It adds the new en;ascii-case comparator
3626 and a new error check for 8bit text in MIME parts. Comparator and
3627 require names are now matched exactly. I enabled the subaddress
3628 extension, but it is not well tested yet (read: it works for me)."
3630 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3631 rework some of the code of TK/09 above to avoid the hardwired use of
3632 "%lld" and "long long". Replaced the call to snprintf() with a call to
3635 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3636 records point to non-existent hosts", "retry timeout exceeded", and
3637 "retry time not reached for any host after a long failure period".
3639 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3640 experimental DomainKeys support:
3642 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3643 (2) On an error such as an illegally used "control", the wrong name for
3644 the control was given.
3646 These problems did NOT occur unless DomainKeys support was compiled.
3648 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3650 PH/24 Added ${if match_ip condition.
3652 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3653 databases so that it will be absolutely obvious if a crash occurs in the
3654 DB library. This is a regular occurrence (often caused by mis-matched
3657 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3658 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3659 picked up on a user's system that detects such things. There doesn't seem
3660 to be a gcc warning option for this - only an attribute that has to be
3661 put on the function's prototype. It seems that in Fedora Core 4 they have
3662 set this on a number of new functions. No doubt there will be more in due
3665 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3666 specify a transport. However, if an address that was verified by such a
3667 router was the subject of a callout, Exim crashed because it tried to
3668 read the rcpt_include_affixes from the non-existent transport. Now it
3669 just assumes that the setting of that option is false. This bug was
3670 introduced by 4.51/PH/31.
3672 PH/28 Changed -d+all to exclude +memory, because that information is very
3673 rarely of interest, but it makes the output a lot bigger. People tend to
3674 do -d+all out of habit.
3676 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3677 code in os-type was giving problems when libc.so lives in lib64, like on
3680 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3681 aren't the modern standard, and it seems that some systems' include files
3682 don't always have them. Exim was already checking for some of the newer
3683 ones like T_AAAA, and defining it itself. I've added checks for all the
3684 record types that Exim uses.
3686 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3687 not automatically generating a new one, as it is supposed to. This
3688 prevented TLS from working. If the file did exist, but contained invalid
3689 data, a new version was generated, as expected. It was only the case of a
3690 non-existent file that was broken.
3692 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3693 with a change in libdomainkeys > 0.64.
3695 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3696 from DNS. If the selector record carries the flag, it now has
3697 precedence over the domain-wide flag.
3699 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3701 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3702 the use of a port name that isn't defined in /etc/services) occurred, the
3703 message was deferred as in a normal delivery, and thus remained on the
3704 spool, instead of being failed because of the mua_wrapper setting. This
3705 is now fixed, and I tidied up some of the mua_wrapper messages at the
3708 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3709 the messages in a hash of arrays rather than using individual hashes.
3710 This is a bit cleaner and results in dramatic memory savings, albeit
3711 at a slight CPU cost.
3713 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3714 as requested by Marc Sherman.
3716 SC/10 Eximstats: added histograms for user specified patterns as requested
3719 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3721 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3722 fopen() in the content-scanning modules that did not already have it.
3728 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3729 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3731 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3733 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3735 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3736 to test Sieve filters that use "vacation".
3738 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3739 that changes the way the GnuTLS parameters are stored in the cache file.
3740 The new format can be generated externally. For backward compatibility,
3741 if the data in the cache doesn't make sense, Exim assumes it has read an
3742 old-format file, and it generates new data and writes a new file. This
3743 means that you can't go back to an older release without removing the
3746 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3747 work if there are any delivery delays because "one_time" forces the
3748 parent to be marked "delivered", so its unseen clone is never tried
3749 again. For this reason, Exim now forbids the simultaneous setting of
3752 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3753 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3754 entry if you want to know the details. The fix, however, neglected to
3755 consider the case when local delivery batching is involved. The test for
3756 "previously delivered" was not happening when checking to see if an
3757 address could be batched with a previous (undelivered) one; under
3758 certain circumstances this could lead to multiple deliveries to the same
3761 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3762 in its include files, and this causes problems building Exim.
3764 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3765 header_syntax) but Exim was just ignoring anything given after a slash.
3766 In particular, this caused confusion with an attempt to use "verify =
3767 reverse_host_lookup/defer_ok". An error is now given when options are
3768 supplied for verify items that do not have them. (Maybe reverse_host_
3769 lookup should have a defer_ok option, but that's a different point.)
3771 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3772 defined by RFC 821) to 2048, because there were problems with some AUTH
3773 commands, and RFC 1869 says the size should be increased for extended
3774 SMTP commands that take arguments.
3776 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3779 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3780 "unknown" error; now it says that the functionality isn't in the binary.
3782 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3783 an address' error message when a string expansion fails (syntax or
3784 whatever). Otherwise the password may appear in the log. Following change
3785 PH/42 below, there is no longer a chance of it appearing in a bounce
3788 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3790 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3791 of its addresses was ever tried. (Bugzilla bug #2.)
3793 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3794 the result incorrectly in the debug output. (It correctly added a newline
3795 to what was transported.)
3797 TF/01 Added $received_time.
3799 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3800 commented out examples of how to interface to a virus scanner and to
3801 SpamAssassin. Also added commented examples of av_scanner and
3802 spamd_address settings.
3804 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3805 and controls are allowed in which ACLs. There were a couple of minor
3806 errors. Some of the entries in the conditions table (which is a table of
3807 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3808 negation of where the condition IS allowed.
3810 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3812 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3813 header file does not have a version number, so I've had to invent a new
3814 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3815 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3816 radiusclient), but it was contributed by a Radius user.
3818 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3819 files or directories, for OpenSSL.
3821 PH/19 When an Exim process that is running as root has to create an Exim log
3822 file, it does so in a subprocess that runs as exim:exim so as to get the
3823 ownership right at creation (otherwise, other Exim processes might see
3824 the file with the wrong ownership). There was no test for failure of this
3825 fork() call, which would lead to the process getting stuck as it waited
3826 for a non-existent subprocess. Forks do occasionally fail when resources
3827 run out. I reviewed all the other calls to fork(); they all seem to check
3830 PH/20 When checking for unexpected SMTP input at connect time (before writing
3831 the banner), Exim was not dealing correctly with a non-positive return
3832 from the read() function. If the client had disconnected by this time,
3833 the result was a log entry for a synchronization error with an empty
3834 string after "input=" when read() returned zero. If read() returned -1
3835 (an event I could not check), uninitialized data bytes were printed.
3836 There were reports of junk text (parts of files, etc) appearing after
3839 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3841 PH/22 Added support for macro redefinition, and (re)definition in between
3842 driver and ACL definitions.
3844 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3845 forgetting to use the resulting value; it was using the unexpanded value.
3847 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3848 hadn't been configured. The fix is from Juergen Kreileder, who
3849 understands it better than I do:
3851 "Here's what I see happening with three configured cyrus_sasl
3852 authenticators configured (plain, login, cram-md5):
3854 On startup auth_cyrus_sasl_init() gets called for each of these.
3855 This means three calls to sasl_listmech() without a specified mech_list.
3856 => SASL tests which mechs of all available mechs actually work
3857 => three warnings about OTP not working
3858 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3860 With the patch, sasl_listmech() also gets called three times. But now
3861 SASL's mech_list option is set to the server_mech specified in the the
3862 authenticator. Or in other words, the answer from sasl_listmech()
3863 gets limited to just the mech you're testing for (which is different
3865 => the return list contains just 'plain' or 'login', 'cram-md5' or
3866 nothing depending on the value of ob->server_mech.
3868 I've just tested the patch: Authentication still works fine,
3869 unavailable mechs specified in the exim configuration are still
3870 caught, and the auth.log warnings about OTP are gone."
3872 PH/25 When debugging is enabled, the contents of the command line are added
3873 to the debugging output, even when log_selector=+arguments is not
3876 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3877 answer is "GNU", and only if the return is "GNU/something" is the answer
3880 PH/27 $acl_verify_message is now set immediately after the failure of a
3881 verification in an ACL, and so is available in subsequent modifiers. In
3882 particular, the message can be preserved by coding like this:
3884 warn !verify = sender
3885 set acl_m0 = $acl_verify_message
3887 Previously, $acl_verify_message was set only while expanding "message"
3888 and "log_message" when a very denied access.
3890 PH/28 Modified OS/os.c-Linux with
3892 -#ifndef OS_LOAD_AVERAGE
3893 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3895 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3896 nomenclature these days.)
3898 PH/29 Installed patch from the Sieve maintainer that adds the options
3899 sieve_useraddress and sieve_subaddress to the redirect router.
3901 PH/30 In these circumstances:
3902 . Two addresses routed to the same list of hosts;
3903 . First host does not offer TLS;
3904 . First host accepts first address;
3905 . First host gives temporary error to second address;
3906 . Second host offers TLS and a TLS session is established;
3907 . Second host accepts second address.
3908 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3909 and peerdn, if requested) that were in fact used only for the second
3912 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3913 attention to any local part prefix or suffix that was matched by the
3914 router that accepted the address. It now behaves in the same way as it
3915 does for delivery: the affixes are removed from the local part unless
3916 rcpt_include_affixes is set on the transport.
3918 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3919 timeout during the DATA phase of an incoming message.
3921 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3922 applied a patch sanctioned by the Sieve maintainer.
3924 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3925 the uid or gid is negative. A case of a negative gid caused this to be
3926 noticed. The fix allows for either to be negative.
3928 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3929 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3932 PH/36 The change PH/12 above was broken. Fixed it.
3934 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3935 the grounds that routing the same address twice would always produce the
3936 same answer. This might have been true once, but it is certainly no
3937 longer true now. Routing a child address may depend on the previous
3938 routing that produced that child. Some complicated redirection strategies
3939 went wrong when messages had multiple recipients, and made Exim's
3940 behaviour dependent on the order in which the addresses were given.
3942 I have moved the duplicate checking until after the routing is complete.
3943 Exim scans the addresses that are assigned to local and remote
3944 transports, and removes any duplicates. This means that more work will be
3945 done, as duplicates will always all be routed, but duplicates are
3946 presumably rare, so I don't expect this is of any significance.
3948 For deliveries to pipes, files, and autoreplies, the duplicate checking
3949 still happens during the routing process, since they are not going to be
3952 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3953 It corrects a timeout issue with spamd. This is Ian's comment: "The
3954 background is that sometimes spamd either never reads data from a
3955 connection it has accepted, or it never writes response data. The exiscan
3956 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3957 blindly assumes that writes won't block so it may never time out."
3959 PH/39 Allow G after quota size as well as K and M.
3961 PH/40 The value set for $authenticated_id in an authenticator may not contain
3962 binary zeroes or newlines because the value is written to log lines and
3963 to spool files. There was no check on this. Now the value is run through
3964 the string_printing() function so that such characters are converted to
3965 printable escape sequences.
3967 PH/41 $message_linecount is a new variable that contains the total number of
3968 lines in the message. Compare $body_linecount, which is the count for the
3971 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3972 bounce and delay warning messages, except in certain special cases, which
3975 (a) An SMTP error message from a remote host;
3976 (b) A message specified in a :fail: redirection;
3977 (c) A message specified in a "fail" command in a system filter;
3978 (d) A message specified in a FAIL return from the queryprogram router;
3979 (e) A message specified by the cannot_route_message router option.
3981 In these cases only, Exim does include the error details in bounce and
3982 warning messages. There are also a few cases where bland messages such
3983 as "unrouteable address" or "local delivery error" are given.
3985 PH/43 $value is now also set for the "else" part of a ${run expansion.
3987 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3988 being worked on, but at least Exim now implements the latest version to
3991 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3992 process to complete was treated as a delivery failure, a timeout while
3993 writing the message to the pipe was logged, but erroneously treated as a
3994 successful delivery. Such timeouts include transport filter timeouts. For
3995 consistency with the overall process timeout, these timeouts are now
3996 treated as errors, giving rise to delivery failures by default. However,
3997 there is now a new Boolean option for the pipe transport called
3998 timeout_defer, which, if set TRUE, converts the failures into defers for
3999 both kinds of timeout. A transport filter timeout is now identified in
4002 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4003 systems where "make" and "gmake" are different, calling "gmake" at top
4004 level broke things. I've arranged for the value of $(MAKE) to be passed
4005 from the Makefile to this script so that it can call the same version of
4009 A note about Exim versions 4.44 and 4.50
4010 ----------------------------------------
4012 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4013 changes of various kinds. As a consequence, a big documentation update was
4014 needed. This delayed the release for rather longer than seemed good, especially
4015 in the light of a couple of (minor) security issues. Therefore, the changes
4016 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4017 release. So 4.44 and 4.50 are in effect two different branches that both start
4020 I have left the 4.50 change log unchanged; it contains all the changes since
4021 4.43. The change log for 4.44 is below; many of its items are identical to
4022 those for 4.50. This seems to be the most sensible way to preserve the
4023 historical information.
4029 1. Minor wording change to the doc/README.SIEVE file.
4031 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4032 computation of the current number of files was incorrect.
4034 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4035 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4038 4. Give more explanation in the error message when the command for a transport
4039 filter fails to execute.
4041 5. There are several places where Exim runs a non-Exim command in a
4042 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4043 was being done only for the command run by the queryprogram router. It is
4044 now done for all such subprocesses. The other cases are: ${run, transport
4045 filters, and the commands run by the lmtp and pipe transports.
4047 6. Added CONFIGURE_GROUP build-time option.
4049 7. Some older OS have a limit of 256 on the maximum number of file
4050 descriptors. Exim was using setrlimit() to set 1000 as a large value
4051 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4052 systems. I've change it so that if it can't get 1000, it tries for 256.
4054 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4055 was an oversight, and furthermore, ever since the addition of extra
4056 controls (e.g. 4.43/32), the checks on when to allow different forms of
4057 "control" were broken. There should now be diagnostics for all cases when a
4058 control that does not make sense is encountered.
4060 9. Added the /retain_sender option to "control=submission".
4062 10. $recipients is now available in the predata ACL (oversight).
4064 11. Tidy the search cache before the fork to do a delivery from a message
4065 received from the command line. Otherwise the child will trigger a lookup
4066 failure and thereby defer the delivery if it tries to use (for example) a
4067 cached ldap connection that the parent has called unbind on.
4069 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4070 of $address_data from the recipient verification was clobbered by the
4071 sender verification.
4073 13. The value of address_data from a sender verification is now available in
4074 $sender_address_data in subsequent conditions in the ACL statement.
4076 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4078 15. Added a new option "connect=<time>" to callout options, to set a different
4081 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4082 was its contents. (It was OK if the option was not defined at all.)
4084 17. A "Completed" log line is now written for messages that are removed from
4085 the spool by the -Mrm option.
4087 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4088 information about exactly what failed.
4090 19. Added -dd to debug only the daemon process.
4092 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4093 handles timeouts, both on the server side and network timeouts. Renamed the
4094 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4096 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4097 It is now set to "smtps".
4099 22. $host_address is now set to the target address during the checking of
4100 ignore_target_hosts.
4102 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4103 being passed; this would have caused $sender_host_name to have been used if
4104 matching the list had actually called for a host name (not very likely,
4105 since this list is usually IP addresses). A host name is now passed as
4108 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4109 code that specifies a non-restarting handler (typically sigaction() in
4110 modern systems) in an attempt to fix a rare and obscure crash bug.
4112 25. Narrowed the window for a race in the daemon that could cause it to ignore
4113 SIGCHLD signals. This is not a major problem, because they are used only to
4114 wake it up if nothing else does.
4116 26. A malformed maildirsize file could cause Exim to calculate negative values
4117 for the mailbox size or file count. Odd effects could occur as a result.
4118 The maildirsize information is now recalculated if the size or filecount
4121 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4122 support for a long time. Removed HAVE_SYS_VFS_H.
4124 28. Installed the latest version of exipick from John Jetmore.
4126 29. In an address list, if the pattern was not a regular expression, an empty
4127 subject address (from a bounce message) matched only if the pattern was an
4128 empty string. Non-empty patterns were not even tested. This was the wrong
4129 because it is perfectly reasonable to use an empty address as part of a
4130 database query. An empty address is now tested by patterns that are
4131 lookups. However, all the other forms of pattern expect the subject to
4132 contain a local part and a domain, and therefore, for them, an empty
4133 address still always fails if the pattern is not itself empty.
4135 30. Exim went into a mad DNS loop when attempting to do a callout where the
4136 host was specified on an smtp transport, and looking it up yielded more
4137 than one IP address.
4139 31. Re-factored the code for checking spool and log partition space into a
4140 function that finds that data and another that does the check. The former
4141 is then used to implement four new variables: $spool_space, $log_space,
4142 $spool_inodes, and $log_inodes.
4144 32. The RFC2047 encoding function was originally intended for short strings
4145 such as real names; it was not keeping to the 75-character limit for
4146 encoded words that the RFC imposes. It now respects the limit, and
4147 generates multiple encoded words if necessary. To be on the safe side, I
4148 have increased the buffer size for the ${rfc2047: expansion operator from
4151 33. It is now permitted to omit both strings after an "if" condition; if the
4152 condition is true, the result is "true". As before, when the second string
4153 is omitted, a false condition yields an empty string. This makes it less
4154 cumbersome to write custom ACL and router conditions.
4156 34. Failure to deliver a bounce message always caused it to be frozen, even if
4157 there was an errors_to setting on the router. The errors_to setting is now
4160 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4161 canonical form (fully expanded) before being placed in
4162 $sender_host_address.
4164 36. The table in the code that translates DNS record types into text (T_A to
4165 "A" for instance) was missing entries for NS and CNAME. It is just possible
4166 that this could have caused confusion if both these types were looked up
4167 for the same domain, because the text type is used as part of Exim's
4168 per-process caching. But the chance of anyone hitting this buglet seems
4171 37. The dnsdb lookup has been extended in a number of ways.
4173 (1) There is a new type, "zns", which walks up the domain tree until it
4174 finds some nameserver records. It should be used with care.
4176 (2) There is a new type, "mxh", which is like "mx" except that it returns
4177 just the host names, not the priorities.
4179 (3) It is now possible to give a list of domains (or IP addresses) to be
4180 looked up. The behaviour when one of the lookups defers can be
4181 controlled by a keyword.
4183 (4) It is now possible to specify the separator character for use when
4184 multiple records are returned.
4186 38. The dnslists ACL condition has been extended: it is now possible to supply
4187 a list of IP addresses and/or domains to be looked up in a particular DNS
4190 39. Added log_selector=+queue_time_overall.
4192 40. When running the queue in the test harness, wait just a tad after forking a
4193 delivery process, to get repeatability of debugging output.
4195 41. Include certificate and key file names in error message when GnuTLS fails
4196 to set them up, because the GnuTLS error message doesn't include the name
4197 of the failing file when there is a problem reading it.
4199 42. Allow both -bf and -bF in the same test run.
4201 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4203 44. The "Exiscan patch" is now merged into the mainline Exim source.
4205 45. Sometimes the final signoff response after QUIT could fail to get
4206 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4207 < 0 before doing a fflush(). This bug looks as though it goes back to the
4208 introduction of TLS in release 3.20, but "sometimes" must have been rare
4209 because the tests only now provoked it.
4211 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4212 (this can affect the format of dates).
4214 47. exim_tidydb, when checking for the continued existence of a message for
4215 which it has found a message-specific retry record, was not finding
4216 messages that were in split spool directories. Consequently, it was
4217 deleting retry records that should have stayed in existence.
4219 48. Steve fixed some bugs in eximstats.
4221 49. The SPA authentication driver was not abandoning authentication and moving
4222 on to the next authenticator when an expansion was forced to fail,
4223 contradicting the general specification for all authenticators. Instead it
4224 was generating a temporary error. It now behaves as specified.
4226 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4227 (the order specifies the preference for clients). The order is now AES256,
4228 AES128, 3DES, ARCFOUR128.
4230 51. Small patch to Sieve code - explicitly set From: when generating an
4233 52. Exim crashed if a remote delivery caused a very long error message to be
4234 recorded - for instance if somebody sent an entire SpamAssassin report back
4235 as a large number of 550 error lines. This bug was coincidentally fixed by
4236 increasing the size of one of Exim's internal buffers (big_buffer) that
4237 happened as part of the Exiscan merge. However, to be on the safe side, I
4238 have made the code more robust (and fixed the comments that describe what
4241 53. Now that there can be additional text after "Completed" in log lines (if
4242 the queue_time_overall log selector is set), a one-byte patch to exigrep
4243 was needed to allow it to recognize "Completed" as not the last thing in
4246 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4247 patch that reportedly fixes this has been added. I am not expert enough to
4248 create a test for it. This is what the patch creator wrote:
4250 "I found a little strange behaviour of ldap code when working with
4251 Windows 2003 AD Domain, where users was placed in more than one
4252 Organization Units. When I tried to give exim partial DN, the exit code
4253 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4254 But simultaneously result of request was absolutely normal ldap result,
4255 so I produce this patch..."
4257 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4258 REFERENCE, so I have modified the code to exclude the patch when that macro
4261 55. Some experimental protocols are using DNS PTR records for new purposes. The
4262 keys for these records are domain names, not reversed IP addresses. The
4263 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4264 leaves it alone. Component reversal etc. now happens only for IP addresses.
4267 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4269 57. Double the size of the debug message buffer (to 2048) so that more of very
4270 long debug lines gets shown.
4272 58. The exicyclog utility now does better if the number of log files to keep
4273 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4275 59. Two changes related to the smtp_active_hostname option:
4277 (1) $smtp_active_hostname is now available as a variable.
4278 (2) The default for smtp_banner uses $smtp_active_hostname instead
4279 of $primary_hostname.
4281 60. The host_aton() function is supposed to be passed a string that is known
4282 to be a valid IP address. However, in the case of IPv6 addresses, it was
4283 not checking this. This is a hostage to fortune. Exim now panics and dies
4284 if the condition is not met. A case was found where this could be provoked
4285 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4286 components; fortuitously, this particular loophole had already been fixed
4287 by change 4.50/55 above.
4289 If there are any other similar loopholes, the new check in host_aton()
4290 itself should stop them being exploited. The report I received stated that
4291 data on the command line could provoke the exploit when Exim was running as
4292 exim, but did not say which command line option was involved. All I could
4293 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4294 running as the user.
4297 61. There was a buffer overflow vulnerability in the SPA authentication code
4298 (which came originally from the Samba project). I have added a test to the
4299 spa_base64_to_bits() function which I hope fixes it.
4302 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4303 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4305 63. The daemon start-up calls getloadavg() while still root for those OS that
4306 need the first call to be done as root, but it missed one case: when
4307 deliver_queue_load_max is set with deliver_drop_privilege. This is
4308 necessary for the benefit of the queue runner, because there is no re-exec
4309 when deliver_drop_privilege is set.
4311 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4312 This has been fixed.
4314 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4315 was in use, was not putting the data itself into the right store pool;
4316 consequently, it could be overwritten for a subsequent message in the same
4317 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4320 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4322 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4323 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4324 the function were treating the return as a boolean value, which happened to
4325 work because 0=false and not-0=true, but is not correct code.
4327 68. The host_aton() function was not handling scoped IPv6 addresses (those
4328 with, for example, "%eth0" on the end) correctly.
4330 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4331 negated items (that is, ~something) in unsigned ints. Some compilers
4332 apparently mutter when there is no cast.
4334 70. If an address verification called from an ACL failed, and did not produce a
4335 user-specific message (i.e. there was only a "system" message), nothing was
4336 put in $acl_verify_message. In this situation, it now puts the system
4339 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4340 session; change 4.31/43 added the unwanted input to the log line - except
4341 that it did not do this in the start of session case. It now does.
4343 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4344 This is wrong and can cause the other end to generate a synchronization
4345 error if it is another Exim or anything else that does the synchronization
4346 check. A QUIT command is no longer sent after a timeout.
4348 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4349 during host lookups.
4351 74. The defer_ok option of callout verification was not working if it was used
4352 when verifying addresses in header lines, that is, for this case:
4354 verify = header_sender/callout=defer_ok
4356 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4357 those file descriptors could be used for SMTP connections. If anything
4358 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4359 could be sent to the SMTP client, causing chaos. The daemon now opens
4360 stdin, stdout, and stderr to /dev/null when it puts itself into the
4363 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4364 log by default. The user can override this with suitable Perl magic.
4366 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4367 the log message when discard triggers, was not working for the DATA ACL or
4368 for the non-SMTP ACL.
4370 78. Error message wording change in sieve.c.
4372 79. If smtp_accept_max_per_host was set, the number of connections could be
4373 restricted to fewer than expected, because the daemon was trying to set up
4374 a new connection before checking whether the processes handling previous
4375 connections had finished. The check for completed processes is now done
4376 earlier. On busy systems, this bug wouldn't be noticed because something
4377 else would have woken the daemon, and it would have reaped the completed
4380 80. If a message was submitted locally by a user whose login name contained one
4381 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4382 It caused a spool format error. I have fixed the spool reading code. A
4383 related problem was that the "from" clause in the Received: line became
4384 illegal because of the space(s). It is now covered by ${quote_local_part.
4386 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4389 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4391 83. Patch from Sieve maintainer to fix unterminated string problem in
4392 "vacation" handling.
4394 84. Some minor changes to the Linux configuration files to help with other
4395 OS variants using glibc.
4397 85. One more patch for Sieve to update vacation handling to latest spec.
4400 ----------------------------------------------------
4401 See the note above about the 4.44 and 4.50 releases.
4402 ----------------------------------------------------
4408 1. Change 4.43/35 introduced a bug that caused file counts to be
4409 incorrectly computed when quota_filecount was set in an appendfile
4412 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4413 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4416 3. Give more explanation in the error message when the command for a transport
4417 filter fails to execute.
4419 4. There are several places where Exim runs a non-Exim command in a
4420 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4421 was being done only for the command run by the queryprogram router. It is
4422 now done for all such subprocesses. The other cases are: ${run, transport
4423 filters, and the commands run by the lmtp and pipe transports.
4425 5. Some older OS have a limit of 256 on the maximum number of file
4426 descriptors. Exim was using setrlimit() to set 1000 as a large value
4427 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4428 systems. I've change it so that if it can't get 1000, it tries for 256.
4430 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4431 was an oversight, and furthermore, ever since the addition of extra
4432 controls (e.g. 4.43/32), the checks on when to allow different forms of
4433 "control" were broken. There should now be diagnostics for all cases when a
4434 control that does not make sense is encountered.
4436 7. $recipients is now available in the predata ACL (oversight).
4438 8. Tidy the search cache before the fork to do a delivery from a message
4439 received from the command line. Otherwise the child will trigger a lookup
4440 failure and thereby defer the delivery if it tries to use (for example) a
4441 cached ldap connection that the parent has called unbind on.
4443 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4444 of $address_data from the recipient verification was clobbered by the
4445 sender verification.
4447 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4448 was its contents. (It was OK if the option was not defined at all.)
4450 11. A "Completed" log line is now written for messages that are removed from
4451 the spool by the -Mrm option.
4453 12. $host_address is now set to the target address during the checking of
4454 ignore_target_hosts.
4456 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4457 being passed; this would have caused $sender_host_name to have been used if
4458 matching the list had actually called for a host name (not very likely,
4459 since this list is usually IP addresses). A host name is now passed as
4462 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4463 code that specifies a non-restarting handler (typically sigaction() in
4464 modern systems) in an attempt to fix a rare and obscure crash bug.
4466 15. Narrowed the window for a race in the daemon that could cause it to ignore
4467 SIGCHLD signals. This is not a major problem, because they are used only to
4468 wake it up if nothing else does.
4470 16. A malformed maildirsize file could cause Exim to calculate negative values
4471 for the mailbox size or file count. Odd effects could occur as a result.
4472 The maildirsize information is now recalculated if the size or filecount
4475 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4476 support for a long time. Removed HAVE_SYS_VFS_H.
4478 18. Updated exipick to current release from John Jetmore.
4480 19. Allow an empty sender to be matched against a lookup in an address list.
4481 Previously the only cases considered were a regular expression, or an
4484 20. Exim went into a mad DNS lookup loop when doing a callout where the
4485 host was specified on the transport, if the DNS lookup yielded more than
4488 21. The RFC2047 encoding function was originally intended for short strings
4489 such as real names; it was not keeping to the 75-character limit for
4490 encoded words that the RFC imposes. It now respects the limit, and
4491 generates multiple encoded words if necessary. To be on the safe side, I
4492 have increased the buffer size for the ${rfc2047: expansion operator from
4495 22. Failure to deliver a bounce message always caused it to be frozen, even if
4496 there was an errors_to setting on the router. The errors_to setting is now
4499 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4500 canonical form (fully expanded) before being placed in
4501 $sender_host_address.
4503 24. Updated eximstats to version 1.33
4505 25. Include certificate and key file names in error message when GnuTLS fails
4506 to set them up, because the GnuTLS error message doesn't include the name
4507 of the failing file when there is a problem reading it.
4509 26. Expand error message when OpenSSL has problems setting up cert/key files.
4512 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4513 (this can affect the format of dates).
4515 28. exim_tidydb, when checking for the continued existence of a message for
4516 which it has found a message-specific retry record, was not finding
4517 messages that were in split spool directories. Consequently, it was
4518 deleting retry records that should have stayed in existence.
4520 29. eximstats updated to version 1.35
4521 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4522 1.35 - bugfix such that pie charts by volume are generated correctly
4524 30. The SPA authentication driver was not abandoning authentication and moving
4525 on to the next authenticator when an expansion was forced to fail,
4526 contradicting the general specification for all authenticators. Instead it
4527 was generating a temporary error. It now behaves as specified.
4529 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4530 (the order specifies the preference for clients). The order is now AES256,
4531 AES128, 3DES, ARCFOUR128.
4533 31. Small patch to Sieve code - explicitly set From: when generating an
4536 32. Exim crashed if a remote delivery caused a very long error message to be
4537 recorded - for instance if somebody sent an entire SpamAssassin report back
4538 as a large number of 550 error lines. This bug was coincidentally fixed by
4539 increasing the size of one of Exim's internal buffers (big_buffer) that
4540 happened as part of the Exiscan merge. However, to be on the safe side, I
4541 have made the code more robust (and fixed the comments that describe what
4544 33. Some experimental protocols are using DNS PTR records for new purposes. The
4545 keys for these records are domain names, not reversed IP addresses. The
4546 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4547 leaves it alone. Component reversal etc. now happens only for IP addresses.
4550 34. The host_aton() function is supposed to be passed a string that is known
4551 to be a valid IP address. However, in the case of IPv6 addresses, it was
4552 not checking this. This is a hostage to fortune. Exim now panics and dies
4553 if the condition is not met. A case was found where this could be provoked
4554 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4555 components; fortuitously, this particular loophole had already been fixed
4556 by change 4.50/55 or 4.44/33 above.
4558 If there are any other similar loopholes, the new check in host_aton()
4559 itself should stop them being exploited. The report I received stated that
4560 data on the command line could provoke the exploit when Exim was running as
4561 exim, but did not say which command line option was involved. All I could
4562 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4563 running as the user.
4566 35. There was a buffer overflow vulnerability in the SPA authentication code
4567 (which came originally from the Samba project). I have added a test to the
4568 spa_base64_to_bits() function which I hope fixes it.
4571 36. The daemon start-up calls getloadavg() while still root for those OS that
4572 need the first call to be done as root, but it missed one case: when
4573 deliver_queue_load_max is set with deliver_drop_privilege. This is
4574 necessary for the benefit of the queue runner, because there is no re-exec
4575 when deliver_drop_privilege is set.
4577 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4578 was in use, was not putting the data itself into the right store pool;
4579 consequently, it could be overwritten for a subsequent message in the same
4580 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4583 38. Sometimes the final signoff response after QUIT could fail to get
4584 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4585 < 0 before doing a fflush(). This bug looks as though it goes back to the
4586 introduction of TLS in release 3.20, but "sometimes" must have been rare
4587 because the tests only now provoked it.
4593 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4594 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4595 Now it returns an integer. A number of calls were still expecting a T/F
4596 return. Fortuitously, in all cases, the tests worked in OK situations,
4597 which is the norm. However, things would have gone wrong on any write
4598 failures on the smtp file descriptor. This function is used when sending
4599 messages over SMTP and also when doing verify callouts.
4601 2. When Exim is called to do synchronous delivery of a locally submitted
4602 message (the -odf or -odi options), it no longer closes stderr before doing
4605 3. Implemented the mua_wrapper option.
4607 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4609 5. Implemented the functions header_remove(), header_testname(),
4610 header_add_at_position(), and receive_remove_recipient(), and exported them
4613 6. If an ACL "warn" statement specified the addition of headers, Exim already
4614 inserted X-ACL-Warn: at the start if there was no header name. However, it
4615 was not making this test for the second and subsequent header lines if
4616 there were newlines in the string. This meant that an invalid header could
4617 be inserted if Exim was badly configured.
4619 7. Allow an ACL "warn" statement to add header lines at the start or after all
4620 the Received: headers, as well as at the end.
4622 8. Added the rcpt_4xx retry error code.
4624 9. Added postmaster_mailfrom=xxx to callout verification option.
4626 10. Added mailfrom=xxxx to the callout verification option, for verify=
4629 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4630 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4632 12. Inserted some casts to stop certain compilers warning when using pointer
4633 differences as field lengths or precisions in printf-type calls (mostly
4634 affecting debugging statements).
4636 13. Added optional readline() support for -be (dynamically loaded).
4638 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4639 same clock tick as a message's arrival, so that its received time was the
4640 same as the "first fail" time on the retry record, and that message
4641 remained on the queue past the ultimate address timeout, every queue runner
4642 would try a delivery (because it was past the ultimate address timeout) but
4643 after another failure, the ultimate address timeout, which should have then
4644 bounced the address, did not kick in. This was a "< instead of <=" error;
4645 in most cases the first failure would have been in the next clock tick
4646 after the received time, and all would be well.
4648 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4649 being recognized when the domain list was tested by the match_domain
4650 condition in an expansion string.
4652 16. Added the ${str2b64: operator.
4654 17. Exim was always calling setrlimit() to set a large limit for the number of
4655 processes, without checking whether the existing limit was already
4656 adequate. (It did check for the limit on file descriptors.) Furthermore,
4657 errors from getrlimit() and setrlimit() were being ignored. Now they are
4658 logged to the main and panic logs, but Exim does carry on, to try to do its
4659 job under whatever limits there are.
4661 18. Imported PCRE 5.0.
4663 19. Trivial typo in log message " temporarily refused connection" (the leading
4666 20. If the log selector return_path_on_delivery was set and an address was
4667 redirected to /dev/null, the delivery process crashed because it assumed
4668 that a return path would always be set for a "successful" delivery. In this
4669 case, the whole delivery is bypassed as an optimization, and therefore no
4672 21. Internal re-arrangement: the function for sending a challenge and reading
4673 a response while authentication was assuming a zero-terminated challenge
4674 string. It's now changed to take a pointer and a length, to allow for
4675 binary data in such strings.
4677 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4679 23. Exim was not respecting finduser_retries when seeking the login of the
4680 uid under which it was called; it was always trying 10 times. (The default
4681 setting of finduser_retries is zero.) Also, it was sleeping after the final
4682 failure, which is pointless.
4684 24. Implemented tls_on_connect_ports.
4686 25. Implemented acl_smtp_predata.
4688 26. If the domain in control=submission is set empty, Exim assumes that the
4689 authenticated id is a complete email address when it generates From: or
4690 Sender: header lines.
4692 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4693 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4694 chown and chgrp in /bin and hostname in /usr/bin.
4696 28. Exim was keeping the "process log" file open after each use, just as it
4697 does for the main log. This opens the possibility of it remaining open for
4698 long periods when the USR1 signal hits a daemon. Occasional processlog
4699 errors were reported, that could have been caused by this. Anyway, it seems
4700 much more sensible not to leave this file open at all, so that is what now
4703 29. The long-running daemon process does not normally write to the log once it
4704 has entered its main loop, and it closes the log before doing so. This is
4705 so that log files can straightforwardly be renamed and moved. However,
4706 there are a couple of unusual error situations where the daemon does write
4707 log entries, and I had neglected to close the log afterwards.
4709 30. The text of an SMTP error response that was received during a remote
4710 delivery was being truncated at 512 bytes. This is too short for some of
4711 the long messages that one sometimes sees. I've increased the limit to
4714 31. It is now possible to make retry rules that apply only when a message has a
4715 specific sender, in particular, an empty sender.
4717 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4718 it possible to be selective about when SMTP synchronization is enforced.
4720 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4722 32. Implemented hosts_connection_nolog.
4724 33. Added an ACL for QUIT.
4726 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4729 35. Added mailbox_size and mailbox_filecount to appendfile.
4731 36. Added control = no_multiline_responses to ACLs.
4733 37. There was a bug in the logic of the code that waits for the clock to tick
4734 in the case where the clock went backwards by a substantial amount such
4735 that the microsecond fraction of "now" was more than the microsecond
4736 fraction of "then" (but the whole seconds number was less).
4738 38. Added support for the libradius Radius client library this is found on
4739 FreeBSD (previously only the radiusclient library was supported).
4745 1. When certain lookups returned multiple values in the form name=value, the
4746 quoting of the values was not always being done properly. Specifically:
4747 (a) If the value started with a double quote, but contained no whitespace,
4749 (b) If the value contained whitespace other than a space character (i.e.
4750 tabs or newlines or carriage returns) it was not quoted.
4751 This fix has been applied to the mysql and pgsql lookups by writing a
4752 separate quoting function and calling it from the lookup code. The fix
4753 should probably also be applied to nisplus, ibase and oracle lookups, but
4754 since I cannot test any of those, I have not disturbed their existing code.
4756 2. A hit in the callout cache for a specific address caused a log line with no
4757 reason for rejecting RCPT. Now it says "Previous (cached) callout
4758 verification failure".
4760 3. There was an off-by-one bug in the queryprogram router. An over-long
4761 return line was truncated at 256 instead of 255 characters, thereby
4762 overflowing its buffer with the terminating zero. As well as fixing this, I
4763 have increased the buffer size to 1024 (and made a note to document this).
4765 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4766 when Exim is waiting for an SMTP response from a remote server, Exim
4767 restarts its select() call on the socket, thereby resetting its timeout.
4768 This is not a problem when such interrupts are rare. Somebody set up a cron
4769 job to run exiwhat every 2 minutes, which is less than the normal select()
4770 timeout (5 or 10 minutes). This meant that the select() timeout never
4771 kicked in because it was always reset. I have fixed this by comparing the
4772 time when an interrupt arrives with the time at the start of the first call
4773 to select(). If more time than the timeout has elapsed, the interrupt is
4774 treated as a timeout.
4776 5. Some internal re-factoring in preparation for the addition of Sieve
4777 extensions (by MH). In particular, the "personal" test is moved to a
4778 separate function, and given an option for scanning Cc: and Bcc: (which is
4779 not set for Exim filters).
4781 6. When Exim created an email address using the login of the caller as the
4782 local part (e.g. when creating a From: or Sender: header line), it was not
4783 quoting the local part when it contained special characters such as @.
4785 7. Installed new OpenBSD configuration files.
4787 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4788 try to make them clearer.
4790 9. Callout options, other than the timeout value, were being ignored when
4791 verifying sender addresses in header lines. For example, when using
4793 verify = header_sender/callout=no_cache
4795 the cache was (incorrectly) being used.
4797 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4798 only the Cygwin environment.
4800 11. When return_path_on_delivery was set as a log selector, if different remote
4801 addresses in the same message used different return paths and parallel
4802 remote delivery occurred, the wrong values would sometimes be logged.
4803 (Whenever a remote delivery process finished, the return path value from
4804 the most recently started remote delivery process was logged.)
4806 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4807 lines when AUTH and/or TLS are in use. This is the "received protocol"
4808 field. Exim used to use "asmtp" for authenticated SMTP, without any
4809 indication (in the protocol name) for TLS use. Now it follows the RFC and
4810 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4811 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4812 names appear in log lines as well as in Received: header lines.
4814 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4815 extensions, and comparison tests, and to fix some bugs.
4817 14. Changes to the "personal" filter test:
4819 (1) The test was buggy in that it was just doing the equivalent of
4820 "contains" tests on header lines. For example, if a user's address was
4821 anne@some.where, the "personal" test would incorrectly be true for
4823 To: susanne@some.where
4825 This test is now done by extracting each address from the header in turn,
4826 and checking the entire address. Other tests that are part of "personal"
4827 are now done using regular expressions (for example, to check local parts
4828 of addresses in From: header lines).
4830 (2) The list of non-personal local parts in From: addresses has been
4831 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4832 taken from the Sieve specification recommendations.
4834 (3) If the message contains any header line starting with "List-" it is
4835 treated as non-personal.
4837 (4) The test for "circular" in the Subject: header line has been removed
4838 because it now seems ill-conceived.
4840 15. Minor typos in src/EDITME comments corrected.
4842 16. Installed latest exipick from John Jetmore.
4844 17. If headers_add on a router specified a text string that was too long for
4845 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4846 of string_sprintf() is now avoided.
4848 18. $message_body_size was not set (it was always zero) when running the DATA
4849 ACL and the local_scan() function.
4851 19. For the "mail" command in an Exim filter, no default was being set for
4852 the once_repeat time, causing a random time value to be used if "once" was
4853 specified. (If the value happened to be <= 0, no repeat happened.) The
4854 default is now 0s, meaning "never repeat". The "vacation" command was OK
4855 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4856 (I found it when inspecting the code).
4858 20. There is now an overall timeout for performing a callout verification. It
4859 defaults to 4 times the callout timeout, which applies to individual SMTP
4860 commands during the callout. The overall timeout applies when there is more
4861 than one host that can be tried. The timeout is checked before trying the
4862 next host. This prevents very long delays if there are a large number of
4863 hosts and all are timing out (e.g. when the network connections are timing
4864 out). The value of the overall timeout can be changed by specifying an
4865 additional sub-option for "callout", called "maxwait". For example:
4867 verify = sender/callout=5s,maxwait=20s
4869 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4870 to the end before writing, but this should make it even safer).
4872 22. Exim was forgetting that it had advertised PIPELINING for the second and
4873 subsequent messages on an SMTP connection. It was also not resetting its
4874 memory on STARTTLS and an internal HELO.
4876 23. When Exim logs an SMTP synchronization error within a session, it now
4877 records whether PIPELINING has been advertised or not.
4879 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4880 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4881 rather than long int.
4883 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4885 26. Added the never_mail option to autoreply.
4891 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4892 crash if the getsockname() call failed; this can happen if a connection is
4893 closed very soon after it is established. The problem was simply in the
4894 order in which certain operations were done, causing Exim to try to write
4895 to the SMTP stream before it had set up the file descriptor. The bug has
4896 been fixed by making things happen in the correct order.
4902 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4903 before the connection was closed, thus losing the rejection response.
4905 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4906 some early Solaris releases, but causes trouble in current releases where
4907 socklen_t is defined.
4909 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4912 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4915 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4916 that releasing the top of it at the end releases what was used for sub-
4917 expansions (unless the block got too big). However, discard this block if
4918 the first thing is a variable or header, so that we can use its block when
4919 it is dynamic (useful for very large $message_headers, for example).
4921 6. Lookups now cache *every* query, not just the most recent. A new, separate
4922 store pool is used for this. It can be recovered when all lookup caches are
4923 flushed. Lookups now release memory at the end of their result strings.
4924 This has involved some general refactoring of the lookup sources.
4926 7. Some code has been added to the store_xxx() functions to reduce the amount
4927 of flapping under certain conditions.
4929 8. log_incoming_interface used to affect only the <= reception log lines. Now
4930 it causes the local interface and port to be added to several more SMTP log
4931 lines, for example "SMTP connection from", and rejection lines.
4933 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4935 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4937 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4938 could be overwritten at the end of the current message (or the start of a
4939 new message if it was set in a HELO ACL). The value is now preserved for
4940 the duration of the SMTP connection.
4942 12. If a transport had a headers_rewrite setting, and a matching header line
4943 contained an unqualified address, that address was qualified, even if it
4944 did not match any rewriting rules. The underlying bug was that the values
4945 of the flags that permit the existence of unqualified sender and recipient
4946 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4947 non-local messages, and by -bnq for local messages) were not being
4948 preserved with the message after it was received.
4950 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4951 "next input=" as part of the text comprising the host identity instead of
4952 the correct text. The code was using the same buffer for two different
4953 strings. However, depending on which order the printing function evaluated
4954 its arguments, the bug did not always show up. Under Linux, for example, my
4955 test suite worked just fine.
4957 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4958 doesn't work with some older versions of Perl. It has been changed to "my",
4959 which in any case is probably the better facility to use.
4961 15. A really picky compiler found some instances of statements for creating
4962 error messages that either had too many or two few arguments for the format
4965 16. The size of the buffer for calls to the DNS resolver has been increased
4966 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4967 for addresses that have a lot of PTR records. This alleviates a problem; it
4968 does not fully solve it.
4970 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4971 buffer now truncates the list and logs the incident, which is the same
4972 action as happens when Exim is looking up a host name and its aliases.
4973 Previously in this situation something unpredictable would happen;
4974 sometimes it was "internal error: store_reset failed".
4976 18. If a server dropped the connection unexpectedly when an Exim client was
4977 using GnuTLS and trying to read a response, the client delivery process
4978 crashed while trying to generate an error log message.
4980 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4981 string, for example:
4983 warn message = H1: something\nH2: something
4985 the text was added as a single header line from Exim's point of view
4986 though it ended up OK in the delivered message. However, searching for the
4987 second and subsequent header lines using $h_h2: did not work. This has been
4988 fixed. Similarly, if a system filter added multiple headers in this way,
4989 the routers could not see them.
4991 20. Expanded the error message when iplsearch is called with an invalid key to
4992 suggest using net-iplsearch in a host list.
4994 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4995 ACLs are no longer actually imposed (and a message to that effect is
4998 22. If a "gecos" field in a passwd entry contained escaped characters, in
4999 particular, if it contained a \" sequence, Exim got it wrong when building
5000 a From: or a Sender: header from that name. A second bug also caused
5001 incorrect handling when an unquoted " was present following a character
5002 that needed quoting.
5004 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5005 was not being matched caselessly.
5007 24. Arranged for all hyphens in the exim.8 source to be escaped with
5010 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5011 properly. Recipient callout cache records were still being keyed to include
5012 the sender, even when use_sender was set false. This led to far more
5013 callouts that were necessary. The sender is no longer included in the key
5014 when use_sender is false.
5016 26. Added "control = submission" modifier to ACLs.
5018 27. Added the ${base62d: operator to decode base 62 numbers.
5020 28. dnsdb lookups can now access SRV records.
5022 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5023 the configuration file.
5025 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5026 (-v) mode. This makes the output for a verbose queue run more intelligible.
5028 31. Added a use_postmaster feature to recipient callouts.
5030 32. Added the $body_zerocount variable, containing the number of binary zero
5031 bytes in the message body.
5033 33. The time of last modification of the "new" subdirectory is now used as the
5034 "mailbox time last read" when there is a quota error for a maildir
5037 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5039 35. Added +ignore_unknown as a special item in host lists.
5041 36. Code for decoding IPv6 addresses in host lists is now included, even if
5042 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5043 address was recognized as an IP address, but was then not correctly decoded
5044 into binary, causing unexpected and incorrect effects when compared with
5051 1. Very minor rewording of debugging text in manualroute to say "list of
5052 hosts" instead of "hostlist".
5054 2. If verify=header_syntax was set, and a header line with an unqualified
5055 address (no domain) and a large number of spaces between the end of the
5056 name and the colon was received, the reception process suffered a buffer
5057 overflow, and (when I tested it) crashed. This was caused by some obsolete
5058 code that should have been removed. The fix is to remove it!
5060 3. When running in the test harness, delay a bit after writing a bounce
5061 message to get a bit more predictability in the log output.
5063 4. Added a call to search_tidyup() just before forking a reception process. In
5064 theory, someone could use a lookup in the expansion of smtp_accept_max_
5065 per_host which, without the tidyup, could leave open a database connection.
5067 5. Added the variables $recipient_data and $sender_data which get set from a
5068 lookup success in an ACL "recipients" or "senders" condition, or a router
5069 "senders" option, similar to $domain_data and $local_part_data.
5071 6. Moved the writing of debug_print from before to after the "senders" test
5074 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5075 problems for message scanning, either using a data ACL, or using
5076 local_scan() because the Received: header was not generated till after they
5077 were called (in order to set the time as the time of reception completion).
5078 I have revised the way this works. The header is now generated after the
5079 body is received, but before the ACL or local_scan() are called. After they
5080 are run, the timestamp in the header is updated.
5086 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5087 before starting a queue runner without re-exec. This happened only when
5088 deliver_drop_privilege was set or when the Exim user was set to root. The
5089 effect of the bug was that timeouts during subsequent deliveries caused
5090 crashes instead of being properly handled. The handler is now left at its
5091 default (and expected) setting.
5093 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5094 message, again when deliver_drop_privilege is set or Exim is run as root.
5095 The bug described in (1) was not present in this case, but the tidying up
5096 of the other signals was missing. I have made the two cases consistent.
5098 3. The ignore_target_hosts setting on a manualroute router was being ignored
5099 for hosts that were looked up using the /MX notation.
5101 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5104 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5105 operated on the sender address. After changing the $sender_address to <>
5106 for the sender address verify, Exim was re-instated it as the original
5107 (before rewriting) address, but remembering that it had rewritten it, so it
5108 wasn't rewriting it again. This bug also had the effect of breaking the
5109 sender address verification caching when the sender address was rewritten.
5111 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5112 This has been changed so that if the ip literal address matches
5113 ignore_target_hosts, the router declines.
5115 7. Added expansion conditions match_domain, match_address, and match_local_
5116 part (NOT match_host).
5118 8. The placeholder for the Received: header didn't have a length field set.
5120 9. Added code to Exim itself and to exim_lock to test for a specific race
5121 condition that could lead to file corruption when using MBX delivery. The
5122 issue is with the lockfile that is created in /tmp. If this file is removed
5123 after a process has opened it but before that process has acquired a lock,
5124 there is the potential for a second process to recreate the file and also
5125 acquire a lock. This could lead to two Exim processes writing to the file
5126 at the same time. The added code performs the same test as UW imapd; it
5127 checks after acquiring the lock that its file descriptor still refers to
5128 the same named file.
5130 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5131 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5134 11. Added the smtp_active_hostname option. If used, this will typically be made
5135 to depend on the incoming interface address. Because $interface_address is
5136 not set up until the daemon has forked a reception process, error responses
5137 that can happen earlier (such as "too many connections") no longer contain
5140 12. If an expansion in a condition on a "warn" statement fails because a lookup
5141 defers, the "warn" statement is abandoned, and the next ACL statement is
5142 processed. Previously this caused the whole ACL to be aborted.
5144 13. Added the iplsearch lookup type.
5146 14. Added ident_timeout as a log selector.
5148 15. Added tls_certificate_verified as a log selector.
5150 16. Added a global option tls_require_ciphers (compare the smtp transport
5151 option of the same name). This controls incoming TLS connections.
5153 17. I finally figured out how to make tls_require_ciphers do a similar thing
5154 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5155 before starting the TLS session.
5157 18. Tabs are now shown as \t in -bP output.
5159 19. If the log selector return_path_on_delivery was set, Exim crashed when
5160 bouncing a message because it had too many Received: header lines.
5162 20. If two routers both had headers_remove settings, and the first one included
5163 a superfluous trailing colon, the final name in the first list and the
5164 first name in the second list were incorrectly joined into one item (with a
5165 colon in the middle).
5171 1. Added -C and -D options to the exinext utility, mainly to make it easier
5172 to include in the automated testing, but these could be helpful when
5173 multiple configurations are in use.
5175 2. The exinext utility was not formatting the output nicely when there was
5176 an alternate port involved in the retry record key, nor when there was a
5177 message id as well (for retries that were specific to a specific message
5178 and a specific host). It was also confused by IPv6 addresses, because of
5179 the additional colons they contain. I have fixed the IPv4 problem, and
5180 patched it up to do a reasonable job for IPv6.
5182 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5183 delivery, the log line now contains "pipelined" if PIPELINING was used.
5185 4. An SMTP transport process used to panic and die if the bind() call to set
5186 an explicit outgoing interface failed. This has been changed; it is now
5187 treated in the same way as a connect() failure.
5189 5. A reference to $sender_host_name in the part of a conditional expansion
5190 that was being skipped was still causing a DNS lookup. This no longer
5193 6. The def: expansion condition was not recognizing references to header lines
5194 that used bh_ and bheader_.
5196 7. Added the _cache feature to named lists.
5198 8. The code for checking quota_filecount in the appendfile transport was
5199 allowing one more file than it should have been.
5201 9. For compatibility with Sendmail, the command line option
5209 and sets the incoming protocol and host name (for trusted callers). The
5210 host name and its colon can be omitted when only the protocol is to be set.
5211 Note the Exim already has two private options, -pd and -ps, that refer to
5212 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5213 "s", but I don't think that's a major issue.
5215 10. A number of refactoring changes to the code, none of which should affect
5218 (a) The number of logging options was getting close to filling up the
5219 32-bit word that was used as a bit map. I have split them into two classes:
5220 those that are passed in the argument to log_write(), and those that are
5221 only ever tested independently outside of that function. These are now in
5222 separate 32-bit words, so there is plenty of room for expansion again.
5223 There is no change in the user interface or the logging behaviour.
5225 (b) When building, for example, log lines, the code previously used a
5226 macro that called string_cat() twice, in order to add two strings. This is
5227 not really sufficiently general. Furthermore, there was one instance where
5228 it was actually wrong because one of the argument was used twice, and in
5229 one call a function was used. (As it happened, calling the function twice
5230 did not affect the overall behaviour.) The macro has been replaced by a
5231 function that can join an arbitrary number of extra strings onto a growing
5234 (c) The code for expansion conditions now uses a table and a binary chop
5235 instead of a serial search (which was left over from when there were very
5236 few conditions). Also, it now recognizes conditions like "pam" even when
5237 the relevant support is not compiled in: a suitably worded error message is
5238 given if an attempt is made to use such a condition.
5240 11. Added ${time_interval:xxxxx}.
5242 12. A bug was causing one of the ddress fields not to be passed back correctly
5243 from remote delivery subprocesses. The field in question was not being
5244 subsequently used, so this caused to problems in practice.
5246 13. Added new log selectors queue_time and deliver_time.
5248 14. Might have fixed a bug in maildirsizefile handling that threw up
5249 "unexpected character" debug warnings, and recalculated the data
5250 unnecessarily. In any case, I expanded the warning message to give more
5253 15. Added the message "Restricted characters in address" to the statements in
5254 the default ACL that block characters like @ and % in local parts.
5256 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5257 Three changes have been made:
5259 (a) There was a serious bug; a negative response to MAIL caused the whole
5260 recipient domain to be cached as invalid, thereby blocking all messages
5261 to all local parts at the same domain, from all senders. This bug has
5262 been fixed. The domain is no longer cached after a negative response to
5263 MAIL if the sender used is not empty.
5265 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5268 (c) A new callout option, "use_sender" has been added for people who want
5269 the modified behaviour.
5275 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5278 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5279 indeed breaks things for older releases.
5281 3. Added additional logging to the case where there is a problem reading data
5282 from a filter that is running in a subprocess using a pipe, in order to
5283 try to track down a specific problem.
5285 4. Testing facility fudge: when running in the test harness and attempting
5286 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5287 getting "No route to host". Convert this to a timeout.
5289 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5292 6. Some OS don't have socklen_t but use size_t instead. This affects the
5293 fifth argument of getsockopt() amongst other things. This is now
5294 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5295 can be set for individual OS. I have set it for SunOS5, OSF1, and
5296 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5297 some earlier ones do not.
5299 7. Change 4.30/15 was not doing the test caselessly.
5301 8. The standard form for an IPv6 address literal was being rejected by address
5302 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5303 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5304 this, as well as the form without the "IPv6" on the front (but only when
5305 address literals are enabled, of course).
5307 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5309 10. Exim crashed if a message with an empty sender address specified by -f
5310 encountered a router with an errors_to setting. This could be provoked only
5311 by a command such as
5315 where an empty string was supplied; "<>" did not hit this bug.
5317 11. Installed PCRE release 4.5.
5319 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5320 remained set. It is now erased.
5322 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5323 times from message ids (which are base 36 rather than the normal 62).
5325 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5326 were being counted as actual protocol errors, and logged if the log
5327 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5328 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5329 and DATA following a set of rejected RCPTs do not count as protocol errors.
5330 In other words, Exim assumes they were pipelined, though this may not
5331 actually be the case. Of course, in all cases the client gets an
5332 appropriate error code.
5334 15. If a lookup fails in an ACL condition, a message about the failure may
5335 be available; it is used if testing the ACL cannot continue, because most
5336 such messages specify what the cause of the deferral is. However, some
5337 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5338 that caused an old message to be retained and used if a later statement
5339 caused a defer, replacing the real cause of the deferral.
5341 16. If an IP address had so many PTR records that the DNS lookup buffer
5342 was not large enough to hold them, Exim could crash while trying to process
5343 the truncated data. It now detects and logs this case.
5345 17. Further to 4.21/58, another change has been made: if (and only if) the
5346 first line of a message (the first header line) ends with CRLF, a bare LF
5347 in a subsequent header line has a space inserted after it, so as not to
5348 terminate the header.
5350 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5351 unnecessarily, used atoi() instead of strtol(), and didn't check the
5352 termination when getting file sizes from file names by regex.
5354 19. Completely re-implemented the support for maildirsize files, in the light
5355 of a number of problems with the previous contributed implementation
5356 (4.30/29). In particular:
5358 . If the quota is zero, the maildirsize file is maintained, but no quota is
5361 . If the maildir directory does not exist, it is created before any attempt
5362 to write a maildirsize file.
5364 . The quota value in the file is just a cache; if the quota is changed in
5365 the transport, the new value overrides.
5367 . A regular expression is available for excluding directories from the
5370 20. The autoreply transport checks the characters in options that define the
5371 message's headers; it allows continued headers, but it was checking with
5372 isspace() after an embedded newline instead of explicitly looking for a
5375 21. If all the "regular" hosts to which an address was routed had passed their
5376 expiry times, and had not reached their retry times, the address was
5377 bounced, even if fallback hosts were defined. Now Exim should go on to try
5380 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5381 equivalent code in the SMTP transport. Some hosts send humungous responses
5382 to HELO/EHLO, more than 1024 it seems.
5384 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5385 gives compiler warnings in some environments. I've now done it "properly",
5388 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5389 (because of problems with the built-in one) was declared to return uschar *
5390 instead of char *, causing compiler failure.
5392 25. Fixed a file descriptor leak when processing alias/forward files.
5394 26. Fixed a minor format string issue in dbfn.c.
5396 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5398 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5399 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5402 29. When a sender address is verified, it is cached, to save repeating the test
5403 when there is more than one recipient in a message. However, when the
5404 verification involves a callout, it is possible for different callout
5405 options to be set for different recipients. It is too complicated to keep
5406 track of this in the cache, so now Exim always runs a verification when a
5407 callout is required, relying on the callout cache for the optimization.
5408 The overhead is duplication of the address routing, but this should not be
5411 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5412 to be verified with callout=postmaster, and the main callout worked but the
5413 postmaster check failed, the verification correctly failed. However, if a
5414 subsequent RCPT command asked for sender verification *without* the
5415 postmaster check, incorrect caching caused this verification also to fail,
5418 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5419 it was not caching the DNS options (qualify_single, search_parents) that
5420 were used when the lookup failed. A subsequent lookup with different
5421 options therefore always gave the same answer, though there were cases
5422 where it should not have. (Example: a "domains = !$mx_any" option on a
5423 dnslookup router: the "domains" option is always processed without any
5424 widening, but the router might have qualify_single set.) Now Exim uses the
5425 cached value only when the same options are set.
5427 32. Added John Jetmore's "exipick" utility to the distribution.
5429 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5430 than a timeout (e.g. a certificate is required, and is not provided), an
5431 Exim server now closes the connection immediately. Previously it waited for
5432 the client to close - but if the client is SSL, it seems that they each
5433 wait for each other, leading to a delay before one of them times out.
5435 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5436 maintained 0.8.x compatibility because I don't think many are using it, and
5437 it is clearly obsolete.
5439 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5442 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5443 client certificate was expired. A simple patch fixes this, though I don't
5444 understand the full logic of why the verify callback is called multiple
5447 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5448 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5449 which causes problems with some clients (such as the Certicom SSL Plus
5450 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5451 disables the coutermeasure allowing Eudora to connect."
5453 38. Exim was not checking that a write() to a log file succeeded. This could
5454 lead to Bad Things if a log got too big, in particular if it hit a file
5455 size limit. Exim now panics and dies if it cannot write to a log file, just
5456 as it does if it cannot open a log file.
5458 39. Modified OS/Makefile-Linux so that it now contains
5460 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5462 The two -D definitions ensure that Exim is compiled with large file
5463 support, which makes it possible to handle log files that are bigger than
5466 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5467 instance) a domain was checked against a named list that involved a lookup,
5468 causing $domain_data to be set, then another domain was checked against the
5469 same list, then the first domain was re-checked, the value of $domain_data
5470 after the final check could be wrong. In particular, if the second check
5471 failed, it could be set empty. This bug probably also applied to
5474 41. The strip_trailing_dot option was not being applied to the address given
5475 with the -f command-line option.
5477 42. The code for reading a message's header from the spool was incrementing
5478 $received_count, but never initializing it. This meant that the value was
5479 incorrect (doubled) while delivering a message in the same process in which
5480 it was received. In the most common configuration of Exim, this never
5481 happens - a fresh exec is done - but it can happen when
5482 deliver_drop_privilege is set.
5484 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5485 it now includes up to 150 characters of the unexpected data in the log
5488 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5489 and building data strings. The size of both of these buffers was 10 000
5490 bytes - far larger than anybody would *ever* want, thought I. Needless to
5491 say, somebody hit the limit. I have increased the maximum line length to
5492 20 000 and the maximum data length of concatenated lines to 100 000. I have
5493 also fixed two bugs, because there was no checking on these buffers. Tsk,
5494 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5495 buffer is too small.
5497 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5498 lsearch lookups. Now it does.
5500 46. When parsing a route_list item in a manualroute router, a fixed-length
5501 buffer was used for the list of hosts. I made this 1024 bytes long,
5502 thinking that nobody would ever have a list of hosts that long. Wrong.
5503 Somebody had a whole pile of complicated expansion conditions, and the
5504 string was silently truncated, leading to an expansion error. It turns out
5505 that it is easier to change to an unlimited length (owing to other changes
5506 that have happened since this code was originally written) than to build
5507 structure for giving a limitation error. The length of the item that
5508 expands into the list of hosts is now unlimited.
5510 47. The lsearch lookup could not handle data where the length of text line was
5511 more than 4095 characters. Such lines were truncated, leading to shortened
5512 data being returned. It should now handle lines of any length.
5514 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5515 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5518 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5519 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5520 no longer happens in comment lines. A list of replacements is now placed
5521 at the head of all of the source files, except those whose only change is
5522 to replace PERL_COMMAND in the very first #! line.
5524 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5525 messages on the queue, with a bottom-up merge sort, using code contributed
5526 by Michael Haardt. This should make operations like -bp somewhat faster on
5527 large queues. It won't affect queue runners, except when queue_run_in_order
5530 51. Installed eximstats 1.31 in the distribution.
5532 52. Added support for SRV lookups to the dnslookup router.
5534 53. If an ACL referred to $message_body or $message_body_end, the value was not
5535 reset for any messages that followed in the same SMTP session.
5537 54. The store-handling optimization for building very long strings was not
5538 differentiating between the different store pools. I don't think this
5539 actually made any difference in practice, but I've tidied it.
5541 55. While running the routers to verify a sender address, $sender_address
5542 was still set to the sender address. This is wrong, because when routing to
5543 send a bounce to the sender, it would be empty. Therefore, I have changed
5544 it so that, while verifying a sender address, $sender_address is set to <>.
5545 (There is no change to what happens when verifying a recipient address.)
5547 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5548 target A or AAAA records (if not already returned) without resetting the
5549 qualify_single or search_parents options of the DNS resolver. These are
5550 inappropriate in this case because the targets of MX and SRV records must
5551 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5552 target that, when qualified, matched something in the local domain. These
5553 two options are now turned off when doing these lookups.
5555 57. It seems that at least some releases of Reiserfs (which does not have the
5556 concept of a fixed number of inodes) returns zero and not -1 for the
5557 number of available inodes. This interacted badly with check_spool_inodes,
5558 which assumed that -1 was the "no such thing" setting. What I have done is
5559 to check that the total number of inodes is greater than zero before doing
5560 the test of how many are available.
5562 58. When a "warn" ACL statement has a log_message modifier, the message is
5563 remembered, and not repeated. This is to avoid a lot of repetition when a
5564 message has many recipients that cause the same warning to be written.
5565 Howewer, Exim was preserving the list of already written lines for an
5566 entire SMTP session, which doesn't seem right. The memory is now reset if a
5567 new message is started.
5569 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5570 debugging output unless log_rewrite was also set.
5572 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5573 of (int)(handle) when we know that handle contains (void *)(-1).
5575 61. The Exim daemon panic-logs an error return when it closes the incoming
5576 connection. However "connection reset by peer" seems to be common, and
5577 isn't really an error worthy of noting specially, so that particular error
5580 62. When Exim is trying to find all the local interfaces, it used to panic and
5581 die if the ioctl to get the interface flags failed. However, it seems that
5582 on at least one OS (Solaris 9) it is possible to have an interface that is
5583 included in the list of interfaces, but for which you get a failure error
5584 for this call. This happens when the interface is not "plumbed" into a
5585 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5586 failure of the "get flags" call assumes that the interface is down.
5588 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5589 makes life easier for people who are doing arithmetic on fields extracted
5590 from dates, where you often get leading zeros that should not be
5591 interpreted as octal.
5593 64. Added qualify_domain to the redirect router, to override the global
5596 65. If a pathologically long header line contained very many addresses (the
5597 report of this problem mentioned 10 000) and each of them was rewritten,
5598 Exim could use up a very large amount of memory. (It kept on making new
5599 copies of the header line as it rewrote, and never released the old ones.)
5600 At the expense of a bit more processing, the header rewriting function has
5601 been changed so that it no longer eats memory in this way.
5603 66. The generation of the Received: header has been moved from the time that a
5604 message starts to be received, to the time that it finishes. The timestamp
5605 in the Received: header should now be very close to that of the <= log
5606 line. There are two side-effects of this change:
5608 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5609 logged header lines no longer include the local Received: line, because
5610 it has not yet been created. The same applies to a copy of the message
5611 that is returned to a non-SMTP sender when a message is rejected.
5613 (b) When a filter file is tested using -bf, no additional Received: header
5614 is added to the test message. After some thought, I decided that this
5617 This change does not affect the value of $received_for. It is still set
5618 after address rewriting, but before local_scan() is called.
5620 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5622 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5623 gave an unhelpful panic error message, and a defer error. I have managed to
5624 change this behaviour so that it now rejects any supplied certificate,
5625 which seems right, as the list of acceptable certificates is empty.
5627 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5628 gave an unhelpful defer error. I have not managed to make this reject any
5629 supplied certificates, but the error message it gives is "no certificate
5630 supplied", which is not helpful.
5632 70. exigrep's output now also includes lines that are not associated with any
5633 message, but which match the given pattern. Implemented by a patch from
5634 Martin Sluka, which also tidied up the Perl a bit.
5636 71. Recipient callout verification, like sender verification, was using <> in
5637 the MAIL FROM command. This isn't really the right thing, since the actual
5638 sender may affect whether the remote host accepts the recipient or not. I
5639 have changed it to use the actual sender in the callout; this means that
5640 the cache record is now keyed on a recipient/sender pair, not just the
5641 recipient address. There doesn't seem to be a real danger of callout loops,
5642 since a callout by the remote host to check the sender would use <>.
5643 [SEE ABOVE: changed after hitting problems.]
5645 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5646 temporary errors. However, in the case of such a code being given after
5647 the end of a data transmission (i.e. after ".") Exim was failing to write
5648 a retry record for the message. (Yes, there was some broken host that was
5649 actually sending 8xx at this point.)
5651 73. An unknown lookup type in a host list could cause Exim to panic-die when
5652 the list was checked. (An example that provoked this was putting <; in the
5653 middle of a list instead of at the start.) If this happened during a DATA
5654 ACL check, a -D file could be left lying around. This kind of configuration
5655 error no longer causes Exim to die; instead it causes a defer errror. The
5656 incident is still logged to the main and panic logs.
5658 74. Buglet left over from Exim 3 conversion. The message "too many messages
5659 in one connection" was written to the rejectlog but not the mainlog, except
5660 when address rewriting (yes!) was being logged.
5662 75. Added write_rejectlog option.
5664 76. When a system filter was run not as root (that is, when system_filter_user
5665 was set), the values of the $n variables were not being returned to the
5666 main process; thus, they were not subsequently available in the $sn
5669 77. Added +return_path_on_delivery log selector.
5671 78. A connection timeout was being treated differently from recipients deferred
5672 when testing hosts_max_try with a message that was older than the host's
5673 retry timeout. (The host should not be counted, thus allowing all hosts to
5674 be tried at least once before bouncing.) This may have been the cause of an
5675 occasionally reported bug whereby a message would remain on the queue
5676 longer than the retry timeout, but would be bounced if a delivery was
5677 forced. I say "may" because I never totally pinned down the problem;
5678 setting up timeout/retry tests is difficult. See also the next item.
5680 79. The ultimate address timeout was not being applied to errors that involved
5681 a combination of host plus message (for example, a timeout on a MAIL
5682 command). When an address resolved to a number of possible hosts, and they
5683 were not all tried for each delivery (e.g. because of hosts_max_try), a
5684 message could remain on the queue longer than the retry timeout.
5686 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5689 81. Fixed an obscure SMTP outgoing bug which required at least the following
5690 conditions: (a) there was another message waiting for the same server;
5691 (b) the server returned 5xx to all RCPT commands in the first message so
5692 that the message was not completed; (c) the server dropped the connection
5693 or gave a negative response to the RSET that Exim sends to abort the
5694 transaction. The observed case was a dropped connection after DATA that had
5695 been sent in pipelining mode. That is, the server had advertised PIPELINING
5696 but was not implementing it correctly. The effect of the bug was incorrect
5697 behaviour, such as trying another host, and this could lead to a crash.
5703 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5704 and daemon.c were passed as pointers to ints; they should have been
5705 pointers to socklen_t variables (which are typically unsigned ints).
5707 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5710 3. Fixed a really odd bug that affected only the testing scheme; patching a
5711 certain fixed string in the binary changed the value of another string that
5712 happened to be identical to the end of the original first string.
5714 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5715 name", it returns that address as the IP address. On some operating
5716 systems (e.g. Solaris), it also passes back the IP address string as the
5717 "host name". However, on others (e.g. Linux), it passes back an empty
5718 string. Exim wasn't checking for this, and was changing the host name to an
5719 empty string, assuming it had been canonicized.
5721 5. Although rare, it is permitted to have more than one PTR record for a given
5722 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5723 all the names associated with an address, because they do in Solaris.
5724 However, it seems that they do not in Linux for data that comes from the
5725 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5726 I found this out when I moved to a new Linux workstation and tried to run
5727 the Exim test suite.
5729 To get round this problem I have changed the code so that it now does its
5730 own call to the DNS to look up PTR records when searching for a host name.
5731 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5732 addresses that are only in /etc/hosts are still found.
5734 This behaviour is, however, controlled by an option called host_lookup_
5735 order, which defaults to "bydns:byaddr". If people want to use the other
5736 order, or indeed, just use one or the other means of lookup, they can
5737 specify it in this variable.
5739 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5740 some operating systems, this comes back from gethostbyaddr() as an empty
5741 string, and this is what Exim used to test for. However, it seems that in
5742 other systems, "." is yielded. Exim now tests for this case too.
5744 7. The values of check_spool_space and check_log_space are now held internally
5745 as a number of kilobytes instead of an absolute number of bytes. If a
5746 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5747 kilobyte. This means that much larger values can be stored.
5749 8. Exim monitor: an attempt to get the action menu when not actually pointing
5750 at a message produces an empty menu entitled "No message selected". This
5751 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5752 no entries in it ("Shell widget menu has zero width and/or height"). So I
5753 have added a single, blank menu entry in this case.
5755 9. Added ${quote_local_part.
5757 10. MIME decoding is now applied to the contents of Subject: header lines when
5760 11. Now that a reference to $sender_host_address automatically causes a reverse
5761 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5762 host lookup before query-style lookups in lists that might use this
5763 variable. This has therefore been abolished, and the "net-" prefix is no
5764 longer necessary for query-style lookups.
5766 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5767 have been a typo for LFLAGS, so it has been changed.
5769 13. The install script calls Exim with "-C /dev/null" in order to find the
5770 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5771 to be output. Howeve, since Exim outputs its version number before the
5772 error, it didn't break the script. It just looked ugly. I fixed this by
5773 always allowing "-C /dev/null" if the caller is root.
5775 14. Ignore overlarge ACL variable number when reading spool file - insurance
5776 against a later release with more variables having written the file.
5778 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5779 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5780 this, as well as the form without the "IPv6" on the front.
5782 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5783 OS/Makefile-Darwin file.
5785 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5786 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5788 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5789 variable. However, some OS use ints for inodes. I've added cast to long int
5790 to get rid of the compiler warning.
5792 19. I had forgotten to lock out "/../" in configuration file names when
5793 ALT_CONFIG_PREFIX was set.
5795 20. Routers used for verification do not need to specify transports. However,
5796 if such a router generated a host list, and callout was configured, Exim
5797 crashed, because it could not find a port number from the (non-existent)
5798 transport. It now assumes port 25 in this circumstance.
5800 21. Added the -t option to exigrep.
5802 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5803 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5804 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5805 them is set, LOOKUP_LSEARCH is forced.
5807 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5809 24. Added sender and host information to the "rejected by local_scan()" log
5810 line; previously there was no indication of these.
5812 25. Added .include_if_exists.
5814 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5815 Linux. It turns out that not all file systems support this. Apparently some
5816 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5817 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5818 not supported on the file descriptor, is now ignored when Exim is trying to
5819 sync a directory. This applies only to Linux.
5821 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5823 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5824 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5827 29. Added support for maildirsize files from supplied patch (modified a bit).
5829 30. The use of :fail: followed by an empty string could lead Exim to respond to
5830 sender verification failures with (e.g.):
5832 550 Verification failed for <xxx>
5833 550 Sender verify failed
5835 where the first response line was missing the '-' that indicates it is not
5836 the final line of the response.
5838 31. The loop for finding the name of the user that called Exim had a hardwired
5839 limit of 10; it now uses the value of finduser_retries, which is used for
5840 all other user lookups.
5842 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5845 33. Exim was neglecting to zero errno before one call of strtol() when
5846 expanding a string and expecting an integer value. On some systems this
5847 resulted in spurious "integer overflow" errors. Also, it was casting the
5848 result into an int without checking.
5850 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5851 did not work. The code looks as if it has *never* worked, though it appears
5852 to have been documented since at least releast 1.62. I have made it work.
5854 35. The "timeout_DNS" error in retry rules, also documented since at least
5855 1.62, also never worked. As it isn't clear exactly what this means, and
5856 clearly it isn't a major issue, I have abolished the feature by treating it
5857 as "timeout", and writing a warning to the main and panic logs.
5859 36. The display of retry rules for -brt wasn't always showing the error code
5862 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5863 timeout_connect_A, timeout_connect_MX.
5865 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5866 to the empty sender.
5868 39. The daemon was not analysing the content of -oX till after it had closed
5869 stderr and disconnected from the controlling terminal. This meant that any
5870 syntax errors were only noted on the panic log, and the return code from
5871 the command was 0. By re-arranging the code a little, I've made the
5872 decoding happen first, so such errors now appear on stderr, and the return
5873 code is 1. However, the actual setting up of the sockets still happens in
5874 the disconnected process, so errors there are still only recorded on the
5877 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5878 connections (as happens on some IP stacks) was logged at start up time as
5879 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5880 it from "IPv6 and IPv4", which means that two separate sockets are being
5883 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5884 says whether AF_INET or AF_INET6 was passed as an argument.
5886 42. Exiwhat output was messed up when time zones were included in log
5889 43. Exiwhat now gives more information about the daemon's listening ports,
5890 and whether -tls-on-connect was used.
5892 44. The "port" option of the smtp transport is now expanded.
5894 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5895 silently ignored. Now an error message is written to the main and panic
5898 46. There's a new ACL modifier called "logwrite" which writes to a log file
5899 as soon as it is encountered.
5901 47. Added $local_user_uid and $local_user_gid at routing time.
5903 48. Exim crashed when trying to verify a sender address that was being
5906 49. Exim was recognizing only a space character after ".include". It now also
5907 recognizes a tab character.
5909 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5910 extracting the relevant information from the specification. The man page no
5911 longer contains scrambled data for the -d option, and I've added a section
5912 at the front about calling Exim under different names.
5914 51. Added "extra_headers" argument to the "mail" command in filter files.
5916 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5919 53. Installed eximstats 1.29.
5921 54. Added transport_filter_timeout as a generic transport option.
5923 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5924 Cc: header lines. This was required by RFC 822, but it not required by RFC
5927 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5928 incoming messages that did not have them. Now it does so only if the
5929 message originates locally, that is, if there is no associated remote host
5930 address. When Resent- header lines are present, this applies to the Resent-
5931 lines rather than the non-Resent- lines.
5933 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5934 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5936 58. Messages for configuration errors now include the name of the main
5937 configuration file - useful now that there may be more than one file in a
5938 list (.included file names were always shown).
5940 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5941 for those rare installations that do not start the daemon as root or run it
5942 setuid root. I've cut out the call to initgroups() if the daemon is not
5945 60. The Exim user and group can now be bound into the binary as text strings
5946 that are looked up at the start of Exim's processing.
5948 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5950 62. Added $mailstore_basename variable.
5952 63. Installed patch to sieve.c from Michael Haardt.
5954 64. When Exim failed to open the panic log after failing to open the main log,
5955 the original message it was trying to log was written to stderr and debug
5956 output, but if they were not available (the usual case in production), it
5957 was lost. Now it is written to syslog before the two lines that record the
5958 failures to open the logs.
5960 65. Users' Exim filters run in subprocesses under the user's uid. It is
5961 possible for a "deliver" command or an alias in a "personal" command to
5962 provoke an address rewrite. If logging of address rewriting is configured,
5963 this fails because the process is not running as root or exim. There may be
5964 a better way of dealing with this, but for the moment (because 4.30 needs
5965 to be released), I have disabled address rewrite logging when running a
5966 filter in a non-root, non-exim process.
5972 1. The buildconfig auxiliary program wasn't quoting the value set for
5973 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5974 not defined. This bug was masked in 4.22 by the effect that was fixed in
5977 2. Some messages that were rejected after a message id was allocated were
5978 shown as "incomplete" by exigrep. It no longer does this for messages that
5979 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5981 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5982 have allow_domain_literals set, the ID did not get logged in the <= line.
5983 Domain literals are now always recognized in Message-ID: header lines.
5985 4. The first argument for a ${extract expansion item is the key name or field
5986 number. Leading and trailing spaces in this item were not being ignored,
5987 causing some misleading effects.
5989 5. When deliver_drop_privilege was set, single queue runner processes started
5990 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5991 same command in the process it spins off) were not dropping privilege.
5993 6. When the daemon running as "exim" started a queue runner, it always
5994 re-executed Exim in the spun-off process. This is a waste of effort when
5995 deliver_drop_privilege is set. The new process now just calls the
5996 queue-runner function directly.
6002 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6005 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6006 ignored. Though the use of -oP was forcing the writing of a pid file, it
6007 was always written to the default place.
6009 3. If the message "no IP address found for host xxxx" is generated during
6010 incoming verification, it is now followed by identification of the incoming
6011 connection (so you can more easily find what provoked it).
6013 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6015 5. Added some features to "harden" Exim a bit more against certain attacks:
6017 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6018 be put in Local/Makefile. This is like the never_users runtime option,
6019 but it cannot be overridden. The default setting is "root".
6021 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6022 prefix string with which any file named in a -C command line option
6025 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6026 is retained for -C and -D only if the caller of Exim is root. Without
6027 it, the exim user may also use -C and -D and retain privilege.
6029 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6030 command line option is disabled.
6032 6. Macro names set by the -D option must start with an upper case letter, just
6033 like macro names defined in the configuration file.
6035 7. Added "dereference=" facility to LDAP.
6037 8. Two instances of the typo "uknown" in the source files are fixed.
6039 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6040 the Configure-Makefile script screwed up while processing it.
6042 10. Incorporated PCRE 4.4.
6044 11. The SMTP synchronization check was not operating right at the start of an
6045 SMTP session. For example, it could not catch a HELO sent before the client
6046 waited for the greeting. There is now a check for outstanding input at the
6047 point when the greeting is written. Because of the duplex, asynchronous
6048 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6049 way, but not yet received, when the check is performed.
6051 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6052 on TCP/IP sockets, because this apparently causes some broken clients to
6055 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6056 unchanged) from the Cygwin maintainer.
6058 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6059 was supported instead of testing for mbx. Effectively a typo.
6061 15. The spa authenticator server code was not checking that the input it
6062 received was valid base64.
6064 16. The debug output line for the "set" modifier in ACLs was not showing the
6065 name of the variable that was being set.
6067 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6069 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6070 lookup on its own. Something else was needed to trigger the lookup. For
6071 example, a match in host_lookup or the need for a host name in a host list.
6072 Now, if $sender_host_name is referenced and the host name has not yet been
6073 looked up, a lookup is performed. If the lookup fails, the variable remains
6074 empty, and $host_lookup_failed is set to "1".
6076 19. Added "eqi" as a case-independent comparison operator.
6078 20. The saslauthd authentication condition could segfault if neither service
6079 nor realm was specified.
6081 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6082 error message that was logged was misleading, and incoming SMTP
6083 connections were dropped. The message is now more accurate, and temporary
6084 errors are given to SMTP connections.
6086 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6087 (except RSET) with a 421 error, until QUIT is received. However, it was
6088 failing to send a response to QUIT.
6090 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6091 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6092 ACL is now run after the helo_try_verify_hosts code.
6094 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6095 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6096 case-independent, so other case variants are also recognized). Apparently
6097 some systems use these upper case variants.
6099 25. If more than two messages were waiting for the same host, and a transport
6100 filter was specified for the transport, Exim sent two messages over the
6101 same TCP/IP connection, and then failed with "socket operation on non-
6102 socket" when it tried to send the third.
6104 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6106 27. The extern definition of crypt16() in expand.c was not being excluded when
6107 the OS had its own crypt16() function.
6109 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6110 as a preferred synonym for return_size_limit, both as an option and as an
6113 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6115 30. Changed the default configuration ACL to relax the local part checking rule
6116 for addresses that are not in any local domains. For these addresses,
6117 slashes and pipe symbols are allowed within local parts, but the sequence
6118 /../ is explicitly forbidden.
6120 31. SPA server authentication was not clearing the challenge buffer before
6123 32. log_message in a "warn" ACL statement was writing to the reject log as
6124 well as to the main log, which contradicts the documentation and doesn't
6125 seem right (because no rejection is happening). So I have stopped it.
6127 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6128 However, I am unable to do any testing of this.
6130 34. Fixed an infelicity in the appendfile transport. When checking directories
6131 for a mailbox, to see if any needed to be created, it was accidentally
6132 using path names with one or more superfluous leading slashes; tracing
6133 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6135 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6136 contents are added to the log line that is written for every discarded
6137 recipient. (Previously a log_message setting was ignored.)
6139 36. The ${quote: operator now quotes the string if it is empty.
6141 37. The install script runs exim in order to find its version number. If for
6142 some reason other than non-existence or emptiness, which it checks, it
6143 could not run './exim', it was installing it with an empty version number,
6144 i.e. as "exim-". This error state is now caught, and the installation is
6147 38. An argument was missing from the function that creates an error message
6148 when Exim fails to connect to the socket for saslauthd authentication.
6149 This could cause Exim to crash, or give a corrupted message.
6151 39. Added isip, isip4, and isip6 to ${if conditions.
6153 40. The ACL variables $acl_xx are now saved with the message, and can be
6154 accessed later in routers, transports, and filters.
6156 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6157 strings in the file are not string-expanded.
6159 42. If a MAIL command specified a SIZE value that was too large to fit into an
6160 int variable, the check against message_size_limit failed. Such values are
6161 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6162 day this will have to be increased, but I don't think I want to be around
6163 when emails are that large.
6170 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6171 iconv() is not standard in FreeBSD.
6173 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6174 IPv6 enabled. The observed symptom was a segmentation fault on return from
6175 the function os_common_find_running_interfaces() in src/os.c.
6177 3. In the check_special_case() function in daemon.c I had used "errno" as an
6178 argument name, which causes warnings on some systems. This was basically a
6179 typo, since it was named "eno" in the comments!
6181 4. The code that waits for the clock to tick (at a resolution of some fraction
6182 of a second) so as to ensure message-id uniqueness was always waiting for
6183 at least one whole tick, when it could have waited for less. [This is
6184 almost certainly not relevant at current processor speeds, where it is
6185 unlikely to ever wait at all. But we try to future-proof.]
6187 5. The function that sleeps for a time interval that includes fractions of a
6188 second contained a race. It did not block SIGALRM between setting the
6189 timer, and suspending (a couple of lines later). If the interval was short
6190 and the sigsuspend() was delayed until after it had expired, the suspension
6191 never ended. On busy systems this could lead to processes getting stuck for
6194 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6195 process, before it forks any delivery processes. The open lookup caching
6196 mechanism meant that the open file or database connection was passed into
6197 the delivery process. The problem was that delivery processes always tidy
6198 up cached lookup data. This could cause a problem for the next delivery
6199 process started by the queue runner, because the external queue runner
6200 process does not know about the closure. So the next delivery process
6201 still has data in the lookup cache. In the case of a file lookup, there was
6202 no problem because closing a file descriptor in a subprocess doesn't affect
6203 the parent. However, if the lookup was caching a connection to a database,
6204 the connection was closed, and the second delivery process was likely to
6205 see errors such as "PGSQL: query failed: server closed the connection
6206 unexpectedly". The problem has been fixed by closing all cached lookups
6207 in a queue runner before running a delivery process.
6209 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6210 seem to have the "const" qualifier which it has on other OS. I've
6213 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6214 *of the same type* (client or server) with the same public name that an
6215 error should be diagnosed.
6217 9. When Exim looked up a host name for an IP address, but failed to find the
6218 original IP address when looking up the host name (a safety check), it
6219 output the message "<ip address> does not match any IP for NULL", which was
6220 confusing, to say the least. The bug was that the host name should have
6221 appeared instead of "NULL".
6223 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6224 user that is built into the binary, and the -C or -D options is used, root
6225 privilege is dropped before the configuration file is read. In addition,
6226 logging is switched to stderr instead of the normal log files. If the
6227 configuration then re-defines the Exim user, the unprivileged environment
6228 is probably not what is expected, so Exim logs a panic warning message (but
6231 However, if deliver_drop_privilege is set, the unprivileged state may well
6232 be exactly what is intended, so the warning has been cut out in that case,
6233 and Exim is allowed to try to write to its normal log files.
6239 1. smtp_return_error_details was not giving details for temporary sender
6240 or receiver verification errors.
6242 2. Diagnose a configuration error if two authenticators have the same public
6245 3. Exim used not to create the message log file for a message until the first
6246 delivery attempt. This could be confusing when incoming messages were held
6247 for policy or load reasons. The message log file is now created at the time
6248 the message is received, and an initial "Received" line is written to it.
6250 4. The automatically generated man page for command line options had a minor
6251 bug that caused no ill effects; however, a more serious problem was that
6252 the procedure for building the man page automatically didn't always
6253 operate. Consequently, release 4.20 contains an out-of-date version. This
6254 shouldn't happen again.
6256 5. When building Exim with embedded Perl support, the script that builds the
6257 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6258 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6260 6. The freeze_tell option was not being used for messages that were frozen on
6261 arrival, either by an ACL or by local_scan().
6263 7. Added the smtp_incomplete_transaction log selector.
6265 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6266 was accepting AUTH without a new EHLO.
6268 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6269 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6272 10. Logging of TCP/IP connections (when configured) now happens in the main
6273 daemon process instead of the child process, so that the TCP/IP connection
6274 count is more accurate (but it can never be perfect).
6276 11. The use of "drop" in a nested ACL was not being handled correctly in the
6277 outer ACL. Now, if condition failure induced by the nested "drop" causes
6278 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6279 or "require"), the connection is dropped.
6281 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6282 that yield "discard" can now be used with an "accept" or a "discard" verb,
6283 but an error is generated for any others (because I can't see a useful way
6284 to define what should happen).
6286 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6287 are now processed in the same way as lines in the Exim configuration file.
6288 In particular, continuation lines are supported.
6290 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6292 15. Added -ti meaning -t -i.
6294 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6295 domains, and warn by logging if others are found.
6297 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6298 returned from GIFCONF when Exim is trying to find the list of interfaces on
6299 a host. The code in os.c has been modified to copy each ifreq to an aligned
6300 structure in all cases.
6302 Also, in some cases, the returned ifreq's were being copied to a 'struct
6303 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6304 means the last couple of bytes of an IPv6 address could be chopped if the
6305 ifreq contained only a normal sockaddr (14 bytes storage).
6307 18. Named domain lists were not supported in the hosts_treat_as_local option.
6308 An entry such as +xxxx was not recognized, and was treated as a literal
6311 19. Ensure that header lines added by a DATA ACL are included in the reject log
6312 if the ACL subsequently rejects the message.
6314 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6315 MD5 (which is deprecated).
6317 21. When testing a filter file using -bf, Exim was writing a message when it
6318 took the sender from a "From " line in the message, but it was not doing so
6319 when it took $return_path from a Return-Path: header line. It now does.
6321 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6322 with a valid header line field name (a series of printing characters
6323 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6325 23. Changed "disc" in the source to "disk" to conform to the documentation and
6326 the book and for uniformity.
6328 24. Ignore Sendmail's -Ooption=value command line item.
6330 25. When execve() failed while trying to run a command in a pipe transport,
6331 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6332 could be confused with a return value of 69 from the command itself. This
6333 has been changed to 127, the value the shell returns if it is asked to run
6334 a non-existent command. The wording for the related log line suggests a
6335 non-existent command as the problem.
6337 26. If received_header_text expands to an empty string, do not add a Received:
6338 header line to the message. (Well, it adds a token one on the spool, but
6339 marks it "old" so that it doesn't get used or transmitted.)
6341 27. Installed eximstats 1.28 (addition of -nt option).
6343 28. There was no check for failure on the call to getsockname() in the daemon
6344 code. This can fail if there is a shortage of resources on the system, with
6345 ENOMEM, for example. A temporary error is now given on failure.
6347 29. Contrary to the C standard, it seems that in some environments, the
6348 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6349 program. Exim now does this explicitly; it affects the formatting of
6350 timestamps using strftime().
6352 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6353 complaints. I've now initialized all the variables, to avoid this.
6355 32. Header lines added by a system filter were not being "seen" during
6356 transport-time rewrites.
6358 33. The info_callback() function passed to OpenSSL is set up with type void
6359 (*)(SSL *, int, int), as described somewhere. However, when calling the
6360 function (actually a macro) that sets it up, the type void(*)() is
6361 expected. I've put in a cast to prevent warnings from picky compilers.
6363 34. If a DNS black list lookup found a CNAME record, but there were no A
6364 records associated with the domain it pointed at, Exim crashed.
6366 35. If a DNS black list lookup returned more than one A record, Exim ignored
6367 all but the first. It now scans all returned addresses if a particular IP
6368 value is being sought. In this situation, the contents of the
6369 $dnslist_value variable are a list of all the addresses, separated by a
6372 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6373 to accept a host name and all its aliases if the forward lookup for any of
6374 them yielded the IP address of the incoming connection. Now it accepts only
6375 those names whose forward lookup yields the correct IP address. Any other
6376 names are discarded. This closes a loophole whereby a rogue DNS
6377 administrator could create reverse DNS records to break through a
6378 wildcarded host restriction in an ACL.
6380 37. If a user filter or a system filter that ran in a subprocess used any of
6381 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6382 the wrong values were passed to the pipe command ($thisaddress had the
6383 value of $0, $0 had the value of $1, etc). This bug was introduced by
6384 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6386 38. Improved the line breaking for long SMTP error messages from ACLs.
6387 Previously, if there was no break point between 40 and 75 characters, Exim
6388 left the rest of the message alone. Two changes have been made: (a) I've
6389 reduced the minimum length to 35 characters; (b) if it can't find a break
6390 point between 35 and 75 characters, it looks ahead and uses the first one
6391 that it finds. This may give the occasional overlong line, but at least the
6392 remaining text gets split now.
6394 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6395 file descriptors might be low, and that setting 1000 would always raise it.
6396 It turns out that in some environments, the limit is already over 1000 and
6397 that lowering it causes trouble. So now Exim takes care not to decrease it.
6399 40. When delivering a message, the value of $return_path is set to $sender_
6400 address at the start of routing (routers may change the value). By an
6401 oversight, this default was not being set up when an address was tested by
6402 -bt or -bv, which affected the outcome if any router or filter referred to
6405 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6406 log only when "message" or "log_message" are set. However, if one of the
6407 conditions was an address verification, or a call to a nested ACL, the
6408 messages generated by the underlying test were being passed through. This
6409 no longer happens. The underlying message is available in $acl_verify_
6410 message for both "message" and "log_message" expansions, so it can be
6411 passed through if needed.
6413 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6414 new expansion $bh_ to give the encoded byte string without charset
6415 translation. Translation happens only if iconv() is available; HAVE_ICONV
6416 indicates this at build time. HEADERS_CHARSET gives the charset to
6417 translate to; headers_charset can change it in the configuration, and
6418 "headers charset" can change it in an individual filter file.
6420 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6421 that creates RFC 2047 encoded "words" labels them as that charset instead
6422 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6423 expansion operator; (ii) when Exim creates a From: line for a local
6424 message; (iii) when a header line is rewritten to include a "phrase" part.
6426 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6427 buggy, causing it to skip the first lines of messages whose message ID
6428 ended in 'D'. This would not have bitten before Exim release 4.14, because
6429 message IDs were unlikely to end in 'D' before then. The effect was to have
6430 incorrect size information for certain domains.
6432 45. #include "config.h" was missing at the start of the crypt16.c module. This
6433 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6436 46. If there was a timeout during a "random" callout check, Exim treated it as
6437 a failure of the random address, and carried on sending RSET and the real
6438 address. If the delay was just some slowness somewhere, the response to the
6439 original RCPT would be taken as a response to RSET and so on, causing
6440 mayhem of various kinds.
6442 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6443 when I implemented it. It didn't allow for the fact that some option values
6444 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6445 the right test for positive values.
6447 48. Domain names in DNS records are case-independent. Exim always looks them up
6448 in lower case. Some resolvers return domain names in exactly the case they
6449 appear in the zone file, that is, they may contain uppercase letters. Not
6450 all resolvers do this - some return always lower case. Exim was treating a
6451 change of case by a resolver as a change of domain, similar to a widening
6452 of a domain abbreviation. This triggered its re-routing code and so it was
6453 trying to route what was effectively the same domain again. This normally
6454 caused routing to fail (because the router wouldn't handle the domain
6455 twice). Now Exim checks for this case specially, and just changes the
6456 casing of the domain that it ultimately uses when it transmits the message
6459 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6462 50. If a filter generated a file delivery with a non-absolute name (possible if
6463 no home directory exists for the router), the forbid_file option was not
6466 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6467 the existing equality matching.
6469 52. Exim was using ints instead of ino_t variables in some places where it was
6470 dealing with inode numbers.
6472 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6473 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6474 called TMPDIR, and if it finds it is different, it changes its value.
6476 54. The smtp_printf() function is now made available to local_scan() so
6477 additional output lines can be written before returning. There is also an
6478 smtp_fflush() function to enable the detection of a dropped connection.
6479 The variables smtp_input and smtp_batched_input are exported to
6482 55. Changed the default runtime configuration: the message "Unknown user"
6483 has been removed from the ACL, and instead placed on the localuser router,
6484 using the cannot_route_message feature. This means that any verification
6485 failures that generate their own messages won't get overridden. Similarly,
6486 the "Unrouteable address" message that was in the ACL for unverifiable
6487 relay addresses has also been removed.
6489 56. Added hosts_avoid_esmtp to the smtp transport.
6491 57. The exicyclog script was not checking for the esoteric option
6492 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6493 will work only if exicyclog is run under the appropriate euid.
6495 58. Following a discussion on the list, the rules by which Exim recognises line
6496 endings on incoming messages have been changed. The -dropcr and drop_cr
6497 options are now no-ops, retained only for backwards compatibility. The
6498 following line terminators are recognized: LF CRLF CR. However, special
6499 processing applies to CR:
6501 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6502 nor a local message in the state where . is a terminator.
6504 (ii) If a bare CR is encountered in a header line, an extra space is added
6505 after the line terminator so as not to end the header. The reasoning
6506 behind this is that bare CRs in header lines are most likely either
6507 to be mistakes, or people trying to play silly games.
6509 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6510 was being incorrectly given as 18 bytes larger than it should have been.
6511 This is a VOB (very old bug).
6513 60. This may never have affected anything current, but just in case it has:
6514 When the local host is found other than at the start of a list of hosts,
6515 the local host, those with the same MX, and any that follow, are discarded.
6516 When the list in question was part of a longer list of hosts, the following
6517 hosts (not currently being processed) were also being discarded. This no
6518 longer happens. I'm not sure if this situation could ever has previously
6521 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6524 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6525 header. This is something that is recommended in a new Internet Draft, and
6526 is something that is documented as being done by Sendmail. There are two
6527 possible values. For messages generated by the autoreply transport, Exim
6530 Auto-Submitted: auto-replied
6532 whereas for all other generated messages (e.g. bounces) it adds
6534 Auto-Submitted: auto-generated
6536 63. The "personal" condition in filters now includes a test for the
6537 Auto-Submitted: header. If it contains the string "auto-" the message it
6538 not considered personal.
6540 64. Added rcpt_include_affixes as a generic transport option.
6542 65. Added queue_only_override (default true).
6544 66. Added the syslog_duplication option.
6546 67. If what should have been the first header line of a message consisted of
6547 a space followed by a colon, Exim was mis-interpreting it as a header line.
6548 It isn't of course - it is syntactically invalid and should therefore be
6549 treated as the start of the message body. The misbehaviour could have
6550 caused a number of strange effects, including loss of data in subsequent
6551 header lines, and spool format errors.
6553 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6554 client host had authenticated. This control can now be exercised by an ACL
6555 for more flexibility.
6557 69. By default, callouts do not happen when testing with -bh. There is now a
6558 variant, -bhc, which does actually run the callout code, including
6559 consulting and updating the callout cache.
6561 70. Added support for saslauthd authentication, courtesy of Alexander
6564 71. If statvfs() failed on the spool or log directories while checking their
6565 size for availability, Exim confusingly gave the error "space shortage".
6566 Furthermore, in debugging mode it crashed with a floating point exception.
6567 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6568 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6569 serious problem, Exim now writes to the main and panic logs when this
6570 happens, with details of the failure. It then refuses to accept the
6571 incoming message, giving the message "spool directory problem" or "log
6572 directory problem" with a 421 code for SMTP messages.
6574 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6575 0, 1, and 2 exist, because some OS complain for execs without them (see
6576 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6577 descriptors. However, the code omitted to check that the open succeeded,
6578 causing mysterious errors if for some reason the permissions on /dev/null
6579 got screwed. Now Exim writes a message to the main and panic logs, and
6580 bombs out if it can't open /dev/null.
6582 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6583 interact so that it is all more flexible. It is supposed to remain
6584 backwards compatible. Also added extra_local_interfaces.
6586 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6587 to bomb out with an assertion failure - to the client this appears as a
6588 connection drop. This problem occurs in the part of the code that was taken
6589 from the Samba project. Fortunately, the assertion is in a very simple
6590 function, so I have fixed this by reproducing the function inline in the
6591 one place where it is called, and arranging for authentication to fail
6592 instead of killing the process with assert().
6594 75. The SPA client code was not working when the server requested OEM rather
6595 than Unicode encoding.
6597 76. Added code to make require_files with a specific uid setting more usable in
6598 the case where statting the file as root fails - usually a non-root-mounted
6599 NFS file system. When this happens and the failure is EACCES, Exim now
6600 forks a subprocess and does the per-uid checking as the relevant uid.
6602 77. Added process_log_path.
6604 78. If log_file_path was not explicitly set, a setting of check_log_space or
6605 check_log_inodes was ignored.
6607 79. If a space check for the spool or log partitions fails, the incident is now
6608 logged. Of course, in the latter case the data may get lost...
6610 80. Added the %p formatting code to string_format() so that it can be used to
6611 print addresses in debug_print(). Adjusted all the address printing in the
6612 debugging in store.c to use %p rather than %d.
6614 81. There was a concern that a line of code in smtp_in.c could overflow a
6615 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6616 initially expressed, the concern was not well-founded, because trailing
6617 spaces are removed early. However, if the trailing spaces were followed by
6618 a NULL, they did not get removed, so the overflow was possible. Two fixes
6621 (a) I re-wrote the offending code in a cleaner fashion.
6622 (b) If an incoming SMTP command contains a NULL character, it is rejected
6625 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6626 runs initgroups(), so that if the Exim user is in any additional groups,
6627 they will be used during message reception.
6633 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob