1 Change log file for Exim from version 4.21
2 -------------------------------------------
6 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
7 than -2003 did; needs libidn2 in addition to linidn.
12 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
13 supports it and a size is available (ie. the sending peer gave us one).
15 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
16 years of being deprecated). The replacements are the ACLs
17 acl_smtp_mime and acl_not_smtp_mime.
19 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
20 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
21 or even an in-clear connection were permitted. Now, if the host lookup
22 was dnssec and dane was requested then the host is only used if the
23 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
24 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
25 if one fails this test.
26 This means that a poorly-configured remote DNS will make it incommunicado;
27 but it protects against a DNS-interception attack on it.
29 JH/04 Bug 1810: make continued-use of an open smtp transport connection
30 non-noisy when a race steals the message being considered.
32 JH/05 If main configuration option tls_certificate is unset, generate a
33 self-signed certificate for inbound TLS connections.
35 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
36 in rewrites and routers.
38 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
39 and logged a warning sing 4.83; now they are a configuration file error.
41 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
42 (lacking @domain). Apply the same qualification processing as RCPT.
44 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
46 JH/10 Support ${sha256:} applied to a string (as well as the previous
49 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
50 a cutthrough deliver is pending, as we always want to make a connection.
51 This also avoids re-routing the message when later placing the cutthrough
52 connection after a verify cache hit.
53 Do not update it with the verify result either.
55 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
56 when routing results in more than one destination address.
58 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
59 signing (which inhibits the cutthrough capability). Previously only
60 the presence of an option was tested; now an expansion evaluating as
61 empty is permissible (obviously it should depend only on data available
62 when the cutthrough connection is made).
64 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
65 the relevant preceding SMTP command did not note the pipelining mode.
67 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
68 Previously they were not counted.
70 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
71 as one having no matching records. Previously we deferred the message
72 that needed the lookup.
74 JH/17 Fakereject: previously logged as a norml message arrival "<="; now
75 distinguished as "(=".
77 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
78 for missing MX records. Previously it only worked for missing A records.
80 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
82 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
83 after the data-go-ahead and data-ack. Patch from Jason Betts.
85 JH/21 Bug 1846: Send DMARC forensic reports for reject and quaratine results,
86 even for a "none" policy. Patch from Tony Meyer.
88 JH/22 Fix continued use of a connection for further deliveries. If a port was
89 specified by a router, it must also match for the delivery to be
92 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
93 When one of the recipients of a message was unsuitable for the connection
94 (has no matching addresses), we lost track of needing to mark it
95 deferred. As a result mail would be lost.
97 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
99 JH/25 Decoding ACL controls is now done using a binary search; the source code
100 takes up less space and should be simpler to maintain. Merge the ACL
101 condition decode tables also, with similar effect.
103 JH/26 Fix problem with one_time used on a redirect router which returned the
104 parent address unchanged. A retry would see the parent address marked as
105 delivered, so not attempt the (identical) child. As a result mail would
108 JH/27 Fix a possible security hole, wherein a process operating with the Exim
109 UID can gain a root shell. Credit to http://www.halfdog.net/ for
110 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
113 JH/28 Enable {spool,log} filesystem space and inode checks as default.
114 Main config options check_{log,spool}_{inodes,space} are now
115 100 inodes, 10MB unless set otherwise in the configuration.
117 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
118 Previously it only applied to the main-section connection policy
121 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
123 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
124 by me. Added RFC7919 DH primes as an alternative.
126 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
127 are not in the system include path.
129 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
130 GnuTLS, when a session startup failed (eg because the client disconnected)
131 Exim did stdio operations after fclose. This was exposed by a recent
132 change which nulled out the file handle after the fclose.
134 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
135 signed directly by the cert-signing cert, rather than an intermediate
136 OCSP-signing cert. This is the model used by LetsEncrypt.
138 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
143 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
144 and 3.4.4 - once the server is enabled to respond to an OCSP request
145 it does even when not requested, resulting in a stapling non-aware
146 client dropping the TLS connection.
148 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
149 support variable-length bit vectors. No functional change.
151 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
152 The I= interface field on outgoing lines is now after the H= remote
153 host field, same as incoming lines. There is a separate
154 outgoing_interface log selector which allows you to disable the
157 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
158 If not running log_selector +smtp_connection the mainlog would be held
159 open indefinitely after a "too many connections" event, including to a
160 deleted file after a log rotate. Leave the per net connection logging
161 leaving it open for efficiency as that will be quickly detected by the
162 check on the next write.
164 HS/01 Bug 1671: Fix post transport crash.
165 Processing the wait-<transport> messages could crash the delivery
166 process if the message IDs didn't exist for some reason. When
167 using 'split_spool_directory=yes' the construction of the spool
168 file name failed already, exposing the same netto behaviour.
170 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
171 mime_regex ACL conditions.
173 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
174 to DSN fail messages (bounces): remote IP, remote greeting, remote response
175 to HELO, local diagnostic string.
177 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
178 log line to debug. Even when configured with a tls authenticator many
179 client connections are expected to not authenticate in this way, so
180 an authenticate fail is not an error.
182 HS/02 Add the Exim version string to the process info. This way exiwhat
183 gives some more detail about the running daemon.
185 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
186 matter for fast-change records such as DNSBLs.
188 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
189 retry record, even if constant. There may be multiple transports with
190 different interface settings and the retry behaviour needs to be kept
193 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
195 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
197 JH/10 Bug 840: fix log_defer_output option of pipe transport
199 JH/11 Bug 830: use same host for all RCPTS of a message, even under
200 hosts_randomize. This matters a lot when combined with mua_wrapper.
202 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
203 ${quote_pgsql:<string>} operator.
205 JH/13 Bug 1708: avoid misaligned access in cached lookup.
207 JH/14 Change header file name for freeradius-client. Relevant if compiling
208 with Radius support; from the Gentoo tree and checked under Fedora.
210 JH/15 Bug 1712: Introduce $prdr_requested flag variable
212 JH/16 Bug 1714: Permit an empty string as expansion result for transport
213 option transport_filter, meaning no filtering.
215 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
217 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
218 defaults to "*" (all hosts). The variable is now available when not built
219 with TLS, default unset, mainly to enable keeping the testuite sane.
220 If a server certificate is not supplied (via tls_certificate) an error is
221 logged, and clients will find TLS connections fail on startup. Presumably
222 they will retry in-clear.
223 Packagers of Exim are strongly encouraged to create a server certificate
224 at installation time.
226 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
227 with the $config_file variable.
229 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
230 in transport context, after the attempt, and per-recipient. The latter type
231 is per host attempted. The event data is the error message, and the errno
232 information encodes the lookup type (A vs. MX) used for the (first) host,
233 and the trailing two digits of the smtp 4xx reponse.
235 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
236 to write to mainlog (or rejectlog, paniclog) in the window between file
237 creation and permissions/ownership being changed. Particularly affects
238 installations where exicyclog is run as root, rather than exim user;
239 result is that the running daemon panics and dies.
241 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
243 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
244 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
245 "pri" and "weight". Note that the previous implicit priority given by the
246 list order is no longer honoured.
248 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
251 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
252 by defining SUPPORT_SOCKS.
254 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
255 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
256 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
257 variables are renamed to proxy_{local,external}_{address,port}.
259 JH/27 Move Internationalisation support from Experimental to mainline, enabled
260 for a build by defining SUPPORT_I18N
262 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
263 of the query string, and make ${quote_redis:} do that quoting.
265 JH/29 Move Events support from Experimental to mainline, enabled by default
266 and removable for a build by defining DISABLE_EVENT.
268 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
270 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
271 cached by the daemon.
273 JH/32 Move Redis support from Experimental to mainline, enabled for a build
274 by defining LOOKUP_REDIS. The libhiredis library is required.
276 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
277 keys are given for lookup.
279 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
280 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
281 only supported when built with TLS support. The PolarSSL SHA routines
282 are still used when the TLS library is too old for convenient support.
284 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
285 openssl_options), for security. OpenSSL forces this from version 1.1.0
286 server-side so match that on older versions.
288 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
289 allocation for $value could be released as the expansion processing
290 concluded, but leaving the global pointer active for it.
292 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
293 and to use the domains and local_parts ACL conditions.
295 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
296 incorrectly not doubled on cutthrough transmission, hence seen as a
297 body-termination at the receiving system - resulting in truncated mails.
298 Commonly the sender saw a TCP-level error, and retransmitted the message
299 via the normal store-and-forward channel. This could result in duplicates
300 received - but deduplicating mailstores were liable to retain only the
301 initial truncated version.
303 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
305 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
307 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
308 we're in there, support oversigning also; bug 1309.
310 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
312 HS/04 Add support for keep_environment and add_environment options.
314 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
315 either intentional arithmetic overflow during PRNG, or testing config-
318 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
319 delivery resulted in actual delivery. Cancel cutthrough before DATA
322 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
323 rejects a recipient: pass the reject to the originator.
325 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
326 Many were false-positives and ignorable, but it's worth fixing the
329 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
330 for the new environment-manipulation done at startup. Move the routines
331 from being local to tls.c to being global via the os.c file.
333 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
334 an extract embedded as result-arg for a map, the first arg for extract
335 is unavailable so we cannot tell if this is a numbered or keyed
336 extraction. Accept either.
342 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
345 JH/02 The smtp transport option "multi_domain" is now expanded.
347 JH/03 The smtp transport now requests PRDR by default, if the server offers
350 JH/04 Certificate name checking on server certificates, when exim is a client,
351 is now done by default. The transport option tls_verify_cert_hostnames
352 can be used to disable this per-host. The build option
353 EXPERIMENTAL_CERTNAMES is withdrawn.
355 JH/05 The value of the tls_verify_certificates smtp transport and main options
356 default to the word "system" to access the system default CA bundle.
357 For GnuTLS, only version 3.0.20 or later.
359 JH/06 Verification of the server certificate for a TLS connection is now tried
360 (but not required) by default. The verification status is now logged by
361 default, for both outbound TLS and client-certificate supplying inbound
364 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
367 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
368 Status Notification (bounce) messages are now MIME format per RFC 3464.
369 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
370 under the control of the dsn_advertise_hosts option, and routers may
371 have a dsn_lasthop option.
373 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
374 default, modifiable by a malware= option. The list separator for
375 the options can now be changed in the usual way. Bug 68.
377 JH/10 The smtp_receive_timeout main option is now expanded before use.
379 JH/11 The incoming_interface log option now also enables logging of the
380 local interface on delivery outgoing connections.
382 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
383 if the interface and destination host and port all match.
385 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
388 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
389 Patch from Andrew Lewis.
391 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
392 now supports optional time-restrictions, weighting, and priority
393 modifiers per server. Patch originally by <rommer@active.by>.
395 JH/16 The spamd_address main option now supports a mixed list of local
396 and remote servers. Remote servers can be IPv6 addresses, and
397 specify a port-range.
399 JH/17 Bug 68: The spamd_address main option now supports an optional
400 timeout value per server.
402 JH/18 Bug 1581: Router and transport options headers_add/remove can
403 now have the list separator specified.
405 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
408 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
411 JH/21 Support for the A6 type of dns record is withdrawn.
413 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
414 rather than the verbs used.
416 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
417 from 255 to 1024 chars.
419 JH/24 Verification callouts now attempt to use TLS by default.
421 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
422 are generic router options now. The defaults didn't change.
424 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
425 Original patch from Alexander Shikoff, worked over by JH.
427 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
428 files if ZCAT_COMMAND is not executable.
430 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
432 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
434 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
435 Normally benign, it bites when the pair was led to by a CNAME;
436 modern usage is to not canoicalize the domain to a CNAME target
437 (and we were inconsistent anyway for A-only vs AAAA+A).
439 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
441 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
442 when evaluating $sender_host_dnssec.
444 JH/31 Check the HELO verification lookup for DNSSEC, adding new
445 $sender_helo_dnssec variable.
447 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
449 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
451 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
453 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
454 documented as working, but never had. Support all but $spam_report.
456 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
457 added for tls authenticator.
459 HS/03 Add perl_taintmode main config option
464 TL/01 When running the test suite, the README says that variables such as
465 no_msglog_check are global and can be placed anywhere in a specific
466 test's script, however it was observed that placement needed to be near
467 the beginning for it to behave that way. Changed the runtest perl
468 script to read through the entire script once to detect and set these
469 variables, reset to the beginning of the script, and then run through
470 the script parsing/test process like normal.
472 TL/02 The BSD's have an arc4random API. One of the functions to induce
473 adding randomness was arc4random_stir(), but it has been removed in
474 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
475 function when detected.
477 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
478 cause callback expansion.
480 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
481 syntax errors in an expansion can be treated as a string instead of
482 logging or causing an error, due to the internal use of bool_lax
483 instead of bool when processing it.
485 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
486 server certificates when making smtp deliveries.
488 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
490 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
492 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
494 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
495 Merged patch from Sebastian Wiedenroth.
497 JH/05 Fix results-pipe from transport process. Several recipients, combined
498 with certificate use, exposed issues where response data items split
499 over buffer boundaries were not parsed properly. This eventually
500 resulted in duplicates being sent. This issue only became common enough
501 to notice due to the introduction of conection certificate information,
502 the item size being so much larger. Found and fixed by Wolfgang Breyha.
504 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
505 size buffer was used, resulting in syntax errors when an expansion
508 JH/07 Add support for directories of certificates when compiled with a GnuTLS
509 version 3.3.6 or later.
511 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
512 is EXPERIMENTAL_EVENT, the main-configuration and transport options
513 both become "event_action", the variables become $event_name, $event_data
514 and $event_defer_errno. There is a new variable $verify_mode, usable in
515 routers, transports and related events. The tls:cert event is now also
516 raised for inbound connections, if the main configuration event_action
519 TL/06 In test suite, disable OCSP for old versions of openssl which contained
520 early OCSP support, but no stapling (appears to be less than 1.0.0).
522 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
523 server certificate names available under the smtp transport option
524 "tls_verify_cert_hostname" now do not permit multi-component wildcard
527 JH/10 Time-related extraction expansions from certificates now use the main
528 option "timezone" setting for output formatting, and are consistent
529 between OpenSSL and GnuTLS compilations. Bug 1541.
531 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
532 encoded parameter in the incoming message. Bug 1558.
534 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
535 include certificate info, eximon was claiming there were spoolfile
538 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
540 JH/14 Log delivery-related information more consistently, using the sequence
541 "H=<name> [<ip>]" wherever possible.
543 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
544 are problematic for Debian distribution, omit them from the release
547 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
549 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
552 JH/17 Fix a null-indirection in certextract expansions when a nondefault
553 output list separator was used.
558 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
559 checkers that were complaining about end of non-void function with no
562 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
563 This was a regression intruduced in 4.83 by another bugfix.
565 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
567 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
568 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
574 TF/01 Correctly close the server side of TLS when forking for delivery.
576 When a message was received over SMTP with TLS, Exim failed to clear up
577 the incoming connection properly after forking off the child process to
578 deliver the message. In some situations the subsequent outgoing
579 delivery connection happened to have the same fd number as the incoming
580 connection previously had. Exim would try to use TLS and fail, logging
581 a "Bad file descriptor" error.
583 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
584 utilities have not been installed.
586 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
587 temporary space as the ACL may create new global variables.
589 TL/01 LDAP support uses per connection or global context settings, depending
590 upon the detected version of the libraries at build time.
592 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
593 to extract and use the src ip:port in logging and expansions as if it
594 were a direct connection from the outside internet. PPv2 support was
595 updated based on HAProxy spec change in May 2014.
597 JH/02 Add ${listextract {number}{list}{success}{fail}}.
599 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
600 Properly escape header and check for NULL return.
602 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
605 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
607 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
608 characters in header names, implemented as a verify condition.
609 Contributed by Michael Fischer v. Mollard.
611 TL/05 Rename SPF condition results err_perm and err_temp to standardized
612 results permerror and temperror. Previous values are deprecated but
613 still accepted. In a future release, err_perm and err_temp will be
614 completely removed, which will be a backward incompatibility if the
615 ACL tests for either of these two old results. Patch contributed by
616 user bes-internal on the mailing list.
618 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
620 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
621 selectors, in both main and reject logs.
623 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
626 JH/07 Add malware type "sock" for talking to simple daemon.
628 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
630 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
631 routers/transports under cutthrough routing.
633 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
634 numbers. Touch up "bool" conditional to keep the same definition.
636 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
638 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
640 JH/12 Expand items in router/transport headers_add or headers_remove lists
641 individually rather than the list as a whole. Bug 1452.
643 Required for reasonable handling of multiple headers_ options when
644 they may be empty; requires that headers_remove items with embedded
645 colons must have them doubled (or the list-separator changed).
647 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
648 view the policy declared in the DMARC record. Currently, $dmarc_status
649 is a combined value of both the record presence and the result of the
652 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
654 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
655 dnslookup router and the smtp transport (applying to the forward
658 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
659 of ldap servers used for a specific lookup. Patch provided by Heiko
662 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
663 New variable $lookup_dnssec_authenticated for observability.
665 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
666 Patch submitted by Lars Timman.
668 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
670 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
671 Requires trusted mode and valid format message id, aborts otherwise.
672 Patch contributed by Heiko Schlichting.
674 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
675 certextract with support for various fields. Bug 1358.
677 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
678 is requested by default, modifiable by smtp transport option
681 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
682 operate on certificate variables to give certificate fingerprints
683 Also new ${sha256:cert_variable}.
685 JH/23 The PRDR feature is moved from being Experimental into the mainline.
687 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
688 Christian Aistleitner.
690 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
692 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
693 file. Patch from Wolfgang Breyha.
695 JH/25 Expand the coverage of the delivery $host and $host_address to
696 client authenticators run in verify callout. Bug 1476.
698 JH/26 Port service names are now accepted for tls_on_connect_ports, to
699 align with daemon_smtp_ports. Bug 72.
701 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
702 support and error reporting did not work properly.
704 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
705 and is readable. Patch from Andrew Colin Kissa.
707 TL/14 Enhance documentation of ${run expansion and how it parses the
708 commandline after expansion, particularly in the case when an
709 unquoted variable expansion results in an empty value.
711 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
713 PP/02 Fix internal collision of T_APL on systems which support RFC3123
714 by renaming away from it. Addresses GH issue 15, reported by
717 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
719 TL/15 SECURITY: prevent double expansion in math comparison functions
720 (can expand unsanitized data). Not remotely exploitable.
727 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
729 PP/02 Make -n do something, by making it not do something.
730 When combined with -bP, the name of an option is not output.
732 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
735 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
736 $sender_host_name and config options to manage this, and basic check
739 PP/05 DSCP support for outbound connections and control modifier for inbound.
741 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
742 (Only plugin which currently uses this is kerberos4, which nobody should
743 be using, but we should make it available and other future plugins might
744 conceivably use it, even though it would break NAT; stuff *should* be
745 using channel bindings instead).
747 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
748 name; added for Sendmail compatibility; requires admin caller.
749 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
750 just ignore it); requires trusted caller.
751 Also parse but ignore: -Ac -Am -X<logfile>
754 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
756 TL/02 Add +smtp_confirmation as a default logging option.
758 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
759 Patch by Magnus Holmgren from 2007-02-20.
761 TL/04 Bugzilla 1281 - Spec typo.
762 Bugzilla 1283 - Spec typo.
763 Bugzilla 1290 - Spec grammar fixes.
765 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
767 TL/06 Add Experimental DMARC support using libopendmarc libraries.
769 TL/07 Fix an out of order global option causing a segfault. Reported to dev
770 mailing list by by Dmitry Isaikin.
772 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
774 JH/02 Support "G" suffix to numbers in ${if comparisons.
776 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
778 NM/01 Bugzilla 1197 - Spec typo
779 Bugzilla 1196 - Spec examples corrections
781 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
783 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
784 gnutls_enable_pkcs11, but renamed to more accurately indicate its
787 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
788 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
790 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
791 "acl {{name}{arg}...}", and optional args on acl condition
794 JH/05 Permit multiple router/transport headers_add/remove lines.
796 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
798 JH/07 Avoid using a waiting database for a single-message-only transport.
799 Performance patch from Paul Fisher. Bugzilla 1262.
801 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
804 JH/09 Add $headers_added variable, with content from use of ACL modifier
805 add_header (but not yet added to the message). Bugzilla 199.
807 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
808 Pulled from Bugzilla 817 by Wolfgang Breyha.
810 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
812 (nb: this is the same fix as in Exim 4.80.1)
814 JH/11 Add A= logging on delivery lines, and a client_set_id option on
817 JH/12 Add optional authenticated_sender logging to A= and a log_selector
820 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
822 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
823 advertise SMTP AUTH mechanism to us, instead of a generic
824 protocol violation error. Also, make Exim more robust to bad
825 data from the Dovecot auth socket.
827 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
829 When a queue runner is handling a message, Exim first routes the
830 recipient addresses, during which it prunes them based on the retry
831 hints database. After that it attempts to deliver the message to
832 any remaining recipients. It then updates the hints database using
835 So if a recipient address works intermittently, it can get repeatedly
836 deferred at routing time. The retry hints record remains fresh so the
837 address never reaches the final cutoff time.
839 This is a fairly common occurrence when a user is bumping up against
840 their storage quota. Exim had some logic in its local delivery code
841 to deal with this. However it did not apply to per-recipient defers
842 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
844 This change adds a proper retry rule check during routing so that the
845 final cutoff time is checked against the message's age. We only do
846 this check if there is an address retry record and there is not a
847 domain retry record; this implies that previous attempts to handle
848 the address had the retry_use_local_parts option turned on. We use
849 this as an approximation for the destination being like a local
850 delivery, as in LMTP.
852 I suspect this new check makes the old local delivery cutoff check
853 redundant, but I have not verified this so I left the code in place.
855 TF/02 Correct gecos expansion when From: is a prefix of the username.
857 Test 0254 submits a message to Exim with the header
861 When I ran the test suite under the user fanf2, Exim expanded
862 the header to contain my full name, whereas it should have added
863 a Resent-Sender: header. It erroneously treats any prefix of the
864 username as equal to the username.
866 This change corrects that bug.
868 GF/01 DCC debug and logging tidyup
869 Error conditions log to paniclog rather than rejectlog.
870 Debug lines prefixed by "DCC: " to remove any ambiguity.
872 TF/03 Avoid unnecessary rebuilds of lookup-related code.
874 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
875 Bug spotted by Jeremy Harris; was flawed since initial commit.
876 Would have resulted in OCSP responses post-SNI triggering an Exim
877 NULL dereference and crash.
879 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
881 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
882 Bug detection, analysis and fix by Samuel Thibault.
883 Bugzilla 1331, Debian bug #698092.
885 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
887 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
888 Server implementation by Todd Lyons, client by JH.
889 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
890 config variable "prdr_enable" controls whether the server
891 advertises the facility. If the client requests PRDR a new
892 acl_data_smtp_prdr ACL is called once for each recipient, after
893 the body content is received and before the acl_smtp_data ACL.
894 The client is controlled by bolth of: a hosts_try_prdr option
895 on the smtp transport, and the server advertisement.
896 Default client logging of deliveries and rejections involving
897 PRDR are flagged with the string "PRDR".
899 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
900 fclose(). Diagnosis by Todd Lyons.
902 PP/17 Update configure.default to handle IPv6 localhost better.
903 Patch by Alain Williams (plus minor tweaks).
906 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
907 This is now consistent with GnuTLS, and is now documented: the
908 previous undocumented portable approach to treating the option as
909 unset was to force an expansion failure. That still works, and
910 an empty string is now equivalent.
912 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
913 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
914 not performing validation itself.
916 PP/20 Added force_command boolean option to pipe transport.
917 Patch from Nick Koston, of cPanel Inc.
919 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
922 TF/04 Added udpsend ACL modifer and hexquote expansion operator
924 PP/21 Fix eximon continuous updating with timestamped log-files.
925 Broken in a format-string cleanup in 4.80, missed when I repaired the
926 other false fix of the same issue.
927 Report and fix from Heiko Schlichting.
930 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
931 Report from Prashanth Katuri.
933 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
934 It's SecureTransport, so affects any MacOS clients which use the
935 system-integrated TLS libraries, including email clients.
937 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
938 using a MIME ACL for non-SMTP local injection.
939 Report and assistance in diagnosis by Warren Baker.
941 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
943 JH/16 Fix comparisons for 64b. Bugzilla 1385.
945 TL/09 Add expansion variable $authenticated_fail_id to keep track of
946 last id that failed so it may be referenced in subsequent ACL's.
948 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
951 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
952 ldap library initialization, allowing self-signed CA's to be
953 used. Also properly sets require_cert option later in code by
954 using NULL (global ldap config) instead of ldap handle (per
955 session). Bug diagnosis and testing by alxgomz.
957 TL/12 Enhanced documentation in the ratelimit.pl script provided in
958 the src/util/ subdirectory.
960 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
961 renamed to Transport Post Delivery Action by Jeremy Harris, as
964 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
965 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
966 redis_servers = needs to be configured which will be used by the redis
967 lookup. Patch from Warren Baker, of The Packet Hub.
969 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
971 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
972 hostname or reverse DNS when processing a host list. Used suggestions
973 from multiple comments on this bug.
975 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
977 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
978 Missed a few lines, added it to make the runtest require no keyboard
981 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
982 contains upper case chars. Make router use caseful_local_part.
984 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
985 support when GnuTLS has been built with p11-kit.
991 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
993 This, or similar/improved, will also be change PP/11 of 4.82.
999 PP/01 Handle short writes when writing local log-files.
1000 In practice, only affects FreeBSD (8 onwards).
1001 Bugzilla 1053, with thanks to Dmitry Isaikin.
1003 NM/01 Bugzilla 949 - Documentation tweak
1005 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1008 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1010 PP/02 Implemented gsasl authenticator.
1012 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1014 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1015 `pkg-config foo` for cflags/libs.
1017 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1018 with rest of GSASL and with heimdal_gssapi.
1020 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1021 `pkg-config foo` for cflags/libs for the TLS implementation.
1023 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1024 properties get this fed in as external SSF. A number of robustness
1025 and debugging improvements to the cyrus_sasl authenticator.
1027 PP/08 cyrus_sasl server now expands the server_realm option.
1029 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1030 Patch by Jeremy Harris.
1032 PP/10 Added dbmjz lookup type.
1034 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1036 PP/12 MAIL args handles TAB as well as SP, for better interop with
1037 non-compliant senders.
1038 Analysis and variant patch by Todd Lyons.
1040 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1041 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1042 Patch from Dirk Mueller <dmueller@suse.com>
1044 PP/13 tls_peerdn now print-escaped for spool files.
1045 Observed some $tls_peerdn in wild which contained \n, which resulted
1046 in spool file corruption.
1048 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1049 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1050 or write after TLS renegotiation, which otherwise led to messages
1053 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1054 as a tracking header (ie: a signed header comes before the signature).
1055 Patch from Wolfgang Breyha.
1057 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1058 comma-sep list; embedded commas doubled.
1060 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1062 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1064 Report and patch from Dmitry Banschikov.
1066 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1067 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1068 failures. We appear to now support TLS1.1+ with Exim.
1070 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1071 lets Exim select keys and certificates based upon TLS SNI from client.
1072 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1073 before an outbound SMTP session. New log_selector, +tls_sni.
1075 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1076 NULL dereference. Report and patch from Alun Jones.
1078 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1079 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1080 Not seeing resolver debug output on NetBSD, but suspect this is a
1081 resolver implementation change.
1083 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1084 Left warnings. Added "eximon gdb" invocation mode.
1086 PP/21 Defaulting "accept_8bitmime" to true, not false.
1088 PP/22 Added -bw for inetd wait mode support.
1090 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1091 locate the relevant includes and libraries. Made this the default.
1093 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1094 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1096 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1097 This may cause build issues on older platforms.
1099 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1100 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1101 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1102 Added SNI support via GnuTLS too.
1103 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1105 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1107 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1108 Applied second patch from Janne, implementing suggestion to default
1109 multiple-strings-in-record handling to match SPF spec.
1111 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1113 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1114 read-only, out of scope).
1115 Patch from Wolfgang Breyha, report from Stuart Northfield.
1117 PP/29 Fix three issues highlighted by clang analyser static analysis.
1118 Only crash-plausible issue would require the Cambridge-specific
1119 iplookup router and a misconfiguration.
1120 Report from Marcin Mirosław.
1122 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1124 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1125 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1126 As part of this, removing so much warning spew let me fix some minor
1127 real issues in debug logging.
1129 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1130 assignment on my part. Fixed.
1132 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1133 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1134 Janne Snabb (who went above and beyond: thank you).
1136 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1137 string otherwise requires a connection and a bunch more work and it's
1138 relatively easy to get wrong. Should also expose TLS library linkage
1141 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1142 64-bit ${eval} (JH/03).
1144 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1145 GNU libc to support some of the 64-bit stuff, should not lead to
1146 conflicts. Defined before os.h is pulled in, so if a given platform
1147 needs to override this, it can.
1149 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1150 protection layer was required, which is not implemented.
1151 Bugzilla 1254, patch from Wolfgang Breyha.
1153 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1154 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1155 tls_dhparam take prime identifiers. Also unbreak combination of
1156 OpenSSL+DH_params+TLSSNI.
1158 PP/39 Disable SSLv2 by default in OpenSSL support.
1164 PP/01 Solaris build fix for Oracle's LDAP libraries.
1165 Bugzilla 1109, patch from Stephen Usher.
1167 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1169 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1172 TF/02 Fix a couple more cases where we did not log the error message
1173 when unlink() failed. See also change 4.74-TF/03.
1175 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1176 lock up or crash if it happened to be inside a call to libc when it
1177 got a SIGUSR1 from exiwhat.
1179 The SIGUSR1 handler appends the current process status to the process
1180 log which is later printed by exiwhat. It used to use the general
1181 purpose logging code to do this, but several functions it calls are
1182 not safe for signals.
1184 The new output code in the SIGUSR1 handler is specific to the process
1185 log, and simple enough that it's easy to inspect for signal safety.
1186 Removing some special cases also simplifies the general logging code.
1187 Removing the spurious timestamps from the process log simplifies
1190 TF/04 Improved ratelimit ACL condition.
1192 The /noupdate option has been deprecated in favour of /readonly which
1193 has clearer semantics. The /leaky, /strict, and /readonly update modes
1194 are mutually exclusive. The update mode is no longer included in the
1195 database key; it just determines when the database is updated. (This
1196 means that when you upgrde Exim will forget old rate measurements.)
1198 Exim now checks that the per_* options are used with an update mode that
1199 makes sense for the current ACL. For example, when Exim is processing a
1200 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1201 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1202 must specify per_mail/readonly. If you omit the update mode it defaults to
1203 /leaky where that makes sense (as before) or /readonly where required.
1205 The /noupdate option is now undocumented but still supported for
1206 backwards compatibility. It is equivalent to /readonly except that in
1207 ACLs where /readonly is required you may specify /leaky/noupdate or
1208 /strict/noupdate which are treated the same as /readonly.
1210 A useful new feature is the /count= option. This is a generalization
1211 of the per_byte option, so that you can measure the throughput of other
1212 aggregate values. For example, the per_byte option is now equivalent
1213 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1215 The per_rcpt option has been generalized using the /count= mechanism
1216 (though it's more complicated than the per_byte equivalence). When it is
1217 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1218 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1219 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1220 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1221 Note that using per_rcpt with a non-readonly update mode in more than
1222 one ACL will cause the recipients to be double-counted. (The per_mail
1223 and per_byte options don't have this problem.)
1225 The handling of very low rates has changed slightly. If the computed rate
1226 is less than the event's count (usually one) then this event is the first
1227 after a long gap. In this case the rate is set to the same as this event's
1228 count, so that the first message of a spam run is counted properly.
1230 The major new feature is a mechanism for counting the rate of unique
1231 events. The new per_addr option counts the number of different
1232 recipients that someone has sent messages to in the last time period. It
1233 behaves like per_rcpt if all the recipient addresses are different, but
1234 duplicate recipient addresses do not increase the measured rate. Like
1235 the /count= option this is a general mechanism, so the per_addr option
1236 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1237 example, measure the rate that a client uses different sender addresses
1238 with the options per_mail/unique=$sender_address. There are further
1239 details in the main documentation.
1241 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1243 TF/06 Removed a few PCRE remnants.
1245 TF/07 Automatically extract Exim's version number from tags in the git
1246 repository when doing development or release builds.
1248 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1249 Bugzilla 879. Patch from Paul Fisher.
1251 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1252 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1255 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1257 PP/05 Variable $av_failed, true if the AV scanner deferred.
1258 Bugzilla 1078. Patch from John Horne.
1260 PP/06 Stop make process more reliably on build failure.
1261 Bugzilla 1087. Patch from Heiko Schlittermann.
1263 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1264 Bugzilla 1089. Patch from Heiko Schlittermann.
1266 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1267 Bugzilla 1131. Patch from Holger Weiß.
1269 PP/09 Handle IPv6 addresses with SPF.
1270 Bugzilla 860. Patch from Wolfgang Breyha.
1272 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1274 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1277 PP/11 match_* no longer expand right-hand-side by default.
1278 New compile-time build option, EXPAND_LISTMATCH_RHS.
1279 New expansion conditions, "inlist", "inlisti".
1281 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1283 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1285 PP/14 fix log_write() format string regression from TF/03.
1286 Bugzilla 1152. Patch from Dmitry Isaikin.
1292 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1294 PP/02 Harmonised TLS library version reporting; only show if debugging.
1295 Layout now matches that introduced for other libraries in 4.74 PP/03.
1297 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1299 PP/04 New "dns_use_edns0" global option.
1301 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1304 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1305 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1307 TK/01 Updated PolarSSL code to 0.14.2.
1308 Bugzilla 1097. Patch from Andreas Metzler.
1310 PP/07 Catch divide-by-zero in ${eval:...}.
1311 Fixes bugzilla 1102.
1313 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1316 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1317 format-string attack -- SECURITY: remote arbitrary code execution.
1319 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1320 time unintentionally subject to list matching rules, letting the header
1321 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1322 arbitrary string expansion). This allowed for information disclosure.
1324 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1325 INT_MIN/-1 -- value coerced to INT_MAX.
1331 NM/01 Workround for PCRE version dependency in version reporting
1334 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1335 This fixes portability to compilers other than gcc, notably
1336 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1338 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1339 makefiles for portability to HP-UX and POSIX correctness.
1341 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1342 Also via indented variable definition in the Makefile.
1343 (Debugging by Oliver Heesakkers).
1345 PP/02 Restore caching of spamd results with expanded spamd_address.
1346 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1348 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1349 Improves build reliability. Fix from: Frank Elsner
1351 NM/02 Fix wide character breakage in the rfc2047 coding
1352 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1354 NM/03 Allow underscore in dnslist lookups
1355 Fixes bug 1026. Patch from Graeme Fowler
1357 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1358 Code patches from Adam Ciarcinski of NetBSD.
1360 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1363 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1364 is logged, to avoid truncation. Patch from John Horne.
1366 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1367 Patch from Jakob Hirsch.
1369 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1370 SQL string expansion failure details.
1371 Patch from Andrey Oktyabrski.
1373 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1374 Patch from Simon Arlott.
1376 PP/09 New lookups functionality failed to compile on old gcc which rejects
1377 extern declarations in function scope.
1378 Patch from Oliver Fleischmann
1380 PP/10 Use sig_atomic_t for flags set from signal handlers.
1381 Check getgroups() return and improve debugging.
1382 Fixed developed for diagnosis in bug 927 (which turned out to be
1385 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1386 Patch from Mark Zealey.
1388 PP/12 Bugzilla 1056: Improved spamd server selection.
1389 Patch from Mark Zealey.
1391 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1392 Based on patch from Heiko Schlittermann.
1394 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1395 Patch from Uwe Doering, sign-off by Michael Haardt.
1397 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1398 variable declaration deep within a block. Bug and patch from
1401 PP/15 lookups-Makefile IRIX compatibilty coercion.
1403 PP/16 Make DISABLE_DKIM build knob functional.
1405 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1406 Patch by Simon Arlott
1408 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1409 variable argument macros. Our copy now differs from upstream.
1415 TF/01 Failure to get a lock on a hints database can have serious
1416 consequences so log it to the panic log.
1418 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1419 controlled using the smtp_confirmation log selector.
1421 TF/03 Include the error message when we fail to unlink a spool file.
1423 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1424 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1425 for maintaining out-of-tree patches for some time.
1427 PP/01 Bugzilla 139: Documentation and portability issues.
1428 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1429 Handle per-OS dynamic-module compilation flags.
1431 PP/02 Let /dev/null have normal permissions.
1432 The 4.73 fixes were a little too stringent and complained about the
1433 permissions on /dev/null. Exempt it from some checks.
1434 Reported by Andreas M. Kirchwitz.
1436 PP/03 Report version information for many libraries, including
1437 Exim version information for dynamically loaded libraries. Created
1438 version.h, now support a version extension string for distributors
1439 who patch heavily. Dynamic module ABI change.
1441 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1442 privilege escalation vulnerability whereby the Exim run-time user
1443 can cause root to append content of the attacker's choosing to
1446 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1449 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1450 If dropping privileges for untrusted macros, we disabled normal logging
1451 on the basis that it would fail; for the Exim run-time user, this is not
1452 the case, and it resulted in successful deliveries going unlogged.
1453 Fixed. Reported by Andreas Metzler.
1459 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1460 only prepend for the Resent-* case. Fixes regression introduced in
1461 Exim 4.70 by NM/22 for Bugzilla 607.
1463 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1464 increasing numbers of administrators be bitten by this.
1466 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1468 PP/03 Bugzilla 994: added openssl_options main configuration option.
1470 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1472 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1474 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1476 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1477 without a peer certificate, leading to a segfault because of an
1478 assumption that peers always have certificates. Be a little more
1479 paranoid. Problem reported by Martin Tscholak.
1481 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1482 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1483 NB: ClamAV planning to remove STREAM in "middle of 2010".
1484 CL also introduces -bmalware, various -d+acl logging additions and
1485 more caution in buffer sizes.
1487 PP/09 Implemented reverse_ip expansion operator.
1489 PP/10 Bugzilla 937: provide a "debug" ACL control.
1491 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1493 PP/12 Bugzilla 973: Implement --version.
1495 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1497 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1499 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1501 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1502 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1503 ignore trailing whitespace.
1505 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1507 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1510 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1511 Notification from Dr Andrew Aitchison.
1513 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1514 ExtendedDetectionInfo response format.
1515 Notification from John Horne.
1517 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1520 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1521 XSL and documented dependency on system catalogs, with examples of how
1524 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1527 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1528 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1529 configuration file which is writeable by the Exim user or group.
1531 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1532 of configuration files to cover files specified with the -C option if
1533 they are going to be used with root privileges, not just the default
1536 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1537 option (effectively making it always true).
1539 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1540 files to be used while preserving root privileges.
1542 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1543 that rogue child processes cannot use them.
1545 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1546 run-time user, instead of root.
1548 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1549 Exim run-time user without dropping privileges.
1551 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1552 result string, instead of calling string_vformat() twice with the same
1555 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1556 for other users. Others should always drop root privileges if they use
1557 -C on the command line, even for a whitelisted configure file.
1559 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1561 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1567 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1568 $data_path, and $header_path variables; fixed documentation bugs and
1571 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1572 exipick to access non-standard spools, including the "frozen" queue
1575 NM/01 Bugzilla 965: Support mysql stored procedures.
1576 Patch from Alain Williams
1578 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1580 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1581 Patch from Andreas Metzler
1583 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1584 Patch from Kirill Miazine
1586 NM/05 Bugzilla 671: Added umask to procmail example.
1588 JJ/03 installed exipick 20100323.0, fixing doc bug
1590 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1591 directory. Notification and patch from Dan Rosenberg.
1593 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1595 TK/02 Improve log output when DKIM signing operation fails.
1597 MH/01 Treat the transport option dkim_domain as a colon separated
1598 list, not as a single string, and sign the message with each element,
1599 omitting multiple occurences of the same signer.
1601 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1602 Bugzilla 985, 986. Patch by Simon Arlott
1604 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1605 Patch by Simon Arlott
1607 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1608 MBX locking. Notification from Dan Rosenberg.
1614 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1616 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1618 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1620 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1622 NM/04 Bugzilla 915: Flag broken perl installation during build.
1628 TK/01 Added patch by Johannes Berg that expands the main option
1629 "spamd_address" if it starts with a dollar sign.
1631 TK/02 Write list of recipients to X-Envelope-Sender header when building
1632 the mbox-format spool file for content scanning (suggested by Jakob
1635 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1636 (http://www.dcc-servers.net/) support via dccifd. Activated by
1637 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1639 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1640 by Mark Daniel Reidel <mr@df.eu>.
1642 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1643 When building exim an external PCRE library is now needed -
1644 PCRE is a system library on the majority of modern systems.
1645 See entry on PCRE_LIBS in EDITME file.
1647 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1648 conversation. Added nologin parameter to request.
1649 Patch contributed by Kirill Miazine.
1651 TF/01 Do not log submission mode rewrites if they do not change the address.
1653 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1655 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1656 log files in place. Contributed by Roberto Lima.
1658 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1660 TF/03 Bugzilla 615: When checking the local_parts router precondition
1661 after a local_part_suffix or local_part_prefix option, Exim now
1662 does not use the address's named list lookup cache, since this
1663 contains cached lookups for the whole local part.
1665 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1666 Robert Millan. Documentation is in experimental-spec.txt.
1668 TF/04 Bugzilla 668: Fix parallel build (make -j).
1670 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1672 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1673 Patch provided by Jan Srzednicki.
1675 TF/05 Leading white space used to be stripped from $spam_report which
1676 wrecked the formatting. Now it is preserved.
1678 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1679 that they are available at delivery time.
1681 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1683 TF/08 TLS error reporting now respects the incoming_interface and
1684 incoming_port log selectors.
1686 TF/09 Produce a more useful error message if an SMTP transport's hosts
1687 setting expands to an empty string.
1689 NM/06 Bugzilla 744: EXPN did not work under TLS.
1690 Patch provided by Phil Pennock.
1692 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1693 Patch provided by Richard Godbee.
1695 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1696 acl_smtp_notquit, added index entry.
1698 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1699 Patch provided by Eugene Bujak.
1701 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1702 accept(). Patch provided by Maxim Dounin.
1704 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1705 Patch provided by Phil Pennock.
1707 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1709 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1710 Patch provided by Brad "anomie" Jorsch.
1712 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1713 Patch provided by Dean Brooks.
1715 TK/05 Add native DKIM support (does not depend on external libraries).
1717 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1718 Patch provided by Graeme Fowler.
1720 NM/16 Bugzilla 851: Documentation example syntax fix.
1722 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1724 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1727 NM/19 Bugzilla 745: TLS version reporting.
1728 Patch provided by Phil Pennock.
1730 NM/20 Bugzilla 167: bool: condition support.
1731 Patch provided by Phil Pennock.
1733 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1734 clients. Patch provided by Phil Pennock.
1736 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1737 Patch provided by Brad "anomie" Jorsch.
1739 NM/23 Bugzilla 687: Fix misparses in eximstats.
1740 Patch provided by Heiko Schlittermann.
1742 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1743 Patch provided by Heiko Schlittermann.
1745 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1746 plus update to original patch.
1748 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1750 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1751 Patch provided by David Brownlee.
1753 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1755 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1757 NM/30 Bugzilla 888: TLS documentation bugfixes.
1759 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1761 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1762 Unlike the original bugzilla I have changed all shell scripts in src tree.
1764 NM/33 Bugzilla 898: Transport filter timeout fix.
1765 Patch by Todd Rinaldo.
1767 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1768 Patch by Serge Demonchaux.
1770 NM/35 Bugzilla 39: Base64 decode bug fixes.
1771 Patch by Jakob Hirsch.
1773 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1775 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1777 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1779 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1785 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1786 ALT-N's libdkim that I have put here:
1787 http://duncanthrax.net/exim-experimental/
1789 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1790 were called 'true' and it seems that C99 defines that as a reserved
1791 keyword to be used with 'bool' variable types. That means you could
1792 not include C99-style headers which use bools without triggering
1793 build errors in sieve.c.
1795 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1796 as mailq or other aliases. Changed the --help handling significantly
1797 to do whats expected. exim_usage() emits usage/help information.
1799 SC/01 Added the -bylocaldomain option to eximstats.
1801 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1803 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1805 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1811 PH/01 Another patch from the Sieve maintainer.
1813 PH/02 When an IPv6 address is converted to a string for single-key lookup
1814 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1815 dots are used instead of colons so that keys in lsearch files need not
1816 contain colons. This was done some time before quoting was made available
1817 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1818 (notated using the quote facility) so as to distinguish them from IPv4
1819 keys. This meant that lookups for IP addresses in host lists did not work
1820 for iplsearch lookups.
1822 This has been fixed by arranging for IPv6 addresses to be expressed with
1823 colons if the lookup type is iplsearch. This is not incompatible, because
1824 previously such lookups could never work.
1826 The situation is now rather anomolous, since one *can* have colons in
1827 ordinary lsearch keys. However, making the change in all cases is
1828 incompatible and would probably break a number of configurations.
1830 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1833 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1834 conversion specification without a maximum field width, thereby enabling
1835 a rogue spamd server to cause a buffer overflow. While nobody in their
1836 right mind would setup Exim to query an untrusted spamd server, an
1837 attacker that gains access to a server running spamd could potentially
1838 exploit this vulnerability to run arbitrary code as the Exim user.
1840 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1841 $primary_hostname instead of what libspf2 thinks the hosts name is.
1843 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1844 a directory entry by the name of the lookup key. Previously, if a
1845 symlink pointed to a non-existing file or a file in a directory that
1846 Exim lacked permissions to read, a lookup for a key matching that
1847 symlink would fail. Now it is enough that a matching directory entry
1848 exists, symlink or not. (Bugzilla 503.)
1850 PH/03 The body_linecount and body_zerocount variables are now exported in the
1853 PH/04 Added the $dnslist_matched variable.
1855 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1856 This means they are set thereafter only if the connection becomes
1859 PH/06 Added the client_condition to authenticators so that some can be skipped
1860 by clients under certain conditions.
1862 PH/07 The error message for a badly-placed control=no_multiline_responses left
1863 "_responses" off the end of the name.
1865 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1867 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1868 (without spaces) instead of just copying the configuration text.
1870 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1872 PH/11 Added $max_received_linelength.
1874 PH/12 Added +ignore_defer and +include_defer to host lists.
1876 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1877 way in which PCRE > 7.0 is built.
1879 PH/14 Implemented queue_only_load_latch.
1881 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1882 MAIL command. The effect was to mangle the value on 64-bit systems.
1884 PH/16 Another patch from the Sieve maintainer.
1886 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1888 PH/18 If a system quota error occurred while trying to create the file for
1889 a maildir delivery, the message "Mailbox is full" was not appended to the
1890 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1891 only to a quota excession during the actual writing of the file.
1893 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1894 characters?) which causes problems in log lines. The DN values are now
1895 passed through string_printing() before being added to log lines.
1897 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1898 and InterBase are left for another time.)
1900 PH/21 Added message_body_newlines option.
1902 PH/22 Guard against possible overflow in moan_check_errorcopy().
1904 PH/23 POSIX allows open() to be a macro; guard against that.
1906 PH/24 If the recipient of an error message contained an @ in the local part
1907 (suitably quoted, of course), incorrect values were put in $domain and
1908 $local_part during the evaluation of errors_copy.
1914 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1915 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1918 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1919 issue a MAIL command.
1921 PH/02 In an ACL statement such as
1923 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1925 if a client was not listed at all, or was listed with a value other than
1926 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1927 the condition was not true (as it should be), so access was not denied.
1928 The bug was that the ! inversion was incorrectly passed on to the second
1929 item. This has been fixed.
1931 PH/03 Added additional dnslists conditions == and =& which are different from
1932 = and & when the dns lookup returns more than one IP address.
1934 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1935 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1937 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1938 FSYNC, which compiles an option called disable_fsync that allows for
1939 bypassing fsync(). The documentation is heavily laced with warnings.
1941 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1943 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1944 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1945 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1946 including adding "make clean"; (3) Added -fPIC when compiling the test
1947 dynamically loaded module, to get rid of a warning.
1949 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1950 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1951 = 0s. The bug is otherwise harmless.
1953 PH/07 There was a bug in the dovecot authenticator such that the value of
1954 $auth1 could be overwritten, and so not correctly preserved, after a
1955 successful authentication. This usually meant that the value preserved by
1956 the server_setid option was incorrect.
1958 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1960 PH/09 Installed PCRE release 7.0.
1962 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1963 run for batched SMTP input. It is now run at the start of every message
1964 in the batch. While fixing this I discovered that the process information
1965 (output by running exiwhat) was not always getting set for -bs and -bS
1966 input. This is fixed, and it now also says "batched" for BSMTP.
1968 PH/11 Added control=no_pipelining.
1970 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1971 patch, slightly modified), and move the expansion of helo_data till after
1972 the connection is made in the smtp transport (so it can use these
1975 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1977 PH/14 Added log_selector = +pid.
1979 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1981 PH/16 Add ${if forany and ${if forall.
1983 PH/17 Added dsn_from option to vary the From: line in DSNs.
1985 PH/18 Flush SMTP output before performing a callout, unless control =
1986 no_callout_flush is set.
1988 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1989 was true (the default) a successful delivery failed to delete the retry
1990 item, thus causing premature timeout of the address. The bug is now
1993 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1995 PH/21 Long custom messages for fakedefer and fakereject are now split up
1996 into multiline reponses in the same way that messages for "deny" and
1997 other ACL rejections are.
1999 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2000 with slight modification.
2002 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2003 draft, changing the syntax and factoring some duplicate code".
2005 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2006 for deliveries of the second and subsequent messages over the same SMTP
2009 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2010 ${reduce, with only minor "tidies".
2012 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2014 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2015 expansion side effects.
2017 PH/27 When a message times out after an over-quota error from an Exim-imposed
2018 quota, the bounce message says "mailbox is full". This message was not
2019 being given when it was a system quota that was exceeded. It now should
2022 MH/03 Made $recipients available in local_scan(). local_scan() already has
2023 better access to the recipient list through recipients_list[], but
2024 $recipients can be useful in postmaster-provided expansion strings.
2026 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2027 in the case of a MAIL command with additional options following the
2028 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2029 were accidentally chopped off.
2031 PH/29 SMTP synchronization checks are implemented when a command is read -
2032 there is a check that no more input is waiting when there shouldn't be
2033 any. However, for some commands, a delay in an ACL can mean that it is
2034 some time before the response is written. In this time, more input might
2035 arrive, invalidly. So now there are extra checks after an ACL has run for
2036 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2037 pipelining has not been advertised.
2039 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2041 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2042 returned in the SMTP response when smtp_return_error_details was false.
2043 This has been fixed.
2045 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2046 instead of the C I/O that was originally supplied, because problems were
2047 reported on Solaris.
2049 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2050 Exim which did not show up earlier: it was assuming that a call to
2051 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2052 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2053 was a macro that became an assignment, so it seemed to work. This has
2054 changed to a proper function call with a void return, hence the compile
2055 error. Exim's code has been fixed.
2057 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2060 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2062 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2064 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2065 because it turns out to be too big or there is a timeout) there is no
2066 "Completed" line in the log. When some messages of this type were
2067 selected by exigrep, they were listed as "not completed". Others were
2068 picked up by some special patterns. I have improved the selection
2069 criteria to be more general.
2071 PH/38 The host_find_failed option in the manualroute router can now be set
2072 to "ignore", to completely ignore a host whose IP address cannot be
2073 found. If all hosts are ignored, the behaviour is controlled by the new
2074 host_all_ignored option.
2076 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2077 homing or because of multiple MX records with /mx) generated more than
2078 one IP address, and the following item turned out to be the local host,
2079 all the secondary addresses of the first item were incorrectly removed
2080 from the list, along with the local host and any following hosts (which
2081 is what is supposed to happen).
2083 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2084 whoever called Exim into the -H file. In the case of the daemon it was
2085 behaving confusingly. When first started, it used values for whoever
2086 started the daemon, but after a SIGHUP it used the Exim user (because it
2087 calls itself on a restart). I have changed the code so that it now always
2090 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2091 message are rejected with the same error (e.g. no authentication or bad
2092 sender address), and a DATA command is nevertheless sent (as can happen
2093 with PIPELINING or a stupid MUA), the error message that was given to the
2094 RCPT commands is included in the rejection of the DATA command. This is
2095 intended to be helpful for MUAs that show only the final error to their
2098 PH/42 Another patch from the Sieve maintainer.
2100 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2101 Eximstats - Fixed some broken HTML links and added missing column headers
2103 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2104 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2106 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2108 PH/43 Yet another patch from the Sieve maintainer.
2110 PH/44 I found a way to check for a TCP/IP connection going away before sending
2111 the response to the final '.' that terminates a message, but only in the
2112 case where the client has not sent further data following the '.'
2113 (unfortunately, this is allowed). However, in many cases there won't be
2114 any further data because there won't be any more messages to send. A call
2115 to select() can be used: if it shows that the input is "ready", there is
2116 either input waiting, or the socket has been closed. An attempt to read
2117 the next input character can distinguish the two cases. Previously, Exim
2118 would have sent an OK response which the client would never have see.
2119 This could lead to message repetition. This fix should cure that, at
2120 least in a lot of common cases.
2122 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2123 advertised in response to EHLO.
2129 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2130 fixed by 4.65/MH/01 (is this a record?) are fixed:
2132 (i) An empty string was always treated as zero by the numeric comparison
2133 operators. This behaviour has been restored.
2135 (ii) It is documented that the numeric comparison operators always treat
2136 their arguments as decimal numbers. This was broken in that numbers
2137 starting with 0 were being interpreted as octal.
2139 While fixing these problems I realized that there was another issue that
2140 hadn't been noticed. Values of message_size_limit (both the global option
2141 and the transport option) were treated as octal if they started with 0.
2142 The documentation was vague. These values are now always treated as
2143 decimal, and I will make that clear in the documentation.
2149 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2150 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2153 MH/01 Don't check that the operands of numeric comparison operators are
2154 integers when their expansion is in "skipping" mode (fixes bug
2155 introduced by 4.64-PH/07).
2157 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2158 child addresses, Exim now panics and dies. Previously, because the count
2159 is held in a short int, deliveries were likely to be lost. As such a
2160 large number of recipients for a single message is ridiculous
2161 (performance will be very, very poor), I have chosen to impose a limit
2162 rather than extend the field.
2168 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2169 leftover -K file (the existence of which was triggered by #402).
2170 While we were at it, introduced process PID as part of the -K
2171 filename. This should rule out race conditions when creating
2174 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2175 processing considerably. Previous code took too long for large mails,
2176 triggering a timeout which in turn triggers #401.
2178 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2179 in the DK code in transports.c. sendfile() is not really portable,
2180 hence the _LINUX specificness.
2182 TF/01 In the add_headers option to the mail command in an Exim filter,
2183 there was a bug that Exim would claim a syntax error in any
2184 header after the first one which had an odd number of characters
2187 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2188 callout verification, Exim cached a "reject" for the entire domain. This
2189 is correct for most verifications, but it is not correct for a recipient
2190 verification with use_sender or use_postmaster set, because in that case
2191 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2192 case of MAIL FROM:<> rejection from other early rejections (e.g.
2193 rejection of HELO). When verifying a recipient using a non-null MAIL
2194 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2195 Whatever the result of the callout, the value of the domain cache is
2196 left unchanged (for any other kind of callout, getting as far as trying
2197 RCPT means that the domain itself is ok).
2199 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2202 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2203 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2204 session, the daemon ignores ECONNECT errors and logs others; it now
2205 ignores EPIPE as well.
2207 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2208 (quoted-printable decoding).
2210 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2211 later the small subsequent patch to fix an introduced bug.
2213 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2215 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2217 PH/08 An error is now given if message_size_limit is specified negative.
2219 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2220 to be given (somewhat) arbitrary names.
2222 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2225 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2226 miscellaneous code fixes
2228 PH/10 Added the log_reject_target ACL modifier to specify where to log
2231 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2232 hostname. This is wrong, because it relates to the incoming message (and
2233 probably the interface on which it is arriving) and not to the outgoing
2234 callout (which could be using a different interface). This has been
2235 changed to use the value of the helo_data option from the smtp transport
2236 instead - this is what is used when a message is actually being sent. If
2237 there is no remote transport (possible with a router that sets up host
2238 addresses), $smtp_active_hostname is used.
2240 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2241 tweaks were necessary in order to get it to work (see also 21 below):
2242 (a) The code assumed that strncpy() returns a negative number on buffer
2243 overflow, which isn't the case. Replaced with Exim's string_format()
2245 (b) There were several signed/unsigned issues. I just did the minimum
2246 hacking in of casts. There is scope for a larger refactoring.
2247 (c) The code used strcasecmp() which is not a standard C function.
2248 Replaced with Exim's strcmpic() function.
2249 (d) The code set only $1; it now sets $auth1 as well.
2250 (e) A simple test gave the error "authentication client didn't specify
2251 service in request". It would seem that Dovecot has changed its
2252 interface. Fortunately there's a specification; I followed it and
2253 changed what the client sends and it appears to be working now.
2255 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2258 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2259 address A is aliased to B and C, where B exists and C does not. Without
2260 -v the output is "A verified" because verification stops after a
2261 successful redirection if more than one address is generated. However,
2262 with -v the child addresses are also verified. Exim was outputting "A
2263 failed to verify" and then showing the successful verification for C,
2264 with its parentage. It now outputs "B failed to verify", showing B's
2265 parentage before showing the successful verification of C.
2267 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2268 look up a TXT record in a specific list after matching in a combined
2271 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2272 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2273 they consult the DNS. I had assumed they would set it the way they
2274 wanted; and indeed my experiments on Linux seem to show that in some
2275 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2276 To be on the safe side, however, I have now made the interface to
2277 host_find_byname() similar to host_find_bydns(), with an argument
2278 containing the DNS resolver options. The host_find_byname() function now
2279 sets these options at its start, just as host_find_bydns() does. The smtp
2280 transport options dns_qualify_single and dns_search_parents are passed to
2281 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2282 of host_find_byname() use the default settings of RES_DEFNAMES
2283 (qualify_single) but not RES_DNSRCH (search_parents).
2285 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2286 spool_read_header() do less string testing, by means of a preliminary
2287 switch on the second character of optional "-foo" lines. (This is
2288 overdue, caused by the large number of possibilities that now exist.
2289 Originally there were few.) While I was there, I also converted the
2290 str(n)cmp tests so they don't re-test the leading "-" and the first
2291 character, in the hope this might squeeze out yet more improvement.
2293 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2294 flag allowing group syntax was set by the header_syntax check but not
2295 turned off, possible causing trouble later; (2) The flag was not being
2296 set at all for the header_verify test, causing "group"-style headers to
2297 be rejected. I have now set it in this case, and also caused header_
2298 verify to ignore an empty address taken from a group. While doing this, I
2299 came across some other cases where the code for allowing group syntax
2300 while scanning a header line wasn't quite right (mostly, not resetting
2301 the flag correctly in the right place). These bugs could have caused
2302 trouble for malformed header lines. I hope it is now all correct.
2304 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2305 with the "reply" argument non-NULL. The code, however (which originally
2306 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2307 but it didn't always do it. This confused somebody who was copying the
2308 code for some other use. I have removed all the tests.
2310 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2311 feature that was used to support insecure browsers during the U.S. crypto
2312 embargo. It requires special client support, and Exim is probably the
2313 only MTA that supported it -- and would never use it because real RSA is
2314 always available. This code has been removed, because it had the bad
2315 effect of slowing Exim down by computing (never used) parameters for the
2316 RSA_EXPORT functionality.
2318 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2319 authenticator to fail if there's a tab character in the incoming data
2320 (there should never be unless someone is messing about, as it's supposed
2321 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2322 if the connection is using TLS or if the remote IP is the same as the
2323 local IP, and the "valid-client-cert option" if a client certificate has
2326 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2327 authenticators. This can be used for authorization after authentication
2328 succeeds. (In the case of plaintext, it servers for both authentication
2331 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2332 if any retry times were supplied.
2334 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2335 connection, where there is no client IP address to check. In this
2336 situation, the verify now always succeeds.
2338 PH/25 Applied John Jetmore's -Mset patch.
2340 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2342 PH/27 In a string expansion for a processed (not raw) header when multiple
2343 headers of the same name were present, leading whitespace was being
2344 removed from all of them, but trailing whitespace was being removed only
2345 from the last one. Now trailing whitespace is removed from each header
2346 before concatenation. Completely empty headers in a concatenation (as
2347 before) are ignored.
2349 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2350 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2352 PH/29 [Removed. This was a change that I later backed out, and forgot to
2353 correct the ChangeLog entry (that I had efficiently created) before
2354 committing the later change.]
2356 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2357 address errors (4xx response to RCPT) over the same connection as other
2358 messages routed to the same hosts. Such deliveries are always "forced",
2359 so retry times are not inspected. This resulted in far too many retries
2360 for the affected addresses. The effect occurred only when there were more
2361 hosts than the hosts_max_try setting in the smtp transport when it had
2362 the 4xx errors. Those hosts that it had tried were not added to the list
2363 of hosts for which the message was waiting, so if all were tried, there
2364 was no problem. Two fixes have been applied:
2366 (i) If there are any address or message errors in an SMTP delivery, none
2367 of the hosts (tried or untried) are now added to the list of hosts
2368 for which the message is waiting, so the message should not be a
2369 candidate for sending over the same connection that was used for a
2370 successful delivery of some other message. This seems entirely
2371 reasonable: after all the message is NOT "waiting for some host".
2372 This is so "obvious" that I'm not sure why it wasn't done
2373 previously. Hope I haven't missed anything, but it can't do any
2374 harm, as the worst effect is to miss an optimization.
2376 (ii) If, despite (i), such a delivery is accidentally attempted, the
2377 routing retry time is respected, so at least it doesn't keep
2378 hammering the server.
2380 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2381 in ${readsocket because some servers need this prod.
2383 PH/32 Added some extra debug output when updating a wait-xxx database.
2385 PH/33 The hint "could be header name not terminated by colon", which has been
2386 given for certain expansion errors for a long time, was not being given
2387 for the ${if def:h_colon_omitted{... case.
2389 PH/34 The spec says: "With one important exception, whenever a domain list is
2390 being scanned, $domain contains the subject domain." There was at least
2391 one case where this was not true.
2393 PH/35 The error "getsockname() failed: connection reset by peer" was being
2394 written to the panic log as well as the main log, but it isn't really
2395 panic-worthy as it just means the connection died rather early on. I have
2396 removed the panic log writing for the ECONNRESET error when getsockname()
2399 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2400 runs only) independently of the message's sender address. This meant
2401 that, if the 4xx error was in fact related to the sender, a different
2402 message to the same recipient with a different sender could confuse
2403 things. In particualar, this can happen when sending to a greylisting
2404 server, but other circumstances could also provoke similar problems.
2405 I have changed the default so that the retry time for these errors is now
2406 based a combination of the sender and recipient addresses. This change
2407 can be overridden by setting address_retry_include_sender=false in the
2410 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2411 remote server are returned as part of bounce messages. This was not
2412 happening for LMTP over a pipe (the lmtp transport), but now it is the
2413 same for both kinds of LMTP.
2415 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2416 in header lines that were in fact CNAMEs. This is no longer the case.
2418 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2419 and queue runs started by the daemon processed all messages. This has
2420 been fixed so that -R and -S can now usefully be given with -q<time>.
2422 PH/40 Import PCRE release 6.7 (fixes some bugs).
2424 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2426 PH/42 Give an error if -q is specified more than once.
2428 PH/43 Renamed the variables $interface_address and $interface_port as
2429 $received_ip_address and $received_port, to make it clear that these
2430 values apply to message reception, and not to the outgoing interface when
2431 a message is delivered. (The old names remain recognized, of course.)
2433 PH/44 There was no timeout on the connect() call when using a Unix domain
2434 socket in the ${readsocket expansion. There now is.
2436 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2437 be meaningful with "accept".
2439 SC/01 Eximstats V1.43
2440 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2442 SC/02 Eximstats V1.44
2443 Use a glob alias rather than an array ref in the generated
2444 parser. This improves both readability and performance.
2446 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2447 Collect SpamAssassin and rejection statistics.
2448 Don't display local sender or destination tables unless
2449 there is data to show.
2450 Added average volumes into the top table text output.
2452 SC/04 Eximstats V1.46
2453 Collect data on the number of addresses (recipients)
2454 as well as the number of messages.
2456 SC/05 Eximstats V1.47
2457 Added 'Message too big' to the list of mail rejection
2458 reasons (thanks to Marco Gaiarin).
2460 SC/06 Eximstats V1.48
2461 Mainlog lines which have GMT offsets and are too short to
2462 have a flag are now skipped.
2464 SC/07 Eximstats V1.49 (Alain Williams)
2465 Added the -emptyok flag.
2467 SC/08 Eximstats V1.50
2468 Fixes for obtaining the IP address from reject messages.
2470 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2471 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2472 whitesspace changes from 4.64-PH/27
2474 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2477 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2478 are found, allow negative numbers in numeric criteria)
2480 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2482 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2483 to match changes made in 4.64-PH/43
2485 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2487 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2488 "domain" is actually an IP address, and force a failure. This locks out
2489 those revolvers/nameservers that support "A-for-A" lookups, in
2490 contravention of the specifications.
2492 PH/48 When a host name was looked up from an IP address, and the subsequent
2493 forward lookup of the name timed out, the host name was left in
2494 $sender_host_name, contrary to the specification.
2496 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2497 restricted to single-key lookups, Exim was not diagnosing an error if
2498 * or *@ was used with a query-style lookup.
2500 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2502 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2503 long ago, but noone interested enough thought of it. Let's just say that
2504 the "1.1" means that there are some new functions that weren't there at
2505 some point in the past.
2507 PH/51 Error processing for expansion failure of helo_data from an smtp
2508 transport during callout processing was broken.
2510 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2511 tested/used via the -bh/-bhc/-bs options.
2513 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2514 bug, fixed in subsequent PCRE releases).
2516 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2517 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2519 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2525 SC/01 Use a glob alias rather than an array ref in eximstats generated
2526 parser. This improves both readability and performance.
2528 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2529 Don't display local sender or destination tables in eximstats unless
2530 there is data to show.
2531 Added average volumes into the eximstats top table text output.
2533 SC/03 Collect data on the number of addresses (recipients) as well
2534 as the number of messages in eximstats.
2536 TF/01 Correct an error in the documentation for the redirect router. Exim
2537 does (usually) call initgroups() when daemonizing.
2539 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2540 with consistent privilege compared to when running as a daemon.
2542 TF/03 Note in the spec that $authenticated_id is not set for local
2543 submissions from trusted users.
2545 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2546 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2548 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2549 by adding some example configuration directives to the default
2550 configuration file. A little bit of work is required to uncomment the
2551 directives and define how usernames and passwords are checked, but
2552 there is now a framework to start from.
2554 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2555 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2556 without this. I don't know how relevant this is to other LDAP libraries.
2558 PH/02 Add the verb name to the "unknown ACL verb" error.
2560 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2562 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2564 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2565 directory not expanded when it should be if an expanded home directory
2566 was set for the address (which is overridden by the transport).
2568 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2571 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2572 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2573 because it is too late at that time, and has no effect.
2575 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2576 security issue with \' (bugzilla #107). I could not use the
2577 PQescapeStringConn() function, because it needs a PGconn value as one of
2580 PH/08 When testing addresses using -bt, indicate those final addresses that
2581 are duplicates that would not cause an additional delivery. At least one
2582 person was confused, thinking that -bt output corresponded to deliveries.
2583 (Suppressing duplicates isn't a good idea as you lose the information
2584 about possibly different redirections that led to the duplicates.)
2586 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2587 systems where poll() doesn't work, in particular OS X.
2589 PH/10 Added more information to debugging output for retry time not reached.
2591 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2592 operations in malware.c.
2594 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2597 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2598 syslog_duplication set false, log lines that would normally be written
2599 both the the main log and to the reject log were not written to syslog at
2602 PH/14 In the default configuration, change the use of "message" in ACL warn
2603 statements to "add_header".
2605 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2606 not followed by a command (e.g. "seen endif").
2608 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2609 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2612 PH/17 Added extra conditions to the default value of delay_warning_condition
2616 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2617 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2618 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2621 The Auto-Submitted: and various List- headers are standardised, whereas I
2622 don't think Precedence: ever was.
2624 PH/18 Refactored debugging code in route_finduser() to show more information,
2625 in particular, the error code if getpwnam() issues one.
2627 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2628 This is apparently needed in addition to the PH/07 change above to avoid
2629 any possible encoding problems.
2631 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2632 but not after initializing Perl.
2634 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2635 output them only if debugging. By default they are written stderr,
2636 apparently, which is not desirable.
2638 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2641 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2644 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2646 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2647 authenticated or an ident call has been made. Suppress the default
2648 values for $authenticated_id and $authenticated_sender (but permit -oMai
2649 and -oMas) when testing with -bh.
2651 PH/24 Re-jigged the order of the tests in the default configuration so that the
2652 tests for valid domains and recipients precede the DNS black list and CSA
2653 tests, on the grounds that those ones are more expensive.
2655 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2656 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2657 This bug exists in every version of Exim that I still have, right back to
2660 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2661 However, an attempt to turn on case-sensitivity in a regex key by
2662 including (?-i) didn't work because the subject string was already
2663 lowercased, and the effects were non-intuitive. It turns out that a
2664 one-line patch can be used to allow (?-i) to work as expected.
2670 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2671 other effects) broke the use of negated acl sub-conditions.
2673 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2676 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2677 "Deny" causes Exim to reject the incoming connection with a 554 error.
2678 Unfortunately, if there is a major crisis, such as a disk failure,
2679 tcp-wrappers gives "deny", whereas what one would like would be some
2680 kind of temporary error. A kludge has been added to help with this.
2681 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2682 554 error is used if errno is still zero or contains ENOENT (which occurs
2683 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2686 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2688 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2689 errors. Otherwise a message that provokes a temporary error (when other
2690 messages do not) can cause a whole host to time out.
2692 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2693 addresses were routed directly to files or pipes from a redirect router.
2694 File deliveries just didn't batch; pipe deliveries might have suffered
2697 PH/06 A failure to get a lock for a hints database would erroneously always say
2698 "Failed to get write lock", even when it was really a read lock.
2700 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2701 of 0600. This has been changed to use the value of the lockfile_mode
2702 option (which defaults to 0600).
2704 PH/08 Applied small patch from the Sieve maintainer.
2706 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2707 folder from quota calculations, a direct delivery into this folder messed
2708 up the contents of the maildirsize file. This was because the regex was
2709 used only to exclude .Trash (or whatever) when the size of the mailbox
2710 was calculated. There was no check that a delivery was happening into an
2711 excluded directory. This bug has been fixed by ignoring all quota
2712 processing for deliveries into excluded directories.
2714 PH/10 Added the maildirfolder_create_regex option to appendfile.
2720 PH/01 The code for finding all the local interface addresses on a FreeBSD
2721 system running IPv6 was broken. This may well have applied to all BSD
2722 systems, as well as to others that have similar system calls. The broken
2723 code found IPv4 interfaces correctly, but gave incorrect values for the
2724 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2725 that it would not match correctly against @[] and not recognize the IPv6
2728 PH/02 The ipliteral router was not recognizing addresses of the form user@
2729 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2731 PH/03 Added disable_ipv6.
2733 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2734 decoded form, because it is most often used to construct To: headers
2735 lines in autoreplies, and the decoded form may well be syntactically
2736 invalid. However, $reply_address has leading white space removed, and all
2737 newlines turned into spaces so that the autoreply transport does not
2740 PH/05 If group was specified without a user on a router, and no group or user
2741 was specified on a transport, the group from the router was ignored.
2743 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2744 for visible compile-time settings that can be used to change these
2745 numbers, for those that want even more. Backwards compatibility with old
2746 spool files has been maintained. However, going back to a previous Exim
2747 release will lost any variables that are in spool files.
2749 PH/07 Two small changes when running in the test harness: increase delay when
2750 passing a TCP/IP connection to a new process, in case the original
2751 process has to generate a bounce, and remove special handling of
2752 127.0.0.2 (sic), which is no longer necessary.
2754 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2755 be the same on different OS.
2757 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2760 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2761 whether --show-vars was specified or not
2763 JJ/02 exipick: Added support for new ACL variable spool format introduced
2766 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2767 syntactically invalid From: or Reply-to: line, and a filter used this to
2768 generate an autoreply, and therefore failed to obtain an address for the
2769 autoreply, Exim could try to deliver to a non-existent relative file
2770 name, causing unrelated and misleading errors. What now happens is that
2771 it logs this as a hard delivery error, but does not attempt to create a
2774 PH/11 The exinext utility has a -C option for testing purposes, but although
2775 the given file was scanned by exinext itself; it wasn't being passed on
2776 when Exim was called.
2778 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2779 an end-of-file indication when reading a command response.
2781 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2782 compiled. In many other places in Exim, IPv6 addresses are always
2783 recognized, so I have changed this. It also means that IPv4 domain
2784 literals of the form [IPV4:n.n.n.n] are now always recognized.
2786 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2787 used if the router is not running as root, for example, when verifying at
2788 ACL time, or when using -bh. The debugging output from this situation was
2789 non-existent - all you got was a failure to exec. I have made two
2792 (a) Failures to set uid/gid, the current directory, or a process leader
2793 in a subprocess such as that created by queryprogram now generate
2794 suitable debugging ouput when -d is set.
2796 (b) The queryprogram router detects when it is not running as root,
2797 outputs suitable debugging information if -d is set, and then runs
2798 the subprocess without attempting to change uid/gid.
2800 PH/15 Minor change to Makefile for building test_host (undocumented testing
2803 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2804 additional section of a DNS packet that returns MX or SRV records.
2805 Instead, it always explicitly searches for A/AAAA records. This avoids
2806 major problems that occur when a DNS server includes only records of one
2807 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2808 fixed another bug: if SRV records were looked up and the corresponding
2809 address records were *not* found in the additional section, the port
2810 values from the SRV records were lost.
2812 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2813 using the correct key (the original address) when searching the retry
2814 rules in order to find which one to use for generating the retry hint.
2816 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2817 adding the default one. Similarly, if it contains a Reply-To: header, the
2818 errors_reply_to option, if set, is not used.
2820 PH/19 When calculating a retry time, Exim used to measure the "time since
2821 failure" by looking at the "first failed" field in the retry record. Now
2822 it does not use this if it is later than than the arrival time of the
2823 message. Instead it uses the arrival time. This makes for better
2824 behaviour in cases where some deliveries succeed, thus re-setting the
2825 "first failed" field. An example is a quota failure for a huge message
2826 when small messages continue to be delivered. Without this change, the
2827 "time since failure" will always be short, possible causing more frequent
2828 delivery attempts for the huge message than are intended.
2829 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2831 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2832 $1, $2, $3) because the numerical variables can be reset during some
2833 expansion items (e.g. "match"), thereby losing the authentication data.
2835 PH/21 Make -bV show the size of off_t variables so that the test suite can
2836 decide whether to run tests for quotas > 2G.
2838 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2839 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2840 filecount value is greater than 2G or if a quota value is greater than 2G
2841 on a system where the size of off_t is not greater than 4, a panic error
2844 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2845 never match. The debug and -bh output now contains an explicit error
2846 message indicating a malformed IPv4 address or mask.
2848 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2849 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2850 PH/23 above applies.
2852 PH/25 Do not write to syslog when running in the test harness. The only
2853 occasion when this arises is a failure to open the main or panic logs
2854 (for which there is an explicit test).
2856 PH/26 Added the /no_tell option to "control=freeze".
2858 PH/27 If a host name lookup failed very early in a connection, for example, if
2859 the IP address matched host_lookup and the reverse lookup yielded a name
2860 that did not have a forward lookup, an error message of the form "no IP
2861 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2862 could be logged. Now it outputs the IP address instead of "NULL".
2864 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2865 allows the sender and the authenticated sender to be set when
2866 submitting a message from within Exim. Since child_open_exim() is
2867 documented for local_scan(), the new function should be too.
2869 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2870 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2871 results in an empty string is now treated as unset.
2873 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2875 PH/31 Added sender_verify_fail logging option.
2877 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2878 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2879 tidied the source and removed it altogether.
2881 PH/33 When a queue run was abandoned because the load average was too high, a
2882 log line was always written; now it is written only if the queue_run log
2883 selector is set. In addition, the log line for abandonment now contains
2884 information about the queue run such as the pid. This is always present
2885 in "start" and "stop" lines but was omitted from the "abandon" line.
2887 PH/34 Omit spaces between a header name and the colon in the error message that
2888 is given when verify = headers_syntax fails (if there are lots of them,
2889 the message gets confusing).
2891 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2892 names, as there are now some PTR records that contain slashes. This check
2893 is only to protect against broken name servers that fall over on strange
2894 characters, so the fact that it applies to all lookups doesn't matter.
2896 PH/36 Now that the new test suite is complete, we can remove some of the
2897 special code in Exim that was needed for the old test suite. For example,
2898 sorting DNS records because real resolvers return them in an arbitrary
2899 order. The new test suite's fake resolver always returns records in the
2902 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2903 bounces) except when queue_only is set, to avoid logging races between
2904 the different processes.
2906 PH/38 Panic-die if .include specifies a non-absolute path.
2908 PH/39 A tweak to the "H" retry rule from its user.
2910 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2911 a label. They prevented compilation on older perls.
2913 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2914 a warning to be raised on newish perls.
2916 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2917 on queue. Changes to match documented behaviour of showing count of
2918 messages matching specified criteria.
2920 PH/40 Changed the default ident timeout from 30s to 5s.
2922 PH/41 Added support for the use of login_cap features, on those BSD systems
2923 that have them, for controlling the resources used by pipe deliveries.
2925 PH/42 The content-scanning code uses fopen() to create files in which to put
2926 message data. Previously it was not paying any attention to the mode of
2927 the files. Exim runs with umask(0) because the rest of the code creates
2928 files with open(), and sets the required mode explicitly. Thus, these
2929 files were ending up world-writeable. This was not a big issue, because,
2930 being within the spool directory, they were not world-accessible. I have
2931 created a function called modefopen, which takes an additional mode
2932 argument. It sets umask(777), creates the file, chmods it to the required
2933 mode, then resets the umask. All the relevant calls to fopen() in the
2934 content scanning code have been changed to use this function.
2936 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2937 to 24 hours. This avoids potential overflow problems when processing G
2938 and H retry rules. I suspect nobody ever tinkers with this value.
2940 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2942 PH/45 When the plaintext authenticator is running as a client, the server's
2943 challenges are checked to ensure they are valid base64 strings. By
2944 default, the authentication attempt is cancelled if an invalid string is
2945 received. Setting client_ignore_invalid_base64 true ignores these errors.
2946 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2947 they are received. Thus, the responses can be made to depend on the
2948 challenges. If an invalid string is ignored, an empty string is placed in
2951 PH/46 Messages that are created by the autoreply transport now contains a
2952 References: header, in accordance with RFCs 2822 and 3834.
2954 PH/47 Added authenticated_sender_force to the smtp transport.
2956 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2958 PH/49 Installed latest patch from the Sieve maintainer.
2960 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2961 was also set, the appendfile transport was unnecessarily scanning a
2962 directory of message files (e.g. for maildir delivery) to find the count
2963 of files (along with the size), even though it did not need this
2964 information. It now does the scan only if it needs to find either the
2965 size of the count of files.
2967 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2969 PH/52 Two bugs concerned with error handling when the smtp transport is
2972 (i) Exim was not creating retry information for temporary errors given
2973 for individual recipients after the DATA command when the smtp transport
2974 was used in LMTP mode. This meant that they could be retried too
2975 frequently, and not timed out correctly.
2977 (ii) Exim was setting the flag that allows error details to be returned
2978 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2979 recipients that were returned after the DATA command.
2981 PH/53 This is related to PH/52, but is more general: for any failing address,
2982 when detailed error information was permitted to be returned to the
2983 sender, but the error was temporary, then after the final timeout, only
2984 "retry timeout exceeded" was returned. Now it returns the full error as
2985 well as "retry timeout exceeded".
2987 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2988 do this, and (what is worse) MTAs that accept it.
2990 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2991 will now be deprecated.
2993 PH/56 New os.c-cygwin from the Cygwin maintainer.
2995 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2996 formats (previously only available in exim formats via -bpr, -bpru,
2997 and -bpra. Now also available in native and exiqgrep formats)
2999 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3000 with very large, slow to parse queues
3002 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3004 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3006 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3007 responses to authentication challenges, though it was showing the
3008 challenges; (ii) I've removed the CR characters from the debug output for
3011 PH/58 Allow for the insertion of a newline as well as a space when a string
3012 is turned into more than one encoded-word during RFC 2047 encoding. The
3013 Sieve code now uses this.
3015 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3016 data_4xx, lost_connection, tls_required.
3018 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3019 message was being sent as an SMTP response.
3021 PH/61 Add -l and -k options to exicyclog.
3023 PH/62 When verifying, if an address was redirected to one new address, so that
3024 verification continued, and the new address failed or deferred after
3025 having set something in $address_data, the value of $address_data was not
3026 passed back to the ACL. This was different to the case when no
3027 redirection occurred. The value is now passed back in both cases.
3029 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3030 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3031 use login_cap.h, so on its own it isn't the distinguishing feature. The
3032 new name refers directly to the setclassresources() function.
3034 PH/65 Added configuration files for NetBSD3.
3036 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3038 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3039 is preferred over IPv4.
3041 PH/68 The bounce_return_message and bounce_return_body options were not being
3042 honoured for bounces generated during the reception of non-SMTP messages.
3043 In particular, this applied to messages rejected by the ACL. This bug has
3044 been fixed. However, if bounce_return_message is true and bounce_return_
3045 body is false, the headers that are returned for a non-SMTP message
3046 include only those that have been read before the error was detected.
3047 (In the case of an ACL rejection, they have all been read.)
3049 PH/69 The HTML version of the specification is now built in a directory called
3050 spec_html instead of spec.html, because the latter looks like a path with
3051 a MIME-type, and this confuses some software.
3053 PH/70 Catch two compiler warnings in sieve.c.
3055 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3056 function verify_get_ident() calls ip_connect() to connect a socket, but
3057 if the "connect()" function timed out, ip_connect() used to close the
3058 socket. However, verify_get_ident() also closes the socket later, and in
3059 between Exim writes to the log, which may get opened at this point. When
3060 the socket was closed in ip_connect(), the log could get the same file
3061 descriptor number as the socket. This naturally causes chaos. The fix is
3062 not to close the socket in ip_connect(); the socket should be closed by
3063 the function that creates it. There was only one place in the code where
3064 this was missing, in the iplookup router, which I don't think anybody now
3065 uses, but I've fixed it anyway.
3067 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3068 well as to direct DNS lookups. Otherwise the handling of names in host
3069 lists is inconsistent and therefore confusing.
3075 PH/01 Two changes to the default runtime configuration:
3077 (1) Move the checks for relay_from_hosts and authenticated clients from
3078 after to before the (commented out) DNS black list checks.
3080 (2) Add control=submission to the relay_from_hosts and authenticated
3081 clients checks, on the grounds that messages accepted by these
3082 statements are most likely to be submissions.
3084 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3086 (1) Generate an error if the third argument for the ${prvs expansion is
3089 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3092 (3) Reset the variables that are obtained from the first argument of
3093 ${prvscheck and used in the second argument before leaving the code,
3094 because their memory is reclaimed, so using them afterwards may do
3097 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3098 one (it's much easier than Tom thought :-).
3100 (5) Because of (4), we can now allow for the use of $prvscheck_result
3101 inside the third argument.
3103 PH/03 For some reason, the default setting of PATH when running a command from
3104 a pipe transport was just "/usr/bin". I have changed it to
3107 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3108 anything to be listed in the output from -bV.
3110 PH/05 When a filter generated an autoreply, the entire To: header line was
3111 quoted in the delivery log line, like this:
3113 => >A.N.Other <ano@some.domain> <original@ddress> ...
3115 This has been changed so that it extracts the operative address. There
3116 may be more than one such address. If so, they are comma-separated, like
3119 => >ano@some.domain,ona@other.domain <original@ddress> ...
3121 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3122 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3123 not being looked up in the rDNS to get a host name, Exim was showing the
3124 IP address twice in Received: lines, even though the IP addresses were
3125 identical. For example:
3127 Received: from [1.2.3.4] (helo=[1.2.3.4])
3129 However, if the real host name was known, it was omitting the HELO data
3130 if it matched the actual IP address. This has been tidied up so that it
3131 doesn't show the same IP address twice.
3133 PH/07 When both +timestamp and +memory debugging was on, the value given by
3134 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3135 called by the debug printing, thereby overwriting the timestamp buffer.
3136 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3138 PH/08 When the original message was included in an autoreply transport, it
3139 always said "this is a copy of the message, including all the headers",
3140 even if body_only or headers_only was set. It now gives an appropriate
3143 PH/09 Applied a patch from the Sieve maintainer which:
3145 o fixes some comments
3146 o adds the (disabled) notify extension core
3147 o adds some debug output for the result of if/elsif tests
3148 o points to the current vacation draft in the documentation
3149 and documents the missing references header update
3153 o fixes a bug in processing the envelope test (when testing
3154 multiple envelope elements, the last element determinted the
3157 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3158 Electronic Mail") by including:
3160 Auto-submitted: auto-generated
3162 in the messages that it generates (bounce messages and others, such as
3163 warnings). In the case of bounce messages for non-SMTP mesages, there was
3164 also a typo: it was using "Auto_submitted" (underscore instead of
3165 hyphen). Since every message generated by Exim is necessarily in response
3166 to another message, thes have all been changed to:
3168 Auto-Submitted: auto-replied
3170 in accordance with these statements in the RFC:
3172 The auto-replied keyword:
3174 - SHOULD be used on messages sent in direct response to another
3175 message by an automatic process,
3177 - MUST NOT be used on manually-generated messages,
3179 - MAY be used on Delivery Status Notifications (DSNs) and Message
3180 Disposition Notifications (MDNs),
3182 - MUST NOT be used on messages generated by automatic or periodic
3183 processes, except for messages which are automatic responses to
3186 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3187 to the default Received: header definition.
3189 PH/12 Added log selector acl_warn_skipped (default on).
3191 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3192 variables because (a) they are in the wrong storage pool and (b) even if
3193 they were copied, it wouldn't work properly because of the caching.
3195 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3196 checking when decoding. Apparently there are clients that generate
3197 overlong encoded strings. Why am I not surprised?
3199 PH/15 If the first argument of "${if match_address" was not empty, but did not
3200 contain an "@" character, Exim crashed. Now it writes a panic log message
3201 and treats the condition as false.
3203 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3205 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3206 extension "envelope-auth". The code is finished and in agreement with
3207 other implementations, but there is no documentation so far and in fact,
3208 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3209 not changing the active code.
3211 Print executed "if" and "elsif" statements when debugging is used. This
3212 helps a great deal to understand what a filter does.
3214 Document more things not specified clearly in RFC3028. I had all this
3215 sorted out, when out of a sudden new issues came to my mind. Oops."
3217 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3220 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3221 canonical form (as documented). However, after a host name lookup from
3222 the IP address, check_host() was doing a simple string comparison with
3223 addresses acquired from the DNS when checking that the found name did
3224 have the original IP as one of its addresses. Since any found IPv6
3225 addresses are likely to be in abbreviated form, the comparison could
3226 fail. Luckily, there already exists a function for doing the comparison
3227 by converting both addresses to binary, so now that is used instead of
3228 the text comparison.
3230 PH/20 There was another similar case to PH/19, when a complete host name was
3231 given in a host list; looking up its IP address could give an abbreviated
3232 form, whereas the current host's name might or might not be abbreviated.
3233 The same fix has been applied.
3239 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3240 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3243 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3244 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3246 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3248 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3249 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3250 Also, for "Auto-Submitted", treat anything other than "no" as
3251 non-personal, in accordance with RFC 3834. (Previously it treated
3252 anything starting "auto-" as non-personal.)
3254 TF/01 The control=submission/name=... option had a problem with syntax
3255 errors if the name included a slash character. The /name= option
3256 now slurps the rest of the string, so it can include any characters
3257 but it must come last in the list of options (after /sender_retain
3260 PH/05 Some modifications to the interface to the fake nameserver for the new
3268 TK/01 Added the "success_on_redirect" address verification option. See
3269 NewStuff for rationale and an example.
3271 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3273 PH/02 Patch to exigrep to allow it to work on syslog lines.
3275 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3276 fread() to skip over the body file's header line, because in Cygwin the
3277 header line is locked and is inaccessible.
3279 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3280 co-exist for some time) to make it clear that it is the Exim ID that is
3281 referenced, not the Message-ID: header line.
3283 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3284 string_format() function, because snprintf() does not exist on all
3287 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3288 ${stat: expansion item.
3290 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3291 protocol synchronization error", to keep the pedants happy.
3293 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3294 well as for IRIX systems, when gcc is being used. See the host.c source
3297 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3299 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3302 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3303 transport and to the smtp transport in LMTP mode.
3305 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3307 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3308 run a filter in a subprocess. This could lead to confusion in subsequent
3309 lookups in the parent process. There should also be a search_tidyup() at
3310 the end of the subprocess.
3312 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3313 only if the host matched helo_try_verify_hosts, which caused the
3314 verification to occur when the EHLO/HELO command was issued. The ACL just
3315 tested the remembered result. Now, if a previous verification attempt has
3316 not happened, "verify = helo" does it there and then.
3318 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3320 TK/03 Fix log output including CR from clamd.
3322 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3323 exist provoked a memory error which could cause a segfault.
3325 PH/15 Installed PCRE 6.2
3327 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3329 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3330 of the problem. Specifically, suggested +O2 rather than +O1 for the
3333 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3335 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3336 record, the port was not correctly logged when the outgoing_port log
3337 selector was set (it logged the transort's default port).
3339 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3340 fallback_hosts, and "hosts" in the smtp transport.
3342 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3343 host errors such as "Connection refused".
3345 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3346 authentication with radiusclient 0.4.9:
3348 - Error returned from rc_read_config was caught wrongly
3349 - Username/password not passed on to radius server due to wrong length.
3351 The presumption is that some radiusclient API changes for 4.51/PH/17
3352 were not taken care of correctly. The code is still untested by me (my
3353 Linux distribution still has 0.3.2 of radiusclient), but it was
3354 contributed by a Radius user.
3356 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3357 expanding the "port" option of the smtp transport.
3359 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3360 while reading a MIME header. Thanks to Tom Hughes for a patch.
3362 PH/24 Include config.h inside local_scan.h so that configuration settings are
3365 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3366 that in an ACL for RCPT (for example), you can examine exactly what was
3369 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3370 commands, but it was not correctly comparing the address with the actual
3371 client host address. Thus, it would show the EHLO address in Received:
3372 header lines when this was not necessary.
3374 PH/27 Added the % operator to ${eval:}.
3376 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3377 it should be ignoring failures (because with -C, for example, it has lost
3378 privilege). It wasn't ignoring creation failures other than "already
3381 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3384 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3385 because an input line was too long, either on its own, or by virtue of
3386 too many continuations, the temporary file was not being removed, and the
3387 return code was incorrect.
3389 PH/31 Missing "BOOL" in function definition in filtertest.c.
3391 PH/32 Applied Sieve patches from the maintainer.
3393 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3395 PH/33 Added "verify = not_blind".
3397 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3398 Local/Makefile (with some defaults set). These are used in built scripts
3399 such as exicyclog, but they have never been used in the exim_install
3400 script (though there are many overriding facilities there). I have
3401 arranged that the exim_install script now takes note of these two
3404 PH/35 Installed configuration files for Dragonfly.
3406 PH/36 When a locally submitted message by a trusted user did not contain a
3407 From: header, and the sender address was obtained from -f or from an SMTP
3408 MAIL command, and the trusted user did not use -F to supply a sender
3409 name, $originator_name was incorrectly used when constructing a From:
3410 header. Furthermore, $originator_name was used for submission mode
3411 messages from external hosts without From: headers in a similar way,
3412 which is clearly wrong.
3414 PH/37 Added control=suppress_local_fixups.
3416 PH/38 When log_selector = +received_sender was set, and the addition of the
3417 sender made the log line's construction buffer exactly full, or one byte
3418 less than full, an overflow happened when the terminating "\n" was
3421 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3422 when the result of a list match is failure because a DNS lookup failed.
3424 PH/40 RM_COMMAND is now used in the building process.
3426 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3427 the "build-* directories that it finds.
3429 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3430 address in a domain literal was a prefix of an interface address.
3432 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3433 when verifying a sender address, unless rewrite_headers is false.
3435 PH/44 Wrote a long comment about why errors_to addresses are verified as
3436 recipients, not senders.
3438 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3439 the ratelimit ACL was added.
3441 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3443 PH/46 Added extra information about PostgreSQL errors to the error string.
3445 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3446 avoiding the need to install special zones in a real server. This is
3447 backwards compatible; if it can't find the fake resolver, it drops back.
3448 Thus, both old and new test suites can be run.
3450 TF/02 Added util/ratelimit.pl
3452 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3453 clock is set back in time.
3455 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3456 Candler <B.Candler@pobox.com>.
3458 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3459 OK for addresses that are the result of redirections.
3461 PH/48 A number of further additions for the benefit of the new test suite,
3462 including a fake gethostbyname() that interfaces to the fake DNS resolver
3465 TF/06 The fix for widen_domains has also been applied to qualify_single and
3466 search_parents which are the other dnslookup options that can cause
3469 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3472 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3474 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3475 service" if it encountered an MX record with an empty target hostname.
3476 The message is now "an MX or SRV record indicated no SMTP service".
3478 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3479 if the DNS of the sending site is misconfigured. This is quite a
3480 common situation. This change restores the behaviour of treating a
3481 helo verification defer as a failure.
3483 PH/51 If self=fail was set on a router, the bounce message did not include the
3484 actual error message.
3490 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3492 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3493 command itself ended in error, the underlying message about the transport
3494 filter timeout was being overwritten with the pipe command error. Now the
3495 underlying error message should be appended to the second error message.
3497 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3499 PH/02 Reduce the amount of output that "make" produces by default. Full output
3500 can still be requested.
3502 PH/03 The warning log line about a condition test deferring for a "warn" verb
3503 was being output only once per connection, rather than after each
3504 occurrence (because it was using the same function as for successful
3505 "warn" verbs). This seems wrong, so I have changed it.
3507 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3508 it should not have, which might have caused a crash in the right
3509 circumstances, but probably never did.
3511 PH/04 Installed a modified version of Tony Finch's patch to make submission
3512 mode fix the return path as well as the Sender: header line, and to
3513 add a /name= option so that you can make the user's friendly name appear
3516 TF/03 Added the control = fakedefer ACL modifier.
3518 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3519 Mark Lowes for thorough testing.
3521 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3523 TK/03 Merged latest SRS patch from Miles Wilton.
3525 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3526 with the definition in sysexits.h (which is #included earlier).
3527 Fortunately, Exim does not actually use EX_OK. The code used to try to
3528 preserve the sysexits.h value, by assumimg that macro definitions were
3529 scanned for macro replacements. I have been disabused of this notion,
3530 so now the code just undefines EX_OK before #including unistd.h.
3532 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3533 in the smtp transport. When a block could not be written in a single
3534 write() function, the timeout was being re-applied to each part-write.
3535 This seems wrong - if the receiver was accepting one byte at a time it
3536 would take for ever. The timeout is now adjusted when this happens. It
3537 doesn't have to be particularly precise.
3539 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3540 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3542 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3543 without a domain if the check to <postmaster@domain> fails.
3545 SC/01 Eximstats: added -xls and the ability to specify output files
3546 (patch written by Frank Heydlauf).
3548 SC/02 Eximstats: use FileHandles for outputing results.
3550 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3552 SC/04 Eximstats: fixed display of large numbers with -nvr option
3554 SC/05 Eximstats: fixed merging of reports with empty tables.
3556 SC/06 Eximstats: added the -include_original_destination flag
3558 SC/07 Eximstats: removed tabs and trailing whitespace.
3560 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3562 TK/06 MBOX spool code: Add real "From " MBOX separator line
3563 so the .eml file is really in mbox format (even though
3564 most programs do not really care). Patch from Alex Miller.
3566 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3567 The latter is generated from $received_to and is only set if the
3568 message has one envelope recipient. SA can use these headers,
3569 obviously out-of-the-box. Patch from Alex Miller.
3571 PH/08 The ${def test on a variable was returning false if the variable's
3572 value was "0", contrary to what the specification has always said!
3573 The result should be true unless the variable is empty.
3575 PH/09 The syntax error of a character other than { following "${if
3576 def:variable_name" (after optional whitespace) was not being diagnosed.
3577 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3578 accidental colon was present, for example, could give incorrect results.
3580 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3581 result is used (not including appendfile, where other changes are about
3584 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3585 This involved changing a lot of size variables from int to off_t. It
3586 should work with maildirs and everything.
3588 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3589 spamd dying while we are connected to it.
3591 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3594 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3595 function for BDB 4.3.
3597 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3599 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3600 the sources for PCRE have been re-arranged and more files are now
3603 PH/15 The code I had for printing potentially long long variables in PH/11
3604 above was not the best (it lost precision). The length of off_t variables
3605 is now inspected at build time, and an appropriate printing format (%ld
3606 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3607 to be "long long int" or "long int". This is needed for the internal
3608 formatting function string_vformat().
3610 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3611 the configuration file to be ":syslog", then the script "guesses" where
3612 the logs files are, rather than using the compiled in default. In our
3613 case the guess is not the same as the compiled default, so the script
3614 suddenly stopped working when I started to use syslog. The patch checks
3615 to see if log_file_path is "". If so, it attempts to read it from exim
3616 with no configuration file to get the compiled in version, before it
3617 falls back to the previous guessing code."
3619 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3620 implementing BATV in an Exim configuration. See NewStuff for the gory
3623 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3624 Makefile that are specific to HP-UX.
3626 PH/18 If the "use_postmaster" option was set for a recipient callout together
3627 with the "random" option, the postmaster address was used as the MAIL
3628 FROM address for the random test, but not for the subsequent recipient
3629 test. It is now used for both.
3631 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3632 patch removes a few documentation additions to RFC 3028, because the
3633 latest draft now contains them. It adds the new en;ascii-case comparator
3634 and a new error check for 8bit text in MIME parts. Comparator and
3635 require names are now matched exactly. I enabled the subaddress
3636 extension, but it is not well tested yet (read: it works for me)."
3638 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3639 rework some of the code of TK/09 above to avoid the hardwired use of
3640 "%lld" and "long long". Replaced the call to snprintf() with a call to
3643 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3644 records point to non-existent hosts", "retry timeout exceeded", and
3645 "retry time not reached for any host after a long failure period".
3647 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3648 experimental DomainKeys support:
3650 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3651 (2) On an error such as an illegally used "control", the wrong name for
3652 the control was given.
3654 These problems did NOT occur unless DomainKeys support was compiled.
3656 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3658 PH/24 Added ${if match_ip condition.
3660 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3661 databases so that it will be absolutely obvious if a crash occurs in the
3662 DB library. This is a regular occurrence (often caused by mis-matched
3665 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3666 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3667 picked up on a user's system that detects such things. There doesn't seem
3668 to be a gcc warning option for this - only an attribute that has to be
3669 put on the function's prototype. It seems that in Fedora Core 4 they have
3670 set this on a number of new functions. No doubt there will be more in due
3673 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3674 specify a transport. However, if an address that was verified by such a
3675 router was the subject of a callout, Exim crashed because it tried to
3676 read the rcpt_include_affixes from the non-existent transport. Now it
3677 just assumes that the setting of that option is false. This bug was
3678 introduced by 4.51/PH/31.
3680 PH/28 Changed -d+all to exclude +memory, because that information is very
3681 rarely of interest, but it makes the output a lot bigger. People tend to
3682 do -d+all out of habit.
3684 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3685 code in os-type was giving problems when libc.so lives in lib64, like on
3688 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3689 aren't the modern standard, and it seems that some systems' include files
3690 don't always have them. Exim was already checking for some of the newer
3691 ones like T_AAAA, and defining it itself. I've added checks for all the
3692 record types that Exim uses.
3694 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3695 not automatically generating a new one, as it is supposed to. This
3696 prevented TLS from working. If the file did exist, but contained invalid
3697 data, a new version was generated, as expected. It was only the case of a
3698 non-existent file that was broken.
3700 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3701 with a change in libdomainkeys > 0.64.
3703 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3704 from DNS. If the selector record carries the flag, it now has
3705 precedence over the domain-wide flag.
3707 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3709 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3710 the use of a port name that isn't defined in /etc/services) occurred, the
3711 message was deferred as in a normal delivery, and thus remained on the
3712 spool, instead of being failed because of the mua_wrapper setting. This
3713 is now fixed, and I tidied up some of the mua_wrapper messages at the
3716 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3717 the messages in a hash of arrays rather than using individual hashes.
3718 This is a bit cleaner and results in dramatic memory savings, albeit
3719 at a slight CPU cost.
3721 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3722 as requested by Marc Sherman.
3724 SC/10 Eximstats: added histograms for user specified patterns as requested
3727 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3729 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3730 fopen() in the content-scanning modules that did not already have it.
3736 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3737 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3739 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3741 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3743 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3744 to test Sieve filters that use "vacation".
3746 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3747 that changes the way the GnuTLS parameters are stored in the cache file.
3748 The new format can be generated externally. For backward compatibility,
3749 if the data in the cache doesn't make sense, Exim assumes it has read an
3750 old-format file, and it generates new data and writes a new file. This
3751 means that you can't go back to an older release without removing the
3754 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3755 work if there are any delivery delays because "one_time" forces the
3756 parent to be marked "delivered", so its unseen clone is never tried
3757 again. For this reason, Exim now forbids the simultaneous setting of
3760 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3761 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3762 entry if you want to know the details. The fix, however, neglected to
3763 consider the case when local delivery batching is involved. The test for
3764 "previously delivered" was not happening when checking to see if an
3765 address could be batched with a previous (undelivered) one; under
3766 certain circumstances this could lead to multiple deliveries to the same
3769 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3770 in its include files, and this causes problems building Exim.
3772 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3773 header_syntax) but Exim was just ignoring anything given after a slash.
3774 In particular, this caused confusion with an attempt to use "verify =
3775 reverse_host_lookup/defer_ok". An error is now given when options are
3776 supplied for verify items that do not have them. (Maybe reverse_host_
3777 lookup should have a defer_ok option, but that's a different point.)
3779 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3780 defined by RFC 821) to 2048, because there were problems with some AUTH
3781 commands, and RFC 1869 says the size should be increased for extended
3782 SMTP commands that take arguments.
3784 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3787 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3788 "unknown" error; now it says that the functionality isn't in the binary.
3790 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3791 an address' error message when a string expansion fails (syntax or
3792 whatever). Otherwise the password may appear in the log. Following change
3793 PH/42 below, there is no longer a chance of it appearing in a bounce
3796 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3798 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3799 of its addresses was ever tried. (Bugzilla bug #2.)
3801 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3802 the result incorrectly in the debug output. (It correctly added a newline
3803 to what was transported.)
3805 TF/01 Added $received_time.
3807 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3808 commented out examples of how to interface to a virus scanner and to
3809 SpamAssassin. Also added commented examples of av_scanner and
3810 spamd_address settings.
3812 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3813 and controls are allowed in which ACLs. There were a couple of minor
3814 errors. Some of the entries in the conditions table (which is a table of
3815 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3816 negation of where the condition IS allowed.
3818 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3820 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3821 header file does not have a version number, so I've had to invent a new
3822 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3823 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3824 radiusclient), but it was contributed by a Radius user.
3826 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3827 files or directories, for OpenSSL.
3829 PH/19 When an Exim process that is running as root has to create an Exim log
3830 file, it does so in a subprocess that runs as exim:exim so as to get the
3831 ownership right at creation (otherwise, other Exim processes might see
3832 the file with the wrong ownership). There was no test for failure of this
3833 fork() call, which would lead to the process getting stuck as it waited
3834 for a non-existent subprocess. Forks do occasionally fail when resources
3835 run out. I reviewed all the other calls to fork(); they all seem to check
3838 PH/20 When checking for unexpected SMTP input at connect time (before writing
3839 the banner), Exim was not dealing correctly with a non-positive return
3840 from the read() function. If the client had disconnected by this time,
3841 the result was a log entry for a synchronization error with an empty
3842 string after "input=" when read() returned zero. If read() returned -1
3843 (an event I could not check), uninitialized data bytes were printed.
3844 There were reports of junk text (parts of files, etc) appearing after
3847 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3849 PH/22 Added support for macro redefinition, and (re)definition in between
3850 driver and ACL definitions.
3852 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3853 forgetting to use the resulting value; it was using the unexpanded value.
3855 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3856 hadn't been configured. The fix is from Juergen Kreileder, who
3857 understands it better than I do:
3859 "Here's what I see happening with three configured cyrus_sasl
3860 authenticators configured (plain, login, cram-md5):
3862 On startup auth_cyrus_sasl_init() gets called for each of these.
3863 This means three calls to sasl_listmech() without a specified mech_list.
3864 => SASL tests which mechs of all available mechs actually work
3865 => three warnings about OTP not working
3866 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3868 With the patch, sasl_listmech() also gets called three times. But now
3869 SASL's mech_list option is set to the server_mech specified in the the
3870 authenticator. Or in other words, the answer from sasl_listmech()
3871 gets limited to just the mech you're testing for (which is different
3873 => the return list contains just 'plain' or 'login', 'cram-md5' or
3874 nothing depending on the value of ob->server_mech.
3876 I've just tested the patch: Authentication still works fine,
3877 unavailable mechs specified in the exim configuration are still
3878 caught, and the auth.log warnings about OTP are gone."
3880 PH/25 When debugging is enabled, the contents of the command line are added
3881 to the debugging output, even when log_selector=+arguments is not
3884 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3885 answer is "GNU", and only if the return is "GNU/something" is the answer
3888 PH/27 $acl_verify_message is now set immediately after the failure of a
3889 verification in an ACL, and so is available in subsequent modifiers. In
3890 particular, the message can be preserved by coding like this:
3892 warn !verify = sender
3893 set acl_m0 = $acl_verify_message
3895 Previously, $acl_verify_message was set only while expanding "message"
3896 and "log_message" when a very denied access.
3898 PH/28 Modified OS/os.c-Linux with
3900 -#ifndef OS_LOAD_AVERAGE
3901 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3903 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3904 nomenclature these days.)
3906 PH/29 Installed patch from the Sieve maintainer that adds the options
3907 sieve_useraddress and sieve_subaddress to the redirect router.
3909 PH/30 In these circumstances:
3910 . Two addresses routed to the same list of hosts;
3911 . First host does not offer TLS;
3912 . First host accepts first address;
3913 . First host gives temporary error to second address;
3914 . Second host offers TLS and a TLS session is established;
3915 . Second host accepts second address.
3916 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3917 and peerdn, if requested) that were in fact used only for the second
3920 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3921 attention to any local part prefix or suffix that was matched by the
3922 router that accepted the address. It now behaves in the same way as it
3923 does for delivery: the affixes are removed from the local part unless
3924 rcpt_include_affixes is set on the transport.
3926 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3927 timeout during the DATA phase of an incoming message.
3929 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3930 applied a patch sanctioned by the Sieve maintainer.
3932 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3933 the uid or gid is negative. A case of a negative gid caused this to be
3934 noticed. The fix allows for either to be negative.
3936 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3937 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3940 PH/36 The change PH/12 above was broken. Fixed it.
3942 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3943 the grounds that routing the same address twice would always produce the
3944 same answer. This might have been true once, but it is certainly no
3945 longer true now. Routing a child address may depend on the previous
3946 routing that produced that child. Some complicated redirection strategies
3947 went wrong when messages had multiple recipients, and made Exim's
3948 behaviour dependent on the order in which the addresses were given.
3950 I have moved the duplicate checking until after the routing is complete.
3951 Exim scans the addresses that are assigned to local and remote
3952 transports, and removes any duplicates. This means that more work will be
3953 done, as duplicates will always all be routed, but duplicates are
3954 presumably rare, so I don't expect this is of any significance.
3956 For deliveries to pipes, files, and autoreplies, the duplicate checking
3957 still happens during the routing process, since they are not going to be
3960 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3961 It corrects a timeout issue with spamd. This is Ian's comment: "The
3962 background is that sometimes spamd either never reads data from a
3963 connection it has accepted, or it never writes response data. The exiscan
3964 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3965 blindly assumes that writes won't block so it may never time out."
3967 PH/39 Allow G after quota size as well as K and M.
3969 PH/40 The value set for $authenticated_id in an authenticator may not contain
3970 binary zeroes or newlines because the value is written to log lines and
3971 to spool files. There was no check on this. Now the value is run through
3972 the string_printing() function so that such characters are converted to
3973 printable escape sequences.
3975 PH/41 $message_linecount is a new variable that contains the total number of
3976 lines in the message. Compare $body_linecount, which is the count for the
3979 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3980 bounce and delay warning messages, except in certain special cases, which
3983 (a) An SMTP error message from a remote host;
3984 (b) A message specified in a :fail: redirection;
3985 (c) A message specified in a "fail" command in a system filter;
3986 (d) A message specified in a FAIL return from the queryprogram router;
3987 (e) A message specified by the cannot_route_message router option.
3989 In these cases only, Exim does include the error details in bounce and
3990 warning messages. There are also a few cases where bland messages such
3991 as "unrouteable address" or "local delivery error" are given.
3993 PH/43 $value is now also set for the "else" part of a ${run expansion.
3995 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3996 being worked on, but at least Exim now implements the latest version to
3999 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4000 process to complete was treated as a delivery failure, a timeout while
4001 writing the message to the pipe was logged, but erroneously treated as a
4002 successful delivery. Such timeouts include transport filter timeouts. For
4003 consistency with the overall process timeout, these timeouts are now
4004 treated as errors, giving rise to delivery failures by default. However,
4005 there is now a new Boolean option for the pipe transport called
4006 timeout_defer, which, if set TRUE, converts the failures into defers for
4007 both kinds of timeout. A transport filter timeout is now identified in
4010 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4011 systems where "make" and "gmake" are different, calling "gmake" at top
4012 level broke things. I've arranged for the value of $(MAKE) to be passed
4013 from the Makefile to this script so that it can call the same version of
4017 A note about Exim versions 4.44 and 4.50
4018 ----------------------------------------
4020 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4021 changes of various kinds. As a consequence, a big documentation update was
4022 needed. This delayed the release for rather longer than seemed good, especially
4023 in the light of a couple of (minor) security issues. Therefore, the changes
4024 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4025 release. So 4.44 and 4.50 are in effect two different branches that both start
4028 I have left the 4.50 change log unchanged; it contains all the changes since
4029 4.43. The change log for 4.44 is below; many of its items are identical to
4030 those for 4.50. This seems to be the most sensible way to preserve the
4031 historical information.
4037 1. Minor wording change to the doc/README.SIEVE file.
4039 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4040 computation of the current number of files was incorrect.
4042 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4043 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4046 4. Give more explanation in the error message when the command for a transport
4047 filter fails to execute.
4049 5. There are several places where Exim runs a non-Exim command in a
4050 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4051 was being done only for the command run by the queryprogram router. It is
4052 now done for all such subprocesses. The other cases are: ${run, transport
4053 filters, and the commands run by the lmtp and pipe transports.
4055 6. Added CONFIGURE_GROUP build-time option.
4057 7. Some older OS have a limit of 256 on the maximum number of file
4058 descriptors. Exim was using setrlimit() to set 1000 as a large value
4059 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4060 systems. I've change it so that if it can't get 1000, it tries for 256.
4062 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4063 was an oversight, and furthermore, ever since the addition of extra
4064 controls (e.g. 4.43/32), the checks on when to allow different forms of
4065 "control" were broken. There should now be diagnostics for all cases when a
4066 control that does not make sense is encountered.
4068 9. Added the /retain_sender option to "control=submission".
4070 10. $recipients is now available in the predata ACL (oversight).
4072 11. Tidy the search cache before the fork to do a delivery from a message
4073 received from the command line. Otherwise the child will trigger a lookup
4074 failure and thereby defer the delivery if it tries to use (for example) a
4075 cached ldap connection that the parent has called unbind on.
4077 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4078 of $address_data from the recipient verification was clobbered by the
4079 sender verification.
4081 13. The value of address_data from a sender verification is now available in
4082 $sender_address_data in subsequent conditions in the ACL statement.
4084 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4086 15. Added a new option "connect=<time>" to callout options, to set a different
4089 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4090 was its contents. (It was OK if the option was not defined at all.)
4092 17. A "Completed" log line is now written for messages that are removed from
4093 the spool by the -Mrm option.
4095 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4096 information about exactly what failed.
4098 19. Added -dd to debug only the daemon process.
4100 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4101 handles timeouts, both on the server side and network timeouts. Renamed the
4102 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4104 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4105 It is now set to "smtps".
4107 22. $host_address is now set to the target address during the checking of
4108 ignore_target_hosts.
4110 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4111 being passed; this would have caused $sender_host_name to have been used if
4112 matching the list had actually called for a host name (not very likely,
4113 since this list is usually IP addresses). A host name is now passed as
4116 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4117 code that specifies a non-restarting handler (typically sigaction() in
4118 modern systems) in an attempt to fix a rare and obscure crash bug.
4120 25. Narrowed the window for a race in the daemon that could cause it to ignore
4121 SIGCHLD signals. This is not a major problem, because they are used only to
4122 wake it up if nothing else does.
4124 26. A malformed maildirsize file could cause Exim to calculate negative values
4125 for the mailbox size or file count. Odd effects could occur as a result.
4126 The maildirsize information is now recalculated if the size or filecount
4129 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4130 support for a long time. Removed HAVE_SYS_VFS_H.
4132 28. Installed the latest version of exipick from John Jetmore.
4134 29. In an address list, if the pattern was not a regular expression, an empty
4135 subject address (from a bounce message) matched only if the pattern was an
4136 empty string. Non-empty patterns were not even tested. This was the wrong
4137 because it is perfectly reasonable to use an empty address as part of a
4138 database query. An empty address is now tested by patterns that are
4139 lookups. However, all the other forms of pattern expect the subject to
4140 contain a local part and a domain, and therefore, for them, an empty
4141 address still always fails if the pattern is not itself empty.
4143 30. Exim went into a mad DNS loop when attempting to do a callout where the
4144 host was specified on an smtp transport, and looking it up yielded more
4145 than one IP address.
4147 31. Re-factored the code for checking spool and log partition space into a
4148 function that finds that data and another that does the check. The former
4149 is then used to implement four new variables: $spool_space, $log_space,
4150 $spool_inodes, and $log_inodes.
4152 32. The RFC2047 encoding function was originally intended for short strings
4153 such as real names; it was not keeping to the 75-character limit for
4154 encoded words that the RFC imposes. It now respects the limit, and
4155 generates multiple encoded words if necessary. To be on the safe side, I
4156 have increased the buffer size for the ${rfc2047: expansion operator from
4159 33. It is now permitted to omit both strings after an "if" condition; if the
4160 condition is true, the result is "true". As before, when the second string
4161 is omitted, a false condition yields an empty string. This makes it less
4162 cumbersome to write custom ACL and router conditions.
4164 34. Failure to deliver a bounce message always caused it to be frozen, even if
4165 there was an errors_to setting on the router. The errors_to setting is now
4168 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4169 canonical form (fully expanded) before being placed in
4170 $sender_host_address.
4172 36. The table in the code that translates DNS record types into text (T_A to
4173 "A" for instance) was missing entries for NS and CNAME. It is just possible
4174 that this could have caused confusion if both these types were looked up
4175 for the same domain, because the text type is used as part of Exim's
4176 per-process caching. But the chance of anyone hitting this buglet seems
4179 37. The dnsdb lookup has been extended in a number of ways.
4181 (1) There is a new type, "zns", which walks up the domain tree until it
4182 finds some nameserver records. It should be used with care.
4184 (2) There is a new type, "mxh", which is like "mx" except that it returns
4185 just the host names, not the priorities.
4187 (3) It is now possible to give a list of domains (or IP addresses) to be
4188 looked up. The behaviour when one of the lookups defers can be
4189 controlled by a keyword.
4191 (4) It is now possible to specify the separator character for use when
4192 multiple records are returned.
4194 38. The dnslists ACL condition has been extended: it is now possible to supply
4195 a list of IP addresses and/or domains to be looked up in a particular DNS
4198 39. Added log_selector=+queue_time_overall.
4200 40. When running the queue in the test harness, wait just a tad after forking a
4201 delivery process, to get repeatability of debugging output.
4203 41. Include certificate and key file names in error message when GnuTLS fails
4204 to set them up, because the GnuTLS error message doesn't include the name
4205 of the failing file when there is a problem reading it.
4207 42. Allow both -bf and -bF in the same test run.
4209 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4211 44. The "Exiscan patch" is now merged into the mainline Exim source.
4213 45. Sometimes the final signoff response after QUIT could fail to get
4214 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4215 < 0 before doing a fflush(). This bug looks as though it goes back to the
4216 introduction of TLS in release 3.20, but "sometimes" must have been rare
4217 because the tests only now provoked it.
4219 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4220 (this can affect the format of dates).
4222 47. exim_tidydb, when checking for the continued existence of a message for
4223 which it has found a message-specific retry record, was not finding
4224 messages that were in split spool directories. Consequently, it was
4225 deleting retry records that should have stayed in existence.
4227 48. Steve fixed some bugs in eximstats.
4229 49. The SPA authentication driver was not abandoning authentication and moving
4230 on to the next authenticator when an expansion was forced to fail,
4231 contradicting the general specification for all authenticators. Instead it
4232 was generating a temporary error. It now behaves as specified.
4234 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4235 (the order specifies the preference for clients). The order is now AES256,
4236 AES128, 3DES, ARCFOUR128.
4238 51. Small patch to Sieve code - explicitly set From: when generating an
4241 52. Exim crashed if a remote delivery caused a very long error message to be
4242 recorded - for instance if somebody sent an entire SpamAssassin report back
4243 as a large number of 550 error lines. This bug was coincidentally fixed by
4244 increasing the size of one of Exim's internal buffers (big_buffer) that
4245 happened as part of the Exiscan merge. However, to be on the safe side, I
4246 have made the code more robust (and fixed the comments that describe what
4249 53. Now that there can be additional text after "Completed" in log lines (if
4250 the queue_time_overall log selector is set), a one-byte patch to exigrep
4251 was needed to allow it to recognize "Completed" as not the last thing in
4254 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4255 patch that reportedly fixes this has been added. I am not expert enough to
4256 create a test for it. This is what the patch creator wrote:
4258 "I found a little strange behaviour of ldap code when working with
4259 Windows 2003 AD Domain, where users was placed in more than one
4260 Organization Units. When I tried to give exim partial DN, the exit code
4261 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4262 But simultaneously result of request was absolutely normal ldap result,
4263 so I produce this patch..."
4265 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4266 REFERENCE, so I have modified the code to exclude the patch when that macro
4269 55. Some experimental protocols are using DNS PTR records for new purposes. The
4270 keys for these records are domain names, not reversed IP addresses. The
4271 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4272 leaves it alone. Component reversal etc. now happens only for IP addresses.
4275 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4277 57. Double the size of the debug message buffer (to 2048) so that more of very
4278 long debug lines gets shown.
4280 58. The exicyclog utility now does better if the number of log files to keep
4281 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4283 59. Two changes related to the smtp_active_hostname option:
4285 (1) $smtp_active_hostname is now available as a variable.
4286 (2) The default for smtp_banner uses $smtp_active_hostname instead
4287 of $primary_hostname.
4289 60. The host_aton() function is supposed to be passed a string that is known
4290 to be a valid IP address. However, in the case of IPv6 addresses, it was
4291 not checking this. This is a hostage to fortune. Exim now panics and dies
4292 if the condition is not met. A case was found where this could be provoked
4293 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4294 components; fortuitously, this particular loophole had already been fixed
4295 by change 4.50/55 above.
4297 If there are any other similar loopholes, the new check in host_aton()
4298 itself should stop them being exploited. The report I received stated that
4299 data on the command line could provoke the exploit when Exim was running as
4300 exim, but did not say which command line option was involved. All I could
4301 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4302 running as the user.
4305 61. There was a buffer overflow vulnerability in the SPA authentication code
4306 (which came originally from the Samba project). I have added a test to the
4307 spa_base64_to_bits() function which I hope fixes it.
4310 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4311 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4313 63. The daemon start-up calls getloadavg() while still root for those OS that
4314 need the first call to be done as root, but it missed one case: when
4315 deliver_queue_load_max is set with deliver_drop_privilege. This is
4316 necessary for the benefit of the queue runner, because there is no re-exec
4317 when deliver_drop_privilege is set.
4319 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4320 This has been fixed.
4322 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4323 was in use, was not putting the data itself into the right store pool;
4324 consequently, it could be overwritten for a subsequent message in the same
4325 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4328 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4330 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4331 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4332 the function were treating the return as a boolean value, which happened to
4333 work because 0=false and not-0=true, but is not correct code.
4335 68. The host_aton() function was not handling scoped IPv6 addresses (those
4336 with, for example, "%eth0" on the end) correctly.
4338 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4339 negated items (that is, ~something) in unsigned ints. Some compilers
4340 apparently mutter when there is no cast.
4342 70. If an address verification called from an ACL failed, and did not produce a
4343 user-specific message (i.e. there was only a "system" message), nothing was
4344 put in $acl_verify_message. In this situation, it now puts the system
4347 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4348 session; change 4.31/43 added the unwanted input to the log line - except
4349 that it did not do this in the start of session case. It now does.
4351 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4352 This is wrong and can cause the other end to generate a synchronization
4353 error if it is another Exim or anything else that does the synchronization
4354 check. A QUIT command is no longer sent after a timeout.
4356 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4357 during host lookups.
4359 74. The defer_ok option of callout verification was not working if it was used
4360 when verifying addresses in header lines, that is, for this case:
4362 verify = header_sender/callout=defer_ok
4364 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4365 those file descriptors could be used for SMTP connections. If anything
4366 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4367 could be sent to the SMTP client, causing chaos. The daemon now opens
4368 stdin, stdout, and stderr to /dev/null when it puts itself into the
4371 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4372 log by default. The user can override this with suitable Perl magic.
4374 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4375 the log message when discard triggers, was not working for the DATA ACL or
4376 for the non-SMTP ACL.
4378 78. Error message wording change in sieve.c.
4380 79. If smtp_accept_max_per_host was set, the number of connections could be
4381 restricted to fewer than expected, because the daemon was trying to set up
4382 a new connection before checking whether the processes handling previous
4383 connections had finished. The check for completed processes is now done
4384 earlier. On busy systems, this bug wouldn't be noticed because something
4385 else would have woken the daemon, and it would have reaped the completed
4388 80. If a message was submitted locally by a user whose login name contained one
4389 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4390 It caused a spool format error. I have fixed the spool reading code. A
4391 related problem was that the "from" clause in the Received: line became
4392 illegal because of the space(s). It is now covered by ${quote_local_part.
4394 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4397 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4399 83. Patch from Sieve maintainer to fix unterminated string problem in
4400 "vacation" handling.
4402 84. Some minor changes to the Linux configuration files to help with other
4403 OS variants using glibc.
4405 85. One more patch for Sieve to update vacation handling to latest spec.
4408 ----------------------------------------------------
4409 See the note above about the 4.44 and 4.50 releases.
4410 ----------------------------------------------------
4416 1. Change 4.43/35 introduced a bug that caused file counts to be
4417 incorrectly computed when quota_filecount was set in an appendfile
4420 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4421 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4424 3. Give more explanation in the error message when the command for a transport
4425 filter fails to execute.
4427 4. There are several places where Exim runs a non-Exim command in a
4428 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4429 was being done only for the command run by the queryprogram router. It is
4430 now done for all such subprocesses. The other cases are: ${run, transport
4431 filters, and the commands run by the lmtp and pipe transports.
4433 5. Some older OS have a limit of 256 on the maximum number of file
4434 descriptors. Exim was using setrlimit() to set 1000 as a large value
4435 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4436 systems. I've change it so that if it can't get 1000, it tries for 256.
4438 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4439 was an oversight, and furthermore, ever since the addition of extra
4440 controls (e.g. 4.43/32), the checks on when to allow different forms of
4441 "control" were broken. There should now be diagnostics for all cases when a
4442 control that does not make sense is encountered.
4444 7. $recipients is now available in the predata ACL (oversight).
4446 8. Tidy the search cache before the fork to do a delivery from a message
4447 received from the command line. Otherwise the child will trigger a lookup
4448 failure and thereby defer the delivery if it tries to use (for example) a
4449 cached ldap connection that the parent has called unbind on.
4451 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4452 of $address_data from the recipient verification was clobbered by the
4453 sender verification.
4455 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4456 was its contents. (It was OK if the option was not defined at all.)
4458 11. A "Completed" log line is now written for messages that are removed from
4459 the spool by the -Mrm option.
4461 12. $host_address is now set to the target address during the checking of
4462 ignore_target_hosts.
4464 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4465 being passed; this would have caused $sender_host_name to have been used if
4466 matching the list had actually called for a host name (not very likely,
4467 since this list is usually IP addresses). A host name is now passed as
4470 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4471 code that specifies a non-restarting handler (typically sigaction() in
4472 modern systems) in an attempt to fix a rare and obscure crash bug.
4474 15. Narrowed the window for a race in the daemon that could cause it to ignore
4475 SIGCHLD signals. This is not a major problem, because they are used only to
4476 wake it up if nothing else does.
4478 16. A malformed maildirsize file could cause Exim to calculate negative values
4479 for the mailbox size or file count. Odd effects could occur as a result.
4480 The maildirsize information is now recalculated if the size or filecount
4483 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4484 support for a long time. Removed HAVE_SYS_VFS_H.
4486 18. Updated exipick to current release from John Jetmore.
4488 19. Allow an empty sender to be matched against a lookup in an address list.
4489 Previously the only cases considered were a regular expression, or an
4492 20. Exim went into a mad DNS lookup loop when doing a callout where the
4493 host was specified on the transport, if the DNS lookup yielded more than
4496 21. The RFC2047 encoding function was originally intended for short strings
4497 such as real names; it was not keeping to the 75-character limit for
4498 encoded words that the RFC imposes. It now respects the limit, and
4499 generates multiple encoded words if necessary. To be on the safe side, I
4500 have increased the buffer size for the ${rfc2047: expansion operator from
4503 22. Failure to deliver a bounce message always caused it to be frozen, even if
4504 there was an errors_to setting on the router. The errors_to setting is now
4507 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4508 canonical form (fully expanded) before being placed in
4509 $sender_host_address.
4511 24. Updated eximstats to version 1.33
4513 25. Include certificate and key file names in error message when GnuTLS fails
4514 to set them up, because the GnuTLS error message doesn't include the name
4515 of the failing file when there is a problem reading it.
4517 26. Expand error message when OpenSSL has problems setting up cert/key files.
4520 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4521 (this can affect the format of dates).
4523 28. exim_tidydb, when checking for the continued existence of a message for
4524 which it has found a message-specific retry record, was not finding
4525 messages that were in split spool directories. Consequently, it was
4526 deleting retry records that should have stayed in existence.
4528 29. eximstats updated to version 1.35
4529 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4530 1.35 - bugfix such that pie charts by volume are generated correctly
4532 30. The SPA authentication driver was not abandoning authentication and moving
4533 on to the next authenticator when an expansion was forced to fail,
4534 contradicting the general specification for all authenticators. Instead it
4535 was generating a temporary error. It now behaves as specified.
4537 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4538 (the order specifies the preference for clients). The order is now AES256,
4539 AES128, 3DES, ARCFOUR128.
4541 31. Small patch to Sieve code - explicitly set From: when generating an
4544 32. Exim crashed if a remote delivery caused a very long error message to be
4545 recorded - for instance if somebody sent an entire SpamAssassin report back
4546 as a large number of 550 error lines. This bug was coincidentally fixed by
4547 increasing the size of one of Exim's internal buffers (big_buffer) that
4548 happened as part of the Exiscan merge. However, to be on the safe side, I
4549 have made the code more robust (and fixed the comments that describe what
4552 33. Some experimental protocols are using DNS PTR records for new purposes. The
4553 keys for these records are domain names, not reversed IP addresses. The
4554 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4555 leaves it alone. Component reversal etc. now happens only for IP addresses.
4558 34. The host_aton() function is supposed to be passed a string that is known
4559 to be a valid IP address. However, in the case of IPv6 addresses, it was
4560 not checking this. This is a hostage to fortune. Exim now panics and dies
4561 if the condition is not met. A case was found where this could be provoked
4562 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4563 components; fortuitously, this particular loophole had already been fixed
4564 by change 4.50/55 or 4.44/33 above.
4566 If there are any other similar loopholes, the new check in host_aton()
4567 itself should stop them being exploited. The report I received stated that
4568 data on the command line could provoke the exploit when Exim was running as
4569 exim, but did not say which command line option was involved. All I could
4570 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4571 running as the user.
4574 35. There was a buffer overflow vulnerability in the SPA authentication code
4575 (which came originally from the Samba project). I have added a test to the
4576 spa_base64_to_bits() function which I hope fixes it.
4579 36. The daemon start-up calls getloadavg() while still root for those OS that
4580 need the first call to be done as root, but it missed one case: when
4581 deliver_queue_load_max is set with deliver_drop_privilege. This is
4582 necessary for the benefit of the queue runner, because there is no re-exec
4583 when deliver_drop_privilege is set.
4585 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4586 was in use, was not putting the data itself into the right store pool;
4587 consequently, it could be overwritten for a subsequent message in the same
4588 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4591 38. Sometimes the final signoff response after QUIT could fail to get
4592 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4593 < 0 before doing a fflush(). This bug looks as though it goes back to the
4594 introduction of TLS in release 3.20, but "sometimes" must have been rare
4595 because the tests only now provoked it.
4601 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4602 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4603 Now it returns an integer. A number of calls were still expecting a T/F
4604 return. Fortuitously, in all cases, the tests worked in OK situations,
4605 which is the norm. However, things would have gone wrong on any write
4606 failures on the smtp file descriptor. This function is used when sending
4607 messages over SMTP and also when doing verify callouts.
4609 2. When Exim is called to do synchronous delivery of a locally submitted
4610 message (the -odf or -odi options), it no longer closes stderr before doing
4613 3. Implemented the mua_wrapper option.
4615 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4617 5. Implemented the functions header_remove(), header_testname(),
4618 header_add_at_position(), and receive_remove_recipient(), and exported them
4621 6. If an ACL "warn" statement specified the addition of headers, Exim already
4622 inserted X-ACL-Warn: at the start if there was no header name. However, it
4623 was not making this test for the second and subsequent header lines if
4624 there were newlines in the string. This meant that an invalid header could
4625 be inserted if Exim was badly configured.
4627 7. Allow an ACL "warn" statement to add header lines at the start or after all
4628 the Received: headers, as well as at the end.
4630 8. Added the rcpt_4xx retry error code.
4632 9. Added postmaster_mailfrom=xxx to callout verification option.
4634 10. Added mailfrom=xxxx to the callout verification option, for verify=
4637 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4638 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4640 12. Inserted some casts to stop certain compilers warning when using pointer
4641 differences as field lengths or precisions in printf-type calls (mostly
4642 affecting debugging statements).
4644 13. Added optional readline() support for -be (dynamically loaded).
4646 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4647 same clock tick as a message's arrival, so that its received time was the
4648 same as the "first fail" time on the retry record, and that message
4649 remained on the queue past the ultimate address timeout, every queue runner
4650 would try a delivery (because it was past the ultimate address timeout) but
4651 after another failure, the ultimate address timeout, which should have then
4652 bounced the address, did not kick in. This was a "< instead of <=" error;
4653 in most cases the first failure would have been in the next clock tick
4654 after the received time, and all would be well.
4656 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4657 being recognized when the domain list was tested by the match_domain
4658 condition in an expansion string.
4660 16. Added the ${str2b64: operator.
4662 17. Exim was always calling setrlimit() to set a large limit for the number of
4663 processes, without checking whether the existing limit was already
4664 adequate. (It did check for the limit on file descriptors.) Furthermore,
4665 errors from getrlimit() and setrlimit() were being ignored. Now they are
4666 logged to the main and panic logs, but Exim does carry on, to try to do its
4667 job under whatever limits there are.
4669 18. Imported PCRE 5.0.
4671 19. Trivial typo in log message " temporarily refused connection" (the leading
4674 20. If the log selector return_path_on_delivery was set and an address was
4675 redirected to /dev/null, the delivery process crashed because it assumed
4676 that a return path would always be set for a "successful" delivery. In this
4677 case, the whole delivery is bypassed as an optimization, and therefore no
4680 21. Internal re-arrangement: the function for sending a challenge and reading
4681 a response while authentication was assuming a zero-terminated challenge
4682 string. It's now changed to take a pointer and a length, to allow for
4683 binary data in such strings.
4685 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4687 23. Exim was not respecting finduser_retries when seeking the login of the
4688 uid under which it was called; it was always trying 10 times. (The default
4689 setting of finduser_retries is zero.) Also, it was sleeping after the final
4690 failure, which is pointless.
4692 24. Implemented tls_on_connect_ports.
4694 25. Implemented acl_smtp_predata.
4696 26. If the domain in control=submission is set empty, Exim assumes that the
4697 authenticated id is a complete email address when it generates From: or
4698 Sender: header lines.
4700 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4701 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4702 chown and chgrp in /bin and hostname in /usr/bin.
4704 28. Exim was keeping the "process log" file open after each use, just as it
4705 does for the main log. This opens the possibility of it remaining open for
4706 long periods when the USR1 signal hits a daemon. Occasional processlog
4707 errors were reported, that could have been caused by this. Anyway, it seems
4708 much more sensible not to leave this file open at all, so that is what now
4711 29. The long-running daemon process does not normally write to the log once it
4712 has entered its main loop, and it closes the log before doing so. This is
4713 so that log files can straightforwardly be renamed and moved. However,
4714 there are a couple of unusual error situations where the daemon does write
4715 log entries, and I had neglected to close the log afterwards.
4717 30. The text of an SMTP error response that was received during a remote
4718 delivery was being truncated at 512 bytes. This is too short for some of
4719 the long messages that one sometimes sees. I've increased the limit to
4722 31. It is now possible to make retry rules that apply only when a message has a
4723 specific sender, in particular, an empty sender.
4725 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4726 it possible to be selective about when SMTP synchronization is enforced.
4728 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4730 32. Implemented hosts_connection_nolog.
4732 33. Added an ACL for QUIT.
4734 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4737 35. Added mailbox_size and mailbox_filecount to appendfile.
4739 36. Added control = no_multiline_responses to ACLs.
4741 37. There was a bug in the logic of the code that waits for the clock to tick
4742 in the case where the clock went backwards by a substantial amount such
4743 that the microsecond fraction of "now" was more than the microsecond
4744 fraction of "then" (but the whole seconds number was less).
4746 38. Added support for the libradius Radius client library this is found on
4747 FreeBSD (previously only the radiusclient library was supported).
4753 1. When certain lookups returned multiple values in the form name=value, the
4754 quoting of the values was not always being done properly. Specifically:
4755 (a) If the value started with a double quote, but contained no whitespace,
4757 (b) If the value contained whitespace other than a space character (i.e.
4758 tabs or newlines or carriage returns) it was not quoted.
4759 This fix has been applied to the mysql and pgsql lookups by writing a
4760 separate quoting function and calling it from the lookup code. The fix
4761 should probably also be applied to nisplus, ibase and oracle lookups, but
4762 since I cannot test any of those, I have not disturbed their existing code.
4764 2. A hit in the callout cache for a specific address caused a log line with no
4765 reason for rejecting RCPT. Now it says "Previous (cached) callout
4766 verification failure".
4768 3. There was an off-by-one bug in the queryprogram router. An over-long
4769 return line was truncated at 256 instead of 255 characters, thereby
4770 overflowing its buffer with the terminating zero. As well as fixing this, I
4771 have increased the buffer size to 1024 (and made a note to document this).
4773 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4774 when Exim is waiting for an SMTP response from a remote server, Exim
4775 restarts its select() call on the socket, thereby resetting its timeout.
4776 This is not a problem when such interrupts are rare. Somebody set up a cron
4777 job to run exiwhat every 2 minutes, which is less than the normal select()
4778 timeout (5 or 10 minutes). This meant that the select() timeout never
4779 kicked in because it was always reset. I have fixed this by comparing the
4780 time when an interrupt arrives with the time at the start of the first call
4781 to select(). If more time than the timeout has elapsed, the interrupt is
4782 treated as a timeout.
4784 5. Some internal re-factoring in preparation for the addition of Sieve
4785 extensions (by MH). In particular, the "personal" test is moved to a
4786 separate function, and given an option for scanning Cc: and Bcc: (which is
4787 not set for Exim filters).
4789 6. When Exim created an email address using the login of the caller as the
4790 local part (e.g. when creating a From: or Sender: header line), it was not
4791 quoting the local part when it contained special characters such as @.
4793 7. Installed new OpenBSD configuration files.
4795 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4796 try to make them clearer.
4798 9. Callout options, other than the timeout value, were being ignored when
4799 verifying sender addresses in header lines. For example, when using
4801 verify = header_sender/callout=no_cache
4803 the cache was (incorrectly) being used.
4805 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4806 only the Cygwin environment.
4808 11. When return_path_on_delivery was set as a log selector, if different remote
4809 addresses in the same message used different return paths and parallel
4810 remote delivery occurred, the wrong values would sometimes be logged.
4811 (Whenever a remote delivery process finished, the return path value from
4812 the most recently started remote delivery process was logged.)
4814 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4815 lines when AUTH and/or TLS are in use. This is the "received protocol"
4816 field. Exim used to use "asmtp" for authenticated SMTP, without any
4817 indication (in the protocol name) for TLS use. Now it follows the RFC and
4818 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4819 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4820 names appear in log lines as well as in Received: header lines.
4822 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4823 extensions, and comparison tests, and to fix some bugs.
4825 14. Changes to the "personal" filter test:
4827 (1) The test was buggy in that it was just doing the equivalent of
4828 "contains" tests on header lines. For example, if a user's address was
4829 anne@some.where, the "personal" test would incorrectly be true for
4831 To: susanne@some.where
4833 This test is now done by extracting each address from the header in turn,
4834 and checking the entire address. Other tests that are part of "personal"
4835 are now done using regular expressions (for example, to check local parts
4836 of addresses in From: header lines).
4838 (2) The list of non-personal local parts in From: addresses has been
4839 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4840 taken from the Sieve specification recommendations.
4842 (3) If the message contains any header line starting with "List-" it is
4843 treated as non-personal.
4845 (4) The test for "circular" in the Subject: header line has been removed
4846 because it now seems ill-conceived.
4848 15. Minor typos in src/EDITME comments corrected.
4850 16. Installed latest exipick from John Jetmore.
4852 17. If headers_add on a router specified a text string that was too long for
4853 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4854 of string_sprintf() is now avoided.
4856 18. $message_body_size was not set (it was always zero) when running the DATA
4857 ACL and the local_scan() function.
4859 19. For the "mail" command in an Exim filter, no default was being set for
4860 the once_repeat time, causing a random time value to be used if "once" was
4861 specified. (If the value happened to be <= 0, no repeat happened.) The
4862 default is now 0s, meaning "never repeat". The "vacation" command was OK
4863 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4864 (I found it when inspecting the code).
4866 20. There is now an overall timeout for performing a callout verification. It
4867 defaults to 4 times the callout timeout, which applies to individual SMTP
4868 commands during the callout. The overall timeout applies when there is more
4869 than one host that can be tried. The timeout is checked before trying the
4870 next host. This prevents very long delays if there are a large number of
4871 hosts and all are timing out (e.g. when the network connections are timing
4872 out). The value of the overall timeout can be changed by specifying an
4873 additional sub-option for "callout", called "maxwait". For example:
4875 verify = sender/callout=5s,maxwait=20s
4877 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4878 to the end before writing, but this should make it even safer).
4880 22. Exim was forgetting that it had advertised PIPELINING for the second and
4881 subsequent messages on an SMTP connection. It was also not resetting its
4882 memory on STARTTLS and an internal HELO.
4884 23. When Exim logs an SMTP synchronization error within a session, it now
4885 records whether PIPELINING has been advertised or not.
4887 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4888 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4889 rather than long int.
4891 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4893 26. Added the never_mail option to autoreply.
4899 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4900 crash if the getsockname() call failed; this can happen if a connection is
4901 closed very soon after it is established. The problem was simply in the
4902 order in which certain operations were done, causing Exim to try to write
4903 to the SMTP stream before it had set up the file descriptor. The bug has
4904 been fixed by making things happen in the correct order.
4910 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4911 before the connection was closed, thus losing the rejection response.
4913 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4914 some early Solaris releases, but causes trouble in current releases where
4915 socklen_t is defined.
4917 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4920 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4923 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4924 that releasing the top of it at the end releases what was used for sub-
4925 expansions (unless the block got too big). However, discard this block if
4926 the first thing is a variable or header, so that we can use its block when
4927 it is dynamic (useful for very large $message_headers, for example).
4929 6. Lookups now cache *every* query, not just the most recent. A new, separate
4930 store pool is used for this. It can be recovered when all lookup caches are
4931 flushed. Lookups now release memory at the end of their result strings.
4932 This has involved some general refactoring of the lookup sources.
4934 7. Some code has been added to the store_xxx() functions to reduce the amount
4935 of flapping under certain conditions.
4937 8. log_incoming_interface used to affect only the <= reception log lines. Now
4938 it causes the local interface and port to be added to several more SMTP log
4939 lines, for example "SMTP connection from", and rejection lines.
4941 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4943 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4945 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4946 could be overwritten at the end of the current message (or the start of a
4947 new message if it was set in a HELO ACL). The value is now preserved for
4948 the duration of the SMTP connection.
4950 12. If a transport had a headers_rewrite setting, and a matching header line
4951 contained an unqualified address, that address was qualified, even if it
4952 did not match any rewriting rules. The underlying bug was that the values
4953 of the flags that permit the existence of unqualified sender and recipient
4954 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4955 non-local messages, and by -bnq for local messages) were not being
4956 preserved with the message after it was received.
4958 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4959 "next input=" as part of the text comprising the host identity instead of
4960 the correct text. The code was using the same buffer for two different
4961 strings. However, depending on which order the printing function evaluated
4962 its arguments, the bug did not always show up. Under Linux, for example, my
4963 test suite worked just fine.
4965 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4966 doesn't work with some older versions of Perl. It has been changed to "my",
4967 which in any case is probably the better facility to use.
4969 15. A really picky compiler found some instances of statements for creating
4970 error messages that either had too many or two few arguments for the format
4973 16. The size of the buffer for calls to the DNS resolver has been increased
4974 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4975 for addresses that have a lot of PTR records. This alleviates a problem; it
4976 does not fully solve it.
4978 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4979 buffer now truncates the list and logs the incident, which is the same
4980 action as happens when Exim is looking up a host name and its aliases.
4981 Previously in this situation something unpredictable would happen;
4982 sometimes it was "internal error: store_reset failed".
4984 18. If a server dropped the connection unexpectedly when an Exim client was
4985 using GnuTLS and trying to read a response, the client delivery process
4986 crashed while trying to generate an error log message.
4988 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4989 string, for example:
4991 warn message = H1: something\nH2: something
4993 the text was added as a single header line from Exim's point of view
4994 though it ended up OK in the delivered message. However, searching for the
4995 second and subsequent header lines using $h_h2: did not work. This has been
4996 fixed. Similarly, if a system filter added multiple headers in this way,
4997 the routers could not see them.
4999 20. Expanded the error message when iplsearch is called with an invalid key to
5000 suggest using net-iplsearch in a host list.
5002 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5003 ACLs are no longer actually imposed (and a message to that effect is
5006 22. If a "gecos" field in a passwd entry contained escaped characters, in
5007 particular, if it contained a \" sequence, Exim got it wrong when building
5008 a From: or a Sender: header from that name. A second bug also caused
5009 incorrect handling when an unquoted " was present following a character
5010 that needed quoting.
5012 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5013 was not being matched caselessly.
5015 24. Arranged for all hyphens in the exim.8 source to be escaped with
5018 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5019 properly. Recipient callout cache records were still being keyed to include
5020 the sender, even when use_sender was set false. This led to far more
5021 callouts that were necessary. The sender is no longer included in the key
5022 when use_sender is false.
5024 26. Added "control = submission" modifier to ACLs.
5026 27. Added the ${base62d: operator to decode base 62 numbers.
5028 28. dnsdb lookups can now access SRV records.
5030 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5031 the configuration file.
5033 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5034 (-v) mode. This makes the output for a verbose queue run more intelligible.
5036 31. Added a use_postmaster feature to recipient callouts.
5038 32. Added the $body_zerocount variable, containing the number of binary zero
5039 bytes in the message body.
5041 33. The time of last modification of the "new" subdirectory is now used as the
5042 "mailbox time last read" when there is a quota error for a maildir
5045 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5047 35. Added +ignore_unknown as a special item in host lists.
5049 36. Code for decoding IPv6 addresses in host lists is now included, even if
5050 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5051 address was recognized as an IP address, but was then not correctly decoded
5052 into binary, causing unexpected and incorrect effects when compared with
5059 1. Very minor rewording of debugging text in manualroute to say "list of
5060 hosts" instead of "hostlist".
5062 2. If verify=header_syntax was set, and a header line with an unqualified
5063 address (no domain) and a large number of spaces between the end of the
5064 name and the colon was received, the reception process suffered a buffer
5065 overflow, and (when I tested it) crashed. This was caused by some obsolete
5066 code that should have been removed. The fix is to remove it!
5068 3. When running in the test harness, delay a bit after writing a bounce
5069 message to get a bit more predictability in the log output.
5071 4. Added a call to search_tidyup() just before forking a reception process. In
5072 theory, someone could use a lookup in the expansion of smtp_accept_max_
5073 per_host which, without the tidyup, could leave open a database connection.
5075 5. Added the variables $recipient_data and $sender_data which get set from a
5076 lookup success in an ACL "recipients" or "senders" condition, or a router
5077 "senders" option, similar to $domain_data and $local_part_data.
5079 6. Moved the writing of debug_print from before to after the "senders" test
5082 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5083 problems for message scanning, either using a data ACL, or using
5084 local_scan() because the Received: header was not generated till after they
5085 were called (in order to set the time as the time of reception completion).
5086 I have revised the way this works. The header is now generated after the
5087 body is received, but before the ACL or local_scan() are called. After they
5088 are run, the timestamp in the header is updated.
5094 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5095 before starting a queue runner without re-exec. This happened only when
5096 deliver_drop_privilege was set or when the Exim user was set to root. The
5097 effect of the bug was that timeouts during subsequent deliveries caused
5098 crashes instead of being properly handled. The handler is now left at its
5099 default (and expected) setting.
5101 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5102 message, again when deliver_drop_privilege is set or Exim is run as root.
5103 The bug described in (1) was not present in this case, but the tidying up
5104 of the other signals was missing. I have made the two cases consistent.
5106 3. The ignore_target_hosts setting on a manualroute router was being ignored
5107 for hosts that were looked up using the /MX notation.
5109 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5112 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5113 operated on the sender address. After changing the $sender_address to <>
5114 for the sender address verify, Exim was re-instated it as the original
5115 (before rewriting) address, but remembering that it had rewritten it, so it
5116 wasn't rewriting it again. This bug also had the effect of breaking the
5117 sender address verification caching when the sender address was rewritten.
5119 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5120 This has been changed so that if the ip literal address matches
5121 ignore_target_hosts, the router declines.
5123 7. Added expansion conditions match_domain, match_address, and match_local_
5124 part (NOT match_host).
5126 8. The placeholder for the Received: header didn't have a length field set.
5128 9. Added code to Exim itself and to exim_lock to test for a specific race
5129 condition that could lead to file corruption when using MBX delivery. The
5130 issue is with the lockfile that is created in /tmp. If this file is removed
5131 after a process has opened it but before that process has acquired a lock,
5132 there is the potential for a second process to recreate the file and also
5133 acquire a lock. This could lead to two Exim processes writing to the file
5134 at the same time. The added code performs the same test as UW imapd; it
5135 checks after acquiring the lock that its file descriptor still refers to
5136 the same named file.
5138 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5139 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5142 11. Added the smtp_active_hostname option. If used, this will typically be made
5143 to depend on the incoming interface address. Because $interface_address is
5144 not set up until the daemon has forked a reception process, error responses
5145 that can happen earlier (such as "too many connections") no longer contain
5148 12. If an expansion in a condition on a "warn" statement fails because a lookup
5149 defers, the "warn" statement is abandoned, and the next ACL statement is
5150 processed. Previously this caused the whole ACL to be aborted.
5152 13. Added the iplsearch lookup type.
5154 14. Added ident_timeout as a log selector.
5156 15. Added tls_certificate_verified as a log selector.
5158 16. Added a global option tls_require_ciphers (compare the smtp transport
5159 option of the same name). This controls incoming TLS connections.
5161 17. I finally figured out how to make tls_require_ciphers do a similar thing
5162 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5163 before starting the TLS session.
5165 18. Tabs are now shown as \t in -bP output.
5167 19. If the log selector return_path_on_delivery was set, Exim crashed when
5168 bouncing a message because it had too many Received: header lines.
5170 20. If two routers both had headers_remove settings, and the first one included
5171 a superfluous trailing colon, the final name in the first list and the
5172 first name in the second list were incorrectly joined into one item (with a
5173 colon in the middle).
5179 1. Added -C and -D options to the exinext utility, mainly to make it easier
5180 to include in the automated testing, but these could be helpful when
5181 multiple configurations are in use.
5183 2. The exinext utility was not formatting the output nicely when there was
5184 an alternate port involved in the retry record key, nor when there was a
5185 message id as well (for retries that were specific to a specific message
5186 and a specific host). It was also confused by IPv6 addresses, because of
5187 the additional colons they contain. I have fixed the IPv4 problem, and
5188 patched it up to do a reasonable job for IPv6.
5190 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5191 delivery, the log line now contains "pipelined" if PIPELINING was used.
5193 4. An SMTP transport process used to panic and die if the bind() call to set
5194 an explicit outgoing interface failed. This has been changed; it is now
5195 treated in the same way as a connect() failure.
5197 5. A reference to $sender_host_name in the part of a conditional expansion
5198 that was being skipped was still causing a DNS lookup. This no longer
5201 6. The def: expansion condition was not recognizing references to header lines
5202 that used bh_ and bheader_.
5204 7. Added the _cache feature to named lists.
5206 8. The code for checking quota_filecount in the appendfile transport was
5207 allowing one more file than it should have been.
5209 9. For compatibility with Sendmail, the command line option
5217 and sets the incoming protocol and host name (for trusted callers). The
5218 host name and its colon can be omitted when only the protocol is to be set.
5219 Note the Exim already has two private options, -pd and -ps, that refer to
5220 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5221 "s", but I don't think that's a major issue.
5223 10. A number of refactoring changes to the code, none of which should affect
5226 (a) The number of logging options was getting close to filling up the
5227 32-bit word that was used as a bit map. I have split them into two classes:
5228 those that are passed in the argument to log_write(), and those that are
5229 only ever tested independently outside of that function. These are now in
5230 separate 32-bit words, so there is plenty of room for expansion again.
5231 There is no change in the user interface or the logging behaviour.
5233 (b) When building, for example, log lines, the code previously used a
5234 macro that called string_cat() twice, in order to add two strings. This is
5235 not really sufficiently general. Furthermore, there was one instance where
5236 it was actually wrong because one of the argument was used twice, and in
5237 one call a function was used. (As it happened, calling the function twice
5238 did not affect the overall behaviour.) The macro has been replaced by a
5239 function that can join an arbitrary number of extra strings onto a growing
5242 (c) The code for expansion conditions now uses a table and a binary chop
5243 instead of a serial search (which was left over from when there were very
5244 few conditions). Also, it now recognizes conditions like "pam" even when
5245 the relevant support is not compiled in: a suitably worded error message is
5246 given if an attempt is made to use such a condition.
5248 11. Added ${time_interval:xxxxx}.
5250 12. A bug was causing one of the ddress fields not to be passed back correctly
5251 from remote delivery subprocesses. The field in question was not being
5252 subsequently used, so this caused to problems in practice.
5254 13. Added new log selectors queue_time and deliver_time.
5256 14. Might have fixed a bug in maildirsizefile handling that threw up
5257 "unexpected character" debug warnings, and recalculated the data
5258 unnecessarily. In any case, I expanded the warning message to give more
5261 15. Added the message "Restricted characters in address" to the statements in
5262 the default ACL that block characters like @ and % in local parts.
5264 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5265 Three changes have been made:
5267 (a) There was a serious bug; a negative response to MAIL caused the whole
5268 recipient domain to be cached as invalid, thereby blocking all messages
5269 to all local parts at the same domain, from all senders. This bug has
5270 been fixed. The domain is no longer cached after a negative response to
5271 MAIL if the sender used is not empty.
5273 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5276 (c) A new callout option, "use_sender" has been added for people who want
5277 the modified behaviour.
5283 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5286 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5287 indeed breaks things for older releases.
5289 3. Added additional logging to the case where there is a problem reading data
5290 from a filter that is running in a subprocess using a pipe, in order to
5291 try to track down a specific problem.
5293 4. Testing facility fudge: when running in the test harness and attempting
5294 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5295 getting "No route to host". Convert this to a timeout.
5297 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5300 6. Some OS don't have socklen_t but use size_t instead. This affects the
5301 fifth argument of getsockopt() amongst other things. This is now
5302 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5303 can be set for individual OS. I have set it for SunOS5, OSF1, and
5304 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5305 some earlier ones do not.
5307 7. Change 4.30/15 was not doing the test caselessly.
5309 8. The standard form for an IPv6 address literal was being rejected by address
5310 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5311 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5312 this, as well as the form without the "IPv6" on the front (but only when
5313 address literals are enabled, of course).
5315 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5317 10. Exim crashed if a message with an empty sender address specified by -f
5318 encountered a router with an errors_to setting. This could be provoked only
5319 by a command such as
5323 where an empty string was supplied; "<>" did not hit this bug.
5325 11. Installed PCRE release 4.5.
5327 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5328 remained set. It is now erased.
5330 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5331 times from message ids (which are base 36 rather than the normal 62).
5333 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5334 were being counted as actual protocol errors, and logged if the log
5335 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5336 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5337 and DATA following a set of rejected RCPTs do not count as protocol errors.
5338 In other words, Exim assumes they were pipelined, though this may not
5339 actually be the case. Of course, in all cases the client gets an
5340 appropriate error code.
5342 15. If a lookup fails in an ACL condition, a message about the failure may
5343 be available; it is used if testing the ACL cannot continue, because most
5344 such messages specify what the cause of the deferral is. However, some
5345 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5346 that caused an old message to be retained and used if a later statement
5347 caused a defer, replacing the real cause of the deferral.
5349 16. If an IP address had so many PTR records that the DNS lookup buffer
5350 was not large enough to hold them, Exim could crash while trying to process
5351 the truncated data. It now detects and logs this case.
5353 17. Further to 4.21/58, another change has been made: if (and only if) the
5354 first line of a message (the first header line) ends with CRLF, a bare LF
5355 in a subsequent header line has a space inserted after it, so as not to
5356 terminate the header.
5358 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5359 unnecessarily, used atoi() instead of strtol(), and didn't check the
5360 termination when getting file sizes from file names by regex.
5362 19. Completely re-implemented the support for maildirsize files, in the light
5363 of a number of problems with the previous contributed implementation
5364 (4.30/29). In particular:
5366 . If the quota is zero, the maildirsize file is maintained, but no quota is
5369 . If the maildir directory does not exist, it is created before any attempt
5370 to write a maildirsize file.
5372 . The quota value in the file is just a cache; if the quota is changed in
5373 the transport, the new value overrides.
5375 . A regular expression is available for excluding directories from the
5378 20. The autoreply transport checks the characters in options that define the
5379 message's headers; it allows continued headers, but it was checking with
5380 isspace() after an embedded newline instead of explicitly looking for a
5383 21. If all the "regular" hosts to which an address was routed had passed their
5384 expiry times, and had not reached their retry times, the address was
5385 bounced, even if fallback hosts were defined. Now Exim should go on to try
5388 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5389 equivalent code in the SMTP transport. Some hosts send humungous responses
5390 to HELO/EHLO, more than 1024 it seems.
5392 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5393 gives compiler warnings in some environments. I've now done it "properly",
5396 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5397 (because of problems with the built-in one) was declared to return uschar *
5398 instead of char *, causing compiler failure.
5400 25. Fixed a file descriptor leak when processing alias/forward files.
5402 26. Fixed a minor format string issue in dbfn.c.
5404 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5406 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5407 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5410 29. When a sender address is verified, it is cached, to save repeating the test
5411 when there is more than one recipient in a message. However, when the
5412 verification involves a callout, it is possible for different callout
5413 options to be set for different recipients. It is too complicated to keep
5414 track of this in the cache, so now Exim always runs a verification when a
5415 callout is required, relying on the callout cache for the optimization.
5416 The overhead is duplication of the address routing, but this should not be
5419 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5420 to be verified with callout=postmaster, and the main callout worked but the
5421 postmaster check failed, the verification correctly failed. However, if a
5422 subsequent RCPT command asked for sender verification *without* the
5423 postmaster check, incorrect caching caused this verification also to fail,
5426 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5427 it was not caching the DNS options (qualify_single, search_parents) that
5428 were used when the lookup failed. A subsequent lookup with different
5429 options therefore always gave the same answer, though there were cases
5430 where it should not have. (Example: a "domains = !$mx_any" option on a
5431 dnslookup router: the "domains" option is always processed without any
5432 widening, but the router might have qualify_single set.) Now Exim uses the
5433 cached value only when the same options are set.
5435 32. Added John Jetmore's "exipick" utility to the distribution.
5437 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5438 than a timeout (e.g. a certificate is required, and is not provided), an
5439 Exim server now closes the connection immediately. Previously it waited for
5440 the client to close - but if the client is SSL, it seems that they each
5441 wait for each other, leading to a delay before one of them times out.
5443 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5444 maintained 0.8.x compatibility because I don't think many are using it, and
5445 it is clearly obsolete.
5447 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5450 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5451 client certificate was expired. A simple patch fixes this, though I don't
5452 understand the full logic of why the verify callback is called multiple
5455 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5456 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5457 which causes problems with some clients (such as the Certicom SSL Plus
5458 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5459 disables the coutermeasure allowing Eudora to connect."
5461 38. Exim was not checking that a write() to a log file succeeded. This could
5462 lead to Bad Things if a log got too big, in particular if it hit a file
5463 size limit. Exim now panics and dies if it cannot write to a log file, just
5464 as it does if it cannot open a log file.
5466 39. Modified OS/Makefile-Linux so that it now contains
5468 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5470 The two -D definitions ensure that Exim is compiled with large file
5471 support, which makes it possible to handle log files that are bigger than
5474 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5475 instance) a domain was checked against a named list that involved a lookup,
5476 causing $domain_data to be set, then another domain was checked against the
5477 same list, then the first domain was re-checked, the value of $domain_data
5478 after the final check could be wrong. In particular, if the second check
5479 failed, it could be set empty. This bug probably also applied to
5482 41. The strip_trailing_dot option was not being applied to the address given
5483 with the -f command-line option.
5485 42. The code for reading a message's header from the spool was incrementing
5486 $received_count, but never initializing it. This meant that the value was
5487 incorrect (doubled) while delivering a message in the same process in which
5488 it was received. In the most common configuration of Exim, this never
5489 happens - a fresh exec is done - but it can happen when
5490 deliver_drop_privilege is set.
5492 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5493 it now includes up to 150 characters of the unexpected data in the log
5496 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5497 and building data strings. The size of both of these buffers was 10 000
5498 bytes - far larger than anybody would *ever* want, thought I. Needless to
5499 say, somebody hit the limit. I have increased the maximum line length to
5500 20 000 and the maximum data length of concatenated lines to 100 000. I have
5501 also fixed two bugs, because there was no checking on these buffers. Tsk,
5502 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5503 buffer is too small.
5505 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5506 lsearch lookups. Now it does.
5508 46. When parsing a route_list item in a manualroute router, a fixed-length
5509 buffer was used for the list of hosts. I made this 1024 bytes long,
5510 thinking that nobody would ever have a list of hosts that long. Wrong.
5511 Somebody had a whole pile of complicated expansion conditions, and the
5512 string was silently truncated, leading to an expansion error. It turns out
5513 that it is easier to change to an unlimited length (owing to other changes
5514 that have happened since this code was originally written) than to build
5515 structure for giving a limitation error. The length of the item that
5516 expands into the list of hosts is now unlimited.
5518 47. The lsearch lookup could not handle data where the length of text line was
5519 more than 4095 characters. Such lines were truncated, leading to shortened
5520 data being returned. It should now handle lines of any length.
5522 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5523 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5526 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5527 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5528 no longer happens in comment lines. A list of replacements is now placed
5529 at the head of all of the source files, except those whose only change is
5530 to replace PERL_COMMAND in the very first #! line.
5532 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5533 messages on the queue, with a bottom-up merge sort, using code contributed
5534 by Michael Haardt. This should make operations like -bp somewhat faster on
5535 large queues. It won't affect queue runners, except when queue_run_in_order
5538 51. Installed eximstats 1.31 in the distribution.
5540 52. Added support for SRV lookups to the dnslookup router.
5542 53. If an ACL referred to $message_body or $message_body_end, the value was not
5543 reset for any messages that followed in the same SMTP session.
5545 54. The store-handling optimization for building very long strings was not
5546 differentiating between the different store pools. I don't think this
5547 actually made any difference in practice, but I've tidied it.
5549 55. While running the routers to verify a sender address, $sender_address
5550 was still set to the sender address. This is wrong, because when routing to
5551 send a bounce to the sender, it would be empty. Therefore, I have changed
5552 it so that, while verifying a sender address, $sender_address is set to <>.
5553 (There is no change to what happens when verifying a recipient address.)
5555 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5556 target A or AAAA records (if not already returned) without resetting the
5557 qualify_single or search_parents options of the DNS resolver. These are
5558 inappropriate in this case because the targets of MX and SRV records must
5559 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5560 target that, when qualified, matched something in the local domain. These
5561 two options are now turned off when doing these lookups.
5563 57. It seems that at least some releases of Reiserfs (which does not have the
5564 concept of a fixed number of inodes) returns zero and not -1 for the
5565 number of available inodes. This interacted badly with check_spool_inodes,
5566 which assumed that -1 was the "no such thing" setting. What I have done is
5567 to check that the total number of inodes is greater than zero before doing
5568 the test of how many are available.
5570 58. When a "warn" ACL statement has a log_message modifier, the message is
5571 remembered, and not repeated. This is to avoid a lot of repetition when a
5572 message has many recipients that cause the same warning to be written.
5573 Howewer, Exim was preserving the list of already written lines for an
5574 entire SMTP session, which doesn't seem right. The memory is now reset if a
5575 new message is started.
5577 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5578 debugging output unless log_rewrite was also set.
5580 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5581 of (int)(handle) when we know that handle contains (void *)(-1).
5583 61. The Exim daemon panic-logs an error return when it closes the incoming
5584 connection. However "connection reset by peer" seems to be common, and
5585 isn't really an error worthy of noting specially, so that particular error
5588 62. When Exim is trying to find all the local interfaces, it used to panic and
5589 die if the ioctl to get the interface flags failed. However, it seems that
5590 on at least one OS (Solaris 9) it is possible to have an interface that is
5591 included in the list of interfaces, but for which you get a failure error
5592 for this call. This happens when the interface is not "plumbed" into a
5593 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5594 failure of the "get flags" call assumes that the interface is down.
5596 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5597 makes life easier for people who are doing arithmetic on fields extracted
5598 from dates, where you often get leading zeros that should not be
5599 interpreted as octal.
5601 64. Added qualify_domain to the redirect router, to override the global
5604 65. If a pathologically long header line contained very many addresses (the
5605 report of this problem mentioned 10 000) and each of them was rewritten,
5606 Exim could use up a very large amount of memory. (It kept on making new
5607 copies of the header line as it rewrote, and never released the old ones.)
5608 At the expense of a bit more processing, the header rewriting function has
5609 been changed so that it no longer eats memory in this way.
5611 66. The generation of the Received: header has been moved from the time that a
5612 message starts to be received, to the time that it finishes. The timestamp
5613 in the Received: header should now be very close to that of the <= log
5614 line. There are two side-effects of this change:
5616 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5617 logged header lines no longer include the local Received: line, because
5618 it has not yet been created. The same applies to a copy of the message
5619 that is returned to a non-SMTP sender when a message is rejected.
5621 (b) When a filter file is tested using -bf, no additional Received: header
5622 is added to the test message. After some thought, I decided that this
5625 This change does not affect the value of $received_for. It is still set
5626 after address rewriting, but before local_scan() is called.
5628 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5630 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5631 gave an unhelpful panic error message, and a defer error. I have managed to
5632 change this behaviour so that it now rejects any supplied certificate,
5633 which seems right, as the list of acceptable certificates is empty.
5635 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5636 gave an unhelpful defer error. I have not managed to make this reject any
5637 supplied certificates, but the error message it gives is "no certificate
5638 supplied", which is not helpful.
5640 70. exigrep's output now also includes lines that are not associated with any
5641 message, but which match the given pattern. Implemented by a patch from
5642 Martin Sluka, which also tidied up the Perl a bit.
5644 71. Recipient callout verification, like sender verification, was using <> in
5645 the MAIL FROM command. This isn't really the right thing, since the actual
5646 sender may affect whether the remote host accepts the recipient or not. I
5647 have changed it to use the actual sender in the callout; this means that
5648 the cache record is now keyed on a recipient/sender pair, not just the
5649 recipient address. There doesn't seem to be a real danger of callout loops,
5650 since a callout by the remote host to check the sender would use <>.
5651 [SEE ABOVE: changed after hitting problems.]
5653 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5654 temporary errors. However, in the case of such a code being given after
5655 the end of a data transmission (i.e. after ".") Exim was failing to write
5656 a retry record for the message. (Yes, there was some broken host that was
5657 actually sending 8xx at this point.)
5659 73. An unknown lookup type in a host list could cause Exim to panic-die when
5660 the list was checked. (An example that provoked this was putting <; in the
5661 middle of a list instead of at the start.) If this happened during a DATA
5662 ACL check, a -D file could be left lying around. This kind of configuration
5663 error no longer causes Exim to die; instead it causes a defer errror. The
5664 incident is still logged to the main and panic logs.
5666 74. Buglet left over from Exim 3 conversion. The message "too many messages
5667 in one connection" was written to the rejectlog but not the mainlog, except
5668 when address rewriting (yes!) was being logged.
5670 75. Added write_rejectlog option.
5672 76. When a system filter was run not as root (that is, when system_filter_user
5673 was set), the values of the $n variables were not being returned to the
5674 main process; thus, they were not subsequently available in the $sn
5677 77. Added +return_path_on_delivery log selector.
5679 78. A connection timeout was being treated differently from recipients deferred
5680 when testing hosts_max_try with a message that was older than the host's
5681 retry timeout. (The host should not be counted, thus allowing all hosts to
5682 be tried at least once before bouncing.) This may have been the cause of an
5683 occasionally reported bug whereby a message would remain on the queue
5684 longer than the retry timeout, but would be bounced if a delivery was
5685 forced. I say "may" because I never totally pinned down the problem;
5686 setting up timeout/retry tests is difficult. See also the next item.
5688 79. The ultimate address timeout was not being applied to errors that involved
5689 a combination of host plus message (for example, a timeout on a MAIL
5690 command). When an address resolved to a number of possible hosts, and they
5691 were not all tried for each delivery (e.g. because of hosts_max_try), a
5692 message could remain on the queue longer than the retry timeout.
5694 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5697 81. Fixed an obscure SMTP outgoing bug which required at least the following
5698 conditions: (a) there was another message waiting for the same server;
5699 (b) the server returned 5xx to all RCPT commands in the first message so
5700 that the message was not completed; (c) the server dropped the connection
5701 or gave a negative response to the RSET that Exim sends to abort the
5702 transaction. The observed case was a dropped connection after DATA that had
5703 been sent in pipelining mode. That is, the server had advertised PIPELINING
5704 but was not implementing it correctly. The effect of the bug was incorrect
5705 behaviour, such as trying another host, and this could lead to a crash.
5711 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5712 and daemon.c were passed as pointers to ints; they should have been
5713 pointers to socklen_t variables (which are typically unsigned ints).
5715 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5718 3. Fixed a really odd bug that affected only the testing scheme; patching a
5719 certain fixed string in the binary changed the value of another string that
5720 happened to be identical to the end of the original first string.
5722 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5723 name", it returns that address as the IP address. On some operating
5724 systems (e.g. Solaris), it also passes back the IP address string as the
5725 "host name". However, on others (e.g. Linux), it passes back an empty
5726 string. Exim wasn't checking for this, and was changing the host name to an
5727 empty string, assuming it had been canonicized.
5729 5. Although rare, it is permitted to have more than one PTR record for a given
5730 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5731 all the names associated with an address, because they do in Solaris.
5732 However, it seems that they do not in Linux for data that comes from the
5733 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5734 I found this out when I moved to a new Linux workstation and tried to run
5735 the Exim test suite.
5737 To get round this problem I have changed the code so that it now does its
5738 own call to the DNS to look up PTR records when searching for a host name.
5739 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5740 addresses that are only in /etc/hosts are still found.
5742 This behaviour is, however, controlled by an option called host_lookup_
5743 order, which defaults to "bydns:byaddr". If people want to use the other
5744 order, or indeed, just use one or the other means of lookup, they can
5745 specify it in this variable.
5747 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5748 some operating systems, this comes back from gethostbyaddr() as an empty
5749 string, and this is what Exim used to test for. However, it seems that in
5750 other systems, "." is yielded. Exim now tests for this case too.
5752 7. The values of check_spool_space and check_log_space are now held internally
5753 as a number of kilobytes instead of an absolute number of bytes. If a
5754 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5755 kilobyte. This means that much larger values can be stored.
5757 8. Exim monitor: an attempt to get the action menu when not actually pointing
5758 at a message produces an empty menu entitled "No message selected". This
5759 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5760 no entries in it ("Shell widget menu has zero width and/or height"). So I
5761 have added a single, blank menu entry in this case.
5763 9. Added ${quote_local_part.
5765 10. MIME decoding is now applied to the contents of Subject: header lines when
5768 11. Now that a reference to $sender_host_address automatically causes a reverse
5769 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5770 host lookup before query-style lookups in lists that might use this
5771 variable. This has therefore been abolished, and the "net-" prefix is no
5772 longer necessary for query-style lookups.
5774 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5775 have been a typo for LFLAGS, so it has been changed.
5777 13. The install script calls Exim with "-C /dev/null" in order to find the
5778 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5779 to be output. Howeve, since Exim outputs its version number before the
5780 error, it didn't break the script. It just looked ugly. I fixed this by
5781 always allowing "-C /dev/null" if the caller is root.
5783 14. Ignore overlarge ACL variable number when reading spool file - insurance
5784 against a later release with more variables having written the file.
5786 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5787 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5788 this, as well as the form without the "IPv6" on the front.
5790 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5791 OS/Makefile-Darwin file.
5793 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5794 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5796 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5797 variable. However, some OS use ints for inodes. I've added cast to long int
5798 to get rid of the compiler warning.
5800 19. I had forgotten to lock out "/../" in configuration file names when
5801 ALT_CONFIG_PREFIX was set.
5803 20. Routers used for verification do not need to specify transports. However,
5804 if such a router generated a host list, and callout was configured, Exim
5805 crashed, because it could not find a port number from the (non-existent)
5806 transport. It now assumes port 25 in this circumstance.
5808 21. Added the -t option to exigrep.
5810 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5811 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5812 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5813 them is set, LOOKUP_LSEARCH is forced.
5815 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5817 24. Added sender and host information to the "rejected by local_scan()" log
5818 line; previously there was no indication of these.
5820 25. Added .include_if_exists.
5822 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5823 Linux. It turns out that not all file systems support this. Apparently some
5824 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5825 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5826 not supported on the file descriptor, is now ignored when Exim is trying to
5827 sync a directory. This applies only to Linux.
5829 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5831 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5832 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5835 29. Added support for maildirsize files from supplied patch (modified a bit).
5837 30. The use of :fail: followed by an empty string could lead Exim to respond to
5838 sender verification failures with (e.g.):
5840 550 Verification failed for <xxx>
5841 550 Sender verify failed
5843 where the first response line was missing the '-' that indicates it is not
5844 the final line of the response.
5846 31. The loop for finding the name of the user that called Exim had a hardwired
5847 limit of 10; it now uses the value of finduser_retries, which is used for
5848 all other user lookups.
5850 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5853 33. Exim was neglecting to zero errno before one call of strtol() when
5854 expanding a string and expecting an integer value. On some systems this
5855 resulted in spurious "integer overflow" errors. Also, it was casting the
5856 result into an int without checking.
5858 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5859 did not work. The code looks as if it has *never* worked, though it appears
5860 to have been documented since at least releast 1.62. I have made it work.
5862 35. The "timeout_DNS" error in retry rules, also documented since at least
5863 1.62, also never worked. As it isn't clear exactly what this means, and
5864 clearly it isn't a major issue, I have abolished the feature by treating it
5865 as "timeout", and writing a warning to the main and panic logs.
5867 36. The display of retry rules for -brt wasn't always showing the error code
5870 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5871 timeout_connect_A, timeout_connect_MX.
5873 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5874 to the empty sender.
5876 39. The daemon was not analysing the content of -oX till after it had closed
5877 stderr and disconnected from the controlling terminal. This meant that any
5878 syntax errors were only noted on the panic log, and the return code from
5879 the command was 0. By re-arranging the code a little, I've made the
5880 decoding happen first, so such errors now appear on stderr, and the return
5881 code is 1. However, the actual setting up of the sockets still happens in
5882 the disconnected process, so errors there are still only recorded on the
5885 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5886 connections (as happens on some IP stacks) was logged at start up time as
5887 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5888 it from "IPv6 and IPv4", which means that two separate sockets are being
5891 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5892 says whether AF_INET or AF_INET6 was passed as an argument.
5894 42. Exiwhat output was messed up when time zones were included in log
5897 43. Exiwhat now gives more information about the daemon's listening ports,
5898 and whether -tls-on-connect was used.
5900 44. The "port" option of the smtp transport is now expanded.
5902 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5903 silently ignored. Now an error message is written to the main and panic
5906 46. There's a new ACL modifier called "logwrite" which writes to a log file
5907 as soon as it is encountered.
5909 47. Added $local_user_uid and $local_user_gid at routing time.
5911 48. Exim crashed when trying to verify a sender address that was being
5914 49. Exim was recognizing only a space character after ".include". It now also
5915 recognizes a tab character.
5917 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5918 extracting the relevant information from the specification. The man page no
5919 longer contains scrambled data for the -d option, and I've added a section
5920 at the front about calling Exim under different names.
5922 51. Added "extra_headers" argument to the "mail" command in filter files.
5924 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5927 53. Installed eximstats 1.29.
5929 54. Added transport_filter_timeout as a generic transport option.
5931 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5932 Cc: header lines. This was required by RFC 822, but it not required by RFC
5935 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5936 incoming messages that did not have them. Now it does so only if the
5937 message originates locally, that is, if there is no associated remote host
5938 address. When Resent- header lines are present, this applies to the Resent-
5939 lines rather than the non-Resent- lines.
5941 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5942 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5944 58. Messages for configuration errors now include the name of the main
5945 configuration file - useful now that there may be more than one file in a
5946 list (.included file names were always shown).
5948 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5949 for those rare installations that do not start the daemon as root or run it
5950 setuid root. I've cut out the call to initgroups() if the daemon is not
5953 60. The Exim user and group can now be bound into the binary as text strings
5954 that are looked up at the start of Exim's processing.
5956 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5958 62. Added $mailstore_basename variable.
5960 63. Installed patch to sieve.c from Michael Haardt.
5962 64. When Exim failed to open the panic log after failing to open the main log,
5963 the original message it was trying to log was written to stderr and debug
5964 output, but if they were not available (the usual case in production), it
5965 was lost. Now it is written to syslog before the two lines that record the
5966 failures to open the logs.
5968 65. Users' Exim filters run in subprocesses under the user's uid. It is
5969 possible for a "deliver" command or an alias in a "personal" command to
5970 provoke an address rewrite. If logging of address rewriting is configured,
5971 this fails because the process is not running as root or exim. There may be
5972 a better way of dealing with this, but for the moment (because 4.30 needs
5973 to be released), I have disabled address rewrite logging when running a
5974 filter in a non-root, non-exim process.
5980 1. The buildconfig auxiliary program wasn't quoting the value set for
5981 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5982 not defined. This bug was masked in 4.22 by the effect that was fixed in
5985 2. Some messages that were rejected after a message id was allocated were
5986 shown as "incomplete" by exigrep. It no longer does this for messages that
5987 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5989 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5990 have allow_domain_literals set, the ID did not get logged in the <= line.
5991 Domain literals are now always recognized in Message-ID: header lines.
5993 4. The first argument for a ${extract expansion item is the key name or field
5994 number. Leading and trailing spaces in this item were not being ignored,
5995 causing some misleading effects.
5997 5. When deliver_drop_privilege was set, single queue runner processes started
5998 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5999 same command in the process it spins off) were not dropping privilege.
6001 6. When the daemon running as "exim" started a queue runner, it always
6002 re-executed Exim in the spun-off process. This is a waste of effort when
6003 deliver_drop_privilege is set. The new process now just calls the
6004 queue-runner function directly.
6010 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6013 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6014 ignored. Though the use of -oP was forcing the writing of a pid file, it
6015 was always written to the default place.
6017 3. If the message "no IP address found for host xxxx" is generated during
6018 incoming verification, it is now followed by identification of the incoming
6019 connection (so you can more easily find what provoked it).
6021 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6023 5. Added some features to "harden" Exim a bit more against certain attacks:
6025 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6026 be put in Local/Makefile. This is like the never_users runtime option,
6027 but it cannot be overridden. The default setting is "root".
6029 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6030 prefix string with which any file named in a -C command line option
6033 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6034 is retained for -C and -D only if the caller of Exim is root. Without
6035 it, the exim user may also use -C and -D and retain privilege.
6037 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6038 command line option is disabled.
6040 6. Macro names set by the -D option must start with an upper case letter, just
6041 like macro names defined in the configuration file.
6043 7. Added "dereference=" facility to LDAP.
6045 8. Two instances of the typo "uknown" in the source files are fixed.
6047 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6048 the Configure-Makefile script screwed up while processing it.
6050 10. Incorporated PCRE 4.4.
6052 11. The SMTP synchronization check was not operating right at the start of an
6053 SMTP session. For example, it could not catch a HELO sent before the client
6054 waited for the greeting. There is now a check for outstanding input at the
6055 point when the greeting is written. Because of the duplex, asynchronous
6056 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6057 way, but not yet received, when the check is performed.
6059 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6060 on TCP/IP sockets, because this apparently causes some broken clients to
6063 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6064 unchanged) from the Cygwin maintainer.
6066 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6067 was supported instead of testing for mbx. Effectively a typo.
6069 15. The spa authenticator server code was not checking that the input it
6070 received was valid base64.
6072 16. The debug output line for the "set" modifier in ACLs was not showing the
6073 name of the variable that was being set.
6075 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6077 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6078 lookup on its own. Something else was needed to trigger the lookup. For
6079 example, a match in host_lookup or the need for a host name in a host list.
6080 Now, if $sender_host_name is referenced and the host name has not yet been
6081 looked up, a lookup is performed. If the lookup fails, the variable remains
6082 empty, and $host_lookup_failed is set to "1".
6084 19. Added "eqi" as a case-independent comparison operator.
6086 20. The saslauthd authentication condition could segfault if neither service
6087 nor realm was specified.
6089 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6090 error message that was logged was misleading, and incoming SMTP
6091 connections were dropped. The message is now more accurate, and temporary
6092 errors are given to SMTP connections.
6094 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6095 (except RSET) with a 421 error, until QUIT is received. However, it was
6096 failing to send a response to QUIT.
6098 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6099 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6100 ACL is now run after the helo_try_verify_hosts code.
6102 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6103 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6104 case-independent, so other case variants are also recognized). Apparently
6105 some systems use these upper case variants.
6107 25. If more than two messages were waiting for the same host, and a transport
6108 filter was specified for the transport, Exim sent two messages over the
6109 same TCP/IP connection, and then failed with "socket operation on non-
6110 socket" when it tried to send the third.
6112 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6114 27. The extern definition of crypt16() in expand.c was not being excluded when
6115 the OS had its own crypt16() function.
6117 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6118 as a preferred synonym for return_size_limit, both as an option and as an
6121 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6123 30. Changed the default configuration ACL to relax the local part checking rule
6124 for addresses that are not in any local domains. For these addresses,
6125 slashes and pipe symbols are allowed within local parts, but the sequence
6126 /../ is explicitly forbidden.
6128 31. SPA server authentication was not clearing the challenge buffer before
6131 32. log_message in a "warn" ACL statement was writing to the reject log as
6132 well as to the main log, which contradicts the documentation and doesn't
6133 seem right (because no rejection is happening). So I have stopped it.
6135 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6136 However, I am unable to do any testing of this.
6138 34. Fixed an infelicity in the appendfile transport. When checking directories
6139 for a mailbox, to see if any needed to be created, it was accidentally
6140 using path names with one or more superfluous leading slashes; tracing
6141 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6143 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6144 contents are added to the log line that is written for every discarded
6145 recipient. (Previously a log_message setting was ignored.)
6147 36. The ${quote: operator now quotes the string if it is empty.
6149 37. The install script runs exim in order to find its version number. If for
6150 some reason other than non-existence or emptiness, which it checks, it
6151 could not run './exim', it was installing it with an empty version number,
6152 i.e. as "exim-". This error state is now caught, and the installation is
6155 38. An argument was missing from the function that creates an error message
6156 when Exim fails to connect to the socket for saslauthd authentication.
6157 This could cause Exim to crash, or give a corrupted message.
6159 39. Added isip, isip4, and isip6 to ${if conditions.
6161 40. The ACL variables $acl_xx are now saved with the message, and can be
6162 accessed later in routers, transports, and filters.
6164 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6165 strings in the file are not string-expanded.
6167 42. If a MAIL command specified a SIZE value that was too large to fit into an
6168 int variable, the check against message_size_limit failed. Such values are
6169 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6170 day this will have to be increased, but I don't think I want to be around
6171 when emails are that large.
6178 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6179 iconv() is not standard in FreeBSD.
6181 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6182 IPv6 enabled. The observed symptom was a segmentation fault on return from
6183 the function os_common_find_running_interfaces() in src/os.c.
6185 3. In the check_special_case() function in daemon.c I had used "errno" as an
6186 argument name, which causes warnings on some systems. This was basically a
6187 typo, since it was named "eno" in the comments!
6189 4. The code that waits for the clock to tick (at a resolution of some fraction
6190 of a second) so as to ensure message-id uniqueness was always waiting for
6191 at least one whole tick, when it could have waited for less. [This is
6192 almost certainly not relevant at current processor speeds, where it is
6193 unlikely to ever wait at all. But we try to future-proof.]
6195 5. The function that sleeps for a time interval that includes fractions of a
6196 second contained a race. It did not block SIGALRM between setting the
6197 timer, and suspending (a couple of lines later). If the interval was short
6198 and the sigsuspend() was delayed until after it had expired, the suspension
6199 never ended. On busy systems this could lead to processes getting stuck for
6202 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6203 process, before it forks any delivery processes. The open lookup caching
6204 mechanism meant that the open file or database connection was passed into
6205 the delivery process. The problem was that delivery processes always tidy
6206 up cached lookup data. This could cause a problem for the next delivery
6207 process started by the queue runner, because the external queue runner
6208 process does not know about the closure. So the next delivery process
6209 still has data in the lookup cache. In the case of a file lookup, there was
6210 no problem because closing a file descriptor in a subprocess doesn't affect
6211 the parent. However, if the lookup was caching a connection to a database,
6212 the connection was closed, and the second delivery process was likely to
6213 see errors such as "PGSQL: query failed: server closed the connection
6214 unexpectedly". The problem has been fixed by closing all cached lookups
6215 in a queue runner before running a delivery process.
6217 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6218 seem to have the "const" qualifier which it has on other OS. I've
6221 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6222 *of the same type* (client or server) with the same public name that an
6223 error should be diagnosed.
6225 9. When Exim looked up a host name for an IP address, but failed to find the
6226 original IP address when looking up the host name (a safety check), it
6227 output the message "<ip address> does not match any IP for NULL", which was
6228 confusing, to say the least. The bug was that the host name should have
6229 appeared instead of "NULL".
6231 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6232 user that is built into the binary, and the -C or -D options is used, root
6233 privilege is dropped before the configuration file is read. In addition,
6234 logging is switched to stderr instead of the normal log files. If the
6235 configuration then re-defines the Exim user, the unprivileged environment
6236 is probably not what is expected, so Exim logs a panic warning message (but
6239 However, if deliver_drop_privilege is set, the unprivileged state may well
6240 be exactly what is intended, so the warning has been cut out in that case,
6241 and Exim is allowed to try to write to its normal log files.
6247 1. smtp_return_error_details was not giving details for temporary sender
6248 or receiver verification errors.
6250 2. Diagnose a configuration error if two authenticators have the same public
6253 3. Exim used not to create the message log file for a message until the first
6254 delivery attempt. This could be confusing when incoming messages were held
6255 for policy or load reasons. The message log file is now created at the time
6256 the message is received, and an initial "Received" line is written to it.
6258 4. The automatically generated man page for command line options had a minor
6259 bug that caused no ill effects; however, a more serious problem was that
6260 the procedure for building the man page automatically didn't always
6261 operate. Consequently, release 4.20 contains an out-of-date version. This
6262 shouldn't happen again.
6264 5. When building Exim with embedded Perl support, the script that builds the
6265 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6266 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6268 6. The freeze_tell option was not being used for messages that were frozen on
6269 arrival, either by an ACL or by local_scan().
6271 7. Added the smtp_incomplete_transaction log selector.
6273 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6274 was accepting AUTH without a new EHLO.
6276 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6277 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6280 10. Logging of TCP/IP connections (when configured) now happens in the main
6281 daemon process instead of the child process, so that the TCP/IP connection
6282 count is more accurate (but it can never be perfect).
6284 11. The use of "drop" in a nested ACL was not being handled correctly in the
6285 outer ACL. Now, if condition failure induced by the nested "drop" causes
6286 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6287 or "require"), the connection is dropped.
6289 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6290 that yield "discard" can now be used with an "accept" or a "discard" verb,
6291 but an error is generated for any others (because I can't see a useful way
6292 to define what should happen).
6294 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6295 are now processed in the same way as lines in the Exim configuration file.
6296 In particular, continuation lines are supported.
6298 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6300 15. Added -ti meaning -t -i.
6302 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6303 domains, and warn by logging if others are found.
6305 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6306 returned from GIFCONF when Exim is trying to find the list of interfaces on
6307 a host. The code in os.c has been modified to copy each ifreq to an aligned
6308 structure in all cases.
6310 Also, in some cases, the returned ifreq's were being copied to a 'struct
6311 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6312 means the last couple of bytes of an IPv6 address could be chopped if the
6313 ifreq contained only a normal sockaddr (14 bytes storage).
6315 18. Named domain lists were not supported in the hosts_treat_as_local option.
6316 An entry such as +xxxx was not recognized, and was treated as a literal
6319 19. Ensure that header lines added by a DATA ACL are included in the reject log
6320 if the ACL subsequently rejects the message.
6322 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6323 MD5 (which is deprecated).
6325 21. When testing a filter file using -bf, Exim was writing a message when it
6326 took the sender from a "From " line in the message, but it was not doing so
6327 when it took $return_path from a Return-Path: header line. It now does.
6329 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6330 with a valid header line field name (a series of printing characters
6331 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6333 23. Changed "disc" in the source to "disk" to conform to the documentation and
6334 the book and for uniformity.
6336 24. Ignore Sendmail's -Ooption=value command line item.
6338 25. When execve() failed while trying to run a command in a pipe transport,
6339 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6340 could be confused with a return value of 69 from the command itself. This
6341 has been changed to 127, the value the shell returns if it is asked to run
6342 a non-existent command. The wording for the related log line suggests a
6343 non-existent command as the problem.
6345 26. If received_header_text expands to an empty string, do not add a Received:
6346 header line to the message. (Well, it adds a token one on the spool, but
6347 marks it "old" so that it doesn't get used or transmitted.)
6349 27. Installed eximstats 1.28 (addition of -nt option).
6351 28. There was no check for failure on the call to getsockname() in the daemon
6352 code. This can fail if there is a shortage of resources on the system, with
6353 ENOMEM, for example. A temporary error is now given on failure.
6355 29. Contrary to the C standard, it seems that in some environments, the
6356 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6357 program. Exim now does this explicitly; it affects the formatting of
6358 timestamps using strftime().
6360 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6361 complaints. I've now initialized all the variables, to avoid this.
6363 32. Header lines added by a system filter were not being "seen" during
6364 transport-time rewrites.
6366 33. The info_callback() function passed to OpenSSL is set up with type void
6367 (*)(SSL *, int, int), as described somewhere. However, when calling the
6368 function (actually a macro) that sets it up, the type void(*)() is
6369 expected. I've put in a cast to prevent warnings from picky compilers.
6371 34. If a DNS black list lookup found a CNAME record, but there were no A
6372 records associated with the domain it pointed at, Exim crashed.
6374 35. If a DNS black list lookup returned more than one A record, Exim ignored
6375 all but the first. It now scans all returned addresses if a particular IP
6376 value is being sought. In this situation, the contents of the
6377 $dnslist_value variable are a list of all the addresses, separated by a
6380 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6381 to accept a host name and all its aliases if the forward lookup for any of
6382 them yielded the IP address of the incoming connection. Now it accepts only
6383 those names whose forward lookup yields the correct IP address. Any other
6384 names are discarded. This closes a loophole whereby a rogue DNS
6385 administrator could create reverse DNS records to break through a
6386 wildcarded host restriction in an ACL.
6388 37. If a user filter or a system filter that ran in a subprocess used any of
6389 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6390 the wrong values were passed to the pipe command ($thisaddress had the
6391 value of $0, $0 had the value of $1, etc). This bug was introduced by
6392 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6394 38. Improved the line breaking for long SMTP error messages from ACLs.
6395 Previously, if there was no break point between 40 and 75 characters, Exim
6396 left the rest of the message alone. Two changes have been made: (a) I've
6397 reduced the minimum length to 35 characters; (b) if it can't find a break
6398 point between 35 and 75 characters, it looks ahead and uses the first one
6399 that it finds. This may give the occasional overlong line, but at least the
6400 remaining text gets split now.
6402 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6403 file descriptors might be low, and that setting 1000 would always raise it.
6404 It turns out that in some environments, the limit is already over 1000 and
6405 that lowering it causes trouble. So now Exim takes care not to decrease it.
6407 40. When delivering a message, the value of $return_path is set to $sender_
6408 address at the start of routing (routers may change the value). By an
6409 oversight, this default was not being set up when an address was tested by
6410 -bt or -bv, which affected the outcome if any router or filter referred to
6413 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6414 log only when "message" or "log_message" are set. However, if one of the
6415 conditions was an address verification, or a call to a nested ACL, the
6416 messages generated by the underlying test were being passed through. This
6417 no longer happens. The underlying message is available in $acl_verify_
6418 message for both "message" and "log_message" expansions, so it can be
6419 passed through if needed.
6421 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6422 new expansion $bh_ to give the encoded byte string without charset
6423 translation. Translation happens only if iconv() is available; HAVE_ICONV
6424 indicates this at build time. HEADERS_CHARSET gives the charset to
6425 translate to; headers_charset can change it in the configuration, and
6426 "headers charset" can change it in an individual filter file.
6428 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6429 that creates RFC 2047 encoded "words" labels them as that charset instead
6430 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6431 expansion operator; (ii) when Exim creates a From: line for a local
6432 message; (iii) when a header line is rewritten to include a "phrase" part.
6434 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6435 buggy, causing it to skip the first lines of messages whose message ID
6436 ended in 'D'. This would not have bitten before Exim release 4.14, because
6437 message IDs were unlikely to end in 'D' before then. The effect was to have
6438 incorrect size information for certain domains.
6440 45. #include "config.h" was missing at the start of the crypt16.c module. This
6441 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6444 46. If there was a timeout during a "random" callout check, Exim treated it as
6445 a failure of the random address, and carried on sending RSET and the real
6446 address. If the delay was just some slowness somewhere, the response to the
6447 original RCPT would be taken as a response to RSET and so on, causing
6448 mayhem of various kinds.
6450 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6451 when I implemented it. It didn't allow for the fact that some option values
6452 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6453 the right test for positive values.
6455 48. Domain names in DNS records are case-independent. Exim always looks them up
6456 in lower case. Some resolvers return domain names in exactly the case they
6457 appear in the zone file, that is, they may contain uppercase letters. Not
6458 all resolvers do this - some return always lower case. Exim was treating a
6459 change of case by a resolver as a change of domain, similar to a widening
6460 of a domain abbreviation. This triggered its re-routing code and so it was
6461 trying to route what was effectively the same domain again. This normally
6462 caused routing to fail (because the router wouldn't handle the domain
6463 twice). Now Exim checks for this case specially, and just changes the
6464 casing of the domain that it ultimately uses when it transmits the message
6467 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6470 50. If a filter generated a file delivery with a non-absolute name (possible if
6471 no home directory exists for the router), the forbid_file option was not
6474 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6475 the existing equality matching.
6477 52. Exim was using ints instead of ino_t variables in some places where it was
6478 dealing with inode numbers.
6480 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6481 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6482 called TMPDIR, and if it finds it is different, it changes its value.
6484 54. The smtp_printf() function is now made available to local_scan() so
6485 additional output lines can be written before returning. There is also an
6486 smtp_fflush() function to enable the detection of a dropped connection.
6487 The variables smtp_input and smtp_batched_input are exported to
6490 55. Changed the default runtime configuration: the message "Unknown user"
6491 has been removed from the ACL, and instead placed on the localuser router,
6492 using the cannot_route_message feature. This means that any verification
6493 failures that generate their own messages won't get overridden. Similarly,
6494 the "Unrouteable address" message that was in the ACL for unverifiable
6495 relay addresses has also been removed.
6497 56. Added hosts_avoid_esmtp to the smtp transport.
6499 57. The exicyclog script was not checking for the esoteric option
6500 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6501 will work only if exicyclog is run under the appropriate euid.
6503 58. Following a discussion on the list, the rules by which Exim recognises line
6504 endings on incoming messages have been changed. The -dropcr and drop_cr
6505 options are now no-ops, retained only for backwards compatibility. The
6506 following line terminators are recognized: LF CRLF CR. However, special
6507 processing applies to CR:
6509 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6510 nor a local message in the state where . is a terminator.
6512 (ii) If a bare CR is encountered in a header line, an extra space is added
6513 after the line terminator so as not to end the header. The reasoning
6514 behind this is that bare CRs in header lines are most likely either
6515 to be mistakes, or people trying to play silly games.
6517 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6518 was being incorrectly given as 18 bytes larger than it should have been.
6519 This is a VOB (very old bug).
6521 60. This may never have affected anything current, but just in case it has:
6522 When the local host is found other than at the start of a list of hosts,
6523 the local host, those with the same MX, and any that follow, are discarded.
6524 When the list in question was part of a longer list of hosts, the following
6525 hosts (not currently being processed) were also being discarded. This no
6526 longer happens. I'm not sure if this situation could ever has previously
6529 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6532 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6533 header. This is something that is recommended in a new Internet Draft, and
6534 is something that is documented as being done by Sendmail. There are two
6535 possible values. For messages generated by the autoreply transport, Exim
6538 Auto-Submitted: auto-replied
6540 whereas for all other generated messages (e.g. bounces) it adds
6542 Auto-Submitted: auto-generated
6544 63. The "personal" condition in filters now includes a test for the
6545 Auto-Submitted: header. If it contains the string "auto-" the message it
6546 not considered personal.
6548 64. Added rcpt_include_affixes as a generic transport option.
6550 65. Added queue_only_override (default true).
6552 66. Added the syslog_duplication option.
6554 67. If what should have been the first header line of a message consisted of
6555 a space followed by a colon, Exim was mis-interpreting it as a header line.
6556 It isn't of course - it is syntactically invalid and should therefore be
6557 treated as the start of the message body. The misbehaviour could have
6558 caused a number of strange effects, including loss of data in subsequent
6559 header lines, and spool format errors.
6561 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6562 client host had authenticated. This control can now be exercised by an ACL
6563 for more flexibility.
6565 69. By default, callouts do not happen when testing with -bh. There is now a
6566 variant, -bhc, which does actually run the callout code, including
6567 consulting and updating the callout cache.
6569 70. Added support for saslauthd authentication, courtesy of Alexander
6572 71. If statvfs() failed on the spool or log directories while checking their
6573 size for availability, Exim confusingly gave the error "space shortage".
6574 Furthermore, in debugging mode it crashed with a floating point exception.
6575 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6576 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6577 serious problem, Exim now writes to the main and panic logs when this
6578 happens, with details of the failure. It then refuses to accept the
6579 incoming message, giving the message "spool directory problem" or "log
6580 directory problem" with a 421 code for SMTP messages.
6582 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6583 0, 1, and 2 exist, because some OS complain for execs without them (see
6584 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6585 descriptors. However, the code omitted to check that the open succeeded,
6586 causing mysterious errors if for some reason the permissions on /dev/null
6587 got screwed. Now Exim writes a message to the main and panic logs, and
6588 bombs out if it can't open /dev/null.
6590 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6591 interact so that it is all more flexible. It is supposed to remain
6592 backwards compatible. Also added extra_local_interfaces.
6594 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6595 to bomb out with an assertion failure - to the client this appears as a
6596 connection drop. This problem occurs in the part of the code that was taken
6597 from the Samba project. Fortunately, the assertion is in a very simple
6598 function, so I have fixed this by reproducing the function inline in the
6599 one place where it is called, and arranging for authentication to fail
6600 instead of killing the process with assert().
6602 75. The SPA client code was not working when the server requested OEM rather
6603 than Unicode encoding.
6605 76. Added code to make require_files with a specific uid setting more usable in
6606 the case where statting the file as root fails - usually a non-root-mounted
6607 NFS file system. When this happens and the failure is EACCES, Exim now
6608 forks a subprocess and does the per-uid checking as the relevant uid.
6610 77. Added process_log_path.
6612 78. If log_file_path was not explicitly set, a setting of check_log_space or
6613 check_log_inodes was ignored.
6615 79. If a space check for the spool or log partitions fails, the incident is now
6616 logged. Of course, in the latter case the data may get lost...
6618 80. Added the %p formatting code to string_format() so that it can be used to
6619 print addresses in debug_print(). Adjusted all the address printing in the
6620 debugging in store.c to use %p rather than %d.
6622 81. There was a concern that a line of code in smtp_in.c could overflow a
6623 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6624 initially expressed, the concern was not well-founded, because trailing
6625 spaces are removed early. However, if the trailing spaces were followed by
6626 a NULL, they did not get removed, so the overflow was possible. Two fixes
6629 (a) I re-wrote the offending code in a cleaner fashion.
6630 (b) If an incoming SMTP command contains a NULL character, it is rejected
6633 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6634 runs initgroups(), so that if the Exim user is in any additional groups,
6635 they will be used during message reception.
6641 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob