1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
27 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
28 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
29 macro. Patches provided by Josh Soref.
31 JH/05 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
32 the dkim code may be unix-mode line endings rather than smtp wire-format
33 CRLF, so prepend a CR to any bare LF.
35 JH/06 Rationalise the coding for callout smtp conversations and transport ones.
36 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
42 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
43 supports it and a size is available (ie. the sending peer gave us one).
45 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
46 years of being deprecated). The replacements are the ACLs
47 acl_smtp_mime and acl_not_smtp_mime.
49 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
50 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
51 or even an in-clear connection were permitted. Now, if the host lookup
52 was dnssec and dane was requested then the host is only used if the
53 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
54 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
55 if one fails this test.
56 This means that a poorly-configured remote DNS will make it incommunicado;
57 but it protects against a DNS-interception attack on it.
59 JH/04 Bug 1810: make continued-use of an open smtp transport connection
60 non-noisy when a race steals the message being considered.
62 JH/05 If main configuration option tls_certificate is unset, generate a
63 self-signed certificate for inbound TLS connections.
65 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
66 in rewrites and routers.
68 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
69 and logged a warning sing 4.83; now they are a configuration file error.
71 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
72 (lacking @domain). Apply the same qualification processing as RCPT.
74 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
76 JH/10 Support ${sha256:} applied to a string (as well as the previous
79 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
80 a cutthrough deliver is pending, as we always want to make a connection.
81 This also avoids re-routing the message when later placing the cutthrough
82 connection after a verify cache hit.
83 Do not update it with the verify result either.
85 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
86 when routing results in more than one destination address.
88 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
89 signing (which inhibits the cutthrough capability). Previously only
90 the presence of an option was tested; now an expansion evaluating as
91 empty is permissible (obviously it should depend only on data available
92 when the cutthrough connection is made).
94 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
95 the relevant preceding SMTP command did not note the pipelining mode.
97 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
98 Previously they were not counted.
100 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
101 as one having no matching records. Previously we deferred the message
102 that needed the lookup.
104 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
105 distinguished as "(=".
107 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
108 for missing MX records. Previously it only worked for missing A records.
110 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
112 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
113 after the data-go-ahead and data-ack. Patch from Jason Betts.
115 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
116 even for a "none" policy. Patch from Tony Meyer.
118 JH/22 Fix continued use of a connection for further deliveries. If a port was
119 specified by a router, it must also match for the delivery to be
122 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
123 When one of the recipients of a message was unsuitable for the connection
124 (has no matching addresses), we lost track of needing to mark it
125 deferred. As a result mail would be lost.
127 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
129 JH/25 Decoding ACL controls is now done using a binary search; the source code
130 takes up less space and should be simpler to maintain. Merge the ACL
131 condition decode tables also, with similar effect.
133 JH/26 Fix problem with one_time used on a redirect router which returned the
134 parent address unchanged. A retry would see the parent address marked as
135 delivered, so not attempt the (identical) child. As a result mail would
138 JH/27 Fix a possible security hole, wherein a process operating with the Exim
139 UID can gain a root shell. Credit to http://www.halfdog.net/ for
140 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
143 JH/28 Enable {spool,log} filesystem space and inode checks as default.
144 Main config options check_{log,spool}_{inodes,space} are now
145 100 inodes, 10MB unless set otherwise in the configuration.
147 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
148 Previously it only applied to the main-section connection policy
151 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
153 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
154 by me. Added RFC7919 DH primes as an alternative.
156 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
157 are not in the system include path.
159 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
160 GnuTLS, when a session startup failed (eg because the client disconnected)
161 Exim did stdio operations after fclose. This was exposed by a recent
162 change which nulled out the file handle after the fclose.
164 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
165 signed directly by the cert-signing cert, rather than an intermediate
166 OCSP-signing cert. This is the model used by LetsEncrypt.
168 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
170 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
171 an incoming connection.
173 HS/02 Bug 1802: Do not half-close the connection after sending a request
176 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
177 fallback to "prime256v1".
179 JH/34 SECURITY: Use proper copy of DATA command in error message.
180 Could leak key material. Remotely exploitable. CVE-2016-9963.
186 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
187 and 3.4.4 - once the server is enabled to respond to an OCSP request
188 it does even when not requested, resulting in a stapling non-aware
189 client dropping the TLS connection.
191 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
192 support variable-length bit vectors. No functional change.
194 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
195 The I= interface field on outgoing lines is now after the H= remote
196 host field, same as incoming lines. There is a separate
197 outgoing_interface log selector which allows you to disable the
200 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
201 If not running log_selector +smtp_connection the mainlog would be held
202 open indefinitely after a "too many connections" event, including to a
203 deleted file after a log rotate. Leave the per net connection logging
204 leaving it open for efficiency as that will be quickly detected by the
205 check on the next write.
207 HS/01 Bug 1671: Fix post transport crash.
208 Processing the wait-<transport> messages could crash the delivery
209 process if the message IDs didn't exist for some reason. When
210 using 'split_spool_directory=yes' the construction of the spool
211 file name failed already, exposing the same netto behaviour.
213 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
214 mime_regex ACL conditions.
216 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
217 to DSN fail messages (bounces): remote IP, remote greeting, remote response
218 to HELO, local diagnostic string.
220 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
221 log line to debug. Even when configured with a tls authenticator many
222 client connections are expected to not authenticate in this way, so
223 an authenticate fail is not an error.
225 HS/02 Add the Exim version string to the process info. This way exiwhat
226 gives some more detail about the running daemon.
228 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
229 matter for fast-change records such as DNSBLs.
231 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
232 retry record, even if constant. There may be multiple transports with
233 different interface settings and the retry behaviour needs to be kept
236 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
238 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
240 JH/10 Bug 840: fix log_defer_output option of pipe transport
242 JH/11 Bug 830: use same host for all RCPTS of a message, even under
243 hosts_randomize. This matters a lot when combined with mua_wrapper.
245 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
246 ${quote_pgsql:<string>} operator.
248 JH/13 Bug 1708: avoid misaligned access in cached lookup.
250 JH/14 Change header file name for freeradius-client. Relevant if compiling
251 with Radius support; from the Gentoo tree and checked under Fedora.
253 JH/15 Bug 1712: Introduce $prdr_requested flag variable
255 JH/16 Bug 1714: Permit an empty string as expansion result for transport
256 option transport_filter, meaning no filtering.
258 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
260 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
261 defaults to "*" (all hosts). The variable is now available when not built
262 with TLS, default unset, mainly to enable keeping the testsuite sane.
263 If a server certificate is not supplied (via tls_certificate) an error is
264 logged, and clients will find TLS connections fail on startup. Presumably
265 they will retry in-clear.
266 Packagers of Exim are strongly encouraged to create a server certificate
267 at installation time.
269 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
270 with the $config_file variable.
272 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
273 in transport context, after the attempt, and per-recipient. The latter type
274 is per host attempted. The event data is the error message, and the errno
275 information encodes the lookup type (A vs. MX) used for the (first) host,
276 and the trailing two digits of the smtp 4xx response.
278 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
279 to write to mainlog (or rejectlog, paniclog) in the window between file
280 creation and permissions/ownership being changed. Particularly affects
281 installations where exicyclog is run as root, rather than exim user;
282 result is that the running daemon panics and dies.
284 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
286 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
287 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
288 "pri" and "weight". Note that the previous implicit priority given by the
289 list order is no longer honoured.
291 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
294 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
295 by defining SUPPORT_SOCKS.
297 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
298 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
299 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
300 variables are renamed to proxy_{local,external}_{address,port}.
302 JH/27 Move Internationalisation support from Experimental to mainline, enabled
303 for a build by defining SUPPORT_I18N
305 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
306 of the query string, and make ${quote_redis:} do that quoting.
308 JH/29 Move Events support from Experimental to mainline, enabled by default
309 and removable for a build by defining DISABLE_EVENT.
311 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
313 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
314 cached by the daemon.
316 JH/32 Move Redis support from Experimental to mainline, enabled for a build
317 by defining LOOKUP_REDIS. The libhiredis library is required.
319 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
320 keys are given for lookup.
322 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
323 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
324 only supported when built with TLS support. The PolarSSL SHA routines
325 are still used when the TLS library is too old for convenient support.
327 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
328 openssl_options), for security. OpenSSL forces this from version 1.1.0
329 server-side so match that on older versions.
331 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
332 allocation for $value could be released as the expansion processing
333 concluded, but leaving the global pointer active for it.
335 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
336 and to use the domains and local_parts ACL conditions.
338 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
339 incorrectly not doubled on cutthrough transmission, hence seen as a
340 body-termination at the receiving system - resulting in truncated mails.
341 Commonly the sender saw a TCP-level error, and retransmitted the message
342 via the normal store-and-forward channel. This could result in duplicates
343 received - but deduplicating mailstores were liable to retain only the
344 initial truncated version.
346 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
348 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
350 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
351 we're in there, support oversigning also; bug 1309.
353 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
355 HS/04 Add support for keep_environment and add_environment options.
357 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
358 either intentional arithmetic overflow during PRNG, or testing config-
361 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
362 delivery resulted in actual delivery. Cancel cutthrough before DATA
365 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
366 rejects a recipient: pass the reject to the originator.
368 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
369 Many were false-positives and ignorable, but it's worth fixing the
372 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
373 for the new environment-manipulation done at startup. Move the routines
374 from being local to tls.c to being global via the os.c file.
376 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
377 an extract embedded as result-arg for a map, the first arg for extract
378 is unavailable so we cannot tell if this is a numbered or keyed
379 extraction. Accept either.
385 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
388 JH/02 The smtp transport option "multi_domain" is now expanded.
390 JH/03 The smtp transport now requests PRDR by default, if the server offers
393 JH/04 Certificate name checking on server certificates, when exim is a client,
394 is now done by default. The transport option tls_verify_cert_hostnames
395 can be used to disable this per-host. The build option
396 EXPERIMENTAL_CERTNAMES is withdrawn.
398 JH/05 The value of the tls_verify_certificates smtp transport and main options
399 default to the word "system" to access the system default CA bundle.
400 For GnuTLS, only version 3.0.20 or later.
402 JH/06 Verification of the server certificate for a TLS connection is now tried
403 (but not required) by default. The verification status is now logged by
404 default, for both outbound TLS and client-certificate supplying inbound
407 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
410 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
411 Status Notification (bounce) messages are now MIME format per RFC 3464.
412 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
413 under the control of the dsn_advertise_hosts option, and routers may
414 have a dsn_lasthop option.
416 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
417 default, modifiable by a malware= option. The list separator for
418 the options can now be changed in the usual way. Bug 68.
420 JH/10 The smtp_receive_timeout main option is now expanded before use.
422 JH/11 The incoming_interface log option now also enables logging of the
423 local interface on delivery outgoing connections.
425 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
426 if the interface and destination host and port all match.
428 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
431 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
432 Patch from Andrew Lewis.
434 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
435 now supports optional time-restrictions, weighting, and priority
436 modifiers per server. Patch originally by <rommer@active.by>.
438 JH/16 The spamd_address main option now supports a mixed list of local
439 and remote servers. Remote servers can be IPv6 addresses, and
440 specify a port-range.
442 JH/17 Bug 68: The spamd_address main option now supports an optional
443 timeout value per server.
445 JH/18 Bug 1581: Router and transport options headers_add/remove can
446 now have the list separator specified.
448 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
451 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
454 JH/21 Support for the A6 type of dns record is withdrawn.
456 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
457 rather than the verbs used.
459 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
460 from 255 to 1024 chars.
462 JH/24 Verification callouts now attempt to use TLS by default.
464 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
465 are generic router options now. The defaults didn't change.
467 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
468 Original patch from Alexander Shikoff, worked over by JH.
470 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
471 files if ZCAT_COMMAND is not executable.
473 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
475 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
477 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
478 Normally benign, it bites when the pair was led to by a CNAME;
479 modern usage is to not canonicalize the domain to a CNAME target
480 (and we were inconsistent anyway for A-only vs AAAA+A).
482 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
484 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
485 when evaluating $sender_host_dnssec.
487 JH/31 Check the HELO verification lookup for DNSSEC, adding new
488 $sender_helo_dnssec variable.
490 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
492 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
494 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
496 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
497 documented as working, but never had. Support all but $spam_report.
499 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
500 added for tls authenticator.
502 HS/03 Add perl_taintmode main config option
508 TL/01 When running the test suite, the README says that variables such as
509 no_msglog_check are global and can be placed anywhere in a specific
510 test's script, however it was observed that placement needed to be near
511 the beginning for it to behave that way. Changed the runtest perl
512 script to read through the entire script once to detect and set these
513 variables, reset to the beginning of the script, and then run through
514 the script parsing/test process like normal.
516 TL/02 The BSD's have an arc4random API. One of the functions to induce
517 adding randomness was arc4random_stir(), but it has been removed in
518 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
519 function when detected.
521 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
522 cause callback expansion.
524 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
525 syntax errors in an expansion can be treated as a string instead of
526 logging or causing an error, due to the internal use of bool_lax
527 instead of bool when processing it.
529 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
530 server certificates when making smtp deliveries.
532 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
534 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
536 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
538 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
539 Merged patch from Sebastian Wiedenroth.
541 JH/05 Fix results-pipe from transport process. Several recipients, combined
542 with certificate use, exposed issues where response data items split
543 over buffer boundaries were not parsed properly. This eventually
544 resulted in duplicates being sent. This issue only became common enough
545 to notice due to the introduction of connection certificate information,
546 the item size being so much larger. Found and fixed by Wolfgang Breyha.
548 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
549 size buffer was used, resulting in syntax errors when an expansion
552 JH/07 Add support for directories of certificates when compiled with a GnuTLS
553 version 3.3.6 or later.
555 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
556 is EXPERIMENTAL_EVENT, the main-configuration and transport options
557 both become "event_action", the variables become $event_name, $event_data
558 and $event_defer_errno. There is a new variable $verify_mode, usable in
559 routers, transports and related events. The tls:cert event is now also
560 raised for inbound connections, if the main configuration event_action
563 TL/06 In test suite, disable OCSP for old versions of openssl which contained
564 early OCSP support, but no stapling (appears to be less than 1.0.0).
566 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
567 server certificate names available under the smtp transport option
568 "tls_verify_cert_hostname" now do not permit multi-component wildcard
571 JH/10 Time-related extraction expansions from certificates now use the main
572 option "timezone" setting for output formatting, and are consistent
573 between OpenSSL and GnuTLS compilations. Bug 1541.
575 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
576 encoded parameter in the incoming message. Bug 1558.
578 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
579 include certificate info, eximon was claiming there were spoolfile
582 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
584 JH/14 Log delivery-related information more consistently, using the sequence
585 "H=<name> [<ip>]" wherever possible.
587 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
588 are problematic for Debian distribution, omit them from the release
591 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
593 JH/16 Fix string representation of time values on 64bit time_t architectures.
596 JH/17 Fix a null-indirection in certextract expansions when a nondefault
597 output list separator was used.
602 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
603 checkers that were complaining about end of non-void function with no
606 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
607 This was a regression introduced in 4.83 by another bugfix.
609 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
611 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
612 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
618 TF/01 Correctly close the server side of TLS when forking for delivery.
620 When a message was received over SMTP with TLS, Exim failed to clear up
621 the incoming connection properly after forking off the child process to
622 deliver the message. In some situations the subsequent outgoing
623 delivery connection happened to have the same fd number as the incoming
624 connection previously had. Exim would try to use TLS and fail, logging
625 a "Bad file descriptor" error.
627 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
628 utilities have not been installed.
630 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
631 temporary space as the ACL may create new global variables.
633 TL/01 LDAP support uses per connection or global context settings, depending
634 upon the detected version of the libraries at build time.
636 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
637 to extract and use the src ip:port in logging and expansions as if it
638 were a direct connection from the outside internet. PPv2 support was
639 updated based on HAProxy spec change in May 2014.
641 JH/02 Add ${listextract {number}{list}{success}{fail}}.
643 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
644 Properly escape header and check for NULL return.
646 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
649 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
651 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
652 characters in header names, implemented as a verify condition.
653 Contributed by Michael Fischer v. Mollard.
655 TL/05 Rename SPF condition results err_perm and err_temp to standardized
656 results permerror and temperror. Previous values are deprecated but
657 still accepted. In a future release, err_perm and err_temp will be
658 completely removed, which will be a backward incompatibility if the
659 ACL tests for either of these two old results. Patch contributed by
660 user bes-internal on the mailing list.
662 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
664 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
665 selectors, in both main and reject logs.
667 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
670 JH/07 Add malware type "sock" for talking to simple daemon.
672 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
674 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
675 routers/transports under cutthrough routing.
677 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
678 numbers. Touch up "bool" conditional to keep the same definition.
680 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
682 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
684 JH/12 Expand items in router/transport headers_add or headers_remove lists
685 individually rather than the list as a whole. Bug 1452.
687 Required for reasonable handling of multiple headers_ options when
688 they may be empty; requires that headers_remove items with embedded
689 colons must have them doubled (or the list-separator changed).
691 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
692 view the policy declared in the DMARC record. Currently, $dmarc_status
693 is a combined value of both the record presence and the result of the
696 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
698 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
699 dnslookup router and the smtp transport (applying to the forward
702 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
703 of ldap servers used for a specific lookup. Patch provided by Heiko
706 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
707 New variable $lookup_dnssec_authenticated for observability.
709 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
710 Patch submitted by Lars Timman.
712 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
714 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
715 Requires trusted mode and valid format message id, aborts otherwise.
716 Patch contributed by Heiko Schlichting.
718 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
719 certextract with support for various fields. Bug 1358.
721 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
722 is requested by default, modifiable by smtp transport option
725 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
726 operate on certificate variables to give certificate fingerprints
727 Also new ${sha256:cert_variable}.
729 JH/23 The PRDR feature is moved from being Experimental into the mainline.
731 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
732 Christian Aistleitner.
734 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
736 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
737 file. Patch from Wolfgang Breyha.
739 JH/25 Expand the coverage of the delivery $host and $host_address to
740 client authenticators run in verify callout. Bug 1476.
742 JH/26 Port service names are now accepted for tls_on_connect_ports, to
743 align with daemon_smtp_ports. Bug 72.
745 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
746 support and error reporting did not work properly.
748 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
749 and is readable. Patch from Andrew Colin Kissa.
751 TL/14 Enhance documentation of ${run expansion and how it parses the
752 commandline after expansion, particularly in the case when an
753 unquoted variable expansion results in an empty value.
755 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
757 PP/02 Fix internal collision of T_APL on systems which support RFC3123
758 by renaming away from it. Addresses GH issue 15, reported by
761 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
763 TL/15 SECURITY: prevent double expansion in math comparison functions
764 (can expand unsanitized data). Not remotely exploitable.
771 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
773 PP/02 Make -n do something, by making it not do something.
774 When combined with -bP, the name of an option is not output.
776 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
779 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
780 $sender_host_name and config options to manage this, and basic check
783 PP/05 DSCP support for outbound connections and control modifier for inbound.
785 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
786 (Only plugin which currently uses this is kerberos4, which nobody should
787 be using, but we should make it available and other future plugins might
788 conceivably use it, even though it would break NAT; stuff *should* be
789 using channel bindings instead).
791 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
792 name; added for Sendmail compatibility; requires admin caller.
793 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
794 just ignore it); requires trusted caller.
795 Also parse but ignore: -Ac -Am -X<logfile>
798 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
800 TL/02 Add +smtp_confirmation as a default logging option.
802 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
803 Patch by Magnus Holmgren from 2007-02-20.
805 TL/04 Bugzilla 1281 - Spec typo.
806 Bugzilla 1283 - Spec typo.
807 Bugzilla 1290 - Spec grammar fixes.
809 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
811 TL/06 Add Experimental DMARC support using libopendmarc libraries.
813 TL/07 Fix an out of order global option causing a segfault. Reported to dev
814 mailing list by by Dmitry Isaikin.
816 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
818 JH/02 Support "G" suffix to numbers in ${if comparisons.
820 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
822 NM/01 Bugzilla 1197 - Spec typo
823 Bugzilla 1196 - Spec examples corrections
825 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
827 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
828 gnutls_enable_pkcs11, but renamed to more accurately indicate its
831 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
832 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
834 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
835 "acl {{name}{arg}...}", and optional args on acl condition
838 JH/05 Permit multiple router/transport headers_add/remove lines.
840 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
842 JH/07 Avoid using a waiting database for a single-message-only transport.
843 Performance patch from Paul Fisher. Bugzilla 1262.
845 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
848 JH/09 Add $headers_added variable, with content from use of ACL modifier
849 add_header (but not yet added to the message). Bugzilla 199.
851 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
852 Pulled from Bugzilla 817 by Wolfgang Breyha.
854 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
856 (nb: this is the same fix as in Exim 4.80.1)
858 JH/11 Add A= logging on delivery lines, and a client_set_id option on
861 JH/12 Add optional authenticated_sender logging to A= and a log_selector
864 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
866 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
867 advertise SMTP AUTH mechanism to us, instead of a generic
868 protocol violation error. Also, make Exim more robust to bad
869 data from the Dovecot auth socket.
871 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
873 When a queue runner is handling a message, Exim first routes the
874 recipient addresses, during which it prunes them based on the retry
875 hints database. After that it attempts to deliver the message to
876 any remaining recipients. It then updates the hints database using
879 So if a recipient address works intermittently, it can get repeatedly
880 deferred at routing time. The retry hints record remains fresh so the
881 address never reaches the final cutoff time.
883 This is a fairly common occurrence when a user is bumping up against
884 their storage quota. Exim had some logic in its local delivery code
885 to deal with this. However it did not apply to per-recipient defers
886 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
888 This change adds a proper retry rule check during routing so that the
889 final cutoff time is checked against the message's age. We only do
890 this check if there is an address retry record and there is not a
891 domain retry record; this implies that previous attempts to handle
892 the address had the retry_use_local_parts option turned on. We use
893 this as an approximation for the destination being like a local
894 delivery, as in LMTP.
896 I suspect this new check makes the old local delivery cutoff check
897 redundant, but I have not verified this so I left the code in place.
899 TF/02 Correct gecos expansion when From: is a prefix of the username.
901 Test 0254 submits a message to Exim with the header
905 When I ran the test suite under the user fanf2, Exim expanded
906 the header to contain my full name, whereas it should have added
907 a Resent-Sender: header. It erroneously treats any prefix of the
908 username as equal to the username.
910 This change corrects that bug.
912 GF/01 DCC debug and logging tidyup
913 Error conditions log to paniclog rather than rejectlog.
914 Debug lines prefixed by "DCC: " to remove any ambiguity.
916 TF/03 Avoid unnecessary rebuilds of lookup-related code.
918 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
919 Bug spotted by Jeremy Harris; was flawed since initial commit.
920 Would have resulted in OCSP responses post-SNI triggering an Exim
921 NULL dereference and crash.
923 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
925 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
926 Bug detection, analysis and fix by Samuel Thibault.
927 Bugzilla 1331, Debian bug #698092.
929 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
931 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
932 Server implementation by Todd Lyons, client by JH.
933 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
934 config variable "prdr_enable" controls whether the server
935 advertises the facility. If the client requests PRDR a new
936 acl_data_smtp_prdr ACL is called once for each recipient, after
937 the body content is received and before the acl_smtp_data ACL.
938 The client is controlled by both of: a hosts_try_prdr option
939 on the smtp transport, and the server advertisement.
940 Default client logging of deliveries and rejections involving
941 PRDR are flagged with the string "PRDR".
943 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
944 fclose(). Diagnosis by Todd Lyons.
946 PP/17 Update configure.default to handle IPv6 localhost better.
947 Patch by Alain Williams (plus minor tweaks).
950 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
951 This is now consistent with GnuTLS, and is now documented: the
952 previous undocumented portable approach to treating the option as
953 unset was to force an expansion failure. That still works, and
954 an empty string is now equivalent.
956 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
957 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
958 not performing validation itself.
960 PP/20 Added force_command boolean option to pipe transport.
961 Patch from Nick Koston, of cPanel Inc.
963 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
966 TF/04 Added udpsend ACL modifier and hexquote expansion operator
968 PP/21 Fix eximon continuous updating with timestamped log-files.
969 Broken in a format-string cleanup in 4.80, missed when I repaired the
970 other false fix of the same issue.
971 Report and fix from Heiko Schlichting.
974 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
975 Report from Prashanth Katuri.
977 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
978 It's SecureTransport, so affects any MacOS clients which use the
979 system-integrated TLS libraries, including email clients.
981 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
982 using a MIME ACL for non-SMTP local injection.
983 Report and assistance in diagnosis by Warren Baker.
985 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
987 JH/16 Fix comparisons for 64b. Bugzilla 1385.
989 TL/09 Add expansion variable $authenticated_fail_id to keep track of
990 last id that failed so it may be referenced in subsequent ACL's.
992 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
995 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
996 ldap library initialization, allowing self-signed CA's to be
997 used. Also properly sets require_cert option later in code by
998 using NULL (global ldap config) instead of ldap handle (per
999 session). Bug diagnosis and testing by alxgomz.
1001 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1002 the src/util/ subdirectory.
1004 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1005 renamed to Transport Post Delivery Action by Jeremy Harris, as
1008 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1009 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1010 redis_servers = needs to be configured which will be used by the redis
1011 lookup. Patch from Warren Baker, of The Packet Hub.
1013 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1015 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1016 hostname or reverse DNS when processing a host list. Used suggestions
1017 from multiple comments on this bug.
1019 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1021 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1022 Missed a few lines, added it to make the runtest require no keyboard
1025 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1026 contains upper case chars. Make router use caseful_local_part.
1028 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1029 support when GnuTLS has been built with p11-kit.
1035 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1037 This, or similar/improved, will also be change PP/11 of 4.82.
1043 PP/01 Handle short writes when writing local log-files.
1044 In practice, only affects FreeBSD (8 onwards).
1045 Bugzilla 1053, with thanks to Dmitry Isaikin.
1047 NM/01 Bugzilla 949 - Documentation tweak
1049 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1052 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1054 PP/02 Implemented gsasl authenticator.
1056 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1058 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1059 `pkg-config foo` for cflags/libs.
1061 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1062 with rest of GSASL and with heimdal_gssapi.
1064 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1065 `pkg-config foo` for cflags/libs for the TLS implementation.
1067 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1068 properties get this fed in as external SSF. A number of robustness
1069 and debugging improvements to the cyrus_sasl authenticator.
1071 PP/08 cyrus_sasl server now expands the server_realm option.
1073 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1074 Patch by Jeremy Harris.
1076 PP/10 Added dbmjz lookup type.
1078 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1080 PP/12 MAIL args handles TAB as well as SP, for better interop with
1081 non-compliant senders.
1082 Analysis and variant patch by Todd Lyons.
1084 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1085 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1086 Patch from Dirk Mueller <dmueller@suse.com>
1088 PP/13 tls_peerdn now print-escaped for spool files.
1089 Observed some $tls_peerdn in wild which contained \n, which resulted
1090 in spool file corruption.
1092 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1093 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1094 or write after TLS renegotiation, which otherwise led to messages
1097 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1098 as a tracking header (ie: a signed header comes before the signature).
1099 Patch from Wolfgang Breyha.
1101 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1102 comma-sep list; embedded commas doubled.
1104 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1106 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1108 Report and patch from Dmitry Banschikov.
1110 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1111 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1112 failures. We appear to now support TLS1.1+ with Exim.
1114 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1115 lets Exim select keys and certificates based upon TLS SNI from client.
1116 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1117 before an outbound SMTP session. New log_selector, +tls_sni.
1119 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1120 NULL dereference. Report and patch from Alun Jones.
1122 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1123 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1124 Not seeing resolver debug output on NetBSD, but suspect this is a
1125 resolver implementation change.
1127 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1128 Left warnings. Added "eximon gdb" invocation mode.
1130 PP/21 Defaulting "accept_8bitmime" to true, not false.
1132 PP/22 Added -bw for inetd wait mode support.
1134 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1135 locate the relevant includes and libraries. Made this the default.
1137 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1138 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1140 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1141 This may cause build issues on older platforms.
1143 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1144 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1145 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1146 Added SNI support via GnuTLS too.
1147 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1149 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1151 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1152 Applied second patch from Janne, implementing suggestion to default
1153 multiple-strings-in-record handling to match SPF spec.
1155 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1157 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1158 read-only, out of scope).
1159 Patch from Wolfgang Breyha, report from Stuart Northfield.
1161 PP/29 Fix three issues highlighted by clang analyser static analysis.
1162 Only crash-plausible issue would require the Cambridge-specific
1163 iplookup router and a misconfiguration.
1164 Report from Marcin Mirosław.
1166 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1168 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1169 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1170 As part of this, removing so much warning spew let me fix some minor
1171 real issues in debug logging.
1173 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1174 assignment on my part. Fixed.
1176 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1177 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1178 Janne Snabb (who went above and beyond: thank you).
1180 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1181 string otherwise requires a connection and a bunch more work and it's
1182 relatively easy to get wrong. Should also expose TLS library linkage
1185 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1186 64-bit ${eval} (JH/03).
1188 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1189 GNU libc to support some of the 64-bit stuff, should not lead to
1190 conflicts. Defined before os.h is pulled in, so if a given platform
1191 needs to override this, it can.
1193 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1194 protection layer was required, which is not implemented.
1195 Bugzilla 1254, patch from Wolfgang Breyha.
1197 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1198 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1199 tls_dhparam take prime identifiers. Also unbreak combination of
1200 OpenSSL+DH_params+TLSSNI.
1202 PP/39 Disable SSLv2 by default in OpenSSL support.
1208 PP/01 Solaris build fix for Oracle's LDAP libraries.
1209 Bugzilla 1109, patch from Stephen Usher.
1211 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1213 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1216 TF/02 Fix a couple more cases where we did not log the error message
1217 when unlink() failed. See also change 4.74-TF/03.
1219 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1220 lock up or crash if it happened to be inside a call to libc when it
1221 got a SIGUSR1 from exiwhat.
1223 The SIGUSR1 handler appends the current process status to the process
1224 log which is later printed by exiwhat. It used to use the general
1225 purpose logging code to do this, but several functions it calls are
1226 not safe for signals.
1228 The new output code in the SIGUSR1 handler is specific to the process
1229 log, and simple enough that it's easy to inspect for signal safety.
1230 Removing some special cases also simplifies the general logging code.
1231 Removing the spurious timestamps from the process log simplifies
1234 TF/04 Improved ratelimit ACL condition.
1236 The /noupdate option has been deprecated in favour of /readonly which
1237 has clearer semantics. The /leaky, /strict, and /readonly update modes
1238 are mutually exclusive. The update mode is no longer included in the
1239 database key; it just determines when the database is updated. (This
1240 means that when you upgrade Exim will forget old rate measurements.)
1242 Exim now checks that the per_* options are used with an update mode that
1243 makes sense for the current ACL. For example, when Exim is processing a
1244 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1245 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1246 must specify per_mail/readonly. If you omit the update mode it defaults to
1247 /leaky where that makes sense (as before) or /readonly where required.
1249 The /noupdate option is now undocumented but still supported for
1250 backwards compatibility. It is equivalent to /readonly except that in
1251 ACLs where /readonly is required you may specify /leaky/noupdate or
1252 /strict/noupdate which are treated the same as /readonly.
1254 A useful new feature is the /count= option. This is a generalization
1255 of the per_byte option, so that you can measure the throughput of other
1256 aggregate values. For example, the per_byte option is now equivalent
1257 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1259 The per_rcpt option has been generalized using the /count= mechanism
1260 (though it's more complicated than the per_byte equivalence). When it is
1261 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1262 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1263 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1264 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1265 Note that using per_rcpt with a non-readonly update mode in more than
1266 one ACL will cause the recipients to be double-counted. (The per_mail
1267 and per_byte options don't have this problem.)
1269 The handling of very low rates has changed slightly. If the computed rate
1270 is less than the event's count (usually one) then this event is the first
1271 after a long gap. In this case the rate is set to the same as this event's
1272 count, so that the first message of a spam run is counted properly.
1274 The major new feature is a mechanism for counting the rate of unique
1275 events. The new per_addr option counts the number of different
1276 recipients that someone has sent messages to in the last time period. It
1277 behaves like per_rcpt if all the recipient addresses are different, but
1278 duplicate recipient addresses do not increase the measured rate. Like
1279 the /count= option this is a general mechanism, so the per_addr option
1280 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1281 example, measure the rate that a client uses different sender addresses
1282 with the options per_mail/unique=$sender_address. There are further
1283 details in the main documentation.
1285 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1287 TF/06 Removed a few PCRE remnants.
1289 TF/07 Automatically extract Exim's version number from tags in the git
1290 repository when doing development or release builds.
1292 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1293 Bugzilla 879. Patch from Paul Fisher.
1295 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1296 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1299 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1301 PP/05 Variable $av_failed, true if the AV scanner deferred.
1302 Bugzilla 1078. Patch from John Horne.
1304 PP/06 Stop make process more reliably on build failure.
1305 Bugzilla 1087. Patch from Heiko Schlittermann.
1307 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1308 Bugzilla 1089. Patch from Heiko Schlittermann.
1310 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1311 Bugzilla 1131. Patch from Holger Weiß.
1313 PP/09 Handle IPv6 addresses with SPF.
1314 Bugzilla 860. Patch from Wolfgang Breyha.
1316 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1318 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1321 PP/11 match_* no longer expand right-hand-side by default.
1322 New compile-time build option, EXPAND_LISTMATCH_RHS.
1323 New expansion conditions, "inlist", "inlisti".
1325 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1327 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1329 PP/14 fix log_write() format string regression from TF/03.
1330 Bugzilla 1152. Patch from Dmitry Isaikin.
1336 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1338 PP/02 Harmonised TLS library version reporting; only show if debugging.
1339 Layout now matches that introduced for other libraries in 4.74 PP/03.
1341 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1343 PP/04 New "dns_use_edns0" global option.
1345 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1348 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1349 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1351 TK/01 Updated PolarSSL code to 0.14.2.
1352 Bugzilla 1097. Patch from Andreas Metzler.
1354 PP/07 Catch divide-by-zero in ${eval:...}.
1355 Fixes bugzilla 1102.
1357 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1360 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1361 format-string attack -- SECURITY: remote arbitrary code execution.
1363 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1364 time unintentionally subject to list matching rules, letting the header
1365 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1366 arbitrary string expansion). This allowed for information disclosure.
1368 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1369 INT_MIN/-1 -- value coerced to INT_MAX.
1375 NM/01 Workaround for PCRE version dependency in version reporting
1378 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1379 This fixes portability to compilers other than gcc, notably
1380 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1382 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1383 makefiles for portability to HP-UX and POSIX correctness.
1385 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1386 Also via indented variable definition in the Makefile.
1387 (Debugging by Oliver Heesakkers).
1389 PP/02 Restore caching of spamd results with expanded spamd_address.
1390 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1392 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1393 Improves build reliability. Fix from: Frank Elsner
1395 NM/02 Fix wide character breakage in the rfc2047 coding
1396 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1398 NM/03 Allow underscore in dnslist lookups
1399 Fixes bug 1026. Patch from Graeme Fowler
1401 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1402 Code patches from Adam Ciarcinski of NetBSD.
1404 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1407 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1408 is logged, to avoid truncation. Patch from John Horne.
1410 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1411 Patch from Jakob Hirsch.
1413 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1414 SQL string expansion failure details.
1415 Patch from Andrey Oktyabrski.
1417 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1418 Patch from Simon Arlott.
1420 PP/09 New lookups functionality failed to compile on old gcc which rejects
1421 extern declarations in function scope.
1422 Patch from Oliver Fleischmann
1424 PP/10 Use sig_atomic_t for flags set from signal handlers.
1425 Check getgroups() return and improve debugging.
1426 Fixed developed for diagnosis in bug 927 (which turned out to be
1429 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1430 Patch from Mark Zealey.
1432 PP/12 Bugzilla 1056: Improved spamd server selection.
1433 Patch from Mark Zealey.
1435 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1436 Based on patch from Heiko Schlittermann.
1438 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1439 Patch from Uwe Doering, sign-off by Michael Haardt.
1441 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1442 variable declaration deep within a block. Bug and patch from
1445 PP/15 lookups-Makefile IRIX compatibility coercion.
1447 PP/16 Make DISABLE_DKIM build knob functional.
1449 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1450 Patch by Simon Arlott
1452 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1453 variable argument macros. Our copy now differs from upstream.
1459 TF/01 Failure to get a lock on a hints database can have serious
1460 consequences so log it to the panic log.
1462 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1463 controlled using the smtp_confirmation log selector.
1465 TF/03 Include the error message when we fail to unlink a spool file.
1467 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1468 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1469 for maintaining out-of-tree patches for some time.
1471 PP/01 Bugzilla 139: Documentation and portability issues.
1472 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1473 Handle per-OS dynamic-module compilation flags.
1475 PP/02 Let /dev/null have normal permissions.
1476 The 4.73 fixes were a little too stringent and complained about the
1477 permissions on /dev/null. Exempt it from some checks.
1478 Reported by Andreas M. Kirchwitz.
1480 PP/03 Report version information for many libraries, including
1481 Exim version information for dynamically loaded libraries. Created
1482 version.h, now support a version extension string for distributors
1483 who patch heavily. Dynamic module ABI change.
1485 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1486 privilege escalation vulnerability whereby the Exim run-time user
1487 can cause root to append content of the attacker's choosing to
1490 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1493 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1494 If dropping privileges for untrusted macros, we disabled normal logging
1495 on the basis that it would fail; for the Exim run-time user, this is not
1496 the case, and it resulted in successful deliveries going unlogged.
1497 Fixed. Reported by Andreas Metzler.
1503 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1504 only prepend for the Resent-* case. Fixes regression introduced in
1505 Exim 4.70 by NM/22 for Bugzilla 607.
1507 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1508 increasing numbers of administrators be bitten by this.
1510 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1512 PP/03 Bugzilla 994: added openssl_options main configuration option.
1514 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1516 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1518 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1520 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1521 without a peer certificate, leading to a segfault because of an
1522 assumption that peers always have certificates. Be a little more
1523 paranoid. Problem reported by Martin Tscholak.
1525 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1526 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1527 NB: ClamAV planning to remove STREAM in "middle of 2010".
1528 CL also introduces -bmalware, various -d+acl logging additions and
1529 more caution in buffer sizes.
1531 PP/09 Implemented reverse_ip expansion operator.
1533 PP/10 Bugzilla 937: provide a "debug" ACL control.
1535 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1537 PP/12 Bugzilla 973: Implement --version.
1539 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1541 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1543 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1545 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1546 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1547 ignore trailing whitespace.
1549 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1551 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1554 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1555 Notification from Dr Andrew Aitchison.
1557 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1558 ExtendedDetectionInfo response format.
1559 Notification from John Horne.
1561 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1564 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1565 XSL and documented dependency on system catalogs, with examples of how
1568 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1571 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1572 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1573 configuration file which is writeable by the Exim user or group.
1575 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1576 of configuration files to cover files specified with the -C option if
1577 they are going to be used with root privileges, not just the default
1580 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1581 option (effectively making it always true).
1583 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1584 files to be used while preserving root privileges.
1586 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1587 that rogue child processes cannot use them.
1589 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1590 run-time user, instead of root.
1592 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1593 Exim run-time user without dropping privileges.
1595 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1596 result string, instead of calling string_vformat() twice with the same
1599 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1600 for other users. Others should always drop root privileges if they use
1601 -C on the command line, even for a whitelisted configure file.
1603 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1605 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1611 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1612 $data_path, and $header_path variables; fixed documentation bugs and
1615 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1616 exipick to access non-standard spools, including the "frozen" queue
1619 NM/01 Bugzilla 965: Support mysql stored procedures.
1620 Patch from Alain Williams
1622 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1624 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1625 Patch from Andreas Metzler
1627 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1628 Patch from Kirill Miazine
1630 NM/05 Bugzilla 671: Added umask to procmail example.
1632 JJ/03 installed exipick 20100323.0, fixing doc bug
1634 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1635 directory. Notification and patch from Dan Rosenberg.
1637 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1639 TK/02 Improve log output when DKIM signing operation fails.
1641 MH/01 Treat the transport option dkim_domain as a colon separated
1642 list, not as a single string, and sign the message with each element,
1643 omitting multiple occurences of the same signer.
1645 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1646 Bugzilla 985, 986. Patch by Simon Arlott
1648 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1649 Patch by Simon Arlott
1651 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1652 MBX locking. Notification from Dan Rosenberg.
1658 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1660 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1662 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1664 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1666 NM/04 Bugzilla 915: Flag broken perl installation during build.
1672 TK/01 Added patch by Johannes Berg that expands the main option
1673 "spamd_address" if it starts with a dollar sign.
1675 TK/02 Write list of recipients to X-Envelope-Sender header when building
1676 the mbox-format spool file for content scanning (suggested by Jakob
1679 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1680 (http://www.dcc-servers.net/) support via dccifd. Activated by
1681 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1683 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1684 by Mark Daniel Reidel <mr@df.eu>.
1686 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1687 When building exim an external PCRE library is now needed -
1688 PCRE is a system library on the majority of modern systems.
1689 See entry on PCRE_LIBS in EDITME file.
1691 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1692 conversation. Added nologin parameter to request.
1693 Patch contributed by Kirill Miazine.
1695 TF/01 Do not log submission mode rewrites if they do not change the address.
1697 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1699 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1700 log files in place. Contributed by Roberto Lima.
1702 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1704 TF/03 Bugzilla 615: When checking the local_parts router precondition
1705 after a local_part_suffix or local_part_prefix option, Exim now
1706 does not use the address's named list lookup cache, since this
1707 contains cached lookups for the whole local part.
1709 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1710 Robert Millan. Documentation is in experimental-spec.txt.
1712 TF/04 Bugzilla 668: Fix parallel build (make -j).
1714 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1716 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1717 Patch provided by Jan Srzednicki.
1719 TF/05 Leading white space used to be stripped from $spam_report which
1720 wrecked the formatting. Now it is preserved.
1722 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1723 that they are available at delivery time.
1725 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1727 TF/08 TLS error reporting now respects the incoming_interface and
1728 incoming_port log selectors.
1730 TF/09 Produce a more useful error message if an SMTP transport's hosts
1731 setting expands to an empty string.
1733 NM/06 Bugzilla 744: EXPN did not work under TLS.
1734 Patch provided by Phil Pennock.
1736 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1737 Patch provided by Richard Godbee.
1739 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1740 acl_smtp_notquit, added index entry.
1742 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1743 Patch provided by Eugene Bujak.
1745 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1746 accept(). Patch provided by Maxim Dounin.
1748 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1749 Patch provided by Phil Pennock.
1751 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1753 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1754 Patch provided by Brad "anomie" Jorsch.
1756 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1757 Patch provided by Dean Brooks.
1759 TK/05 Add native DKIM support (does not depend on external libraries).
1761 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1762 Patch provided by Graeme Fowler.
1764 NM/16 Bugzilla 851: Documentation example syntax fix.
1766 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1768 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1771 NM/19 Bugzilla 745: TLS version reporting.
1772 Patch provided by Phil Pennock.
1774 NM/20 Bugzilla 167: bool: condition support.
1775 Patch provided by Phil Pennock.
1777 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1778 clients. Patch provided by Phil Pennock.
1780 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1781 Patch provided by Brad "anomie" Jorsch.
1783 NM/23 Bugzilla 687: Fix misparses in eximstats.
1784 Patch provided by Heiko Schlittermann.
1786 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1787 Patch provided by Heiko Schlittermann.
1789 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1790 plus update to original patch.
1792 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1794 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1795 Patch provided by David Brownlee.
1797 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1799 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1801 NM/30 Bugzilla 888: TLS documentation bugfixes.
1803 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1805 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1806 Unlike the original bugzilla I have changed all shell scripts in src tree.
1808 NM/33 Bugzilla 898: Transport filter timeout fix.
1809 Patch by Todd Rinaldo.
1811 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1812 Patch by Serge Demonchaux.
1814 NM/35 Bugzilla 39: Base64 decode bug fixes.
1815 Patch by Jakob Hirsch.
1817 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1819 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1821 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1823 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1829 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1830 ALT-N's libdkim that I have put here:
1831 http://duncanthrax.net/exim-experimental/
1833 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1834 were called 'true' and it seems that C99 defines that as a reserved
1835 keyword to be used with 'bool' variable types. That means you could
1836 not include C99-style headers which use bools without triggering
1837 build errors in sieve.c.
1839 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1840 as mailq or other aliases. Changed the --help handling significantly
1841 to do whats expected. exim_usage() emits usage/help information.
1843 SC/01 Added the -bylocaldomain option to eximstats.
1845 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1847 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1849 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1855 PH/01 Another patch from the Sieve maintainer.
1857 PH/02 When an IPv6 address is converted to a string for single-key lookup
1858 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1859 dots are used instead of colons so that keys in lsearch files need not
1860 contain colons. This was done some time before quoting was made available
1861 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1862 (notated using the quote facility) so as to distinguish them from IPv4
1863 keys. This meant that lookups for IP addresses in host lists did not work
1864 for iplsearch lookups.
1866 This has been fixed by arranging for IPv6 addresses to be expressed with
1867 colons if the lookup type is iplsearch. This is not incompatible, because
1868 previously such lookups could never work.
1870 The situation is now rather anomalous, since one *can* have colons in
1871 ordinary lsearch keys. However, making the change in all cases is
1872 incompatible and would probably break a number of configurations.
1874 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1877 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1878 conversion specification without a maximum field width, thereby enabling
1879 a rogue spamd server to cause a buffer overflow. While nobody in their
1880 right mind would setup Exim to query an untrusted spamd server, an
1881 attacker that gains access to a server running spamd could potentially
1882 exploit this vulnerability to run arbitrary code as the Exim user.
1884 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1885 $primary_hostname instead of what libspf2 thinks the hosts name is.
1887 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1888 a directory entry by the name of the lookup key. Previously, if a
1889 symlink pointed to a non-existing file or a file in a directory that
1890 Exim lacked permissions to read, a lookup for a key matching that
1891 symlink would fail. Now it is enough that a matching directory entry
1892 exists, symlink or not. (Bugzilla 503.)
1894 PH/03 The body_linecount and body_zerocount variables are now exported in the
1897 PH/04 Added the $dnslist_matched variable.
1899 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1900 This means they are set thereafter only if the connection becomes
1903 PH/06 Added the client_condition to authenticators so that some can be skipped
1904 by clients under certain conditions.
1906 PH/07 The error message for a badly-placed control=no_multiline_responses left
1907 "_responses" off the end of the name.
1909 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1911 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1912 (without spaces) instead of just copying the configuration text.
1914 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1916 PH/11 Added $max_received_linelength.
1918 PH/12 Added +ignore_defer and +include_defer to host lists.
1920 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1921 way in which PCRE > 7.0 is built.
1923 PH/14 Implemented queue_only_load_latch.
1925 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1926 MAIL command. The effect was to mangle the value on 64-bit systems.
1928 PH/16 Another patch from the Sieve maintainer.
1930 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1932 PH/18 If a system quota error occurred while trying to create the file for
1933 a maildir delivery, the message "Mailbox is full" was not appended to the
1934 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1935 only to a quota excession during the actual writing of the file.
1937 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1938 characters?) which causes problems in log lines. The DN values are now
1939 passed through string_printing() before being added to log lines.
1941 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1942 and InterBase are left for another time.)
1944 PH/21 Added message_body_newlines option.
1946 PH/22 Guard against possible overflow in moan_check_errorcopy().
1948 PH/23 POSIX allows open() to be a macro; guard against that.
1950 PH/24 If the recipient of an error message contained an @ in the local part
1951 (suitably quoted, of course), incorrect values were put in $domain and
1952 $local_part during the evaluation of errors_copy.
1958 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1959 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1962 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1963 issue a MAIL command.
1965 PH/02 In an ACL statement such as
1967 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1969 if a client was not listed at all, or was listed with a value other than
1970 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1971 the condition was not true (as it should be), so access was not denied.
1972 The bug was that the ! inversion was incorrectly passed on to the second
1973 item. This has been fixed.
1975 PH/03 Added additional dnslists conditions == and =& which are different from
1976 = and & when the dns lookup returns more than one IP address.
1978 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1979 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1981 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1982 FSYNC, which compiles an option called disable_fsync that allows for
1983 bypassing fsync(). The documentation is heavily laced with warnings.
1985 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1987 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1988 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1989 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1990 including adding "make clean"; (3) Added -fPIC when compiling the test
1991 dynamically loaded module, to get rid of a warning.
1993 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1994 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1995 = 0s. The bug is otherwise harmless.
1997 PH/07 There was a bug in the dovecot authenticator such that the value of
1998 $auth1 could be overwritten, and so not correctly preserved, after a
1999 successful authentication. This usually meant that the value preserved by
2000 the server_setid option was incorrect.
2002 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2004 PH/09 Installed PCRE release 7.0.
2006 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2007 run for batched SMTP input. It is now run at the start of every message
2008 in the batch. While fixing this I discovered that the process information
2009 (output by running exiwhat) was not always getting set for -bs and -bS
2010 input. This is fixed, and it now also says "batched" for BSMTP.
2012 PH/11 Added control=no_pipelining.
2014 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2015 patch, slightly modified), and move the expansion of helo_data till after
2016 the connection is made in the smtp transport (so it can use these
2019 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2021 PH/14 Added log_selector = +pid.
2023 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2025 PH/16 Add ${if forany and ${if forall.
2027 PH/17 Added dsn_from option to vary the From: line in DSNs.
2029 PH/18 Flush SMTP output before performing a callout, unless control =
2030 no_callout_flush is set.
2032 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2033 was true (the default) a successful delivery failed to delete the retry
2034 item, thus causing premature timeout of the address. The bug is now
2037 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2039 PH/21 Long custom messages for fakedefer and fakereject are now split up
2040 into multiline responses in the same way that messages for "deny" and
2041 other ACL rejections are.
2043 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2044 with slight modification.
2046 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2047 draft, changing the syntax and factoring some duplicate code".
2049 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2050 for deliveries of the second and subsequent messages over the same SMTP
2053 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2054 ${reduce, with only minor "tidies".
2056 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2058 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2059 expansion side effects.
2061 PH/27 When a message times out after an over-quota error from an Exim-imposed
2062 quota, the bounce message says "mailbox is full". This message was not
2063 being given when it was a system quota that was exceeded. It now should
2066 MH/03 Made $recipients available in local_scan(). local_scan() already has
2067 better access to the recipient list through recipients_list[], but
2068 $recipients can be useful in postmaster-provided expansion strings.
2070 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2071 in the case of a MAIL command with additional options following the
2072 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2073 were accidentally chopped off.
2075 PH/29 SMTP synchronization checks are implemented when a command is read -
2076 there is a check that no more input is waiting when there shouldn't be
2077 any. However, for some commands, a delay in an ACL can mean that it is
2078 some time before the response is written. In this time, more input might
2079 arrive, invalidly. So now there are extra checks after an ACL has run for
2080 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2081 pipelining has not been advertised.
2083 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2085 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2086 returned in the SMTP response when smtp_return_error_details was false.
2087 This has been fixed.
2089 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2090 instead of the C I/O that was originally supplied, because problems were
2091 reported on Solaris.
2093 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2094 Exim which did not show up earlier: it was assuming that a call to
2095 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2096 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2097 was a macro that became an assignment, so it seemed to work. This has
2098 changed to a proper function call with a void return, hence the compile
2099 error. Exim's code has been fixed.
2101 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2104 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2106 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2108 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2109 because it turns out to be too big or there is a timeout) there is no
2110 "Completed" line in the log. When some messages of this type were
2111 selected by exigrep, they were listed as "not completed". Others were
2112 picked up by some special patterns. I have improved the selection
2113 criteria to be more general.
2115 PH/38 The host_find_failed option in the manualroute router can now be set
2116 to "ignore", to completely ignore a host whose IP address cannot be
2117 found. If all hosts are ignored, the behaviour is controlled by the new
2118 host_all_ignored option.
2120 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2121 homing or because of multiple MX records with /mx) generated more than
2122 one IP address, and the following item turned out to be the local host,
2123 all the secondary addresses of the first item were incorrectly removed
2124 from the list, along with the local host and any following hosts (which
2125 is what is supposed to happen).
2127 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2128 whoever called Exim into the -H file. In the case of the daemon it was
2129 behaving confusingly. When first started, it used values for whoever
2130 started the daemon, but after a SIGHUP it used the Exim user (because it
2131 calls itself on a restart). I have changed the code so that it now always
2134 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2135 message are rejected with the same error (e.g. no authentication or bad
2136 sender address), and a DATA command is nevertheless sent (as can happen
2137 with PIPELINING or a stupid MUA), the error message that was given to the
2138 RCPT commands is included in the rejection of the DATA command. This is
2139 intended to be helpful for MUAs that show only the final error to their
2142 PH/42 Another patch from the Sieve maintainer.
2144 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2145 Eximstats - Fixed some broken HTML links and added missing column headers
2147 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2148 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2150 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2152 PH/43 Yet another patch from the Sieve maintainer.
2154 PH/44 I found a way to check for a TCP/IP connection going away before sending
2155 the response to the final '.' that terminates a message, but only in the
2156 case where the client has not sent further data following the '.'
2157 (unfortunately, this is allowed). However, in many cases there won't be
2158 any further data because there won't be any more messages to send. A call
2159 to select() can be used: if it shows that the input is "ready", there is
2160 either input waiting, or the socket has been closed. An attempt to read
2161 the next input character can distinguish the two cases. Previously, Exim
2162 would have sent an OK response which the client would never have see.
2163 This could lead to message repetition. This fix should cure that, at
2164 least in a lot of common cases.
2166 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2167 advertised in response to EHLO.
2173 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2174 fixed by 4.65/MH/01 (is this a record?) are fixed:
2176 (i) An empty string was always treated as zero by the numeric comparison
2177 operators. This behaviour has been restored.
2179 (ii) It is documented that the numeric comparison operators always treat
2180 their arguments as decimal numbers. This was broken in that numbers
2181 starting with 0 were being interpreted as octal.
2183 While fixing these problems I realized that there was another issue that
2184 hadn't been noticed. Values of message_size_limit (both the global option
2185 and the transport option) were treated as octal if they started with 0.
2186 The documentation was vague. These values are now always treated as
2187 decimal, and I will make that clear in the documentation.
2193 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2194 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2197 MH/01 Don't check that the operands of numeric comparison operators are
2198 integers when their expansion is in "skipping" mode (fixes bug
2199 introduced by 4.64-PH/07).
2201 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2202 child addresses, Exim now panics and dies. Previously, because the count
2203 is held in a short int, deliveries were likely to be lost. As such a
2204 large number of recipients for a single message is ridiculous
2205 (performance will be very, very poor), I have chosen to impose a limit
2206 rather than extend the field.
2212 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2213 leftover -K file (the existence of which was triggered by #402).
2214 While we were at it, introduced process PID as part of the -K
2215 filename. This should rule out race conditions when creating
2218 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2219 processing considerably. Previous code took too long for large mails,
2220 triggering a timeout which in turn triggers #401.
2222 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2223 in the DK code in transports.c. sendfile() is not really portable,
2224 hence the _LINUX specificness.
2226 TF/01 In the add_headers option to the mail command in an Exim filter,
2227 there was a bug that Exim would claim a syntax error in any
2228 header after the first one which had an odd number of characters
2231 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2232 callout verification, Exim cached a "reject" for the entire domain. This
2233 is correct for most verifications, but it is not correct for a recipient
2234 verification with use_sender or use_postmaster set, because in that case
2235 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2236 case of MAIL FROM:<> rejection from other early rejections (e.g.
2237 rejection of HELO). When verifying a recipient using a non-null MAIL
2238 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2239 Whatever the result of the callout, the value of the domain cache is
2240 left unchanged (for any other kind of callout, getting as far as trying
2241 RCPT means that the domain itself is ok).
2243 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2246 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2247 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2248 session, the daemon ignores ECONNECT errors and logs others; it now
2249 ignores EPIPE as well.
2251 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2252 (quoted-printable decoding).
2254 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2255 later the small subsequent patch to fix an introduced bug.
2257 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2259 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2261 PH/08 An error is now given if message_size_limit is specified negative.
2263 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2264 to be given (somewhat) arbitrary names.
2266 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2269 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2270 miscellaneous code fixes
2272 PH/10 Added the log_reject_target ACL modifier to specify where to log
2275 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2276 hostname. This is wrong, because it relates to the incoming message (and
2277 probably the interface on which it is arriving) and not to the outgoing
2278 callout (which could be using a different interface). This has been
2279 changed to use the value of the helo_data option from the smtp transport
2280 instead - this is what is used when a message is actually being sent. If
2281 there is no remote transport (possible with a router that sets up host
2282 addresses), $smtp_active_hostname is used.
2284 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2285 tweaks were necessary in order to get it to work (see also 21 below):
2286 (a) The code assumed that strncpy() returns a negative number on buffer
2287 overflow, which isn't the case. Replaced with Exim's string_format()
2289 (b) There were several signed/unsigned issues. I just did the minimum
2290 hacking in of casts. There is scope for a larger refactoring.
2291 (c) The code used strcasecmp() which is not a standard C function.
2292 Replaced with Exim's strcmpic() function.
2293 (d) The code set only $1; it now sets $auth1 as well.
2294 (e) A simple test gave the error "authentication client didn't specify
2295 service in request". It would seem that Dovecot has changed its
2296 interface. Fortunately there's a specification; I followed it and
2297 changed what the client sends and it appears to be working now.
2299 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2302 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2303 address A is aliased to B and C, where B exists and C does not. Without
2304 -v the output is "A verified" because verification stops after a
2305 successful redirection if more than one address is generated. However,
2306 with -v the child addresses are also verified. Exim was outputting "A
2307 failed to verify" and then showing the successful verification for C,
2308 with its parentage. It now outputs "B failed to verify", showing B's
2309 parentage before showing the successful verification of C.
2311 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2312 look up a TXT record in a specific list after matching in a combined
2315 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2316 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2317 they consult the DNS. I had assumed they would set it the way they
2318 wanted; and indeed my experiments on Linux seem to show that in some
2319 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2320 To be on the safe side, however, I have now made the interface to
2321 host_find_byname() similar to host_find_bydns(), with an argument
2322 containing the DNS resolver options. The host_find_byname() function now
2323 sets these options at its start, just as host_find_bydns() does. The smtp
2324 transport options dns_qualify_single and dns_search_parents are passed to
2325 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2326 of host_find_byname() use the default settings of RES_DEFNAMES
2327 (qualify_single) but not RES_DNSRCH (search_parents).
2329 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2330 spool_read_header() do less string testing, by means of a preliminary
2331 switch on the second character of optional "-foo" lines. (This is
2332 overdue, caused by the large number of possibilities that now exist.
2333 Originally there were few.) While I was there, I also converted the
2334 str(n)cmp tests so they don't re-test the leading "-" and the first
2335 character, in the hope this might squeeze out yet more improvement.
2337 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2338 flag allowing group syntax was set by the header_syntax check but not
2339 turned off, possible causing trouble later; (2) The flag was not being
2340 set at all for the header_verify test, causing "group"-style headers to
2341 be rejected. I have now set it in this case, and also caused header_
2342 verify to ignore an empty address taken from a group. While doing this, I
2343 came across some other cases where the code for allowing group syntax
2344 while scanning a header line wasn't quite right (mostly, not resetting
2345 the flag correctly in the right place). These bugs could have caused
2346 trouble for malformed header lines. I hope it is now all correct.
2348 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2349 with the "reply" argument non-NULL. The code, however (which originally
2350 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2351 but it didn't always do it. This confused somebody who was copying the
2352 code for some other use. I have removed all the tests.
2354 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2355 feature that was used to support insecure browsers during the U.S. crypto
2356 embargo. It requires special client support, and Exim is probably the
2357 only MTA that supported it -- and would never use it because real RSA is
2358 always available. This code has been removed, because it had the bad
2359 effect of slowing Exim down by computing (never used) parameters for the
2360 RSA_EXPORT functionality.
2362 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2363 authenticator to fail if there's a tab character in the incoming data
2364 (there should never be unless someone is messing about, as it's supposed
2365 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2366 if the connection is using TLS or if the remote IP is the same as the
2367 local IP, and the "valid-client-cert option" if a client certificate has
2370 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2371 authenticators. This can be used for authorization after authentication
2372 succeeds. (In the case of plaintext, it servers for both authentication
2375 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2376 if any retry times were supplied.
2378 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2379 connection, where there is no client IP address to check. In this
2380 situation, the verify now always succeeds.
2382 PH/25 Applied John Jetmore's -Mset patch.
2384 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2386 PH/27 In a string expansion for a processed (not raw) header when multiple
2387 headers of the same name were present, leading whitespace was being
2388 removed from all of them, but trailing whitespace was being removed only
2389 from the last one. Now trailing whitespace is removed from each header
2390 before concatenation. Completely empty headers in a concatenation (as
2391 before) are ignored.
2393 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2394 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2396 PH/29 [Removed. This was a change that I later backed out, and forgot to
2397 correct the ChangeLog entry (that I had efficiently created) before
2398 committing the later change.]
2400 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2401 address errors (4xx response to RCPT) over the same connection as other
2402 messages routed to the same hosts. Such deliveries are always "forced",
2403 so retry times are not inspected. This resulted in far too many retries
2404 for the affected addresses. The effect occurred only when there were more
2405 hosts than the hosts_max_try setting in the smtp transport when it had
2406 the 4xx errors. Those hosts that it had tried were not added to the list
2407 of hosts for which the message was waiting, so if all were tried, there
2408 was no problem. Two fixes have been applied:
2410 (i) If there are any address or message errors in an SMTP delivery, none
2411 of the hosts (tried or untried) are now added to the list of hosts
2412 for which the message is waiting, so the message should not be a
2413 candidate for sending over the same connection that was used for a
2414 successful delivery of some other message. This seems entirely
2415 reasonable: after all the message is NOT "waiting for some host".
2416 This is so "obvious" that I'm not sure why it wasn't done
2417 previously. Hope I haven't missed anything, but it can't do any
2418 harm, as the worst effect is to miss an optimization.
2420 (ii) If, despite (i), such a delivery is accidentally attempted, the
2421 routing retry time is respected, so at least it doesn't keep
2422 hammering the server.
2424 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2425 in ${readsocket because some servers need this prod.
2427 PH/32 Added some extra debug output when updating a wait-xxx database.
2429 PH/33 The hint "could be header name not terminated by colon", which has been
2430 given for certain expansion errors for a long time, was not being given
2431 for the ${if def:h_colon_omitted{... case.
2433 PH/34 The spec says: "With one important exception, whenever a domain list is
2434 being scanned, $domain contains the subject domain." There was at least
2435 one case where this was not true.
2437 PH/35 The error "getsockname() failed: connection reset by peer" was being
2438 written to the panic log as well as the main log, but it isn't really
2439 panic-worthy as it just means the connection died rather early on. I have
2440 removed the panic log writing for the ECONNRESET error when getsockname()
2443 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2444 runs only) independently of the message's sender address. This meant
2445 that, if the 4xx error was in fact related to the sender, a different
2446 message to the same recipient with a different sender could confuse
2447 things. In particular, this can happen when sending to a greylisting
2448 server, but other circumstances could also provoke similar problems.
2449 I have changed the default so that the retry time for these errors is now
2450 based a combination of the sender and recipient addresses. This change
2451 can be overridden by setting address_retry_include_sender=false in the
2454 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2455 remote server are returned as part of bounce messages. This was not
2456 happening for LMTP over a pipe (the lmtp transport), but now it is the
2457 same for both kinds of LMTP.
2459 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2460 in header lines that were in fact CNAMEs. This is no longer the case.
2462 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2463 and queue runs started by the daemon processed all messages. This has
2464 been fixed so that -R and -S can now usefully be given with -q<time>.
2466 PH/40 Import PCRE release 6.7 (fixes some bugs).
2468 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2470 PH/42 Give an error if -q is specified more than once.
2472 PH/43 Renamed the variables $interface_address and $interface_port as
2473 $received_ip_address and $received_port, to make it clear that these
2474 values apply to message reception, and not to the outgoing interface when
2475 a message is delivered. (The old names remain recognized, of course.)
2477 PH/44 There was no timeout on the connect() call when using a Unix domain
2478 socket in the ${readsocket expansion. There now is.
2480 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2481 be meaningful with "accept".
2483 SC/01 Eximstats V1.43
2484 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2486 SC/02 Eximstats V1.44
2487 Use a glob alias rather than an array ref in the generated
2488 parser. This improves both readability and performance.
2490 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2491 Collect SpamAssassin and rejection statistics.
2492 Don't display local sender or destination tables unless
2493 there is data to show.
2494 Added average volumes into the top table text output.
2496 SC/04 Eximstats V1.46
2497 Collect data on the number of addresses (recipients)
2498 as well as the number of messages.
2500 SC/05 Eximstats V1.47
2501 Added 'Message too big' to the list of mail rejection
2502 reasons (thanks to Marco Gaiarin).
2504 SC/06 Eximstats V1.48
2505 Mainlog lines which have GMT offsets and are too short to
2506 have a flag are now skipped.
2508 SC/07 Eximstats V1.49 (Alain Williams)
2509 Added the -emptyok flag.
2511 SC/08 Eximstats V1.50
2512 Fixes for obtaining the IP address from reject messages.
2514 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2515 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2516 whitespace changes from 4.64-PH/27
2518 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2521 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2522 are found, allow negative numbers in numeric criteria)
2524 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2526 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2527 to match changes made in 4.64-PH/43
2529 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2531 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2532 "domain" is actually an IP address, and force a failure. This locks out
2533 those revolvers/nameservers that support "A-for-A" lookups, in
2534 contravention of the specifications.
2536 PH/48 When a host name was looked up from an IP address, and the subsequent
2537 forward lookup of the name timed out, the host name was left in
2538 $sender_host_name, contrary to the specification.
2540 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2541 restricted to single-key lookups, Exim was not diagnosing an error if
2542 * or *@ was used with a query-style lookup.
2544 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2546 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2547 long ago, but noone interested enough thought of it. Let's just say that
2548 the "1.1" means that there are some new functions that weren't there at
2549 some point in the past.
2551 PH/51 Error processing for expansion failure of helo_data from an smtp
2552 transport during callout processing was broken.
2554 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2555 tested/used via the -bh/-bhc/-bs options.
2557 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2558 bug, fixed in subsequent PCRE releases).
2560 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2561 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2563 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2569 SC/01 Use a glob alias rather than an array ref in eximstats generated
2570 parser. This improves both readability and performance.
2572 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2573 Don't display local sender or destination tables in eximstats unless
2574 there is data to show.
2575 Added average volumes into the eximstats top table text output.
2577 SC/03 Collect data on the number of addresses (recipients) as well
2578 as the number of messages in eximstats.
2580 TF/01 Correct an error in the documentation for the redirect router. Exim
2581 does (usually) call initgroups() when daemonizing.
2583 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2584 with consistent privilege compared to when running as a daemon.
2586 TF/03 Note in the spec that $authenticated_id is not set for local
2587 submissions from trusted users.
2589 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2590 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2592 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2593 by adding some example configuration directives to the default
2594 configuration file. A little bit of work is required to uncomment the
2595 directives and define how usernames and passwords are checked, but
2596 there is now a framework to start from.
2598 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2599 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2600 without this. I don't know how relevant this is to other LDAP libraries.
2602 PH/02 Add the verb name to the "unknown ACL verb" error.
2604 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2606 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2608 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2609 directory not expanded when it should be if an expanded home directory
2610 was set for the address (which is overridden by the transport).
2612 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2615 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2616 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2617 because it is too late at that time, and has no effect.
2619 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2620 security issue with \' (bugzilla #107). I could not use the
2621 PQescapeStringConn() function, because it needs a PGconn value as one of
2624 PH/08 When testing addresses using -bt, indicate those final addresses that
2625 are duplicates that would not cause an additional delivery. At least one
2626 person was confused, thinking that -bt output corresponded to deliveries.
2627 (Suppressing duplicates isn't a good idea as you lose the information
2628 about possibly different redirections that led to the duplicates.)
2630 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2631 systems where poll() doesn't work, in particular OS X.
2633 PH/10 Added more information to debugging output for retry time not reached.
2635 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2636 operations in malware.c.
2638 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2641 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2642 syslog_duplication set false, log lines that would normally be written
2643 both the the main log and to the reject log were not written to syslog at
2646 PH/14 In the default configuration, change the use of "message" in ACL warn
2647 statements to "add_header".
2649 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2650 not followed by a command (e.g. "seen endif").
2652 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2653 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2656 PH/17 Added extra conditions to the default value of delay_warning_condition
2660 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2661 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2662 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2665 The Auto-Submitted: and various List- headers are standardised, whereas I
2666 don't think Precedence: ever was.
2668 PH/18 Refactored debugging code in route_finduser() to show more information,
2669 in particular, the error code if getpwnam() issues one.
2671 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2672 This is apparently needed in addition to the PH/07 change above to avoid
2673 any possible encoding problems.
2675 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2676 but not after initializing Perl.
2678 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2679 output them only if debugging. By default they are written stderr,
2680 apparently, which is not desirable.
2682 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2685 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2688 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2690 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2691 authenticated or an ident call has been made. Suppress the default
2692 values for $authenticated_id and $authenticated_sender (but permit -oMai
2693 and -oMas) when testing with -bh.
2695 PH/24 Re-jigged the order of the tests in the default configuration so that the
2696 tests for valid domains and recipients precede the DNS black list and CSA
2697 tests, on the grounds that those ones are more expensive.
2699 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2700 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2701 This bug exists in every version of Exim that I still have, right back to
2704 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2705 However, an attempt to turn on case-sensitivity in a regex key by
2706 including (?-i) didn't work because the subject string was already
2707 lowercased, and the effects were non-intuitive. It turns out that a
2708 one-line patch can be used to allow (?-i) to work as expected.
2714 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2715 other effects) broke the use of negated acl sub-conditions.
2717 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2720 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2721 "Deny" causes Exim to reject the incoming connection with a 554 error.
2722 Unfortunately, if there is a major crisis, such as a disk failure,
2723 tcp-wrappers gives "deny", whereas what one would like would be some
2724 kind of temporary error. A kludge has been added to help with this.
2725 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2726 554 error is used if errno is still zero or contains ENOENT (which occurs
2727 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2730 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2732 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2733 errors. Otherwise a message that provokes a temporary error (when other
2734 messages do not) can cause a whole host to time out.
2736 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2737 addresses were routed directly to files or pipes from a redirect router.
2738 File deliveries just didn't batch; pipe deliveries might have suffered
2741 PH/06 A failure to get a lock for a hints database would erroneously always say
2742 "Failed to get write lock", even when it was really a read lock.
2744 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2745 of 0600. This has been changed to use the value of the lockfile_mode
2746 option (which defaults to 0600).
2748 PH/08 Applied small patch from the Sieve maintainer.
2750 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2751 folder from quota calculations, a direct delivery into this folder messed
2752 up the contents of the maildirsize file. This was because the regex was
2753 used only to exclude .Trash (or whatever) when the size of the mailbox
2754 was calculated. There was no check that a delivery was happening into an
2755 excluded directory. This bug has been fixed by ignoring all quota
2756 processing for deliveries into excluded directories.
2758 PH/10 Added the maildirfolder_create_regex option to appendfile.
2764 PH/01 The code for finding all the local interface addresses on a FreeBSD
2765 system running IPv6 was broken. This may well have applied to all BSD
2766 systems, as well as to others that have similar system calls. The broken
2767 code found IPv4 interfaces correctly, but gave incorrect values for the
2768 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2769 that it would not match correctly against @[] and not recognize the IPv6
2772 PH/02 The ipliteral router was not recognizing addresses of the form user@
2773 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2775 PH/03 Added disable_ipv6.
2777 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2778 decoded form, because it is most often used to construct To: headers
2779 lines in autoreplies, and the decoded form may well be syntactically
2780 invalid. However, $reply_address has leading white space removed, and all
2781 newlines turned into spaces so that the autoreply transport does not
2784 PH/05 If group was specified without a user on a router, and no group or user
2785 was specified on a transport, the group from the router was ignored.
2787 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2788 for visible compile-time settings that can be used to change these
2789 numbers, for those that want even more. Backwards compatibility with old
2790 spool files has been maintained. However, going back to a previous Exim
2791 release will lost any variables that are in spool files.
2793 PH/07 Two small changes when running in the test harness: increase delay when
2794 passing a TCP/IP connection to a new process, in case the original
2795 process has to generate a bounce, and remove special handling of
2796 127.0.0.2 (sic), which is no longer necessary.
2798 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2799 be the same on different OS.
2801 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2804 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2805 whether --show-vars was specified or not
2807 JJ/02 exipick: Added support for new ACL variable spool format introduced
2810 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2811 syntactically invalid From: or Reply-to: line, and a filter used this to
2812 generate an autoreply, and therefore failed to obtain an address for the
2813 autoreply, Exim could try to deliver to a non-existent relative file
2814 name, causing unrelated and misleading errors. What now happens is that
2815 it logs this as a hard delivery error, but does not attempt to create a
2818 PH/11 The exinext utility has a -C option for testing purposes, but although
2819 the given file was scanned by exinext itself; it wasn't being passed on
2820 when Exim was called.
2822 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2823 an end-of-file indication when reading a command response.
2825 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2826 compiled. In many other places in Exim, IPv6 addresses are always
2827 recognized, so I have changed this. It also means that IPv4 domain
2828 literals of the form [IPV4:n.n.n.n] are now always recognized.
2830 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2831 used if the router is not running as root, for example, when verifying at
2832 ACL time, or when using -bh. The debugging output from this situation was
2833 non-existent - all you got was a failure to exec. I have made two
2836 (a) Failures to set uid/gid, the current directory, or a process leader
2837 in a subprocess such as that created by queryprogram now generate
2838 suitable debugging output when -d is set.
2840 (b) The queryprogram router detects when it is not running as root,
2841 outputs suitable debugging information if -d is set, and then runs
2842 the subprocess without attempting to change uid/gid.
2844 PH/15 Minor change to Makefile for building test_host (undocumented testing
2847 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2848 additional section of a DNS packet that returns MX or SRV records.
2849 Instead, it always explicitly searches for A/AAAA records. This avoids
2850 major problems that occur when a DNS server includes only records of one
2851 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2852 fixed another bug: if SRV records were looked up and the corresponding
2853 address records were *not* found in the additional section, the port
2854 values from the SRV records were lost.
2856 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2857 using the correct key (the original address) when searching the retry
2858 rules in order to find which one to use for generating the retry hint.
2860 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2861 adding the default one. Similarly, if it contains a Reply-To: header, the
2862 errors_reply_to option, if set, is not used.
2864 PH/19 When calculating a retry time, Exim used to measure the "time since
2865 failure" by looking at the "first failed" field in the retry record. Now
2866 it does not use this if it is later than than the arrival time of the
2867 message. Instead it uses the arrival time. This makes for better
2868 behaviour in cases where some deliveries succeed, thus re-setting the
2869 "first failed" field. An example is a quota failure for a huge message
2870 when small messages continue to be delivered. Without this change, the
2871 "time since failure" will always be short, possible causing more frequent
2872 delivery attempts for the huge message than are intended.
2873 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2875 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2876 $1, $2, $3) because the numerical variables can be reset during some
2877 expansion items (e.g. "match"), thereby losing the authentication data.
2879 PH/21 Make -bV show the size of off_t variables so that the test suite can
2880 decide whether to run tests for quotas > 2G.
2882 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2883 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2884 filecount value is greater than 2G or if a quota value is greater than 2G
2885 on a system where the size of off_t is not greater than 4, a panic error
2888 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2889 never match. The debug and -bh output now contains an explicit error
2890 message indicating a malformed IPv4 address or mask.
2892 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2893 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2894 PH/23 above applies.
2896 PH/25 Do not write to syslog when running in the test harness. The only
2897 occasion when this arises is a failure to open the main or panic logs
2898 (for which there is an explicit test).
2900 PH/26 Added the /no_tell option to "control=freeze".
2902 PH/27 If a host name lookup failed very early in a connection, for example, if
2903 the IP address matched host_lookup and the reverse lookup yielded a name
2904 that did not have a forward lookup, an error message of the form "no IP
2905 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2906 could be logged. Now it outputs the IP address instead of "NULL".
2908 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2909 allows the sender and the authenticated sender to be set when
2910 submitting a message from within Exim. Since child_open_exim() is
2911 documented for local_scan(), the new function should be too.
2913 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2914 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2915 results in an empty string is now treated as unset.
2917 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2919 PH/31 Added sender_verify_fail logging option.
2921 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2922 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2923 tidied the source and removed it altogether.
2925 PH/33 When a queue run was abandoned because the load average was too high, a
2926 log line was always written; now it is written only if the queue_run log
2927 selector is set. In addition, the log line for abandonment now contains
2928 information about the queue run such as the pid. This is always present
2929 in "start" and "stop" lines but was omitted from the "abandon" line.
2931 PH/34 Omit spaces between a header name and the colon in the error message that
2932 is given when verify = headers_syntax fails (if there are lots of them,
2933 the message gets confusing).
2935 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2936 names, as there are now some PTR records that contain slashes. This check
2937 is only to protect against broken name servers that fall over on strange
2938 characters, so the fact that it applies to all lookups doesn't matter.
2940 PH/36 Now that the new test suite is complete, we can remove some of the
2941 special code in Exim that was needed for the old test suite. For example,
2942 sorting DNS records because real resolvers return them in an arbitrary
2943 order. The new test suite's fake resolver always returns records in the
2946 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2947 bounces) except when queue_only is set, to avoid logging races between
2948 the different processes.
2950 PH/38 Panic-die if .include specifies a non-absolute path.
2952 PH/39 A tweak to the "H" retry rule from its user.
2954 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2955 a label. They prevented compilation on older perls.
2957 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2958 a warning to be raised on newish perls.
2960 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2961 on queue. Changes to match documented behaviour of showing count of
2962 messages matching specified criteria.
2964 PH/40 Changed the default ident timeout from 30s to 5s.
2966 PH/41 Added support for the use of login_cap features, on those BSD systems
2967 that have them, for controlling the resources used by pipe deliveries.
2969 PH/42 The content-scanning code uses fopen() to create files in which to put
2970 message data. Previously it was not paying any attention to the mode of
2971 the files. Exim runs with umask(0) because the rest of the code creates
2972 files with open(), and sets the required mode explicitly. Thus, these
2973 files were ending up world-writeable. This was not a big issue, because,
2974 being within the spool directory, they were not world-accessible. I have
2975 created a function called modefopen, which takes an additional mode
2976 argument. It sets umask(777), creates the file, chmods it to the required
2977 mode, then resets the umask. All the relevant calls to fopen() in the
2978 content scanning code have been changed to use this function.
2980 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2981 to 24 hours. This avoids potential overflow problems when processing G
2982 and H retry rules. I suspect nobody ever tinkers with this value.
2984 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2986 PH/45 When the plaintext authenticator is running as a client, the server's
2987 challenges are checked to ensure they are valid base64 strings. By
2988 default, the authentication attempt is cancelled if an invalid string is
2989 received. Setting client_ignore_invalid_base64 true ignores these errors.
2990 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2991 they are received. Thus, the responses can be made to depend on the
2992 challenges. If an invalid string is ignored, an empty string is placed in
2995 PH/46 Messages that are created by the autoreply transport now contains a
2996 References: header, in accordance with RFCs 2822 and 3834.
2998 PH/47 Added authenticated_sender_force to the smtp transport.
3000 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3002 PH/49 Installed latest patch from the Sieve maintainer.
3004 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3005 was also set, the appendfile transport was unnecessarily scanning a
3006 directory of message files (e.g. for maildir delivery) to find the count
3007 of files (along with the size), even though it did not need this
3008 information. It now does the scan only if it needs to find either the
3009 size of the count of files.
3011 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3013 PH/52 Two bugs concerned with error handling when the smtp transport is
3016 (i) Exim was not creating retry information for temporary errors given
3017 for individual recipients after the DATA command when the smtp transport
3018 was used in LMTP mode. This meant that they could be retried too
3019 frequently, and not timed out correctly.
3021 (ii) Exim was setting the flag that allows error details to be returned
3022 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3023 recipients that were returned after the DATA command.
3025 PH/53 This is related to PH/52, but is more general: for any failing address,
3026 when detailed error information was permitted to be returned to the
3027 sender, but the error was temporary, then after the final timeout, only
3028 "retry timeout exceeded" was returned. Now it returns the full error as
3029 well as "retry timeout exceeded".
3031 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3032 do this, and (what is worse) MTAs that accept it.
3034 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3035 will now be deprecated.
3037 PH/56 New os.c-cygwin from the Cygwin maintainer.
3039 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3040 formats (previously only available in exim formats via -bpr, -bpru,
3041 and -bpra. Now also available in native and exiqgrep formats)
3043 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3044 with very large, slow to parse queues
3046 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3048 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3050 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3051 responses to authentication challenges, though it was showing the
3052 challenges; (ii) I've removed the CR characters from the debug output for
3055 PH/58 Allow for the insertion of a newline as well as a space when a string
3056 is turned into more than one encoded-word during RFC 2047 encoding. The
3057 Sieve code now uses this.
3059 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3060 data_4xx, lost_connection, tls_required.
3062 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3063 message was being sent as an SMTP response.
3065 PH/61 Add -l and -k options to exicyclog.
3067 PH/62 When verifying, if an address was redirected to one new address, so that
3068 verification continued, and the new address failed or deferred after
3069 having set something in $address_data, the value of $address_data was not
3070 passed back to the ACL. This was different to the case when no
3071 redirection occurred. The value is now passed back in both cases.
3073 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3074 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3075 use login_cap.h, so on its own it isn't the distinguishing feature. The
3076 new name refers directly to the setclassresources() function.
3078 PH/65 Added configuration files for NetBSD3.
3080 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3082 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3083 is preferred over IPv4.
3085 PH/68 The bounce_return_message and bounce_return_body options were not being
3086 honoured for bounces generated during the reception of non-SMTP messages.
3087 In particular, this applied to messages rejected by the ACL. This bug has
3088 been fixed. However, if bounce_return_message is true and bounce_return_
3089 body is false, the headers that are returned for a non-SMTP message
3090 include only those that have been read before the error was detected.
3091 (In the case of an ACL rejection, they have all been read.)
3093 PH/69 The HTML version of the specification is now built in a directory called
3094 spec_html instead of spec.html, because the latter looks like a path with
3095 a MIME-type, and this confuses some software.
3097 PH/70 Catch two compiler warnings in sieve.c.
3099 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3100 function verify_get_ident() calls ip_connect() to connect a socket, but
3101 if the "connect()" function timed out, ip_connect() used to close the
3102 socket. However, verify_get_ident() also closes the socket later, and in
3103 between Exim writes to the log, which may get opened at this point. When
3104 the socket was closed in ip_connect(), the log could get the same file
3105 descriptor number as the socket. This naturally causes chaos. The fix is
3106 not to close the socket in ip_connect(); the socket should be closed by
3107 the function that creates it. There was only one place in the code where
3108 this was missing, in the iplookup router, which I don't think anybody now
3109 uses, but I've fixed it anyway.
3111 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3112 well as to direct DNS lookups. Otherwise the handling of names in host
3113 lists is inconsistent and therefore confusing.
3119 PH/01 Two changes to the default runtime configuration:
3121 (1) Move the checks for relay_from_hosts and authenticated clients from
3122 after to before the (commented out) DNS black list checks.
3124 (2) Add control=submission to the relay_from_hosts and authenticated
3125 clients checks, on the grounds that messages accepted by these
3126 statements are most likely to be submissions.
3128 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3130 (1) Generate an error if the third argument for the ${prvs expansion is
3133 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3136 (3) Reset the variables that are obtained from the first argument of
3137 ${prvscheck and used in the second argument before leaving the code,
3138 because their memory is reclaimed, so using them afterwards may do
3141 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3142 one (it's much easier than Tom thought :-).
3144 (5) Because of (4), we can now allow for the use of $prvscheck_result
3145 inside the third argument.
3147 PH/03 For some reason, the default setting of PATH when running a command from
3148 a pipe transport was just "/usr/bin". I have changed it to
3151 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3152 anything to be listed in the output from -bV.
3154 PH/05 When a filter generated an autoreply, the entire To: header line was
3155 quoted in the delivery log line, like this:
3157 => >A.N.Other <ano@some.domain> <original@ddress> ...
3159 This has been changed so that it extracts the operative address. There
3160 may be more than one such address. If so, they are comma-separated, like
3163 => >ano@some.domain,ona@other.domain <original@ddress> ...
3165 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3166 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3167 not being looked up in the rDNS to get a host name, Exim was showing the
3168 IP address twice in Received: lines, even though the IP addresses were
3169 identical. For example:
3171 Received: from [1.2.3.4] (helo=[1.2.3.4])
3173 However, if the real host name was known, it was omitting the HELO data
3174 if it matched the actual IP address. This has been tidied up so that it
3175 doesn't show the same IP address twice.
3177 PH/07 When both +timestamp and +memory debugging was on, the value given by
3178 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3179 called by the debug printing, thereby overwriting the timestamp buffer.
3180 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3182 PH/08 When the original message was included in an autoreply transport, it
3183 always said "this is a copy of the message, including all the headers",
3184 even if body_only or headers_only was set. It now gives an appropriate
3187 PH/09 Applied a patch from the Sieve maintainer which:
3189 o fixes some comments
3190 o adds the (disabled) notify extension core
3191 o adds some debug output for the result of if/elsif tests
3192 o points to the current vacation draft in the documentation
3193 and documents the missing references header update
3197 o fixes a bug in processing the envelope test (when testing
3198 multiple envelope elements, the last element determined the
3201 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3202 Electronic Mail") by including:
3204 Auto-submitted: auto-generated
3206 in the messages that it generates (bounce messages and others, such as
3207 warnings). In the case of bounce messages for non-SMTP messages, there was
3208 also a typo: it was using "Auto_submitted" (underscore instead of
3209 hyphen). Since every message generated by Exim is necessarily in response
3210 to another message, thes have all been changed to:
3212 Auto-Submitted: auto-replied
3214 in accordance with these statements in the RFC:
3216 The auto-replied keyword:
3218 - SHOULD be used on messages sent in direct response to another
3219 message by an automatic process,
3221 - MUST NOT be used on manually-generated messages,
3223 - MAY be used on Delivery Status Notifications (DSNs) and Message
3224 Disposition Notifications (MDNs),
3226 - MUST NOT be used on messages generated by automatic or periodic
3227 processes, except for messages which are automatic responses to
3230 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3231 to the default Received: header definition.
3233 PH/12 Added log selector acl_warn_skipped (default on).
3235 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3236 variables because (a) they are in the wrong storage pool and (b) even if
3237 they were copied, it wouldn't work properly because of the caching.
3239 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3240 checking when decoding. Apparently there are clients that generate
3241 overlong encoded strings. Why am I not surprised?
3243 PH/15 If the first argument of "${if match_address" was not empty, but did not
3244 contain an "@" character, Exim crashed. Now it writes a panic log message
3245 and treats the condition as false.
3247 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3249 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3250 extension "envelope-auth". The code is finished and in agreement with
3251 other implementations, but there is no documentation so far and in fact,
3252 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3253 not changing the active code.
3255 Print executed "if" and "elsif" statements when debugging is used. This
3256 helps a great deal to understand what a filter does.
3258 Document more things not specified clearly in RFC3028. I had all this
3259 sorted out, when out of a sudden new issues came to my mind. Oops."
3261 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3264 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3265 canonical form (as documented). However, after a host name lookup from
3266 the IP address, check_host() was doing a simple string comparison with
3267 addresses acquired from the DNS when checking that the found name did
3268 have the original IP as one of its addresses. Since any found IPv6
3269 addresses are likely to be in abbreviated form, the comparison could
3270 fail. Luckily, there already exists a function for doing the comparison
3271 by converting both addresses to binary, so now that is used instead of
3272 the text comparison.
3274 PH/20 There was another similar case to PH/19, when a complete host name was
3275 given in a host list; looking up its IP address could give an abbreviated
3276 form, whereas the current host's name might or might not be abbreviated.
3277 The same fix has been applied.
3283 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3284 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3287 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3288 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3290 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3292 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3293 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3294 Also, for "Auto-Submitted", treat anything other than "no" as
3295 non-personal, in accordance with RFC 3834. (Previously it treated
3296 anything starting "auto-" as non-personal.)
3298 TF/01 The control=submission/name=... option had a problem with syntax
3299 errors if the name included a slash character. The /name= option
3300 now slurps the rest of the string, so it can include any characters
3301 but it must come last in the list of options (after /sender_retain
3304 PH/05 Some modifications to the interface to the fake nameserver for the new
3312 TK/01 Added the "success_on_redirect" address verification option. See
3313 NewStuff for rationale and an example.
3315 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3317 PH/02 Patch to exigrep to allow it to work on syslog lines.
3319 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3320 fread() to skip over the body file's header line, because in Cygwin the
3321 header line is locked and is inaccessible.
3323 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3324 co-exist for some time) to make it clear that it is the Exim ID that is
3325 referenced, not the Message-ID: header line.
3327 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3328 string_format() function, because snprintf() does not exist on all
3331 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3332 ${stat: expansion item.
3334 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3335 protocol synchronization error", to keep the pedants happy.
3337 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3338 well as for IRIX systems, when gcc is being used. See the host.c source
3341 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3343 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3346 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3347 transport and to the smtp transport in LMTP mode.
3349 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3351 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3352 run a filter in a subprocess. This could lead to confusion in subsequent
3353 lookups in the parent process. There should also be a search_tidyup() at
3354 the end of the subprocess.
3356 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3357 only if the host matched helo_try_verify_hosts, which caused the
3358 verification to occur when the EHLO/HELO command was issued. The ACL just
3359 tested the remembered result. Now, if a previous verification attempt has
3360 not happened, "verify = helo" does it there and then.
3362 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3364 TK/03 Fix log output including CR from clamd.
3366 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3367 exist provoked a memory error which could cause a segfault.
3369 PH/15 Installed PCRE 6.2
3371 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3373 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3374 of the problem. Specifically, suggested +O2 rather than +O1 for the
3377 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3379 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3380 record, the port was not correctly logged when the outgoing_port log
3381 selector was set (it logged the transort's default port).
3383 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3384 fallback_hosts, and "hosts" in the smtp transport.
3386 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3387 host errors such as "Connection refused".
3389 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3390 authentication with radiusclient 0.4.9:
3392 - Error returned from rc_read_config was caught wrongly
3393 - Username/password not passed on to radius server due to wrong length.
3395 The presumption is that some radiusclient API changes for 4.51/PH/17
3396 were not taken care of correctly. The code is still untested by me (my
3397 Linux distribution still has 0.3.2 of radiusclient), but it was
3398 contributed by a Radius user.
3400 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3401 expanding the "port" option of the smtp transport.
3403 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3404 while reading a MIME header. Thanks to Tom Hughes for a patch.
3406 PH/24 Include config.h inside local_scan.h so that configuration settings are
3409 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3410 that in an ACL for RCPT (for example), you can examine exactly what was
3413 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3414 commands, but it was not correctly comparing the address with the actual
3415 client host address. Thus, it would show the EHLO address in Received:
3416 header lines when this was not necessary.
3418 PH/27 Added the % operator to ${eval:}.
3420 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3421 it should be ignoring failures (because with -C, for example, it has lost
3422 privilege). It wasn't ignoring creation failures other than "already
3425 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3428 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3429 because an input line was too long, either on its own, or by virtue of
3430 too many continuations, the temporary file was not being removed, and the
3431 return code was incorrect.
3433 PH/31 Missing "BOOL" in function definition in filtertest.c.
3435 PH/32 Applied Sieve patches from the maintainer.
3437 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3439 PH/33 Added "verify = not_blind".
3441 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3442 Local/Makefile (with some defaults set). These are used in built scripts
3443 such as exicyclog, but they have never been used in the exim_install
3444 script (though there are many overriding facilities there). I have
3445 arranged that the exim_install script now takes note of these two
3448 PH/35 Installed configuration files for Dragonfly.
3450 PH/36 When a locally submitted message by a trusted user did not contain a
3451 From: header, and the sender address was obtained from -f or from an SMTP
3452 MAIL command, and the trusted user did not use -F to supply a sender
3453 name, $originator_name was incorrectly used when constructing a From:
3454 header. Furthermore, $originator_name was used for submission mode
3455 messages from external hosts without From: headers in a similar way,
3456 which is clearly wrong.
3458 PH/37 Added control=suppress_local_fixups.
3460 PH/38 When log_selector = +received_sender was set, and the addition of the
3461 sender made the log line's construction buffer exactly full, or one byte
3462 less than full, an overflow happened when the terminating "\n" was
3465 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3466 when the result of a list match is failure because a DNS lookup failed.
3468 PH/40 RM_COMMAND is now used in the building process.
3470 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3471 the "build-* directories that it finds.
3473 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3474 address in a domain literal was a prefix of an interface address.
3476 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3477 when verifying a sender address, unless rewrite_headers is false.
3479 PH/44 Wrote a long comment about why errors_to addresses are verified as
3480 recipients, not senders.
3482 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3483 the ratelimit ACL was added.
3485 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3487 PH/46 Added extra information about PostgreSQL errors to the error string.
3489 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3490 avoiding the need to install special zones in a real server. This is
3491 backwards compatible; if it can't find the fake resolver, it drops back.
3492 Thus, both old and new test suites can be run.
3494 TF/02 Added util/ratelimit.pl
3496 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3497 clock is set back in time.
3499 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3500 Candler <B.Candler@pobox.com>.
3502 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3503 OK for addresses that are the result of redirections.
3505 PH/48 A number of further additions for the benefit of the new test suite,
3506 including a fake gethostbyname() that interfaces to the fake DNS resolver
3509 TF/06 The fix for widen_domains has also been applied to qualify_single and
3510 search_parents which are the other dnslookup options that can cause
3513 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3516 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3518 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3519 service" if it encountered an MX record with an empty target hostname.
3520 The message is now "an MX or SRV record indicated no SMTP service".
3522 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3523 if the DNS of the sending site is misconfigured. This is quite a
3524 common situation. This change restores the behaviour of treating a
3525 helo verification defer as a failure.
3527 PH/51 If self=fail was set on a router, the bounce message did not include the
3528 actual error message.
3534 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3536 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3537 command itself ended in error, the underlying message about the transport
3538 filter timeout was being overwritten with the pipe command error. Now the
3539 underlying error message should be appended to the second error message.
3541 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3543 PH/02 Reduce the amount of output that "make" produces by default. Full output
3544 can still be requested.
3546 PH/03 The warning log line about a condition test deferring for a "warn" verb
3547 was being output only once per connection, rather than after each
3548 occurrence (because it was using the same function as for successful
3549 "warn" verbs). This seems wrong, so I have changed it.
3551 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3552 it should not have, which might have caused a crash in the right
3553 circumstances, but probably never did.
3555 PH/04 Installed a modified version of Tony Finch's patch to make submission
3556 mode fix the return path as well as the Sender: header line, and to
3557 add a /name= option so that you can make the user's friendly name appear
3560 TF/03 Added the control = fakedefer ACL modifier.
3562 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3563 Mark Lowes for thorough testing.
3565 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3567 TK/03 Merged latest SRS patch from Miles Wilton.
3569 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3570 with the definition in sysexits.h (which is #included earlier).
3571 Fortunately, Exim does not actually use EX_OK. The code used to try to
3572 preserve the sysexits.h value, by assuming that macro definitions were
3573 scanned for macro replacements. I have been disabused of this notion,
3574 so now the code just undefines EX_OK before #including unistd.h.
3576 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3577 in the smtp transport. When a block could not be written in a single
3578 write() function, the timeout was being re-applied to each part-write.
3579 This seems wrong - if the receiver was accepting one byte at a time it
3580 would take for ever. The timeout is now adjusted when this happens. It
3581 doesn't have to be particularly precise.
3583 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3584 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3586 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3587 without a domain if the check to <postmaster@domain> fails.
3589 SC/01 Eximstats: added -xls and the ability to specify output files
3590 (patch written by Frank Heydlauf).
3592 SC/02 Eximstats: use FileHandles for outputting results.
3594 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3596 SC/04 Eximstats: fixed display of large numbers with -nvr option
3598 SC/05 Eximstats: fixed merging of reports with empty tables.
3600 SC/06 Eximstats: added the -include_original_destination flag
3602 SC/07 Eximstats: removed tabs and trailing whitespace.
3604 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3606 TK/06 MBOX spool code: Add real "From " MBOX separator line
3607 so the .eml file is really in mbox format (even though
3608 most programs do not really care). Patch from Alex Miller.
3610 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3611 The latter is generated from $received_to and is only set if the
3612 message has one envelope recipient. SA can use these headers,
3613 obviously out-of-the-box. Patch from Alex Miller.
3615 PH/08 The ${def test on a variable was returning false if the variable's
3616 value was "0", contrary to what the specification has always said!
3617 The result should be true unless the variable is empty.
3619 PH/09 The syntax error of a character other than { following "${if
3620 def:variable_name" (after optional whitespace) was not being diagnosed.
3621 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3622 accidental colon was present, for example, could give incorrect results.
3624 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3625 result is used (not including appendfile, where other changes are about
3628 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3629 This involved changing a lot of size variables from int to off_t. It
3630 should work with maildirs and everything.
3632 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3633 spamd dying while we are connected to it.
3635 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3638 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3639 function for BDB 4.3.
3641 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3643 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3644 the sources for PCRE have been re-arranged and more files are now
3647 PH/15 The code I had for printing potentially long long variables in PH/11
3648 above was not the best (it lost precision). The length of off_t variables
3649 is now inspected at build time, and an appropriate printing format (%ld
3650 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3651 to be "long long int" or "long int". This is needed for the internal
3652 formatting function string_vformat().
3654 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3655 the configuration file to be ":syslog", then the script "guesses" where
3656 the logs files are, rather than using the compiled in default. In our
3657 case the guess is not the same as the compiled default, so the script
3658 suddenly stopped working when I started to use syslog. The patch checks
3659 to see if log_file_path is "". If so, it attempts to read it from exim
3660 with no configuration file to get the compiled in version, before it
3661 falls back to the previous guessing code."
3663 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3664 implementing BATV in an Exim configuration. See NewStuff for the gory
3667 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3668 Makefile that are specific to HP-UX.
3670 PH/18 If the "use_postmaster" option was set for a recipient callout together
3671 with the "random" option, the postmaster address was used as the MAIL
3672 FROM address for the random test, but not for the subsequent recipient
3673 test. It is now used for both.
3675 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3676 patch removes a few documentation additions to RFC 3028, because the
3677 latest draft now contains them. It adds the new en;ascii-case comparator
3678 and a new error check for 8bit text in MIME parts. Comparator and
3679 require names are now matched exactly. I enabled the subaddress
3680 extension, but it is not well tested yet (read: it works for me)."
3682 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3683 rework some of the code of TK/09 above to avoid the hardwired use of
3684 "%lld" and "long long". Replaced the call to snprintf() with a call to
3687 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3688 records point to non-existent hosts", "retry timeout exceeded", and
3689 "retry time not reached for any host after a long failure period".
3691 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3692 experimental DomainKeys support:
3694 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3695 (2) On an error such as an illegally used "control", the wrong name for
3696 the control was given.
3698 These problems did NOT occur unless DomainKeys support was compiled.
3700 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3702 PH/24 Added ${if match_ip condition.
3704 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3705 databases so that it will be absolutely obvious if a crash occurs in the
3706 DB library. This is a regular occurrence (often caused by mis-matched
3709 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3710 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3711 picked up on a user's system that detects such things. There doesn't seem
3712 to be a gcc warning option for this - only an attribute that has to be
3713 put on the function's prototype. It seems that in Fedora Core 4 they have
3714 set this on a number of new functions. No doubt there will be more in due
3717 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3718 specify a transport. However, if an address that was verified by such a
3719 router was the subject of a callout, Exim crashed because it tried to
3720 read the rcpt_include_affixes from the non-existent transport. Now it
3721 just assumes that the setting of that option is false. This bug was
3722 introduced by 4.51/PH/31.
3724 PH/28 Changed -d+all to exclude +memory, because that information is very
3725 rarely of interest, but it makes the output a lot bigger. People tend to
3726 do -d+all out of habit.
3728 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3729 code in os-type was giving problems when libc.so lives in lib64, like on
3732 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3733 aren't the modern standard, and it seems that some systems' include files
3734 don't always have them. Exim was already checking for some of the newer
3735 ones like T_AAAA, and defining it itself. I've added checks for all the
3736 record types that Exim uses.
3738 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3739 not automatically generating a new one, as it is supposed to. This
3740 prevented TLS from working. If the file did exist, but contained invalid
3741 data, a new version was generated, as expected. It was only the case of a
3742 non-existent file that was broken.
3744 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3745 with a change in libdomainkeys > 0.64.
3747 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3748 from DNS. If the selector record carries the flag, it now has
3749 precedence over the domain-wide flag.
3751 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3753 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3754 the use of a port name that isn't defined in /etc/services) occurred, the
3755 message was deferred as in a normal delivery, and thus remained on the
3756 spool, instead of being failed because of the mua_wrapper setting. This
3757 is now fixed, and I tidied up some of the mua_wrapper messages at the
3760 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3761 the messages in a hash of arrays rather than using individual hashes.
3762 This is a bit cleaner and results in dramatic memory savings, albeit
3763 at a slight CPU cost.
3765 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3766 as requested by Marc Sherman.
3768 SC/10 Eximstats: added histograms for user specified patterns as requested
3771 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3773 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3774 fopen() in the content-scanning modules that did not already have it.
3780 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3781 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3783 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3785 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3787 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3788 to test Sieve filters that use "vacation".
3790 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3791 that changes the way the GnuTLS parameters are stored in the cache file.
3792 The new format can be generated externally. For backward compatibility,
3793 if the data in the cache doesn't make sense, Exim assumes it has read an
3794 old-format file, and it generates new data and writes a new file. This
3795 means that you can't go back to an older release without removing the
3798 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3799 work if there are any delivery delays because "one_time" forces the
3800 parent to be marked "delivered", so its unseen clone is never tried
3801 again. For this reason, Exim now forbids the simultaneous setting of
3804 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3805 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3806 entry if you want to know the details. The fix, however, neglected to
3807 consider the case when local delivery batching is involved. The test for
3808 "previously delivered" was not happening when checking to see if an
3809 address could be batched with a previous (undelivered) one; under
3810 certain circumstances this could lead to multiple deliveries to the same
3813 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3814 in its include files, and this causes problems building Exim.
3816 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3817 header_syntax) but Exim was just ignoring anything given after a slash.
3818 In particular, this caused confusion with an attempt to use "verify =
3819 reverse_host_lookup/defer_ok". An error is now given when options are
3820 supplied for verify items that do not have them. (Maybe reverse_host_
3821 lookup should have a defer_ok option, but that's a different point.)
3823 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3824 defined by RFC 821) to 2048, because there were problems with some AUTH
3825 commands, and RFC 1869 says the size should be increased for extended
3826 SMTP commands that take arguments.
3828 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3831 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3832 "unknown" error; now it says that the functionality isn't in the binary.
3834 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3835 an address' error message when a string expansion fails (syntax or
3836 whatever). Otherwise the password may appear in the log. Following change
3837 PH/42 below, there is no longer a chance of it appearing in a bounce
3840 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3842 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3843 of its addresses was ever tried. (Bugzilla bug #2.)
3845 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3846 the result incorrectly in the debug output. (It correctly added a newline
3847 to what was transported.)
3849 TF/01 Added $received_time.
3851 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3852 commented out examples of how to interface to a virus scanner and to
3853 SpamAssassin. Also added commented examples of av_scanner and
3854 spamd_address settings.
3856 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3857 and controls are allowed in which ACLs. There were a couple of minor
3858 errors. Some of the entries in the conditions table (which is a table of
3859 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3860 negation of where the condition IS allowed.
3862 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3864 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3865 header file does not have a version number, so I've had to invent a new
3866 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3867 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3868 radiusclient), but it was contributed by a Radius user.
3870 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3871 files or directories, for OpenSSL.
3873 PH/19 When an Exim process that is running as root has to create an Exim log
3874 file, it does so in a subprocess that runs as exim:exim so as to get the
3875 ownership right at creation (otherwise, other Exim processes might see
3876 the file with the wrong ownership). There was no test for failure of this
3877 fork() call, which would lead to the process getting stuck as it waited
3878 for a non-existent subprocess. Forks do occasionally fail when resources
3879 run out. I reviewed all the other calls to fork(); they all seem to check
3882 PH/20 When checking for unexpected SMTP input at connect time (before writing
3883 the banner), Exim was not dealing correctly with a non-positive return
3884 from the read() function. If the client had disconnected by this time,
3885 the result was a log entry for a synchronization error with an empty
3886 string after "input=" when read() returned zero. If read() returned -1
3887 (an event I could not check), uninitialized data bytes were printed.
3888 There were reports of junk text (parts of files, etc) appearing after
3891 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3893 PH/22 Added support for macro redefinition, and (re)definition in between
3894 driver and ACL definitions.
3896 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3897 forgetting to use the resulting value; it was using the unexpanded value.
3899 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3900 hadn't been configured. The fix is from Juergen Kreileder, who
3901 understands it better than I do:
3903 "Here's what I see happening with three configured cyrus_sasl
3904 authenticators configured (plain, login, cram-md5):
3906 On startup auth_cyrus_sasl_init() gets called for each of these.
3907 This means three calls to sasl_listmech() without a specified mech_list.
3908 => SASL tests which mechs of all available mechs actually work
3909 => three warnings about OTP not working
3910 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3912 With the patch, sasl_listmech() also gets called three times. But now
3913 SASL's mech_list option is set to the server_mech specified in the the
3914 authenticator. Or in other words, the answer from sasl_listmech()
3915 gets limited to just the mech you're testing for (which is different
3917 => the return list contains just 'plain' or 'login', 'cram-md5' or
3918 nothing depending on the value of ob->server_mech.
3920 I've just tested the patch: Authentication still works fine,
3921 unavailable mechs specified in the exim configuration are still
3922 caught, and the auth.log warnings about OTP are gone."
3924 PH/25 When debugging is enabled, the contents of the command line are added
3925 to the debugging output, even when log_selector=+arguments is not
3928 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3929 answer is "GNU", and only if the return is "GNU/something" is the answer
3932 PH/27 $acl_verify_message is now set immediately after the failure of a
3933 verification in an ACL, and so is available in subsequent modifiers. In
3934 particular, the message can be preserved by coding like this:
3936 warn !verify = sender
3937 set acl_m0 = $acl_verify_message
3939 Previously, $acl_verify_message was set only while expanding "message"
3940 and "log_message" when a very denied access.
3942 PH/28 Modified OS/os.c-Linux with
3944 -#ifndef OS_LOAD_AVERAGE
3945 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3947 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3948 nomenclature these days.)
3950 PH/29 Installed patch from the Sieve maintainer that adds the options
3951 sieve_useraddress and sieve_subaddress to the redirect router.
3953 PH/30 In these circumstances:
3954 . Two addresses routed to the same list of hosts;
3955 . First host does not offer TLS;
3956 . First host accepts first address;
3957 . First host gives temporary error to second address;
3958 . Second host offers TLS and a TLS session is established;
3959 . Second host accepts second address.
3960 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3961 and peerdn, if requested) that were in fact used only for the second
3964 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3965 attention to any local part prefix or suffix that was matched by the
3966 router that accepted the address. It now behaves in the same way as it
3967 does for delivery: the affixes are removed from the local part unless
3968 rcpt_include_affixes is set on the transport.
3970 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3971 timeout during the DATA phase of an incoming message.
3973 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3974 applied a patch sanctioned by the Sieve maintainer.
3976 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3977 the uid or gid is negative. A case of a negative gid caused this to be
3978 noticed. The fix allows for either to be negative.
3980 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3981 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3984 PH/36 The change PH/12 above was broken. Fixed it.
3986 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3987 the grounds that routing the same address twice would always produce the
3988 same answer. This might have been true once, but it is certainly no
3989 longer true now. Routing a child address may depend on the previous
3990 routing that produced that child. Some complicated redirection strategies
3991 went wrong when messages had multiple recipients, and made Exim's
3992 behaviour dependent on the order in which the addresses were given.
3994 I have moved the duplicate checking until after the routing is complete.
3995 Exim scans the addresses that are assigned to local and remote
3996 transports, and removes any duplicates. This means that more work will be
3997 done, as duplicates will always all be routed, but duplicates are
3998 presumably rare, so I don't expect this is of any significance.
4000 For deliveries to pipes, files, and autoreplies, the duplicate checking
4001 still happens during the routing process, since they are not going to be
4004 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4005 It corrects a timeout issue with spamd. This is Ian's comment: "The
4006 background is that sometimes spamd either never reads data from a
4007 connection it has accepted, or it never writes response data. The exiscan
4008 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4009 blindly assumes that writes won't block so it may never time out."
4011 PH/39 Allow G after quota size as well as K and M.
4013 PH/40 The value set for $authenticated_id in an authenticator may not contain
4014 binary zeroes or newlines because the value is written to log lines and
4015 to spool files. There was no check on this. Now the value is run through
4016 the string_printing() function so that such characters are converted to
4017 printable escape sequences.
4019 PH/41 $message_linecount is a new variable that contains the total number of
4020 lines in the message. Compare $body_linecount, which is the count for the
4023 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4024 bounce and delay warning messages, except in certain special cases, which
4027 (a) An SMTP error message from a remote host;
4028 (b) A message specified in a :fail: redirection;
4029 (c) A message specified in a "fail" command in a system filter;
4030 (d) A message specified in a FAIL return from the queryprogram router;
4031 (e) A message specified by the cannot_route_message router option.
4033 In these cases only, Exim does include the error details in bounce and
4034 warning messages. There are also a few cases where bland messages such
4035 as "unrouteable address" or "local delivery error" are given.
4037 PH/43 $value is now also set for the "else" part of a ${run expansion.
4039 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4040 being worked on, but at least Exim now implements the latest version to
4043 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4044 process to complete was treated as a delivery failure, a timeout while
4045 writing the message to the pipe was logged, but erroneously treated as a
4046 successful delivery. Such timeouts include transport filter timeouts. For
4047 consistency with the overall process timeout, these timeouts are now
4048 treated as errors, giving rise to delivery failures by default. However,
4049 there is now a new Boolean option for the pipe transport called
4050 timeout_defer, which, if set TRUE, converts the failures into defers for
4051 both kinds of timeout. A transport filter timeout is now identified in
4054 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4055 systems where "make" and "gmake" are different, calling "gmake" at top
4056 level broke things. I've arranged for the value of $(MAKE) to be passed
4057 from the Makefile to this script so that it can call the same version of
4061 A note about Exim versions 4.44 and 4.50
4062 ----------------------------------------
4064 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4065 changes of various kinds. As a consequence, a big documentation update was
4066 needed. This delayed the release for rather longer than seemed good, especially
4067 in the light of a couple of (minor) security issues. Therefore, the changes
4068 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4069 release. So 4.44 and 4.50 are in effect two different branches that both start
4072 I have left the 4.50 change log unchanged; it contains all the changes since
4073 4.43. The change log for 4.44 is below; many of its items are identical to
4074 those for 4.50. This seems to be the most sensible way to preserve the
4075 historical information.
4081 1. Minor wording change to the doc/README.SIEVE file.
4083 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4084 computation of the current number of files was incorrect.
4086 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4087 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4090 4. Give more explanation in the error message when the command for a transport
4091 filter fails to execute.
4093 5. There are several places where Exim runs a non-Exim command in a
4094 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4095 was being done only for the command run by the queryprogram router. It is
4096 now done for all such subprocesses. The other cases are: ${run, transport
4097 filters, and the commands run by the lmtp and pipe transports.
4099 6. Added CONFIGURE_GROUP build-time option.
4101 7. Some older OS have a limit of 256 on the maximum number of file
4102 descriptors. Exim was using setrlimit() to set 1000 as a large value
4103 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4104 systems. I've change it so that if it can't get 1000, it tries for 256.
4106 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4107 was an oversight, and furthermore, ever since the addition of extra
4108 controls (e.g. 4.43/32), the checks on when to allow different forms of
4109 "control" were broken. There should now be diagnostics for all cases when a
4110 control that does not make sense is encountered.
4112 9. Added the /retain_sender option to "control=submission".
4114 10. $recipients is now available in the predata ACL (oversight).
4116 11. Tidy the search cache before the fork to do a delivery from a message
4117 received from the command line. Otherwise the child will trigger a lookup
4118 failure and thereby defer the delivery if it tries to use (for example) a
4119 cached ldap connection that the parent has called unbind on.
4121 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4122 of $address_data from the recipient verification was clobbered by the
4123 sender verification.
4125 13. The value of address_data from a sender verification is now available in
4126 $sender_address_data in subsequent conditions in the ACL statement.
4128 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4130 15. Added a new option "connect=<time>" to callout options, to set a different
4133 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4134 was its contents. (It was OK if the option was not defined at all.)
4136 17. A "Completed" log line is now written for messages that are removed from
4137 the spool by the -Mrm option.
4139 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4140 information about exactly what failed.
4142 19. Added -dd to debug only the daemon process.
4144 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4145 handles timeouts, both on the server side and network timeouts. Renamed the
4146 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4148 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4149 It is now set to "smtps".
4151 22. $host_address is now set to the target address during the checking of
4152 ignore_target_hosts.
4154 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4155 being passed; this would have caused $sender_host_name to have been used if
4156 matching the list had actually called for a host name (not very likely,
4157 since this list is usually IP addresses). A host name is now passed as
4160 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4161 code that specifies a non-restarting handler (typically sigaction() in
4162 modern systems) in an attempt to fix a rare and obscure crash bug.
4164 25. Narrowed the window for a race in the daemon that could cause it to ignore
4165 SIGCHLD signals. This is not a major problem, because they are used only to
4166 wake it up if nothing else does.
4168 26. A malformed maildirsize file could cause Exim to calculate negative values
4169 for the mailbox size or file count. Odd effects could occur as a result.
4170 The maildirsize information is now recalculated if the size or filecount
4173 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4174 support for a long time. Removed HAVE_SYS_VFS_H.
4176 28. Installed the latest version of exipick from John Jetmore.
4178 29. In an address list, if the pattern was not a regular expression, an empty
4179 subject address (from a bounce message) matched only if the pattern was an
4180 empty string. Non-empty patterns were not even tested. This was the wrong
4181 because it is perfectly reasonable to use an empty address as part of a
4182 database query. An empty address is now tested by patterns that are
4183 lookups. However, all the other forms of pattern expect the subject to
4184 contain a local part and a domain, and therefore, for them, an empty
4185 address still always fails if the pattern is not itself empty.
4187 30. Exim went into a mad DNS loop when attempting to do a callout where the
4188 host was specified on an smtp transport, and looking it up yielded more
4189 than one IP address.
4191 31. Re-factored the code for checking spool and log partition space into a
4192 function that finds that data and another that does the check. The former
4193 is then used to implement four new variables: $spool_space, $log_space,
4194 $spool_inodes, and $log_inodes.
4196 32. The RFC2047 encoding function was originally intended for short strings
4197 such as real names; it was not keeping to the 75-character limit for
4198 encoded words that the RFC imposes. It now respects the limit, and
4199 generates multiple encoded words if necessary. To be on the safe side, I
4200 have increased the buffer size for the ${rfc2047: expansion operator from
4203 33. It is now permitted to omit both strings after an "if" condition; if the
4204 condition is true, the result is "true". As before, when the second string
4205 is omitted, a false condition yields an empty string. This makes it less
4206 cumbersome to write custom ACL and router conditions.
4208 34. Failure to deliver a bounce message always caused it to be frozen, even if
4209 there was an errors_to setting on the router. The errors_to setting is now
4212 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4213 canonical form (fully expanded) before being placed in
4214 $sender_host_address.
4216 36. The table in the code that translates DNS record types into text (T_A to
4217 "A" for instance) was missing entries for NS and CNAME. It is just possible
4218 that this could have caused confusion if both these types were looked up
4219 for the same domain, because the text type is used as part of Exim's
4220 per-process caching. But the chance of anyone hitting this buglet seems
4223 37. The dnsdb lookup has been extended in a number of ways.
4225 (1) There is a new type, "zns", which walks up the domain tree until it
4226 finds some nameserver records. It should be used with care.
4228 (2) There is a new type, "mxh", which is like "mx" except that it returns
4229 just the host names, not the priorities.
4231 (3) It is now possible to give a list of domains (or IP addresses) to be
4232 looked up. The behaviour when one of the lookups defers can be
4233 controlled by a keyword.
4235 (4) It is now possible to specify the separator character for use when
4236 multiple records are returned.
4238 38. The dnslists ACL condition has been extended: it is now possible to supply
4239 a list of IP addresses and/or domains to be looked up in a particular DNS
4242 39. Added log_selector=+queue_time_overall.
4244 40. When running the queue in the test harness, wait just a tad after forking a
4245 delivery process, to get repeatability of debugging output.
4247 41. Include certificate and key file names in error message when GnuTLS fails
4248 to set them up, because the GnuTLS error message doesn't include the name
4249 of the failing file when there is a problem reading it.
4251 42. Allow both -bf and -bF in the same test run.
4253 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4255 44. The "Exiscan patch" is now merged into the mainline Exim source.
4257 45. Sometimes the final signoff response after QUIT could fail to get
4258 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4259 < 0 before doing a fflush(). This bug looks as though it goes back to the
4260 introduction of TLS in release 3.20, but "sometimes" must have been rare
4261 because the tests only now provoked it.
4263 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4264 (this can affect the format of dates).
4266 47. exim_tidydb, when checking for the continued existence of a message for
4267 which it has found a message-specific retry record, was not finding
4268 messages that were in split spool directories. Consequently, it was
4269 deleting retry records that should have stayed in existence.
4271 48. Steve fixed some bugs in eximstats.
4273 49. The SPA authentication driver was not abandoning authentication and moving
4274 on to the next authenticator when an expansion was forced to fail,
4275 contradicting the general specification for all authenticators. Instead it
4276 was generating a temporary error. It now behaves as specified.
4278 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4279 (the order specifies the preference for clients). The order is now AES256,
4280 AES128, 3DES, ARCFOUR128.
4282 51. Small patch to Sieve code - explicitly set From: when generating an
4285 52. Exim crashed if a remote delivery caused a very long error message to be
4286 recorded - for instance if somebody sent an entire SpamAssassin report back
4287 as a large number of 550 error lines. This bug was coincidentally fixed by
4288 increasing the size of one of Exim's internal buffers (big_buffer) that
4289 happened as part of the Exiscan merge. However, to be on the safe side, I
4290 have made the code more robust (and fixed the comments that describe what
4293 53. Now that there can be additional text after "Completed" in log lines (if
4294 the queue_time_overall log selector is set), a one-byte patch to exigrep
4295 was needed to allow it to recognize "Completed" as not the last thing in
4298 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4299 patch that reportedly fixes this has been added. I am not expert enough to
4300 create a test for it. This is what the patch creator wrote:
4302 "I found a little strange behaviour of ldap code when working with
4303 Windows 2003 AD Domain, where users was placed in more than one
4304 Organization Units. When I tried to give exim partial DN, the exit code
4305 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4306 But simultaneously result of request was absolutely normal ldap result,
4307 so I produce this patch..."
4309 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4310 REFERENCE, so I have modified the code to exclude the patch when that macro
4313 55. Some experimental protocols are using DNS PTR records for new purposes. The
4314 keys for these records are domain names, not reversed IP addresses. The
4315 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4316 leaves it alone. Component reversal etc. now happens only for IP addresses.
4319 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4321 57. Double the size of the debug message buffer (to 2048) so that more of very
4322 long debug lines gets shown.
4324 58. The exicyclog utility now does better if the number of log files to keep
4325 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4327 59. Two changes related to the smtp_active_hostname option:
4329 (1) $smtp_active_hostname is now available as a variable.
4330 (2) The default for smtp_banner uses $smtp_active_hostname instead
4331 of $primary_hostname.
4333 60. The host_aton() function is supposed to be passed a string that is known
4334 to be a valid IP address. However, in the case of IPv6 addresses, it was
4335 not checking this. This is a hostage to fortune. Exim now panics and dies
4336 if the condition is not met. A case was found where this could be provoked
4337 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4338 components; fortuitously, this particular loophole had already been fixed
4339 by change 4.50/55 above.
4341 If there are any other similar loopholes, the new check in host_aton()
4342 itself should stop them being exploited. The report I received stated that
4343 data on the command line could provoke the exploit when Exim was running as
4344 exim, but did not say which command line option was involved. All I could
4345 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4346 running as the user.
4349 61. There was a buffer overflow vulnerability in the SPA authentication code
4350 (which came originally from the Samba project). I have added a test to the
4351 spa_base64_to_bits() function which I hope fixes it.
4354 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4355 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4357 63. The daemon start-up calls getloadavg() while still root for those OS that
4358 need the first call to be done as root, but it missed one case: when
4359 deliver_queue_load_max is set with deliver_drop_privilege. This is
4360 necessary for the benefit of the queue runner, because there is no re-exec
4361 when deliver_drop_privilege is set.
4363 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4364 This has been fixed.
4366 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4367 was in use, was not putting the data itself into the right store pool;
4368 consequently, it could be overwritten for a subsequent message in the same
4369 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4372 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4374 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4375 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4376 the function were treating the return as a boolean value, which happened to
4377 work because 0=false and not-0=true, but is not correct code.
4379 68. The host_aton() function was not handling scoped IPv6 addresses (those
4380 with, for example, "%eth0" on the end) correctly.
4382 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4383 negated items (that is, ~something) in unsigned ints. Some compilers
4384 apparently mutter when there is no cast.
4386 70. If an address verification called from an ACL failed, and did not produce a
4387 user-specific message (i.e. there was only a "system" message), nothing was
4388 put in $acl_verify_message. In this situation, it now puts the system
4391 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4392 session; change 4.31/43 added the unwanted input to the log line - except
4393 that it did not do this in the start of session case. It now does.
4395 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4396 This is wrong and can cause the other end to generate a synchronization
4397 error if it is another Exim or anything else that does the synchronization
4398 check. A QUIT command is no longer sent after a timeout.
4400 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4401 during host lookups.
4403 74. The defer_ok option of callout verification was not working if it was used
4404 when verifying addresses in header lines, that is, for this case:
4406 verify = header_sender/callout=defer_ok
4408 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4409 those file descriptors could be used for SMTP connections. If anything
4410 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4411 could be sent to the SMTP client, causing chaos. The daemon now opens
4412 stdin, stdout, and stderr to /dev/null when it puts itself into the
4415 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4416 log by default. The user can override this with suitable Perl magic.
4418 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4419 the log message when discard triggers, was not working for the DATA ACL or
4420 for the non-SMTP ACL.
4422 78. Error message wording change in sieve.c.
4424 79. If smtp_accept_max_per_host was set, the number of connections could be
4425 restricted to fewer than expected, because the daemon was trying to set up
4426 a new connection before checking whether the processes handling previous
4427 connections had finished. The check for completed processes is now done
4428 earlier. On busy systems, this bug wouldn't be noticed because something
4429 else would have woken the daemon, and it would have reaped the completed
4432 80. If a message was submitted locally by a user whose login name contained one
4433 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4434 It caused a spool format error. I have fixed the spool reading code. A
4435 related problem was that the "from" clause in the Received: line became
4436 illegal because of the space(s). It is now covered by ${quote_local_part.
4438 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4441 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4443 83. Patch from Sieve maintainer to fix unterminated string problem in
4444 "vacation" handling.
4446 84. Some minor changes to the Linux configuration files to help with other
4447 OS variants using glibc.
4449 85. One more patch for Sieve to update vacation handling to latest spec.
4452 ----------------------------------------------------
4453 See the note above about the 4.44 and 4.50 releases.
4454 ----------------------------------------------------
4460 1. Change 4.43/35 introduced a bug that caused file counts to be
4461 incorrectly computed when quota_filecount was set in an appendfile
4464 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4465 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4468 3. Give more explanation in the error message when the command for a transport
4469 filter fails to execute.
4471 4. There are several places where Exim runs a non-Exim command in a
4472 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4473 was being done only for the command run by the queryprogram router. It is
4474 now done for all such subprocesses. The other cases are: ${run, transport
4475 filters, and the commands run by the lmtp and pipe transports.
4477 5. Some older OS have a limit of 256 on the maximum number of file
4478 descriptors. Exim was using setrlimit() to set 1000 as a large value
4479 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4480 systems. I've change it so that if it can't get 1000, it tries for 256.
4482 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4483 was an oversight, and furthermore, ever since the addition of extra
4484 controls (e.g. 4.43/32), the checks on when to allow different forms of
4485 "control" were broken. There should now be diagnostics for all cases when a
4486 control that does not make sense is encountered.
4488 7. $recipients is now available in the predata ACL (oversight).
4490 8. Tidy the search cache before the fork to do a delivery from a message
4491 received from the command line. Otherwise the child will trigger a lookup
4492 failure and thereby defer the delivery if it tries to use (for example) a
4493 cached ldap connection that the parent has called unbind on.
4495 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4496 of $address_data from the recipient verification was clobbered by the
4497 sender verification.
4499 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4500 was its contents. (It was OK if the option was not defined at all.)
4502 11. A "Completed" log line is now written for messages that are removed from
4503 the spool by the -Mrm option.
4505 12. $host_address is now set to the target address during the checking of
4506 ignore_target_hosts.
4508 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4509 being passed; this would have caused $sender_host_name to have been used if
4510 matching the list had actually called for a host name (not very likely,
4511 since this list is usually IP addresses). A host name is now passed as
4514 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4515 code that specifies a non-restarting handler (typically sigaction() in
4516 modern systems) in an attempt to fix a rare and obscure crash bug.
4518 15. Narrowed the window for a race in the daemon that could cause it to ignore
4519 SIGCHLD signals. This is not a major problem, because they are used only to
4520 wake it up if nothing else does.
4522 16. A malformed maildirsize file could cause Exim to calculate negative values
4523 for the mailbox size or file count. Odd effects could occur as a result.
4524 The maildirsize information is now recalculated if the size or filecount
4527 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4528 support for a long time. Removed HAVE_SYS_VFS_H.
4530 18. Updated exipick to current release from John Jetmore.
4532 19. Allow an empty sender to be matched against a lookup in an address list.
4533 Previously the only cases considered were a regular expression, or an
4536 20. Exim went into a mad DNS lookup loop when doing a callout where the
4537 host was specified on the transport, if the DNS lookup yielded more than
4540 21. The RFC2047 encoding function was originally intended for short strings
4541 such as real names; it was not keeping to the 75-character limit for
4542 encoded words that the RFC imposes. It now respects the limit, and
4543 generates multiple encoded words if necessary. To be on the safe side, I
4544 have increased the buffer size for the ${rfc2047: expansion operator from
4547 22. Failure to deliver a bounce message always caused it to be frozen, even if
4548 there was an errors_to setting on the router. The errors_to setting is now
4551 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4552 canonical form (fully expanded) before being placed in
4553 $sender_host_address.
4555 24. Updated eximstats to version 1.33
4557 25. Include certificate and key file names in error message when GnuTLS fails
4558 to set them up, because the GnuTLS error message doesn't include the name
4559 of the failing file when there is a problem reading it.
4561 26. Expand error message when OpenSSL has problems setting up cert/key files.
4564 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4565 (this can affect the format of dates).
4567 28. exim_tidydb, when checking for the continued existence of a message for
4568 which it has found a message-specific retry record, was not finding
4569 messages that were in split spool directories. Consequently, it was
4570 deleting retry records that should have stayed in existence.
4572 29. eximstats updated to version 1.35
4573 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4574 1.35 - bugfix such that pie charts by volume are generated correctly
4576 30. The SPA authentication driver was not abandoning authentication and moving
4577 on to the next authenticator when an expansion was forced to fail,
4578 contradicting the general specification for all authenticators. Instead it
4579 was generating a temporary error. It now behaves as specified.
4581 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4582 (the order specifies the preference for clients). The order is now AES256,
4583 AES128, 3DES, ARCFOUR128.
4585 31. Small patch to Sieve code - explicitly set From: when generating an
4588 32. Exim crashed if a remote delivery caused a very long error message to be
4589 recorded - for instance if somebody sent an entire SpamAssassin report back
4590 as a large number of 550 error lines. This bug was coincidentally fixed by
4591 increasing the size of one of Exim's internal buffers (big_buffer) that
4592 happened as part of the Exiscan merge. However, to be on the safe side, I
4593 have made the code more robust (and fixed the comments that describe what
4596 33. Some experimental protocols are using DNS PTR records for new purposes. The
4597 keys for these records are domain names, not reversed IP addresses. The
4598 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4599 leaves it alone. Component reversal etc. now happens only for IP addresses.
4602 34. The host_aton() function is supposed to be passed a string that is known
4603 to be a valid IP address. However, in the case of IPv6 addresses, it was
4604 not checking this. This is a hostage to fortune. Exim now panics and dies
4605 if the condition is not met. A case was found where this could be provoked
4606 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4607 components; fortuitously, this particular loophole had already been fixed
4608 by change 4.50/55 or 4.44/33 above.
4610 If there are any other similar loopholes, the new check in host_aton()
4611 itself should stop them being exploited. The report I received stated that
4612 data on the command line could provoke the exploit when Exim was running as
4613 exim, but did not say which command line option was involved. All I could
4614 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4615 running as the user.
4618 35. There was a buffer overflow vulnerability in the SPA authentication code
4619 (which came originally from the Samba project). I have added a test to the
4620 spa_base64_to_bits() function which I hope fixes it.
4623 36. The daemon start-up calls getloadavg() while still root for those OS that
4624 need the first call to be done as root, but it missed one case: when
4625 deliver_queue_load_max is set with deliver_drop_privilege. This is
4626 necessary for the benefit of the queue runner, because there is no re-exec
4627 when deliver_drop_privilege is set.
4629 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4630 was in use, was not putting the data itself into the right store pool;
4631 consequently, it could be overwritten for a subsequent message in the same
4632 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4635 38. Sometimes the final signoff response after QUIT could fail to get
4636 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4637 < 0 before doing a fflush(). This bug looks as though it goes back to the
4638 introduction of TLS in release 3.20, but "sometimes" must have been rare
4639 because the tests only now provoked it.
4645 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4646 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4647 Now it returns an integer. A number of calls were still expecting a T/F
4648 return. Fortuitously, in all cases, the tests worked in OK situations,
4649 which is the norm. However, things would have gone wrong on any write
4650 failures on the smtp file descriptor. This function is used when sending
4651 messages over SMTP and also when doing verify callouts.
4653 2. When Exim is called to do synchronous delivery of a locally submitted
4654 message (the -odf or -odi options), it no longer closes stderr before doing
4657 3. Implemented the mua_wrapper option.
4659 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4661 5. Implemented the functions header_remove(), header_testname(),
4662 header_add_at_position(), and receive_remove_recipient(), and exported them
4665 6. If an ACL "warn" statement specified the addition of headers, Exim already
4666 inserted X-ACL-Warn: at the start if there was no header name. However, it
4667 was not making this test for the second and subsequent header lines if
4668 there were newlines in the string. This meant that an invalid header could
4669 be inserted if Exim was badly configured.
4671 7. Allow an ACL "warn" statement to add header lines at the start or after all
4672 the Received: headers, as well as at the end.
4674 8. Added the rcpt_4xx retry error code.
4676 9. Added postmaster_mailfrom=xxx to callout verification option.
4678 10. Added mailfrom=xxxx to the callout verification option, for verify=
4681 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4682 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4684 12. Inserted some casts to stop certain compilers warning when using pointer
4685 differences as field lengths or precisions in printf-type calls (mostly
4686 affecting debugging statements).
4688 13. Added optional readline() support for -be (dynamically loaded).
4690 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4691 same clock tick as a message's arrival, so that its received time was the
4692 same as the "first fail" time on the retry record, and that message
4693 remained on the queue past the ultimate address timeout, every queue runner
4694 would try a delivery (because it was past the ultimate address timeout) but
4695 after another failure, the ultimate address timeout, which should have then
4696 bounced the address, did not kick in. This was a "< instead of <=" error;
4697 in most cases the first failure would have been in the next clock tick
4698 after the received time, and all would be well.
4700 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4701 being recognized when the domain list was tested by the match_domain
4702 condition in an expansion string.
4704 16. Added the ${str2b64: operator.
4706 17. Exim was always calling setrlimit() to set a large limit for the number of
4707 processes, without checking whether the existing limit was already
4708 adequate. (It did check for the limit on file descriptors.) Furthermore,
4709 errors from getrlimit() and setrlimit() were being ignored. Now they are
4710 logged to the main and panic logs, but Exim does carry on, to try to do its
4711 job under whatever limits there are.
4713 18. Imported PCRE 5.0.
4715 19. Trivial typo in log message " temporarily refused connection" (the leading
4718 20. If the log selector return_path_on_delivery was set and an address was
4719 redirected to /dev/null, the delivery process crashed because it assumed
4720 that a return path would always be set for a "successful" delivery. In this
4721 case, the whole delivery is bypassed as an optimization, and therefore no
4724 21. Internal re-arrangement: the function for sending a challenge and reading
4725 a response while authentication was assuming a zero-terminated challenge
4726 string. It's now changed to take a pointer and a length, to allow for
4727 binary data in such strings.
4729 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4731 23. Exim was not respecting finduser_retries when seeking the login of the
4732 uid under which it was called; it was always trying 10 times. (The default
4733 setting of finduser_retries is zero.) Also, it was sleeping after the final
4734 failure, which is pointless.
4736 24. Implemented tls_on_connect_ports.
4738 25. Implemented acl_smtp_predata.
4740 26. If the domain in control=submission is set empty, Exim assumes that the
4741 authenticated id is a complete email address when it generates From: or
4742 Sender: header lines.
4744 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4745 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4746 chown and chgrp in /bin and hostname in /usr/bin.
4748 28. Exim was keeping the "process log" file open after each use, just as it
4749 does for the main log. This opens the possibility of it remaining open for
4750 long periods when the USR1 signal hits a daemon. Occasional processlog
4751 errors were reported, that could have been caused by this. Anyway, it seems
4752 much more sensible not to leave this file open at all, so that is what now
4755 29. The long-running daemon process does not normally write to the log once it
4756 has entered its main loop, and it closes the log before doing so. This is
4757 so that log files can straightforwardly be renamed and moved. However,
4758 there are a couple of unusual error situations where the daemon does write
4759 log entries, and I had neglected to close the log afterwards.
4761 30. The text of an SMTP error response that was received during a remote
4762 delivery was being truncated at 512 bytes. This is too short for some of
4763 the long messages that one sometimes sees. I've increased the limit to
4766 31. It is now possible to make retry rules that apply only when a message has a
4767 specific sender, in particular, an empty sender.
4769 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4770 it possible to be selective about when SMTP synchronization is enforced.
4772 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4774 32. Implemented hosts_connection_nolog.
4776 33. Added an ACL for QUIT.
4778 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4781 35. Added mailbox_size and mailbox_filecount to appendfile.
4783 36. Added control = no_multiline_responses to ACLs.
4785 37. There was a bug in the logic of the code that waits for the clock to tick
4786 in the case where the clock went backwards by a substantial amount such
4787 that the microsecond fraction of "now" was more than the microsecond
4788 fraction of "then" (but the whole seconds number was less).
4790 38. Added support for the libradius Radius client library this is found on
4791 FreeBSD (previously only the radiusclient library was supported).
4797 1. When certain lookups returned multiple values in the form name=value, the
4798 quoting of the values was not always being done properly. Specifically:
4799 (a) If the value started with a double quote, but contained no whitespace,
4801 (b) If the value contained whitespace other than a space character (i.e.
4802 tabs or newlines or carriage returns) it was not quoted.
4803 This fix has been applied to the mysql and pgsql lookups by writing a
4804 separate quoting function and calling it from the lookup code. The fix
4805 should probably also be applied to nisplus, ibase and oracle lookups, but
4806 since I cannot test any of those, I have not disturbed their existing code.
4808 2. A hit in the callout cache for a specific address caused a log line with no
4809 reason for rejecting RCPT. Now it says "Previous (cached) callout
4810 verification failure".
4812 3. There was an off-by-one bug in the queryprogram router. An over-long
4813 return line was truncated at 256 instead of 255 characters, thereby
4814 overflowing its buffer with the terminating zero. As well as fixing this, I
4815 have increased the buffer size to 1024 (and made a note to document this).
4817 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4818 when Exim is waiting for an SMTP response from a remote server, Exim
4819 restarts its select() call on the socket, thereby resetting its timeout.
4820 This is not a problem when such interrupts are rare. Somebody set up a cron
4821 job to run exiwhat every 2 minutes, which is less than the normal select()
4822 timeout (5 or 10 minutes). This meant that the select() timeout never
4823 kicked in because it was always reset. I have fixed this by comparing the
4824 time when an interrupt arrives with the time at the start of the first call
4825 to select(). If more time than the timeout has elapsed, the interrupt is
4826 treated as a timeout.
4828 5. Some internal re-factoring in preparation for the addition of Sieve
4829 extensions (by MH). In particular, the "personal" test is moved to a
4830 separate function, and given an option for scanning Cc: and Bcc: (which is
4831 not set for Exim filters).
4833 6. When Exim created an email address using the login of the caller as the
4834 local part (e.g. when creating a From: or Sender: header line), it was not
4835 quoting the local part when it contained special characters such as @.
4837 7. Installed new OpenBSD configuration files.
4839 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4840 try to make them clearer.
4842 9. Callout options, other than the timeout value, were being ignored when
4843 verifying sender addresses in header lines. For example, when using
4845 verify = header_sender/callout=no_cache
4847 the cache was (incorrectly) being used.
4849 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4850 only the Cygwin environment.
4852 11. When return_path_on_delivery was set as a log selector, if different remote
4853 addresses in the same message used different return paths and parallel
4854 remote delivery occurred, the wrong values would sometimes be logged.
4855 (Whenever a remote delivery process finished, the return path value from
4856 the most recently started remote delivery process was logged.)
4858 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4859 lines when AUTH and/or TLS are in use. This is the "received protocol"
4860 field. Exim used to use "asmtp" for authenticated SMTP, without any
4861 indication (in the protocol name) for TLS use. Now it follows the RFC and
4862 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4863 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4864 names appear in log lines as well as in Received: header lines.
4866 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4867 extensions, and comparison tests, and to fix some bugs.
4869 14. Changes to the "personal" filter test:
4871 (1) The test was buggy in that it was just doing the equivalent of
4872 "contains" tests on header lines. For example, if a user's address was
4873 anne@some.where, the "personal" test would incorrectly be true for
4875 To: susanne@some.where
4877 This test is now done by extracting each address from the header in turn,
4878 and checking the entire address. Other tests that are part of "personal"
4879 are now done using regular expressions (for example, to check local parts
4880 of addresses in From: header lines).
4882 (2) The list of non-personal local parts in From: addresses has been
4883 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4884 taken from the Sieve specification recommendations.
4886 (3) If the message contains any header line starting with "List-" it is
4887 treated as non-personal.
4889 (4) The test for "circular" in the Subject: header line has been removed
4890 because it now seems ill-conceived.
4892 15. Minor typos in src/EDITME comments corrected.
4894 16. Installed latest exipick from John Jetmore.
4896 17. If headers_add on a router specified a text string that was too long for
4897 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4898 of string_sprintf() is now avoided.
4900 18. $message_body_size was not set (it was always zero) when running the DATA
4901 ACL and the local_scan() function.
4903 19. For the "mail" command in an Exim filter, no default was being set for
4904 the once_repeat time, causing a random time value to be used if "once" was
4905 specified. (If the value happened to be <= 0, no repeat happened.) The
4906 default is now 0s, meaning "never repeat". The "vacation" command was OK
4907 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4908 (I found it when inspecting the code).
4910 20. There is now an overall timeout for performing a callout verification. It
4911 defaults to 4 times the callout timeout, which applies to individual SMTP
4912 commands during the callout. The overall timeout applies when there is more
4913 than one host that can be tried. The timeout is checked before trying the
4914 next host. This prevents very long delays if there are a large number of
4915 hosts and all are timing out (e.g. when the network connections are timing
4916 out). The value of the overall timeout can be changed by specifying an
4917 additional sub-option for "callout", called "maxwait". For example:
4919 verify = sender/callout=5s,maxwait=20s
4921 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4922 to the end before writing, but this should make it even safer).
4924 22. Exim was forgetting that it had advertised PIPELINING for the second and
4925 subsequent messages on an SMTP connection. It was also not resetting its
4926 memory on STARTTLS and an internal HELO.
4928 23. When Exim logs an SMTP synchronization error within a session, it now
4929 records whether PIPELINING has been advertised or not.
4931 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4932 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4933 rather than long int.
4935 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4937 26. Added the never_mail option to autoreply.
4943 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4944 crash if the getsockname() call failed; this can happen if a connection is
4945 closed very soon after it is established. The problem was simply in the
4946 order in which certain operations were done, causing Exim to try to write
4947 to the SMTP stream before it had set up the file descriptor. The bug has
4948 been fixed by making things happen in the correct order.
4954 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4955 before the connection was closed, thus losing the rejection response.
4957 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4958 some early Solaris releases, but causes trouble in current releases where
4959 socklen_t is defined.
4961 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4964 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4967 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4968 that releasing the top of it at the end releases what was used for sub-
4969 expansions (unless the block got too big). However, discard this block if
4970 the first thing is a variable or header, so that we can use its block when
4971 it is dynamic (useful for very large $message_headers, for example).
4973 6. Lookups now cache *every* query, not just the most recent. A new, separate
4974 store pool is used for this. It can be recovered when all lookup caches are
4975 flushed. Lookups now release memory at the end of their result strings.
4976 This has involved some general refactoring of the lookup sources.
4978 7. Some code has been added to the store_xxx() functions to reduce the amount
4979 of flapping under certain conditions.
4981 8. log_incoming_interface used to affect only the <= reception log lines. Now
4982 it causes the local interface and port to be added to several more SMTP log
4983 lines, for example "SMTP connection from", and rejection lines.
4985 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4987 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4989 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4990 could be overwritten at the end of the current message (or the start of a
4991 new message if it was set in a HELO ACL). The value is now preserved for
4992 the duration of the SMTP connection.
4994 12. If a transport had a headers_rewrite setting, and a matching header line
4995 contained an unqualified address, that address was qualified, even if it
4996 did not match any rewriting rules. The underlying bug was that the values
4997 of the flags that permit the existence of unqualified sender and recipient
4998 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4999 non-local messages, and by -bnq for local messages) were not being
5000 preserved with the message after it was received.
5002 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5003 "next input=" as part of the text comprising the host identity instead of
5004 the correct text. The code was using the same buffer for two different
5005 strings. However, depending on which order the printing function evaluated
5006 its arguments, the bug did not always show up. Under Linux, for example, my
5007 test suite worked just fine.
5009 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5010 doesn't work with some older versions of Perl. It has been changed to "my",
5011 which in any case is probably the better facility to use.
5013 15. A really picky compiler found some instances of statements for creating
5014 error messages that either had too many or two few arguments for the format
5017 16. The size of the buffer for calls to the DNS resolver has been increased
5018 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5019 for addresses that have a lot of PTR records. This alleviates a problem; it
5020 does not fully solve it.
5022 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5023 buffer now truncates the list and logs the incident, which is the same
5024 action as happens when Exim is looking up a host name and its aliases.
5025 Previously in this situation something unpredictable would happen;
5026 sometimes it was "internal error: store_reset failed".
5028 18. If a server dropped the connection unexpectedly when an Exim client was
5029 using GnuTLS and trying to read a response, the client delivery process
5030 crashed while trying to generate an error log message.
5032 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5033 string, for example:
5035 warn message = H1: something\nH2: something
5037 the text was added as a single header line from Exim's point of view
5038 though it ended up OK in the delivered message. However, searching for the
5039 second and subsequent header lines using $h_h2: did not work. This has been
5040 fixed. Similarly, if a system filter added multiple headers in this way,
5041 the routers could not see them.
5043 20. Expanded the error message when iplsearch is called with an invalid key to
5044 suggest using net-iplsearch in a host list.
5046 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5047 ACLs are no longer actually imposed (and a message to that effect is
5050 22. If a "gecos" field in a passwd entry contained escaped characters, in
5051 particular, if it contained a \" sequence, Exim got it wrong when building
5052 a From: or a Sender: header from that name. A second bug also caused
5053 incorrect handling when an unquoted " was present following a character
5054 that needed quoting.
5056 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5057 was not being matched caselessly.
5059 24. Arranged for all hyphens in the exim.8 source to be escaped with
5062 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5063 properly. Recipient callout cache records were still being keyed to include
5064 the sender, even when use_sender was set false. This led to far more
5065 callouts that were necessary. The sender is no longer included in the key
5066 when use_sender is false.
5068 26. Added "control = submission" modifier to ACLs.
5070 27. Added the ${base62d: operator to decode base 62 numbers.
5072 28. dnsdb lookups can now access SRV records.
5074 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5075 the configuration file.
5077 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5078 (-v) mode. This makes the output for a verbose queue run more intelligible.
5080 31. Added a use_postmaster feature to recipient callouts.
5082 32. Added the $body_zerocount variable, containing the number of binary zero
5083 bytes in the message body.
5085 33. The time of last modification of the "new" subdirectory is now used as the
5086 "mailbox time last read" when there is a quota error for a maildir
5089 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5091 35. Added +ignore_unknown as a special item in host lists.
5093 36. Code for decoding IPv6 addresses in host lists is now included, even if
5094 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5095 address was recognized as an IP address, but was then not correctly decoded
5096 into binary, causing unexpected and incorrect effects when compared with
5103 1. Very minor rewording of debugging text in manualroute to say "list of
5104 hosts" instead of "hostlist".
5106 2. If verify=header_syntax was set, and a header line with an unqualified
5107 address (no domain) and a large number of spaces between the end of the
5108 name and the colon was received, the reception process suffered a buffer
5109 overflow, and (when I tested it) crashed. This was caused by some obsolete
5110 code that should have been removed. The fix is to remove it!
5112 3. When running in the test harness, delay a bit after writing a bounce
5113 message to get a bit more predictability in the log output.
5115 4. Added a call to search_tidyup() just before forking a reception process. In
5116 theory, someone could use a lookup in the expansion of smtp_accept_max_
5117 per_host which, without the tidyup, could leave open a database connection.
5119 5. Added the variables $recipient_data and $sender_data which get set from a
5120 lookup success in an ACL "recipients" or "senders" condition, or a router
5121 "senders" option, similar to $domain_data and $local_part_data.
5123 6. Moved the writing of debug_print from before to after the "senders" test
5126 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5127 problems for message scanning, either using a data ACL, or using
5128 local_scan() because the Received: header was not generated till after they
5129 were called (in order to set the time as the time of reception completion).
5130 I have revised the way this works. The header is now generated after the
5131 body is received, but before the ACL or local_scan() are called. After they
5132 are run, the timestamp in the header is updated.
5138 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5139 before starting a queue runner without re-exec. This happened only when
5140 deliver_drop_privilege was set or when the Exim user was set to root. The
5141 effect of the bug was that timeouts during subsequent deliveries caused
5142 crashes instead of being properly handled. The handler is now left at its
5143 default (and expected) setting.
5145 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5146 message, again when deliver_drop_privilege is set or Exim is run as root.
5147 The bug described in (1) was not present in this case, but the tidying up
5148 of the other signals was missing. I have made the two cases consistent.
5150 3. The ignore_target_hosts setting on a manualroute router was being ignored
5151 for hosts that were looked up using the /MX notation.
5153 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5156 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5157 operated on the sender address. After changing the $sender_address to <>
5158 for the sender address verify, Exim was re-instated it as the original
5159 (before rewriting) address, but remembering that it had rewritten it, so it
5160 wasn't rewriting it again. This bug also had the effect of breaking the
5161 sender address verification caching when the sender address was rewritten.
5163 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5164 This has been changed so that if the ip literal address matches
5165 ignore_target_hosts, the router declines.
5167 7. Added expansion conditions match_domain, match_address, and match_local_
5168 part (NOT match_host).
5170 8. The placeholder for the Received: header didn't have a length field set.
5172 9. Added code to Exim itself and to exim_lock to test for a specific race
5173 condition that could lead to file corruption when using MBX delivery. The
5174 issue is with the lockfile that is created in /tmp. If this file is removed
5175 after a process has opened it but before that process has acquired a lock,
5176 there is the potential for a second process to recreate the file and also
5177 acquire a lock. This could lead to two Exim processes writing to the file
5178 at the same time. The added code performs the same test as UW imapd; it
5179 checks after acquiring the lock that its file descriptor still refers to
5180 the same named file.
5182 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5183 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5186 11. Added the smtp_active_hostname option. If used, this will typically be made
5187 to depend on the incoming interface address. Because $interface_address is
5188 not set up until the daemon has forked a reception process, error responses
5189 that can happen earlier (such as "too many connections") no longer contain
5192 12. If an expansion in a condition on a "warn" statement fails because a lookup
5193 defers, the "warn" statement is abandoned, and the next ACL statement is
5194 processed. Previously this caused the whole ACL to be aborted.
5196 13. Added the iplsearch lookup type.
5198 14. Added ident_timeout as a log selector.
5200 15. Added tls_certificate_verified as a log selector.
5202 16. Added a global option tls_require_ciphers (compare the smtp transport
5203 option of the same name). This controls incoming TLS connections.
5205 17. I finally figured out how to make tls_require_ciphers do a similar thing
5206 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5207 before starting the TLS session.
5209 18. Tabs are now shown as \t in -bP output.
5211 19. If the log selector return_path_on_delivery was set, Exim crashed when
5212 bouncing a message because it had too many Received: header lines.
5214 20. If two routers both had headers_remove settings, and the first one included
5215 a superfluous trailing colon, the final name in the first list and the
5216 first name in the second list were incorrectly joined into one item (with a
5217 colon in the middle).
5223 1. Added -C and -D options to the exinext utility, mainly to make it easier
5224 to include in the automated testing, but these could be helpful when
5225 multiple configurations are in use.
5227 2. The exinext utility was not formatting the output nicely when there was
5228 an alternate port involved in the retry record key, nor when there was a
5229 message id as well (for retries that were specific to a specific message
5230 and a specific host). It was also confused by IPv6 addresses, because of
5231 the additional colons they contain. I have fixed the IPv4 problem, and
5232 patched it up to do a reasonable job for IPv6.
5234 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5235 delivery, the log line now contains "pipelined" if PIPELINING was used.
5237 4. An SMTP transport process used to panic and die if the bind() call to set
5238 an explicit outgoing interface failed. This has been changed; it is now
5239 treated in the same way as a connect() failure.
5241 5. A reference to $sender_host_name in the part of a conditional expansion
5242 that was being skipped was still causing a DNS lookup. This no longer
5245 6. The def: expansion condition was not recognizing references to header lines
5246 that used bh_ and bheader_.
5248 7. Added the _cache feature to named lists.
5250 8. The code for checking quota_filecount in the appendfile transport was
5251 allowing one more file than it should have been.
5253 9. For compatibility with Sendmail, the command line option
5261 and sets the incoming protocol and host name (for trusted callers). The
5262 host name and its colon can be omitted when only the protocol is to be set.
5263 Note the Exim already has two private options, -pd and -ps, that refer to
5264 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5265 "s", but I don't think that's a major issue.
5267 10. A number of refactoring changes to the code, none of which should affect
5270 (a) The number of logging options was getting close to filling up the
5271 32-bit word that was used as a bit map. I have split them into two classes:
5272 those that are passed in the argument to log_write(), and those that are
5273 only ever tested independently outside of that function. These are now in
5274 separate 32-bit words, so there is plenty of room for expansion again.
5275 There is no change in the user interface or the logging behaviour.
5277 (b) When building, for example, log lines, the code previously used a
5278 macro that called string_cat() twice, in order to add two strings. This is
5279 not really sufficiently general. Furthermore, there was one instance where
5280 it was actually wrong because one of the argument was used twice, and in
5281 one call a function was used. (As it happened, calling the function twice
5282 did not affect the overall behaviour.) The macro has been replaced by a
5283 function that can join an arbitrary number of extra strings onto a growing
5286 (c) The code for expansion conditions now uses a table and a binary chop
5287 instead of a serial search (which was left over from when there were very
5288 few conditions). Also, it now recognizes conditions like "pam" even when
5289 the relevant support is not compiled in: a suitably worded error message is
5290 given if an attempt is made to use such a condition.
5292 11. Added ${time_interval:xxxxx}.
5294 12. A bug was causing one of the ddress fields not to be passed back correctly
5295 from remote delivery subprocesses. The field in question was not being
5296 subsequently used, so this caused to problems in practice.
5298 13. Added new log selectors queue_time and deliver_time.
5300 14. Might have fixed a bug in maildirsizefile handling that threw up
5301 "unexpected character" debug warnings, and recalculated the data
5302 unnecessarily. In any case, I expanded the warning message to give more
5305 15. Added the message "Restricted characters in address" to the statements in
5306 the default ACL that block characters like @ and % in local parts.
5308 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5309 Three changes have been made:
5311 (a) There was a serious bug; a negative response to MAIL caused the whole
5312 recipient domain to be cached as invalid, thereby blocking all messages
5313 to all local parts at the same domain, from all senders. This bug has
5314 been fixed. The domain is no longer cached after a negative response to
5315 MAIL if the sender used is not empty.
5317 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5320 (c) A new callout option, "use_sender" has been added for people who want
5321 the modified behaviour.
5327 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5330 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5331 indeed breaks things for older releases.
5333 3. Added additional logging to the case where there is a problem reading data
5334 from a filter that is running in a subprocess using a pipe, in order to
5335 try to track down a specific problem.
5337 4. Testing facility fudge: when running in the test harness and attempting
5338 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5339 getting "No route to host". Convert this to a timeout.
5341 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5344 6. Some OS don't have socklen_t but use size_t instead. This affects the
5345 fifth argument of getsockopt() amongst other things. This is now
5346 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5347 can be set for individual OS. I have set it for SunOS5, OSF1, and
5348 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5349 some earlier ones do not.
5351 7. Change 4.30/15 was not doing the test caselessly.
5353 8. The standard form for an IPv6 address literal was being rejected by address
5354 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5355 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5356 this, as well as the form without the "IPv6" on the front (but only when
5357 address literals are enabled, of course).
5359 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5361 10. Exim crashed if a message with an empty sender address specified by -f
5362 encountered a router with an errors_to setting. This could be provoked only
5363 by a command such as
5367 where an empty string was supplied; "<>" did not hit this bug.
5369 11. Installed PCRE release 4.5.
5371 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5372 remained set. It is now erased.
5374 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5375 times from message ids (which are base 36 rather than the normal 62).
5377 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5378 were being counted as actual protocol errors, and logged if the log
5379 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5380 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5381 and DATA following a set of rejected RCPTs do not count as protocol errors.
5382 In other words, Exim assumes they were pipelined, though this may not
5383 actually be the case. Of course, in all cases the client gets an
5384 appropriate error code.
5386 15. If a lookup fails in an ACL condition, a message about the failure may
5387 be available; it is used if testing the ACL cannot continue, because most
5388 such messages specify what the cause of the deferral is. However, some
5389 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5390 that caused an old message to be retained and used if a later statement
5391 caused a defer, replacing the real cause of the deferral.
5393 16. If an IP address had so many PTR records that the DNS lookup buffer
5394 was not large enough to hold them, Exim could crash while trying to process
5395 the truncated data. It now detects and logs this case.
5397 17. Further to 4.21/58, another change has been made: if (and only if) the
5398 first line of a message (the first header line) ends with CRLF, a bare LF
5399 in a subsequent header line has a space inserted after it, so as not to
5400 terminate the header.
5402 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5403 unnecessarily, used atoi() instead of strtol(), and didn't check the
5404 termination when getting file sizes from file names by regex.
5406 19. Completely re-implemented the support for maildirsize files, in the light
5407 of a number of problems with the previous contributed implementation
5408 (4.30/29). In particular:
5410 . If the quota is zero, the maildirsize file is maintained, but no quota is
5413 . If the maildir directory does not exist, it is created before any attempt
5414 to write a maildirsize file.
5416 . The quota value in the file is just a cache; if the quota is changed in
5417 the transport, the new value overrides.
5419 . A regular expression is available for excluding directories from the
5422 20. The autoreply transport checks the characters in options that define the
5423 message's headers; it allows continued headers, but it was checking with
5424 isspace() after an embedded newline instead of explicitly looking for a
5427 21. If all the "regular" hosts to which an address was routed had passed their
5428 expiry times, and had not reached their retry times, the address was
5429 bounced, even if fallback hosts were defined. Now Exim should go on to try
5432 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5433 equivalent code in the SMTP transport. Some hosts send humungous responses
5434 to HELO/EHLO, more than 1024 it seems.
5436 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5437 gives compiler warnings in some environments. I've now done it "properly",
5440 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5441 (because of problems with the built-in one) was declared to return uschar *
5442 instead of char *, causing compiler failure.
5444 25. Fixed a file descriptor leak when processing alias/forward files.
5446 26. Fixed a minor format string issue in dbfn.c.
5448 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5450 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5451 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5454 29. When a sender address is verified, it is cached, to save repeating the test
5455 when there is more than one recipient in a message. However, when the
5456 verification involves a callout, it is possible for different callout
5457 options to be set for different recipients. It is too complicated to keep
5458 track of this in the cache, so now Exim always runs a verification when a
5459 callout is required, relying on the callout cache for the optimization.
5460 The overhead is duplication of the address routing, but this should not be
5463 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5464 to be verified with callout=postmaster, and the main callout worked but the
5465 postmaster check failed, the verification correctly failed. However, if a
5466 subsequent RCPT command asked for sender verification *without* the
5467 postmaster check, incorrect caching caused this verification also to fail,
5470 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5471 it was not caching the DNS options (qualify_single, search_parents) that
5472 were used when the lookup failed. A subsequent lookup with different
5473 options therefore always gave the same answer, though there were cases
5474 where it should not have. (Example: a "domains = !$mx_any" option on a
5475 dnslookup router: the "domains" option is always processed without any
5476 widening, but the router might have qualify_single set.) Now Exim uses the
5477 cached value only when the same options are set.
5479 32. Added John Jetmore's "exipick" utility to the distribution.
5481 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5482 than a timeout (e.g. a certificate is required, and is not provided), an
5483 Exim server now closes the connection immediately. Previously it waited for
5484 the client to close - but if the client is SSL, it seems that they each
5485 wait for each other, leading to a delay before one of them times out.
5487 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5488 maintained 0.8.x compatibility because I don't think many are using it, and
5489 it is clearly obsolete.
5491 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5494 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5495 client certificate was expired. A simple patch fixes this, though I don't
5496 understand the full logic of why the verify callback is called multiple
5499 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5500 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5501 which causes problems with some clients (such as the Certicom SSL Plus
5502 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5503 disables the coutermeasure allowing Eudora to connect."
5505 38. Exim was not checking that a write() to a log file succeeded. This could
5506 lead to Bad Things if a log got too big, in particular if it hit a file
5507 size limit. Exim now panics and dies if it cannot write to a log file, just
5508 as it does if it cannot open a log file.
5510 39. Modified OS/Makefile-Linux so that it now contains
5512 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5514 The two -D definitions ensure that Exim is compiled with large file
5515 support, which makes it possible to handle log files that are bigger than
5518 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5519 instance) a domain was checked against a named list that involved a lookup,
5520 causing $domain_data to be set, then another domain was checked against the
5521 same list, then the first domain was re-checked, the value of $domain_data
5522 after the final check could be wrong. In particular, if the second check
5523 failed, it could be set empty. This bug probably also applied to
5526 41. The strip_trailing_dot option was not being applied to the address given
5527 with the -f command-line option.
5529 42. The code for reading a message's header from the spool was incrementing
5530 $received_count, but never initializing it. This meant that the value was
5531 incorrect (doubled) while delivering a message in the same process in which
5532 it was received. In the most common configuration of Exim, this never
5533 happens - a fresh exec is done - but it can happen when
5534 deliver_drop_privilege is set.
5536 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5537 it now includes up to 150 characters of the unexpected data in the log
5540 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5541 and building data strings. The size of both of these buffers was 10 000
5542 bytes - far larger than anybody would *ever* want, thought I. Needless to
5543 say, somebody hit the limit. I have increased the maximum line length to
5544 20 000 and the maximum data length of concatenated lines to 100 000. I have
5545 also fixed two bugs, because there was no checking on these buffers. Tsk,
5546 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5547 buffer is too small.
5549 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5550 lsearch lookups. Now it does.
5552 46. When parsing a route_list item in a manualroute router, a fixed-length
5553 buffer was used for the list of hosts. I made this 1024 bytes long,
5554 thinking that nobody would ever have a list of hosts that long. Wrong.
5555 Somebody had a whole pile of complicated expansion conditions, and the
5556 string was silently truncated, leading to an expansion error. It turns out
5557 that it is easier to change to an unlimited length (owing to other changes
5558 that have happened since this code was originally written) than to build
5559 structure for giving a limitation error. The length of the item that
5560 expands into the list of hosts is now unlimited.
5562 47. The lsearch lookup could not handle data where the length of text line was
5563 more than 4095 characters. Such lines were truncated, leading to shortened
5564 data being returned. It should now handle lines of any length.
5566 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5567 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5570 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5571 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5572 no longer happens in comment lines. A list of replacements is now placed
5573 at the head of all of the source files, except those whose only change is
5574 to replace PERL_COMMAND in the very first #! line.
5576 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5577 messages on the queue, with a bottom-up merge sort, using code contributed
5578 by Michael Haardt. This should make operations like -bp somewhat faster on
5579 large queues. It won't affect queue runners, except when queue_run_in_order
5582 51. Installed eximstats 1.31 in the distribution.
5584 52. Added support for SRV lookups to the dnslookup router.
5586 53. If an ACL referred to $message_body or $message_body_end, the value was not
5587 reset for any messages that followed in the same SMTP session.
5589 54. The store-handling optimization for building very long strings was not
5590 differentiating between the different store pools. I don't think this
5591 actually made any difference in practice, but I've tidied it.
5593 55. While running the routers to verify a sender address, $sender_address
5594 was still set to the sender address. This is wrong, because when routing to
5595 send a bounce to the sender, it would be empty. Therefore, I have changed
5596 it so that, while verifying a sender address, $sender_address is set to <>.
5597 (There is no change to what happens when verifying a recipient address.)
5599 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5600 target A or AAAA records (if not already returned) without resetting the
5601 qualify_single or search_parents options of the DNS resolver. These are
5602 inappropriate in this case because the targets of MX and SRV records must
5603 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5604 target that, when qualified, matched something in the local domain. These
5605 two options are now turned off when doing these lookups.
5607 57. It seems that at least some releases of Reiserfs (which does not have the
5608 concept of a fixed number of inodes) returns zero and not -1 for the
5609 number of available inodes. This interacted badly with check_spool_inodes,
5610 which assumed that -1 was the "no such thing" setting. What I have done is
5611 to check that the total number of inodes is greater than zero before doing
5612 the test of how many are available.
5614 58. When a "warn" ACL statement has a log_message modifier, the message is
5615 remembered, and not repeated. This is to avoid a lot of repetition when a
5616 message has many recipients that cause the same warning to be written.
5617 However, Exim was preserving the list of already written lines for an
5618 entire SMTP session, which doesn't seem right. The memory is now reset if a
5619 new message is started.
5621 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5622 debugging output unless log_rewrite was also set.
5624 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5625 of (int)(handle) when we know that handle contains (void *)(-1).
5627 61. The Exim daemon panic-logs an error return when it closes the incoming
5628 connection. However "connection reset by peer" seems to be common, and
5629 isn't really an error worthy of noting specially, so that particular error
5632 62. When Exim is trying to find all the local interfaces, it used to panic and
5633 die if the ioctl to get the interface flags failed. However, it seems that
5634 on at least one OS (Solaris 9) it is possible to have an interface that is
5635 included in the list of interfaces, but for which you get a failure error
5636 for this call. This happens when the interface is not "plumbed" into a
5637 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5638 failure of the "get flags" call assumes that the interface is down.
5640 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5641 makes life easier for people who are doing arithmetic on fields extracted
5642 from dates, where you often get leading zeros that should not be
5643 interpreted as octal.
5645 64. Added qualify_domain to the redirect router, to override the global
5648 65. If a pathologically long header line contained very many addresses (the
5649 report of this problem mentioned 10 000) and each of them was rewritten,
5650 Exim could use up a very large amount of memory. (It kept on making new
5651 copies of the header line as it rewrote, and never released the old ones.)
5652 At the expense of a bit more processing, the header rewriting function has
5653 been changed so that it no longer eats memory in this way.
5655 66. The generation of the Received: header has been moved from the time that a
5656 message starts to be received, to the time that it finishes. The timestamp
5657 in the Received: header should now be very close to that of the <= log
5658 line. There are two side-effects of this change:
5660 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5661 logged header lines no longer include the local Received: line, because
5662 it has not yet been created. The same applies to a copy of the message
5663 that is returned to a non-SMTP sender when a message is rejected.
5665 (b) When a filter file is tested using -bf, no additional Received: header
5666 is added to the test message. After some thought, I decided that this
5669 This change does not affect the value of $received_for. It is still set
5670 after address rewriting, but before local_scan() is called.
5672 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5674 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5675 gave an unhelpful panic error message, and a defer error. I have managed to
5676 change this behaviour so that it now rejects any supplied certificate,
5677 which seems right, as the list of acceptable certificates is empty.
5679 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5680 gave an unhelpful defer error. I have not managed to make this reject any
5681 supplied certificates, but the error message it gives is "no certificate
5682 supplied", which is not helpful.
5684 70. exigrep's output now also includes lines that are not associated with any
5685 message, but which match the given pattern. Implemented by a patch from
5686 Martin Sluka, which also tidied up the Perl a bit.
5688 71. Recipient callout verification, like sender verification, was using <> in
5689 the MAIL FROM command. This isn't really the right thing, since the actual
5690 sender may affect whether the remote host accepts the recipient or not. I
5691 have changed it to use the actual sender in the callout; this means that
5692 the cache record is now keyed on a recipient/sender pair, not just the
5693 recipient address. There doesn't seem to be a real danger of callout loops,
5694 since a callout by the remote host to check the sender would use <>.
5695 [SEE ABOVE: changed after hitting problems.]
5697 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5698 temporary errors. However, in the case of such a code being given after
5699 the end of a data transmission (i.e. after ".") Exim was failing to write
5700 a retry record for the message. (Yes, there was some broken host that was
5701 actually sending 8xx at this point.)
5703 73. An unknown lookup type in a host list could cause Exim to panic-die when
5704 the list was checked. (An example that provoked this was putting <; in the
5705 middle of a list instead of at the start.) If this happened during a DATA
5706 ACL check, a -D file could be left lying around. This kind of configuration
5707 error no longer causes Exim to die; instead it causes a defer error. The
5708 incident is still logged to the main and panic logs.
5710 74. Buglet left over from Exim 3 conversion. The message "too many messages
5711 in one connection" was written to the rejectlog but not the mainlog, except
5712 when address rewriting (yes!) was being logged.
5714 75. Added write_rejectlog option.
5716 76. When a system filter was run not as root (that is, when system_filter_user
5717 was set), the values of the $n variables were not being returned to the
5718 main process; thus, they were not subsequently available in the $sn
5721 77. Added +return_path_on_delivery log selector.
5723 78. A connection timeout was being treated differently from recipients deferred
5724 when testing hosts_max_try with a message that was older than the host's
5725 retry timeout. (The host should not be counted, thus allowing all hosts to
5726 be tried at least once before bouncing.) This may have been the cause of an
5727 occasionally reported bug whereby a message would remain on the queue
5728 longer than the retry timeout, but would be bounced if a delivery was
5729 forced. I say "may" because I never totally pinned down the problem;
5730 setting up timeout/retry tests is difficult. See also the next item.
5732 79. The ultimate address timeout was not being applied to errors that involved
5733 a combination of host plus message (for example, a timeout on a MAIL
5734 command). When an address resolved to a number of possible hosts, and they
5735 were not all tried for each delivery (e.g. because of hosts_max_try), a
5736 message could remain on the queue longer than the retry timeout.
5738 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5741 81. Fixed an obscure SMTP outgoing bug which required at least the following
5742 conditions: (a) there was another message waiting for the same server;
5743 (b) the server returned 5xx to all RCPT commands in the first message so
5744 that the message was not completed; (c) the server dropped the connection
5745 or gave a negative response to the RSET that Exim sends to abort the
5746 transaction. The observed case was a dropped connection after DATA that had
5747 been sent in pipelining mode. That is, the server had advertised PIPELINING
5748 but was not implementing it correctly. The effect of the bug was incorrect
5749 behaviour, such as trying another host, and this could lead to a crash.
5755 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5756 and daemon.c were passed as pointers to ints; they should have been
5757 pointers to socklen_t variables (which are typically unsigned ints).
5759 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5762 3. Fixed a really odd bug that affected only the testing scheme; patching a
5763 certain fixed string in the binary changed the value of another string that
5764 happened to be identical to the end of the original first string.
5766 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5767 name", it returns that address as the IP address. On some operating
5768 systems (e.g. Solaris), it also passes back the IP address string as the
5769 "host name". However, on others (e.g. Linux), it passes back an empty
5770 string. Exim wasn't checking for this, and was changing the host name to an
5771 empty string, assuming it had been canonicalized.
5773 5. Although rare, it is permitted to have more than one PTR record for a given
5774 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5775 all the names associated with an address, because they do in Solaris.
5776 However, it seems that they do not in Linux for data that comes from the
5777 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5778 I found this out when I moved to a new Linux workstation and tried to run
5779 the Exim test suite.
5781 To get round this problem I have changed the code so that it now does its
5782 own call to the DNS to look up PTR records when searching for a host name.
5783 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5784 addresses that are only in /etc/hosts are still found.
5786 This behaviour is, however, controlled by an option called host_lookup_
5787 order, which defaults to "bydns:byaddr". If people want to use the other
5788 order, or indeed, just use one or the other means of lookup, they can
5789 specify it in this variable.
5791 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5792 some operating systems, this comes back from gethostbyaddr() as an empty
5793 string, and this is what Exim used to test for. However, it seems that in
5794 other systems, "." is yielded. Exim now tests for this case too.
5796 7. The values of check_spool_space and check_log_space are now held internally
5797 as a number of kilobytes instead of an absolute number of bytes. If a
5798 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5799 kilobyte. This means that much larger values can be stored.
5801 8. Exim monitor: an attempt to get the action menu when not actually pointing
5802 at a message produces an empty menu entitled "No message selected". This
5803 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5804 no entries in it ("Shell widget menu has zero width and/or height"). So I
5805 have added a single, blank menu entry in this case.
5807 9. Added ${quote_local_part.
5809 10. MIME decoding is now applied to the contents of Subject: header lines when
5812 11. Now that a reference to $sender_host_address automatically causes a reverse
5813 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5814 host lookup before query-style lookups in lists that might use this
5815 variable. This has therefore been abolished, and the "net-" prefix is no
5816 longer necessary for query-style lookups.
5818 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5819 have been a typo for LFLAGS, so it has been changed.
5821 13. The install script calls Exim with "-C /dev/null" in order to find the
5822 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5823 to be output. However, since Exim outputs its version number before the
5824 error, it didn't break the script. It just looked ugly. I fixed this by
5825 always allowing "-C /dev/null" if the caller is root.
5827 14. Ignore overlarge ACL variable number when reading spool file - insurance
5828 against a later release with more variables having written the file.
5830 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5831 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5832 this, as well as the form without the "IPv6" on the front.
5834 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5835 OS/Makefile-Darwin file.
5837 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5838 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5840 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5841 variable. However, some OS use ints for inodes. I've added cast to long int
5842 to get rid of the compiler warning.
5844 19. I had forgotten to lock out "/../" in configuration file names when
5845 ALT_CONFIG_PREFIX was set.
5847 20. Routers used for verification do not need to specify transports. However,
5848 if such a router generated a host list, and callout was configured, Exim
5849 crashed, because it could not find a port number from the (non-existent)
5850 transport. It now assumes port 25 in this circumstance.
5852 21. Added the -t option to exigrep.
5854 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5855 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5856 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5857 them is set, LOOKUP_LSEARCH is forced.
5859 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5861 24. Added sender and host information to the "rejected by local_scan()" log
5862 line; previously there was no indication of these.
5864 25. Added .include_if_exists.
5866 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5867 Linux. It turns out that not all file systems support this. Apparently some
5868 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5869 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5870 not supported on the file descriptor, is now ignored when Exim is trying to
5871 sync a directory. This applies only to Linux.
5873 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5875 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5876 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5879 29. Added support for maildirsize files from supplied patch (modified a bit).
5881 30. The use of :fail: followed by an empty string could lead Exim to respond to
5882 sender verification failures with (e.g.):
5884 550 Verification failed for <xxx>
5885 550 Sender verify failed
5887 where the first response line was missing the '-' that indicates it is not
5888 the final line of the response.
5890 31. The loop for finding the name of the user that called Exim had a hardwired
5891 limit of 10; it now uses the value of finduser_retries, which is used for
5892 all other user lookups.
5894 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5897 33. Exim was neglecting to zero errno before one call of strtol() when
5898 expanding a string and expecting an integer value. On some systems this
5899 resulted in spurious "integer overflow" errors. Also, it was casting the
5900 result into an int without checking.
5902 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5903 did not work. The code looks as if it has *never* worked, though it appears
5904 to have been documented since at least release 1.62. I have made it work.
5906 35. The "timeout_DNS" error in retry rules, also documented since at least
5907 1.62, also never worked. As it isn't clear exactly what this means, and
5908 clearly it isn't a major issue, I have abolished the feature by treating it
5909 as "timeout", and writing a warning to the main and panic logs.
5911 36. The display of retry rules for -brt wasn't always showing the error code
5914 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5915 timeout_connect_A, timeout_connect_MX.
5917 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5918 to the empty sender.
5920 39. The daemon was not analysing the content of -oX till after it had closed
5921 stderr and disconnected from the controlling terminal. This meant that any
5922 syntax errors were only noted on the panic log, and the return code from
5923 the command was 0. By re-arranging the code a little, I've made the
5924 decoding happen first, so such errors now appear on stderr, and the return
5925 code is 1. However, the actual setting up of the sockets still happens in
5926 the disconnected process, so errors there are still only recorded on the
5929 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5930 connections (as happens on some IP stacks) was logged at start up time as
5931 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5932 it from "IPv6 and IPv4", which means that two separate sockets are being
5935 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5936 says whether AF_INET or AF_INET6 was passed as an argument.
5938 42. Exiwhat output was messed up when time zones were included in log
5941 43. Exiwhat now gives more information about the daemon's listening ports,
5942 and whether -tls-on-connect was used.
5944 44. The "port" option of the smtp transport is now expanded.
5946 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5947 silently ignored. Now an error message is written to the main and panic
5950 46. There's a new ACL modifier called "logwrite" which writes to a log file
5951 as soon as it is encountered.
5953 47. Added $local_user_uid and $local_user_gid at routing time.
5955 48. Exim crashed when trying to verify a sender address that was being
5958 49. Exim was recognizing only a space character after ".include". It now also
5959 recognizes a tab character.
5961 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5962 extracting the relevant information from the specification. The man page no
5963 longer contains scrambled data for the -d option, and I've added a section
5964 at the front about calling Exim under different names.
5966 51. Added "extra_headers" argument to the "mail" command in filter files.
5968 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5971 53. Installed eximstats 1.29.
5973 54. Added transport_filter_timeout as a generic transport option.
5975 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5976 Cc: header lines. This was required by RFC 822, but it not required by RFC
5979 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5980 incoming messages that did not have them. Now it does so only if the
5981 message originates locally, that is, if there is no associated remote host
5982 address. When Resent- header lines are present, this applies to the Resent-
5983 lines rather than the non-Resent- lines.
5985 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5986 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5988 58. Messages for configuration errors now include the name of the main
5989 configuration file - useful now that there may be more than one file in a
5990 list (.included file names were always shown).
5992 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5993 for those rare installations that do not start the daemon as root or run it
5994 setuid root. I've cut out the call to initgroups() if the daemon is not
5997 60. The Exim user and group can now be bound into the binary as text strings
5998 that are looked up at the start of Exim's processing.
6000 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6002 62. Added $mailstore_basename variable.
6004 63. Installed patch to sieve.c from Michael Haardt.
6006 64. When Exim failed to open the panic log after failing to open the main log,
6007 the original message it was trying to log was written to stderr and debug
6008 output, but if they were not available (the usual case in production), it
6009 was lost. Now it is written to syslog before the two lines that record the
6010 failures to open the logs.
6012 65. Users' Exim filters run in subprocesses under the user's uid. It is
6013 possible for a "deliver" command or an alias in a "personal" command to
6014 provoke an address rewrite. If logging of address rewriting is configured,
6015 this fails because the process is not running as root or exim. There may be
6016 a better way of dealing with this, but for the moment (because 4.30 needs
6017 to be released), I have disabled address rewrite logging when running a
6018 filter in a non-root, non-exim process.
6024 1. The buildconfig auxiliary program wasn't quoting the value set for
6025 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6026 not defined. This bug was masked in 4.22 by the effect that was fixed in
6029 2. Some messages that were rejected after a message id was allocated were
6030 shown as "incomplete" by exigrep. It no longer does this for messages that
6031 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6033 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6034 have allow_domain_literals set, the ID did not get logged in the <= line.
6035 Domain literals are now always recognized in Message-ID: header lines.
6037 4. The first argument for a ${extract expansion item is the key name or field
6038 number. Leading and trailing spaces in this item were not being ignored,
6039 causing some misleading effects.
6041 5. When deliver_drop_privilege was set, single queue runner processes started
6042 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6043 same command in the process it spins off) were not dropping privilege.
6045 6. When the daemon running as "exim" started a queue runner, it always
6046 re-executed Exim in the spun-off process. This is a waste of effort when
6047 deliver_drop_privilege is set. The new process now just calls the
6048 queue-runner function directly.
6054 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6057 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6058 ignored. Though the use of -oP was forcing the writing of a pid file, it
6059 was always written to the default place.
6061 3. If the message "no IP address found for host xxxx" is generated during
6062 incoming verification, it is now followed by identification of the incoming
6063 connection (so you can more easily find what provoked it).
6065 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6067 5. Added some features to "harden" Exim a bit more against certain attacks:
6069 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6070 be put in Local/Makefile. This is like the never_users runtime option,
6071 but it cannot be overridden. The default setting is "root".
6073 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6074 prefix string with which any file named in a -C command line option
6077 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6078 is retained for -C and -D only if the caller of Exim is root. Without
6079 it, the exim user may also use -C and -D and retain privilege.
6081 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6082 command line option is disabled.
6084 6. Macro names set by the -D option must start with an upper case letter, just
6085 like macro names defined in the configuration file.
6087 7. Added "dereference=" facility to LDAP.
6089 8. Two instances of the typo "uknown" in the source files are fixed.
6091 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6092 the Configure-Makefile script screwed up while processing it.
6094 10. Incorporated PCRE 4.4.
6096 11. The SMTP synchronization check was not operating right at the start of an
6097 SMTP session. For example, it could not catch a HELO sent before the client
6098 waited for the greeting. There is now a check for outstanding input at the
6099 point when the greeting is written. Because of the duplex, asynchronous
6100 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6101 way, but not yet received, when the check is performed.
6103 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6104 on TCP/IP sockets, because this apparently causes some broken clients to
6107 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6108 unchanged) from the Cygwin maintainer.
6110 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6111 was supported instead of testing for mbx. Effectively a typo.
6113 15. The spa authenticator server code was not checking that the input it
6114 received was valid base64.
6116 16. The debug output line for the "set" modifier in ACLs was not showing the
6117 name of the variable that was being set.
6119 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6121 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6122 lookup on its own. Something else was needed to trigger the lookup. For
6123 example, a match in host_lookup or the need for a host name in a host list.
6124 Now, if $sender_host_name is referenced and the host name has not yet been
6125 looked up, a lookup is performed. If the lookup fails, the variable remains
6126 empty, and $host_lookup_failed is set to "1".
6128 19. Added "eqi" as a case-independent comparison operator.
6130 20. The saslauthd authentication condition could segfault if neither service
6131 nor realm was specified.
6133 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6134 error message that was logged was misleading, and incoming SMTP
6135 connections were dropped. The message is now more accurate, and temporary
6136 errors are given to SMTP connections.
6138 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6139 (except RSET) with a 421 error, until QUIT is received. However, it was
6140 failing to send a response to QUIT.
6142 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6143 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6144 ACL is now run after the helo_try_verify_hosts code.
6146 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6147 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6148 case-independent, so other case variants are also recognized). Apparently
6149 some systems use these upper case variants.
6151 25. If more than two messages were waiting for the same host, and a transport
6152 filter was specified for the transport, Exim sent two messages over the
6153 same TCP/IP connection, and then failed with "socket operation on non-
6154 socket" when it tried to send the third.
6156 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6158 27. The extern definition of crypt16() in expand.c was not being excluded when
6159 the OS had its own crypt16() function.
6161 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6162 as a preferred synonym for return_size_limit, both as an option and as an
6165 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6167 30. Changed the default configuration ACL to relax the local part checking rule
6168 for addresses that are not in any local domains. For these addresses,
6169 slashes and pipe symbols are allowed within local parts, but the sequence
6170 /../ is explicitly forbidden.
6172 31. SPA server authentication was not clearing the challenge buffer before
6175 32. log_message in a "warn" ACL statement was writing to the reject log as
6176 well as to the main log, which contradicts the documentation and doesn't
6177 seem right (because no rejection is happening). So I have stopped it.
6179 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6180 However, I am unable to do any testing of this.
6182 34. Fixed an infelicity in the appendfile transport. When checking directories
6183 for a mailbox, to see if any needed to be created, it was accidentally
6184 using path names with one or more superfluous leading slashes; tracing
6185 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6187 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6188 contents are added to the log line that is written for every discarded
6189 recipient. (Previously a log_message setting was ignored.)
6191 36. The ${quote: operator now quotes the string if it is empty.
6193 37. The install script runs exim in order to find its version number. If for
6194 some reason other than non-existence or emptiness, which it checks, it
6195 could not run './exim', it was installing it with an empty version number,
6196 i.e. as "exim-". This error state is now caught, and the installation is
6199 38. An argument was missing from the function that creates an error message
6200 when Exim fails to connect to the socket for saslauthd authentication.
6201 This could cause Exim to crash, or give a corrupted message.
6203 39. Added isip, isip4, and isip6 to ${if conditions.
6205 40. The ACL variables $acl_xx are now saved with the message, and can be
6206 accessed later in routers, transports, and filters.
6208 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6209 strings in the file are not string-expanded.
6211 42. If a MAIL command specified a SIZE value that was too large to fit into an
6212 int variable, the check against message_size_limit failed. Such values are
6213 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6214 day this will have to be increased, but I don't think I want to be around
6215 when emails are that large.
6222 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6223 iconv() is not standard in FreeBSD.
6225 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6226 IPv6 enabled. The observed symptom was a segmentation fault on return from
6227 the function os_common_find_running_interfaces() in src/os.c.
6229 3. In the check_special_case() function in daemon.c I had used "errno" as an
6230 argument name, which causes warnings on some systems. This was basically a
6231 typo, since it was named "eno" in the comments!
6233 4. The code that waits for the clock to tick (at a resolution of some fraction
6234 of a second) so as to ensure message-id uniqueness was always waiting for
6235 at least one whole tick, when it could have waited for less. [This is
6236 almost certainly not relevant at current processor speeds, where it is
6237 unlikely to ever wait at all. But we try to future-proof.]
6239 5. The function that sleeps for a time interval that includes fractions of a
6240 second contained a race. It did not block SIGALRM between setting the
6241 timer, and suspending (a couple of lines later). If the interval was short
6242 and the sigsuspend() was delayed until after it had expired, the suspension
6243 never ended. On busy systems this could lead to processes getting stuck for
6246 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6247 process, before it forks any delivery processes. The open lookup caching
6248 mechanism meant that the open file or database connection was passed into
6249 the delivery process. The problem was that delivery processes always tidy
6250 up cached lookup data. This could cause a problem for the next delivery
6251 process started by the queue runner, because the external queue runner
6252 process does not know about the closure. So the next delivery process
6253 still has data in the lookup cache. In the case of a file lookup, there was
6254 no problem because closing a file descriptor in a subprocess doesn't affect
6255 the parent. However, if the lookup was caching a connection to a database,
6256 the connection was closed, and the second delivery process was likely to
6257 see errors such as "PGSQL: query failed: server closed the connection
6258 unexpectedly". The problem has been fixed by closing all cached lookups
6259 in a queue runner before running a delivery process.
6261 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6262 seem to have the "const" qualifier which it has on other OS. I've
6265 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6266 *of the same type* (client or server) with the same public name that an
6267 error should be diagnosed.
6269 9. When Exim looked up a host name for an IP address, but failed to find the
6270 original IP address when looking up the host name (a safety check), it
6271 output the message "<ip address> does not match any IP for NULL", which was
6272 confusing, to say the least. The bug was that the host name should have
6273 appeared instead of "NULL".
6275 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6276 user that is built into the binary, and the -C or -D options is used, root
6277 privilege is dropped before the configuration file is read. In addition,
6278 logging is switched to stderr instead of the normal log files. If the
6279 configuration then re-defines the Exim user, the unprivileged environment
6280 is probably not what is expected, so Exim logs a panic warning message (but
6283 However, if deliver_drop_privilege is set, the unprivileged state may well
6284 be exactly what is intended, so the warning has been cut out in that case,
6285 and Exim is allowed to try to write to its normal log files.
6291 1. smtp_return_error_details was not giving details for temporary sender
6292 or receiver verification errors.
6294 2. Diagnose a configuration error if two authenticators have the same public
6297 3. Exim used not to create the message log file for a message until the first
6298 delivery attempt. This could be confusing when incoming messages were held
6299 for policy or load reasons. The message log file is now created at the time
6300 the message is received, and an initial "Received" line is written to it.
6302 4. The automatically generated man page for command line options had a minor
6303 bug that caused no ill effects; however, a more serious problem was that
6304 the procedure for building the man page automatically didn't always
6305 operate. Consequently, release 4.20 contains an out-of-date version. This
6306 shouldn't happen again.
6308 5. When building Exim with embedded Perl support, the script that builds the
6309 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6310 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6312 6. The freeze_tell option was not being used for messages that were frozen on
6313 arrival, either by an ACL or by local_scan().
6315 7. Added the smtp_incomplete_transaction log selector.
6317 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6318 was accepting AUTH without a new EHLO.
6320 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6321 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6324 10. Logging of TCP/IP connections (when configured) now happens in the main
6325 daemon process instead of the child process, so that the TCP/IP connection
6326 count is more accurate (but it can never be perfect).
6328 11. The use of "drop" in a nested ACL was not being handled correctly in the
6329 outer ACL. Now, if condition failure induced by the nested "drop" causes
6330 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6331 or "require"), the connection is dropped.
6333 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6334 that yield "discard" can now be used with an "accept" or a "discard" verb,
6335 but an error is generated for any others (because I can't see a useful way
6336 to define what should happen).
6338 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6339 are now processed in the same way as lines in the Exim configuration file.
6340 In particular, continuation lines are supported.
6342 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6344 15. Added -ti meaning -t -i.
6346 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6347 domains, and warn by logging if others are found.
6349 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6350 returned from GIFCONF when Exim is trying to find the list of interfaces on
6351 a host. The code in os.c has been modified to copy each ifreq to an aligned
6352 structure in all cases.
6354 Also, in some cases, the returned ifreq's were being copied to a 'struct
6355 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6356 means the last couple of bytes of an IPv6 address could be chopped if the
6357 ifreq contained only a normal sockaddr (14 bytes storage).
6359 18. Named domain lists were not supported in the hosts_treat_as_local option.
6360 An entry such as +xxxx was not recognized, and was treated as a literal
6363 19. Ensure that header lines added by a DATA ACL are included in the reject log
6364 if the ACL subsequently rejects the message.
6366 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6367 MD5 (which is deprecated).
6369 21. When testing a filter file using -bf, Exim was writing a message when it
6370 took the sender from a "From " line in the message, but it was not doing so
6371 when it took $return_path from a Return-Path: header line. It now does.
6373 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6374 with a valid header line field name (a series of printing characters
6375 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6377 23. Changed "disc" in the source to "disk" to conform to the documentation and
6378 the book and for uniformity.
6380 24. Ignore Sendmail's -Ooption=value command line item.
6382 25. When execve() failed while trying to run a command in a pipe transport,
6383 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6384 could be confused with a return value of 69 from the command itself. This
6385 has been changed to 127, the value the shell returns if it is asked to run
6386 a non-existent command. The wording for the related log line suggests a
6387 non-existent command as the problem.
6389 26. If received_header_text expands to an empty string, do not add a Received:
6390 header line to the message. (Well, it adds a token one on the spool, but
6391 marks it "old" so that it doesn't get used or transmitted.)
6393 27. Installed eximstats 1.28 (addition of -nt option).
6395 28. There was no check for failure on the call to getsockname() in the daemon
6396 code. This can fail if there is a shortage of resources on the system, with
6397 ENOMEM, for example. A temporary error is now given on failure.
6399 29. Contrary to the C standard, it seems that in some environments, the
6400 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6401 program. Exim now does this explicitly; it affects the formatting of
6402 timestamps using strftime().
6404 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6405 complaints. I've now initialized all the variables, to avoid this.
6407 32. Header lines added by a system filter were not being "seen" during
6408 transport-time rewrites.
6410 33. The info_callback() function passed to OpenSSL is set up with type void
6411 (*)(SSL *, int, int), as described somewhere. However, when calling the
6412 function (actually a macro) that sets it up, the type void(*)() is
6413 expected. I've put in a cast to prevent warnings from picky compilers.
6415 34. If a DNS black list lookup found a CNAME record, but there were no A
6416 records associated with the domain it pointed at, Exim crashed.
6418 35. If a DNS black list lookup returned more than one A record, Exim ignored
6419 all but the first. It now scans all returned addresses if a particular IP
6420 value is being sought. In this situation, the contents of the
6421 $dnslist_value variable are a list of all the addresses, separated by a
6424 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6425 to accept a host name and all its aliases if the forward lookup for any of
6426 them yielded the IP address of the incoming connection. Now it accepts only
6427 those names whose forward lookup yields the correct IP address. Any other
6428 names are discarded. This closes a loophole whereby a rogue DNS
6429 administrator could create reverse DNS records to break through a
6430 wildcarded host restriction in an ACL.
6432 37. If a user filter or a system filter that ran in a subprocess used any of
6433 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6434 the wrong values were passed to the pipe command ($thisaddress had the
6435 value of $0, $0 had the value of $1, etc). This bug was introduced by
6436 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6438 38. Improved the line breaking for long SMTP error messages from ACLs.
6439 Previously, if there was no break point between 40 and 75 characters, Exim
6440 left the rest of the message alone. Two changes have been made: (a) I've
6441 reduced the minimum length to 35 characters; (b) if it can't find a break
6442 point between 35 and 75 characters, it looks ahead and uses the first one
6443 that it finds. This may give the occasional overlong line, but at least the
6444 remaining text gets split now.
6446 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6447 file descriptors might be low, and that setting 1000 would always raise it.
6448 It turns out that in some environments, the limit is already over 1000 and
6449 that lowering it causes trouble. So now Exim takes care not to decrease it.
6451 40. When delivering a message, the value of $return_path is set to $sender_
6452 address at the start of routing (routers may change the value). By an
6453 oversight, this default was not being set up when an address was tested by
6454 -bt or -bv, which affected the outcome if any router or filter referred to
6457 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6458 log only when "message" or "log_message" are set. However, if one of the
6459 conditions was an address verification, or a call to a nested ACL, the
6460 messages generated by the underlying test were being passed through. This
6461 no longer happens. The underlying message is available in $acl_verify_
6462 message for both "message" and "log_message" expansions, so it can be
6463 passed through if needed.
6465 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6466 new expansion $bh_ to give the encoded byte string without charset
6467 translation. Translation happens only if iconv() is available; HAVE_ICONV
6468 indicates this at build time. HEADERS_CHARSET gives the charset to
6469 translate to; headers_charset can change it in the configuration, and
6470 "headers charset" can change it in an individual filter file.
6472 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6473 that creates RFC 2047 encoded "words" labels them as that charset instead
6474 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6475 expansion operator; (ii) when Exim creates a From: line for a local
6476 message; (iii) when a header line is rewritten to include a "phrase" part.
6478 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6479 buggy, causing it to skip the first lines of messages whose message ID
6480 ended in 'D'. This would not have bitten before Exim release 4.14, because
6481 message IDs were unlikely to end in 'D' before then. The effect was to have
6482 incorrect size information for certain domains.
6484 45. #include "config.h" was missing at the start of the crypt16.c module. This
6485 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6488 46. If there was a timeout during a "random" callout check, Exim treated it as
6489 a failure of the random address, and carried on sending RSET and the real
6490 address. If the delay was just some slowness somewhere, the response to the
6491 original RCPT would be taken as a response to RSET and so on, causing
6492 mayhem of various kinds.
6494 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6495 when I implemented it. It didn't allow for the fact that some option values
6496 may legitimately be negative (e.g. size_addition), and it didn't even do
6497 the right test for positive values.
6499 48. Domain names in DNS records are case-independent. Exim always looks them up
6500 in lower case. Some resolvers return domain names in exactly the case they
6501 appear in the zone file, that is, they may contain uppercase letters. Not
6502 all resolvers do this - some return always lower case. Exim was treating a
6503 change of case by a resolver as a change of domain, similar to a widening
6504 of a domain abbreviation. This triggered its re-routing code and so it was
6505 trying to route what was effectively the same domain again. This normally
6506 caused routing to fail (because the router wouldn't handle the domain
6507 twice). Now Exim checks for this case specially, and just changes the
6508 casing of the domain that it ultimately uses when it transmits the message
6511 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6514 50. If a filter generated a file delivery with a non-absolute name (possible if
6515 no home directory exists for the router), the forbid_file option was not
6518 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6519 the existing equality matching.
6521 52. Exim was using ints instead of ino_t variables in some places where it was
6522 dealing with inode numbers.
6524 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6525 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6526 called TMPDIR, and if it finds it is different, it changes its value.
6528 54. The smtp_printf() function is now made available to local_scan() so
6529 additional output lines can be written before returning. There is also an
6530 smtp_fflush() function to enable the detection of a dropped connection.
6531 The variables smtp_input and smtp_batched_input are exported to
6534 55. Changed the default runtime configuration: the message "Unknown user"
6535 has been removed from the ACL, and instead placed on the localuser router,
6536 using the cannot_route_message feature. This means that any verification
6537 failures that generate their own messages won't get overridden. Similarly,
6538 the "Unrouteable address" message that was in the ACL for unverifiable
6539 relay addresses has also been removed.
6541 56. Added hosts_avoid_esmtp to the smtp transport.
6543 57. The exicyclog script was not checking for the esoteric option
6544 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6545 will work only if exicyclog is run under the appropriate euid.
6547 58. Following a discussion on the list, the rules by which Exim recognises line
6548 endings on incoming messages have been changed. The -dropcr and drop_cr
6549 options are now no-ops, retained only for backwards compatibility. The
6550 following line terminators are recognized: LF CRLF CR. However, special
6551 processing applies to CR:
6553 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6554 nor a local message in the state where . is a terminator.
6556 (ii) If a bare CR is encountered in a header line, an extra space is added
6557 after the line terminator so as not to end the header. The reasoning
6558 behind this is that bare CRs in header lines are most likely either
6559 to be mistakes, or people trying to play silly games.
6561 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6562 was being incorrectly given as 18 bytes larger than it should have been.
6563 This is a VOB (very old bug).
6565 60. This may never have affected anything current, but just in case it has:
6566 When the local host is found other than at the start of a list of hosts,
6567 the local host, those with the same MX, and any that follow, are discarded.
6568 When the list in question was part of a longer list of hosts, the following
6569 hosts (not currently being processed) were also being discarded. This no
6570 longer happens. I'm not sure if this situation could ever has previously
6573 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6576 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6577 header. This is something that is recommended in a new Internet Draft, and
6578 is something that is documented as being done by Sendmail. There are two
6579 possible values. For messages generated by the autoreply transport, Exim
6582 Auto-Submitted: auto-replied
6584 whereas for all other generated messages (e.g. bounces) it adds
6586 Auto-Submitted: auto-generated
6588 63. The "personal" condition in filters now includes a test for the
6589 Auto-Submitted: header. If it contains the string "auto-" the message it
6590 not considered personal.
6592 64. Added rcpt_include_affixes as a generic transport option.
6594 65. Added queue_only_override (default true).
6596 66. Added the syslog_duplication option.
6598 67. If what should have been the first header line of a message consisted of
6599 a space followed by a colon, Exim was mis-interpreting it as a header line.
6600 It isn't of course - it is syntactically invalid and should therefore be
6601 treated as the start of the message body. The misbehaviour could have
6602 caused a number of strange effects, including loss of data in subsequent
6603 header lines, and spool format errors.
6605 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6606 client host had authenticated. This control can now be exercised by an ACL
6607 for more flexibility.
6609 69. By default, callouts do not happen when testing with -bh. There is now a
6610 variant, -bhc, which does actually run the callout code, including
6611 consulting and updating the callout cache.
6613 70. Added support for saslauthd authentication, courtesy of Alexander
6616 71. If statvfs() failed on the spool or log directories while checking their
6617 size for availability, Exim confusingly gave the error "space shortage".
6618 Furthermore, in debugging mode it crashed with a floating point exception.
6619 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6620 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6621 serious problem, Exim now writes to the main and panic logs when this
6622 happens, with details of the failure. It then refuses to accept the
6623 incoming message, giving the message "spool directory problem" or "log
6624 directory problem" with a 421 code for SMTP messages.
6626 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6627 0, 1, and 2 exist, because some OS complain for execs without them (see
6628 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6629 descriptors. However, the code omitted to check that the open succeeded,
6630 causing mysterious errors if for some reason the permissions on /dev/null
6631 got screwed. Now Exim writes a message to the main and panic logs, and
6632 bombs out if it can't open /dev/null.
6634 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6635 interact so that it is all more flexible. It is supposed to remain
6636 backwards compatible. Also added extra_local_interfaces.
6638 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6639 to bomb out with an assertion failure - to the client this appears as a
6640 connection drop. This problem occurs in the part of the code that was taken
6641 from the Samba project. Fortunately, the assertion is in a very simple
6642 function, so I have fixed this by reproducing the function inline in the
6643 one place where it is called, and arranging for authentication to fail
6644 instead of killing the process with assert().
6646 75. The SPA client code was not working when the server requested OEM rather
6647 than Unicode encoding.
6649 76. Added code to make require_files with a specific uid setting more usable in
6650 the case where statting the file as root fails - usually a non-root-mounted
6651 NFS file system. When this happens and the failure is EACCES, Exim now
6652 forks a subprocess and does the per-uid checking as the relevant uid.
6654 77. Added process_log_path.
6656 78. If log_file_path was not explicitly set, a setting of check_log_space or
6657 check_log_inodes was ignored.
6659 79. If a space check for the spool or log partitions fails, the incident is now
6660 logged. Of course, in the latter case the data may get lost...
6662 80. Added the %p formatting code to string_format() so that it can be used to
6663 print addresses in debug_print(). Adjusted all the address printing in the
6664 debugging in store.c to use %p rather than %d.
6666 81. There was a concern that a line of code in smtp_in.c could overflow a
6667 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6668 initially expressed, the concern was not well-founded, because trailing
6669 spaces are removed early. However, if the trailing spaces were followed by
6670 a NULL, they did not get removed, so the overflow was possible. Two fixes
6673 (a) I re-wrote the offending code in a cleaner fashion.
6674 (b) If an incoming SMTP command contains a NULL character, it is rejected
6677 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6678 runs initgroups(), so that if the Exim user is in any additional groups,
6679 they will be used during message reception.
6685 The change log for 4.20 and earlier releases has been archived.