1 Change log file for Exim from version 4.21
2 -------------------------------------------
6 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
7 supports it and a size is available (ie. the sending peer gave us one).
9 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
10 years of being deprecated). The replacements are the ACLs
11 acl_smtp_mime and acl_not_smtp_mime.
13 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
14 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
15 or even an in-clear connection were permitted. Now, if the host lookup
16 was dnssec and dane was requested then the host is only used if the
17 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
18 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
19 if one fails this test.
20 This means that a poorly-configured remote DNS will make it incommunicado;
21 but it protects against a DNS-interception attack on it.
23 JH/04 Bug 1810: make continued-use of an open smtp transport connection
24 non-noisy when a race steals the message being considered.
26 JH/05 If main configuration option tls_certificate is unset, generate a
27 selfsigned certificate for inbound TLS connections.
29 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
30 in rewrites and routers.
32 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
33 and logged a warning sing 4.83; now they are a configuration file error.
35 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
36 (lacking @domain). Apply the same qualification processing as RCPT.
38 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
40 JH/10 Support ${sha256:} applied to a string (as well as the previous
43 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
44 a cutthrough deliver is pending, as we always want to make a connection.
45 This also avoids re-routing the message when later placing the cutthrough
46 connection after a verify cache hit.
47 Do not update it with the verify result either.
49 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
50 when routing results in more than one destination address.
52 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
53 signing (which inhibits the cutthrough capability). Previously only
54 the presence of an option was tested; now an expansion evaluating as
55 empty is permissible (obviously it should depend only on data available
56 when the cutthrough connection is made).
58 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
59 the relevant preceding SMTP command did not note the pipelining mode.
61 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
62 Previously they were not counted.
64 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
65 as one having no matching records. Previously we deferred the message
66 that needed the lookup.
68 JH/17 Fakereject: previously logged as a norml message arrival "<="; now
69 distinguished as "(=".
71 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
72 for missing MX records. Previously it only worked for missing A records.
74 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
76 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
77 after the data-go-ahead and data-ack. Patch from Jason Betts.
79 JH/21 Bug 1846: Send DMARC forensic reports for reject and quaratine results,
80 even for a "none" policy. Patch from Tony Meyer.
82 JH/22 Fix continued use of a connection for further deliveries. If a port was
83 specified by a router, it must also match for the delivery to be
86 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
87 When one of the recipients of a message was unsuitable for the connection
88 (has no matching addresses), we lost track of needing to mark it
89 deferred. As a result mail would be lost.
91 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
96 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
97 and 3.4.4 - once the server is enabled to respond to an OCSP request
98 it does even when not requested, resulting in a stapling non-aware
99 client dropping the TLS connection.
101 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
102 support variable-length bit vectors. No functional change.
104 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
105 The I= interface field on outgoing lines is now after the H= remote
106 host field, same as incoming lines. There is a separate
107 outgoing_interface log selector which allows you to disable the
110 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
111 If not running log_selector +smtp_connection the mainlog would be held
112 open indefinitely after a "too many connections" event, including to a
113 deleted file after a log rotate. Leave the per net connection logging
114 leaving it open for efficiency as that will be quickly detected by the
115 check on the next write.
117 HS/01 Bug 1671: Fix post transport crash.
118 Processing the wait-<transport> messages could crash the delivery
119 process if the message IDs didn't exist for some reason. When
120 using 'split_spool_directory=yes' the construction of the spool
121 file name failed already, exposing the same netto behaviour.
123 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
124 mime_regex ACL conditions.
126 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
127 to DSN fail messages (bounces): remote IP, remote greeting, remote response
128 to HELO, local diagnostic string.
130 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
131 log line to debug. Even when configured with a tls authenticator many
132 client connections are expected to not authenticate in this way, so
133 an authenticate fail is not an error.
135 HS/02 Add the Exim version string to the process info. This way exiwhat
136 gives some more detail about the running daemon.
138 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
139 matter for fast-change records such as DNSBLs.
141 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
142 retry record, even if constant. There may be multiple transports with
143 different interface settings and the retry behaviour needs to be kept
146 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
148 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
150 JH/10 Bug 840: fix log_defer_output option of pipe transport
152 JH/11 Bug 830: use same host for all RCPTS of a message, even under
153 hosts_randomize. This matters a lot when combined with mua_wrapper.
155 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
156 ${quote_pgsql:<string>} operator.
158 JH/13 Bug 1708: avoid misaligned access in cached lookup.
160 JH/14 Change header file name for freeradius-client. Relevant if compiling
161 with Radius support; from the Gentoo tree and checked under Fedora.
163 JH/15 Bug 1712: Introduce $prdr_requested flag variable
165 JH/16 Bug 1714: Permit an empty string as expansion result for transport
166 option transport_filter, meaning no filtering.
168 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
170 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
171 defaults to "*" (all hosts). The variable is now available when not built
172 with TLS, default unset, mainly to enable keeping the testuite sane.
173 If a server certificate is not supplied (via tls_certificate) an error is
174 logged, and clients will find TLS connections fail on startup. Presumably
175 they will retry in-clear.
176 Packagers of Exim are strongly encouraged to create a server certificate
177 at installation time.
179 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
180 with the $config_file variable.
182 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
183 in transport context, after the attempt, and per-recipient. The latter type
184 is per host attempted. The event data is the error message, and the errno
185 information encodes the lookup type (A vs. MX) used for the (first) host,
186 and the trailing two digits of the smtp 4xx reponse.
188 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
189 to write to mainlog (or rejectlog, paniclog) in the window between file
190 creation and permissions/ownership being changed. Particularly affects
191 installations where exicyclog is run as root, rather than exim user;
192 result is that the running daemon panics and dies.
194 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
196 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
197 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
198 "pri" and "weight". Note that the previous implicit priority given by the
199 list order is no longer honoured.
201 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
204 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
205 by defining SUPPORT_SOCKS.
207 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
208 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
209 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
210 variables are renamed to proxy_{local,external}_{address,port}.
212 JH/27 Move Internationalisation support from Experimental to mainline, enabled
213 for a build by defining SUPPORT_I18N
215 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
216 of the query string, and make ${quote_redis:} do that quoting.
218 JH/29 Move Events support from Experimental to mainline, enabled by default
219 and removable for a build by defining DISABLE_EVENT.
221 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
223 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
224 cached by the daemon.
226 JH/32 Move Redis support from Experimental to mainline, enabled for a build
227 by defining LOOKUP_REDIS. The libhiredis library is required.
229 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
230 keys are given for lookup.
232 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
233 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
234 only supported when built with TLS support. The PolarSSL SHA routines
235 are still used when the TLS library is too old for convenient support.
237 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
238 openssl_options), for security. OpenSSL forces this from version 1.1.0
239 server-side so match that on older versions.
241 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
242 allocation for $value could be released as the expansion processing
243 concluded, but leaving the global pointer active for it.
245 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
246 and to use the domains and local_parts ACL conditions.
248 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
249 incorrectly not doubled on cutthrough transmission, hence seen as a
250 body-termination at the receiving system - resulting in truncated mails.
251 Commonly the sender saw a TCP-level error, and retransmitted the message
252 via the normal store-and-forward channel. This could result in duplicates
253 received - but deduplicating mailstores were liable to retain only the
254 initial truncated version.
256 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
258 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
260 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
261 we're in there, support oversigning also; bug 1309.
263 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
265 HS/04 Add support for keep_environment and add_environment options.
267 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
268 either intentional arithmetic overflow during PRNG, or testing config-
271 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
272 delivery resulted in actual delivery. Cancel cutthrough before DATA
275 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
276 rejects a recipient: pass the reject to the originator.
278 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
279 Many were false-positives and ignorable, but it's worth fixing the
282 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
283 for the new environment-manipulation done at startup. Move the routines
284 from being local to tls.c to being global via the os.c file.
286 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
287 an extract embedded as result-arg for a map, the first arg for extract
288 is unavailable so we cannot tell if this is a numbered or keyed
289 extraction. Accept either.
295 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
298 JH/02 The smtp transport option "multi_domain" is now expanded.
300 JH/03 The smtp transport now requests PRDR by default, if the server offers
303 JH/04 Certificate name checking on server certificates, when exim is a client,
304 is now done by default. The transport option tls_verify_cert_hostnames
305 can be used to disable this per-host. The build option
306 EXPERIMENTAL_CERTNAMES is withdrawn.
308 JH/05 The value of the tls_verify_certificates smtp transport and main options
309 default to the word "system" to access the system default CA bundle.
310 For GnuTLS, only version 3.0.20 or later.
312 JH/06 Verification of the server certificate for a TLS connection is now tried
313 (but not required) by default. The verification status is now logged by
314 default, for both outbound TLS and client-certificate supplying inbound
317 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
320 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
321 Status Notification (bounce) messages are now MIME format per RFC 3464.
322 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
323 under the control of the dsn_advertise_hosts option, and routers may
324 have a dsn_lasthop option.
326 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
327 default, modifiable by a malware= option. The list separator for
328 the options can now be changed in the usual way. Bug 68.
330 JH/10 The smtp_receive_timeout main option is now expanded before use.
332 JH/11 The incoming_interface log option now also enables logging of the
333 local interface on delivery outgoing connections.
335 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
336 if the interface and destination host and port all match.
338 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
341 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
342 Patch from Andrew Lewis.
344 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
345 now supports optional time-restrictions, weighting, and priority
346 modifiers per server. Patch originally by <rommer@active.by>.
348 JH/16 The spamd_address main option now supports a mixed list of local
349 and remote servers. Remote servers can be IPv6 addresses, and
350 specify a port-range.
352 JH/17 Bug 68: The spamd_address main option now supports an optional
353 timeout value per server.
355 JH/18 Bug 1581: Router and transport options headers_add/remove can
356 now have the list separator specified.
358 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
361 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
364 JH/21 Support for the A6 type of dns record is withdrawn.
366 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
367 rather than the verbs used.
369 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
370 from 255 to 1024 chars.
372 JH/24 Verification callouts now attempt to use TLS by default.
374 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
375 are generic router options now. The defaults didn't change.
377 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
378 Original patch from Alexander Shikoff, worked over by JH.
380 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
381 files if ZCAT_COMMAND is not executable.
383 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
385 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
387 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
388 Normally benign, it bites when the pair was led to by a CNAME;
389 modern usage is to not canoicalize the domain to a CNAME target
390 (and we were inconsistent anyway for A-only vs AAAA+A).
392 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
394 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
395 when evaluating $sender_host_dnssec.
397 JH/31 Check the HELO verification lookup for DNSSEC, adding new
398 $sender_helo_dnssec variable.
400 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
402 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
404 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
406 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
407 documented as working, but never had. Support all but $spam_report.
409 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
410 added for tls authenticator.
412 HS/03 Add perl_taintmode main config option
417 TL/01 When running the test suite, the README says that variables such as
418 no_msglog_check are global and can be placed anywhere in a specific
419 test's script, however it was observed that placement needed to be near
420 the beginning for it to behave that way. Changed the runtest perl
421 script to read through the entire script once to detect and set these
422 variables, reset to the beginning of the script, and then run through
423 the script parsing/test process like normal.
425 TL/02 The BSD's have an arc4random API. One of the functions to induce
426 adding randomness was arc4random_stir(), but it has been removed in
427 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
428 function when detected.
430 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
431 cause callback expansion.
433 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
434 syntax errors in an expansion can be treated as a string instead of
435 logging or causing an error, due to the internal use of bool_lax
436 instead of bool when processing it.
438 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
439 server certificates when making smtp deliveries.
441 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
443 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
445 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
447 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
448 Merged patch from Sebastian Wiedenroth.
450 JH/05 Fix results-pipe from transport process. Several recipients, combined
451 with certificate use, exposed issues where response data items split
452 over buffer boundaries were not parsed properly. This eventually
453 resulted in duplicates being sent. This issue only became common enough
454 to notice due to the introduction of conection certificate information,
455 the item size being so much larger. Found and fixed by Wolfgang Breyha.
457 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
458 size buffer was used, resulting in syntax errors when an expansion
461 JH/07 Add support for directories of certificates when compiled with a GnuTLS
462 version 3.3.6 or later.
464 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
465 is EXPERIMENTAL_EVENT, the main-configuration and transport options
466 both become "event_action", the variables become $event_name, $event_data
467 and $event_defer_errno. There is a new variable $verify_mode, usable in
468 routers, transports and related events. The tls:cert event is now also
469 raised for inbound connections, if the main configuration event_action
472 TL/06 In test suite, disable OCSP for old versions of openssl which contained
473 early OCSP support, but no stapling (appears to be less than 1.0.0).
475 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
476 server certificate names available under the smtp transport option
477 "tls_verify_cert_hostname" now do not permit multi-component wildcard
480 JH/10 Time-related extraction expansions from certificates now use the main
481 option "timezone" setting for output formatting, and are consistent
482 between OpenSSL and GnuTLS compilations. Bug 1541.
484 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
485 encoded parameter in the incoming message. Bug 1558.
487 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
488 include certificate info, eximon was claiming there were spoolfile
491 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
493 JH/14 Log delivery-related information more consistently, using the sequence
494 "H=<name> [<ip>]" wherever possible.
496 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
497 are problematic for Debian distribution, omit them from the release
500 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
502 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
505 JH/17 Fix a null-indirection in certextract expansions when a nondefault
506 output list separator was used.
511 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
512 checkers that were complaining about end of non-void function with no
515 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
516 This was a regression intruduced in 4.83 by another bugfix.
518 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
520 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
521 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
527 TF/01 Correctly close the server side of TLS when forking for delivery.
529 When a message was received over SMTP with TLS, Exim failed to clear up
530 the incoming connection properly after forking off the child process to
531 deliver the message. In some situations the subsequent outgoing
532 delivery connection happened to have the same fd number as the incoming
533 connection previously had. Exim would try to use TLS and fail, logging
534 a "Bad file descriptor" error.
536 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
537 utilities have not been installed.
539 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
540 temporary space as the ACL may create new global variables.
542 TL/01 LDAP support uses per connection or global context settings, depending
543 upon the detected version of the libraries at build time.
545 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
546 to extract and use the src ip:port in logging and expansions as if it
547 were a direct connection from the outside internet. PPv2 support was
548 updated based on HAProxy spec change in May 2014.
550 JH/02 Add ${listextract {number}{list}{success}{fail}}.
552 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
553 Properly escape header and check for NULL return.
555 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
558 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
560 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
561 characters in header names, implemented as a verify condition.
562 Contributed by Michael Fischer v. Mollard.
564 TL/05 Rename SPF condition results err_perm and err_temp to standardized
565 results permerror and temperror. Previous values are deprecated but
566 still accepted. In a future release, err_perm and err_temp will be
567 completely removed, which will be a backward incompatibility if the
568 ACL tests for either of these two old results. Patch contributed by
569 user bes-internal on the mailing list.
571 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
573 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
574 selectors, in both main and reject logs.
576 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
579 JH/07 Add malware type "sock" for talking to simple daemon.
581 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
583 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
584 routers/transports under cutthrough routing.
586 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
587 numbers. Touch up "bool" conditional to keep the same definition.
589 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
591 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
593 JH/12 Expand items in router/transport headers_add or headers_remove lists
594 individually rather than the list as a whole. Bug 1452.
596 Required for reasonable handling of multiple headers_ options when
597 they may be empty; requires that headers_remove items with embedded
598 colons must have them doubled (or the list-separator changed).
600 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
601 view the policy declared in the DMARC record. Currently, $dmarc_status
602 is a combined value of both the record presence and the result of the
605 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
607 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
608 dnslookup router and the smtp transport (applying to the forward
611 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
612 of ldap servers used for a specific lookup. Patch provided by Heiko
615 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
616 New variable $lookup_dnssec_authenticated for observability.
618 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
619 Patch submitted by Lars Timman.
621 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
623 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
624 Requires trusted mode and valid format message id, aborts otherwise.
625 Patch contributed by Heiko Schlichting.
627 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
628 certextract with support for various fields. Bug 1358.
630 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
631 is requested by default, modifiable by smtp transport option
634 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
635 operate on certificate variables to give certificate fingerprints
636 Also new ${sha256:cert_variable}.
638 JH/23 The PRDR feature is moved from being Experimental into the mainline.
640 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
641 Christian Aistleitner.
643 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
645 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
646 file. Patch from Wolfgang Breyha.
648 JH/25 Expand the coverage of the delivery $host and $host_address to
649 client authenticators run in verify callout. Bug 1476.
651 JH/26 Port service names are now accepted for tls_on_connect_ports, to
652 align with daemon_smtp_ports. Bug 72.
654 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
655 support and error reporting did not work properly.
657 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
658 and is readable. Patch from Andrew Colin Kissa.
660 TL/14 Enhance documentation of ${run expansion and how it parses the
661 commandline after expansion, particularly in the case when an
662 unquoted variable expansion results in an empty value.
664 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
666 PP/02 Fix internal collision of T_APL on systems which support RFC3123
667 by renaming away from it. Addresses GH issue 15, reported by
670 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
672 TL/15 SECURITY: prevent double expansion in math comparison functions
673 (can expand unsanitized data). Not remotely exploitable.
680 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
682 PP/02 Make -n do something, by making it not do something.
683 When combined with -bP, the name of an option is not output.
685 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
688 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
689 $sender_host_name and config options to manage this, and basic check
692 PP/05 DSCP support for outbound connections and control modifier for inbound.
694 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
695 (Only plugin which currently uses this is kerberos4, which nobody should
696 be using, but we should make it available and other future plugins might
697 conceivably use it, even though it would break NAT; stuff *should* be
698 using channel bindings instead).
700 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
701 name; added for Sendmail compatibility; requires admin caller.
702 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
703 just ignore it); requires trusted caller.
704 Also parse but ignore: -Ac -Am -X<logfile>
707 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
709 TL/02 Add +smtp_confirmation as a default logging option.
711 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
712 Patch by Magnus Holmgren from 2007-02-20.
714 TL/04 Bugzilla 1281 - Spec typo.
715 Bugzilla 1283 - Spec typo.
716 Bugzilla 1290 - Spec grammar fixes.
718 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
720 TL/06 Add Experimental DMARC support using libopendmarc libraries.
722 TL/07 Fix an out of order global option causing a segfault. Reported to dev
723 mailing list by by Dmitry Isaikin.
725 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
727 JH/02 Support "G" suffix to numbers in ${if comparisons.
729 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
731 NM/01 Bugzilla 1197 - Spec typo
732 Bugzilla 1196 - Spec examples corrections
734 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
736 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
737 gnutls_enable_pkcs11, but renamed to more accurately indicate its
740 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
741 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
743 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
744 "acl {{name}{arg}...}", and optional args on acl condition
747 JH/05 Permit multiple router/transport headers_add/remove lines.
749 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
751 JH/07 Avoid using a waiting database for a single-message-only transport.
752 Performance patch from Paul Fisher. Bugzilla 1262.
754 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
757 JH/09 Add $headers_added variable, with content from use of ACL modifier
758 add_header (but not yet added to the message). Bugzilla 199.
760 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
761 Pulled from Bugzilla 817 by Wolfgang Breyha.
763 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
765 (nb: this is the same fix as in Exim 4.80.1)
767 JH/11 Add A= logging on delivery lines, and a client_set_id option on
770 JH/12 Add optional authenticated_sender logging to A= and a log_selector
773 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
775 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
776 advertise SMTP AUTH mechanism to us, instead of a generic
777 protocol violation error. Also, make Exim more robust to bad
778 data from the Dovecot auth socket.
780 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
782 When a queue runner is handling a message, Exim first routes the
783 recipient addresses, during which it prunes them based on the retry
784 hints database. After that it attempts to deliver the message to
785 any remaining recipients. It then updates the hints database using
788 So if a recipient address works intermittently, it can get repeatedly
789 deferred at routing time. The retry hints record remains fresh so the
790 address never reaches the final cutoff time.
792 This is a fairly common occurrence when a user is bumping up against
793 their storage quota. Exim had some logic in its local delivery code
794 to deal with this. However it did not apply to per-recipient defers
795 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
797 This change adds a proper retry rule check during routing so that the
798 final cutoff time is checked against the message's age. We only do
799 this check if there is an address retry record and there is not a
800 domain retry record; this implies that previous attempts to handle
801 the address had the retry_use_local_parts option turned on. We use
802 this as an approximation for the destination being like a local
803 delivery, as in LMTP.
805 I suspect this new check makes the old local delivery cutoff check
806 redundant, but I have not verified this so I left the code in place.
808 TF/02 Correct gecos expansion when From: is a prefix of the username.
810 Test 0254 submits a message to Exim with the header
814 When I ran the test suite under the user fanf2, Exim expanded
815 the header to contain my full name, whereas it should have added
816 a Resent-Sender: header. It erroneously treats any prefix of the
817 username as equal to the username.
819 This change corrects that bug.
821 GF/01 DCC debug and logging tidyup
822 Error conditions log to paniclog rather than rejectlog.
823 Debug lines prefixed by "DCC: " to remove any ambiguity.
825 TF/03 Avoid unnecessary rebuilds of lookup-related code.
827 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
828 Bug spotted by Jeremy Harris; was flawed since initial commit.
829 Would have resulted in OCSP responses post-SNI triggering an Exim
830 NULL dereference and crash.
832 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
834 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
835 Bug detection, analysis and fix by Samuel Thibault.
836 Bugzilla 1331, Debian bug #698092.
838 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
840 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
841 Server implementation by Todd Lyons, client by JH.
842 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
843 config variable "prdr_enable" controls whether the server
844 advertises the facility. If the client requests PRDR a new
845 acl_data_smtp_prdr ACL is called once for each recipient, after
846 the body content is received and before the acl_smtp_data ACL.
847 The client is controlled by bolth of: a hosts_try_prdr option
848 on the smtp transport, and the server advertisement.
849 Default client logging of deliveries and rejections involving
850 PRDR are flagged with the string "PRDR".
852 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
853 fclose(). Diagnosis by Todd Lyons.
855 PP/17 Update configure.default to handle IPv6 localhost better.
856 Patch by Alain Williams (plus minor tweaks).
859 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
860 This is now consistent with GnuTLS, and is now documented: the
861 previous undocumented portable approach to treating the option as
862 unset was to force an expansion failure. That still works, and
863 an empty string is now equivalent.
865 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
866 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
867 not performing validation itself.
869 PP/20 Added force_command boolean option to pipe transport.
870 Patch from Nick Koston, of cPanel Inc.
872 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
875 TF/04 Added udpsend ACL modifer and hexquote expansion operator
877 PP/21 Fix eximon continuous updating with timestamped log-files.
878 Broken in a format-string cleanup in 4.80, missed when I repaired the
879 other false fix of the same issue.
880 Report and fix from Heiko Schlichting.
883 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
884 Report from Prashanth Katuri.
886 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
887 It's SecureTransport, so affects any MacOS clients which use the
888 system-integrated TLS libraries, including email clients.
890 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
891 using a MIME ACL for non-SMTP local injection.
892 Report and assistance in diagnosis by Warren Baker.
894 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
896 JH/16 Fix comparisons for 64b. Bugzilla 1385.
898 TL/09 Add expansion variable $authenticated_fail_id to keep track of
899 last id that failed so it may be referenced in subsequent ACL's.
901 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
904 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
905 ldap library initialization, allowing self-signed CA's to be
906 used. Also properly sets require_cert option later in code by
907 using NULL (global ldap config) instead of ldap handle (per
908 session). Bug diagnosis and testing by alxgomz.
910 TL/12 Enhanced documentation in the ratelimit.pl script provided in
911 the src/util/ subdirectory.
913 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
914 renamed to Transport Post Delivery Action by Jeremy Harris, as
917 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
918 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
919 redis_servers = needs to be configured which will be used by the redis
920 lookup. Patch from Warren Baker, of The Packet Hub.
922 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
924 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
925 hostname or reverse DNS when processing a host list. Used suggestions
926 from multiple comments on this bug.
928 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
930 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
931 Missed a few lines, added it to make the runtest require no keyboard
934 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
935 contains upper case chars. Make router use caseful_local_part.
937 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
938 support when GnuTLS has been built with p11-kit.
944 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
946 This, or similar/improved, will also be change PP/11 of 4.82.
952 PP/01 Handle short writes when writing local log-files.
953 In practice, only affects FreeBSD (8 onwards).
954 Bugzilla 1053, with thanks to Dmitry Isaikin.
956 NM/01 Bugzilla 949 - Documentation tweak
958 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
961 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
963 PP/02 Implemented gsasl authenticator.
965 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
967 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
968 `pkg-config foo` for cflags/libs.
970 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
971 with rest of GSASL and with heimdal_gssapi.
973 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
974 `pkg-config foo` for cflags/libs for the TLS implementation.
976 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
977 properties get this fed in as external SSF. A number of robustness
978 and debugging improvements to the cyrus_sasl authenticator.
980 PP/08 cyrus_sasl server now expands the server_realm option.
982 PP/09 Bugzilla 1214 - Log authentication information in reject log.
983 Patch by Jeremy Harris.
985 PP/10 Added dbmjz lookup type.
987 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
989 PP/12 MAIL args handles TAB as well as SP, for better interop with
990 non-compliant senders.
991 Analysis and variant patch by Todd Lyons.
993 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
994 Bug report from Lars Müller <lars@samba.org> (via SUSE),
995 Patch from Dirk Mueller <dmueller@suse.com>
997 PP/13 tls_peerdn now print-escaped for spool files.
998 Observed some $tls_peerdn in wild which contained \n, which resulted
999 in spool file corruption.
1001 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1002 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1003 or write after TLS renegotiation, which otherwise led to messages
1006 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1007 as a tracking header (ie: a signed header comes before the signature).
1008 Patch from Wolfgang Breyha.
1010 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1011 comma-sep list; embedded commas doubled.
1013 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1015 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1017 Report and patch from Dmitry Banschikov.
1019 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1020 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1021 failures. We appear to now support TLS1.1+ with Exim.
1023 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1024 lets Exim select keys and certificates based upon TLS SNI from client.
1025 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1026 before an outbound SMTP session. New log_selector, +tls_sni.
1028 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1029 NULL dereference. Report and patch from Alun Jones.
1031 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1032 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1033 Not seeing resolver debug output on NetBSD, but suspect this is a
1034 resolver implementation change.
1036 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1037 Left warnings. Added "eximon gdb" invocation mode.
1039 PP/21 Defaulting "accept_8bitmime" to true, not false.
1041 PP/22 Added -bw for inetd wait mode support.
1043 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1044 locate the relevant includes and libraries. Made this the default.
1046 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1047 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1049 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1050 This may cause build issues on older platforms.
1052 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1053 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1054 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1055 Added SNI support via GnuTLS too.
1056 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1058 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1060 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1061 Applied second patch from Janne, implementing suggestion to default
1062 multiple-strings-in-record handling to match SPF spec.
1064 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1066 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1067 read-only, out of scope).
1068 Patch from Wolfgang Breyha, report from Stuart Northfield.
1070 PP/29 Fix three issues highlighted by clang analyser static analysis.
1071 Only crash-plausible issue would require the Cambridge-specific
1072 iplookup router and a misconfiguration.
1073 Report from Marcin Mirosław.
1075 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1077 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1078 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1079 As part of this, removing so much warning spew let me fix some minor
1080 real issues in debug logging.
1082 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1083 assignment on my part. Fixed.
1085 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1086 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1087 Janne Snabb (who went above and beyond: thank you).
1089 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1090 string otherwise requires a connection and a bunch more work and it's
1091 relatively easy to get wrong. Should also expose TLS library linkage
1094 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1095 64-bit ${eval} (JH/03).
1097 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1098 GNU libc to support some of the 64-bit stuff, should not lead to
1099 conflicts. Defined before os.h is pulled in, so if a given platform
1100 needs to override this, it can.
1102 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1103 protection layer was required, which is not implemented.
1104 Bugzilla 1254, patch from Wolfgang Breyha.
1106 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1107 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1108 tls_dhparam take prime identifiers. Also unbreak combination of
1109 OpenSSL+DH_params+TLSSNI.
1111 PP/39 Disable SSLv2 by default in OpenSSL support.
1117 PP/01 Solaris build fix for Oracle's LDAP libraries.
1118 Bugzilla 1109, patch from Stephen Usher.
1120 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1122 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1125 TF/02 Fix a couple more cases where we did not log the error message
1126 when unlink() failed. See also change 4.74-TF/03.
1128 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1129 lock up or crash if it happened to be inside a call to libc when it
1130 got a SIGUSR1 from exiwhat.
1132 The SIGUSR1 handler appends the current process status to the process
1133 log which is later printed by exiwhat. It used to use the general
1134 purpose logging code to do this, but several functions it calls are
1135 not safe for signals.
1137 The new output code in the SIGUSR1 handler is specific to the process
1138 log, and simple enough that it's easy to inspect for signal safety.
1139 Removing some special cases also simplifies the general logging code.
1140 Removing the spurious timestamps from the process log simplifies
1143 TF/04 Improved ratelimit ACL condition.
1145 The /noupdate option has been deprecated in favour of /readonly which
1146 has clearer semantics. The /leaky, /strict, and /readonly update modes
1147 are mutually exclusive. The update mode is no longer included in the
1148 database key; it just determines when the database is updated. (This
1149 means that when you upgrde Exim will forget old rate measurements.)
1151 Exim now checks that the per_* options are used with an update mode that
1152 makes sense for the current ACL. For example, when Exim is processing a
1153 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1154 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1155 must specify per_mail/readonly. If you omit the update mode it defaults to
1156 /leaky where that makes sense (as before) or /readonly where required.
1158 The /noupdate option is now undocumented but still supported for
1159 backwards compatibility. It is equivalent to /readonly except that in
1160 ACLs where /readonly is required you may specify /leaky/noupdate or
1161 /strict/noupdate which are treated the same as /readonly.
1163 A useful new feature is the /count= option. This is a generalization
1164 of the per_byte option, so that you can measure the throughput of other
1165 aggregate values. For example, the per_byte option is now equivalent
1166 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1168 The per_rcpt option has been generalized using the /count= mechanism
1169 (though it's more complicated than the per_byte equivalence). When it is
1170 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1171 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1172 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1173 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1174 Note that using per_rcpt with a non-readonly update mode in more than
1175 one ACL will cause the recipients to be double-counted. (The per_mail
1176 and per_byte options don't have this problem.)
1178 The handling of very low rates has changed slightly. If the computed rate
1179 is less than the event's count (usually one) then this event is the first
1180 after a long gap. In this case the rate is set to the same as this event's
1181 count, so that the first message of a spam run is counted properly.
1183 The major new feature is a mechanism for counting the rate of unique
1184 events. The new per_addr option counts the number of different
1185 recipients that someone has sent messages to in the last time period. It
1186 behaves like per_rcpt if all the recipient addresses are different, but
1187 duplicate recipient addresses do not increase the measured rate. Like
1188 the /count= option this is a general mechanism, so the per_addr option
1189 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1190 example, measure the rate that a client uses different sender addresses
1191 with the options per_mail/unique=$sender_address. There are further
1192 details in the main documentation.
1194 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1196 TF/06 Removed a few PCRE remnants.
1198 TF/07 Automatically extract Exim's version number from tags in the git
1199 repository when doing development or release builds.
1201 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1202 Bugzilla 879. Patch from Paul Fisher.
1204 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1205 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1208 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1210 PP/05 Variable $av_failed, true if the AV scanner deferred.
1211 Bugzilla 1078. Patch from John Horne.
1213 PP/06 Stop make process more reliably on build failure.
1214 Bugzilla 1087. Patch from Heiko Schlittermann.
1216 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1217 Bugzilla 1089. Patch from Heiko Schlittermann.
1219 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1220 Bugzilla 1131. Patch from Holger Weiß.
1222 PP/09 Handle IPv6 addresses with SPF.
1223 Bugzilla 860. Patch from Wolfgang Breyha.
1225 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1227 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1230 PP/11 match_* no longer expand right-hand-side by default.
1231 New compile-time build option, EXPAND_LISTMATCH_RHS.
1232 New expansion conditions, "inlist", "inlisti".
1234 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1236 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1238 PP/14 fix log_write() format string regression from TF/03.
1239 Bugzilla 1152. Patch from Dmitry Isaikin.
1245 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1247 PP/02 Harmonised TLS library version reporting; only show if debugging.
1248 Layout now matches that introduced for other libraries in 4.74 PP/03.
1250 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1252 PP/04 New "dns_use_edns0" global option.
1254 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1257 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1258 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1260 TK/01 Updated PolarSSL code to 0.14.2.
1261 Bugzilla 1097. Patch from Andreas Metzler.
1263 PP/07 Catch divide-by-zero in ${eval:...}.
1264 Fixes bugzilla 1102.
1266 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1269 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1270 format-string attack -- SECURITY: remote arbitrary code execution.
1272 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1273 time unintentionally subject to list matching rules, letting the header
1274 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1275 arbitrary string expansion). This allowed for information disclosure.
1277 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1278 INT_MIN/-1 -- value coerced to INT_MAX.
1284 NM/01 Workround for PCRE version dependency in version reporting
1287 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1288 This fixes portability to compilers other than gcc, notably
1289 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1291 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1292 makefiles for portability to HP-UX and POSIX correctness.
1294 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1295 Also via indented variable definition in the Makefile.
1296 (Debugging by Oliver Heesakkers).
1298 PP/02 Restore caching of spamd results with expanded spamd_address.
1299 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1301 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1302 Improves build reliability. Fix from: Frank Elsner
1304 NM/02 Fix wide character breakage in the rfc2047 coding
1305 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1307 NM/03 Allow underscore in dnslist lookups
1308 Fixes bug 1026. Patch from Graeme Fowler
1310 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1311 Code patches from Adam Ciarcinski of NetBSD.
1313 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1316 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1317 is logged, to avoid truncation. Patch from John Horne.
1319 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1320 Patch from Jakob Hirsch.
1322 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1323 SQL string expansion failure details.
1324 Patch from Andrey Oktyabrski.
1326 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1327 Patch from Simon Arlott.
1329 PP/09 New lookups functionality failed to compile on old gcc which rejects
1330 extern declarations in function scope.
1331 Patch from Oliver Fleischmann
1333 PP/10 Use sig_atomic_t for flags set from signal handlers.
1334 Check getgroups() return and improve debugging.
1335 Fixed developed for diagnosis in bug 927 (which turned out to be
1338 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1339 Patch from Mark Zealey.
1341 PP/12 Bugzilla 1056: Improved spamd server selection.
1342 Patch from Mark Zealey.
1344 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1345 Based on patch from Heiko Schlittermann.
1347 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1348 Patch from Uwe Doering, sign-off by Michael Haardt.
1350 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1351 variable declaration deep within a block. Bug and patch from
1354 PP/15 lookups-Makefile IRIX compatibilty coercion.
1356 PP/16 Make DISABLE_DKIM build knob functional.
1358 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1359 Patch by Simon Arlott
1361 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1362 variable argument macros. Our copy now differs from upstream.
1368 TF/01 Failure to get a lock on a hints database can have serious
1369 consequences so log it to the panic log.
1371 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1372 controlled using the smtp_confirmation log selector.
1374 TF/03 Include the error message when we fail to unlink a spool file.
1376 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1377 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1378 for maintaining out-of-tree patches for some time.
1380 PP/01 Bugzilla 139: Documentation and portability issues.
1381 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1382 Handle per-OS dynamic-module compilation flags.
1384 PP/02 Let /dev/null have normal permissions.
1385 The 4.73 fixes were a little too stringent and complained about the
1386 permissions on /dev/null. Exempt it from some checks.
1387 Reported by Andreas M. Kirchwitz.
1389 PP/03 Report version information for many libraries, including
1390 Exim version information for dynamically loaded libraries. Created
1391 version.h, now support a version extension string for distributors
1392 who patch heavily. Dynamic module ABI change.
1394 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1395 privilege escalation vulnerability whereby the Exim run-time user
1396 can cause root to append content of the attacker's choosing to
1399 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1402 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1403 If dropping privileges for untrusted macros, we disabled normal logging
1404 on the basis that it would fail; for the Exim run-time user, this is not
1405 the case, and it resulted in successful deliveries going unlogged.
1406 Fixed. Reported by Andreas Metzler.
1412 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1413 only prepend for the Resent-* case. Fixes regression introduced in
1414 Exim 4.70 by NM/22 for Bugzilla 607.
1416 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1417 increasing numbers of administrators be bitten by this.
1419 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1421 PP/03 Bugzilla 994: added openssl_options main configuration option.
1423 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1425 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1427 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1429 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1430 without a peer certificate, leading to a segfault because of an
1431 assumption that peers always have certificates. Be a little more
1432 paranoid. Problem reported by Martin Tscholak.
1434 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1435 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1436 NB: ClamAV planning to remove STREAM in "middle of 2010".
1437 CL also introduces -bmalware, various -d+acl logging additions and
1438 more caution in buffer sizes.
1440 PP/09 Implemented reverse_ip expansion operator.
1442 PP/10 Bugzilla 937: provide a "debug" ACL control.
1444 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1446 PP/12 Bugzilla 973: Implement --version.
1448 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1450 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1452 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1454 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1455 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1456 ignore trailing whitespace.
1458 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1460 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1463 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1464 Notification from Dr Andrew Aitchison.
1466 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1467 ExtendedDetectionInfo response format.
1468 Notification from John Horne.
1470 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1473 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1474 XSL and documented dependency on system catalogs, with examples of how
1477 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1480 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1481 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1482 configuration file which is writeable by the Exim user or group.
1484 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1485 of configuration files to cover files specified with the -C option if
1486 they are going to be used with root privileges, not just the default
1489 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1490 option (effectively making it always true).
1492 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1493 files to be used while preserving root privileges.
1495 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1496 that rogue child processes cannot use them.
1498 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1499 run-time user, instead of root.
1501 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1502 Exim run-time user without dropping privileges.
1504 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1505 result string, instead of calling string_vformat() twice with the same
1508 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1509 for other users. Others should always drop root privileges if they use
1510 -C on the command line, even for a whitelisted configure file.
1512 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1514 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1520 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1521 $data_path, and $header_path variables; fixed documentation bugs and
1524 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1525 exipick to access non-standard spools, including the "frozen" queue
1528 NM/01 Bugzilla 965: Support mysql stored procedures.
1529 Patch from Alain Williams
1531 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1533 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1534 Patch from Andreas Metzler
1536 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1537 Patch from Kirill Miazine
1539 NM/05 Bugzilla 671: Added umask to procmail example.
1541 JJ/03 installed exipick 20100323.0, fixing doc bug
1543 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1544 directory. Notification and patch from Dan Rosenberg.
1546 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1548 TK/02 Improve log output when DKIM signing operation fails.
1550 MH/01 Treat the transport option dkim_domain as a colon separated
1551 list, not as a single string, and sign the message with each element,
1552 omitting multiple occurences of the same signer.
1554 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1555 Bugzilla 985, 986. Patch by Simon Arlott
1557 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1558 Patch by Simon Arlott
1560 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1561 MBX locking. Notification from Dan Rosenberg.
1567 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1569 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1571 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1573 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1575 NM/04 Bugzilla 915: Flag broken perl installation during build.
1581 TK/01 Added patch by Johannes Berg that expands the main option
1582 "spamd_address" if it starts with a dollar sign.
1584 TK/02 Write list of recipients to X-Envelope-Sender header when building
1585 the mbox-format spool file for content scanning (suggested by Jakob
1588 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1589 (http://www.dcc-servers.net/) support via dccifd. Activated by
1590 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1592 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1593 by Mark Daniel Reidel <mr@df.eu>.
1595 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1596 When building exim an external PCRE library is now needed -
1597 PCRE is a system library on the majority of modern systems.
1598 See entry on PCRE_LIBS in EDITME file.
1600 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1601 conversation. Added nologin parameter to request.
1602 Patch contributed by Kirill Miazine.
1604 TF/01 Do not log submission mode rewrites if they do not change the address.
1606 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1608 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1609 log files in place. Contributed by Roberto Lima.
1611 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1613 TF/03 Bugzilla 615: When checking the local_parts router precondition
1614 after a local_part_suffix or local_part_prefix option, Exim now
1615 does not use the address's named list lookup cache, since this
1616 contains cached lookups for the whole local part.
1618 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1619 Robert Millan. Documentation is in experimental-spec.txt.
1621 TF/04 Bugzilla 668: Fix parallel build (make -j).
1623 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1625 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1626 Patch provided by Jan Srzednicki.
1628 TF/05 Leading white space used to be stripped from $spam_report which
1629 wrecked the formatting. Now it is preserved.
1631 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1632 that they are available at delivery time.
1634 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1636 TF/08 TLS error reporting now respects the incoming_interface and
1637 incoming_port log selectors.
1639 TF/09 Produce a more useful error message if an SMTP transport's hosts
1640 setting expands to an empty string.
1642 NM/06 Bugzilla 744: EXPN did not work under TLS.
1643 Patch provided by Phil Pennock.
1645 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1646 Patch provided by Richard Godbee.
1648 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1649 acl_smtp_notquit, added index entry.
1651 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1652 Patch provided by Eugene Bujak.
1654 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1655 accept(). Patch provided by Maxim Dounin.
1657 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1658 Patch provided by Phil Pennock.
1660 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1662 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1663 Patch provided by Brad "anomie" Jorsch.
1665 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1666 Patch provided by Dean Brooks.
1668 TK/05 Add native DKIM support (does not depend on external libraries).
1670 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1671 Patch provided by Graeme Fowler.
1673 NM/16 Bugzilla 851: Documentation example syntax fix.
1675 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1677 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1680 NM/19 Bugzilla 745: TLS version reporting.
1681 Patch provided by Phil Pennock.
1683 NM/20 Bugzilla 167: bool: condition support.
1684 Patch provided by Phil Pennock.
1686 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1687 clients. Patch provided by Phil Pennock.
1689 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1690 Patch provided by Brad "anomie" Jorsch.
1692 NM/23 Bugzilla 687: Fix misparses in eximstats.
1693 Patch provided by Heiko Schlittermann.
1695 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1696 Patch provided by Heiko Schlittermann.
1698 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1699 plus update to original patch.
1701 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1703 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1704 Patch provided by David Brownlee.
1706 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1708 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1710 NM/30 Bugzilla 888: TLS documentation bugfixes.
1712 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1714 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1715 Unlike the original bugzilla I have changed all shell scripts in src tree.
1717 NM/33 Bugzilla 898: Transport filter timeout fix.
1718 Patch by Todd Rinaldo.
1720 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1721 Patch by Serge Demonchaux.
1723 NM/35 Bugzilla 39: Base64 decode bug fixes.
1724 Patch by Jakob Hirsch.
1726 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1728 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1730 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1732 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1738 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1739 ALT-N's libdkim that I have put here:
1740 http://duncanthrax.net/exim-experimental/
1742 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1743 were called 'true' and it seems that C99 defines that as a reserved
1744 keyword to be used with 'bool' variable types. That means you could
1745 not include C99-style headers which use bools without triggering
1746 build errors in sieve.c.
1748 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1749 as mailq or other aliases. Changed the --help handling significantly
1750 to do whats expected. exim_usage() emits usage/help information.
1752 SC/01 Added the -bylocaldomain option to eximstats.
1754 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1756 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1758 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1764 PH/01 Another patch from the Sieve maintainer.
1766 PH/02 When an IPv6 address is converted to a string for single-key lookup
1767 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1768 dots are used instead of colons so that keys in lsearch files need not
1769 contain colons. This was done some time before quoting was made available
1770 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1771 (notated using the quote facility) so as to distinguish them from IPv4
1772 keys. This meant that lookups for IP addresses in host lists did not work
1773 for iplsearch lookups.
1775 This has been fixed by arranging for IPv6 addresses to be expressed with
1776 colons if the lookup type is iplsearch. This is not incompatible, because
1777 previously such lookups could never work.
1779 The situation is now rather anomolous, since one *can* have colons in
1780 ordinary lsearch keys. However, making the change in all cases is
1781 incompatible and would probably break a number of configurations.
1783 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1786 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1787 conversion specification without a maximum field width, thereby enabling
1788 a rogue spamd server to cause a buffer overflow. While nobody in their
1789 right mind would setup Exim to query an untrusted spamd server, an
1790 attacker that gains access to a server running spamd could potentially
1791 exploit this vulnerability to run arbitrary code as the Exim user.
1793 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1794 $primary_hostname instead of what libspf2 thinks the hosts name is.
1796 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1797 a directory entry by the name of the lookup key. Previously, if a
1798 symlink pointed to a non-existing file or a file in a directory that
1799 Exim lacked permissions to read, a lookup for a key matching that
1800 symlink would fail. Now it is enough that a matching directory entry
1801 exists, symlink or not. (Bugzilla 503.)
1803 PH/03 The body_linecount and body_zerocount variables are now exported in the
1806 PH/04 Added the $dnslist_matched variable.
1808 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1809 This means they are set thereafter only if the connection becomes
1812 PH/06 Added the client_condition to authenticators so that some can be skipped
1813 by clients under certain conditions.
1815 PH/07 The error message for a badly-placed control=no_multiline_responses left
1816 "_responses" off the end of the name.
1818 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1820 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1821 (without spaces) instead of just copying the configuration text.
1823 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1825 PH/11 Added $max_received_linelength.
1827 PH/12 Added +ignore_defer and +include_defer to host lists.
1829 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1830 way in which PCRE > 7.0 is built.
1832 PH/14 Implemented queue_only_load_latch.
1834 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1835 MAIL command. The effect was to mangle the value on 64-bit systems.
1837 PH/16 Another patch from the Sieve maintainer.
1839 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1841 PH/18 If a system quota error occurred while trying to create the file for
1842 a maildir delivery, the message "Mailbox is full" was not appended to the
1843 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1844 only to a quota excession during the actual writing of the file.
1846 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1847 characters?) which causes problems in log lines. The DN values are now
1848 passed through string_printing() before being added to log lines.
1850 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1851 and InterBase are left for another time.)
1853 PH/21 Added message_body_newlines option.
1855 PH/22 Guard against possible overflow in moan_check_errorcopy().
1857 PH/23 POSIX allows open() to be a macro; guard against that.
1859 PH/24 If the recipient of an error message contained an @ in the local part
1860 (suitably quoted, of course), incorrect values were put in $domain and
1861 $local_part during the evaluation of errors_copy.
1867 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1868 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1871 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1872 issue a MAIL command.
1874 PH/02 In an ACL statement such as
1876 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1878 if a client was not listed at all, or was listed with a value other than
1879 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1880 the condition was not true (as it should be), so access was not denied.
1881 The bug was that the ! inversion was incorrectly passed on to the second
1882 item. This has been fixed.
1884 PH/03 Added additional dnslists conditions == and =& which are different from
1885 = and & when the dns lookup returns more than one IP address.
1887 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1888 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1890 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1891 FSYNC, which compiles an option called disable_fsync that allows for
1892 bypassing fsync(). The documentation is heavily laced with warnings.
1894 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1896 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1897 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1898 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1899 including adding "make clean"; (3) Added -fPIC when compiling the test
1900 dynamically loaded module, to get rid of a warning.
1902 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1903 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1904 = 0s. The bug is otherwise harmless.
1906 PH/07 There was a bug in the dovecot authenticator such that the value of
1907 $auth1 could be overwritten, and so not correctly preserved, after a
1908 successful authentication. This usually meant that the value preserved by
1909 the server_setid option was incorrect.
1911 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1913 PH/09 Installed PCRE release 7.0.
1915 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1916 run for batched SMTP input. It is now run at the start of every message
1917 in the batch. While fixing this I discovered that the process information
1918 (output by running exiwhat) was not always getting set for -bs and -bS
1919 input. This is fixed, and it now also says "batched" for BSMTP.
1921 PH/11 Added control=no_pipelining.
1923 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1924 patch, slightly modified), and move the expansion of helo_data till after
1925 the connection is made in the smtp transport (so it can use these
1928 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1930 PH/14 Added log_selector = +pid.
1932 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1934 PH/16 Add ${if forany and ${if forall.
1936 PH/17 Added dsn_from option to vary the From: line in DSNs.
1938 PH/18 Flush SMTP output before performing a callout, unless control =
1939 no_callout_flush is set.
1941 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1942 was true (the default) a successful delivery failed to delete the retry
1943 item, thus causing premature timeout of the address. The bug is now
1946 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1948 PH/21 Long custom messages for fakedefer and fakereject are now split up
1949 into multiline reponses in the same way that messages for "deny" and
1950 other ACL rejections are.
1952 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1953 with slight modification.
1955 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1956 draft, changing the syntax and factoring some duplicate code".
1958 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1959 for deliveries of the second and subsequent messages over the same SMTP
1962 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1963 ${reduce, with only minor "tidies".
1965 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1967 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1968 expansion side effects.
1970 PH/27 When a message times out after an over-quota error from an Exim-imposed
1971 quota, the bounce message says "mailbox is full". This message was not
1972 being given when it was a system quota that was exceeded. It now should
1975 MH/03 Made $recipients available in local_scan(). local_scan() already has
1976 better access to the recipient list through recipients_list[], but
1977 $recipients can be useful in postmaster-provided expansion strings.
1979 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1980 in the case of a MAIL command with additional options following the
1981 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1982 were accidentally chopped off.
1984 PH/29 SMTP synchronization checks are implemented when a command is read -
1985 there is a check that no more input is waiting when there shouldn't be
1986 any. However, for some commands, a delay in an ACL can mean that it is
1987 some time before the response is written. In this time, more input might
1988 arrive, invalidly. So now there are extra checks after an ACL has run for
1989 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1990 pipelining has not been advertised.
1992 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1994 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1995 returned in the SMTP response when smtp_return_error_details was false.
1996 This has been fixed.
1998 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1999 instead of the C I/O that was originally supplied, because problems were
2000 reported on Solaris.
2002 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2003 Exim which did not show up earlier: it was assuming that a call to
2004 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2005 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2006 was a macro that became an assignment, so it seemed to work. This has
2007 changed to a proper function call with a void return, hence the compile
2008 error. Exim's code has been fixed.
2010 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2013 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2015 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2017 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2018 because it turns out to be too big or there is a timeout) there is no
2019 "Completed" line in the log. When some messages of this type were
2020 selected by exigrep, they were listed as "not completed". Others were
2021 picked up by some special patterns. I have improved the selection
2022 criteria to be more general.
2024 PH/38 The host_find_failed option in the manualroute router can now be set
2025 to "ignore", to completely ignore a host whose IP address cannot be
2026 found. If all hosts are ignored, the behaviour is controlled by the new
2027 host_all_ignored option.
2029 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2030 homing or because of multiple MX records with /mx) generated more than
2031 one IP address, and the following item turned out to be the local host,
2032 all the secondary addresses of the first item were incorrectly removed
2033 from the list, along with the local host and any following hosts (which
2034 is what is supposed to happen).
2036 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2037 whoever called Exim into the -H file. In the case of the daemon it was
2038 behaving confusingly. When first started, it used values for whoever
2039 started the daemon, but after a SIGHUP it used the Exim user (because it
2040 calls itself on a restart). I have changed the code so that it now always
2043 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2044 message are rejected with the same error (e.g. no authentication or bad
2045 sender address), and a DATA command is nevertheless sent (as can happen
2046 with PIPELINING or a stupid MUA), the error message that was given to the
2047 RCPT commands is included in the rejection of the DATA command. This is
2048 intended to be helpful for MUAs that show only the final error to their
2051 PH/42 Another patch from the Sieve maintainer.
2053 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2054 Eximstats - Fixed some broken HTML links and added missing column headers
2056 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2057 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2059 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2061 PH/43 Yet another patch from the Sieve maintainer.
2063 PH/44 I found a way to check for a TCP/IP connection going away before sending
2064 the response to the final '.' that terminates a message, but only in the
2065 case where the client has not sent further data following the '.'
2066 (unfortunately, this is allowed). However, in many cases there won't be
2067 any further data because there won't be any more messages to send. A call
2068 to select() can be used: if it shows that the input is "ready", there is
2069 either input waiting, or the socket has been closed. An attempt to read
2070 the next input character can distinguish the two cases. Previously, Exim
2071 would have sent an OK response which the client would never have see.
2072 This could lead to message repetition. This fix should cure that, at
2073 least in a lot of common cases.
2075 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2076 advertised in response to EHLO.
2082 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2083 fixed by 4.65/MH/01 (is this a record?) are fixed:
2085 (i) An empty string was always treated as zero by the numeric comparison
2086 operators. This behaviour has been restored.
2088 (ii) It is documented that the numeric comparison operators always treat
2089 their arguments as decimal numbers. This was broken in that numbers
2090 starting with 0 were being interpreted as octal.
2092 While fixing these problems I realized that there was another issue that
2093 hadn't been noticed. Values of message_size_limit (both the global option
2094 and the transport option) were treated as octal if they started with 0.
2095 The documentation was vague. These values are now always treated as
2096 decimal, and I will make that clear in the documentation.
2102 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2103 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2106 MH/01 Don't check that the operands of numeric comparison operators are
2107 integers when their expansion is in "skipping" mode (fixes bug
2108 introduced by 4.64-PH/07).
2110 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2111 child addresses, Exim now panics and dies. Previously, because the count
2112 is held in a short int, deliveries were likely to be lost. As such a
2113 large number of recipients for a single message is ridiculous
2114 (performance will be very, very poor), I have chosen to impose a limit
2115 rather than extend the field.
2121 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2122 leftover -K file (the existence of which was triggered by #402).
2123 While we were at it, introduced process PID as part of the -K
2124 filename. This should rule out race conditions when creating
2127 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2128 processing considerably. Previous code took too long for large mails,
2129 triggering a timeout which in turn triggers #401.
2131 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2132 in the DK code in transports.c. sendfile() is not really portable,
2133 hence the _LINUX specificness.
2135 TF/01 In the add_headers option to the mail command in an Exim filter,
2136 there was a bug that Exim would claim a syntax error in any
2137 header after the first one which had an odd number of characters
2140 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2141 callout verification, Exim cached a "reject" for the entire domain. This
2142 is correct for most verifications, but it is not correct for a recipient
2143 verification with use_sender or use_postmaster set, because in that case
2144 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2145 case of MAIL FROM:<> rejection from other early rejections (e.g.
2146 rejection of HELO). When verifying a recipient using a non-null MAIL
2147 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2148 Whatever the result of the callout, the value of the domain cache is
2149 left unchanged (for any other kind of callout, getting as far as trying
2150 RCPT means that the domain itself is ok).
2152 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2155 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2156 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2157 session, the daemon ignores ECONNECT errors and logs others; it now
2158 ignores EPIPE as well.
2160 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2161 (quoted-printable decoding).
2163 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2164 later the small subsequent patch to fix an introduced bug.
2166 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2168 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2170 PH/08 An error is now given if message_size_limit is specified negative.
2172 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2173 to be given (somewhat) arbitrary names.
2175 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2178 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2179 miscellaneous code fixes
2181 PH/10 Added the log_reject_target ACL modifier to specify where to log
2184 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2185 hostname. This is wrong, because it relates to the incoming message (and
2186 probably the interface on which it is arriving) and not to the outgoing
2187 callout (which could be using a different interface). This has been
2188 changed to use the value of the helo_data option from the smtp transport
2189 instead - this is what is used when a message is actually being sent. If
2190 there is no remote transport (possible with a router that sets up host
2191 addresses), $smtp_active_hostname is used.
2193 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2194 tweaks were necessary in order to get it to work (see also 21 below):
2195 (a) The code assumed that strncpy() returns a negative number on buffer
2196 overflow, which isn't the case. Replaced with Exim's string_format()
2198 (b) There were several signed/unsigned issues. I just did the minimum
2199 hacking in of casts. There is scope for a larger refactoring.
2200 (c) The code used strcasecmp() which is not a standard C function.
2201 Replaced with Exim's strcmpic() function.
2202 (d) The code set only $1; it now sets $auth1 as well.
2203 (e) A simple test gave the error "authentication client didn't specify
2204 service in request". It would seem that Dovecot has changed its
2205 interface. Fortunately there's a specification; I followed it and
2206 changed what the client sends and it appears to be working now.
2208 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2211 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2212 address A is aliased to B and C, where B exists and C does not. Without
2213 -v the output is "A verified" because verification stops after a
2214 successful redirection if more than one address is generated. However,
2215 with -v the child addresses are also verified. Exim was outputting "A
2216 failed to verify" and then showing the successful verification for C,
2217 with its parentage. It now outputs "B failed to verify", showing B's
2218 parentage before showing the successful verification of C.
2220 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2221 look up a TXT record in a specific list after matching in a combined
2224 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2225 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2226 they consult the DNS. I had assumed they would set it the way they
2227 wanted; and indeed my experiments on Linux seem to show that in some
2228 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2229 To be on the safe side, however, I have now made the interface to
2230 host_find_byname() similar to host_find_bydns(), with an argument
2231 containing the DNS resolver options. The host_find_byname() function now
2232 sets these options at its start, just as host_find_bydns() does. The smtp
2233 transport options dns_qualify_single and dns_search_parents are passed to
2234 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2235 of host_find_byname() use the default settings of RES_DEFNAMES
2236 (qualify_single) but not RES_DNSRCH (search_parents).
2238 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2239 spool_read_header() do less string testing, by means of a preliminary
2240 switch on the second character of optional "-foo" lines. (This is
2241 overdue, caused by the large number of possibilities that now exist.
2242 Originally there were few.) While I was there, I also converted the
2243 str(n)cmp tests so they don't re-test the leading "-" and the first
2244 character, in the hope this might squeeze out yet more improvement.
2246 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2247 flag allowing group syntax was set by the header_syntax check but not
2248 turned off, possible causing trouble later; (2) The flag was not being
2249 set at all for the header_verify test, causing "group"-style headers to
2250 be rejected. I have now set it in this case, and also caused header_
2251 verify to ignore an empty address taken from a group. While doing this, I
2252 came across some other cases where the code for allowing group syntax
2253 while scanning a header line wasn't quite right (mostly, not resetting
2254 the flag correctly in the right place). These bugs could have caused
2255 trouble for malformed header lines. I hope it is now all correct.
2257 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2258 with the "reply" argument non-NULL. The code, however (which originally
2259 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2260 but it didn't always do it. This confused somebody who was copying the
2261 code for some other use. I have removed all the tests.
2263 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2264 feature that was used to support insecure browsers during the U.S. crypto
2265 embargo. It requires special client support, and Exim is probably the
2266 only MTA that supported it -- and would never use it because real RSA is
2267 always available. This code has been removed, because it had the bad
2268 effect of slowing Exim down by computing (never used) parameters for the
2269 RSA_EXPORT functionality.
2271 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2272 authenticator to fail if there's a tab character in the incoming data
2273 (there should never be unless someone is messing about, as it's supposed
2274 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2275 if the connection is using TLS or if the remote IP is the same as the
2276 local IP, and the "valid-client-cert option" if a client certificate has
2279 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2280 authenticators. This can be used for authorization after authentication
2281 succeeds. (In the case of plaintext, it servers for both authentication
2284 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2285 if any retry times were supplied.
2287 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2288 connection, where there is no client IP address to check. In this
2289 situation, the verify now always succeeds.
2291 PH/25 Applied John Jetmore's -Mset patch.
2293 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2295 PH/27 In a string expansion for a processed (not raw) header when multiple
2296 headers of the same name were present, leading whitespace was being
2297 removed from all of them, but trailing whitespace was being removed only
2298 from the last one. Now trailing whitespace is removed from each header
2299 before concatenation. Completely empty headers in a concatenation (as
2300 before) are ignored.
2302 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2303 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2305 PH/29 [Removed. This was a change that I later backed out, and forgot to
2306 correct the ChangeLog entry (that I had efficiently created) before
2307 committing the later change.]
2309 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2310 address errors (4xx response to RCPT) over the same connection as other
2311 messages routed to the same hosts. Such deliveries are always "forced",
2312 so retry times are not inspected. This resulted in far too many retries
2313 for the affected addresses. The effect occurred only when there were more
2314 hosts than the hosts_max_try setting in the smtp transport when it had
2315 the 4xx errors. Those hosts that it had tried were not added to the list
2316 of hosts for which the message was waiting, so if all were tried, there
2317 was no problem. Two fixes have been applied:
2319 (i) If there are any address or message errors in an SMTP delivery, none
2320 of the hosts (tried or untried) are now added to the list of hosts
2321 for which the message is waiting, so the message should not be a
2322 candidate for sending over the same connection that was used for a
2323 successful delivery of some other message. This seems entirely
2324 reasonable: after all the message is NOT "waiting for some host".
2325 This is so "obvious" that I'm not sure why it wasn't done
2326 previously. Hope I haven't missed anything, but it can't do any
2327 harm, as the worst effect is to miss an optimization.
2329 (ii) If, despite (i), such a delivery is accidentally attempted, the
2330 routing retry time is respected, so at least it doesn't keep
2331 hammering the server.
2333 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2334 in ${readsocket because some servers need this prod.
2336 PH/32 Added some extra debug output when updating a wait-xxx database.
2338 PH/33 The hint "could be header name not terminated by colon", which has been
2339 given for certain expansion errors for a long time, was not being given
2340 for the ${if def:h_colon_omitted{... case.
2342 PH/34 The spec says: "With one important exception, whenever a domain list is
2343 being scanned, $domain contains the subject domain." There was at least
2344 one case where this was not true.
2346 PH/35 The error "getsockname() failed: connection reset by peer" was being
2347 written to the panic log as well as the main log, but it isn't really
2348 panic-worthy as it just means the connection died rather early on. I have
2349 removed the panic log writing for the ECONNRESET error when getsockname()
2352 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2353 runs only) independently of the message's sender address. This meant
2354 that, if the 4xx error was in fact related to the sender, a different
2355 message to the same recipient with a different sender could confuse
2356 things. In particualar, this can happen when sending to a greylisting
2357 server, but other circumstances could also provoke similar problems.
2358 I have changed the default so that the retry time for these errors is now
2359 based a combination of the sender and recipient addresses. This change
2360 can be overridden by setting address_retry_include_sender=false in the
2363 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2364 remote server are returned as part of bounce messages. This was not
2365 happening for LMTP over a pipe (the lmtp transport), but now it is the
2366 same for both kinds of LMTP.
2368 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2369 in header lines that were in fact CNAMEs. This is no longer the case.
2371 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2372 and queue runs started by the daemon processed all messages. This has
2373 been fixed so that -R and -S can now usefully be given with -q<time>.
2375 PH/40 Import PCRE release 6.7 (fixes some bugs).
2377 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2379 PH/42 Give an error if -q is specified more than once.
2381 PH/43 Renamed the variables $interface_address and $interface_port as
2382 $received_ip_address and $received_port, to make it clear that these
2383 values apply to message reception, and not to the outgoing interface when
2384 a message is delivered. (The old names remain recognized, of course.)
2386 PH/44 There was no timeout on the connect() call when using a Unix domain
2387 socket in the ${readsocket expansion. There now is.
2389 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2390 be meaningful with "accept".
2392 SC/01 Eximstats V1.43
2393 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2395 SC/02 Eximstats V1.44
2396 Use a glob alias rather than an array ref in the generated
2397 parser. This improves both readability and performance.
2399 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2400 Collect SpamAssassin and rejection statistics.
2401 Don't display local sender or destination tables unless
2402 there is data to show.
2403 Added average volumes into the top table text output.
2405 SC/04 Eximstats V1.46
2406 Collect data on the number of addresses (recipients)
2407 as well as the number of messages.
2409 SC/05 Eximstats V1.47
2410 Added 'Message too big' to the list of mail rejection
2411 reasons (thanks to Marco Gaiarin).
2413 SC/06 Eximstats V1.48
2414 Mainlog lines which have GMT offsets and are too short to
2415 have a flag are now skipped.
2417 SC/07 Eximstats V1.49 (Alain Williams)
2418 Added the -emptyok flag.
2420 SC/08 Eximstats V1.50
2421 Fixes for obtaining the IP address from reject messages.
2423 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2424 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2425 whitesspace changes from 4.64-PH/27
2427 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2430 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2431 are found, allow negative numbers in numeric criteria)
2433 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2435 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2436 to match changes made in 4.64-PH/43
2438 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2440 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2441 "domain" is actually an IP address, and force a failure. This locks out
2442 those revolvers/nameservers that support "A-for-A" lookups, in
2443 contravention of the specifications.
2445 PH/48 When a host name was looked up from an IP address, and the subsequent
2446 forward lookup of the name timed out, the host name was left in
2447 $sender_host_name, contrary to the specification.
2449 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2450 restricted to single-key lookups, Exim was not diagnosing an error if
2451 * or *@ was used with a query-style lookup.
2453 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2455 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2456 long ago, but noone interested enough thought of it. Let's just say that
2457 the "1.1" means that there are some new functions that weren't there at
2458 some point in the past.
2460 PH/51 Error processing for expansion failure of helo_data from an smtp
2461 transport during callout processing was broken.
2463 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2464 tested/used via the -bh/-bhc/-bs options.
2466 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2467 bug, fixed in subsequent PCRE releases).
2469 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2470 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2472 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2478 SC/01 Use a glob alias rather than an array ref in eximstats generated
2479 parser. This improves both readability and performance.
2481 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2482 Don't display local sender or destination tables in eximstats unless
2483 there is data to show.
2484 Added average volumes into the eximstats top table text output.
2486 SC/03 Collect data on the number of addresses (recipients) as well
2487 as the number of messages in eximstats.
2489 TF/01 Correct an error in the documentation for the redirect router. Exim
2490 does (usually) call initgroups() when daemonizing.
2492 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2493 with consistent privilege compared to when running as a daemon.
2495 TF/03 Note in the spec that $authenticated_id is not set for local
2496 submissions from trusted users.
2498 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2499 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2501 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2502 by adding some example configuration directives to the default
2503 configuration file. A little bit of work is required to uncomment the
2504 directives and define how usernames and passwords are checked, but
2505 there is now a framework to start from.
2507 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2508 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2509 without this. I don't know how relevant this is to other LDAP libraries.
2511 PH/02 Add the verb name to the "unknown ACL verb" error.
2513 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2515 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2517 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2518 directory not expanded when it should be if an expanded home directory
2519 was set for the address (which is overridden by the transport).
2521 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2524 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2525 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2526 because it is too late at that time, and has no effect.
2528 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2529 security issue with \' (bugzilla #107). I could not use the
2530 PQescapeStringConn() function, because it needs a PGconn value as one of
2533 PH/08 When testing addresses using -bt, indicate those final addresses that
2534 are duplicates that would not cause an additional delivery. At least one
2535 person was confused, thinking that -bt output corresponded to deliveries.
2536 (Suppressing duplicates isn't a good idea as you lose the information
2537 about possibly different redirections that led to the duplicates.)
2539 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2540 systems where poll() doesn't work, in particular OS X.
2542 PH/10 Added more information to debugging output for retry time not reached.
2544 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2545 operations in malware.c.
2547 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2550 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2551 syslog_duplication set false, log lines that would normally be written
2552 both the the main log and to the reject log were not written to syslog at
2555 PH/14 In the default configuration, change the use of "message" in ACL warn
2556 statements to "add_header".
2558 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2559 not followed by a command (e.g. "seen endif").
2561 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2562 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2565 PH/17 Added extra conditions to the default value of delay_warning_condition
2569 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2570 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2571 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2574 The Auto-Submitted: and various List- headers are standardised, whereas I
2575 don't think Precedence: ever was.
2577 PH/18 Refactored debugging code in route_finduser() to show more information,
2578 in particular, the error code if getpwnam() issues one.
2580 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2581 This is apparently needed in addition to the PH/07 change above to avoid
2582 any possible encoding problems.
2584 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2585 but not after initializing Perl.
2587 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2588 output them only if debugging. By default they are written stderr,
2589 apparently, which is not desirable.
2591 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2594 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2597 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2599 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2600 authenticated or an ident call has been made. Suppress the default
2601 values for $authenticated_id and $authenticated_sender (but permit -oMai
2602 and -oMas) when testing with -bh.
2604 PH/24 Re-jigged the order of the tests in the default configuration so that the
2605 tests for valid domains and recipients precede the DNS black list and CSA
2606 tests, on the grounds that those ones are more expensive.
2608 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2609 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2610 This bug exists in every version of Exim that I still have, right back to
2613 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2614 However, an attempt to turn on case-sensitivity in a regex key by
2615 including (?-i) didn't work because the subject string was already
2616 lowercased, and the effects were non-intuitive. It turns out that a
2617 one-line patch can be used to allow (?-i) to work as expected.
2623 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2624 other effects) broke the use of negated acl sub-conditions.
2626 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2629 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2630 "Deny" causes Exim to reject the incoming connection with a 554 error.
2631 Unfortunately, if there is a major crisis, such as a disk failure,
2632 tcp-wrappers gives "deny", whereas what one would like would be some
2633 kind of temporary error. A kludge has been added to help with this.
2634 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2635 554 error is used if errno is still zero or contains ENOENT (which occurs
2636 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2639 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2641 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2642 errors. Otherwise a message that provokes a temporary error (when other
2643 messages do not) can cause a whole host to time out.
2645 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2646 addresses were routed directly to files or pipes from a redirect router.
2647 File deliveries just didn't batch; pipe deliveries might have suffered
2650 PH/06 A failure to get a lock for a hints database would erroneously always say
2651 "Failed to get write lock", even when it was really a read lock.
2653 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2654 of 0600. This has been changed to use the value of the lockfile_mode
2655 option (which defaults to 0600).
2657 PH/08 Applied small patch from the Sieve maintainer.
2659 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2660 folder from quota calculations, a direct delivery into this folder messed
2661 up the contents of the maildirsize file. This was because the regex was
2662 used only to exclude .Trash (or whatever) when the size of the mailbox
2663 was calculated. There was no check that a delivery was happening into an
2664 excluded directory. This bug has been fixed by ignoring all quota
2665 processing for deliveries into excluded directories.
2667 PH/10 Added the maildirfolder_create_regex option to appendfile.
2673 PH/01 The code for finding all the local interface addresses on a FreeBSD
2674 system running IPv6 was broken. This may well have applied to all BSD
2675 systems, as well as to others that have similar system calls. The broken
2676 code found IPv4 interfaces correctly, but gave incorrect values for the
2677 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2678 that it would not match correctly against @[] and not recognize the IPv6
2681 PH/02 The ipliteral router was not recognizing addresses of the form user@
2682 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2684 PH/03 Added disable_ipv6.
2686 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2687 decoded form, because it is most often used to construct To: headers
2688 lines in autoreplies, and the decoded form may well be syntactically
2689 invalid. However, $reply_address has leading white space removed, and all
2690 newlines turned into spaces so that the autoreply transport does not
2693 PH/05 If group was specified without a user on a router, and no group or user
2694 was specified on a transport, the group from the router was ignored.
2696 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2697 for visible compile-time settings that can be used to change these
2698 numbers, for those that want even more. Backwards compatibility with old
2699 spool files has been maintained. However, going back to a previous Exim
2700 release will lost any variables that are in spool files.
2702 PH/07 Two small changes when running in the test harness: increase delay when
2703 passing a TCP/IP connection to a new process, in case the original
2704 process has to generate a bounce, and remove special handling of
2705 127.0.0.2 (sic), which is no longer necessary.
2707 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2708 be the same on different OS.
2710 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2713 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2714 whether --show-vars was specified or not
2716 JJ/02 exipick: Added support for new ACL variable spool format introduced
2719 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2720 syntactically invalid From: or Reply-to: line, and a filter used this to
2721 generate an autoreply, and therefore failed to obtain an address for the
2722 autoreply, Exim could try to deliver to a non-existent relative file
2723 name, causing unrelated and misleading errors. What now happens is that
2724 it logs this as a hard delivery error, but does not attempt to create a
2727 PH/11 The exinext utility has a -C option for testing purposes, but although
2728 the given file was scanned by exinext itself; it wasn't being passed on
2729 when Exim was called.
2731 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2732 an end-of-file indication when reading a command response.
2734 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2735 compiled. In many other places in Exim, IPv6 addresses are always
2736 recognized, so I have changed this. It also means that IPv4 domain
2737 literals of the form [IPV4:n.n.n.n] are now always recognized.
2739 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2740 used if the router is not running as root, for example, when verifying at
2741 ACL time, or when using -bh. The debugging output from this situation was
2742 non-existent - all you got was a failure to exec. I have made two
2745 (a) Failures to set uid/gid, the current directory, or a process leader
2746 in a subprocess such as that created by queryprogram now generate
2747 suitable debugging ouput when -d is set.
2749 (b) The queryprogram router detects when it is not running as root,
2750 outputs suitable debugging information if -d is set, and then runs
2751 the subprocess without attempting to change uid/gid.
2753 PH/15 Minor change to Makefile for building test_host (undocumented testing
2756 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2757 additional section of a DNS packet that returns MX or SRV records.
2758 Instead, it always explicitly searches for A/AAAA records. This avoids
2759 major problems that occur when a DNS server includes only records of one
2760 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2761 fixed another bug: if SRV records were looked up and the corresponding
2762 address records were *not* found in the additional section, the port
2763 values from the SRV records were lost.
2765 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2766 using the correct key (the original address) when searching the retry
2767 rules in order to find which one to use for generating the retry hint.
2769 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2770 adding the default one. Similarly, if it contains a Reply-To: header, the
2771 errors_reply_to option, if set, is not used.
2773 PH/19 When calculating a retry time, Exim used to measure the "time since
2774 failure" by looking at the "first failed" field in the retry record. Now
2775 it does not use this if it is later than than the arrival time of the
2776 message. Instead it uses the arrival time. This makes for better
2777 behaviour in cases where some deliveries succeed, thus re-setting the
2778 "first failed" field. An example is a quota failure for a huge message
2779 when small messages continue to be delivered. Without this change, the
2780 "time since failure" will always be short, possible causing more frequent
2781 delivery attempts for the huge message than are intended.
2782 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2784 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2785 $1, $2, $3) because the numerical variables can be reset during some
2786 expansion items (e.g. "match"), thereby losing the authentication data.
2788 PH/21 Make -bV show the size of off_t variables so that the test suite can
2789 decide whether to run tests for quotas > 2G.
2791 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2792 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2793 filecount value is greater than 2G or if a quota value is greater than 2G
2794 on a system where the size of off_t is not greater than 4, a panic error
2797 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2798 never match. The debug and -bh output now contains an explicit error
2799 message indicating a malformed IPv4 address or mask.
2801 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2802 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2803 PH/23 above applies.
2805 PH/25 Do not write to syslog when running in the test harness. The only
2806 occasion when this arises is a failure to open the main or panic logs
2807 (for which there is an explicit test).
2809 PH/26 Added the /no_tell option to "control=freeze".
2811 PH/27 If a host name lookup failed very early in a connection, for example, if
2812 the IP address matched host_lookup and the reverse lookup yielded a name
2813 that did not have a forward lookup, an error message of the form "no IP
2814 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2815 could be logged. Now it outputs the IP address instead of "NULL".
2817 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2818 allows the sender and the authenticated sender to be set when
2819 submitting a message from within Exim. Since child_open_exim() is
2820 documented for local_scan(), the new function should be too.
2822 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2823 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2824 results in an empty string is now treated as unset.
2826 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2828 PH/31 Added sender_verify_fail logging option.
2830 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2831 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2832 tidied the source and removed it altogether.
2834 PH/33 When a queue run was abandoned because the load average was too high, a
2835 log line was always written; now it is written only if the queue_run log
2836 selector is set. In addition, the log line for abandonment now contains
2837 information about the queue run such as the pid. This is always present
2838 in "start" and "stop" lines but was omitted from the "abandon" line.
2840 PH/34 Omit spaces between a header name and the colon in the error message that
2841 is given when verify = headers_syntax fails (if there are lots of them,
2842 the message gets confusing).
2844 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2845 names, as there are now some PTR records that contain slashes. This check
2846 is only to protect against broken name servers that fall over on strange
2847 characters, so the fact that it applies to all lookups doesn't matter.
2849 PH/36 Now that the new test suite is complete, we can remove some of the
2850 special code in Exim that was needed for the old test suite. For example,
2851 sorting DNS records because real resolvers return them in an arbitrary
2852 order. The new test suite's fake resolver always returns records in the
2855 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2856 bounces) except when queue_only is set, to avoid logging races between
2857 the different processes.
2859 PH/38 Panic-die if .include specifies a non-absolute path.
2861 PH/39 A tweak to the "H" retry rule from its user.
2863 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2864 a label. They prevented compilation on older perls.
2866 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2867 a warning to be raised on newish perls.
2869 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2870 on queue. Changes to match documented behaviour of showing count of
2871 messages matching specified criteria.
2873 PH/40 Changed the default ident timeout from 30s to 5s.
2875 PH/41 Added support for the use of login_cap features, on those BSD systems
2876 that have them, for controlling the resources used by pipe deliveries.
2878 PH/42 The content-scanning code uses fopen() to create files in which to put
2879 message data. Previously it was not paying any attention to the mode of
2880 the files. Exim runs with umask(0) because the rest of the code creates
2881 files with open(), and sets the required mode explicitly. Thus, these
2882 files were ending up world-writeable. This was not a big issue, because,
2883 being within the spool directory, they were not world-accessible. I have
2884 created a function called modefopen, which takes an additional mode
2885 argument. It sets umask(777), creates the file, chmods it to the required
2886 mode, then resets the umask. All the relevant calls to fopen() in the
2887 content scanning code have been changed to use this function.
2889 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2890 to 24 hours. This avoids potential overflow problems when processing G
2891 and H retry rules. I suspect nobody ever tinkers with this value.
2893 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2895 PH/45 When the plaintext authenticator is running as a client, the server's
2896 challenges are checked to ensure they are valid base64 strings. By
2897 default, the authentication attempt is cancelled if an invalid string is
2898 received. Setting client_ignore_invalid_base64 true ignores these errors.
2899 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2900 they are received. Thus, the responses can be made to depend on the
2901 challenges. If an invalid string is ignored, an empty string is placed in
2904 PH/46 Messages that are created by the autoreply transport now contains a
2905 References: header, in accordance with RFCs 2822 and 3834.
2907 PH/47 Added authenticated_sender_force to the smtp transport.
2909 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2911 PH/49 Installed latest patch from the Sieve maintainer.
2913 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2914 was also set, the appendfile transport was unnecessarily scanning a
2915 directory of message files (e.g. for maildir delivery) to find the count
2916 of files (along with the size), even though it did not need this
2917 information. It now does the scan only if it needs to find either the
2918 size of the count of files.
2920 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2922 PH/52 Two bugs concerned with error handling when the smtp transport is
2925 (i) Exim was not creating retry information for temporary errors given
2926 for individual recipients after the DATA command when the smtp transport
2927 was used in LMTP mode. This meant that they could be retried too
2928 frequently, and not timed out correctly.
2930 (ii) Exim was setting the flag that allows error details to be returned
2931 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2932 recipients that were returned after the DATA command.
2934 PH/53 This is related to PH/52, but is more general: for any failing address,
2935 when detailed error information was permitted to be returned to the
2936 sender, but the error was temporary, then after the final timeout, only
2937 "retry timeout exceeded" was returned. Now it returns the full error as
2938 well as "retry timeout exceeded".
2940 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2941 do this, and (what is worse) MTAs that accept it.
2943 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2944 will now be deprecated.
2946 PH/56 New os.c-cygwin from the Cygwin maintainer.
2948 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2949 formats (previously only available in exim formats via -bpr, -bpru,
2950 and -bpra. Now also available in native and exiqgrep formats)
2952 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2953 with very large, slow to parse queues
2955 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2957 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2959 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2960 responses to authentication challenges, though it was showing the
2961 challenges; (ii) I've removed the CR characters from the debug output for
2964 PH/58 Allow for the insertion of a newline as well as a space when a string
2965 is turned into more than one encoded-word during RFC 2047 encoding. The
2966 Sieve code now uses this.
2968 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2969 data_4xx, lost_connection, tls_required.
2971 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2972 message was being sent as an SMTP response.
2974 PH/61 Add -l and -k options to exicyclog.
2976 PH/62 When verifying, if an address was redirected to one new address, so that
2977 verification continued, and the new address failed or deferred after
2978 having set something in $address_data, the value of $address_data was not
2979 passed back to the ACL. This was different to the case when no
2980 redirection occurred. The value is now passed back in both cases.
2982 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2983 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2984 use login_cap.h, so on its own it isn't the distinguishing feature. The
2985 new name refers directly to the setclassresources() function.
2987 PH/65 Added configuration files for NetBSD3.
2989 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2991 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2992 is preferred over IPv4.
2994 PH/68 The bounce_return_message and bounce_return_body options were not being
2995 honoured for bounces generated during the reception of non-SMTP messages.
2996 In particular, this applied to messages rejected by the ACL. This bug has
2997 been fixed. However, if bounce_return_message is true and bounce_return_
2998 body is false, the headers that are returned for a non-SMTP message
2999 include only those that have been read before the error was detected.
3000 (In the case of an ACL rejection, they have all been read.)
3002 PH/69 The HTML version of the specification is now built in a directory called
3003 spec_html instead of spec.html, because the latter looks like a path with
3004 a MIME-type, and this confuses some software.
3006 PH/70 Catch two compiler warnings in sieve.c.
3008 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3009 function verify_get_ident() calls ip_connect() to connect a socket, but
3010 if the "connect()" function timed out, ip_connect() used to close the
3011 socket. However, verify_get_ident() also closes the socket later, and in
3012 between Exim writes to the log, which may get opened at this point. When
3013 the socket was closed in ip_connect(), the log could get the same file
3014 descriptor number as the socket. This naturally causes chaos. The fix is
3015 not to close the socket in ip_connect(); the socket should be closed by
3016 the function that creates it. There was only one place in the code where
3017 this was missing, in the iplookup router, which I don't think anybody now
3018 uses, but I've fixed it anyway.
3020 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3021 well as to direct DNS lookups. Otherwise the handling of names in host
3022 lists is inconsistent and therefore confusing.
3028 PH/01 Two changes to the default runtime configuration:
3030 (1) Move the checks for relay_from_hosts and authenticated clients from
3031 after to before the (commented out) DNS black list checks.
3033 (2) Add control=submission to the relay_from_hosts and authenticated
3034 clients checks, on the grounds that messages accepted by these
3035 statements are most likely to be submissions.
3037 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3039 (1) Generate an error if the third argument for the ${prvs expansion is
3042 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3045 (3) Reset the variables that are obtained from the first argument of
3046 ${prvscheck and used in the second argument before leaving the code,
3047 because their memory is reclaimed, so using them afterwards may do
3050 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3051 one (it's much easier than Tom thought :-).
3053 (5) Because of (4), we can now allow for the use of $prvscheck_result
3054 inside the third argument.
3056 PH/03 For some reason, the default setting of PATH when running a command from
3057 a pipe transport was just "/usr/bin". I have changed it to
3060 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3061 anything to be listed in the output from -bV.
3063 PH/05 When a filter generated an autoreply, the entire To: header line was
3064 quoted in the delivery log line, like this:
3066 => >A.N.Other <ano@some.domain> <original@ddress> ...
3068 This has been changed so that it extracts the operative address. There
3069 may be more than one such address. If so, they are comma-separated, like
3072 => >ano@some.domain,ona@other.domain <original@ddress> ...
3074 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3075 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3076 not being looked up in the rDNS to get a host name, Exim was showing the
3077 IP address twice in Received: lines, even though the IP addresses were
3078 identical. For example:
3080 Received: from [1.2.3.4] (helo=[1.2.3.4])
3082 However, if the real host name was known, it was omitting the HELO data
3083 if it matched the actual IP address. This has been tidied up so that it
3084 doesn't show the same IP address twice.
3086 PH/07 When both +timestamp and +memory debugging was on, the value given by
3087 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3088 called by the debug printing, thereby overwriting the timestamp buffer.
3089 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3091 PH/08 When the original message was included in an autoreply transport, it
3092 always said "this is a copy of the message, including all the headers",
3093 even if body_only or headers_only was set. It now gives an appropriate
3096 PH/09 Applied a patch from the Sieve maintainer which:
3098 o fixes some comments
3099 o adds the (disabled) notify extension core
3100 o adds some debug output for the result of if/elsif tests
3101 o points to the current vacation draft in the documentation
3102 and documents the missing references header update
3106 o fixes a bug in processing the envelope test (when testing
3107 multiple envelope elements, the last element determinted the
3110 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3111 Electronic Mail") by including:
3113 Auto-submitted: auto-generated
3115 in the messages that it generates (bounce messages and others, such as
3116 warnings). In the case of bounce messages for non-SMTP mesages, there was
3117 also a typo: it was using "Auto_submitted" (underscore instead of
3118 hyphen). Since every message generated by Exim is necessarily in response
3119 to another message, thes have all been changed to:
3121 Auto-Submitted: auto-replied
3123 in accordance with these statements in the RFC:
3125 The auto-replied keyword:
3127 - SHOULD be used on messages sent in direct response to another
3128 message by an automatic process,
3130 - MUST NOT be used on manually-generated messages,
3132 - MAY be used on Delivery Status Notifications (DSNs) and Message
3133 Disposition Notifications (MDNs),
3135 - MUST NOT be used on messages generated by automatic or periodic
3136 processes, except for messages which are automatic responses to
3139 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3140 to the default Received: header definition.
3142 PH/12 Added log selector acl_warn_skipped (default on).
3144 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3145 variables because (a) they are in the wrong storage pool and (b) even if
3146 they were copied, it wouldn't work properly because of the caching.
3148 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3149 checking when decoding. Apparently there are clients that generate
3150 overlong encoded strings. Why am I not surprised?
3152 PH/15 If the first argument of "${if match_address" was not empty, but did not
3153 contain an "@" character, Exim crashed. Now it writes a panic log message
3154 and treats the condition as false.
3156 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3158 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3159 extension "envelope-auth". The code is finished and in agreement with
3160 other implementations, but there is no documentation so far and in fact,
3161 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3162 not changing the active code.
3164 Print executed "if" and "elsif" statements when debugging is used. This
3165 helps a great deal to understand what a filter does.
3167 Document more things not specified clearly in RFC3028. I had all this
3168 sorted out, when out of a sudden new issues came to my mind. Oops."
3170 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3173 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3174 canonical form (as documented). However, after a host name lookup from
3175 the IP address, check_host() was doing a simple string comparison with
3176 addresses acquired from the DNS when checking that the found name did
3177 have the original IP as one of its addresses. Since any found IPv6
3178 addresses are likely to be in abbreviated form, the comparison could
3179 fail. Luckily, there already exists a function for doing the comparison
3180 by converting both addresses to binary, so now that is used instead of
3181 the text comparison.
3183 PH/20 There was another similar case to PH/19, when a complete host name was
3184 given in a host list; looking up its IP address could give an abbreviated
3185 form, whereas the current host's name might or might not be abbreviated.
3186 The same fix has been applied.
3192 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3193 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3196 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3197 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3199 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3201 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3202 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3203 Also, for "Auto-Submitted", treat anything other than "no" as
3204 non-personal, in accordance with RFC 3834. (Previously it treated
3205 anything starting "auto-" as non-personal.)
3207 TF/01 The control=submission/name=... option had a problem with syntax
3208 errors if the name included a slash character. The /name= option
3209 now slurps the rest of the string, so it can include any characters
3210 but it must come last in the list of options (after /sender_retain
3213 PH/05 Some modifications to the interface to the fake nameserver for the new
3221 TK/01 Added the "success_on_redirect" address verification option. See
3222 NewStuff for rationale and an example.
3224 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3226 PH/02 Patch to exigrep to allow it to work on syslog lines.
3228 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3229 fread() to skip over the body file's header line, because in Cygwin the
3230 header line is locked and is inaccessible.
3232 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3233 co-exist for some time) to make it clear that it is the Exim ID that is
3234 referenced, not the Message-ID: header line.
3236 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3237 string_format() function, because snprintf() does not exist on all
3240 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3241 ${stat: expansion item.
3243 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3244 protocol synchronization error", to keep the pedants happy.
3246 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3247 well as for IRIX systems, when gcc is being used. See the host.c source
3250 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3252 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3255 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3256 transport and to the smtp transport in LMTP mode.
3258 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3260 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3261 run a filter in a subprocess. This could lead to confusion in subsequent
3262 lookups in the parent process. There should also be a search_tidyup() at
3263 the end of the subprocess.
3265 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3266 only if the host matched helo_try_verify_hosts, which caused the
3267 verification to occur when the EHLO/HELO command was issued. The ACL just
3268 tested the remembered result. Now, if a previous verification attempt has
3269 not happened, "verify = helo" does it there and then.
3271 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3273 TK/03 Fix log output including CR from clamd.
3275 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3276 exist provoked a memory error which could cause a segfault.
3278 PH/15 Installed PCRE 6.2
3280 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3282 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3283 of the problem. Specifically, suggested +O2 rather than +O1 for the
3286 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3288 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3289 record, the port was not correctly logged when the outgoing_port log
3290 selector was set (it logged the transort's default port).
3292 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3293 fallback_hosts, and "hosts" in the smtp transport.
3295 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3296 host errors such as "Connection refused".
3298 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3299 authentication with radiusclient 0.4.9:
3301 - Error returned from rc_read_config was caught wrongly
3302 - Username/password not passed on to radius server due to wrong length.
3304 The presumption is that some radiusclient API changes for 4.51/PH/17
3305 were not taken care of correctly. The code is still untested by me (my
3306 Linux distribution still has 0.3.2 of radiusclient), but it was
3307 contributed by a Radius user.
3309 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3310 expanding the "port" option of the smtp transport.
3312 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3313 while reading a MIME header. Thanks to Tom Hughes for a patch.
3315 PH/24 Include config.h inside local_scan.h so that configuration settings are
3318 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3319 that in an ACL for RCPT (for example), you can examine exactly what was
3322 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3323 commands, but it was not correctly comparing the address with the actual
3324 client host address. Thus, it would show the EHLO address in Received:
3325 header lines when this was not necessary.
3327 PH/27 Added the % operator to ${eval:}.
3329 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3330 it should be ignoring failures (because with -C, for example, it has lost
3331 privilege). It wasn't ignoring creation failures other than "already
3334 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3337 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3338 because an input line was too long, either on its own, or by virtue of
3339 too many continuations, the temporary file was not being removed, and the
3340 return code was incorrect.
3342 PH/31 Missing "BOOL" in function definition in filtertest.c.
3344 PH/32 Applied Sieve patches from the maintainer.
3346 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3348 PH/33 Added "verify = not_blind".
3350 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3351 Local/Makefile (with some defaults set). These are used in built scripts
3352 such as exicyclog, but they have never been used in the exim_install
3353 script (though there are many overriding facilities there). I have
3354 arranged that the exim_install script now takes note of these two
3357 PH/35 Installed configuration files for Dragonfly.
3359 PH/36 When a locally submitted message by a trusted user did not contain a
3360 From: header, and the sender address was obtained from -f or from an SMTP
3361 MAIL command, and the trusted user did not use -F to supply a sender
3362 name, $originator_name was incorrectly used when constructing a From:
3363 header. Furthermore, $originator_name was used for submission mode
3364 messages from external hosts without From: headers in a similar way,
3365 which is clearly wrong.
3367 PH/37 Added control=suppress_local_fixups.
3369 PH/38 When log_selector = +received_sender was set, and the addition of the
3370 sender made the log line's construction buffer exactly full, or one byte
3371 less than full, an overflow happened when the terminating "\n" was
3374 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3375 when the result of a list match is failure because a DNS lookup failed.
3377 PH/40 RM_COMMAND is now used in the building process.
3379 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3380 the "build-* directories that it finds.
3382 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3383 address in a domain literal was a prefix of an interface address.
3385 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3386 when verifying a sender address, unless rewrite_headers is false.
3388 PH/44 Wrote a long comment about why errors_to addresses are verified as
3389 recipients, not senders.
3391 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3392 the ratelimit ACL was added.
3394 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3396 PH/46 Added extra information about PostgreSQL errors to the error string.
3398 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3399 avoiding the need to install special zones in a real server. This is
3400 backwards compatible; if it can't find the fake resolver, it drops back.
3401 Thus, both old and new test suites can be run.
3403 TF/02 Added util/ratelimit.pl
3405 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3406 clock is set back in time.
3408 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3409 Candler <B.Candler@pobox.com>.
3411 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3412 OK for addresses that are the result of redirections.
3414 PH/48 A number of further additions for the benefit of the new test suite,
3415 including a fake gethostbyname() that interfaces to the fake DNS resolver
3418 TF/06 The fix for widen_domains has also been applied to qualify_single and
3419 search_parents which are the other dnslookup options that can cause
3422 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3425 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3427 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3428 service" if it encountered an MX record with an empty target hostname.
3429 The message is now "an MX or SRV record indicated no SMTP service".
3431 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3432 if the DNS of the sending site is misconfigured. This is quite a
3433 common situation. This change restores the behaviour of treating a
3434 helo verification defer as a failure.
3436 PH/51 If self=fail was set on a router, the bounce message did not include the
3437 actual error message.
3443 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3445 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3446 command itself ended in error, the underlying message about the transport
3447 filter timeout was being overwritten with the pipe command error. Now the
3448 underlying error message should be appended to the second error message.
3450 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3452 PH/02 Reduce the amount of output that "make" produces by default. Full output
3453 can still be requested.
3455 PH/03 The warning log line about a condition test deferring for a "warn" verb
3456 was being output only once per connection, rather than after each
3457 occurrence (because it was using the same function as for successful
3458 "warn" verbs). This seems wrong, so I have changed it.
3460 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3461 it should not have, which might have caused a crash in the right
3462 circumstances, but probably never did.
3464 PH/04 Installed a modified version of Tony Finch's patch to make submission
3465 mode fix the return path as well as the Sender: header line, and to
3466 add a /name= option so that you can make the user's friendly name appear
3469 TF/03 Added the control = fakedefer ACL modifier.
3471 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3472 Mark Lowes for thorough testing.
3474 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3476 TK/03 Merged latest SRS patch from Miles Wilton.
3478 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3479 with the definition in sysexits.h (which is #included earlier).
3480 Fortunately, Exim does not actually use EX_OK. The code used to try to
3481 preserve the sysexits.h value, by assumimg that macro definitions were
3482 scanned for macro replacements. I have been disabused of this notion,
3483 so now the code just undefines EX_OK before #including unistd.h.
3485 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3486 in the smtp transport. When a block could not be written in a single
3487 write() function, the timeout was being re-applied to each part-write.
3488 This seems wrong - if the receiver was accepting one byte at a time it
3489 would take for ever. The timeout is now adjusted when this happens. It
3490 doesn't have to be particularly precise.
3492 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3493 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3495 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3496 without a domain if the check to <postmaster@domain> fails.
3498 SC/01 Eximstats: added -xls and the ability to specify output files
3499 (patch written by Frank Heydlauf).
3501 SC/02 Eximstats: use FileHandles for outputing results.
3503 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3505 SC/04 Eximstats: fixed display of large numbers with -nvr option
3507 SC/05 Eximstats: fixed merging of reports with empty tables.
3509 SC/06 Eximstats: added the -include_original_destination flag
3511 SC/07 Eximstats: removed tabs and trailing whitespace.
3513 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3515 TK/06 MBOX spool code: Add real "From " MBOX separator line
3516 so the .eml file is really in mbox format (even though
3517 most programs do not really care). Patch from Alex Miller.
3519 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3520 The latter is generated from $received_to and is only set if the
3521 message has one envelope recipient. SA can use these headers,
3522 obviously out-of-the-box. Patch from Alex Miller.
3524 PH/08 The ${def test on a variable was returning false if the variable's
3525 value was "0", contrary to what the specification has always said!
3526 The result should be true unless the variable is empty.
3528 PH/09 The syntax error of a character other than { following "${if
3529 def:variable_name" (after optional whitespace) was not being diagnosed.
3530 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3531 accidental colon was present, for example, could give incorrect results.
3533 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3534 result is used (not including appendfile, where other changes are about
3537 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3538 This involved changing a lot of size variables from int to off_t. It
3539 should work with maildirs and everything.
3541 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3542 spamd dying while we are connected to it.
3544 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3547 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3548 function for BDB 4.3.
3550 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3552 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3553 the sources for PCRE have been re-arranged and more files are now
3556 PH/15 The code I had for printing potentially long long variables in PH/11
3557 above was not the best (it lost precision). The length of off_t variables
3558 is now inspected at build time, and an appropriate printing format (%ld
3559 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3560 to be "long long int" or "long int". This is needed for the internal
3561 formatting function string_vformat().
3563 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3564 the configuration file to be ":syslog", then the script "guesses" where
3565 the logs files are, rather than using the compiled in default. In our
3566 case the guess is not the same as the compiled default, so the script
3567 suddenly stopped working when I started to use syslog. The patch checks
3568 to see if log_file_path is "". If so, it attempts to read it from exim
3569 with no configuration file to get the compiled in version, before it
3570 falls back to the previous guessing code."
3572 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3573 implementing BATV in an Exim configuration. See NewStuff for the gory
3576 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3577 Makefile that are specific to HP-UX.
3579 PH/18 If the "use_postmaster" option was set for a recipient callout together
3580 with the "random" option, the postmaster address was used as the MAIL
3581 FROM address for the random test, but not for the subsequent recipient
3582 test. It is now used for both.
3584 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3585 patch removes a few documentation additions to RFC 3028, because the
3586 latest draft now contains them. It adds the new en;ascii-case comparator
3587 and a new error check for 8bit text in MIME parts. Comparator and
3588 require names are now matched exactly. I enabled the subaddress
3589 extension, but it is not well tested yet (read: it works for me)."
3591 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3592 rework some of the code of TK/09 above to avoid the hardwired use of
3593 "%lld" and "long long". Replaced the call to snprintf() with a call to
3596 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3597 records point to non-existent hosts", "retry timeout exceeded", and
3598 "retry time not reached for any host after a long failure period".
3600 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3601 experimental DomainKeys support:
3603 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3604 (2) On an error such as an illegally used "control", the wrong name for
3605 the control was given.
3607 These problems did NOT occur unless DomainKeys support was compiled.
3609 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3611 PH/24 Added ${if match_ip condition.
3613 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3614 databases so that it will be absolutely obvious if a crash occurs in the
3615 DB library. This is a regular occurrence (often caused by mis-matched
3618 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3619 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3620 picked up on a user's system that detects such things. There doesn't seem
3621 to be a gcc warning option for this - only an attribute that has to be
3622 put on the function's prototype. It seems that in Fedora Core 4 they have
3623 set this on a number of new functions. No doubt there will be more in due
3626 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3627 specify a transport. However, if an address that was verified by such a
3628 router was the subject of a callout, Exim crashed because it tried to
3629 read the rcpt_include_affixes from the non-existent transport. Now it
3630 just assumes that the setting of that option is false. This bug was
3631 introduced by 4.51/PH/31.
3633 PH/28 Changed -d+all to exclude +memory, because that information is very
3634 rarely of interest, but it makes the output a lot bigger. People tend to
3635 do -d+all out of habit.
3637 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3638 code in os-type was giving problems when libc.so lives in lib64, like on
3641 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3642 aren't the modern standard, and it seems that some systems' include files
3643 don't always have them. Exim was already checking for some of the newer
3644 ones like T_AAAA, and defining it itself. I've added checks for all the
3645 record types that Exim uses.
3647 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3648 not automatically generating a new one, as it is supposed to. This
3649 prevented TLS from working. If the file did exist, but contained invalid
3650 data, a new version was generated, as expected. It was only the case of a
3651 non-existent file that was broken.
3653 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3654 with a change in libdomainkeys > 0.64.
3656 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3657 from DNS. If the selector record carries the flag, it now has
3658 precedence over the domain-wide flag.
3660 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3662 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3663 the use of a port name that isn't defined in /etc/services) occurred, the
3664 message was deferred as in a normal delivery, and thus remained on the
3665 spool, instead of being failed because of the mua_wrapper setting. This
3666 is now fixed, and I tidied up some of the mua_wrapper messages at the
3669 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3670 the messages in a hash of arrays rather than using individual hashes.
3671 This is a bit cleaner and results in dramatic memory savings, albeit
3672 at a slight CPU cost.
3674 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3675 as requested by Marc Sherman.
3677 SC/10 Eximstats: added histograms for user specified patterns as requested
3680 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3682 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3683 fopen() in the content-scanning modules that did not already have it.
3689 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3690 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3692 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3694 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3696 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3697 to test Sieve filters that use "vacation".
3699 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3700 that changes the way the GnuTLS parameters are stored in the cache file.
3701 The new format can be generated externally. For backward compatibility,
3702 if the data in the cache doesn't make sense, Exim assumes it has read an
3703 old-format file, and it generates new data and writes a new file. This
3704 means that you can't go back to an older release without removing the
3707 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3708 work if there are any delivery delays because "one_time" forces the
3709 parent to be marked "delivered", so its unseen clone is never tried
3710 again. For this reason, Exim now forbids the simultaneous setting of
3713 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3714 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3715 entry if you want to know the details. The fix, however, neglected to
3716 consider the case when local delivery batching is involved. The test for
3717 "previously delivered" was not happening when checking to see if an
3718 address could be batched with a previous (undelivered) one; under
3719 certain circumstances this could lead to multiple deliveries to the same
3722 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3723 in its include files, and this causes problems building Exim.
3725 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3726 header_syntax) but Exim was just ignoring anything given after a slash.
3727 In particular, this caused confusion with an attempt to use "verify =
3728 reverse_host_lookup/defer_ok". An error is now given when options are
3729 supplied for verify items that do not have them. (Maybe reverse_host_
3730 lookup should have a defer_ok option, but that's a different point.)
3732 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3733 defined by RFC 821) to 2048, because there were problems with some AUTH
3734 commands, and RFC 1869 says the size should be increased for extended
3735 SMTP commands that take arguments.
3737 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3740 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3741 "unknown" error; now it says that the functionality isn't in the binary.
3743 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3744 an address' error message when a string expansion fails (syntax or
3745 whatever). Otherwise the password may appear in the log. Following change
3746 PH/42 below, there is no longer a chance of it appearing in a bounce
3749 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3751 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3752 of its addresses was ever tried. (Bugzilla bug #2.)
3754 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3755 the result incorrectly in the debug output. (It correctly added a newline
3756 to what was transported.)
3758 TF/01 Added $received_time.
3760 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3761 commented out examples of how to interface to a virus scanner and to
3762 SpamAssassin. Also added commented examples of av_scanner and
3763 spamd_address settings.
3765 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3766 and controls are allowed in which ACLs. There were a couple of minor
3767 errors. Some of the entries in the conditions table (which is a table of
3768 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3769 negation of where the condition IS allowed.
3771 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3773 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3774 header file does not have a version number, so I've had to invent a new
3775 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3776 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3777 radiusclient), but it was contributed by a Radius user.
3779 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3780 files or directories, for OpenSSL.
3782 PH/19 When an Exim process that is running as root has to create an Exim log
3783 file, it does so in a subprocess that runs as exim:exim so as to get the
3784 ownership right at creation (otherwise, other Exim processes might see
3785 the file with the wrong ownership). There was no test for failure of this
3786 fork() call, which would lead to the process getting stuck as it waited
3787 for a non-existent subprocess. Forks do occasionally fail when resources
3788 run out. I reviewed all the other calls to fork(); they all seem to check
3791 PH/20 When checking for unexpected SMTP input at connect time (before writing
3792 the banner), Exim was not dealing correctly with a non-positive return
3793 from the read() function. If the client had disconnected by this time,
3794 the result was a log entry for a synchronization error with an empty
3795 string after "input=" when read() returned zero. If read() returned -1
3796 (an event I could not check), uninitialized data bytes were printed.
3797 There were reports of junk text (parts of files, etc) appearing after
3800 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3802 PH/22 Added support for macro redefinition, and (re)definition in between
3803 driver and ACL definitions.
3805 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3806 forgetting to use the resulting value; it was using the unexpanded value.
3808 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3809 hadn't been configured. The fix is from Juergen Kreileder, who
3810 understands it better than I do:
3812 "Here's what I see happening with three configured cyrus_sasl
3813 authenticators configured (plain, login, cram-md5):
3815 On startup auth_cyrus_sasl_init() gets called for each of these.
3816 This means three calls to sasl_listmech() without a specified mech_list.
3817 => SASL tests which mechs of all available mechs actually work
3818 => three warnings about OTP not working
3819 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3821 With the patch, sasl_listmech() also gets called three times. But now
3822 SASL's mech_list option is set to the server_mech specified in the the
3823 authenticator. Or in other words, the answer from sasl_listmech()
3824 gets limited to just the mech you're testing for (which is different
3826 => the return list contains just 'plain' or 'login', 'cram-md5' or
3827 nothing depending on the value of ob->server_mech.
3829 I've just tested the patch: Authentication still works fine,
3830 unavailable mechs specified in the exim configuration are still
3831 caught, and the auth.log warnings about OTP are gone."
3833 PH/25 When debugging is enabled, the contents of the command line are added
3834 to the debugging output, even when log_selector=+arguments is not
3837 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3838 answer is "GNU", and only if the return is "GNU/something" is the answer
3841 PH/27 $acl_verify_message is now set immediately after the failure of a
3842 verification in an ACL, and so is available in subsequent modifiers. In
3843 particular, the message can be preserved by coding like this:
3845 warn !verify = sender
3846 set acl_m0 = $acl_verify_message
3848 Previously, $acl_verify_message was set only while expanding "message"
3849 and "log_message" when a very denied access.
3851 PH/28 Modified OS/os.c-Linux with
3853 -#ifndef OS_LOAD_AVERAGE
3854 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3856 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3857 nomenclature these days.)
3859 PH/29 Installed patch from the Sieve maintainer that adds the options
3860 sieve_useraddress and sieve_subaddress to the redirect router.
3862 PH/30 In these circumstances:
3863 . Two addresses routed to the same list of hosts;
3864 . First host does not offer TLS;
3865 . First host accepts first address;
3866 . First host gives temporary error to second address;
3867 . Second host offers TLS and a TLS session is established;
3868 . Second host accepts second address.
3869 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3870 and peerdn, if requested) that were in fact used only for the second
3873 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3874 attention to any local part prefix or suffix that was matched by the
3875 router that accepted the address. It now behaves in the same way as it
3876 does for delivery: the affixes are removed from the local part unless
3877 rcpt_include_affixes is set on the transport.
3879 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3880 timeout during the DATA phase of an incoming message.
3882 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3883 applied a patch sanctioned by the Sieve maintainer.
3885 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3886 the uid or gid is negative. A case of a negative gid caused this to be
3887 noticed. The fix allows for either to be negative.
3889 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3890 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3893 PH/36 The change PH/12 above was broken. Fixed it.
3895 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3896 the grounds that routing the same address twice would always produce the
3897 same answer. This might have been true once, but it is certainly no
3898 longer true now. Routing a child address may depend on the previous
3899 routing that produced that child. Some complicated redirection strategies
3900 went wrong when messages had multiple recipients, and made Exim's
3901 behaviour dependent on the order in which the addresses were given.
3903 I have moved the duplicate checking until after the routing is complete.
3904 Exim scans the addresses that are assigned to local and remote
3905 transports, and removes any duplicates. This means that more work will be
3906 done, as duplicates will always all be routed, but duplicates are
3907 presumably rare, so I don't expect this is of any significance.
3909 For deliveries to pipes, files, and autoreplies, the duplicate checking
3910 still happens during the routing process, since they are not going to be
3913 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3914 It corrects a timeout issue with spamd. This is Ian's comment: "The
3915 background is that sometimes spamd either never reads data from a
3916 connection it has accepted, or it never writes response data. The exiscan
3917 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3918 blindly assumes that writes won't block so it may never time out."
3920 PH/39 Allow G after quota size as well as K and M.
3922 PH/40 The value set for $authenticated_id in an authenticator may not contain
3923 binary zeroes or newlines because the value is written to log lines and
3924 to spool files. There was no check on this. Now the value is run through
3925 the string_printing() function so that such characters are converted to
3926 printable escape sequences.
3928 PH/41 $message_linecount is a new variable that contains the total number of
3929 lines in the message. Compare $body_linecount, which is the count for the
3932 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3933 bounce and delay warning messages, except in certain special cases, which
3936 (a) An SMTP error message from a remote host;
3937 (b) A message specified in a :fail: redirection;
3938 (c) A message specified in a "fail" command in a system filter;
3939 (d) A message specified in a FAIL return from the queryprogram router;
3940 (e) A message specified by the cannot_route_message router option.
3942 In these cases only, Exim does include the error details in bounce and
3943 warning messages. There are also a few cases where bland messages such
3944 as "unrouteable address" or "local delivery error" are given.
3946 PH/43 $value is now also set for the "else" part of a ${run expansion.
3948 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3949 being worked on, but at least Exim now implements the latest version to
3952 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3953 process to complete was treated as a delivery failure, a timeout while
3954 writing the message to the pipe was logged, but erroneously treated as a
3955 successful delivery. Such timeouts include transport filter timeouts. For
3956 consistency with the overall process timeout, these timeouts are now
3957 treated as errors, giving rise to delivery failures by default. However,
3958 there is now a new Boolean option for the pipe transport called
3959 timeout_defer, which, if set TRUE, converts the failures into defers for
3960 both kinds of timeout. A transport filter timeout is now identified in
3963 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3964 systems where "make" and "gmake" are different, calling "gmake" at top
3965 level broke things. I've arranged for the value of $(MAKE) to be passed
3966 from the Makefile to this script so that it can call the same version of
3970 A note about Exim versions 4.44 and 4.50
3971 ----------------------------------------
3973 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3974 changes of various kinds. As a consequence, a big documentation update was
3975 needed. This delayed the release for rather longer than seemed good, especially
3976 in the light of a couple of (minor) security issues. Therefore, the changes
3977 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3978 release. So 4.44 and 4.50 are in effect two different branches that both start
3981 I have left the 4.50 change log unchanged; it contains all the changes since
3982 4.43. The change log for 4.44 is below; many of its items are identical to
3983 those for 4.50. This seems to be the most sensible way to preserve the
3984 historical information.
3990 1. Minor wording change to the doc/README.SIEVE file.
3992 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3993 computation of the current number of files was incorrect.
3995 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3996 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3999 4. Give more explanation in the error message when the command for a transport
4000 filter fails to execute.
4002 5. There are several places where Exim runs a non-Exim command in a
4003 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4004 was being done only for the command run by the queryprogram router. It is
4005 now done for all such subprocesses. The other cases are: ${run, transport
4006 filters, and the commands run by the lmtp and pipe transports.
4008 6. Added CONFIGURE_GROUP build-time option.
4010 7. Some older OS have a limit of 256 on the maximum number of file
4011 descriptors. Exim was using setrlimit() to set 1000 as a large value
4012 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4013 systems. I've change it so that if it can't get 1000, it tries for 256.
4015 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4016 was an oversight, and furthermore, ever since the addition of extra
4017 controls (e.g. 4.43/32), the checks on when to allow different forms of
4018 "control" were broken. There should now be diagnostics for all cases when a
4019 control that does not make sense is encountered.
4021 9. Added the /retain_sender option to "control=submission".
4023 10. $recipients is now available in the predata ACL (oversight).
4025 11. Tidy the search cache before the fork to do a delivery from a message
4026 received from the command line. Otherwise the child will trigger a lookup
4027 failure and thereby defer the delivery if it tries to use (for example) a
4028 cached ldap connection that the parent has called unbind on.
4030 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4031 of $address_data from the recipient verification was clobbered by the
4032 sender verification.
4034 13. The value of address_data from a sender verification is now available in
4035 $sender_address_data in subsequent conditions in the ACL statement.
4037 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4039 15. Added a new option "connect=<time>" to callout options, to set a different
4042 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4043 was its contents. (It was OK if the option was not defined at all.)
4045 17. A "Completed" log line is now written for messages that are removed from
4046 the spool by the -Mrm option.
4048 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4049 information about exactly what failed.
4051 19. Added -dd to debug only the daemon process.
4053 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4054 handles timeouts, both on the server side and network timeouts. Renamed the
4055 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4057 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4058 It is now set to "smtps".
4060 22. $host_address is now set to the target address during the checking of
4061 ignore_target_hosts.
4063 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4064 being passed; this would have caused $sender_host_name to have been used if
4065 matching the list had actually called for a host name (not very likely,
4066 since this list is usually IP addresses). A host name is now passed as
4069 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4070 code that specifies a non-restarting handler (typically sigaction() in
4071 modern systems) in an attempt to fix a rare and obscure crash bug.
4073 25. Narrowed the window for a race in the daemon that could cause it to ignore
4074 SIGCHLD signals. This is not a major problem, because they are used only to
4075 wake it up if nothing else does.
4077 26. A malformed maildirsize file could cause Exim to calculate negative values
4078 for the mailbox size or file count. Odd effects could occur as a result.
4079 The maildirsize information is now recalculated if the size or filecount
4082 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4083 support for a long time. Removed HAVE_SYS_VFS_H.
4085 28. Installed the latest version of exipick from John Jetmore.
4087 29. In an address list, if the pattern was not a regular expression, an empty
4088 subject address (from a bounce message) matched only if the pattern was an
4089 empty string. Non-empty patterns were not even tested. This was the wrong
4090 because it is perfectly reasonable to use an empty address as part of a
4091 database query. An empty address is now tested by patterns that are
4092 lookups. However, all the other forms of pattern expect the subject to
4093 contain a local part and a domain, and therefore, for them, an empty
4094 address still always fails if the pattern is not itself empty.
4096 30. Exim went into a mad DNS loop when attempting to do a callout where the
4097 host was specified on an smtp transport, and looking it up yielded more
4098 than one IP address.
4100 31. Re-factored the code for checking spool and log partition space into a
4101 function that finds that data and another that does the check. The former
4102 is then used to implement four new variables: $spool_space, $log_space,
4103 $spool_inodes, and $log_inodes.
4105 32. The RFC2047 encoding function was originally intended for short strings
4106 such as real names; it was not keeping to the 75-character limit for
4107 encoded words that the RFC imposes. It now respects the limit, and
4108 generates multiple encoded words if necessary. To be on the safe side, I
4109 have increased the buffer size for the ${rfc2047: expansion operator from
4112 33. It is now permitted to omit both strings after an "if" condition; if the
4113 condition is true, the result is "true". As before, when the second string
4114 is omitted, a false condition yields an empty string. This makes it less
4115 cumbersome to write custom ACL and router conditions.
4117 34. Failure to deliver a bounce message always caused it to be frozen, even if
4118 there was an errors_to setting on the router. The errors_to setting is now
4121 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4122 canonical form (fully expanded) before being placed in
4123 $sender_host_address.
4125 36. The table in the code that translates DNS record types into text (T_A to
4126 "A" for instance) was missing entries for NS and CNAME. It is just possible
4127 that this could have caused confusion if both these types were looked up
4128 for the same domain, because the text type is used as part of Exim's
4129 per-process caching. But the chance of anyone hitting this buglet seems
4132 37. The dnsdb lookup has been extended in a number of ways.
4134 (1) There is a new type, "zns", which walks up the domain tree until it
4135 finds some nameserver records. It should be used with care.
4137 (2) There is a new type, "mxh", which is like "mx" except that it returns
4138 just the host names, not the priorities.
4140 (3) It is now possible to give a list of domains (or IP addresses) to be
4141 looked up. The behaviour when one of the lookups defers can be
4142 controlled by a keyword.
4144 (4) It is now possible to specify the separator character for use when
4145 multiple records are returned.
4147 38. The dnslists ACL condition has been extended: it is now possible to supply
4148 a list of IP addresses and/or domains to be looked up in a particular DNS
4151 39. Added log_selector=+queue_time_overall.
4153 40. When running the queue in the test harness, wait just a tad after forking a
4154 delivery process, to get repeatability of debugging output.
4156 41. Include certificate and key file names in error message when GnuTLS fails
4157 to set them up, because the GnuTLS error message doesn't include the name
4158 of the failing file when there is a problem reading it.
4160 42. Allow both -bf and -bF in the same test run.
4162 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4164 44. The "Exiscan patch" is now merged into the mainline Exim source.
4166 45. Sometimes the final signoff response after QUIT could fail to get
4167 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4168 < 0 before doing a fflush(). This bug looks as though it goes back to the
4169 introduction of TLS in release 3.20, but "sometimes" must have been rare
4170 because the tests only now provoked it.
4172 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4173 (this can affect the format of dates).
4175 47. exim_tidydb, when checking for the continued existence of a message for
4176 which it has found a message-specific retry record, was not finding
4177 messages that were in split spool directories. Consequently, it was
4178 deleting retry records that should have stayed in existence.
4180 48. Steve fixed some bugs in eximstats.
4182 49. The SPA authentication driver was not abandoning authentication and moving
4183 on to the next authenticator when an expansion was forced to fail,
4184 contradicting the general specification for all authenticators. Instead it
4185 was generating a temporary error. It now behaves as specified.
4187 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4188 (the order specifies the preference for clients). The order is now AES256,
4189 AES128, 3DES, ARCFOUR128.
4191 51. Small patch to Sieve code - explicitly set From: when generating an
4194 52. Exim crashed if a remote delivery caused a very long error message to be
4195 recorded - for instance if somebody sent an entire SpamAssassin report back
4196 as a large number of 550 error lines. This bug was coincidentally fixed by
4197 increasing the size of one of Exim's internal buffers (big_buffer) that
4198 happened as part of the Exiscan merge. However, to be on the safe side, I
4199 have made the code more robust (and fixed the comments that describe what
4202 53. Now that there can be additional text after "Completed" in log lines (if
4203 the queue_time_overall log selector is set), a one-byte patch to exigrep
4204 was needed to allow it to recognize "Completed" as not the last thing in
4207 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4208 patch that reportedly fixes this has been added. I am not expert enough to
4209 create a test for it. This is what the patch creator wrote:
4211 "I found a little strange behaviour of ldap code when working with
4212 Windows 2003 AD Domain, where users was placed in more than one
4213 Organization Units. When I tried to give exim partial DN, the exit code
4214 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4215 But simultaneously result of request was absolutely normal ldap result,
4216 so I produce this patch..."
4218 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4219 REFERENCE, so I have modified the code to exclude the patch when that macro
4222 55. Some experimental protocols are using DNS PTR records for new purposes. The
4223 keys for these records are domain names, not reversed IP addresses. The
4224 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4225 leaves it alone. Component reversal etc. now happens only for IP addresses.
4228 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4230 57. Double the size of the debug message buffer (to 2048) so that more of very
4231 long debug lines gets shown.
4233 58. The exicyclog utility now does better if the number of log files to keep
4234 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4236 59. Two changes related to the smtp_active_hostname option:
4238 (1) $smtp_active_hostname is now available as a variable.
4239 (2) The default for smtp_banner uses $smtp_active_hostname instead
4240 of $primary_hostname.
4242 60. The host_aton() function is supposed to be passed a string that is known
4243 to be a valid IP address. However, in the case of IPv6 addresses, it was
4244 not checking this. This is a hostage to fortune. Exim now panics and dies
4245 if the condition is not met. A case was found where this could be provoked
4246 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4247 components; fortuitously, this particular loophole had already been fixed
4248 by change 4.50/55 above.
4250 If there are any other similar loopholes, the new check in host_aton()
4251 itself should stop them being exploited. The report I received stated that
4252 data on the command line could provoke the exploit when Exim was running as
4253 exim, but did not say which command line option was involved. All I could
4254 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4255 running as the user.
4258 61. There was a buffer overflow vulnerability in the SPA authentication code
4259 (which came originally from the Samba project). I have added a test to the
4260 spa_base64_to_bits() function which I hope fixes it.
4263 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4264 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4266 63. The daemon start-up calls getloadavg() while still root for those OS that
4267 need the first call to be done as root, but it missed one case: when
4268 deliver_queue_load_max is set with deliver_drop_privilege. This is
4269 necessary for the benefit of the queue runner, because there is no re-exec
4270 when deliver_drop_privilege is set.
4272 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4273 This has been fixed.
4275 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4276 was in use, was not putting the data itself into the right store pool;
4277 consequently, it could be overwritten for a subsequent message in the same
4278 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4281 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4283 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4284 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4285 the function were treating the return as a boolean value, which happened to
4286 work because 0=false and not-0=true, but is not correct code.
4288 68. The host_aton() function was not handling scoped IPv6 addresses (those
4289 with, for example, "%eth0" on the end) correctly.
4291 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4292 negated items (that is, ~something) in unsigned ints. Some compilers
4293 apparently mutter when there is no cast.
4295 70. If an address verification called from an ACL failed, and did not produce a
4296 user-specific message (i.e. there was only a "system" message), nothing was
4297 put in $acl_verify_message. In this situation, it now puts the system
4300 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4301 session; change 4.31/43 added the unwanted input to the log line - except
4302 that it did not do this in the start of session case. It now does.
4304 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4305 This is wrong and can cause the other end to generate a synchronization
4306 error if it is another Exim or anything else that does the synchronization
4307 check. A QUIT command is no longer sent after a timeout.
4309 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4310 during host lookups.
4312 74. The defer_ok option of callout verification was not working if it was used
4313 when verifying addresses in header lines, that is, for this case:
4315 verify = header_sender/callout=defer_ok
4317 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4318 those file descriptors could be used for SMTP connections. If anything
4319 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4320 could be sent to the SMTP client, causing chaos. The daemon now opens
4321 stdin, stdout, and stderr to /dev/null when it puts itself into the
4324 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4325 log by default. The user can override this with suitable Perl magic.
4327 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4328 the log message when discard triggers, was not working for the DATA ACL or
4329 for the non-SMTP ACL.
4331 78. Error message wording change in sieve.c.
4333 79. If smtp_accept_max_per_host was set, the number of connections could be
4334 restricted to fewer than expected, because the daemon was trying to set up
4335 a new connection before checking whether the processes handling previous
4336 connections had finished. The check for completed processes is now done
4337 earlier. On busy systems, this bug wouldn't be noticed because something
4338 else would have woken the daemon, and it would have reaped the completed
4341 80. If a message was submitted locally by a user whose login name contained one
4342 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4343 It caused a spool format error. I have fixed the spool reading code. A
4344 related problem was that the "from" clause in the Received: line became
4345 illegal because of the space(s). It is now covered by ${quote_local_part.
4347 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4350 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4352 83. Patch from Sieve maintainer to fix unterminated string problem in
4353 "vacation" handling.
4355 84. Some minor changes to the Linux configuration files to help with other
4356 OS variants using glibc.
4358 85. One more patch for Sieve to update vacation handling to latest spec.
4361 ----------------------------------------------------
4362 See the note above about the 4.44 and 4.50 releases.
4363 ----------------------------------------------------
4369 1. Change 4.43/35 introduced a bug that caused file counts to be
4370 incorrectly computed when quota_filecount was set in an appendfile
4373 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4374 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4377 3. Give more explanation in the error message when the command for a transport
4378 filter fails to execute.
4380 4. There are several places where Exim runs a non-Exim command in a
4381 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4382 was being done only for the command run by the queryprogram router. It is
4383 now done for all such subprocesses. The other cases are: ${run, transport
4384 filters, and the commands run by the lmtp and pipe transports.
4386 5. Some older OS have a limit of 256 on the maximum number of file
4387 descriptors. Exim was using setrlimit() to set 1000 as a large value
4388 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4389 systems. I've change it so that if it can't get 1000, it tries for 256.
4391 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4392 was an oversight, and furthermore, ever since the addition of extra
4393 controls (e.g. 4.43/32), the checks on when to allow different forms of
4394 "control" were broken. There should now be diagnostics for all cases when a
4395 control that does not make sense is encountered.
4397 7. $recipients is now available in the predata ACL (oversight).
4399 8. Tidy the search cache before the fork to do a delivery from a message
4400 received from the command line. Otherwise the child will trigger a lookup
4401 failure and thereby defer the delivery if it tries to use (for example) a
4402 cached ldap connection that the parent has called unbind on.
4404 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4405 of $address_data from the recipient verification was clobbered by the
4406 sender verification.
4408 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4409 was its contents. (It was OK if the option was not defined at all.)
4411 11. A "Completed" log line is now written for messages that are removed from
4412 the spool by the -Mrm option.
4414 12. $host_address is now set to the target address during the checking of
4415 ignore_target_hosts.
4417 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4418 being passed; this would have caused $sender_host_name to have been used if
4419 matching the list had actually called for a host name (not very likely,
4420 since this list is usually IP addresses). A host name is now passed as
4423 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4424 code that specifies a non-restarting handler (typically sigaction() in
4425 modern systems) in an attempt to fix a rare and obscure crash bug.
4427 15. Narrowed the window for a race in the daemon that could cause it to ignore
4428 SIGCHLD signals. This is not a major problem, because they are used only to
4429 wake it up if nothing else does.
4431 16. A malformed maildirsize file could cause Exim to calculate negative values
4432 for the mailbox size or file count. Odd effects could occur as a result.
4433 The maildirsize information is now recalculated if the size or filecount
4436 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4437 support for a long time. Removed HAVE_SYS_VFS_H.
4439 18. Updated exipick to current release from John Jetmore.
4441 19. Allow an empty sender to be matched against a lookup in an address list.
4442 Previously the only cases considered were a regular expression, or an
4445 20. Exim went into a mad DNS lookup loop when doing a callout where the
4446 host was specified on the transport, if the DNS lookup yielded more than
4449 21. The RFC2047 encoding function was originally intended for short strings
4450 such as real names; it was not keeping to the 75-character limit for
4451 encoded words that the RFC imposes. It now respects the limit, and
4452 generates multiple encoded words if necessary. To be on the safe side, I
4453 have increased the buffer size for the ${rfc2047: expansion operator from
4456 22. Failure to deliver a bounce message always caused it to be frozen, even if
4457 there was an errors_to setting on the router. The errors_to setting is now
4460 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4461 canonical form (fully expanded) before being placed in
4462 $sender_host_address.
4464 24. Updated eximstats to version 1.33
4466 25. Include certificate and key file names in error message when GnuTLS fails
4467 to set them up, because the GnuTLS error message doesn't include the name
4468 of the failing file when there is a problem reading it.
4470 26. Expand error message when OpenSSL has problems setting up cert/key files.
4473 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4474 (this can affect the format of dates).
4476 28. exim_tidydb, when checking for the continued existence of a message for
4477 which it has found a message-specific retry record, was not finding
4478 messages that were in split spool directories. Consequently, it was
4479 deleting retry records that should have stayed in existence.
4481 29. eximstats updated to version 1.35
4482 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4483 1.35 - bugfix such that pie charts by volume are generated correctly
4485 30. The SPA authentication driver was not abandoning authentication and moving
4486 on to the next authenticator when an expansion was forced to fail,
4487 contradicting the general specification for all authenticators. Instead it
4488 was generating a temporary error. It now behaves as specified.
4490 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4491 (the order specifies the preference for clients). The order is now AES256,
4492 AES128, 3DES, ARCFOUR128.
4494 31. Small patch to Sieve code - explicitly set From: when generating an
4497 32. Exim crashed if a remote delivery caused a very long error message to be
4498 recorded - for instance if somebody sent an entire SpamAssassin report back
4499 as a large number of 550 error lines. This bug was coincidentally fixed by
4500 increasing the size of one of Exim's internal buffers (big_buffer) that
4501 happened as part of the Exiscan merge. However, to be on the safe side, I
4502 have made the code more robust (and fixed the comments that describe what
4505 33. Some experimental protocols are using DNS PTR records for new purposes. The
4506 keys for these records are domain names, not reversed IP addresses. The
4507 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4508 leaves it alone. Component reversal etc. now happens only for IP addresses.
4511 34. The host_aton() function is supposed to be passed a string that is known
4512 to be a valid IP address. However, in the case of IPv6 addresses, it was
4513 not checking this. This is a hostage to fortune. Exim now panics and dies
4514 if the condition is not met. A case was found where this could be provoked
4515 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4516 components; fortuitously, this particular loophole had already been fixed
4517 by change 4.50/55 or 4.44/33 above.
4519 If there are any other similar loopholes, the new check in host_aton()
4520 itself should stop them being exploited. The report I received stated that
4521 data on the command line could provoke the exploit when Exim was running as
4522 exim, but did not say which command line option was involved. All I could
4523 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4524 running as the user.
4527 35. There was a buffer overflow vulnerability in the SPA authentication code
4528 (which came originally from the Samba project). I have added a test to the
4529 spa_base64_to_bits() function which I hope fixes it.
4532 36. The daemon start-up calls getloadavg() while still root for those OS that
4533 need the first call to be done as root, but it missed one case: when
4534 deliver_queue_load_max is set with deliver_drop_privilege. This is
4535 necessary for the benefit of the queue runner, because there is no re-exec
4536 when deliver_drop_privilege is set.
4538 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4539 was in use, was not putting the data itself into the right store pool;
4540 consequently, it could be overwritten for a subsequent message in the same
4541 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4544 38. Sometimes the final signoff response after QUIT could fail to get
4545 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4546 < 0 before doing a fflush(). This bug looks as though it goes back to the
4547 introduction of TLS in release 3.20, but "sometimes" must have been rare
4548 because the tests only now provoked it.
4554 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4555 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4556 Now it returns an integer. A number of calls were still expecting a T/F
4557 return. Fortuitously, in all cases, the tests worked in OK situations,
4558 which is the norm. However, things would have gone wrong on any write
4559 failures on the smtp file descriptor. This function is used when sending
4560 messages over SMTP and also when doing verify callouts.
4562 2. When Exim is called to do synchronous delivery of a locally submitted
4563 message (the -odf or -odi options), it no longer closes stderr before doing
4566 3. Implemented the mua_wrapper option.
4568 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4570 5. Implemented the functions header_remove(), header_testname(),
4571 header_add_at_position(), and receive_remove_recipient(), and exported them
4574 6. If an ACL "warn" statement specified the addition of headers, Exim already
4575 inserted X-ACL-Warn: at the start if there was no header name. However, it
4576 was not making this test for the second and subsequent header lines if
4577 there were newlines in the string. This meant that an invalid header could
4578 be inserted if Exim was badly configured.
4580 7. Allow an ACL "warn" statement to add header lines at the start or after all
4581 the Received: headers, as well as at the end.
4583 8. Added the rcpt_4xx retry error code.
4585 9. Added postmaster_mailfrom=xxx to callout verification option.
4587 10. Added mailfrom=xxxx to the callout verification option, for verify=
4590 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4591 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4593 12. Inserted some casts to stop certain compilers warning when using pointer
4594 differences as field lengths or precisions in printf-type calls (mostly
4595 affecting debugging statements).
4597 13. Added optional readline() support for -be (dynamically loaded).
4599 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4600 same clock tick as a message's arrival, so that its received time was the
4601 same as the "first fail" time on the retry record, and that message
4602 remained on the queue past the ultimate address timeout, every queue runner
4603 would try a delivery (because it was past the ultimate address timeout) but
4604 after another failure, the ultimate address timeout, which should have then
4605 bounced the address, did not kick in. This was a "< instead of <=" error;
4606 in most cases the first failure would have been in the next clock tick
4607 after the received time, and all would be well.
4609 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4610 being recognized when the domain list was tested by the match_domain
4611 condition in an expansion string.
4613 16. Added the ${str2b64: operator.
4615 17. Exim was always calling setrlimit() to set a large limit for the number of
4616 processes, without checking whether the existing limit was already
4617 adequate. (It did check for the limit on file descriptors.) Furthermore,
4618 errors from getrlimit() and setrlimit() were being ignored. Now they are
4619 logged to the main and panic logs, but Exim does carry on, to try to do its
4620 job under whatever limits there are.
4622 18. Imported PCRE 5.0.
4624 19. Trivial typo in log message " temporarily refused connection" (the leading
4627 20. If the log selector return_path_on_delivery was set and an address was
4628 redirected to /dev/null, the delivery process crashed because it assumed
4629 that a return path would always be set for a "successful" delivery. In this
4630 case, the whole delivery is bypassed as an optimization, and therefore no
4633 21. Internal re-arrangement: the function for sending a challenge and reading
4634 a response while authentication was assuming a zero-terminated challenge
4635 string. It's now changed to take a pointer and a length, to allow for
4636 binary data in such strings.
4638 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4640 23. Exim was not respecting finduser_retries when seeking the login of the
4641 uid under which it was called; it was always trying 10 times. (The default
4642 setting of finduser_retries is zero.) Also, it was sleeping after the final
4643 failure, which is pointless.
4645 24. Implemented tls_on_connect_ports.
4647 25. Implemented acl_smtp_predata.
4649 26. If the domain in control=submission is set empty, Exim assumes that the
4650 authenticated id is a complete email address when it generates From: or
4651 Sender: header lines.
4653 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4654 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4655 chown and chgrp in /bin and hostname in /usr/bin.
4657 28. Exim was keeping the "process log" file open after each use, just as it
4658 does for the main log. This opens the possibility of it remaining open for
4659 long periods when the USR1 signal hits a daemon. Occasional processlog
4660 errors were reported, that could have been caused by this. Anyway, it seems
4661 much more sensible not to leave this file open at all, so that is what now
4664 29. The long-running daemon process does not normally write to the log once it
4665 has entered its main loop, and it closes the log before doing so. This is
4666 so that log files can straightforwardly be renamed and moved. However,
4667 there are a couple of unusual error situations where the daemon does write
4668 log entries, and I had neglected to close the log afterwards.
4670 30. The text of an SMTP error response that was received during a remote
4671 delivery was being truncated at 512 bytes. This is too short for some of
4672 the long messages that one sometimes sees. I've increased the limit to
4675 31. It is now possible to make retry rules that apply only when a message has a
4676 specific sender, in particular, an empty sender.
4678 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4679 it possible to be selective about when SMTP synchronization is enforced.
4681 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4683 32. Implemented hosts_connection_nolog.
4685 33. Added an ACL for QUIT.
4687 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4690 35. Added mailbox_size and mailbox_filecount to appendfile.
4692 36. Added control = no_multiline_responses to ACLs.
4694 37. There was a bug in the logic of the code that waits for the clock to tick
4695 in the case where the clock went backwards by a substantial amount such
4696 that the microsecond fraction of "now" was more than the microsecond
4697 fraction of "then" (but the whole seconds number was less).
4699 38. Added support for the libradius Radius client library this is found on
4700 FreeBSD (previously only the radiusclient library was supported).
4706 1. When certain lookups returned multiple values in the form name=value, the
4707 quoting of the values was not always being done properly. Specifically:
4708 (a) If the value started with a double quote, but contained no whitespace,
4710 (b) If the value contained whitespace other than a space character (i.e.
4711 tabs or newlines or carriage returns) it was not quoted.
4712 This fix has been applied to the mysql and pgsql lookups by writing a
4713 separate quoting function and calling it from the lookup code. The fix
4714 should probably also be applied to nisplus, ibase and oracle lookups, but
4715 since I cannot test any of those, I have not disturbed their existing code.
4717 2. A hit in the callout cache for a specific address caused a log line with no
4718 reason for rejecting RCPT. Now it says "Previous (cached) callout
4719 verification failure".
4721 3. There was an off-by-one bug in the queryprogram router. An over-long
4722 return line was truncated at 256 instead of 255 characters, thereby
4723 overflowing its buffer with the terminating zero. As well as fixing this, I
4724 have increased the buffer size to 1024 (and made a note to document this).
4726 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4727 when Exim is waiting for an SMTP response from a remote server, Exim
4728 restarts its select() call on the socket, thereby resetting its timeout.
4729 This is not a problem when such interrupts are rare. Somebody set up a cron
4730 job to run exiwhat every 2 minutes, which is less than the normal select()
4731 timeout (5 or 10 minutes). This meant that the select() timeout never
4732 kicked in because it was always reset. I have fixed this by comparing the
4733 time when an interrupt arrives with the time at the start of the first call
4734 to select(). If more time than the timeout has elapsed, the interrupt is
4735 treated as a timeout.
4737 5. Some internal re-factoring in preparation for the addition of Sieve
4738 extensions (by MH). In particular, the "personal" test is moved to a
4739 separate function, and given an option for scanning Cc: and Bcc: (which is
4740 not set for Exim filters).
4742 6. When Exim created an email address using the login of the caller as the
4743 local part (e.g. when creating a From: or Sender: header line), it was not
4744 quoting the local part when it contained special characters such as @.
4746 7. Installed new OpenBSD configuration files.
4748 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4749 try to make them clearer.
4751 9. Callout options, other than the timeout value, were being ignored when
4752 verifying sender addresses in header lines. For example, when using
4754 verify = header_sender/callout=no_cache
4756 the cache was (incorrectly) being used.
4758 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4759 only the Cygwin environment.
4761 11. When return_path_on_delivery was set as a log selector, if different remote
4762 addresses in the same message used different return paths and parallel
4763 remote delivery occurred, the wrong values would sometimes be logged.
4764 (Whenever a remote delivery process finished, the return path value from
4765 the most recently started remote delivery process was logged.)
4767 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4768 lines when AUTH and/or TLS are in use. This is the "received protocol"
4769 field. Exim used to use "asmtp" for authenticated SMTP, without any
4770 indication (in the protocol name) for TLS use. Now it follows the RFC and
4771 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4772 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4773 names appear in log lines as well as in Received: header lines.
4775 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4776 extensions, and comparison tests, and to fix some bugs.
4778 14. Changes to the "personal" filter test:
4780 (1) The test was buggy in that it was just doing the equivalent of
4781 "contains" tests on header lines. For example, if a user's address was
4782 anne@some.where, the "personal" test would incorrectly be true for
4784 To: susanne@some.where
4786 This test is now done by extracting each address from the header in turn,
4787 and checking the entire address. Other tests that are part of "personal"
4788 are now done using regular expressions (for example, to check local parts
4789 of addresses in From: header lines).
4791 (2) The list of non-personal local parts in From: addresses has been
4792 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4793 taken from the Sieve specification recommendations.
4795 (3) If the message contains any header line starting with "List-" it is
4796 treated as non-personal.
4798 (4) The test for "circular" in the Subject: header line has been removed
4799 because it now seems ill-conceived.
4801 15. Minor typos in src/EDITME comments corrected.
4803 16. Installed latest exipick from John Jetmore.
4805 17. If headers_add on a router specified a text string that was too long for
4806 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4807 of string_sprintf() is now avoided.
4809 18. $message_body_size was not set (it was always zero) when running the DATA
4810 ACL and the local_scan() function.
4812 19. For the "mail" command in an Exim filter, no default was being set for
4813 the once_repeat time, causing a random time value to be used if "once" was
4814 specified. (If the value happened to be <= 0, no repeat happened.) The
4815 default is now 0s, meaning "never repeat". The "vacation" command was OK
4816 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4817 (I found it when inspecting the code).
4819 20. There is now an overall timeout for performing a callout verification. It
4820 defaults to 4 times the callout timeout, which applies to individual SMTP
4821 commands during the callout. The overall timeout applies when there is more
4822 than one host that can be tried. The timeout is checked before trying the
4823 next host. This prevents very long delays if there are a large number of
4824 hosts and all are timing out (e.g. when the network connections are timing
4825 out). The value of the overall timeout can be changed by specifying an
4826 additional sub-option for "callout", called "maxwait". For example:
4828 verify = sender/callout=5s,maxwait=20s
4830 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4831 to the end before writing, but this should make it even safer).
4833 22. Exim was forgetting that it had advertised PIPELINING for the second and
4834 subsequent messages on an SMTP connection. It was also not resetting its
4835 memory on STARTTLS and an internal HELO.
4837 23. When Exim logs an SMTP synchronization error within a session, it now
4838 records whether PIPELINING has been advertised or not.
4840 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4841 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4842 rather than long int.
4844 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4846 26. Added the never_mail option to autoreply.
4852 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4853 crash if the getsockname() call failed; this can happen if a connection is
4854 closed very soon after it is established. The problem was simply in the
4855 order in which certain operations were done, causing Exim to try to write
4856 to the SMTP stream before it had set up the file descriptor. The bug has
4857 been fixed by making things happen in the correct order.
4863 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4864 before the connection was closed, thus losing the rejection response.
4866 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4867 some early Solaris releases, but causes trouble in current releases where
4868 socklen_t is defined.
4870 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4873 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4876 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4877 that releasing the top of it at the end releases what was used for sub-
4878 expansions (unless the block got too big). However, discard this block if
4879 the first thing is a variable or header, so that we can use its block when
4880 it is dynamic (useful for very large $message_headers, for example).
4882 6. Lookups now cache *every* query, not just the most recent. A new, separate
4883 store pool is used for this. It can be recovered when all lookup caches are
4884 flushed. Lookups now release memory at the end of their result strings.
4885 This has involved some general refactoring of the lookup sources.
4887 7. Some code has been added to the store_xxx() functions to reduce the amount
4888 of flapping under certain conditions.
4890 8. log_incoming_interface used to affect only the <= reception log lines. Now
4891 it causes the local interface and port to be added to several more SMTP log
4892 lines, for example "SMTP connection from", and rejection lines.
4894 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4896 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4898 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4899 could be overwritten at the end of the current message (or the start of a
4900 new message if it was set in a HELO ACL). The value is now preserved for
4901 the duration of the SMTP connection.
4903 12. If a transport had a headers_rewrite setting, and a matching header line
4904 contained an unqualified address, that address was qualified, even if it
4905 did not match any rewriting rules. The underlying bug was that the values
4906 of the flags that permit the existence of unqualified sender and recipient
4907 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4908 non-local messages, and by -bnq for local messages) were not being
4909 preserved with the message after it was received.
4911 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4912 "next input=" as part of the text comprising the host identity instead of
4913 the correct text. The code was using the same buffer for two different
4914 strings. However, depending on which order the printing function evaluated
4915 its arguments, the bug did not always show up. Under Linux, for example, my
4916 test suite worked just fine.
4918 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4919 doesn't work with some older versions of Perl. It has been changed to "my",
4920 which in any case is probably the better facility to use.
4922 15. A really picky compiler found some instances of statements for creating
4923 error messages that either had too many or two few arguments for the format
4926 16. The size of the buffer for calls to the DNS resolver has been increased
4927 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4928 for addresses that have a lot of PTR records. This alleviates a problem; it
4929 does not fully solve it.
4931 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4932 buffer now truncates the list and logs the incident, which is the same
4933 action as happens when Exim is looking up a host name and its aliases.
4934 Previously in this situation something unpredictable would happen;
4935 sometimes it was "internal error: store_reset failed".
4937 18. If a server dropped the connection unexpectedly when an Exim client was
4938 using GnuTLS and trying to read a response, the client delivery process
4939 crashed while trying to generate an error log message.
4941 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4942 string, for example:
4944 warn message = H1: something\nH2: something
4946 the text was added as a single header line from Exim's point of view
4947 though it ended up OK in the delivered message. However, searching for the
4948 second and subsequent header lines using $h_h2: did not work. This has been
4949 fixed. Similarly, if a system filter added multiple headers in this way,
4950 the routers could not see them.
4952 20. Expanded the error message when iplsearch is called with an invalid key to
4953 suggest using net-iplsearch in a host list.
4955 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4956 ACLs are no longer actually imposed (and a message to that effect is
4959 22. If a "gecos" field in a passwd entry contained escaped characters, in
4960 particular, if it contained a \" sequence, Exim got it wrong when building
4961 a From: or a Sender: header from that name. A second bug also caused
4962 incorrect handling when an unquoted " was present following a character
4963 that needed quoting.
4965 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4966 was not being matched caselessly.
4968 24. Arranged for all hyphens in the exim.8 source to be escaped with
4971 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4972 properly. Recipient callout cache records were still being keyed to include
4973 the sender, even when use_sender was set false. This led to far more
4974 callouts that were necessary. The sender is no longer included in the key
4975 when use_sender is false.
4977 26. Added "control = submission" modifier to ACLs.
4979 27. Added the ${base62d: operator to decode base 62 numbers.
4981 28. dnsdb lookups can now access SRV records.
4983 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4984 the configuration file.
4986 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4987 (-v) mode. This makes the output for a verbose queue run more intelligible.
4989 31. Added a use_postmaster feature to recipient callouts.
4991 32. Added the $body_zerocount variable, containing the number of binary zero
4992 bytes in the message body.
4994 33. The time of last modification of the "new" subdirectory is now used as the
4995 "mailbox time last read" when there is a quota error for a maildir
4998 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5000 35. Added +ignore_unknown as a special item in host lists.
5002 36. Code for decoding IPv6 addresses in host lists is now included, even if
5003 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5004 address was recognized as an IP address, but was then not correctly decoded
5005 into binary, causing unexpected and incorrect effects when compared with
5012 1. Very minor rewording of debugging text in manualroute to say "list of
5013 hosts" instead of "hostlist".
5015 2. If verify=header_syntax was set, and a header line with an unqualified
5016 address (no domain) and a large number of spaces between the end of the
5017 name and the colon was received, the reception process suffered a buffer
5018 overflow, and (when I tested it) crashed. This was caused by some obsolete
5019 code that should have been removed. The fix is to remove it!
5021 3. When running in the test harness, delay a bit after writing a bounce
5022 message to get a bit more predictability in the log output.
5024 4. Added a call to search_tidyup() just before forking a reception process. In
5025 theory, someone could use a lookup in the expansion of smtp_accept_max_
5026 per_host which, without the tidyup, could leave open a database connection.
5028 5. Added the variables $recipient_data and $sender_data which get set from a
5029 lookup success in an ACL "recipients" or "senders" condition, or a router
5030 "senders" option, similar to $domain_data and $local_part_data.
5032 6. Moved the writing of debug_print from before to after the "senders" test
5035 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5036 problems for message scanning, either using a data ACL, or using
5037 local_scan() because the Received: header was not generated till after they
5038 were called (in order to set the time as the time of reception completion).
5039 I have revised the way this works. The header is now generated after the
5040 body is received, but before the ACL or local_scan() are called. After they
5041 are run, the timestamp in the header is updated.
5047 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5048 before starting a queue runner without re-exec. This happened only when
5049 deliver_drop_privilege was set or when the Exim user was set to root. The
5050 effect of the bug was that timeouts during subsequent deliveries caused
5051 crashes instead of being properly handled. The handler is now left at its
5052 default (and expected) setting.
5054 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5055 message, again when deliver_drop_privilege is set or Exim is run as root.
5056 The bug described in (1) was not present in this case, but the tidying up
5057 of the other signals was missing. I have made the two cases consistent.
5059 3. The ignore_target_hosts setting on a manualroute router was being ignored
5060 for hosts that were looked up using the /MX notation.
5062 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5065 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5066 operated on the sender address. After changing the $sender_address to <>
5067 for the sender address verify, Exim was re-instated it as the original
5068 (before rewriting) address, but remembering that it had rewritten it, so it
5069 wasn't rewriting it again. This bug also had the effect of breaking the
5070 sender address verification caching when the sender address was rewritten.
5072 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5073 This has been changed so that if the ip literal address matches
5074 ignore_target_hosts, the router declines.
5076 7. Added expansion conditions match_domain, match_address, and match_local_
5077 part (NOT match_host).
5079 8. The placeholder for the Received: header didn't have a length field set.
5081 9. Added code to Exim itself and to exim_lock to test for a specific race
5082 condition that could lead to file corruption when using MBX delivery. The
5083 issue is with the lockfile that is created in /tmp. If this file is removed
5084 after a process has opened it but before that process has acquired a lock,
5085 there is the potential for a second process to recreate the file and also
5086 acquire a lock. This could lead to two Exim processes writing to the file
5087 at the same time. The added code performs the same test as UW imapd; it
5088 checks after acquiring the lock that its file descriptor still refers to
5089 the same named file.
5091 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5092 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5095 11. Added the smtp_active_hostname option. If used, this will typically be made
5096 to depend on the incoming interface address. Because $interface_address is
5097 not set up until the daemon has forked a reception process, error responses
5098 that can happen earlier (such as "too many connections") no longer contain
5101 12. If an expansion in a condition on a "warn" statement fails because a lookup
5102 defers, the "warn" statement is abandoned, and the next ACL statement is
5103 processed. Previously this caused the whole ACL to be aborted.
5105 13. Added the iplsearch lookup type.
5107 14. Added ident_timeout as a log selector.
5109 15. Added tls_certificate_verified as a log selector.
5111 16. Added a global option tls_require_ciphers (compare the smtp transport
5112 option of the same name). This controls incoming TLS connections.
5114 17. I finally figured out how to make tls_require_ciphers do a similar thing
5115 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5116 before starting the TLS session.
5118 18. Tabs are now shown as \t in -bP output.
5120 19. If the log selector return_path_on_delivery was set, Exim crashed when
5121 bouncing a message because it had too many Received: header lines.
5123 20. If two routers both had headers_remove settings, and the first one included
5124 a superfluous trailing colon, the final name in the first list and the
5125 first name in the second list were incorrectly joined into one item (with a
5126 colon in the middle).
5132 1. Added -C and -D options to the exinext utility, mainly to make it easier
5133 to include in the automated testing, but these could be helpful when
5134 multiple configurations are in use.
5136 2. The exinext utility was not formatting the output nicely when there was
5137 an alternate port involved in the retry record key, nor when there was a
5138 message id as well (for retries that were specific to a specific message
5139 and a specific host). It was also confused by IPv6 addresses, because of
5140 the additional colons they contain. I have fixed the IPv4 problem, and
5141 patched it up to do a reasonable job for IPv6.
5143 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5144 delivery, the log line now contains "pipelined" if PIPELINING was used.
5146 4. An SMTP transport process used to panic and die if the bind() call to set
5147 an explicit outgoing interface failed. This has been changed; it is now
5148 treated in the same way as a connect() failure.
5150 5. A reference to $sender_host_name in the part of a conditional expansion
5151 that was being skipped was still causing a DNS lookup. This no longer
5154 6. The def: expansion condition was not recognizing references to header lines
5155 that used bh_ and bheader_.
5157 7. Added the _cache feature to named lists.
5159 8. The code for checking quota_filecount in the appendfile transport was
5160 allowing one more file than it should have been.
5162 9. For compatibility with Sendmail, the command line option
5170 and sets the incoming protocol and host name (for trusted callers). The
5171 host name and its colon can be omitted when only the protocol is to be set.
5172 Note the Exim already has two private options, -pd and -ps, that refer to
5173 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5174 "s", but I don't think that's a major issue.
5176 10. A number of refactoring changes to the code, none of which should affect
5179 (a) The number of logging options was getting close to filling up the
5180 32-bit word that was used as a bit map. I have split them into two classes:
5181 those that are passed in the argument to log_write(), and those that are
5182 only ever tested independently outside of that function. These are now in
5183 separate 32-bit words, so there is plenty of room for expansion again.
5184 There is no change in the user interface or the logging behaviour.
5186 (b) When building, for example, log lines, the code previously used a
5187 macro that called string_cat() twice, in order to add two strings. This is
5188 not really sufficiently general. Furthermore, there was one instance where
5189 it was actually wrong because one of the argument was used twice, and in
5190 one call a function was used. (As it happened, calling the function twice
5191 did not affect the overall behaviour.) The macro has been replaced by a
5192 function that can join an arbitrary number of extra strings onto a growing
5195 (c) The code for expansion conditions now uses a table and a binary chop
5196 instead of a serial search (which was left over from when there were very
5197 few conditions). Also, it now recognizes conditions like "pam" even when
5198 the relevant support is not compiled in: a suitably worded error message is
5199 given if an attempt is made to use such a condition.
5201 11. Added ${time_interval:xxxxx}.
5203 12. A bug was causing one of the ddress fields not to be passed back correctly
5204 from remote delivery subprocesses. The field in question was not being
5205 subsequently used, so this caused to problems in practice.
5207 13. Added new log selectors queue_time and deliver_time.
5209 14. Might have fixed a bug in maildirsizefile handling that threw up
5210 "unexpected character" debug warnings, and recalculated the data
5211 unnecessarily. In any case, I expanded the warning message to give more
5214 15. Added the message "Restricted characters in address" to the statements in
5215 the default ACL that block characters like @ and % in local parts.
5217 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5218 Three changes have been made:
5220 (a) There was a serious bug; a negative response to MAIL caused the whole
5221 recipient domain to be cached as invalid, thereby blocking all messages
5222 to all local parts at the same domain, from all senders. This bug has
5223 been fixed. The domain is no longer cached after a negative response to
5224 MAIL if the sender used is not empty.
5226 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5229 (c) A new callout option, "use_sender" has been added for people who want
5230 the modified behaviour.
5236 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5239 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5240 indeed breaks things for older releases.
5242 3. Added additional logging to the case where there is a problem reading data
5243 from a filter that is running in a subprocess using a pipe, in order to
5244 try to track down a specific problem.
5246 4. Testing facility fudge: when running in the test harness and attempting
5247 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5248 getting "No route to host". Convert this to a timeout.
5250 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5253 6. Some OS don't have socklen_t but use size_t instead. This affects the
5254 fifth argument of getsockopt() amongst other things. This is now
5255 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5256 can be set for individual OS. I have set it for SunOS5, OSF1, and
5257 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5258 some earlier ones do not.
5260 7. Change 4.30/15 was not doing the test caselessly.
5262 8. The standard form for an IPv6 address literal was being rejected by address
5263 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5264 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5265 this, as well as the form without the "IPv6" on the front (but only when
5266 address literals are enabled, of course).
5268 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5270 10. Exim crashed if a message with an empty sender address specified by -f
5271 encountered a router with an errors_to setting. This could be provoked only
5272 by a command such as
5276 where an empty string was supplied; "<>" did not hit this bug.
5278 11. Installed PCRE release 4.5.
5280 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5281 remained set. It is now erased.
5283 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5284 times from message ids (which are base 36 rather than the normal 62).
5286 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5287 were being counted as actual protocol errors, and logged if the log
5288 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5289 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5290 and DATA following a set of rejected RCPTs do not count as protocol errors.
5291 In other words, Exim assumes they were pipelined, though this may not
5292 actually be the case. Of course, in all cases the client gets an
5293 appropriate error code.
5295 15. If a lookup fails in an ACL condition, a message about the failure may
5296 be available; it is used if testing the ACL cannot continue, because most
5297 such messages specify what the cause of the deferral is. However, some
5298 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5299 that caused an old message to be retained and used if a later statement
5300 caused a defer, replacing the real cause of the deferral.
5302 16. If an IP address had so many PTR records that the DNS lookup buffer
5303 was not large enough to hold them, Exim could crash while trying to process
5304 the truncated data. It now detects and logs this case.
5306 17. Further to 4.21/58, another change has been made: if (and only if) the
5307 first line of a message (the first header line) ends with CRLF, a bare LF
5308 in a subsequent header line has a space inserted after it, so as not to
5309 terminate the header.
5311 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5312 unnecessarily, used atoi() instead of strtol(), and didn't check the
5313 termination when getting file sizes from file names by regex.
5315 19. Completely re-implemented the support for maildirsize files, in the light
5316 of a number of problems with the previous contributed implementation
5317 (4.30/29). In particular:
5319 . If the quota is zero, the maildirsize file is maintained, but no quota is
5322 . If the maildir directory does not exist, it is created before any attempt
5323 to write a maildirsize file.
5325 . The quota value in the file is just a cache; if the quota is changed in
5326 the transport, the new value overrides.
5328 . A regular expression is available for excluding directories from the
5331 20. The autoreply transport checks the characters in options that define the
5332 message's headers; it allows continued headers, but it was checking with
5333 isspace() after an embedded newline instead of explicitly looking for a
5336 21. If all the "regular" hosts to which an address was routed had passed their
5337 expiry times, and had not reached their retry times, the address was
5338 bounced, even if fallback hosts were defined. Now Exim should go on to try
5341 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5342 equivalent code in the SMTP transport. Some hosts send humungous responses
5343 to HELO/EHLO, more than 1024 it seems.
5345 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5346 gives compiler warnings in some environments. I've now done it "properly",
5349 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5350 (because of problems with the built-in one) was declared to return uschar *
5351 instead of char *, causing compiler failure.
5353 25. Fixed a file descriptor leak when processing alias/forward files.
5355 26. Fixed a minor format string issue in dbfn.c.
5357 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5359 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5360 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5363 29. When a sender address is verified, it is cached, to save repeating the test
5364 when there is more than one recipient in a message. However, when the
5365 verification involves a callout, it is possible for different callout
5366 options to be set for different recipients. It is too complicated to keep
5367 track of this in the cache, so now Exim always runs a verification when a
5368 callout is required, relying on the callout cache for the optimization.
5369 The overhead is duplication of the address routing, but this should not be
5372 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5373 to be verified with callout=postmaster, and the main callout worked but the
5374 postmaster check failed, the verification correctly failed. However, if a
5375 subsequent RCPT command asked for sender verification *without* the
5376 postmaster check, incorrect caching caused this verification also to fail,
5379 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5380 it was not caching the DNS options (qualify_single, search_parents) that
5381 were used when the lookup failed. A subsequent lookup with different
5382 options therefore always gave the same answer, though there were cases
5383 where it should not have. (Example: a "domains = !$mx_any" option on a
5384 dnslookup router: the "domains" option is always processed without any
5385 widening, but the router might have qualify_single set.) Now Exim uses the
5386 cached value only when the same options are set.
5388 32. Added John Jetmore's "exipick" utility to the distribution.
5390 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5391 than a timeout (e.g. a certificate is required, and is not provided), an
5392 Exim server now closes the connection immediately. Previously it waited for
5393 the client to close - but if the client is SSL, it seems that they each
5394 wait for each other, leading to a delay before one of them times out.
5396 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5397 maintained 0.8.x compatibility because I don't think many are using it, and
5398 it is clearly obsolete.
5400 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5403 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5404 client certificate was expired. A simple patch fixes this, though I don't
5405 understand the full logic of why the verify callback is called multiple
5408 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5409 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5410 which causes problems with some clients (such as the Certicom SSL Plus
5411 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5412 disables the coutermeasure allowing Eudora to connect."
5414 38. Exim was not checking that a write() to a log file succeeded. This could
5415 lead to Bad Things if a log got too big, in particular if it hit a file
5416 size limit. Exim now panics and dies if it cannot write to a log file, just
5417 as it does if it cannot open a log file.
5419 39. Modified OS/Makefile-Linux so that it now contains
5421 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5423 The two -D definitions ensure that Exim is compiled with large file
5424 support, which makes it possible to handle log files that are bigger than
5427 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5428 instance) a domain was checked against a named list that involved a lookup,
5429 causing $domain_data to be set, then another domain was checked against the
5430 same list, then the first domain was re-checked, the value of $domain_data
5431 after the final check could be wrong. In particular, if the second check
5432 failed, it could be set empty. This bug probably also applied to
5435 41. The strip_trailing_dot option was not being applied to the address given
5436 with the -f command-line option.
5438 42. The code for reading a message's header from the spool was incrementing
5439 $received_count, but never initializing it. This meant that the value was
5440 incorrect (doubled) while delivering a message in the same process in which
5441 it was received. In the most common configuration of Exim, this never
5442 happens - a fresh exec is done - but it can happen when
5443 deliver_drop_privilege is set.
5445 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5446 it now includes up to 150 characters of the unexpected data in the log
5449 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5450 and building data strings. The size of both of these buffers was 10 000
5451 bytes - far larger than anybody would *ever* want, thought I. Needless to
5452 say, somebody hit the limit. I have increased the maximum line length to
5453 20 000 and the maximum data length of concatenated lines to 100 000. I have
5454 also fixed two bugs, because there was no checking on these buffers. Tsk,
5455 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5456 buffer is too small.
5458 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5459 lsearch lookups. Now it does.
5461 46. When parsing a route_list item in a manualroute router, a fixed-length
5462 buffer was used for the list of hosts. I made this 1024 bytes long,
5463 thinking that nobody would ever have a list of hosts that long. Wrong.
5464 Somebody had a whole pile of complicated expansion conditions, and the
5465 string was silently truncated, leading to an expansion error. It turns out
5466 that it is easier to change to an unlimited length (owing to other changes
5467 that have happened since this code was originally written) than to build
5468 structure for giving a limitation error. The length of the item that
5469 expands into the list of hosts is now unlimited.
5471 47. The lsearch lookup could not handle data where the length of text line was
5472 more than 4095 characters. Such lines were truncated, leading to shortened
5473 data being returned. It should now handle lines of any length.
5475 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5476 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5479 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5480 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5481 no longer happens in comment lines. A list of replacements is now placed
5482 at the head of all of the source files, except those whose only change is
5483 to replace PERL_COMMAND in the very first #! line.
5485 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5486 messages on the queue, with a bottom-up merge sort, using code contributed
5487 by Michael Haardt. This should make operations like -bp somewhat faster on
5488 large queues. It won't affect queue runners, except when queue_run_in_order
5491 51. Installed eximstats 1.31 in the distribution.
5493 52. Added support for SRV lookups to the dnslookup router.
5495 53. If an ACL referred to $message_body or $message_body_end, the value was not
5496 reset for any messages that followed in the same SMTP session.
5498 54. The store-handling optimization for building very long strings was not
5499 differentiating between the different store pools. I don't think this
5500 actually made any difference in practice, but I've tidied it.
5502 55. While running the routers to verify a sender address, $sender_address
5503 was still set to the sender address. This is wrong, because when routing to
5504 send a bounce to the sender, it would be empty. Therefore, I have changed
5505 it so that, while verifying a sender address, $sender_address is set to <>.
5506 (There is no change to what happens when verifying a recipient address.)
5508 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5509 target A or AAAA records (if not already returned) without resetting the
5510 qualify_single or search_parents options of the DNS resolver. These are
5511 inappropriate in this case because the targets of MX and SRV records must
5512 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5513 target that, when qualified, matched something in the local domain. These
5514 two options are now turned off when doing these lookups.
5516 57. It seems that at least some releases of Reiserfs (which does not have the
5517 concept of a fixed number of inodes) returns zero and not -1 for the
5518 number of available inodes. This interacted badly with check_spool_inodes,
5519 which assumed that -1 was the "no such thing" setting. What I have done is
5520 to check that the total number of inodes is greater than zero before doing
5521 the test of how many are available.
5523 58. When a "warn" ACL statement has a log_message modifier, the message is
5524 remembered, and not repeated. This is to avoid a lot of repetition when a
5525 message has many recipients that cause the same warning to be written.
5526 Howewer, Exim was preserving the list of already written lines for an
5527 entire SMTP session, which doesn't seem right. The memory is now reset if a
5528 new message is started.
5530 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5531 debugging output unless log_rewrite was also set.
5533 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5534 of (int)(handle) when we know that handle contains (void *)(-1).
5536 61. The Exim daemon panic-logs an error return when it closes the incoming
5537 connection. However "connection reset by peer" seems to be common, and
5538 isn't really an error worthy of noting specially, so that particular error
5541 62. When Exim is trying to find all the local interfaces, it used to panic and
5542 die if the ioctl to get the interface flags failed. However, it seems that
5543 on at least one OS (Solaris 9) it is possible to have an interface that is
5544 included in the list of interfaces, but for which you get a failure error
5545 for this call. This happens when the interface is not "plumbed" into a
5546 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5547 failure of the "get flags" call assumes that the interface is down.
5549 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5550 makes life easier for people who are doing arithmetic on fields extracted
5551 from dates, where you often get leading zeros that should not be
5552 interpreted as octal.
5554 64. Added qualify_domain to the redirect router, to override the global
5557 65. If a pathologically long header line contained very many addresses (the
5558 report of this problem mentioned 10 000) and each of them was rewritten,
5559 Exim could use up a very large amount of memory. (It kept on making new
5560 copies of the header line as it rewrote, and never released the old ones.)
5561 At the expense of a bit more processing, the header rewriting function has
5562 been changed so that it no longer eats memory in this way.
5564 66. The generation of the Received: header has been moved from the time that a
5565 message starts to be received, to the time that it finishes. The timestamp
5566 in the Received: header should now be very close to that of the <= log
5567 line. There are two side-effects of this change:
5569 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5570 logged header lines no longer include the local Received: line, because
5571 it has not yet been created. The same applies to a copy of the message
5572 that is returned to a non-SMTP sender when a message is rejected.
5574 (b) When a filter file is tested using -bf, no additional Received: header
5575 is added to the test message. After some thought, I decided that this
5578 This change does not affect the value of $received_for. It is still set
5579 after address rewriting, but before local_scan() is called.
5581 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5583 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5584 gave an unhelpful panic error message, and a defer error. I have managed to
5585 change this behaviour so that it now rejects any supplied certificate,
5586 which seems right, as the list of acceptable certificates is empty.
5588 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5589 gave an unhelpful defer error. I have not managed to make this reject any
5590 supplied certificates, but the error message it gives is "no certificate
5591 supplied", which is not helpful.
5593 70. exigrep's output now also includes lines that are not associated with any
5594 message, but which match the given pattern. Implemented by a patch from
5595 Martin Sluka, which also tidied up the Perl a bit.
5597 71. Recipient callout verification, like sender verification, was using <> in
5598 the MAIL FROM command. This isn't really the right thing, since the actual
5599 sender may affect whether the remote host accepts the recipient or not. I
5600 have changed it to use the actual sender in the callout; this means that
5601 the cache record is now keyed on a recipient/sender pair, not just the
5602 recipient address. There doesn't seem to be a real danger of callout loops,
5603 since a callout by the remote host to check the sender would use <>.
5604 [SEE ABOVE: changed after hitting problems.]
5606 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5607 temporary errors. However, in the case of such a code being given after
5608 the end of a data transmission (i.e. after ".") Exim was failing to write
5609 a retry record for the message. (Yes, there was some broken host that was
5610 actually sending 8xx at this point.)
5612 73. An unknown lookup type in a host list could cause Exim to panic-die when
5613 the list was checked. (An example that provoked this was putting <; in the
5614 middle of a list instead of at the start.) If this happened during a DATA
5615 ACL check, a -D file could be left lying around. This kind of configuration
5616 error no longer causes Exim to die; instead it causes a defer errror. The
5617 incident is still logged to the main and panic logs.
5619 74. Buglet left over from Exim 3 conversion. The message "too many messages
5620 in one connection" was written to the rejectlog but not the mainlog, except
5621 when address rewriting (yes!) was being logged.
5623 75. Added write_rejectlog option.
5625 76. When a system filter was run not as root (that is, when system_filter_user
5626 was set), the values of the $n variables were not being returned to the
5627 main process; thus, they were not subsequently available in the $sn
5630 77. Added +return_path_on_delivery log selector.
5632 78. A connection timeout was being treated differently from recipients deferred
5633 when testing hosts_max_try with a message that was older than the host's
5634 retry timeout. (The host should not be counted, thus allowing all hosts to
5635 be tried at least once before bouncing.) This may have been the cause of an
5636 occasionally reported bug whereby a message would remain on the queue
5637 longer than the retry timeout, but would be bounced if a delivery was
5638 forced. I say "may" because I never totally pinned down the problem;
5639 setting up timeout/retry tests is difficult. See also the next item.
5641 79. The ultimate address timeout was not being applied to errors that involved
5642 a combination of host plus message (for example, a timeout on a MAIL
5643 command). When an address resolved to a number of possible hosts, and they
5644 were not all tried for each delivery (e.g. because of hosts_max_try), a
5645 message could remain on the queue longer than the retry timeout.
5647 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5650 81. Fixed an obscure SMTP outgoing bug which required at least the following
5651 conditions: (a) there was another message waiting for the same server;
5652 (b) the server returned 5xx to all RCPT commands in the first message so
5653 that the message was not completed; (c) the server dropped the connection
5654 or gave a negative response to the RSET that Exim sends to abort the
5655 transaction. The observed case was a dropped connection after DATA that had
5656 been sent in pipelining mode. That is, the server had advertised PIPELINING
5657 but was not implementing it correctly. The effect of the bug was incorrect
5658 behaviour, such as trying another host, and this could lead to a crash.
5664 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5665 and daemon.c were passed as pointers to ints; they should have been
5666 pointers to socklen_t variables (which are typically unsigned ints).
5668 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5671 3. Fixed a really odd bug that affected only the testing scheme; patching a
5672 certain fixed string in the binary changed the value of another string that
5673 happened to be identical to the end of the original first string.
5675 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5676 name", it returns that address as the IP address. On some operating
5677 systems (e.g. Solaris), it also passes back the IP address string as the
5678 "host name". However, on others (e.g. Linux), it passes back an empty
5679 string. Exim wasn't checking for this, and was changing the host name to an
5680 empty string, assuming it had been canonicized.
5682 5. Although rare, it is permitted to have more than one PTR record for a given
5683 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5684 all the names associated with an address, because they do in Solaris.
5685 However, it seems that they do not in Linux for data that comes from the
5686 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5687 I found this out when I moved to a new Linux workstation and tried to run
5688 the Exim test suite.
5690 To get round this problem I have changed the code so that it now does its
5691 own call to the DNS to look up PTR records when searching for a host name.
5692 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5693 addresses that are only in /etc/hosts are still found.
5695 This behaviour is, however, controlled by an option called host_lookup_
5696 order, which defaults to "bydns:byaddr". If people want to use the other
5697 order, or indeed, just use one or the other means of lookup, they can
5698 specify it in this variable.
5700 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5701 some operating systems, this comes back from gethostbyaddr() as an empty
5702 string, and this is what Exim used to test for. However, it seems that in
5703 other systems, "." is yielded. Exim now tests for this case too.
5705 7. The values of check_spool_space and check_log_space are now held internally
5706 as a number of kilobytes instead of an absolute number of bytes. If a
5707 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5708 kilobyte. This means that much larger values can be stored.
5710 8. Exim monitor: an attempt to get the action menu when not actually pointing
5711 at a message produces an empty menu entitled "No message selected". This
5712 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5713 no entries in it ("Shell widget menu has zero width and/or height"). So I
5714 have added a single, blank menu entry in this case.
5716 9. Added ${quote_local_part.
5718 10. MIME decoding is now applied to the contents of Subject: header lines when
5721 11. Now that a reference to $sender_host_address automatically causes a reverse
5722 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5723 host lookup before query-style lookups in lists that might use this
5724 variable. This has therefore been abolished, and the "net-" prefix is no
5725 longer necessary for query-style lookups.
5727 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5728 have been a typo for LFLAGS, so it has been changed.
5730 13. The install script calls Exim with "-C /dev/null" in order to find the
5731 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5732 to be output. Howeve, since Exim outputs its version number before the
5733 error, it didn't break the script. It just looked ugly. I fixed this by
5734 always allowing "-C /dev/null" if the caller is root.
5736 14. Ignore overlarge ACL variable number when reading spool file - insurance
5737 against a later release with more variables having written the file.
5739 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5740 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5741 this, as well as the form without the "IPv6" on the front.
5743 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5744 OS/Makefile-Darwin file.
5746 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5747 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5749 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5750 variable. However, some OS use ints for inodes. I've added cast to long int
5751 to get rid of the compiler warning.
5753 19. I had forgotten to lock out "/../" in configuration file names when
5754 ALT_CONFIG_PREFIX was set.
5756 20. Routers used for verification do not need to specify transports. However,
5757 if such a router generated a host list, and callout was configured, Exim
5758 crashed, because it could not find a port number from the (non-existent)
5759 transport. It now assumes port 25 in this circumstance.
5761 21. Added the -t option to exigrep.
5763 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5764 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5765 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5766 them is set, LOOKUP_LSEARCH is forced.
5768 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5770 24. Added sender and host information to the "rejected by local_scan()" log
5771 line; previously there was no indication of these.
5773 25. Added .include_if_exists.
5775 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5776 Linux. It turns out that not all file systems support this. Apparently some
5777 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5778 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5779 not supported on the file descriptor, is now ignored when Exim is trying to
5780 sync a directory. This applies only to Linux.
5782 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5784 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5785 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5788 29. Added support for maildirsize files from supplied patch (modified a bit).
5790 30. The use of :fail: followed by an empty string could lead Exim to respond to
5791 sender verification failures with (e.g.):
5793 550 Verification failed for <xxx>
5794 550 Sender verify failed
5796 where the first response line was missing the '-' that indicates it is not
5797 the final line of the response.
5799 31. The loop for finding the name of the user that called Exim had a hardwired
5800 limit of 10; it now uses the value of finduser_retries, which is used for
5801 all other user lookups.
5803 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5806 33. Exim was neglecting to zero errno before one call of strtol() when
5807 expanding a string and expecting an integer value. On some systems this
5808 resulted in spurious "integer overflow" errors. Also, it was casting the
5809 result into an int without checking.
5811 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5812 did not work. The code looks as if it has *never* worked, though it appears
5813 to have been documented since at least releast 1.62. I have made it work.
5815 35. The "timeout_DNS" error in retry rules, also documented since at least
5816 1.62, also never worked. As it isn't clear exactly what this means, and
5817 clearly it isn't a major issue, I have abolished the feature by treating it
5818 as "timeout", and writing a warning to the main and panic logs.
5820 36. The display of retry rules for -brt wasn't always showing the error code
5823 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5824 timeout_connect_A, timeout_connect_MX.
5826 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5827 to the empty sender.
5829 39. The daemon was not analysing the content of -oX till after it had closed
5830 stderr and disconnected from the controlling terminal. This meant that any
5831 syntax errors were only noted on the panic log, and the return code from
5832 the command was 0. By re-arranging the code a little, I've made the
5833 decoding happen first, so such errors now appear on stderr, and the return
5834 code is 1. However, the actual setting up of the sockets still happens in
5835 the disconnected process, so errors there are still only recorded on the
5838 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5839 connections (as happens on some IP stacks) was logged at start up time as
5840 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5841 it from "IPv6 and IPv4", which means that two separate sockets are being
5844 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5845 says whether AF_INET or AF_INET6 was passed as an argument.
5847 42. Exiwhat output was messed up when time zones were included in log
5850 43. Exiwhat now gives more information about the daemon's listening ports,
5851 and whether -tls-on-connect was used.
5853 44. The "port" option of the smtp transport is now expanded.
5855 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5856 silently ignored. Now an error message is written to the main and panic
5859 46. There's a new ACL modifier called "logwrite" which writes to a log file
5860 as soon as it is encountered.
5862 47. Added $local_user_uid and $local_user_gid at routing time.
5864 48. Exim crashed when trying to verify a sender address that was being
5867 49. Exim was recognizing only a space character after ".include". It now also
5868 recognizes a tab character.
5870 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5871 extracting the relevant information from the specification. The man page no
5872 longer contains scrambled data for the -d option, and I've added a section
5873 at the front about calling Exim under different names.
5875 51. Added "extra_headers" argument to the "mail" command in filter files.
5877 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5880 53. Installed eximstats 1.29.
5882 54. Added transport_filter_timeout as a generic transport option.
5884 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5885 Cc: header lines. This was required by RFC 822, but it not required by RFC
5888 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5889 incoming messages that did not have them. Now it does so only if the
5890 message originates locally, that is, if there is no associated remote host
5891 address. When Resent- header lines are present, this applies to the Resent-
5892 lines rather than the non-Resent- lines.
5894 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5895 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5897 58. Messages for configuration errors now include the name of the main
5898 configuration file - useful now that there may be more than one file in a
5899 list (.included file names were always shown).
5901 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5902 for those rare installations that do not start the daemon as root or run it
5903 setuid root. I've cut out the call to initgroups() if the daemon is not
5906 60. The Exim user and group can now be bound into the binary as text strings
5907 that are looked up at the start of Exim's processing.
5909 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5911 62. Added $mailstore_basename variable.
5913 63. Installed patch to sieve.c from Michael Haardt.
5915 64. When Exim failed to open the panic log after failing to open the main log,
5916 the original message it was trying to log was written to stderr and debug
5917 output, but if they were not available (the usual case in production), it
5918 was lost. Now it is written to syslog before the two lines that record the
5919 failures to open the logs.
5921 65. Users' Exim filters run in subprocesses under the user's uid. It is
5922 possible for a "deliver" command or an alias in a "personal" command to
5923 provoke an address rewrite. If logging of address rewriting is configured,
5924 this fails because the process is not running as root or exim. There may be
5925 a better way of dealing with this, but for the moment (because 4.30 needs
5926 to be released), I have disabled address rewrite logging when running a
5927 filter in a non-root, non-exim process.
5933 1. The buildconfig auxiliary program wasn't quoting the value set for
5934 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5935 not defined. This bug was masked in 4.22 by the effect that was fixed in
5938 2. Some messages that were rejected after a message id was allocated were
5939 shown as "incomplete" by exigrep. It no longer does this for messages that
5940 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5942 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5943 have allow_domain_literals set, the ID did not get logged in the <= line.
5944 Domain literals are now always recognized in Message-ID: header lines.
5946 4. The first argument for a ${extract expansion item is the key name or field
5947 number. Leading and trailing spaces in this item were not being ignored,
5948 causing some misleading effects.
5950 5. When deliver_drop_privilege was set, single queue runner processes started
5951 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5952 same command in the process it spins off) were not dropping privilege.
5954 6. When the daemon running as "exim" started a queue runner, it always
5955 re-executed Exim in the spun-off process. This is a waste of effort when
5956 deliver_drop_privilege is set. The new process now just calls the
5957 queue-runner function directly.
5963 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5966 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5967 ignored. Though the use of -oP was forcing the writing of a pid file, it
5968 was always written to the default place.
5970 3. If the message "no IP address found for host xxxx" is generated during
5971 incoming verification, it is now followed by identification of the incoming
5972 connection (so you can more easily find what provoked it).
5974 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5976 5. Added some features to "harden" Exim a bit more against certain attacks:
5978 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5979 be put in Local/Makefile. This is like the never_users runtime option,
5980 but it cannot be overridden. The default setting is "root".
5982 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5983 prefix string with which any file named in a -C command line option
5986 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5987 is retained for -C and -D only if the caller of Exim is root. Without
5988 it, the exim user may also use -C and -D and retain privilege.
5990 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5991 command line option is disabled.
5993 6. Macro names set by the -D option must start with an upper case letter, just
5994 like macro names defined in the configuration file.
5996 7. Added "dereference=" facility to LDAP.
5998 8. Two instances of the typo "uknown" in the source files are fixed.
6000 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6001 the Configure-Makefile script screwed up while processing it.
6003 10. Incorporated PCRE 4.4.
6005 11. The SMTP synchronization check was not operating right at the start of an
6006 SMTP session. For example, it could not catch a HELO sent before the client
6007 waited for the greeting. There is now a check for outstanding input at the
6008 point when the greeting is written. Because of the duplex, asynchronous
6009 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6010 way, but not yet received, when the check is performed.
6012 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6013 on TCP/IP sockets, because this apparently causes some broken clients to
6016 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6017 unchanged) from the Cygwin maintainer.
6019 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6020 was supported instead of testing for mbx. Effectively a typo.
6022 15. The spa authenticator server code was not checking that the input it
6023 received was valid base64.
6025 16. The debug output line for the "set" modifier in ACLs was not showing the
6026 name of the variable that was being set.
6028 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6030 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6031 lookup on its own. Something else was needed to trigger the lookup. For
6032 example, a match in host_lookup or the need for a host name in a host list.
6033 Now, if $sender_host_name is referenced and the host name has not yet been
6034 looked up, a lookup is performed. If the lookup fails, the variable remains
6035 empty, and $host_lookup_failed is set to "1".
6037 19. Added "eqi" as a case-independent comparison operator.
6039 20. The saslauthd authentication condition could segfault if neither service
6040 nor realm was specified.
6042 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6043 error message that was logged was misleading, and incoming SMTP
6044 connections were dropped. The message is now more accurate, and temporary
6045 errors are given to SMTP connections.
6047 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6048 (except RSET) with a 421 error, until QUIT is received. However, it was
6049 failing to send a response to QUIT.
6051 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6052 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6053 ACL is now run after the helo_try_verify_hosts code.
6055 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6056 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6057 case-independent, so other case variants are also recognized). Apparently
6058 some systems use these upper case variants.
6060 25. If more than two messages were waiting for the same host, and a transport
6061 filter was specified for the transport, Exim sent two messages over the
6062 same TCP/IP connection, and then failed with "socket operation on non-
6063 socket" when it tried to send the third.
6065 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6067 27. The extern definition of crypt16() in expand.c was not being excluded when
6068 the OS had its own crypt16() function.
6070 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6071 as a preferred synonym for return_size_limit, both as an option and as an
6074 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6076 30. Changed the default configuration ACL to relax the local part checking rule
6077 for addresses that are not in any local domains. For these addresses,
6078 slashes and pipe symbols are allowed within local parts, but the sequence
6079 /../ is explicitly forbidden.
6081 31. SPA server authentication was not clearing the challenge buffer before
6084 32. log_message in a "warn" ACL statement was writing to the reject log as
6085 well as to the main log, which contradicts the documentation and doesn't
6086 seem right (because no rejection is happening). So I have stopped it.
6088 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6089 However, I am unable to do any testing of this.
6091 34. Fixed an infelicity in the appendfile transport. When checking directories
6092 for a mailbox, to see if any needed to be created, it was accidentally
6093 using path names with one or more superfluous leading slashes; tracing
6094 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6096 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6097 contents are added to the log line that is written for every discarded
6098 recipient. (Previously a log_message setting was ignored.)
6100 36. The ${quote: operator now quotes the string if it is empty.
6102 37. The install script runs exim in order to find its version number. If for
6103 some reason other than non-existence or emptiness, which it checks, it
6104 could not run './exim', it was installing it with an empty version number,
6105 i.e. as "exim-". This error state is now caught, and the installation is
6108 38. An argument was missing from the function that creates an error message
6109 when Exim fails to connect to the socket for saslauthd authentication.
6110 This could cause Exim to crash, or give a corrupted message.
6112 39. Added isip, isip4, and isip6 to ${if conditions.
6114 40. The ACL variables $acl_xx are now saved with the message, and can be
6115 accessed later in routers, transports, and filters.
6117 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6118 strings in the file are not string-expanded.
6120 42. If a MAIL command specified a SIZE value that was too large to fit into an
6121 int variable, the check against message_size_limit failed. Such values are
6122 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6123 day this will have to be increased, but I don't think I want to be around
6124 when emails are that large.
6131 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6132 iconv() is not standard in FreeBSD.
6134 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6135 IPv6 enabled. The observed symptom was a segmentation fault on return from
6136 the function os_common_find_running_interfaces() in src/os.c.
6138 3. In the check_special_case() function in daemon.c I had used "errno" as an
6139 argument name, which causes warnings on some systems. This was basically a
6140 typo, since it was named "eno" in the comments!
6142 4. The code that waits for the clock to tick (at a resolution of some fraction
6143 of a second) so as to ensure message-id uniqueness was always waiting for
6144 at least one whole tick, when it could have waited for less. [This is
6145 almost certainly not relevant at current processor speeds, where it is
6146 unlikely to ever wait at all. But we try to future-proof.]
6148 5. The function that sleeps for a time interval that includes fractions of a
6149 second contained a race. It did not block SIGALRM between setting the
6150 timer, and suspending (a couple of lines later). If the interval was short
6151 and the sigsuspend() was delayed until after it had expired, the suspension
6152 never ended. On busy systems this could lead to processes getting stuck for
6155 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6156 process, before it forks any delivery processes. The open lookup caching
6157 mechanism meant that the open file or database connection was passed into
6158 the delivery process. The problem was that delivery processes always tidy
6159 up cached lookup data. This could cause a problem for the next delivery
6160 process started by the queue runner, because the external queue runner
6161 process does not know about the closure. So the next delivery process
6162 still has data in the lookup cache. In the case of a file lookup, there was
6163 no problem because closing a file descriptor in a subprocess doesn't affect
6164 the parent. However, if the lookup was caching a connection to a database,
6165 the connection was closed, and the second delivery process was likely to
6166 see errors such as "PGSQL: query failed: server closed the connection
6167 unexpectedly". The problem has been fixed by closing all cached lookups
6168 in a queue runner before running a delivery process.
6170 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6171 seem to have the "const" qualifier which it has on other OS. I've
6174 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6175 *of the same type* (client or server) with the same public name that an
6176 error should be diagnosed.
6178 9. When Exim looked up a host name for an IP address, but failed to find the
6179 original IP address when looking up the host name (a safety check), it
6180 output the message "<ip address> does not match any IP for NULL", which was
6181 confusing, to say the least. The bug was that the host name should have
6182 appeared instead of "NULL".
6184 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6185 user that is built into the binary, and the -C or -D options is used, root
6186 privilege is dropped before the configuration file is read. In addition,
6187 logging is switched to stderr instead of the normal log files. If the
6188 configuration then re-defines the Exim user, the unprivileged environment
6189 is probably not what is expected, so Exim logs a panic warning message (but
6192 However, if deliver_drop_privilege is set, the unprivileged state may well
6193 be exactly what is intended, so the warning has been cut out in that case,
6194 and Exim is allowed to try to write to its normal log files.
6200 1. smtp_return_error_details was not giving details for temporary sender
6201 or receiver verification errors.
6203 2. Diagnose a configuration error if two authenticators have the same public
6206 3. Exim used not to create the message log file for a message until the first
6207 delivery attempt. This could be confusing when incoming messages were held
6208 for policy or load reasons. The message log file is now created at the time
6209 the message is received, and an initial "Received" line is written to it.
6211 4. The automatically generated man page for command line options had a minor
6212 bug that caused no ill effects; however, a more serious problem was that
6213 the procedure for building the man page automatically didn't always
6214 operate. Consequently, release 4.20 contains an out-of-date version. This
6215 shouldn't happen again.
6217 5. When building Exim with embedded Perl support, the script that builds the
6218 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6219 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6221 6. The freeze_tell option was not being used for messages that were frozen on
6222 arrival, either by an ACL or by local_scan().
6224 7. Added the smtp_incomplete_transaction log selector.
6226 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6227 was accepting AUTH without a new EHLO.
6229 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6230 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6233 10. Logging of TCP/IP connections (when configured) now happens in the main
6234 daemon process instead of the child process, so that the TCP/IP connection
6235 count is more accurate (but it can never be perfect).
6237 11. The use of "drop" in a nested ACL was not being handled correctly in the
6238 outer ACL. Now, if condition failure induced by the nested "drop" causes
6239 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6240 or "require"), the connection is dropped.
6242 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6243 that yield "discard" can now be used with an "accept" or a "discard" verb,
6244 but an error is generated for any others (because I can't see a useful way
6245 to define what should happen).
6247 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6248 are now processed in the same way as lines in the Exim configuration file.
6249 In particular, continuation lines are supported.
6251 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6253 15. Added -ti meaning -t -i.
6255 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6256 domains, and warn by logging if others are found.
6258 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6259 returned from GIFCONF when Exim is trying to find the list of interfaces on
6260 a host. The code in os.c has been modified to copy each ifreq to an aligned
6261 structure in all cases.
6263 Also, in some cases, the returned ifreq's were being copied to a 'struct
6264 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6265 means the last couple of bytes of an IPv6 address could be chopped if the
6266 ifreq contained only a normal sockaddr (14 bytes storage).
6268 18. Named domain lists were not supported in the hosts_treat_as_local option.
6269 An entry such as +xxxx was not recognized, and was treated as a literal
6272 19. Ensure that header lines added by a DATA ACL are included in the reject log
6273 if the ACL subsequently rejects the message.
6275 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6276 MD5 (which is deprecated).
6278 21. When testing a filter file using -bf, Exim was writing a message when it
6279 took the sender from a "From " line in the message, but it was not doing so
6280 when it took $return_path from a Return-Path: header line. It now does.
6282 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6283 with a valid header line field name (a series of printing characters
6284 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6286 23. Changed "disc" in the source to "disk" to conform to the documentation and
6287 the book and for uniformity.
6289 24. Ignore Sendmail's -Ooption=value command line item.
6291 25. When execve() failed while trying to run a command in a pipe transport,
6292 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6293 could be confused with a return value of 69 from the command itself. This
6294 has been changed to 127, the value the shell returns if it is asked to run
6295 a non-existent command. The wording for the related log line suggests a
6296 non-existent command as the problem.
6298 26. If received_header_text expands to an empty string, do not add a Received:
6299 header line to the message. (Well, it adds a token one on the spool, but
6300 marks it "old" so that it doesn't get used or transmitted.)
6302 27. Installed eximstats 1.28 (addition of -nt option).
6304 28. There was no check for failure on the call to getsockname() in the daemon
6305 code. This can fail if there is a shortage of resources on the system, with
6306 ENOMEM, for example. A temporary error is now given on failure.
6308 29. Contrary to the C standard, it seems that in some environments, the
6309 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6310 program. Exim now does this explicitly; it affects the formatting of
6311 timestamps using strftime().
6313 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6314 complaints. I've now initialized all the variables, to avoid this.
6316 32. Header lines added by a system filter were not being "seen" during
6317 transport-time rewrites.
6319 33. The info_callback() function passed to OpenSSL is set up with type void
6320 (*)(SSL *, int, int), as described somewhere. However, when calling the
6321 function (actually a macro) that sets it up, the type void(*)() is
6322 expected. I've put in a cast to prevent warnings from picky compilers.
6324 34. If a DNS black list lookup found a CNAME record, but there were no A
6325 records associated with the domain it pointed at, Exim crashed.
6327 35. If a DNS black list lookup returned more than one A record, Exim ignored
6328 all but the first. It now scans all returned addresses if a particular IP
6329 value is being sought. In this situation, the contents of the
6330 $dnslist_value variable are a list of all the addresses, separated by a
6333 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6334 to accept a host name and all its aliases if the forward lookup for any of
6335 them yielded the IP address of the incoming connection. Now it accepts only
6336 those names whose forward lookup yields the correct IP address. Any other
6337 names are discarded. This closes a loophole whereby a rogue DNS
6338 administrator could create reverse DNS records to break through a
6339 wildcarded host restriction in an ACL.
6341 37. If a user filter or a system filter that ran in a subprocess used any of
6342 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6343 the wrong values were passed to the pipe command ($thisaddress had the
6344 value of $0, $0 had the value of $1, etc). This bug was introduced by
6345 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6347 38. Improved the line breaking for long SMTP error messages from ACLs.
6348 Previously, if there was no break point between 40 and 75 characters, Exim
6349 left the rest of the message alone. Two changes have been made: (a) I've
6350 reduced the minimum length to 35 characters; (b) if it can't find a break
6351 point between 35 and 75 characters, it looks ahead and uses the first one
6352 that it finds. This may give the occasional overlong line, but at least the
6353 remaining text gets split now.
6355 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6356 file descriptors might be low, and that setting 1000 would always raise it.
6357 It turns out that in some environments, the limit is already over 1000 and
6358 that lowering it causes trouble. So now Exim takes care not to decrease it.
6360 40. When delivering a message, the value of $return_path is set to $sender_
6361 address at the start of routing (routers may change the value). By an
6362 oversight, this default was not being set up when an address was tested by
6363 -bt or -bv, which affected the outcome if any router or filter referred to
6366 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6367 log only when "message" or "log_message" are set. However, if one of the
6368 conditions was an address verification, or a call to a nested ACL, the
6369 messages generated by the underlying test were being passed through. This
6370 no longer happens. The underlying message is available in $acl_verify_
6371 message for both "message" and "log_message" expansions, so it can be
6372 passed through if needed.
6374 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6375 new expansion $bh_ to give the encoded byte string without charset
6376 translation. Translation happens only if iconv() is available; HAVE_ICONV
6377 indicates this at build time. HEADERS_CHARSET gives the charset to
6378 translate to; headers_charset can change it in the configuration, and
6379 "headers charset" can change it in an individual filter file.
6381 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6382 that creates RFC 2047 encoded "words" labels them as that charset instead
6383 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6384 expansion operator; (ii) when Exim creates a From: line for a local
6385 message; (iii) when a header line is rewritten to include a "phrase" part.
6387 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6388 buggy, causing it to skip the first lines of messages whose message ID
6389 ended in 'D'. This would not have bitten before Exim release 4.14, because
6390 message IDs were unlikely to end in 'D' before then. The effect was to have
6391 incorrect size information for certain domains.
6393 45. #include "config.h" was missing at the start of the crypt16.c module. This
6394 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6397 46. If there was a timeout during a "random" callout check, Exim treated it as
6398 a failure of the random address, and carried on sending RSET and the real
6399 address. If the delay was just some slowness somewhere, the response to the
6400 original RCPT would be taken as a response to RSET and so on, causing
6401 mayhem of various kinds.
6403 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6404 when I implemented it. It didn't allow for the fact that some option values
6405 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6406 the right test for positive values.
6408 48. Domain names in DNS records are case-independent. Exim always looks them up
6409 in lower case. Some resolvers return domain names in exactly the case they
6410 appear in the zone file, that is, they may contain uppercase letters. Not
6411 all resolvers do this - some return always lower case. Exim was treating a
6412 change of case by a resolver as a change of domain, similar to a widening
6413 of a domain abbreviation. This triggered its re-routing code and so it was
6414 trying to route what was effectively the same domain again. This normally
6415 caused routing to fail (because the router wouldn't handle the domain
6416 twice). Now Exim checks for this case specially, and just changes the
6417 casing of the domain that it ultimately uses when it transmits the message
6420 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6423 50. If a filter generated a file delivery with a non-absolute name (possible if
6424 no home directory exists for the router), the forbid_file option was not
6427 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6428 the existing equality matching.
6430 52. Exim was using ints instead of ino_t variables in some places where it was
6431 dealing with inode numbers.
6433 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6434 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6435 called TMPDIR, and if it finds it is different, it changes its value.
6437 54. The smtp_printf() function is now made available to local_scan() so
6438 additional output lines can be written before returning. There is also an
6439 smtp_fflush() function to enable the detection of a dropped connection.
6440 The variables smtp_input and smtp_batched_input are exported to
6443 55. Changed the default runtime configuration: the message "Unknown user"
6444 has been removed from the ACL, and instead placed on the localuser router,
6445 using the cannot_route_message feature. This means that any verification
6446 failures that generate their own messages won't get overridden. Similarly,
6447 the "Unrouteable address" message that was in the ACL for unverifiable
6448 relay addresses has also been removed.
6450 56. Added hosts_avoid_esmtp to the smtp transport.
6452 57. The exicyclog script was not checking for the esoteric option
6453 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6454 will work only if exicyclog is run under the appropriate euid.
6456 58. Following a discussion on the list, the rules by which Exim recognises line
6457 endings on incoming messages have been changed. The -dropcr and drop_cr
6458 options are now no-ops, retained only for backwards compatibility. The
6459 following line terminators are recognized: LF CRLF CR. However, special
6460 processing applies to CR:
6462 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6463 nor a local message in the state where . is a terminator.
6465 (ii) If a bare CR is encountered in a header line, an extra space is added
6466 after the line terminator so as not to end the header. The reasoning
6467 behind this is that bare CRs in header lines are most likely either
6468 to be mistakes, or people trying to play silly games.
6470 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6471 was being incorrectly given as 18 bytes larger than it should have been.
6472 This is a VOB (very old bug).
6474 60. This may never have affected anything current, but just in case it has:
6475 When the local host is found other than at the start of a list of hosts,
6476 the local host, those with the same MX, and any that follow, are discarded.
6477 When the list in question was part of a longer list of hosts, the following
6478 hosts (not currently being processed) were also being discarded. This no
6479 longer happens. I'm not sure if this situation could ever has previously
6482 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6485 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6486 header. This is something that is recommended in a new Internet Draft, and
6487 is something that is documented as being done by Sendmail. There are two
6488 possible values. For messages generated by the autoreply transport, Exim
6491 Auto-Submitted: auto-replied
6493 whereas for all other generated messages (e.g. bounces) it adds
6495 Auto-Submitted: auto-generated
6497 63. The "personal" condition in filters now includes a test for the
6498 Auto-Submitted: header. If it contains the string "auto-" the message it
6499 not considered personal.
6501 64. Added rcpt_include_affixes as a generic transport option.
6503 65. Added queue_only_override (default true).
6505 66. Added the syslog_duplication option.
6507 67. If what should have been the first header line of a message consisted of
6508 a space followed by a colon, Exim was mis-interpreting it as a header line.
6509 It isn't of course - it is syntactically invalid and should therefore be
6510 treated as the start of the message body. The misbehaviour could have
6511 caused a number of strange effects, including loss of data in subsequent
6512 header lines, and spool format errors.
6514 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6515 client host had authenticated. This control can now be exercised by an ACL
6516 for more flexibility.
6518 69. By default, callouts do not happen when testing with -bh. There is now a
6519 variant, -bhc, which does actually run the callout code, including
6520 consulting and updating the callout cache.
6522 70. Added support for saslauthd authentication, courtesy of Alexander
6525 71. If statvfs() failed on the spool or log directories while checking their
6526 size for availability, Exim confusingly gave the error "space shortage".
6527 Furthermore, in debugging mode it crashed with a floating point exception.
6528 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6529 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6530 serious problem, Exim now writes to the main and panic logs when this
6531 happens, with details of the failure. It then refuses to accept the
6532 incoming message, giving the message "spool directory problem" or "log
6533 directory problem" with a 421 code for SMTP messages.
6535 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6536 0, 1, and 2 exist, because some OS complain for execs without them (see
6537 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6538 descriptors. However, the code omitted to check that the open succeeded,
6539 causing mysterious errors if for some reason the permissions on /dev/null
6540 got screwed. Now Exim writes a message to the main and panic logs, and
6541 bombs out if it can't open /dev/null.
6543 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6544 interact so that it is all more flexible. It is supposed to remain
6545 backwards compatible. Also added extra_local_interfaces.
6547 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6548 to bomb out with an assertion failure - to the client this appears as a
6549 connection drop. This problem occurs in the part of the code that was taken
6550 from the Samba project. Fortunately, the assertion is in a very simple
6551 function, so I have fixed this by reproducing the function inline in the
6552 one place where it is called, and arranging for authentication to fail
6553 instead of killing the process with assert().
6555 75. The SPA client code was not working when the server requested OEM rather
6556 than Unicode encoding.
6558 76. Added code to make require_files with a specific uid setting more usable in
6559 the case where statting the file as root fails - usually a non-root-mounted
6560 NFS file system. When this happens and the failure is EACCES, Exim now
6561 forks a subprocess and does the per-uid checking as the relevant uid.
6563 77. Added process_log_path.
6565 78. If log_file_path was not explicitly set, a setting of check_log_space or
6566 check_log_inodes was ignored.
6568 79. If a space check for the spool or log partitions fails, the incident is now
6569 logged. Of course, in the latter case the data may get lost...
6571 80. Added the %p formatting code to string_format() so that it can be used to
6572 print addresses in debug_print(). Adjusted all the address printing in the
6573 debugging in store.c to use %p rather than %d.
6575 81. There was a concern that a line of code in smtp_in.c could overflow a
6576 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6577 initially expressed, the concern was not well-founded, because trailing
6578 spaces are removed early. However, if the trailing spaces were followed by
6579 a NULL, they did not get removed, so the overflow was possible. Two fixes
6582 (a) I re-wrote the offending code in a cleaner fashion.
6583 (b) If an incoming SMTP command contains a NULL character, it is rejected
6586 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6587 runs initgroups(), so that if the Exim user is in any additional groups,
6588 they will be used during message reception.
6594 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob